a systematic literature review of chronic kidney disease

Loading metrics

Open Access

Peer-reviewed

Research Article

Symptom burden and health-related quality of life in chronic kidney disease: A global systematic review and meta-analysis

Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Visualization, Writing – original draft, Writing – review & editing

Affiliation Centre for Patient Reported Outcomes Research, Institute of Applied Health Research, University of Birmingham, Birmingham, United Kingdom

Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing

Affiliation National Institute for Health Research Applied Research Collaboration West Midlands, Institute of Applied Health Research, University of Birmingham, Birmingham, United Kingdom

Roles Conceptualization, Funding acquisition, Writing – original draft, Writing – review & editing

Affiliations Centre for Patient Reported Outcomes Research, Institute of Applied Health Research, University of Birmingham, Birmingham, United Kingdom, National Institute for Health Research Applied Research Collaboration West Midlands, Institute of Applied Health Research, University of Birmingham, Birmingham, United Kingdom, National Institute for Health Research Birmingham Biomedical Research Centre, University of Birmingham, Birmingham, United Kingdom

Roles Conceptualization, Data curation, Methodology, Writing – original draft, Writing – review & editing

Affiliations Centre for Patient Reported Outcomes Research, Institute of Applied Health Research, University of Birmingham, Birmingham, United Kingdom, Department of Renal Medicine, Queen Elizabeth Hospital Birmingham, University Hospitals Birmingham, Birmingham, United Kingdom

Roles Conceptualization, Funding acquisition, Investigation, Methodology, Supervision, Writing – original draft, Writing – review & editing

Affiliations Centre for Patient Reported Outcomes Research, Institute of Applied Health Research, University of Birmingham, Birmingham, United Kingdom, National Institute for Health Research Birmingham Biomedical Research Centre, University of Birmingham, Birmingham, United Kingdom, National Institute for Health Research Surgical Reconstruction and Microbiology Research Centre, University of Birmingham, Birmingham, United Kingdom

Roles Conceptualization, Funding acquisition, Investigation, Methodology, Writing – original draft, Writing – review & editing

Affiliations Centre for Patient Reported Outcomes Research, Institute of Applied Health Research, University of Birmingham, Birmingham, United Kingdom, Department of Renal Medicine, Queen Elizabeth Hospital Birmingham, University Hospitals Birmingham, Birmingham, United Kingdom, Institute of Inflammation and Ageing, University of Birmingham, Birmingham, United Kingdom

Affiliations Centre for Patient Reported Outcomes Research, Institute of Applied Health Research, University of Birmingham, Birmingham, United Kingdom, National Institute for Health Research Birmingham Biomedical Research Centre, University of Birmingham, Birmingham, United Kingdom

Roles Conceptualization, Formal analysis, Funding acquisition, Methodology, Writing – original draft, Writing – review & editing

Affiliation Psychometric Laboratory for Health Sciences, University of Leeds, Leeds, United Kingdom

Affiliation Department of Child and Family Welfare, Faculty of Behavioural and Social Sciences, University of Groningen, Groningen, The Netherlands

Affiliation MD Anderson Center for INSPiRED Cancer Care, University of Texas, Houston, Texas, United States of America

Roles Conceptualization, Formal analysis, Methodology, Writing – original draft, Writing – review & editing

Affiliation Centre for Inflammation Research, University of Edinburgh, Edinburgh, United Kingdom

Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Visualization, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

Affiliations Centre for Patient Reported Outcomes Research, Institute of Applied Health Research, University of Birmingham, Birmingham, United Kingdom, School of Allied Health and Community, University of Worcester, Worcester, United Kingdom

• Benjamin R. Fletcher, 
• Sarah Damery, 
• Olalekan Lee Aiyegbusi, 
• Nicola Anderson, 
• Melanie Calvert, 
• Paul Cockwell, 
• James Ferguson, 
• Mike Horton, 
• Muirne C. S. Paap, 

• Published: April 6, 2022
• https://doi.org/10.1371/journal.pmed.1003954
• Peer Review
• Reader Comments

The importance of patient-reported outcome measurement in chronic kidney disease (CKD) populations has been established. However, there remains a lack of research that has synthesised data around CKD-specific symptom and health-related quality of life (HRQOL) burden globally, to inform focused measurement of the most relevant patient-important information in a way that minimises patient burden. The aim of this review was to synthesise symptom prevalence/severity and HRQOL data across the following CKD clinical groups globally: (1) stage 1–5 and not on renal replacement therapy (RRT), (2) receiving dialysis, or (3) in receipt of a kidney transplant.

Methods and findings

MEDLINE, PsycINFO, and CINAHL were searched for English-language cross-sectional/longitudinal studies reporting prevalence and/or severity of symptoms and/or HRQOL in CKD, published between January 2000 and September 2021, including adult patients with CKD, and measuring symptom prevalence/severity and/or HRQOL using a patient-reported outcome measure (PROM). Random effects meta-analyses were used to pool data, stratified by CKD group: not on RRT, receiving dialysis, or in receipt of a kidney transplant. Methodological quality of included studies was assessed using the Joanna Briggs Institute Critical Appraisal Checklist for Studies Reporting Prevalence Data, and an exploration of publication bias performed. The search identified 1,529 studies, of which 449, with 199,147 participants from 62 countries, were included in the analysis. Studies used 67 different symptom and HRQOL outcome measures, which provided data on 68 reported symptoms. Random effects meta-analyses highlighted the considerable symptom and HRQOL burden associated with CKD, with fatigue particularly prevalent, both in patients not on RRT (14 studies, 4,139 participants: 70%, 95% CI 60%–79%) and those receiving dialysis (21 studies, 2,943 participants: 70%, 95% CI 64%–76%). A number of symptoms were significantly ( p < 0.05 after adjustment for multiple testing) less prevalent and/or less severe within the post-transplantation population, which may suggest attribution to CKD (fatigue, depression, itching, poor mobility, poor sleep, and dry mouth). Quality of life was commonly lower in patients on dialysis (36-Item Short Form Health Survey [SF-36] Mental Component Summary [MCS] 45.7 [95% CI 45.5–45.8]; SF-36 Physical Component Summary [PCS] 35.5 [95% CI 35.3–35.6]; 91 studies, 32,105 participants for MCS and PCS) than in other CKD populations (patients not on RRT: SF-36 MCS 66.6 [95% CI 66.5–66.6], p = 0.002; PCS 66.3 [95% CI 66.2–66.4], p = 0.002; 39 studies, 24,600 participants; transplant: MCS 50.0 [95% CI 49.9–50.1], p = 0.002; PCS 48.0 [95% CI 47.9–48.1], p = 0.002; 39 studies, 9,664 participants). Limitations of the analysis are the relatively few studies contributing to symptom severity estimates and inconsistent use of PROMs (different measures and time points) across the included literature, which hindered interpretation.

Conclusions

The main findings highlight the considerable symptom and HRQOL burden associated with CKD. The synthesis provides a detailed overview of the symptom/HRQOL profile across clinical groups, which may support healthcare professionals when discussing, measuring, and managing the potential treatment burden associated with CKD.

Protocol registration

PROSPERO CRD42020164737.

Author summary

Why was this study done.

• Chronic kidney disease (CKD) is a common disease globally.
• Patients with CKD have a reduced quality of life and a greater risk of hospitalisation, heart problems, and death.
• Monitoring patient symptoms and quality of life can provide important information to help optimise CKD management.
• There is a lack of clear evidence on differences in patient quality of life between CKD groups and which symptoms are experienced most often and/or are most severe.

What did the researchers do and find?

• We reviewed 449 studies that included 199,147 patients from 62 countries.
• Patients with CKD reported a range of common and/or severe symptoms; the exact symptom burden depended on the stage of the disease and how it was being treated. Fatigue, however, was a very common and severe symptom in all patient groups.
• Quality of life for patients with CKD was significantly lower than for individuals without the disease, and was worst in patients receiving dialysis.
• In general, patients who had received a kidney transplant experienced fewer and less severe symptoms and had an improved quality of life, but this was still worse than that of people without CKD.

What do these findings mean?

• Symptom burden and negative impact on quality of life are considerable for people with CKD, especially for those receiving dialysis treatment.
• The findings of this study will support healthcare professionals when discussing, measuring, and managing the potential treatment burden associated with CKD.
• This global review of symptoms in patients with CKD will help in the selection of symptoms for inclusion in remote monitoring to identify patients in need of intervention.

Citation: Fletcher BR, Damery S, Aiyegbusi OL, Anderson N, Calvert M, Cockwell P, et al. (2022) Symptom burden and health-related quality of life in chronic kidney disease: A global systematic review and meta-analysis. PLoS Med 19(4): e1003954. https://doi.org/10.1371/journal.pmed.1003954

Academic Editor: Sanjay Basu, Harvard Medical School, UNITED STATES

Received: March 19, 2021; Accepted: February 23, 2022; Published: April 6, 2022

Copyright: © 2022 Fletcher et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the manuscript and its Supporting information files.

Funding: This review was funded as part of a study funded by Kidney Research UK (Stoneygate Research Award KS_RP_013_20180914; www.kidneyresearchuk.org ). The grant was awarded to DK, and funded BF full time. The funders did not play any role in study design, data collection/analysis, decision to publish, or preparation of the manuscript.

Competing interests: I have read the journal’s policy and the authors of this manuscript have the following competing interests: MC is Director of the Birmingham Health Partners Centre for Regulatory Science and Innovation, Director of the Centre for the Centre for Patient Reported Outcomes Research and is a National Institute for Health Research (NIHR) Senior Investigator. MC receives funding from the National Institute for Health Research (NIHR) Birmingham Biomedical Research Centre, the NIHR Surgical Reconstruction and Microbiology Research Centre and NIHR ARC West Midlands at the at the University of Birmingham and University Hospitals Birmingham NHS Foundation Trust NIHR/UKRI., Health Data Research UK, Innovate UK (part of UK Research and Innovation), Macmillan Cancer Support, UCB Pharma, Janssen, GSK and Gilead. MC has received personal fees from Astellas Aparito Ltd, CIS Oncology, Takeda, Merck, Daiichi Sankyo, Glaukos, GSK and the Patient-Centered Outcomes Research Institute (PCORI) outside the submitted work. OLA is supported by the NIHR Birmingham BRC, West Midlands, Birmingham; UKRI, Janssen, Gilead and GSK. He declares personal fees from Gilead Sciences Ltd, Merck and GSK outside the submitted work. DK reports grants from Macmillan Cancer Support, Innovate UK, the NIHR, NIHR Birmingham Biomedical Research Centre, and NIHR SRMRC at the University of Birmingham and University Hospitals Birmingham NHS Foundation Trust, and personal fees from Merck outside the submitted work.

Abbreviations: BDI-II, Beck Depression Inventory; CKD, chronic kidney disease; EQ-5D, EuroQoL–5 Dimension; FDR, false discovery rate; HADS, Hospital Anxiety and Depression Scale; HRQOL, health-related quality of life; KDQOL, Kidney Disease Quality of Life; KSS, Kidney Disease Quality of Life Summary Score; MCS, Mental Component Summary; PCS, Physical Component Summary; PROM, patient-reported outcome measure; RRT, renal replacement therapy; SF-12, 12-Item Short Form Health Survey; SF-36, 36-Item Short Form Health Survey

Introduction

Chronic kidney disease (CKD) has an estimated global prevalence of 9.1% (700 million people) and is associated with a major increased risk of early death for those affected, with 4.6% of deaths annually attributable to impaired kidney function [ 1 ]. In addition, CKD represents a substantial health economic burden, with advancement from stage 3 to 4/5 associated with a 1.3-to 4.2-fold increase in costs, and progression to end-stage renal disease (ESRD) estimated to cost $20,000–$100,000 per patient per year [ 2 ].

Patients with CKD experience an increased risk of hospitalisation and mortality [ 3 ], and reduced health-related quality of life (HRQOL), which is independently associated with cardiovascular disease events and death [ 4 – 6 ]. Patient-reported outcomes, including HRQOL and symptoms, are often identified by patients with CKD as more important to them than clinical outcomes such as survival [ 7 ].

There has been an increasing move towards models of remote and virtual care for patients with CKD [ 8 ], accelerated by the emergence of COVID-19, within which, capture of symptom and HRQOL data are seen as key adjuncts to support optimal care [ 9 – 11 ]. However, there remains a lack of research that has synthesised global data on CKD-specific symptom and HRQOL burden to inform collection of the most relevant patient-important data in a way that minimises patient burden.

The aim of this study was to (1) produce a comprehensive and consolidated global synthesis of symptom prevalence/severity and HRQOL across CKD treatment groups, (2) explore which symptom/HRQOL domains are modified by CKD and may be attributable to the disease, and (3) determine which patient-reported outcome measures (PROMs) are currently available to capture symptom prevalence/severity and HRQOL in CKD.

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed ( S1 Appendix ), and the study protocol was registered with PROSPERO (CRD42020164737) ( S9 Appendix ). The protocol and analysis methods were developed prospectively.

Search strategy and data sources

The following databases were searched from 1 January 2000 until 6 February 2020 (searches updated on 1 September 2021): Ovid MEDLINE, Ovid PsycINFO, and EBSCO CINAHL (for the full search strategy, see S2 Appendix ). Two authors (2 of BRF, SD, and DK) independently assessed selected articles for eligibility at the title/abstract and full-text stages, with disagreements resolved through discussion.

Inclusion and exclusion criteria

Studies were included if they met the following criteria: (1) published in or after January 2000; (2) written in English; (3) included adult (≥18 years) patients with CKD at stage 1–5 not receiving renal replacement therapy (RRT), those receiving dialysis, or those in receipt of a kidney transplant; (4) reported prevalence and/or severity of symptoms and/or HRQOL measured using a PROM; and (5) used a cross-sectional or longitudinal study design. Studies were excluded if they (1) were editorials, conference abstracts, reports of qualitative findings, or systematic reviews; (2) solely reported symptoms or HRQOL that were not self-reported by patients (e.g., clinician-reported); or (3) were not reported in English.

Data extraction

The following data were independently extracted into a pre-piloted spreadsheet by 2 authors (2 of BRF, SD, and DK) and disagreements resolved through discussion: study information (year conducted, country of origin, single/multi-centre, cross-sectional/longitudinal design), study population (inclusion/exclusion criteria, CKD stage, estimated glomerular filtration rate [eGFR], co-morbidity indices, demographics), and study outcomes (measures used, symptom prevalence and severity, HRQOL PROM scores). Where possible, we also attempted to extract data from the included studies that were collected from contemporaneous (ideally matched) non-CKD control populations.

Quality appraisal

Methodological quality of included studies was assessed using the Joanna Briggs Institute Critical Appraisal Checklist for Studies Reporting Prevalence Data [ 12 ]. Studies were assessed for adequacy of sampling (frame and method), sample size and description, data analysis, and comparability of outcomes across studies. Two authors independently conducted the appraisal (BRF and SD), and disagreements were resolved through discussion, or by a third reviewer (DK) where required.

Data analysis

We followed established guidelines for systematic reviews of observational epidemiological studies reporting prevalence with a focus on estimating the global burden of disease [ 12 ]. As outlined in the protocol, prevalence/severity data were pooled using either a random or fixed effects model depending on the heterogeneity of the included studies. Heterogeneity was determined using Cochran’s Q test at a significance level of 0.10. Heterogeneity was quantified using the I 2 statistic (acceptable heterogeneity defined as I 2 < 70%) [ 13 ]. All analyses had high heterogeneity; therefore, a random effects model was used. Subgroup analysis was performed based on the stage of CKD (categorised as not on RRT, on dialysis, or in receipt of a kidney transplant), if there were 3 or more studies within a subgroup. Publication bias was assessed where meta-analyses included 10 or more studies using Egger’s test for funnel plot asymmetry, with a significance level of p < 0.05 [ 14 ]. Using Stata, the metaprop (meta-analysis of binomial data) command was used to summarise prevalence data, and the metan command was used to summarise severity and HRQOL scores [ 15 , 16 ].

To aid comparison of symptom severity data provided across different outcome measures, all mean severity scores were converted to a 0–100 scale, where a higher score indicates greater severity. For HRQOL scores, 100 represents the best possible quality of life. For example, the Beck Depression Inventory (BDI-II) scale results in a severity score of 0–63; therefore, a score of 43 would convert to 68.3 on a 0–100 scale: 43/63 × 100. Symptom severity scores were also combined using random effects meta-analysis.

A weighted composite summary score for the Kidney Disease Quality of Life (KDQOL) instrument—KDQOL Summary Score (KSS)—was calculated by combining the ‘symptoms and problems’ (12 items), ‘effects of kidney disease’ (8 items), and ‘burden of kidney disease’ (4 items) domains. This summary score was calculated using the recommended method reported by Peipert et al. [ 17 ], in which mean scores and standard deviations for each of the 3 domains were combined, weighted by the number of items per domain.

Presentation of symptom prevalence and severity focused on 2 areas: those symptoms that were most prevalent/severe across populations and those symptoms that were significantly different between populations not receiving RRT and those receiving dialysis or transplantation. The latter area is important, as it could provide insight into those symptoms that may be attributable to changes in renal function and may provide potential targets for symptom tracking in CKD.

Exploratory subgroup analysis was used to compare prevalence and score (severity and HRQOL) estimates between groups in meta-analyses (not on RRT versus dialysis, not on RRT versus transplant, and dialysis versus transplant). To account for multiple testing, sharpened false discovery rate (FDR) q -values were computed [ 18 ], and adjusted p -values are reported, with a significance level of p < 0.05. All analyses were conducted using Stata (version 15.0).

Role of the funder

The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the report and had final responsibility for the decision to submit for publication.

Included studies

Searches identified 1,521 records after deduplication, and an additional 8 were identified through citation searches of included studies. Following title/abstract screening, 631 full-text articles were obtained, with 182 excluded at this stage, leaving 449 studies for inclusion in the final syntheses ( Table 1 ). Information on individual studies included in this review (outcomes used, study design, country of origin, population, and risk of bias) is included in S3 Appendix . The full lists of included and excluded studies are provided in S4 and S5 Appendices, respectively. The PRISMA diagram is shown in Fig 1 .

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

CKD, chronic kidney disease; PROM, patient-reported outcome measure.

https://doi.org/10.1371/journal.pmed.1003954.g001

https://doi.org/10.1371/journal.pmed.1003954.t001

There was a total of 199,147 participants involved in the included studies (median 146, IQR 85 to 267, range 9 to 18,015). Studies were conducted in 62 countries, with the most studies in the following countries: US, 43 (10%); Brazil, 43 (10%); UK, 36 (8%); Turkey, 30 (7%); and China, 29 (6%); the majority of studies were conducted at a single centre (251, 56%). Most studies were cross-sectional in design (385, 86%). Patients with CKD stage 1–5 who were not on RRT were included in 126 (28%) studies. The staging of patients not receiving RRT was as follows: 29 studies, stage 1; 44 studies, stage 2; 80 studies, stage 3; 92 studies, stage 4; and 98 studies, stage 5. Patients receiving dialysis were included in 274 (61%) studies (explicitly stated as haemodialysis in 228, and peritoneal dialysis in 118). Patients in receipt of a kidney transplant were included in 139 (31%) studies.

Outcome measures

The included studies utilised 67 different PROMs to collect information on symptoms and HRQOL. Eleven measures were reported in 10 or more individual studies: the 36-Item Short Form Health Survey (SF-36) in 227 studies, KDQOL in 100 studies, the 12-Item Short Form Health Survey (SF-12) in 52 studies, the BDI-II in 51 studies, the World Health Organization Quality of Life (WHOQOL-BREF) instrument in 33 studies, EuroQoL–5 Dimension (EQ-5D) in 28 studies, the Hospital Anxiety and Depression Scale (HADS) in 22 studies, the Centre for Epidemiologic Studies Depression Scale (CES-D) in 11 studies, the Patient Health Questionnaire–9 (PHQ-9) in 10 studies, the Integrated Palliative care Outcome Scale–Renal (IPOS-Renal) in 10 studies, and the Pittsburgh Sleep Quality Index (PSQI) in 10 studies.

A total of 68 different symptoms were measured across 54 PROMs (mean number of items per PROM = 22, range 1–90). No single PROM measured the majority of reported symptoms across the CKD population. The PROMs with the most comprehensive symptom coverage included the CKD Symptom Burden Index (44% of symptoms), the Dialysis Symptom Index (41%), the Memorial Symptom Assessment Scale–Short Form (33%), the Modified Transplant Symptom Occurrence and Symptom Distress Scale (33%), and the Chronic Kidney Disease Symptom Index (32%). There was little consistency across measures; some focused on a single symptom (e.g., BDI-II: depression), others included a number of symptom subdomains (e.g., HADS: anxiety and depression), and some included multiple questions, each tackling a different symptom (e.g., Disease Symptom Index: 30 individual symptom questions).

Symptom prevalence and severity

Data on symptom prevalence and severity were extracted from 181 studies. Pooled summary data are available in Table 2 . Symptom prevalence data were available for 45 symptoms in patients not on RRT, for 42 symptoms in patients receiving dialysis, and for 27 symptoms in the transplant population. Symptom severity data were available for 18 symptoms in patients not on RRT, for 33 symptoms in patients receiving dialysis, and for 22 symptoms in transplant patients. Data for symptom prevalence and severity are shown in Figs 2 – 7 .

https://doi.org/10.1371/journal.pmed.1003954.g002

https://doi.org/10.1371/journal.pmed.1003954.g003

https://doi.org/10.1371/journal.pmed.1003954.g004

Scores represent mean severity scores converted to a 0–100 scale, where a higher score indicates greater severity; vertical line at 50 for reference. RRT, renal replacement therapy.

https://doi.org/10.1371/journal.pmed.1003954.g005

Scores represent mean severity scores converted to a 0–100 scale, where a higher score indicates greater severity; vertical line at 50 for reference.

https://doi.org/10.1371/journal.pmed.1003954.g006

https://doi.org/10.1371/journal.pmed.1003954.g007

https://doi.org/10.1371/journal.pmed.1003954.t002

Patients not on RRT and those receiving dialysis shared a similar profile of the most prevalent symptoms. For example, symptoms with a reported prevalence of >50% in both populations included fatigue and poor mobility, and symptoms with a reported prevalence of >45% in both populations included bone/joint pain, general pain, poor sleep, sexual dysfunction, heartburn, muscle cramps, itching, and dry skin. Fewer data were available for transplant patients; however, indigestion, abdominal pain, constipation, muscle weakness, and muscle cramps were most prevalent, being present in >50% of patients.

For patients not on RRT, the most severe symptoms included sexual dysfunction, anxiety, itching, and depression. Pain was the most severe symptom in dialysis, followed by fatigue, dry skin, and bone/joint pain. For transplant recipients, the most severe symptoms were change in appearance, blurred vision, and excessive appetite.

Trends in prevalence/severity across clinical groups and attribution

Within the included studies, data from contemporaneous non-CKD control populations were limited, and available for only 17 symptoms for prevalence and 4 for severity.

Prevalence was higher in CKD patients compared to healthy controls for 14 of 17 symptoms (bone/joint pain, fatigue, trouble with memory, muscle cramps, itching, restless legs, muscle weakness, constipation, shortness of breath, anxiety, depression, decreased appetite, diarrhoea, and abdominal pain), and lower than controls for 1 symptom (stress).

Fatigue was the most prevalent symptom in patients not on RRT and in those on dialysis. Fatigue was also the second most severe symptom in dialysis patients (adjusted severity score 51.5, 95% CI 29.1–33.8). Fig 8 displays the full results of the meta-analysis of fatigue prevalence across CKD clinical groups including controls. Fatigue prevalence in controls was 34% (95% CI 0%–70%). In comparison, fatigue prevalence was significantly higher (FDR-sharpened q -value 0.021) in patients with CKD not on RRT (70%, 95% CI 60%–79%) and in dialysis patients (70%, 95% CI 64%–76%; FDR-sharpened q -value 0.021); fatigue prevalence was significantly lower in transplant patients (48%, 95% CI 32%–63%; FDR-sharpened q -value 0.005) than in patients on RRT or dialysis, although notably not as low as in controls. A number of other symptoms followed this prevalence pattern across clinical groups. All symptom prevalence and severity data are available in S6 and S7 Appendices, and all pairwise comparisons between groups including FDR-sharpened q -values are available in S8 Appendix . Fig 9 includes the point estimates for symptom prevalence reported across the 3 study populations (with control data where available).

See S4 Appendix for full references for included studies. ES, effect size; HD, haemodialysis; PD, peritoneal dialysis; RRT, renal replacement therapy.

https://doi.org/10.1371/journal.pmed.1003954.g008

CKD, chronic kidney disease; RRT, renal replacement therapy.

https://doi.org/10.1371/journal.pmed.1003954.g009

Patients not on RRT and those receiving dialysis had similar profiles of prevalence across most symptoms. However, exploratory subgroup analysis (including correction for multiple testing) highlighted 2 symptoms that were significantly more prevalent in the dialysis population than in patients not on RTT: depression and stress (FDR-sharpened q -value 0.021 in both cases). Seven further symptoms showed tendencies towards greater prevalence (>10% difference) in the dialysis population but did not reach adjusted statistical significance: weight loss, muscle weakness, hiccups, heartburn, changes in skin, trouble with memory, and dry skin. Drowsiness demonstrated a tendency towards lower prevalence in the dialysis population (>10% difference), but again the difference did not reach significance.

When compared to patients not on RRT and dialysis patients, the following symptoms were significantly less prevalent in patients who had received a kidney transplant: muscle weakness, fatigue, poor sleep, itching, decreased appetite, depression, dry mouth, and poor mobility (FDR-sharpened q -values 0.005–0.037). Overall, compared to the kidney transplant population, symptom prevalence was higher in patients not on RRT and patients on dialysis for 31 of 50 comparisons. However, there were 2 symptoms that reversed this pattern, constipation and indigestion, which were both significantly more prevalent in the transplant population (FDR-sharpened q -values 0.005–0.013).

Data on HRQOL were extracted from 361 articles. The Medical Outcomes Study SF-12 and SF-36 were reported in 52 and 227 studies, respectively, KDQOL in 100, the World Health Organization Quality of Life (WHOQOL-BREF) instrument in 33, and EQ-5D in 28.

Pooled scores are shown for SF-12/SF-36, KDQOL, and EQ-5D in Table 3 and Fig 10 . For all scores, a higher number represents better quality of life (0–100 scale for SF-12/SF-36 and KDQOL, and possible range of −0.224 to 1 for EQ-5D).

CKD, chronic kidney disease; EQ5D, EuroQoL–5 Dimension; KSS, Kidney Disease Quality of Life Summary Score; MCS, Mental Component Summary; No. pops, number of populations; PCS, Physical Component Summary; SF12, 12-Item Short Form Health Survey; SF36, 36-Item Short Form Health Survey.

https://doi.org/10.1371/journal.pmed.1003954.g010

https://doi.org/10.1371/journal.pmed.1003954.t003

Within the generic measures of HRQOL, SF-12/SF-36 and EQ-5D, where data were available, scores were highest in controls (EQ-5D index 0.95, 95% CI 0.95–0.95) and lowest in patients on dialysis (EQ-5D index 0.78, 95% CI 0.77–0.79; SF-36 Mental Component Summary [MCS] 45.7, Physical Component Summary [PCS] 35.5; SF-12 MCS 45.4, PCS 35.2). HRQOL scores were higher in patients receiving a kidney transplant (EQ-5D index 0.84, 95% CI 0.82–0.86; SF-36 MCS 50.0, PCS 48.0; SF-12 MCS 48.2, PCS 44.8), and higher still for patients not on RRT (EQ-5D index 0.88, 95% CI 0.88–0.88; SF-36 MCS 66.6, PCS 66.3; SF-12 MCS 49.8, PCS 47.5).

For the disease-specific KDQOL measure, the KSS was 73.0 in patients not on RRT, 64.6 in patients receiving dialysis, and highest in transplant patients (84.0). This pattern was similar in the KDQOL ‘effects of kidney disease’ (not on RRT, 71.7; dialysis, 63.2; transplant, 87.5) and ‘burden of kidney disease’ subscales (not on RRT, 50.6; dialysis, 41.7; transplant, 72.0). The burden of kidney disease subscale includes items related to how much kidney disease interferes with daily life, or makes the respondent feel like a burden. The effects of kidney disease subscale includes items exploring respondents’ perceived dependency on clinicians, stress/worries, and bother associated with treatment/dietary restrictions. In the ‘symptoms and problems’ subscale, for patients not on RRT and those in receipt of a transplant, scores were homologous (85.9 and 86.1, respectively), whilst pooled scores for the dialysis population were lower (73.6). The symptoms and problems subscale items measured how bothered respondents were by certain symptoms (e.g., sore muscles, chest pain, cramps, itchy/dry skin, and fatigue) or problems associated with dialysis access.

All exploratory subgroup analyses comparing HRQOL scores between populations showed statistically significant differences, with the exception of the KDQOL symptoms and problems subscale comparison between patients not on RRT and in receipt of a transplant (FDR-sharpened q -value 0.107) (HRQOL subgroup analyses available in S8 Appendix ).

Quality appraisal of included studies

Results of the quality appraisal are shown in Fig 11 and for individual studies in S3 Appendix .

Yes = appraised as adequate; no = appraised as not adequate.

https://doi.org/10.1371/journal.pmed.1003954.g011

Whilst the majority of studies used random sampling or approached all patients on a clinic/registry list (59.5%), some used convenience or consecutive sampling (28.3%), or methods were unclear (12.2%). Sample size was deemed adequate in 61.9% of studies, and response rate in 57.7%. PROMs for symptoms or HRQOL that allowed comparison with other studies in CKD were used in 93.3% of studies (i.e., the measure had been used in CKD before, either as identified in this review or in the author description of previous use). Statistical analysis was reported in sufficient detail in 92.2% of studies.

There was evidence of publication bias (Egger’s test for funnel plot asymmetry) for 3 of 32 symptom prevalence meta-analyses including ≥10 studies; the 3 symptoms were poor sleep, numbness in hands/feet, and anxiety. No evidence of publication bias was found in meta-analyses of HRQOL data. All publication bias analyses are available in S6 – S8 Appendices.

In this study, our first aim was to provide a global synthesis of symptom/HRQOL burden across all CKD stages. Overall, patients with CKD had a significantly increased symptom burden, and lower quality of life, compared to individuals without CKD. Patients reported a range of common and/or severe symptoms, with the precise configuration depending on the stage of CKD and RRT treatment modality received. Fatigue, however, was a very common and severe symptom in all patient groups. Symptom burden and quality of life were worst in patients receiving dialysis. In general, patients who had received a kidney transplant experienced fewer and less severe symptoms, and had an improved quality of life, compared to patients with CKD not receiving RRT or patients receiving treatment with dialysis. Transplantation, however, did not restore quality of life to levels seen in those without CKD.

Identification of the burden of symptoms for patients with advanced CKD is important: Many of the symptoms reported can be mitigated by changes in clinical management [ 19 ]. This synthesis will support clinicians and patients in consultations, to ensure that all potential symptoms associated with CKD are recognised, facilitating shared decision-making regarding management [ 20 ]. In addition, the data highlight key variations in symptom burden between kidney disease modalities, allowing clinicians to identify fundamental differences and administer appropriate treatment. Moreover, the data may highlight key domains appropriate for inclusion within routine remote monitoring, a tool increasingly employed in clinical practice to support timely intervention in response to patient deterioration [ 8 – 11 ]. Recent reviews report on some symptoms, but do not quantify the symptom burden or accurately identify differences in symptom burden between treatment states [ 21 , 22 ]. The results of this study address this shortfall and provide information that has direct implications for clinical practice.

Our results concur with previous research regarding the prevalence of a number of symptoms, particularly in patients with stage 4/5 CKD. Almutary and colleagues conducted a 2013 review of symptom burden in CKD (stage 4/5 not on RRT/on dialysis), finding that the most prevalent symptoms were fatigue, feeling drowsy, pain, pruritus, and dry skin [ 21 ]. Murtagh et al. conducted a systematic review of the prevalence of symptoms in patients with end-stage renal disease, with the following prevalent symptoms highlighted: fatigue/tiredness, pruritus, constipation, anorexia, pain, sleep disturbance, anxiety, dyspnea, nausea, restless legs, and depression [ 22 ]. Differences between these studies and ours may be explained by our far larger sample size, resulting in much more precise estimates. Our study also has the advantage of comparing data across CKD clinical groups and including additional information regarding HRQOL and symptom severity.

Our second aim was to explore which symptom/HRQOL domains are modified by CKD and may be attributable to the disease. To the best of our knowledge, ours is the first study to attempt this. Our results suggest there was a significantly lower prevalence of many symptoms in kidney transplant patients, compared to patients not on RRT and dialysis patients, which may suggest attribution; these symptoms included muscle weakness, fatigue, poor sleep, itching, decreased appetite, depression, dry mouth, and poor mobility. Collecting routine data on which symptoms/HRQOL are prevalent, impactful, and directly attributable to CKD is vital to improve understanding of individuals’ experience of illness and to target treatment/support. This information can also be used alongside existing clinical data in discussions with patients to help better prepare them for CKD progression and to inform shared decisions around treatment [ 10 ]. We also found that some symptoms did not differ significantly across clinical groups and therefore may be largely unrelated to kidney function specifically, or the confidence intervals may preclude accurate interpretation. Further research may be required to explore the symptoms with broad confidence intervals, and hence greater uncertainty.

The third aim of the study was to determine which current PROMs may capture patient-important symptom/HRQOL information in a way that minimises patient burden. In total, 54 PROMs were used to collect data on symptoms across the included studies, and we found little consistency in the measures. This may be a consequence of the fact that no single tool measured >45% of symptoms reported in the population. This is problematic. At present, comprehensive measurement of symptoms would require that patients complete multiple PROMs, which may include large numbers of items or may have items that overlap. Such PROM ‘item burden’ has been widely recognised as an important threat to adherence [ 23 ]. Given the high number of symptoms experienced by patients with CKD, the use of contemporary psychometrics, encompassing item response theory (IRT) and computerised adaptive testing (CAT), may be warranted in order to develop new measures that capture sufficient information regarding all patient-important symptoms, whilst minimising questionnaire burden [ 24 ]. CATs efficiently select questions from an IRT-calibrated item bank that are targeted to an individual’s ability/trait level using an adaptive algorithm, minimising the number of items administered, for example, the Patient-Reported Outcomes Measurement Information System (PROMIS) physical function CAT [ 25 – 27 ]. The findings of this systematic review will contribute to the construction of an item bank as part of the RCAT (Renal Computerized Adaptive Test) study [ 28 ].

Strengths and limitations

Our findings present a comprehensive overview of the differences in symptom and HRQOL prevalence and severity between patients with CKD stage 1–5 not on RRT, patients receiving dialysis, and transplant patients. In addition, where contemporaneous data were available, we were able to compare with controls. The study included >190,000 patients with CKD (from 62 countries) throughout the trajectory of the disease and its treatment. A large quantity of data was available from different settings, and this synthesis presents strong evidence for the ongoing and considerable impact of CKD on patients’ lives. The review included data collected using PROMs only, hence data provided from CKD patients themselves. It is now widely understood that PROMs are patient centric, and provide information that is as important, if not more so, to patients than solely focusing on clinical outcomes [ 29 – 32 ].

Data were most frequently from cross-sectional studies, and whilst this is useful in providing an estimate of the prevalence and impact of symptoms at a population level, it does not address the day-to-day variation experienced by CKD patients and makes it challenging to draw robust conclusions around longitudinal patterns of symptom burden during the course of the disease.

A limitation is that we excluded non-English-language papers, meaning some potentially relevant studies may not have been included in our analysis. In addition, many symptom severity estimates came from single studies, and inconsistent use of PROMs across this literature hindered interpretation, especially with regard to clinical significance. This necessitated standardisation of severity scores onto a 0–100 scale to support meaningful synthesis [ 33 ]. Moreover, whilst there was information available in some of the studies on the severity of CKD, symptom burden with respect of excretory kidney function was limited, so the evolution of symptoms during the progression of CKD could not be assessed.

A further limitation was the considerable heterogeneity of included studies. This was not unexpected, as heterogeneity can be a common problem in systematic reviews of global prevalence data [ 12 ]. However, we followed established guidelines in our analysis [ 12 ], which suggest that in the presence of significant heterogeneity, random effects meta-analysis may be an appropriate method of generating a distribution that allows estimation of population differences with a quantifiable degree of probability.

Future research

The findings of this review highlight several areas that warrant further research. In particular, additional high-quality studies exploring symptom severity are required in order to generate more precise estimates. We also found that there were fewer studies exploring symptom burden/HRQOL in the transplant population compared to other CKD groups. Finally, historical and ongoing use of many different symptom/HRQOL PROMs across studies poses particular challenges for those wishing to synthesise data. Future studies should focus on more consistent use of recommended outcome measures, such as those included in internationally endorsed core outcome sets [ 34 , 35 ], to facilitate comparisons between studies and enhance the generalisability of findings.

This systematic review provides a detailed overview of the symptom/HRQOL profile across CKD clinical groups, with fatigue particularly prevalent, both in patients not on RRT and in those receiving dialysis. A number of symptoms were less prevalent and/or severe within the post-transplantation population, which may suggest attribution to CKD. HRQOL in patients with CKD was significantly worse than in individuals without the disease, particularly so in patients receiving dialysis. In general, patients receiving a transplant experienced lower symptom prevalence and severity and improved disease-specific quality of life, but this still did not reach the level of HRQOL of people without CKD. The findings of this review may support healthcare professionals when discussing, measuring, and managing the potential treatment burden associated with CKD.

Supporting information

S1 appendix. prisma checklist..

https://doi.org/10.1371/journal.pmed.1003954.s001

S2 Appendix. Ovid MEDLINE search strategy.

https://doi.org/10.1371/journal.pmed.1003954.s002

S3 Appendix. Information on included studies.

https://doi.org/10.1371/journal.pmed.1003954.s003

S4 Appendix. List of included studies.

https://doi.org/10.1371/journal.pmed.1003954.s004

S5 Appendix. List of excluded studies.

https://doi.org/10.1371/journal.pmed.1003954.s005

S6 Appendix. Symptom prevalence across studies.

https://doi.org/10.1371/journal.pmed.1003954.s006

S7 Appendix. Symptom severity across studies.

https://doi.org/10.1371/journal.pmed.1003954.s007

S8 Appendix. Pairwise group comparisons.

https://doi.org/10.1371/journal.pmed.1003954.s008

S9 Appendix. Review protocol.

https://doi.org/10.1371/journal.pmed.1003954.s009

S10 Appendix. Summary of symptom outcome measures.

https://doi.org/10.1371/journal.pmed.1003954.s010

• View Article
• PubMed/NCBI
• Google Scholar
• 28. Institute of Applied Health Research. RCAT—renal computerised adaptive test. Birmingham: University of Birmingham; 2022 [cited 2022 Mar 7]. https://tinyurl.com/13vdwh0m .

Guideline Development Process

KDOQI strives to make clinical practice guideline development as transparent and efficient as possible. It generally takes 18-24 months from development of a scope of work to submission of the manuscript to AJKD. Our process is as follows: 

• National Kidney Foundation Scientific Advisory Board approves the topic for guideline development or update.
• KDOQI Chairs select a Work Group Chair/Co-Chair based on previous work in the area, a clear understanding of the guideline development process, and willingness and ability to commit time to the project.
• Scope of Work document is developed by the Work Group Chair/Co-Chair
• Work group chair/co-chairs invite work group members with the aim of creating a diverse, multidisciplinary group. Work group may include representation from nephrology, vascular surgery, pediatric nephrology, pharmacy, nutrition, social work, nursing, general practice, transplantation and/or other pertinent medical subspecialties. 
• Individuals being considered for work group membership declare any potentially conflicting interests and activities by written disclosure. Disclosures include all current and planned commercial, non-commercial, intellectual, institutional, and patient/public activities relating to the scope of the guideline.
• KDOQI contracts with an evidence review team (ERT), generally an AHRQ Evidence based Practice Center , to carry out a formal systematic review.
• Scope of Work is refined by the ERT and wider workgroup. ( Beginning in 2017, KDOQI aims to publish Guideline Scope documents for public review )
• Formulating the research questions and developing an analytic framework
• Pre-specifying the approach and methods
• Assessing abstracts for inclusion/exclusion
• The ERT produces an evidence report with findings from the literature review. This is submitted to AJKD for publication.
• The work group uses these findings to make clinical practice guidelines (high quality evidence) and clinical practice recommendations (lower quality evidence, consensus based recommendations), and grades them using the GRADE methodology . Each guideline and recommendation is published along with a rationale, including a summary of relevant available evidence (and gaps in evidence), and an explanation of potential benefits and harms.
• The work group also provides recommendations for further research and limitations of current data within each chapter. View research recommendations from all past KDOQI guidelines ( download PDF )  

KDOQI Guideline Peer Review Process

Internal review.

• Once the guideline text is signed off by the Work Group Chairs, recommendations are formatted and circulated to KDOQI Leadership and the National Kidney Foundation Scientific Advisory Board for review. Reviewers are asked to use an online form that presents each recommendation along with a box to check “agree” or “disagree” and space for comment.
• At the discretion of the KDOQI Chair, select outside experts may be invited to comment at this stage.
• If the guideline is jointly produced with another organization, that organization will carry out an internal review simultaneously. 
• Reviewer comments are collated by KDOQI staff and sent to the Work Group for discussion and possible edits. 
• Following edits, the guideline document is made available for public review.

External (Public) Review

• Throughout the latter stages of guideline development, a link to register for public review is posted to the NKF guidelines web page.
• The document is sent to registered reviewers and publicized to other individuals and groups with an interest in the topic.
• Public reviewers are provided with a link to the online form that presents each recommendation along with a box to check “agree” or “disagree” and space for comment.
• Individual Work Group members are asked to address comments pertaining to recommendations and sections they have authored.
• The Work Group Chairs coordinate the final rewriting of the guideline document based on the public review comments.

The updated guideline is submitted for publication in the American Journal of Kidney Diseases (AJKD)

KDOQI Guideline Development Timeline 

KDOQI continually updates a selection of its original clinical practice guidelines. Updates are initiated when enough new evidence becomes available of high enough quality to change current recommendations, e.g. a recommended intervention causes previously unknown substantial harm, a new intervention is significantly superior to a previously recommended intervention, or a recommendation can be applied to a new population.  

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Can functional motor capacity influence mortality in advanced chronic kidney disease patients.

1. Introduction

2. materials and methods, 2.1. study of functional capacity and muscle strength, 2.1.1. functional capacity was assessed using the following tests, short physical performance battery test (sppb), 6-minute walk test (6mwt), timed up and go test (tutg), sit-to-stand test (sts5), 2.1.2. muscle strength was assessed with the hand grip test, 2.2. comorbidity, 2.3. frailty, 2.4. nutritional status study, 2.4.1. body composition study: bioimpedance, 2.4.2. laboratory parameters, 2.5. statistical analysis, 3.1. general characteristics of the study population, 3.2. mortality outcome, 4. discussion, 5. conclusions, author contributions, institutional review board statement, informed consent statement, data availability statement, acknowledgments, conflicts of interest.

• Ahmed, M.M.; Shafat, Z.; Tazyeen, S.; Ali, R.; Almashjary, M.N.; Al-Raddadi, R.; Harakeh, S.; Alam, A.; Haque, S.; Ishrat, R. Identification of pathogenic genes associated with CKD: An integrated bioinformatics approach. Front. Genet. 2022 , 13 , 891055. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Kovesdy, C.P. Epidemiology of chronic kidney disease: An update 2022. Kidney Int. Suppl. 2022 , 12 , 7–11. [ Google Scholar ] [ CrossRef ]
• Mora-Gutiérrez, J.M.; Slon Roblero, M.F.; Castaño Bilbao, I.; Izquierdo Bautista, D.; Arteaga Coloma, J.; Martínez Velilla, N. Enfermedad renal crónica en el paciente anciano [Chronic kidney disease in the elderly patient]. Rev. Esp. Geriatr. Gerontol. 2017 , 52 , 152–158. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Corsonello, A.; Mattace-Raso, F.; Tap, L.; Maggio, M.; Zerbinati, L.; Guarasci, F.; Cozza, A.; D’Alia, S.; Soraci, L.; Corigliano, V.; et al. Design and methodology of the chronic kidney disease as a dysmetabolic determinant of disability among older people (CKD-3D) study: A multicenter cohort observational study. Aging Clin. Exp. Res. 2021 , 33 , 2445–2451. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Bhat, P.R.; Urooj, A.; Nalloor, S. Changes in body composition in relation to estimated glomerular filtration rate and physical activity in predialysis chronic kidney disease. Chronic Dis. Transl. Med. 2022 , 8 , 305–313. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Guo, Y.; Zhang, M.; Ye, T.; Wang, Z.; Yao, Y. Application of Bioelectrical Impedance Analysis in Nutritional Management of Patients with Chronic Kidney Disease. Nutrients 2023 , 15 , 3941. [ Google Scholar ] [ CrossRef ]
• Reinders, I.; Visser, M.; Schaap, L. Body weight and body composition in old age and their relationship with frailty. Curr. Opin. Clin. Nutr. Metab. Care 2017 , 20 , 11–15. [ Google Scholar ] [ CrossRef ]
• Mishra, M.; Wu, J.; Kane, A.E.; Howlett, S.E. The intersection of frailty and metabolism. Cell Metab. 2024 , 36 , 893–911. [ Google Scholar ] [ CrossRef ]
• Arena, R.; Cahalin, L.P.; Borghi-Silva, A.; Phillips, S.A. Improving functional capacity in heart failure: The need for a multifaceted approach. Curr. Opin. Cardiol. 2014 , 29 , 467–474. [ Google Scholar ] [ CrossRef ]
• Kennard, A.; Glasgow, N.; Rainsford, S.; Talaulikar, G. Frailty in chronic kidney disease: Challenges in nephrology practice. A review of current literature. Intern. Med. J. 2023 , 53 , 465–472. [ Google Scholar ] [ CrossRef ]
• Yang, L.; He, Y.; Li, X. Physical function and all-cause mortality in patients with chronic kidney disease and end-stage renal disease: A systematic review and meta-analysis. Int. Urol. Nephrol. 2023 , 55 , 1219–1228. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ju, S.H.; Yi, H.S. Clinical features and molecular mechanism of muscle wasting in end stage renal disease. BMB Rep. 2023 , 56 , 426–438. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Nogueira, Á.; Álvarez, G.; Russo, F.; San-José, B.; Sánchez-Tomero, J.A.; Barril, G. Is SPPB useful as a screening method of functional capacity in patients with advanced chronic kidney disease? Nephrology 2019 , 39 , 489–496. [ Google Scholar ] [ CrossRef ]
• Schrauben, S.J.; Chang, A.R. Functional Status in CKD: What Measures to Use? Kidney360 2021 , 2 , 608–610. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Painter, P.; Marcus, R.L. Assessing physical function and physical activity in patients with CKD. Clin. J. Am. Soc. Nephrol. 2013 , 8 , 861–872. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Guralnik, J.M.; Ferrucci, L.; Simonsick, E.M.; Salive, M.E.; Wallace, R.B. Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N. Engl. J. Med. 1995 , 332 , 556–561. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Guralnik, J.M.; Simonsick, E.M.; Ferrucci, L.; Glynn, R.J.; Berkman, L.F.; Blazer, D.G.; Scherr, P.A.; Wallace, R.B. A short physical performance battery assessing lower extremity function: Association with self-reported disability and prediction of mortality and nursing home admission. J. Gerontol. Med. Sci. 1994 , 49 , M85–M94. [ Google Scholar ] [ CrossRef ]
• ATS Committee on Proficiency Standards for Clinical Pulmonary Function Laboratories. ATS statement: Guidelines for the six-minute walk test. Am. J. Respir. Crit. Care Med. 2002 , 166 , 111–117, Erratum in Am. J. Respir. Crit. Care Med. 2016 , 193 , 1185. [ Google Scholar ]
• Matos Casano, H.A.; Anjum, F. Six-Minute Walk Test. In StatPearls ; StatPearls Publishing: Treasure Island, FL, USA, 2023. [ Google Scholar ]
• Podsiadlo, D.; Richardson, S. The timed “Up & Go”: A test of basic functional mobility for frail elderly persons. J. Am. Geriatr. Soc. 1991 , 39 , 142–148. [ Google Scholar ]
• Csuka, M.; McCarty, D.J. Simple method for measurement of lower extremity muscle strength. Am. J. Med. 1985 , 78 , 77–81. [ Google Scholar ] [ CrossRef ]
• Cruz-Montecinos, C.; Castro, R.T.M.; Otto-Yáñez, M.; Barros-Poblete, M.M.; Valencia, C.; Campos, A.; Jadue, L.; Barros, M.; Solis-Navarro, L.M.; Resqueti, V. Which sit-to-stand test best differentiates functional capacity in older people? Am. J. Phys. Med. Rehabil. 2024 . [ Google Scholar ] [ CrossRef ]
• Alcazar, J.; Losa-Reyna, J.; Rodriguez-Lopez, C.; Alfaro-Acha, A.; Rodriguez-Mañas, L.; Ara, I.; García-García, F.J.; Alegre, L.M. The sit-to-stand muscle power test: An easy, inexpensive and portable procedure to assess muscle power in older people. Exp. Gerontol. 2018 , 112 , 38–43. [ Google Scholar ] [ CrossRef ]
• Leal, V.O.; Mafra, D.; Fouque, D.; Anjos, L.A. Use of handgrip strength in the assessment of the muscle function of chronic kidney disease patients on dialysis: A systematic review. Nephrol. Dial. Transplant. 2011 , 26 , 1354–1360. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Oliveira, M.C.; Bufarah, M.N.B.; Balbi, A.L. Handgrip strength in end stage of renal disease: A narrative review. Nutrire 2018 , 43 , 14. [ Google Scholar ] [ CrossRef ]
• Charlson, M.E.; Pompei, P.; Ales, K.L.; MacKenzie, C.R. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J. Chronic Dis. 1987 , 40 , 373–383. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Fried, L.P.; Tangen, C.M.; Walston, J.; Newman, A.B.; Hirsch, C.; Gottdiener, J.; Seeman, T.; Tracy, R.; Kop, W.J.; Burke, G.; et al. Frailty in older adults: Evidence for a phenotype. J. Gerontol. A Biol. Sci. Med. Sci. 2001 , 56 , M146–M156. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Eyre, S.; Stenberg, J.; Wallengren, O.; Keane, D.; Avesani, C.M.; Bosaeus, I.; Clyne, N.; Heimbürger, O.; Indurain, A.; Johansson, A.; et al. Bioimpedance analysis in patients with chronic kidney disease. J. Ren. Care 2023 , 49 , 147–157. [ Google Scholar ] [ CrossRef ]
• McGinlay, J.M.; Payne, R.B. Serum albumin by dye-binding: Bromocresol green or bromocresol purple? The case for conservatism. Ann. Clin. Biochem. 1988 , 25 , 417–421. [ Google Scholar ] [ CrossRef ]
• Serra-Prat, M.; Lorenzo, I.; Papiol, M.; Palomera, E.; Bartolomé, M.; Pleguezuelos, E.; Burdoy, E. Intracellular Water Content in Lean Mass as an Indicator of Muscle Quality in an Older Obese Population. J. Clin. Med. 2020 , 9 , 1580. [ Google Scholar ] [ CrossRef ]
• Tian, M.; Yuan, J.; Yu, F.; He, P.; Zhang, Q.; Zha, Y. Decreased intracellular water is associated with sarcopenic obesity in chronic haemodialysis patients. BMC Geriatr. 2023 , 23 , 630. [ Google Scholar ] [ CrossRef ]
• Nogueira, Á.; Álvarez, G.; Barril, G. Impact of the Nutrition-Inflammation Status on the Functionality of Patients with Chronic Kidney Disease. Nutrients 2022 , 14 , 4745. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Barril, G.; Nogueira, A.; Alvarez-García, G.; Núñez, A.; Sánchez-González, C.; Ruperto, M. Nutritional Predictors of Mortality after 10 Years of Follow-Up in Patients with Chronic Kidney Disease at a Multidisciplinary Unit of Advanced Chronic Kidney Disease. Nutrients 2022 , 14 , 3848. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ekramzadeh, M.; Santoro, D.; Kopple, J.D. The Effect of Nutrition and Exercise on Body Composition, Exercise Capacity, and Physical Functioning in Advanced CKD Patients. Nutrients 2022 , 14 , 2129. [ Google Scholar ] [ CrossRef ]
• Hanna, R.M.; Ghobry, L.; Wassef, O.; Rhee, C.M.; Kalantar-Zadeh, K. A Practical Approach to Nutrition, Protein-Energy Wasting, Sarcopenia, and Cachexia in Patients with Chronic Kidney Disease. Blood Purif. 2020 , 49 , 202–211. [ Google Scholar ] [ CrossRef ]
• Cruz-Jentoft, A.J.; Bahat, G.; Bauer, J.; Boirie, Y.; Bruyère, O.; Cederholm, T.; Cooper, C.; Landi, F.; Rolland, Y.; Sayer, A.A.; et al. Sarcopenia: Revised European consensus on definition and diagnosis. Age Ageing 2019 , 48 , 16–31, Erratum in Age Ageing 2019 , 48 , 601. [ Google Scholar ] [ CrossRef ]
• Hwang, S.H.; Lee, D.H.; Min, J.; Jeon, J.Y. Handgrip Strength as a Predictor of All-Cause Mortality in Patients With Chronic Kidney Disease Undergoing Dialysis: A Meta-Analysis of Prospective Cohort Studies. J. Ren. Nutr. 2019 , 29 , 471–479. [ Google Scholar ] [ CrossRef ]
• Wilkinson, T.J.; Miksza, J.; Yates, T.; Lightfoot, C.J.; Baker, L.A.; Watson, E.L.; Zaccardi, F.; Smith, A.C. Association of sarcopenia with mortality and end-stage renal disease in those with chronic kidney disease: A UK Biobank study. J. Cachexia Sarcopenia Muscle 2021 , 12 , 586–598. [ Google Scholar ] [ CrossRef ]
• de Fátima Ribeiro Silva, C.; Ohara, D.G.; Matos, A.P.; Pinto, A.C.P.N.; Pegorari, M.S. Short Physical Performance Battery as a Measure of Physical Performance and Mortality Predictor in Older Adults: A Comprehensive Literature Review. Int. J. Environ. Res. Public Health 2021 , 18 , 10612. [ Google Scholar ] [ CrossRef ]
• Pavasini, R.; Guralnik, J.; Brown, J.C.; di Bari, M.; Cesari, M.; Landi, F.; Vaes, B.; Legrand, D.; Verghese, J.; Wang, C.; et al. Short Physical Performance Battery and all-cause mortality: Systematic review and meta-analysis. BMC Med. 2016 , 14 , 215. [ Google Scholar ] [ CrossRef ]
• Belkin, M.D.; Doerfler, R.M.; Wagner l Zhan m Fink, J.C. Associations of Performance-Based Functional Assessments and Adverse Outcomes in CKD. Kidney360 2021 , 2 , 629–638. [ Google Scholar ] [ CrossRef ]
• Corsonello, A.; Soraci, L.; Ärnlöv, J.; Carlsson, A.C.; Mattace-Raso, F.; Tap, L.; Formiga, F.; Moreno-González, R.; Kostka, T.; Guligowska, A.; et al. The relevance of geriatric assessments on the association between chronic kidney disease stages and mortality among older people: A secondary analysis of a multicentre cohort study. Age Ageing 2022 , 51 , afac168. [ Google Scholar ] [ CrossRef ]
• Lattanzio, F.; Corsonello, A.; Montesanto, A.; Abbatecola, A.M.; Lofaro, D.; Passarino, G.; Fusco, S.; Corica, F.; Pedone, C.; Maggio, M.; et al. Disentangling the Impact of Chronic Kidney Disease, Anemia, and Mobility Limitation on Mortality in Older Patients Discharged from Hospital. J. Gerontol. A Biol. Sci. Med. Sci. 2015 , 70 , 1120–1127. [ Google Scholar ] [ CrossRef ]
• Soto, R.; Díaz, L.A.; Rivas, V.; Fuentes-López, E.; Zalaquett, M.; Bruera, M.J.; González, C.; Mezzano, G.; Benítez, C. Frailty and reduced gait speed are independently related to mortality of cirrhotic patients in long-term follow-up. Ann. Hepatol. 2021 , 25 , 100327. [ Google Scholar ] [ CrossRef ]
• Vestergaard, S.; Patel, K.V.; Bandinelli, S.; Ferrucci, L.; Guralnik, J.M. Characteristics of 400-meter walk test performance and subsequent mortality in older adults. Rejuvenation Res. 2009 , 12 , 177–184. [ Google Scholar ] [ CrossRef ]
• Souza, R.; Channick, R.N.; Delcroix, M.; Galiè, N.; Ghofrani, H.-A.; Jansa, P.; Le Brun, F.-O.; Mehta, S.; Perchenet, L.; Pulido, T.; et al. Association between six-minute walk distance and long-term outcomes in patients with pulmonary arterial hypertension: Data from the randomized SERAPHIN trial. PLoS ONE 2018 , 13 , e0193226. [ Google Scholar ] [ CrossRef ]
• Roshanravan, B.; Robinson-Cohen, C.; Patel, K.V.; Ayers, E.; Littman, A.J.; de Boer, I.H.; Ikizler, T.A.; Himmelfarb, J.; Katzel, L.I.; Kestenbaum, B.; et al. Association between physical performance and all-cause mortality in CKD. J. Am. Soc. Nephrol. 2013 , 24 , 822–830. [ Google Scholar ] [ CrossRef ]
• Díaz, P.; Orellana, J.; Soto, C. Incremental Shuttle Walking Test (Incremental Shuttle Walking Test) in healthy children. Rev. Chil. Enfermedades Respir. 2018 , 34 , 160–164. [ Google Scholar ]
• Tudor-Locke, C.; Craig, C.L.; Thyfault, J.P.; Spence, J.C. A step-defined sedentary lifestyle index: <5000 steps/day. Appl. Physiol. Nutr. Metab. 2013 , 38 , 100–114. [ Google Scholar ] [ PubMed ]
• Paluch, A.E.; Gabriel, K.P.; Fulton, J.E.; Lewis, C.E.; Schreiner, P.J.; Sternfeld, B.; Sidney, S.; Siddique, J.; Whitaker, K.M.; Carnethon, M.R. Steps per Day and All-Cause Mortality in Middle-aged Adults in the Coronary Artery Risk Development in Young Adults Study. JAMA Netw. Open 2021 , 4 , e2124516. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Hall, K.S.; Hyde, E.T.; Bassett, D.R.; Carlson, S.A.; Carnethon, M.R.; Ekelund, U.; Evenson, K.R.; Galuska, D.A.; Kraus, W.E.; Lee, I.M.; et al. Systematic review of the prospective association of daily step counts with risk of mortality, cardiovascular disease, and dysglycemia. Int. J. Behav. Nutr. Phys. Act 2020 , 17 , 78. [ Google Scholar ] [ CrossRef ]
• Wang, Y.; Nie, J.; Ferrari, G.; Rey-Lopez, J.P.; Rezende, L.F.M. Association of Physical Activity Intensity with Mortality: A National Cohort Study of 403,681 US Adults. JAMA Intern. Med. 2021 , 181 , 203–211. [ Google Scholar ] [ CrossRef ]
• Kim, J.C.; Kalantar-Zadeh, K.; Kopple, J.D. Frailty and protein-energy wasting in elderly patients with end stage kidney disease. J. Am. Soc. Nephrol. 2013 , 24 , 337–351. [ Google Scholar ] [ CrossRef ]
• Clarke, A.L.; Zaccardi, F.; Gould, D.W.; Hull, K.L.; Smith, A.C.; Burton, J.O.; Yates, T. Association of self-reported physical function with survival in patients with chronic kidney disease. Clin. Kidney J. 2019 , 12 , 122–128. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Sato, T.; Kohzuki, M.; Ono, M.; Muto, M.; Osugi, T.; Kawamura, K.; Naganuma, W.; Sato, M.; Shishito, N. Association between physical activity and change in renal function in patients after acute myocardial infarction. PLoS ONE 2019 , 14 , e0212100. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Clarkson, M.J.; Bennett, P.N.; Fraser, S.F.; Warmington, S.A. Exercise interventions for improving objective physical function in patients with end-stage kidney disease on dialysis: A systematic review and meta-analysis. Am. J. Physiol. Renal Physiol. 2019 , 316 , F856–F872. [ Google Scholar ] [ CrossRef ]
• Joo, Y.S.; Jhee, J.H.; Kim, H.W.; Han, S.H.; Yoo, T.H.; Kang, S.W.; Park, J.T. Physical performance and chronic kidney disease development in elderly adults: Results from a nationwide cohort study. Aging 2020 , 12 , 17393–17417. [ Google Scholar ] [ CrossRef ]
• Bohannon, R.W.; Crouch, R. 1-Minute Sit-to-Stand Test: Systematic review of procedures, performance, and clinimetric properties. J. Cardiopulm. Rehabil. Prev. 2019 , 39 , 2–8. [ Google Scholar ] [ CrossRef ]
• Höglund, J.; Boström, C.; Sundh, J. Six-Minute Walking Test and 30 Seconds Chair-Stand-Test as Predictors of Mortality in COPD—A Cohort Study. Int. J. Chron. Obstruct Pulmon Dis. 2022 , 17 , 2461–2469. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Medina-Mirapeix, F.; Valera-Novella, E.; Morera-Balaguer, J.; Bernabeu-Mora, R. Prognostic value of the five-repetition sit-to-stand test for mortality in people with chronic obstructive pulmonary disease. Ann. Phys. Rehabil. Med. 2022 , 65 , 101598. [ Google Scholar ] [ CrossRef ]
• Morishita, S.; Tsubaki, A.; Shirai, N. Physical function was related to mortality in patients with chronic kidney disease and dialysis. Hemodial. Int. 2017 , 21 , 483–489. [ Google Scholar ] [ CrossRef ]

Click here to enlarge figure

Share and Cite

Nogueira-Pérez, Á.; Ruiz-López-Alvarado, P.; Barril-Cuadrado, G. Can Functional Motor Capacity Influence Mortality in Advanced Chronic Kidney Disease Patients? Nutrients 2024 , 16 , 2689. https://doi.org/10.3390/nu16162689

Nogueira-Pérez Á, Ruiz-López-Alvarado P, Barril-Cuadrado G. Can Functional Motor Capacity Influence Mortality in Advanced Chronic Kidney Disease Patients? Nutrients . 2024; 16(16):2689. https://doi.org/10.3390/nu16162689

Nogueira-Pérez, Ángel, Paloma Ruiz-López-Alvarado, and Guillermina Barril-Cuadrado. 2024. "Can Functional Motor Capacity Influence Mortality in Advanced Chronic Kidney Disease Patients?" Nutrients 16, no. 16: 2689. https://doi.org/10.3390/nu16162689

Article Metrics

Further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Published: 01 August 2024

Making advance care planning easier for adults with kidney disease and their clinicians

• Ryan D. McMahan   ORCID: orcid.org/0000-0001-5572-0513 1 , 2 &
• Rebecca L. Sudore   ORCID: orcid.org/0000-0003-4436-2209 1 , 2  

Nature Reviews Nephrology ( 2024 ) Cite this article

47 Accesses

5 Altmetric

Metrics details

• End-stage renal disease
• Quality of life
• Renal replacement therapy

Advance care planning (ACP) has evolved from a narrow focus on end-of-life preference, such as resuscitation, to a continuum of care planning across the life course. Older adults with kidney disease have high morbidity and mortality, and easy-to-use tools can make ACP easier for patients and clinicians.

This is a preview of subscription content, access via your institution

Access options

Access Nature and 54 other Nature Portfolio journals

Get Nature+, our best-value online-access subscription

24,99 € / 30 days

cancel any time

Subscribe to this journal

Receive 12 print issues and online access

195,33 € per year

only 16,28 € per issue

Buy this article

• Purchase on Springer Link
• Instant access to full article PDF

Prices may be subject to local taxes which are calculated during checkout

United States Renal Data System. 2023 USRDS Annual Data Report: Epidemiology of Kidney Disease in the United States (National Institute of Diabetes and Digestive and Kidney Diseases, 2023).

Sudore, R. L. & Fried, T. R. Redefining the “planning” in advance care planning: preparing for end-of-life decision making. Ann. Intern. Med. 153 , 256–261 (2010).

Article   PubMed   PubMed Central   Google Scholar  

Hickman, S. E., Lum, H. D., Walling, A. M., Savoy, A. & Sudore, R. L. The care planning umbrella: the evolution of advance care planning. J. Am. Geriatr. Soc. 71 , 2350–2356 (2023).

McMahan, R. D., Hickman, S. E. & Sudore, R. L. What clinicians and researchers should know about the evolving field of advance care planning: a narrative review. J. Gen. Intern. Med. 39 , 652–660 (2024).

McMahan, R. D., Tellez, I. & Sudore, R. L. Deconstructing the complexities of advance care planning outcomes: what do we know and where do we go? A scoping review. J. Am. Geriatr. Soc. 69 , 234–244 (2021).

Article   PubMed   Google Scholar  

Wachterman, M. W. et al. One-year mortality after dialysis initiation among older adults. JAMA Intern. Med. 179 , 987–990 (2019).

Rennels, C. F. et al. PREPARE For Your Care and easy-to-read advance directives increase real-time goal concordant care. J. Am. Geriatr. Soc. 71 , 668–670 (2023).

Fried, T. R. Giving up on the objective of providing goal-concordant care: advance care planning for improving caregiver outcomes. J. Am. Geriatr. Soc. 70 , 3006–3011 (2022).

Ghanem, S. et al. Patient-nephrologist prognostic awareness and discordance in end stage renal disease on renal replacement therapy. Int. Urol. Nephrol. 52 , 765–773 (2020).

O’Hare, A. M. et al. Provider perspectives on advance care planning for patients with kidney disease: whose job is it anyway? Clin. J. Am. Soc. Nephrol. 11 , 855–866 (2016).

Zupanc, S. N., Durieux, B. N., Walling, A. M. & Lindvall, C. Bolstering advance care planning measurement using natural language processing. J. Palliat. Med. 27 , 447–450 (2024).

Song, M. K. et al. Effectiveness of an advance care planning intervention in adults receiving dialysis and their families: a cluster randomized clinical trial. JAMA Netw. Open 7 , e2351511 (2024).

Lakin, J. R. et al. Creating KidneyPal: a specialty-aligned palliative care service for people with kidney disease. J. Pain Symptom Manage. 64 , e331–e339 (2022).

Davis, J. L. & Davison, S. N. Hard choices, better outcomes: a review of shared decision-making and patient decision aids around dialysis initiation and conservative kidney management. Curr. Opin. Nephrol. Hypertens. 26 , 205–213 (2017).

Ladin, K. et al. Effectiveness of an intervention to improve decision making for older patients with advanced chronic kidney disease : a randomized controlled trial. Ann. Intern. Med. 176 , 29–38 (2023).

Download references

Author information

Authors and affiliations.

University of California, San Francisco School of Medicine, San Francisco, CA, USA

Ryan D. McMahan & Rebecca L. Sudore

San Francisco Veterans Affairs Health Care System, San Francisco, CA, USA

You can also search for this author in PubMed   Google Scholar

Corresponding authors

Correspondence to Ryan D. McMahan or Rebecca L. Sudore .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Related links.

ACP Decisions: https://www.acpdecisions.org

Five Wishes: https://www.fivewishes.org

KidneyPal: https://www.brighamandwomens.org/medicine/renal/services/kidney-palliative-care

PREPARE For Your Care: https://prepareforyourcare.org

Serious Illness Care Program: https://www.ariadnelabs.org/serious-illness-care

The Conversation Project: https://theconversationproject.org

VitalTalk: https://www.vitaltalk.org

Rights and permissions

Reprints and permissions

About this article

Cite this article.

McMahan, R.D., Sudore, R.L. Making advance care planning easier for adults with kidney disease and their clinicians. Nat Rev Nephrol (2024). https://doi.org/10.1038/s41581-024-00871-9

Download citation

Published : 01 August 2024

DOI : https://doi.org/10.1038/s41581-024-00871-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

• Internal Medicine
• Kidney Diseases
• Chronic Kidney Disease

Treatments for Chronic Kidney Disease: A Systematic Literature Review of Randomized Controlled Trials

• December 2021
• Advances in Therapy 39(11)
• CC BY-NC 4.0

• AstraZeneca
• This person is not on ResearchGate, or hasn't claimed this research yet.

• University of Leicester

Abstract and Figures

Discover the world's research

• 25+ million members
• 160+ million publication pages
• 2.3+ billion citations

• Stephen Nolan
• Juan-Jesus Carrero

• Qiangqiang Gu
• Ronda Esper
• Nicole Maeser
• Vikri Haikal
• Soebagijo Adi Soelistijo

• Steven Fishbane
• Supriya R. Kumar
• Roberto Pecoits-Filho
• J VET PHARMACOL THER
• Jonathan Elliott
• Mark A. Oyama
• J AM SOC NEPHROL
• Hiddo J.L. Heerspink
• David Cherney

• Jazmine Benjamin

• Luana Alves Lima
• Monique Cassiano Lopes
• Thaise Fontes Almeida
• Nathália Campos Moreira Leal Ribeiro
• Tope Olufade
• Lois Lamerato

• Janani Rangaswami
• Oliver Darlington
• Carissa Dickerson
• Molly Murton
• Danielle Goff-Leggett

• Katherine R. Tuttle
• HEALTH QUAL LIFE OUT
• Jacie T. Cooper

• Phil McFarlane
• Ali Ihsan Yagoglu

• Selahattin Erdem
• Ali Ihsan Gunal
• Kenichi Shikata
• Masakazu Haneda
• Toshiharu Ninomiya
• Hirofumi Makino
• George Bakris

• NEW ENGL J MED

• Milton Packer
• Stefan D. Anker
• Javed Butler

• Anushree Tiku

• Recruit researchers
• Join for free
• Login Email Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google Welcome back! Please log in. Email · Hint Tip: Most researchers use their institutional email address as their ResearchGate login Password Forgot password? Keep me logged in Log in or Continue with Google No account? Sign up

• DOI: 10.1080/10408398.2024.2384658
• Corpus ID: 271598019

Effects of dietary compounds on nuclear factor erythroid 2-related factor 2 (Nrf2) modulation in chronic kidney disease: a systematic review of clinical trials.

• Pricilla de Castro Trigueira , K. S. Coutinho-Wolino , +5 authors Milena Barcza Stockler-Pinto
• Published in Critical reviews in food… 31 July 2024

72 References

Systematic review of the nuclear factor erythroid 2-related factor 2 (nrf2) system in human chronic kidney disease: alterations, interventions, and relation to morbidity., effects of resveratrol supplementation in nrf2 and nf-κb expressions in nondialyzed chronic kidney disease patients: a randomized, double-blind, placebo-controlled, crossover clinical trial., nonpharmacologic strategies to modulate nuclear factor erythroid 2-related factor 2 pathway in chronic kidney disease., sulforaphane supplementation did not modulate nrf2 and nf-kb mrna expressions in hemodialysis patients., impact of curcumin supplementation on expression of inflammatory transcription factors in hemodialysis patients: a pilot randomized, double-blind, controlled study., brazil nut consumption modulates nrf2 expression in hemodialysis patients: a pilot study., nutritional strategies to modulate inflammation and oxidative stress pathways via activation of the master antioxidant switch nrf2., antioxidant and anti-inflammatory effects of curcumin/turmeric supplementation in adults: a grade-assessed systematic review and dose-response meta-analysis of randomized controlled trials., phytochemical activators of nrf2: a review of therapeutic strategies in diabetes, nrf2 activation in chronic kidney disease: promises and pitfalls, related papers.

Showing 1 through 3 of 0 Related Papers

Advertisement

A Narrative Review of Chronic Kidney Disease in Clinical Practice: Current Challenges and Future Perspectives

• Open access
• Published: 05 November 2021
• Volume 39 , pages 33–43, ( 2022 )

Cite this article

You have full access to this open access article

• Marc Evans 1 ,
• Ruth D. Lewis 2 ,
• Angharad R. Morgan 2 ,
• Martin B. Whyte 3 ,
• Wasim Hanif 4 ,
• Stephen C. Bain 5 ,
• Sarah Davies 6 ,
• Umesh Dashora 7 ,
• Zaheer Yousef 8 ,
• Dipesh C. Patel 9 &
• W. David Strain 10  

16k Accesses

68 Citations

11 Altmetric

Explore all metrics

Chronic kidney disease (CKD) is a complex disease which affects approximately 13% of the world’s population. Over time, CKD can cause renal dysfunction and progression to end-stage kidney disease and cardiovascular disease. Complications associated with CKD may contribute to the acceleration of disease progression and the risk of cardiovascular-related morbidities. Early CKD is asymptomatic, and symptoms only present at later stages when complications of the disease arise, such as a decline in kidney function and the presence of other comorbidities associated with the disease. In advanced stages of the disease, when kidney function is significantly impaired, patients can only be treated with dialysis or a transplant. With limited treatment options available, an increasing prevalence of both the elderly population and comorbidities associated with the disease, the prevalence of CKD is set to rise. This review discusses the current challenges and the unmet patient need in CKD.

Similar content being viewed by others

Chronic Kidney Disease: Management

Chronic Kidney Disease: Overview

Cardiovascular disease in patients with chronic kidney disease

Avoid common mistakes on your manuscript.

Introduction

Chronic kidney disease (CKD) is a complex and multifaceted disease, causing renal dysfunction and progression to end-stage kidney disease (ESKD) and cardiovascular disease. Complications associated with the disease contribute to the acceleration of CKD progression and risk of cardiovascular-related morbidities.

Despite its high prevalence and the clinical and economic burden of its associated complications, disease awareness remains profoundly low. Worldwide, only 6% of the general population and 10% of the high‐risk population are aware of their CKD statuses [ 1 ]. In addition, CKD recognition in primary care settings is also suboptimal, ranging from 6% to 50%, dependent upon primary care specialty, severity of disease, and experience. Awareness of CKD remains low in part because CKD is usually silent until its late stages. However, diagnosis of CKD during the later stages results in fewer opportunities to prevent adverse outcomes. Physician awareness of CKD is critical for the early implementation of evidence-based therapies that can slow progression of renal dysfunction, prevent metabolic complications, and reduce cardiovascular-related outcomes.

Currently CKD is not curable, and management of the disease relies on treatments which prevent CKD progression and cardiovascular disease. Despite available treatments, a residual risk of adverse events and CKD progression remains. This article reviews the challenges associated with CKD and the treatments available for patients, highlighting the unmet need for cardio-renal protection in patients with CKD.

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

CKD Prevalence

CKD is a global health problem. A meta-analysis of observational studies estimating CKD prevalence showed that approximately 13.4% of the world’s population has CKD [ 2 ]. The majority, 79%, were at late stages of the disease (stage 3–5); however, the actual proportion of people with early CKD (stage 1 or 2) is likely to be much higher since early kidney disease is clinically silent [ 3 ].

Prevalence of CKD appears to be growing rapidly both in the UK and in the Western world. Based on the 2012 subnational population projections for England [ 4 ], the number of people with CKD stage 3–5 is projected to exceed 4 million by 2036 [ 5 ]. This rise in CKD prevalence is due to an increased aging population and prevalence of type 2 diabetes (T2DM), obesity, hypertension and cardiovascular disease that contribute to CKD [ 6 , 7 , 8 ].

The World Health Organization (WHO) estimated that the annual, global number of deaths caused directly by CKD is 5–10 million [ 9 ]. The presence of CKD advances mortality of comorbidities such as cardiovascular diseases, T2DM, hypertension, and infection with human immunodeficiency virus (HIV), malaria and Covid-19, thereby indirectly adding to CKD mortality [ 9 , 10 ]. A contributing cause of high morbidity and mortality associated with CKD is a lack of awareness of the disease, by both patients and providers [ 11 , 12 ]. Early stages of CKD are clinically silent and patients have no symptoms. Lack of treatment at this stage allows CKD to progress through to advanced stages of the disease, where patients may present complications and/or cardiovascular-related comorbidities, or ESKD. Raising awareness of CKD is therefore paramount to allow for early intervention and reduce the risk of comorbidities and mortality.

Classification of CKD

In order to better manage CKD and provide better care for patients, the classification of CKD was developed by the National Kidney Foundation Kidney Disease Outcomes Quality Initiative [ 13 ] and the international guideline group Kidney Disease Improving Global Outcomes (KDIGO) [ 14 ]. CKD stratification is based upon the estimated glomerular filtration rate (eGFR) and albuminuria.

There are six eGFR categories. An eGFR of less than 60 mL/min per 1.73 m 2 for more than 3 months is indicative of impaired renal function and the severity of kidney damage increases with decreasing eGFR measurements. Patients with early onset of the disease, stage 1–2, have normal to mild decreased levels of eGFR (60 to ≥ 90 mL/min per 1.73 m 2 ). Patients with stage 3a–3b have mild to moderate decreased levels of eGFR (45–59 mL/min per 1.73 m 2 , respectively). Severely decreased levels of eGFR, stage 4–5 (15–29 to < 15 mL/min per 1.73 m 2 , respectively), are indicative of advanced stages of the disease and kidney failure.

Stratification also comprises three categories of albuminuria. Patients with an albumin to creatinine ratio (ACR) of 3 to at most 30 mg/mmol are classified as having microalbuminuria and at moderate risk of adverse outcomes. Those with ACR of greater than 30 mg/mmol are classified as having macroalbuminuria and being severely at risk of developing adverse events [ 15 ]. The eGFR and albuminuria categories independently predict adverse outcomes for patients with CKD, and the combination of both increases this risk further [ 16 ]. The CKD classification system aids clinicians in carrying out accurate assessments of CKD severity and other complications which helps to inform decisions associated with the management and monitoring of patients [ 3 , 17 , 18 ].

Clinical Burden of CKD

CKD is a complex disease, involving both non-modifiable (e.g. older age, family history and ethnicity) and modifiable risk factors (e.g. T2DM, hypertension and dyslipidaemia) which are responsible for the initiation of early CKD, CKD progression (stage 3–5) and ESKD.

In early stages of CKD (stage 1–2), factors such as hypertension, obesity and T2DM can trigger kidney function impairment. This causes glomerular/interstitial damage and results in impaired glomerular filtration, leading to decreased eGFR and increased albuminuria. At this stage, even though clinical symptoms do not present, the presence of additional risk factors, including hypertension, hyperglycaemia, smoking, obesity, dyslipidaemia and cardiovascular disease, may accelerate CKD progression and result in ESKD.

As the disease progresses, the clinical and economic burden of CKD increases (Fig.  1 ) as complications such as CKD mineral bone disorder, anaemia, hypertension and hyperkalaemia may occur and advanced stages of CKD, stage 4–5, ensue. Clinical symptoms, such as fatigue, itching of the skin, bone or joint pain, muscle cramps and swollen ankles, feet or hands, are often present at this stage [ 19 ]. Further deterioration of kidney function causes tubular and glomerular hypertrophy, sclerosis and fibrosis, leading to a significant reduction in eGFR, extreme albuminuria and kidney failure.

A schematic diagram showing the association between CKD progression and clinical and economic burden. Symptoms of CKD typically present during advanced stages of the disease where patients are at increased risk of cardiovascular disease and other comorbidities

Even though CKD progression may lead to kidney failure and renal death, patients with CKD are more likely to die from cardiovascular-related complications before reaching ESKD [ 20 ]. A study using data from a meta-analysis involving 1.4 million individuals found a significant increased risk of cardiovascular-related mortality, even in stage 2 of CKD (eGFR levels < 90 mL/min per 1.73 m 2 ) [ 16 , 21 , 22 ].

As the disease progresses, the risk of cardiovascular disease is markedly increased, such that 50% of patients with late-stage CKD, stage 4–5, have cardiovascular disease. The risk of atrial fibrillation (AF) and acute coronary syndrome (ACS) is doubled in patients with eGFR < 60 mL/min per 1.73 m 2 . AF is associated with a threefold higher risk of progression to ESKD. The incidence of heart failure (HF) is also threefold greater in patients with eGFR < 60 mL/min per 1.73 m 2 compared with > 90 mL/min per 1.73 m 2 and HF is associated with CKD progression, hospitalisation and death [ 23 ].

The increased risk of cardiovascular disease in patients with CKD is due in part to the traditional risk factors associated with cardiovascular disease such as hypertension, T2DM and dyslipidaemia. For instance, a large observational database linked study (Third National Health and Nutrition Examination Survey (NHANES) III) found a strong association between CKD and T2DM combined and an increased risk of mortality [ 24 ]. In this study, the authors observed a 31.1% mortality rate in patients with CKD and diabetes, compared to 11.5% in people with diabetes only. An observational study using both US and UK linked databases showed that the presence of both CKD and T2DM was related to increased risk of major adverse cardiac events (MACE), HF and arrhythmogenic cardiomyopathy (ACM) [ 25 ]. This risk was further elevated in older patients with atherosclerotic cardiovascular disease [ 25 ]. Similarly, the presence of both CKD and T2DM leads to a significant increased risk of all-cause and cardiovascular-related mortality versus T2DM alone [ 24 ].

The direct renal effect on cardiovascular disease is due to generalised inflammatory change, cardiac remodelling, narrowing of the arteries and vascular calcification, both contributing to the acceleration of vascular ageing and atherosclerotic processes, and leading to myocardial infarction, stroke and HF [ 26 ].

Together, these studies highlight the strong relationship which exists between CKD progression, number of comorbidities and heightened risk of cardiovascular disease and cardiovascular-related mortality.

Economic Burden of CKD

In addition to the clinical burden, management of CKD also requires significant healthcare resources and utilisation. In 2009–2010, the estimated cost of CKD to the National Health Service (NHS) in England was £1.45 billion [ 27 ]. Furthermore, in 2016, US Medicare combined expenditure for CKD and ESKD exceeded $114 billion (£86 billion) [ 28 ].

Although estimating the true cost of early CKD is difficult because of the lack of data available for unreported cases, CKD progression is associated with increased healthcare costs [ 29 , 30 ]. A study by Honeycutt et al. combined laboratory data from NHANES with expenditure data from Medicare and found that costs of CKD management increased with disease progression [ 29 ]. Estimated annual medical costs of CKD per person were not significant at stage 1, $1700 at stage 2, $3500 at stage 3 and $12,700 at stage 4.

Healthcare costs associated with early CKD are more likely to be from the sequalae of comorbid disease rather than kidney disease. Hence, patients with CKD stage 1 or 2 are at increased risk of hospitalisation if they also have T2DM (9%), cardiovascular disease (more than twofold), and both cardiovascular disease and T2DM (approximately fourfold) [ 31 ].

ESKD accounts for the largest proportion of CKD management costs. In 2009–2010, 50% of the overall CKD cost to NHS (England) was due to renal replacement therapy (RRT), which accounted for 2% of the CKD population [ 27 ]. The other 50% included renal primary care costs, such as treatment costs for hypertension and tests, consultation costs, non-renal care attributable to CKD and renal secondary care costs. Approximately £174 million was estimated for the annual cost of myocardial infarctions and strokes associated with CKD [ 27 ].

More recently, an economic analysis investigated the burden associated with the management of cardiovascular-related morbidity and mortality in CKD, according to the KDIGO categorisation of both eGFR and albuminuria [ 15 ]. Decreased eGFR levels increased both the risk of adverse clinical outcomes and economic costs, and albuminuria elevated this risk significantly. Furthermore, CKD progression correlated with increased CKD management costs and bed days. Stage 5 CKD (versus stage 1 (or without) CKD) per 1000 patient years was associated with £435,000 in additional costs and 1017 bed days.

The significant economic burden associated with CKD progression and ESKD highlights the importance of optimising CKD management and the unmet need for better treatment options in slowing disease progression in this patient population. Thus, early detection and intervention to slow the progression of the disease has the potential to make substantial savings in healthcare costs.

Current CKD Management Strategies

KDIGO and National Institute for Health and Care Excellence (NICE) have produced detailed guidelines for the evaluation and management of CKD [ 3 , 32 , 33 ]. Both recommend implementing strategies for early diagnosis of the disease in order to reduce the risk of cardiovascular disease, attenuate CKD progression and decrease the incidence of ESKD in this patient population. CKD is a complex disease and thus treatment requires a multifaceted approach utilising both non-pharmacological, e.g. diet and exercise regimes and pharmacological interventions such as antihypertensive and antihyperglycemic drugs [ 34 ]. There has, however, been no significant breakthrough in this area for over 2 decades.

The effect of lifestyle intervention on reducing disease progression is still unclear, although increased physical activity has been shown to slow the rate of eGFR decline [ 35 ] and ESKD progression [ 36 ], improve eGFR levels [ 35 ] and albuminuria [ 37 ], and reduce mortality in patients with CKD [ 35 , 38 , 39 , 40 ]. Similarly, diet regimes such as low-protein diet or Mediterranean diet reduce renal function decline and mortality rate in CKD [ 41 , 42 ]. Hence, dietary advice is recommended in accordance with CKD severity to control for daily calorie, salt, potassium, phosphate and protein intake [ 3 , 33 ]. However, patients with consistently elevated serum phosphate levels or metabolic acidosis [low serum bicarbonate levels (< 22 mmol/l)], associated with increased risk of CKD progression and death, may be treated with phosphate binding agents (e.g. aluminium hydroxide and calcium carbonate) or sodium bicarbonate, respectively [ 3 ].

To reduce the risk of cardiovascular disease, KDIGO and NICE recommend active lipid management and blood pressure control [ 33 , 43 , 44 ]. In early CKD stages 1 and 2, statins are recommended for all patients over 50 years of age, whilst in stage 3 and advanced stages of the disease, stage 4–5 (eGFR < 60 mL/min per 1.73 m 2 ), a combination of statins and ezetimibe is advised [ 43 ].

Management of hypertension includes a target blood pressure of less than 140/90 mmHg for patients with CKD and hypertension and less than 130/80 mmHg for patients with CKD and T2DM, and also in patients with albuminuria [ 3 , 32 ], alongside blood pressure lowering therapies and renin–angiotensin–aldosterone system (RAAS) blocking agents, such as angiotensin receptor blockers (ARB) or angiotensin-converting enzyme inhibitors (ACEi). As such, RAAS inhibitors (RAASi) are currently recommended to treat patients with diabetes, hypertension and albuminuria in CKD [ 45 ]. These RAAS blocking agents confer both renal and cardiovascular protection and are recommended as first-line treatment to treat hypertension in patients with CKD [ 34 , 46 ].

The clinical benefits of RAASi have been demonstrated in patients with CKD with and without diabetes [ 47 , 48 , 49 ]. These clinical benefits are in addition to their effects on reducing blood pressure and albuminuria, including a reduction in eGFR decline and a decreased risk of ESKD cardiac-related morbidity and all-cause mortality [ 47 , 48 , 49 ]. Nevertheless, despite their benefits, RAASi treatment can induce hyperkalaemia, and patients are often advised to reduce RAASi dosage or even discontinue their treatment, which prevents optimum clinical benefits of RAASi therapy being reached. In this instance, combination therapy with potassium binding agents, such as patiromer and sodium zirconium cyclosilicate, may be used alongside RAASi therapy to reduce RAASi-associated hyperkalaemia.

However long-term trials will be required to determine their effect on cardiovascular morbidity and mortality in CKD [ 50 , 51 , 52 ]. Despite these therapies being the mainstay of CKD management, there is still a residual risk of CKD progression and an unmet need for new treatments.

Novel/Emerging Treatments for CKD Management

Over the last 2 years, novel therapeutic approaches for CKD management have emerged, with particular attention on mineralocorticoid receptor antagonists (MRAs) and sodium–glucose co-transporter 2 (SGLT2) inhibitors. The clinical effectiveness of finerenone, a selective oral, non-steroidal MRA, has recently been demonstrated to lower risks of CKD progression and cardiovascular events in diabetic kidney disease (DKD) [ 53 ]. Finerenone is under review for approval by the European Medicines Agency (EMA) and US Food and Drug Administration (FDA).

Of these new and emerging therapies, SGTL2i offers the most clinical benefit with both cardiovascular and renal protective effects, independent of glucose lowering. Clinical trials of SGTL2 in T2DM with and without CKD overall showed a 14–31% reduction in cardiovascular endpoints including hospitalisation for HF and MACE and a 34–37% reduction in hard renal-specific clinical endpoints including a sustained reduction in eGFR, progression of albuminuria and progression to ESKD [ 54 , 55 , 56 , 57 , 58 ]. CREDENCE, was a double-blind, multicentre, randomised trial in diabetic patients with albuminuric CKD (eGFR 30 to < 90 mL/min per 1.73 m 2 and ACR ≥ 30 mg/mmol) [ 57 ]. In this trial, canagliflozin reduced the relative risk of the composite of ESKD, doubling of serum creatinine and renal-related mortality by 34%, relative risk of ESKD by 34% and risk of cardiovascular-related morbidity, including myocardial infarction and stroke, and mortality.

The SGTL2i dapagliflozin has proven its effectiveness in slowing CKD progression in addition to reducing cardiovascular risk in early stages of CKD. The DECLARE-TIMI58 trial involved 17,160 diabetic patients with established atherosclerotic cardiovascular disease and early-stage CKD (mean eGFR was 85.2 mL/min per 1.73 m 2 ) and were randomised to receive either dapagliflozin or placebo. Following a median follow-up of 4.2 years, there was a significant reduction in renal composite endpoints with dapagliflozin versus placebo, with a 46% reduction in sustained decline of at least 40% eGFR to less than 60 mL/min per 1.73 m 2 and a reduction in ESKD (defined as dialysis for at least 90 days, kidney transplantation, or confirmed sustained eGFR < 15 mL/min per 1.73 m 2 ) or renal death.

More recently, these cardiovascular and renal protective effects of SGTLT2i have also been demonstrated in a broad range of patients with more advanced stages of CKD (mean eGFR was 43.1 ± 12.4 mL/min per 1.73 m 2 ) without diabetes [ 58 , 59 ]. In the DAPA-CKD trial, many patients were without diabetes, including IgA nephropathy, ischemic/hypertension nephropathy and other glomerulonephritis [ 59 ]. Patients receiving dapagliflozin had a 39% relative risk reduction in the primary composite outcomes of a sustained decline in eGFR of at least 50%, ESKD and renal- or cardiovascular-related mortality and a 31% relative risk reduction of all-cause mortality compared to placebo [ 58 , 60 ]. Safety outcomes from clinical trials of dapagliflozin have also shown similar incidences of adverse events in both placebo and dapagliflozin arms [ 58 , 61 ].

The clinical benefits and safety outcomes from these trials highlight the potential use of SGTL2i in reducing cardiovascular burden and CKD progression in a broad range of CKD aetiologies at early and late stages where there is an unmet need. Currently, SGTL2i class drugs, including canagliflozin, dapagliflozin and empagliflozin, are approved by the US FDA for the treatment of T2DM and, more recently, dapagliflozin and canagliflozin for CKD and DKD respectively [ 62 , 63 ]. In addition, SGTL2i has been recommended for approval in the European Union (EU) by the Committee for Medicinal Products for Human Use (CHMP) of the EMA, for the treatment of CKD in adults with and without T2D [ 64 ]. Hence, there is now a need to raise awareness of the clinical applicability of these drugs in CKD to ensure full utilisation and maximum benefits are met, for both patients and providers.

This narrative review has summarised some of the key challenges associated with CKD. Early stages of the disease are clinically silent which prevents early intervention to slow the progression of the disease and allows progression of CKD and ESKD. At advanced stages of the disease, when clinical symptoms are present, patients with CKD are already at heightened risk of cardiovascular-related morbidity and mortality. Hence, advanced stages of CKD and ESKD are associated with poor outcomes and a significant clinical and economic burden.

At present, there are no treatments to cure CKD; as such, strategies for CKD management have been developed to target the modifiable risk factors in order to reduce cardiovascular disease morbidity in patients with CKD and slow the progression of CKD to ESKD. However, despite available treatment options, residual risk of adverse events and CKD progression remain; hence, an unmet need exists in CKD treatment. SGTL2i have the potential to fill this gap, with recent evidence from clinical trials showing a reduction in cardiovascular and renal adverse endpoints in a broad range of patients with CKD.

Ene-Iordache B, Perico N, Bikbov B, et al. Chronic kidney disease and cardiovascular risk in six regions of the world (ISN-KDDC): a cross-sectional study. Lancet Glob Health. 2016;4(5):e307–19.

Article   PubMed   Google Scholar  

Hill NR, Fatoba ST, Oke JL, et al. Global prevalence of chronic kidney disease–a systematic review and meta-analysis. PLoS ONE. 2016;11(7):e0158765.

Article   PubMed   PubMed Central   CAS   Google Scholar  

Kidney International Organisation. KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease. https://kdigo.org/wp-content/uploads/2017/02/KDIGO_2012_CKD_GL.pdf . 2012. Accessed May 2021.

Office for National Statistics: Subnational population projections for England: 2012-based. https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationprojections/bulletins/subnationalpopulationprojectionsforengland/2014-05-29 . 2014. Accessed May 2021.

Public Health England: Chronic kidney disease prevalence model. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/612303/ChronickidneydiseaseCKDprevalencemodelbriefing.pdf . 2014. Accessed May 2021.

Jha V, Garcia-Garcia G, Iseki K, et al. Chronic kidney disease: global dimension and perspectives. Lancet. 2013;382(9888):260–72.

Go AS, Chertow GM, Fan D, McCulloch CE, Hsu C-Y. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med. 2004;351(13):1296–305.

Article   CAS   PubMed   Google Scholar  

Thomas R, Kanso A, Sedor JR. Chronic kidney disease and its complications. Primary care Clin Office Pract. 2008;35(2):329–44.

Article   Google Scholar  

World Health Organization. The global burden of kidney disease and the sustainable development goals. https://www.who.int/bulletin/volumes/96/6/17-206441/en/ . 2018. Accessed Apr 2021.

Pecly IMD, Azevedo RB, Muxfeldt ES, et al. COVID-19 and chronic kidney disease: a comprehensive review. J Bras Nefrol. 2021;43(3):383–99.

Tuot DS, Wong KK, Velasquez A, et al. CKD awareness in the general population: performance of CKD-specific questions. Kidney Med. 2019;1(2):43–50.

Article   PubMed   PubMed Central   Google Scholar  

Plantinga LC, Tuot DS, Powe NR. Awareness of chronic kidney disease among patients and providers. Adv Chronic Kidney Dis. 2010;17(3):225–36.

Levey AS, Coresh J, Bolton K, et al. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis. 2002;39(2 Suppl 1):S1–266.

Google Scholar  

Levey AS, Eckardt K-U, Tsukamoto Y, et al. Definition and classification of chronic kidney disease: a position statement from Kidney Disease: Improving Global Outcomes (KDIGO). Kidney Int. 2005;67(6):2089–100.

Darlington O, Dickerson C, Evans M, et al. Costs and healthcare resource use associated with risk of cardiovascular morbidity in patients with chronic kidney Disease: evidence from a systematic literature review. Adv Ther. 2021;38(2):994–1010.

Gansevoort RT, Correa-Rotter R, Hemmelgarn BR, et al. Chronic kidney disease and cardiovascular risk: epidemiology, mechanisms, and prevention. Lancet. 2013;382(9889):339–52.

Tonelli M, Muntner P, Lloyd A, et al. Using proteinuria and estimated glomerular filtration rate to classify risk in patients with chronic kidney disease: a cohort study. Ann Intern Med. 2011;154(1):12–21.

Levey AS, Tangri N, Stevens LA. Classification of chronic kidney disease: a step forward. Ann Intern Med. 2011;154(1):65–7.

Abdel-Kader K, Unruh ML, Weisbord SD. Symptom burden, depression, and quality of life in chronic and end-stage kidney disease. Clin J Am Soc Nephrol. 2009;4(6):1057–64.

Tuegel C, Bansal N. Heart failure in patients with kidney disease. Heart. 2017;103(23):1848–53.

Chronic Kidney Disease Prognosis Consortium. Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet. 2010;375(9731):2073–81.

Van Der Velde M, Matsushita K, Coresh J, et al. Lower estimated glomerular filtration rate and higher albuminuria are associated with all-cause and cardiovascular mortality. A collaborative meta-analysis of high-risk population cohorts. Kidney Int. 2011;79(12):1341–52.

Article   PubMed   CAS   Google Scholar  

Kottgen A, Russell SD, Loehr LR, et al. Reduced kidney function as a risk factor for incident heart failure: the atherosclerosis risk in communities (ARIC) study. J Am Soc Nephrol. 2007;18(4):1307–15.

Afkarian M, Sachs MC, Kestenbaum B, et al. Kidney disease and increased mortality risk in type 2 diabetes. J Am Soc Nephrol. 2013;24(2):302–8.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Cherney DZ, Repetto E, Wheeler DC, et al. Impact of cardio-renal-metabolic comorbidities on cardiovascular outcomes and mortality in type 2 diabetes mellitus. Am J Nephrol. 2020;51(1):74–82.

Stevens P, O’Donoghue D, De Lusignan S, Van Vlymen J, Klebe B, Middleton R, et al. Chronic kidney disease management in the United Kingdom: NEOERICA project results. Kidney Int. 2007;72(1):92–9.

Kerr M, Bray B, Medcalf J, O’Donoghue DJ, Matthews B. Estimating the financial cost of chronic kidney disease to the NHS in England. Nephrol Dial Transplant. 2012;27(suppl_3):73–80.

Saran R, Robinson B, Abbott KC, et al. US renal data system 2017 annual data report: epidemiology of kidney disease in the United States. Am J Kidney Dis. 2018;71(3):A7.

Honeycutt AA, Segel JE, Zhuo X, Hoerger TJ, Imai K, Williams D. Medical costs of CKD in the Medicare population. J Am Soc Nephrol. 2013;24(9):1478–83.

Smith DH, Gullion CM, Nichols G, Keith DS, Brown JB. Cost of medical care for chronic kidney disease and comorbidity among enrollees in a large HMO population. J Am Soc Nephrol. 2004;15(5):1300–6.

Wang V, Vilme H, Maciejewski ML, Boulware LE. The economic burden of chronic kidney disease and end-stage renal disease. Semin Nephrol. 2016;36:319–30.

National Institute for Health Care Excellence. Chronic kidney disease in adults: assessment and management. https://www.nice.org.uk/guidance/cg182 . 2014. Accessed Apr 2021.

National Institute for Health and Care Excellence. Guideline chronic kidney disease. https://www.nice.org.uk/guidance/gid-ng10118/documents/draft-guideline . 2021. Accessed June 2021.

National Institute for Health Care Excellence: Management of chronic kidney disease. https://pathways.nice.org.uk/pathways/chronic-kidney-disease . Accessed Apr 2021.

Chen I-R, Wang S-M, Liang C-C, et al. Association of walking with survival and RRT among patients with CKD stages 3–5. Clin J Am Soc Nephrol. 2014;9(7):1183–9.

Robinson-Cohen C, Littman AJ, Duncan GE, et al. Physical activity and change in estimated GFR among persons with CKD. J Am Soc Nephrol. 2014;25(2):399–406.

Hellberg M, Höglund P, Svensson P, Clyne N. Randomized controlled trial of exercise in CKD—the RENEXC study. Kidney Int Rep. 2019;4(7):963–76.

MacKinnon HJ, Wilkinson TJ, Clarke AL, et al. The association of physical function and physical activity with all-cause mortality and adverse clinical outcomes in nondialysis chronic kidney disease: a systematic review. Therap Adv Chronic Dis. 2018;9(11):209–26.

Beddhu S, Baird BC, Zitterkoph J, Neilson J, Greene T. Physical activity and mortality in chronic kidney disease (NHANES III). Clin J Am Soc Nephrol. 2009;4(12):1901–6.

Clarke AL, Zaccardi F, Gould DW, et al. Association of self-reported physical function with survival in patients with chronic kidney disease. Clin Kidney J. 2019;12(1):122–8.

Chauveau P, Aparicio M, Bellizzi V, et al. Mediterranean diet as the diet of choice for patients with chronic kidney disease. Nephrol Dial Transplant. 2018;33(5):725–35.

Rhee CM, Ahmadi SF, Kovesdy CP, Kalantar-Zadeh K. Low-protein diet for conservative management of chronic kidney disease: a systematic review and meta-analysis of controlled trials. J Cachexia Sarcopenia Muscle. 2018;9(2):235–45.

Wanner C, Tonelli M. KDIGO Clinical Practice Guideline for Lipid Management in CKD: summary of recommendation statements and clinical approach to the patient. Kidney Int. 2014;85(6):1303–9.

National Institute for Health and Care Excellence. Cardiovasvcular disease: risk assessment and reduction, including lipid modification. https://www.nice.org.uk/guidance/cg181/resources/cardiovascular-disease-risk-assessment-and-reduction-including-lipid-modification-pdf-35109807660997 . 2014. Accessed Apr 2021.

de Boer IH, Caramori ML, Chan JC, et al. KDIGO 2020 clinical practice guideline for diabetes management in chronic kidney disease. Kidney Int. 2020;98(4):S1–115.

National Collaborating Centre for Chronic Conditions. Chronic kidney disease: national clinical guideline for early identification and management in adults in primary and secondary care. London: Royal College of Physicians; 2008.

Jafar TH, Schmid CH, Landa M, et al. Angiotensin-converting enzyme inhibitors and progression of nondiabetic renal disease: a meta-analysis of patient-level data. Ann Intern Med. 2001;135(2):73–87.

Evans M, Bain SC, Hogan S, Bilous RW. Irbesartan delays progression of nephropathy as measured by estimated glomerular filtration rate: post hoc analysis of the Irbesartan Diabetic Nephropathy Trial. Nephrol Dial Transplant. 2012;27(6):2255–63.

Xie X, Liu Y, Perkovic V, et al. Renin-angiotensin system inhibitors and kidney and cardiovascular outcomes in patients with CKD: a Bayesian network meta-analysis of randomized clinical trials. Am J Kidney Dis. 2016;67(5):728–41.

Wai H-T, Meah N, Katira R. Novel potassium binders: a clinical update. https://bjcardio.co.uk/2021/04/novel-potassium-binders-a-clinical-update-3/ . Accessed June 2021.

National Institute for Health and Care Excellence. Sodium zirconium cyclosilicate for treating hyperkalaemia. Technology appraisal guidance TA599. https://www.nice.org.uk/guidance/TA599 (2019). Accessed April 2021.

National Institute for Health and Care Excellence. Patiromer for treating hyperkalaemia. https://www.nice.org.uk/guidance/ta623/chapter/1-Recommendations . 2020. Accessed May 2021.

Bakris GL, Agarwal R, Anker SD, et al. Effect of finerenone on chronic kidney disease outcomes in type 2 diabetes. N Engl J Med. 2020;383(23):2219–29.

Zinman B, Wanner C, Lachin J. Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes. N Engl J Med. 2015;373(22):2117–28.

Neal B, Perkovic V, Mahaffey KW, et al. Canagliflozin and cardiovascular and renal events in type 2 diabetes. N Engl J Med. 2017;377(7):644–57.

Wiviott SD, Raz I, Bonaca MP, et al. Dapagliflozin and cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2019;380(4):347–57.

Perkovic V, Jardine MJ, Neal B, et al. Canagliflozin and renal outcomes in type 2 diabetes and nephropathy. N Engl J Med. 2019;380(24):2295–306.

Heerspink HJ, Stefánsson BV, Correa-Rotter R, et al. Dapagliflozin in patients with chronic kidney disease. N Engl J Med. 2020;383(15):1436–46.

Wheeler DC, Stefansson BV, Batiushin M, et al. The dapagliflozin and prevention of adverse outcomes in chronic kidney disease (DAPA-CKD) trial: baseline characteristics. Nephrol Dial Transplant. 2020;35(10):1700–11.

Heerspink HJ, Sjöström CD, Jongs N, et al. Effects of dapagliflozin on mortality in patients with chronic kidney disease: a pre-specified analysis from the DAPA-CKD randomized controlled trial. Eur Heart J. 2021;42(13):1216–27.

McMurray JJ, Solomon SD, Inzucchi SE, et al. Dapagliflozin in patients with heart failure and reduced ejection fraction. N Engl J Med. 2019;381(21):1995–2008.

U.S. Food & Drug Administration (FDA). FDA approves treatment for chronic kidney disease. https://www.fda.gov/news-events/press-announcements/fda-approves-treatment-chronic-kidney-disease . 2021. Accessed May 2021.

Janssen Pharmaceutical of Johnson & Johnson. U.S. FDA approves INVOKANA® (canagliflozin) to treat diabetic kidney disease (DKD) and reduce the risk of hospitalization for heart failure in patients with type 2 diabetes (T2D) and DKD. https://www.prnewswire.com/news-releases/us-fda-approves-invokana-canagliflozin-to-treat-diabetic-kidney-disease-dkd-and-reduce-the-risk-of-hospitalization-for-heart-failure-in-patients-with-type-2-diabetes-t2d-and-dkd-300927348.html . 2019. Accessed May 2021.

European Medicines Agency. Forxiga: Summary of opinion (post authorisation). https://www.ema.europa.eu/en/medicines/human/EPAR/forxiga . 2021. Accessed May 2021.

Download references

Acknowledgements

This manuscript was supported by a grant from AstraZeneca UK Ltd. in respect of medical writing and publication costs (the journal’s Rapid Service and Open Access Fees). AstraZeneca has not influenced the content of the publication, and has reviewed this document for factual accuracy only.

All named authors meet the International Committee of Medical Journal Editors (ICMJE) criteria for authorship for this article, take responsibility for the integrity of the work as a whole, and have given their approval for this version to be published.

Authors’ Contributions

All authors have been involved in design of this review. Marc Evans, Ruth D. Lewis, and Angharad R. Morgan produced the primary manuscript. All authors have contributed to the drafting and revision of the manuscript and have approved the final version for publication. Marc Evans is responsible for the integrity of the work as a whole.

Disclosures

Marc Evans reports honoraria from AstraZeneca, Novo Nordisk, Takeda and NAPP, and research support from Novo Nordisk outside the submitted work. Ruth D. Lewis and Angharad R Morgan are employees of Health Economics and Outcomes Research Ltd., Cardiff, UK who received fees from AstraZeneca in relation to this study. Martin B. Whyte reports investigator-led research grants from Sanofi, Eli Lilly and AstraZeneca and personal fees from AstraZeneca, Boehringer Ingelheim and MSD outside the submitted work. Wasim Hanif reports grants and personal fees from AstraZeneca, grants and personal fees from Boerhinger Inglhiem, grants and personal fees from NAPP, from MSD, outside the submitted work. Stephen C. Bain reports personal fees and other from Abbott, personal fees and other from AstraZeneca, personal fees and other from Boehringer Ingelheim, personal fees and other from Eli Lilly, personal fees and other from Merck Sharp & Dohme, personal fees and other from Novo Nordisk, personal fees and other from Sanofi-aventis, other from Cardiff University, other from Doctors.net, other from Elsevier, other from Onmedica, other from Omnia-Med, other from Medscape, other from All-Wales Medicines Strategy Group, other from National Institute for Health and Care Excellence (NICE) UK, and other from Glycosmedia, outside the submitted work. PAK reports personal fees for lecturing from AstraZeneca, Boehringer Inglhiem, NAPP, MundiPharma and Novo Nordisk outside the submitted work. Sarah Davies has received honorarium from AstraZeneca, Boehringer Ingelheim, Lilly, Novo Nordisk, Takeda, MSD, NAPP, Bayer and Roche for attending and participating in educational events and advisory boards, outside the submitted work. Umesh Dashora reports personal fees from AstraZeneca, NAPP, Sanofi, Boehringer Inglhiem, Lilly and Novo Nordisk, outside the submitted work. Zaheer Yousef reports personal fees from AstraZeneca, personal fees from Lilly, personal fees from Boehringer Ingelheim and personal fees from Novartis outside the submitted work. Dipesh C. Patel reports personal fees from AstraZeneca, personal fees from Boehringer Ingelheim, personal fees from Eli Lilly, non-financial support from NAPP, personal fees from Novo Nordisk, personal fees from MSD, personal fees and non-financial support from Sanofi outside the submitted work. In addition, DCP is an executive committee member of the Association of British Clinical Diabetologists and member of the CaReMe UK group. W. David Strain holds research grants from Bayer, Novo Nordisk and Novartis and has received speaker honoraria from AstraZeneca, Bayer, Bristol-Myers Squibb, Merck, NAPP, Novartis, Novo Nordisk and Takeda. WDS is supported by the NIHR Exeter Clinical Research Facility and the NIHR Collaboration for Leadership in Applied Health Research and Care (CLAHRC) for the South West Peninsula.

Compliance with Ethics Guidelines

This article is based on previously conducted studies and does not contain any studies with human participants or animals performed by any of the authors.

Author information

Authors and affiliations.

Diabetes Resource Centre, University Hospital Llandough, Penlan Road, Llandough, Cardiff, CF64 2XX, UK

Health Economics and Outcomes Research Ltd, Cardiff, UK

Ruth D. Lewis & Angharad R. Morgan

Faculty of Health and Medical Sciences, University of Surrey, Guildford, UK

Martin B. Whyte

University Hospitals Birmingham NHS Foundation Trust, Birmingham, UK

Wasim Hanif

Diabetes Research Unit, Swansea University Medical School, Swansea, UK

Stephen C. Bain

Woodlands Medical Centre, Cardiff, UK

Sarah Davies

East Sussex Healthcare NHS Trust, Eastbourne, UK

Umesh Dashora

Department of Cardiology, University Hospital of Wales and Cardiff University, Cardiff, UK

Zaheer Yousef

Division of Medicine, Department of Diabetes, University College London, Royal Free Campus, London, UK

Dipesh C. Patel

Diabetes and Vascular Research Centre, University of Exeter Medical School, Exeter, UK

W. David Strain

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Marc Evans .

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, which permits any non-commercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc/4.0/ .

Reprints and permissions

About this article

Evans, M., Lewis, R.D., Morgan, A.R. et al. A Narrative Review of Chronic Kidney Disease in Clinical Practice: Current Challenges and Future Perspectives. Adv Ther 39 , 33–43 (2022). https://doi.org/10.1007/s12325-021-01927-z

Download citation

Received : 24 August 2021

Accepted : 15 September 2021

Published : 05 November 2021

Issue Date : January 2022

DOI : https://doi.org/10.1007/s12325-021-01927-z

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Cardiovascular disease
• Chronic kidney disease
• Diabetic kidney disease
• Sodium–glucose co-transporter 2 inhibitors
• Find a journal
• Publish with us
• Track your research

• Open access
• Published: 12 August 2024

Systematic literature review of the somatic comorbidities experienced by adults with phenylketonuria

• Kaleigh B. Whitehall   ORCID: orcid.org/0009-0009-9090-3580 1 ,
• Sarah Rose 1 ,
• Gillian E. Clague 1 ,
• Kirsten K. Ahring 2 ,
• Deborah A. Bilder 3 ,
• Cary O. Harding 4 ,
• Álvaro Hermida 5 ,
• Anita Inwood 6 ,
• Nicola Longo 7 ,
• François Maillot 8 ,
• Ania C. Muntau 9 ,
• André L. S. Pessoa 10 , 11 ,
• Júlio C. Rocha 12 , 13 , 14 ,
• Fran Rohr 15 ,
• Serap Sivri 16 ,
• Jack Said 17 ,
• Sheun Oshinbolu 18 &
• Gillian C. Sibbring 17  

Orphanet Journal of Rare Diseases volume  19 , Article number:  293 ( 2024 ) Cite this article

Metrics details

Phenylketonuria (PKU) is an inborn error of phenylalanine (Phe) metabolism that, if untreated, causes Phe accumulation in the brain leading to neurophysiologic alterations and poor outcomes. Lifelong management centers on dietary Phe restriction, yet long-term complete metabolic control is unachievable for many adults. High blood Phe levels or chronic Phe and intact protein restriction in the diet may lead to somatic comorbidities. A systematic literature review was conducted to evaluate somatic comorbidities experienced by adults with PKU.

Clinical and observational studies reporting somatic comorbidities experienced by individuals with PKU aged ≥ 16 years (or classified as adults) evaluating a Phe-restricted diet with or without pharmacologic therapy versus no therapeutic intervention (including healthy controls), or pharmacologic therapy versus a Phe-restricted diet alone, were identified. PubMed® was searched (February 1, 2022 and updated November 1, 2023), using a pre-defined search strategy, followed by two-stage screening and data extraction. Included studies were grouped by PKU population comparison.

1185 records were screened; 51 studies across 12,602 individuals were extracted. Bone-related abnormalities were the most reported outcome ( n  = 21); several outcome measures were used. Original study groupings included: Phe-restricted diet versus healthy controls or reference values ( n  = 40); treatment-adherent versus those non-adherent ( n  = 12). Additional groups added as part of a protocol amendment included: different Phe-restricted diets ( n  = 4); severe versus less severe disease ( n  = 5). Vote counting indicated a higher burden of ≥ 1 comorbidity (or outcome measure) for the Phe-restricted diet group by 37 of 38 studies included in the analysis of Phe-restricted diet versus healthy controls; higher burden in healthy controls was reported in 12 studies. Vote counting was similar between those treatment adherent ( n  = 7) versus non-adherent ( n  = 10).

Conclusions

Adults with PKU have a higher comorbidity burden than a non-PKU population. More robust studies are needed to better understand the relationship between effective metabolic control and comorbidity burden, using consistent outcome measures.

This SLR was supported by BioMarin Pharmaceutical Inc., Novato, CA, and is registered with the Research Registry (reviewregistry1476).

Phenylketonuria (PKU), as a colloquial term for phenylalanine hydroxylase (PAH) deficiency (OMIM# 261600), is an autosomal recessive inborn error of amino acid metabolism. PKU is caused by pathogenic variants in the gene encoding PAH, impairing enzyme function such that PAH cannot metabolize phenylalanine (Phe) to tyrosine normally. Phe accumulates in the blood, crossing the blood–brain barrier at high concentrations with toxic effects. Phe also competes with other large neutral amino acids for transport across the blood–brain barrier by the L-type amino acid transporter 1 (LAT1); high concentrations of Phe may block transport of other LAT1 substrates into the brain, including tyrosine and tryptophan, important for neurotransmission [ 1 ]. If left untreated, PKU is associated with poor neurologic, neurocognitive, and neuropsychiatric outcomes [ 1 , 2 ].

Recognized as exhibiting a spectrum of severity, the most severe form of PKU is often referred to as classical PKU (cPKU) and is defined as little or no PAH activity and untreated blood Phe levels typically > 1200 µmol/L at the time of diagnosis (normal blood Phe level is 50–110 µmol/L). An individual’s specific genetic variation determines the degree of PAH activity; variants only partially inhibiting PAH activity result in a milder form of PKU or mild hyperphenylalaninemia (HPA) [ 2 ].

The goal of treating PKU is to achieve and maintain appropriate blood Phe levels recommended by the United States and European guidelines [ 3 , 4 ]. To control blood Phe levels, individuals with PKU are placed on a lifelong prescribed medical intervention termed medical nutrition therapy (MNT) [ 4 ], which involves severely restricting the natural intake of protein and replacing it with a Phe-free, amino acid-based medical food to supplement the reduced protein intake, and provide a source of energy and other nutrients. Supplements might include modified low-protein foods and Phe-free medical food beverages, Phe-free amino acid mixture, medical foods derived from glycomacropeptide, and protein substitutes [ 4 ].

Importantly, studies have shown that not all patients, including adolescents and adults, are able to achieve blood Phe levels within guideline-recommended target ranges [ 5 , 6 ]. Even with active management, blood Phe levels may remain uncontrolled, especially as patients age. An online survey conducted in the United States estimated that 67% of adults with PKU had blood Phe levels in excess of the upper limit of the American College of Medical Genetics target levels [ 7 ]. Patients with cPKU have the most difficulty controlling blood Phe levels with MNT, and control is considered suboptimal when compared with patients with mild PKU [ 6 ].

Pharmaceutical intervention with sapropterin dihydrochloride (KUVAN®; BioMarin Pharmaceutical Inc., Novato, CA, USA), a derivative of the PAH cofactor, tetrahydrobiopterin (BH4), may be used in conjunction with a Phe-restricted diet, for individuals who are deemed responsive. For adults (or patients ≥ 16 years old outside of the United States) with uncontrolled blood Phe levels despite existing management, pegvaliase (PALYNZIQ®; BioMarin Pharmaceutical Inc., Novato, CA, USA), a PAH substitute [ 8 , 9 ], may be an option to achieve appropriate blood Phe levels, without requiring patients to maintain a Phe-restricted diet [ 10 ].

Lifelong treatment of PKU is recommended by guidelines [ 3 , 4 ]. Early intervention prevents the severe and irreversible intellectual impairment caused by elevated blood Phe levels in childhood and adolescence [ 11 ], but adherence to dietary restrictions is challenging, and the number of patients achieving target blood Phe levels tends to diminish with age [ 2 , 7 ]. Uncontrolled Phe levels are also associated with adverse neurocognitive and neuropsychiatric outcomes in adults [ 12 , 13 , 14 ]. Meta-analysis of cognitive function in adults with PKU has shown impairment in reasoning, visuo-spatial attention speed, sustained attention, and visuo-motor control, despite early initiation of treatment [ 14 ], and meta-analysis of neuropsychiatric symptoms in adults with PKU has shown that inattention, hyperactivity, depression, and anxiety exceed general population estimates [ 13 ]. Neurocognitive and neuropsychiatric symptoms associated with Phe accumulation may make it more difficult for patients to adhere to dietary restriction, which in turn can lead to poor blood Phe control and worsening of symptoms [ 4 ].

An emerging body of literature suggests that the impact of PKU on an individual’s health may extend beyond symptoms of a neurocognitive and neuropsychiatric nature. Comorbidities across various organ systems have been reported in adults with PKU, with health insurance claims-based studies suggesting a higher prevalence of somatic comorbidities compared with a general population [ 15 , 16 ].

High blood Phe levels may impact biologic mechanisms that are related to increased risk of comorbid conditions such as obesity, renal disease, metabolic dysfunction, and cardiovascular complications [ 16 ]. Due to the impact on different organ systems, the etiology is complex and multifactorial [ 17 ]. Retrospective analysis of insurance claims data has enabled researchers to generate hypotheses for development of certain comorbidities based on their knowledge of PKU and the associated dietary management [ 15 , 16 ]. Better understanding of the etiology of somatic comorbidities associated with PKU and identification of factors other than high blood Phe that may be preventable or amenable to treatment, together with effective metabolic control, could aid in reducing the burden of illness and healthcare costs. However, the first step is to investigate differences in somatic comorbidity burden, not only between adults with PKU and the general population but also among adults with PKU receiving different therapeutic interventions, those adherent to treatment or not, and with different disease severity.

A systematic literature review (SLR) has been conducted to evaluate the published evidence on the somatic comorbidities experienced by adults with PKU. The SLR aims to further characterize the physical health burden of PKU and provide insight into the impact of differences in therapeutic interventions, adherence to treatment, and differences in disease severity on the somatic comorbidity burden.

Materials and methods

The SLR is registered with the Research Registry with the unique identifying number review registry 1476 and is reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and Synthesis Without Meta-analysis (SWiM) guidelines [ 18 , 19 ].

Eligibility criteria

Eligibility criteria for the inclusion of studies to address the research question were established using the Population, Intervention, Comparator, Outcome, Study design (PICOS) framework (Table  1 ).

Peer-reviewed observational studies (cohort, case–control, cross-sectional, surveys) and clinical trials in individuals ≥ 16 years of age (or defined as adults by the study) either confirmed or described as having PKU were included. The cut-off for adult age was chosen to be ≥ 16 years because adolescent age differs globally. Eligible studies were those evaluating a Phe-restricted diet with or without pharmacologic therapy (sapropterin dihydrochloride or pegvaliase) versus no form of therapeutic intervention (including healthy controls or reference values); or pharmacologic therapy versus a Phe-restricted diet alone. Studies comparing different Phe-restricted diets (e.g., different modified low-protein foods and Phe-free medical food beverages, Phe-free amino acid mixture, medical foods derived from glycomacropeptide) and those comparing populations of individuals with more severe disease and those with less severe disease, were also considered eligible in a protocol amendment that was made during the full-text review stage. Outcomes of interest were defined in the study protocol as the prevalence and/or severity of somatic comorbidities in individuals with PKU, but in practice, any measure of somatic comorbidity experienced by individuals with PKU in eligible studies was considered for inclusion.

Studies carried out exclusively in a population of individuals identified as children or adolescents were excluded; however, otherwise eligible studies with mixed-age populations were included whether or not results were presented separately for adults. Non-human studies and in vitro studies, single-cohort studies in individuals with PKU who were untreated or in individuals with PKU on Phe-restricted diet that were not compared with a healthy control population or reference values, or in individuals with PKU treated with pharmacologic therapy who were not compared with those on Phe-restricted diet alone, were excluded. Secondary literature sources, including narrative review articles, letters, editorials, and commentaries were also excluded, as were therapy recommendations, clinical guidelines, congress abstracts, and non-peer-reviewed literature.

Information sources and search strategy

Literature was retrieved via the PubMed® interface. No date restrictions were applied to the search, thus publications in English from MEDLINE earliest coverage to November 1, 2023 were included (search conducted on February 1, 2022 and updated on November 1, 2023) [ 20 ]. A pre-defined systematic search strategy (Additional file 1 : Table S1) was designed to identify relevant records; the search string included a combination of free text and Medical Subject Headings (MeSH) search terms based on the inclusion and exclusion criteria documented in Table  1 . To maximize the identification of relevant articles, the search string included general terms for comorbidity and burden, as well as specific somatic comorbidity types guided by the results of a previous (unpublished) literature review that was available to authors designing the study.

Backwards citation searching was employed to identify additional papers of interest in the reference lists of relevant systematic reviews and meta-analyses that were retrieved as part of the systematic literature search and search update. Duplicate records were removed during screening.

Selection process

Study screening was carried out by a small team of reviewers to identify records eligible for data extraction according to the PICOS framework. A small number of records ( n  = 10) were screened independently by the reviewing team in a pilot phase, and results compared and discussed to assess concordance of eligibility decisions and ensure relevance and utility of the inclusion and exclusion criteria used to screen records.

A two-stage screening process was then applied to identify records eligible for data extraction according to the inclusion and exclusion criteria. Records were screened once by title and abstract and selected for full-text review if they met all inclusion criteria or if it was unclear whether all inclusion criteria were met; records were only rejected if it was clear that at least one of the inclusion criteria was not met (termed positive exclusion methodology). Records considered potentially eligible were screened once by full text to confirm eligibility for data extraction. Concordance of eligibility decisions was assessed at both screening stages, whereby 10% of records underwent a second independent screen and any discrepancies in first and second reviewer opinion were discussed with a third reviewer to achieve a consensus decision.

Screening was carried out within the DistillerSR Inc. (Ottawa, ON, Canada) workflow management software and used to view records for review, indicate conflicts between and record reviewers’ decisions, including reasons for exclusion. The continuous artificial intelligence reprioritization feature was utilized to continuously re-order the screening of records based on previous screening decisions.

Data collection and data outcomes

Data extraction from eligible studies was conducted by one reviewer into a pre-designed data extraction spreadsheet (Microsoft Excel®; Microsoft Corporation, Redmond, WA, USA). All data that related to outcomes of interest, including any measure or description of any somatic comorbidities experienced by individuals with PKU, were extracted, as well as other data items including study design, geographic coverage, year of publication, main study conclusions and limitations. Statistical comparisons between groups were also recorded, when available. Studies reporting insufficient data to satisfy inclusion criteria (e.g., data were not reported separately for group of interest) were rejected. Extracted data were checked for accuracy by an independent reviewer.

Grouping studies for synthesis

Studies were grouped by PKU population to allow synthesis of data according to the populations identified in the research question (i.e., those on a Phe-restricted diet with or without pharmacologic therapy [sapropterin dihydrochloride or pegvaliase] versus healthy controls or reference values; those on a Phe-restricted diet who were adherent versus non-adherent; those on different Phe-restricted diets; those on a Phe-restricted diet with more severe PKU versus those with HPA or less severe PKU).

Studies were then grouped according to comorbidity type (abdominal or pelvic pain, bone-related abnormalities, cancer, cardiovascular outcomes, chronic obstructive pulmonary disease [COPD]/asthma, dermatologic disorders, diabetes, gastrointestinal disorders, hypertension, migraine/headache, musculoskeletal outcomes, nutritional outcomes, overweight/obesity, or other) to allow synthesis of data by specific comorbidity type.

Data synthesis

The breadth of measures and numbers of studies in each population grouping that reported outcome measures for the same comorbidity type were analyzed to identify appropriate methods for data synthesis (e.g., meta-analysis or synthesis without meta-analysis) to determine intervention effects.

Bone mineral density (BMD) Z-scores, where a low BMD Z-score is considered an indicator of bone-related abnormality, was the only outcome measure considered appropriate for meta-analysis of effect estimates due to sufficiency and homogeneity in clinical outcomes, methodological approach, and statistics reported, and this analysis is reported separately. Meta-analysis of effect estimates was not considered appropriate for the other somatic, comorbidity types due to extensive heterogeneity in clinical outcomes used, the definitions of clinical outcomes used, how the clinical outcome was measured, and study design including interventions and comparators. Vote counting was considered an acceptable alternative method given it allows the direction of effect to be determined using all available evidence, for example, even when there is no consistent effect measure or data reported across studies.

Vote counting was conducted according to the methods described in the Cochrane handbook and reported according to the SWiM guidelines [ 19 , 21 ]. A standardized binary metric was created by allocating votes to individual studies according to the direction of a higher comorbidity burden (i.e., either a higher burden in the direction of the ‘intervention’ [for example, individuals with PKU on a Phe-restricted diet with or without pharmacologic therapy] or a higher burden in the direction of the ‘comparator’ [for example, healthy controls or reference values]), regardless of the statistical significance clinical relevance of differences between the groups. The number of votes allocated to the intervention population was then compared with the number allocated to the comparator population, to determine the direction of effect, and was visualized using an effect direction plot, in line with guidance from the Cochrane handbook [ 21 ]. Studies were prioritized for data synthesis based on directness in relation to the research question and availability of data. No assessment of certainty of the evidence was undertaken given it is difficult to assess consistency of effects when vote counting is undertaken.

Study selection and characteristics

The PRISMA flow diagram (Fig. 1 ) shows the results of the study selection process. The PubMed® search identified 1185 unique records. Subsequently, 473 records were considered potentially eligible for inclusion and 53 studies were confirmed as eligible for inclusion. Five included studies from the PubMed® search were SLRs and used for backwards citation searching only, revealing five additional papers of interest, of which two were confirmed as eligible for inclusion. One of these studies was an SLR and backwards citation-searching revealed two publications of interest, of which only one met inclusion criteria. Overall, a total of 57 studies met the PICOS criteria. Data were extracted from 51 studies spanning 12,602 individuals (excluding the six SLRs) and included in the synthesis. Reasons for exclusion of studies at each stage of the selection process are listed in Fig. 1 .

PRISMA diagram showing article selection process. Articles were excluded on a hierarchical basis, in the order that questions were asked (i.e., if the answer to the first question was no, this was given as the main reason for exclusion, but articles may have met or not met other criteria). Abbreviation: SLR, systematic literature review. a Five systematic reviews were identified via the database search and used for backwards citation-searching only plus one additional systematic review identified via backwards citation-searching that was then used for further backwards citation-searching; b Includes studies that did not present outcomes in a meaningful way which answered one or more of the pre-specified research questions; c Other studies include open interventional trials, pooled analyses, and cost analyses

Of the 51 studies included in the synthesis, most were of observational design, including cross-sectional studies ( n  = 31, 61%), retrospective cohort studies ( n  = 9, 18%) and case-controlled studies ( n  = 5, 10%) (Additional file 2 : Figure S1). Most studies were conducted in European countries ( n  = 40) and/or in North America ( n  = 6) (Additional file 3 : Figure S2).

More than 13 different comorbidity types were reported across the 51 studies; bone-related abnormalities were the most reported ( n  = 23), followed by overweight/obesity ( n  = 18), nutritional outcomes ( n  = 16), and cardiovascular outcomes ( n  = 9) (Fig. 2 ). Migraine/headache and cancer were reported in a consistent manner across the studies, whereas outcome measures for bone-related abnormalities, cardiovascular outcomes, and dermatologic disorders were reported with a high degree of inconsistency resulting in heterogeneity. One study used the Charlson Comorbidity Index (CCI) to report multiple comorbidities [ 22 ].

Distribution of studies by comorbidity type. Abbreviations: COPD, chronic obstructive pulmonary disease. a Other comorbidities include: acute upper respiratory infections of multiple and unspecified sites; allergic and chronic rhinitis; anemia; adverse events, not elsewhere classified; calculus of kidneys; Charlson Comorbidity Index score; chronic kidney disease; congenital deformities of feet; dizziness and giddiness; dorsalgia; esophageal disorders; gallbladder disease; grip force; gynecological symptoms; hypothyroidism; menopausal and other perimenopausal disorders; metabolic syndrome; ophthalmological symptoms; other disorders of the urinary system; other hypothyroidism; other non-inflammatory disorders of the vagina; other non-toxic goiter; other and unspecified dorsopathies; other and unspecified soft tissue disorders; otolaryngological symptoms; refraction and accommodation disorders; renal insufficiency with hypertension; renal insufficiency without hypertension; thyroid function; upper respiratory traction infection; varicose veins of lower extremities; vasomotor and allergic rhinitis

Vote counting was used to determine the direction of comorbidity burden in two PKU population comparisons: those on a Phe-restricted diet with or without pharmacologic therapy versus healthy controls or reference values, and those on a Phe-restricted diet who were treatment-adherent versus a non-adherent population. For all other PKU population comparisons, data synthesis by vote counting was not feasible due to a low number of studies.

Individuals with PKU on a Phe-restricted diet with or without pharmacologic therapy versus healthy controls or reference values

Of the 40 studies comparing individuals with PKU on a Phe-restricted diet, with or without pharmacologic therapy, with healthy controls or reference values (Table  2 ), two studies were excluded from the vote counting because it was not possible to confirm treatment with a Phe-restricted diet in the full study population [ 16 , 23 ].

A higher burden of ≥ 1 comorbidity (or outcome measure) in individuals with PKU on a Phe-restricted diet with or without pharmacologic therapy was indicated by 37 of 38 studies included in the vote-counting analysis, versus a higher burden of ≥ 1 comorbidity (or outcome measure) in healthy controls or reference values in 12 studies (Fig. 3 ). The most commonly reported somatic comorbidities with a higher burden in those on a Phe-restricted diet with or without pharmacologic therapy were bone-related abnormalities ( n  = 21), nutritional outcomes ( n  = 9), overweight/obesity ( n  = 8), and cardiovascular outcomes ( n  = 5). The most commonly reported somatic comorbidities with a higher burden in healthy controls or reference values were overweight/obesity ( n  = 7), bone-related abnormalities ( n  = 3), and nutritional outcomes ( n  = 3).

Burden of somatic comorbidities in individuals with PKU on a Phe-restricted diet versus healthy controls or reference values as assessed by vote counting. Abbreviations: COPD, chronic obstructive pulmonary disease; Phe, phenylalanine; PKU, phenylketonuria. Note: Total number of studies = 38. a Studies with a higher burden of ≥ 1 comorbidity or outcome measure, for a given comorbidity category, in individuals with PKU who adhered to a Phe-restricted diet. b Studies with a higher burden of ≥ 1 comorbidity or outcome measure, for a given comorbidity category, in healthy control individuals or a normal reference population. Some studies reported more than one comorbidity or outcome measure per category. Studies reporting a differing direction of effect between comorbidities or outcome measures within a category, are indicated below. Details of studies with consistent direction of effect are not listed below (but are included in Table  2 ). Vote counting was conducted on the basis of numerical differences in the direction of effect, regardless of statistical significance or clinical relevance. Abdominal and pelvic pain: Higher burden in PKU ( n  = 2) [ 15 , 22 ]. Bone-related abnormalities: Higher burden in PKU ( n  = 21) [ 15 , 22 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 ]; negative BMD for distal radius, total body, and trabecular bone; proximal radius, total body, and worse measures of bone geometry and strength in PKU group [ 24 ], lumbar and femoral BMD Z-score < –2 in 5.0% and 7.0% of all patients, negative median BMD in adults for hip bone, higher percentage of all patients with fracture history in PKU group [ 26 ], lower vitamin D status, higher concentrations of all bone resorption markers, lower concentrations of all bone formation markers except alkaline phosphatase, and higher calcium and phosphorus excretion in PKU group [ 37 ]; higher burden in controls ( n  = 3) [ 24 , 26 , 37 ], positive BMD for proximal radius cortical bone in PKU group [ 24 ], positive median BMD in adults for femur in PKU group [ 26 ], higher concentration of alkaline phosphatase in PKU group [ 24 , 37 ]. Cardiovascular outcomes: Higher burden in PKU ( n  = 5) [ 15 , 22 , 43 , 46 , 47 ], higher arterial stiffness in PKU group [ 47 ]; higher burden in controls ( n  = 1), higher intima media thickness in control group [ 47 ]. COPD/asthma: Higher burden in PKU ( n  = 2) [ 15 , 22 ]. Dermatologic disorders: Higher burden in PKU ( n  = 1) [ 15 ]. Diabetes: Higher burden in PKU ( n  = 3) [ 15 , 22 , 44 ]. Gastrointestinal disorders: Higher burden in PKU ( n  = 2) [ 15 , 22 ], numerically higher frequency of diverticular disease of intestine in individuals with PKU versus non-PKU control [ 22 ]; higher burden in controls ( n  = 1) [ 22 ], numerically higher frequency of gastritis and duodenitis in non-PKU controls versus individuals with PKU and numerically higher frequency of constipation in non-PKU controls compared with the early-diagnosed PKU subgroup only (null effect on constipation between the overall PKU group and non-PKU controls) [ 22 ]. Hypertension: Higher burden in PKU (n = 2) [ 15 , 22 ]. Musculoskeletal outcomes: Higher burden in PKU ( n  = 3) [ 22 , 24 , 33 ], muscle size and performance were preserved in individuals with PKU and regression lines were comparable to the reference population (null effect, excluded from vote counting [ 24 ]). Nutritional outcomes: Higher burden in PKU ( n  = 9) [ 22 , 26 , 45 , 48 , 49 , 54 , 56 , 58 , 59 ], decreased concentration of vitamin B12 in relaxed diet and unrestricted diet groups versus control [ 54 ], concentration of vitamin D, selenium and zinc below reference range [ 26 ], individuals with PKU were less likely to achieve adequate choline intake compared with controls [ 56 ]; higher burden in controls ( n  = 3) [ 26 , 33 , 54 ], increased concentration of vitamin B12 in strict diet group and increased concentration of folate in all diet groups versus control (within or above normal range) [ 54 ], concentration of magnesium, folate, vitamin B12 and B6 above reference range [ 26 , 54 ], analysis only considered the PKU population who consumed adequate protein substitute without Phe and maintained strict metabolic follow-up [ 33 ]; one study investigating mean probability of adequacy for vitamin B6, B12, and folate reported a null effect for individuals with PKU on Phe-restricted diet with medical food and dietary supplements versus healthy controls (excluded from vote counting) [ 56 ]. Overweight/obesity: Higher burden in PKU ( n  = 8) [ 15 , 22 , 43 , 44 , 50 , 51 , 52 , 53 ], percentage of females with BMI > 30 kg/m2 was higher than in all UK countries assessed, percentage of females with BMI > 25 kg/m2 was higher than in Northern Ireland only [ 53 ], percentages of individuals with PKU who were obese was higher than in the general population in 2/6 centers [ 52 ], fat-free mass (Kg) was numerically lower in individuals with PKU versus healthy control [ 51 ], no controls below normal range for BMI as opposed to PKU group; higher burden in controls ( n  = 7) [ 33 , 50 , 51 , 52 , 53 , 55 , 56 ], adults > 16 years subgroup had higher prevalence of overweight/obesity in control versus PKU [ 55 ] but percentage body fat was equal (excluded from vote counting) [ 55 ], percentage of males with BMI > 25 and > 30 kg/m2 was higher than in all UK countries assessed, percentage of females with BMI > 30 kg/m2 was higher than in England and Scotland only [ 53 ], percentages of individuals with PKU who were overweight and obese were lower than those in the general population in 5/6 and 3/6 study centers, respectively (percentage of individuals with PKU who were obese was the same as that for the general population in 1/6 centers (excluded from vote counting) [ 52 ], bodyweight and BMI was numerically lower in individuals with PKU versus healthy control but both groups were only borderline overweight, percentage fat-free mass was numerically higher in individuals with PKU versus healthy control [ 51 ], bodyweight, percentage fat mass, and BMI were numerically less in individuals with PKU than healthy controls and BMI of more controls was above normal range, percentage fat-free mass was higher in individuals with PKU than in healthy controls [ 50 ]. Other: Higher burden in PKU ( n  = 3) [ 15 , 22 , 57 ]; higher burden in controls ( n  = 1) [ 55 ]. In the majority of studies, all individuals were on a Phe-restricted diet, with the following exceptions: one study ( n  = 83) of which 31 were on an unrestricted diet – no formal protein restriction and not taking amino acid supplements, 30 were on a relaxed diet – total protein intake of approximately 1 g/kg/d (50% from natural protein/ 50% from amino acid, vitamin and mineral supplements), and 22 were on a strict low-Phe diet, including amino acid, vitamin, and mineral supplements [ 54 ]; one study with a mixture of individuals on and not on a Phe-restricted diet [ 25 ]; one study in which some individuals were on sapropterin dihydrochloride or pegvaliase in addition to a Phe-restricted diet [ 25 ]; one study in which some individuals received sapropterin dihydrochloride, some were on a Phe-restricted diet, and, for some, it was not clear whether they were on a Phe-restricted diet or not [ 15 ]; one study in which some individuals were treated with sapropterin dihydrochloride in addition to dietary treatment [ 46 ]; one study ( n  = 164) in which the majority of individuals were on a Phe-restricted diet, up to 20 adults received additional BH4, and up to 11 adults received BH4 alone [ 52 ]; one study ( n  = 1911) in which 29% of individuals received amino acid supplementation and 5% received sapropterin dihydrochloride (unclear if remaining individuals were on Phe-restricted diet) [ 22 ]; one study in which 80% of individuals were on Phe-restricted diet and 20% were on sapropterin dihydrochloride with or without amino acid supplementation [ 32 ]; one study in which individuals on BH4 were excluded and it was unclear whether individuals were on a Phe-restricted diet or not [ 51 ]

Bone-related abnormalities encompass a range of features; abnormalities reported in individuals with PKU included reduced BMD, measured by Z-scores [ 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 ] or g/cm 2 [ 33 ]; presence of osteopenia/osteoporosis [ 16 , 22 , 27 , 29 , 30 ]; lower distal radius [ 24 ]; reduced cortical thickness and strength-strain index [ 24 ]; greater risk of fracture [ 26 , 34 , 35 ]; reduced levels of bone formation markers and/or increased levels of bone resorption markers [ 30 , 36 , 37 , 38 ]; higher prevalence of osteoarthritis of the knee [ 15 ]; higher prevalence of spondylosis [ 15 ]; and increased presence of osteoclastogenesis [ 39 , 40 ]. Bone-related abnormalities reported in healthy controls included reduced BMD [ 24 , 26 ]; reduced cortical density [ 24 ]; and high levels of bone formation markers [ 37 ]. The most reported outcome measures for bone-related abnormalities (reported in ≥ 4 studies) were Z-scores ( n  = 10), markers for bone resorption and bone formation ( n  = 4), and prevalence of osteopenia/osteoporosis ( n  = 6) (Fig. 4 ). There was no numerical difference between groups in femoral BMD in one study [ 33 ]; therefore, this was not captured in the vote counting. Statistical significance of the difference in bone-related abnormalities between those on a Phe-restricted diet with or without pharmacologic therapy versus healthy controls or reference values was assessed in 17 [ 15 , 16 , 22 , 24 , 27 , 28 , 29 , 30 , 33 , 34 , 35 , 36 , 37 , 39 , 40 , 41 , 42 ] of the 22 studies (including one study [ 16 ] not included in the vote counting), of which the majority showed a higher burden of ≥ 1 outcome measure in the PKU group; results are reported in Fig. 4 and Table  2 .

Overview of measures used to report bone-related abnormalities in individuals with PKU on a Phe-restricted diet versus healthy controls or reference values. Abbreviations: BMD, bone mineral density; OC, osteoclastogenesis; Phe, phenylalanine; PKU, phenylketonuria; PR, prevalence ratio. Studies in bold font showed a statistically significant difference between groups. All 21 studies indicated a higher clinical burden of ≥ 1 outcome measure in the PKU group (or a particular subgroup) compared with healthy controls; with 15 studies reporting a statistically significant difference [ 16 , 22 , 24 , 27 , 28 , 29 , 30 , 34 , 35 , 36 , 37 , 39 , 40 , 41 , 42 ], two studies that did not find a statistically significant difference for any outcome measure [ 15 , 33 ], and four studies that did not test for statistical significance between PKU group and controls [ 25 , 26 , 31 , 32 ]; in seven studies the difference between groups was not statistically significant for all outcome measures [ 29 ], outcomes [ 24 , 27 , 28 , 30 , 37 ], or in the comparison of the overall PKU population [ 35 ]. a Units: osteocalcin (μg/L), bone alkaline phosphatase (BAP; μg/L), deoxypyridinoline (μmol/mol creatine), calcium/creatine index (no units reported); b Units: osteocalcin (ng/mL), BAP (U/I), intact parathyroid hormone (pg/mL), 1,25 (OH) 2 vitamin D (pg/mL), 25 (OH) vitamin D (ng/mL), urinary deoxypyridinoline (nmol/mmol creatinine), urinary N-telopeptides of type collagen (nmol/mmol creatinine), ICTP (pyridinoline cross-linked telopeptide domain of type I collagen; ng/mL), osteoprotegerin (pmol/L), urinary calcium/creatine index (mmol/mmol creatinine), urinary phosphorus/creatine index (mmol/mmol creatinine); c Units: osteocalcin (no reported units), BAP (μg/L); d Units: BAP (μg/L); e Lifetime fracture prevalence was measured as percentage of the population; f Risk of fracture was measured between 0 and 20 years of age using a Kaplan–Meier graph (cumulative proportion with fracture vs age)

Of ten studies reporting nutritional outcomes included in the vote-counting analysis shown in Fig. 3 , nine studies reported a higher burden of ≥ 1 outcome measure in those on a Phe-restricted diet compared with healthy controls or reference values [ 22 , 26 , 45 , 48 , 49 , 54 , 56 , 58 , 59 ] whereas three studies reported a higher burden of ≥ 1 outcome measure in healthy controls or reference values [ 26 , 33 , 54 ]; results are reported in Table  2 . This included significantly lower free carnitine concentrations [ 59 ]; significantly higher percentages of individuals with vitamin D deficiency and iron deficiency anemia [ 22 ] abnormal concentrations of vitamin B12, methylmalonic acid, and homocysteine [ 58 ]; higher concentrations of folate [ 33 , 48 , 54 ], cobalamin, and homocysteine [ 48 ]; concentrations of magnesium, folate, vitamin B12, and vitamin B6 above or within the reference range [ 26 ]; higher concentrations of vitamin B12 and vitamin D3 and lower concentrations of homocysteine that were within reference range (but the differences between groups were not statistically significant) [ 33 ]; lower concentrations of vitamin B12 [ 45 , 49 , 54 ], vitamin B6 [ 49 ], selenium, pre-albumin, folate, vitamin D, ferritin and zinc (although the difference versus the normal reference was not statistically significant [ 45 ]); concentrations of vitamin D and selenium within or below the reference range [ 26 ]; and a lower likelihood of achieving adequate choline intake but a very low probability of achieving inadequate intake of vitamin B6, vitamin B12, folate, and methionine in both groups [ 56 ]. One study [ 45 ] observed significant correlations between changes in nutritional outcomes and participant age (≤ 18 years versus > 18 years): total protein and pre-albumin levels increased with age ( p  = 0.002 and p  < 0.0001, respectively), whereas calcium and phosphorus decreased with age ( p  = 0.015 and p  < 0.0001, respectively). In the same study, vitamin B12 levels were significantly lower in BH4-treated versus BH4-untreated participants [ 45 ].

Twelve studies reported outcome measures relating to overweight/obesity [ 15 , 16 , 22 , 33 , 43 , 44 , 50 , 51 , 52 , 53 , 55 ]; results are reported in Table  2 and Fig. 3 (one study was excluded from the vote-counting analysis [ 16 ]). Four studies [ 15 , 16 , 22 , 43 ] reported a significantly higher body mass index (BMI), prevalence ratio, or percentage of individuals who were overweight and/or obese in those with PKU versus healthy controls. The statistical significance of the difference between groups was maintained for a subgroup of early diagnosed individuals in one of these studies [ 22 ]; one study [ 44 ] reported numerically higher proportions of overweight/obesity among the PKU population versus healthy controls (39% versus 25%) but the statistical significance of the difference was not reported; in one study [ 52 ] the rate of obesity in females with PKU was higher than in the respective general (non-PKU) population in four of six centers, but the overall rate of overweight participants was lower in five of six centers studied; and being overweight was more common and obesity was less common in individuals with PKU compared with the reference dataset in one study [ 56 ]. However, two studies [ 53 , 55 ] reported no significant difference in the proportions of the PKU population who were overweight/obese compared with the control population and two studies reported no significant difference in bodyweight, BMI, total fat mass, total fat-free mass [ 33 , 50 ], appendicular fat-free mass, appendicular fat-free mass index, and waist circumference (WC) [ 33 ] between the PKU population on a Phe-restricted diet and heathy controls. Three studies did not assess statistical significance of the differences between populations: in the first study [ 52 ], the proportion of obese females with PKU was higher than in the respective general (non-PKU) population in four of six centers studied, but numerically lower proportions of obese individuals were reported overall in three of six centers, higher proportions were reported in two of six centers, and the same proportion was reported in one of six centers in the PKU group versus the control group, while a lower proportion of overweight/obese individuals in the PKU group versus the control group was observed in five of six centers; the second study [ 51 ] reported numerically lower WC and BMI and a numerically higher percentage of fat-free mass in individuals with PKU versus controls, but numerically lower absolute fat-fee mass (kg) in those with PKU versus controls; and the third study [ 56 ] reported numerically higher percentages of overweight individuals and numerically lower percentages of obese individuals in the PKU group versus controls.

Five studies reported cardiovascular outcome measures [ 15 , 22 , 43 , 46 , 47 ], including an increase in arterial stiffness markers ( n  = 2) [ 46 , 47 ], increased prevalence of ischemic heart disease ( n  = 2) [ 15 , 22 ], higher heart rate and blood pressure ( n  = 1) [ 43 ], and increase in carotid intima media thickness ( n  = 1) [ 47 ]. Statistical significance of the difference in cardiovascular outcomes between those on a Phe-restricted diet with or without pharmacologic therapy versus healthy controls or reference values was assessed in four [ 15 , 22 , 43 , 46 ] out of five [ 15 , 22 , 43 , 46 , 47 ] studies and included significantly higher resting heart rate and systolic blood pressure [ 43 ], markers of arterial stiffness [ 46 ], and prevalence of chronic ischemic heart disease [ 15 , 22 ] in individuals with PKU versus healthy controls; results are reported in Table  2 and Fig. 3 .

Individuals with PKU on pharmacologic therapy with or without a Phe-restricted diet versus those on Phe-restricted diet alone

One study [ 56 ] investigated choline nutriture in adults and children with PKU receiving pegvaliase ( n  = 33 adults), sapropterin dihydrochloride ( n  = 21 adults), or dietary therapy alone ( n  = 17 adults). This study found that adults receiving pegvaliase were most likely to exceed adequate intake of choline (14.82%, standard error [SE] 4.48), while adults on dietary therapy alone were least likely (5.59%, SE 2.95). In general, however, there was a very low probability of inadequate intake of nutrients affecting choline metabolism (vitamin B6, vitamin B12, folate, and methionine) among adults with PKU. In this study [ 56 ], the pegvaliase group had the highest percentage of overweight/obesity compared with those on dietary therapy alone and those on sapropterin dihydrochloride (81.8%, 64.7%, and 61.9%, respectively), driven by a higher rate of obesity than in individuals on either sapropterin dihydrochloride or dietary therapy alone (48.5%, 38.1%, and 23.5%, respectively); however, the percentage of individuals who were overweight was highest in the group on dietary therapy alone (41.2% versus 33.3% and 23.8% for those on pegvaliase and sapropterin dihydrochloride, respectively).

Individuals with PKU adherent to a Phe-restricted diet versus a non-adherent population

Twelve studies comparing individuals with PKU who adhered to a Phe-restricted diet with a non-adherent population were included in the SLR; however, one study [ 50 ] was excluded from the vote-counting analysis because numerical data were not reported (Table  3 ). There were seven [ 29 , 32 , 33 , 45 , 60 , 61 , 62 ] studies that reported a higher burden of ≥ 1 comorbidity (or outcome measure) in those who adhered to a Phe-restricted diet compared with those non-adherent (Fig. 5 ). There were ten studies [ 5 , 32 , 33 , 45 , 47 , 54 , 60 , 61 , 62 , 63 ] that reported a higher burden of ≥ 1 comorbidity (or outcome measures) in those who did not adhere to a Phe-restricted diet compared with those who adhered to diet.

Burden of somatic comorbidities in individuals with PKU adherent to versus those not adherent to a Phe-restricted diet as assessed by vote counting. Abbreviations: COPD, chronic obstructive pulmonary disease; Phe, phenylalanine; PKU, phenylketonuria. Note: Total number of studies = 11. a Studies with a higher burden of ≥ 1 comorbidity or outcome measure, for a given comorbidity category, in individuals with PKU who adhered to a Phe-restricted diet. b Studies with a higher burden of ≥ 1 comorbidity or outcome measure, for a given comorbidity category, in individuals with PKU who did not adhere to a Phe-restricted diet. Studies reporting more than one comorbidity or outcome measure per category, or those with a differing direction of effect between comorbidities or outcome measures within a category, are indicated below. Vote counting was conducted on the basis of numerical differences in the direction of effect, regardless of statistical significance or clinical relevance. Bone-related abnormalities: higher burden in adherent ( n  = 2) [ 29 , 33 ], lower lumbar, femoral neck, and total body BMD Z-scores [ 29 ], lower spine BMD and null effect for femoral BMD (excluded from vote counting) [ 33 ]; higher burden in non-adherent ( n  = 1) [ 63 ]. Cancer: higher burden in adherent ( n  = 1), higher incidence in discontinued and restarted (group 3) compared with never treated (group 4) and off-diet (group 2) [ 60 ]; higher burden in non-adherent ( n  = 1), higher incidence in off diet (group 2) compared with adherent since birth (group 1) [ 60 ]; no reports of cancer in either group (excluded from vote counting) [ 61 ]. Cardiovascular outcomes: higher burden in non-adherent ( n  = 2) [ 60 , 61 ], heart disease in larger proportion of participants [ 61 ], higher incidence of cardiovascular symptoms in off-diet (group 2) compared with discontinued and restarted (group 3) and adhered since birth (group 1), and higher incidence in never treated (group 4) compared with all other groups [ 60 ]. COPD/asthma: higher burden in adherent (n = 1), higher incidence of asthma in discontinued and restarted (group 3) than in off-diet (group 2) and never treated (group 4) [ 60 ]; higher burden in non-adherent ( n  = 2) [ 60 , 61 ], higher incidence of asthma in never treated (group 4) than in adhered since birth (group 1) [ 60 ], asthma reported in larger proportion of participants [ 61 ]. Dermatologic disorders: higher burden in adherent ( n  = 1), higher incidence of dermatologic symptoms in discontinued and restarted (group 3) than in off-diet (group 2) and never treated (group 4) [ 60 ]; higher burden in non-adherent ( n  = 2) [ 60 , 61 ], eczema reported in larger proportion of participants [ 61 ], higher incidence of dermatologic disorders in off diet (group 2) compared with adherent since birth (group 1), higher incidence of dermatologic disorders in never treated (group 4) compared with adherent since birth (group 1) and off-diet (group 2) [ 60 ]. Gastrointestinal disorders: higher burden in adherent ( n  = 1), higher incidence in discontinued and restarted (group 3) compared with off diet (group 2) and never treated (group 4) [ 60 ]; higher burden in non-adherent ( n  = 1), higher incidence in discontinued and restarted (group 3) compared with adherent since birth (group 1) [ 60 ]. Hypertension: higher burden in adherent ( n  = 1), hypertension reported in larger proportion of participants [ 61 ]. Migraine/headache: higher burden in adherent ( n  = 1), higher incidence of headaches in adherent since birth (group 1) compared with off-diet (group 2) and higher incidence of headaches in discontinued and restarted (group 3) compared with off-diet (group 2) and never treated (group 4) [ 60 ]; higher burden in non-adherent ( n  = 2) [ 60 , 61 ], higher incidence of headaches in discontinued and restarted (group 3) compared with adherent since birth (group 1) [ 60 ], headaches reported in larger proportion of participants [ 61 ]. Musculoskeletal outcomes: higher burden in adherent ( n  = 1), decreased left and right hand-grip strength in adherent versus non-adherent [ 33 ]; higher burden in non-adherent ( n  = 2) [ 60 , 63 ], higher incidence of arthritis/musculoskeletal symptoms in discontinued and restarted (group 3) compared with adherent since birth (group 1) [ 60 ]. Nutritional outcomes: higher burden in adherent ( n  = 3) [ 32 , 45 , 62 ], decreased concentrations of total protein and pre-albumin [ 45 ], lower concentrations of vitamin B12 and niacin in males [ 62 ], lower vitamin B12 in controlled versus uncontrolled population and almost significant increase in percentage of individuals with vitamin B12 deficiency [ 32 ]; higher burden in non-adherent ( n  = 5) [ 5 , 33 , 45 , 54 , 62 ], lower concentrations of phosphorus and vitamin B12 [ 45 ], lower concentrations of vitamin B12 and niacin in females as well as all other nutrients measured [ 62 ], lower intakes of iron, zinc, vitamin D3, magnesium, calcium, selenium, iodine, vitamin C, vitamin A, and copper, which were below UK Reference, and lower intakes of thiamin, riboflavin, niacin, vitamin B6, and phosphorus, which met UK Reference [ 5 ], lower concentrations of vitamin B12 and folate but levels were within or above normal range [ 5 , 45 , 54 , 62 ], significantly lower serum vitamin D3 and vitamin B12, below reference range, versus above reference range in adherent, lower serum folic acid and higher serum homocysteine but both within reference range [ 33 ]. Overweight/obesity: higher burden in adherent ( n  = 2) [ 33 , 62 ], obesity reported in larger proportion of participants [ 61 ], increased total fat free mass in non-adherent versus adherent [ 33 ]; higher burden in non-adherent ( n  = 3) [ 32 , 33 , 63 ], increased total fat mass, bodyweight, WC and BMI in non-adherent versus adherent (BMI within normal range for adherent), and decreased appendicular fat free mass in non-adherent versus adherent [ 33 ], significantly increased BMI in uncontrolled vs controlled (total population and women only), numerical increase in BMI of uncontrolled vs controlled men (controlled groups within normal range) [ 32 ]. Other: higher burden in adherent ( n  = 1) [ 60 ], higher incidence of otolaryngologic symptoms in adherent since birth (group 1) compared with off diet (group 2) and never treated (group 4), and gynecologic symptoms in adherent since birth (group 1) compared with never treated (group 4), higher incidence of arthritis/musculoskeletal symptoms in discontinued and restarted (group 3) compared with never treated (group 4) and off-diet (group 2), higher incidence of ophthalmologic and gynecologic symptoms in adherent since birth (group 1) compared with never treated (group 4) [ 60 ]; higher burden in non-adherent ( n  = 2) [ 57 , 60 ], higher incidence of gynecologic and ophthalmologic symptoms in off diet (group 2) compared with adherent since birth (group 1), higher incidence of otolaryngologic symptoms in discontinued and restarted (group 3) compared with adherent since birth (group 1), higher incidence of ophthalmologic symptoms in off diet (group 2) compared with discontinued and restarted (group 3), higher incidence of gynecologic symptoms in discontinued and restarted (group 3) compared with never treated (group 4), and higher incidence of ophthalmologic symptoms in never treated (group 4) compared with adherent since birth (group 1) [ 60 ], poorer thyroid function as measured by serum TSH, UIC and UIC/Cr [ 57 ] .  Definitions of adherence versus non-adherence: Adamczyk et al. 2011. [ 63 ], all individuals on Phe-restricted diet from within the first month of life, with blood Phe level assessment at least every second month: subgroup 2a (adherent) had recommended blood Phe levels for treated patients (2–10 mg/dL for children > 12 years), subgroup 2b (non-adherent) had blood Phe levels above the recommended level; Crujeiras et al. [ 45 ], those with high adherence to a natural protein restricted diet and supplementation with Phe-free amino acids mixture versus those with low adherence; Dios-Fuentes et al. [ 32 ], good metabolic control was defined as Phe levels < 600 µmol/L; Green et al. [ 5 ], minimum of 20 g protein equivalent from a low-Phe protein substitute per day for ≥ 1 month prior to inclusion with good adherence versus maximum of 20 g protein equivalent from a low-Phe protein substitute per day for ≥ 1 month prior to inclusion and blood Phe ≥ 600 µmol/L (of n  = 14 in this group: n  = 2 with 20 g of protein equivalent and no natural protein restriction; n  = 1 with low protein diet but no low-Phe protein substitute; n  = 11 with unrestricted diet and no low-Phe protein substitute); Guest et al. [ 60 ], remaining on Phe-restricted diet since < 1 year of age (group 1) versus discontinuation by 15–25 years of age (group 2) versus those off diet by 15–25 years of age but restarted diet at a mean of 30 years of age (group 3) versus those never treated (group 4); Koch et al. [ 61 ], Phe-restricted diet from infancy until ≥ 10 years of age and taking medical food as the primary protein source versus discontinuation of dietary restriction by age 10; Moden-Moses et al. [ 29 ], classified as diet-adherent or non-adherent based on self-report; Rojas-Agurto et al. [ 33 ], participants with a neonatal diagnosis of PKU, who continued with nutritional treatment, received an adequate supply of protein substitute without Phe, and kept strict follow-up were categorized as adherent, participants with a neonatal diagnosis of PKU, who discontinued the protein substitute and micronutrient supplementation (calcium, iron, and zinc) at 18 years of age and stopped attending metabolic control appointments; Robinson et al. [ 54 ], strict low-Phe diet with amino acid, mineral, and vitamin supplements versus no formal protein restriction and no amino acid vitamin and mineral supplementation (those on a total protein intake of approximately 1 g/kg/d with roughly 50% of this from natural protein and 50% from amino acid, mineral, and vitamin supplements were not included in the vote counting); Schulz et al. [ 62 ] taking amino acid mixture versus not taking amino acid mixture; Sumanszki et al. [ 57 ], mean blood Phe concentration for the 12-month period prior to the study < 600 μmol/L versus > 600 μmol/L

Six studies investigating nutritional outcomes were included in the analysis of adherent versus non-adherent PKU populations [ 5 , 32 , 33 , 45 , 54 , 62 ], with five studies reporting a higher burden of ≥ 1 outcome measure in those who were non-adherent [ 5 , 33 , 45 , 54 , 62 ] and three reporting a higher burden in those who were adherent [ 32 , 45 , 62 ]. In one study [ 45 ], the impact of dietary adherence differed with respect to the nutritional outcome measured. Total protein and serum pre-albumin concentrations were significantly lower in those with high adherence to diet versus those with low adherence ( p  = 0.0072 and p  = 0.00011, respectively), whereas concentrations of phosphorus and vitamin B12 were significantly lower in those with low adherence versus those with high adherence ( p  < 0.0001 and p  = 0.03, respectively) [ 45 ]. A second study [ 62 ] reported lower intake of all vitamins and minerals measured (statistically significant differences, except for potassium and phosphorus) in those not taking amino acid mixture (AAM) compared with those who adhered to AAM, except for intakes of vitamin B12 and niacin in males, which were higher in those not taking AAM, but the difference was only statistically significant for vitamin B12). Significantly lower intakes of many micronutrients were reported in non-adherent compared with adherent groups in a third study [ 5 ]; however, intakes of manganese, potassium, vitamin B12, sodium, chloride, and folate were similar between groups. In a fourth study [ 33 ], significantly lower (below reference levels) serum vitamin D3 ( p  < 0.01) and vitamin B12 ( p  = 0.03) were reported in individuals who had discontinued the protein substitute at 18 years of age and stopped attending metabolic control appointments (non-adherent group) compared with those who had continued adherence to a Phe-restricted diet with an adequate protein substitute (above reference levels). Levels of folic acid were lower and homocysteine levels were higher in the non-adherent group compared with the adherent group (but the difference between groups was not statistically significant and levels of both nutrients in both groups were within the reference range).

Although concentrations of vitamin B12 and folate measured in a fifth study [ 54 ] were lower in the unrestricted diet group than in the strict low-Phe diet group, statistical significance was only assessed versus the control population. In the sixth study [ 32 ], vitamin B12 levels were higher in the group of individuals with uncontrolled Phe levels (non-adherent) than in the group with controlled Phe levels (adherent) and this group included a higher percentage of individuals with vitamin B12 deficiency that was almost statistically significant ( p  = 0.053).

Four studies [ 33 , 60 , 61 , 63 ] investigated several types of comorbidities in diet-adherent versus non-adherent populations, including cardiovascular outcomes, migraine/headaches, cancer, COPD/asthma, and dermatologic outcomes [ 60 , 61 ], overweight/obesity [ 33 , 61 , 63 , 64 ], musculoskeletal outcomes [ 33 , 60 , 63 ], bone-related abnormalities [ 33 , 63 ], hypertension [ 61 ], gastrointestinal outcomes, and other outcomes [ 60 ]. The direction of the higher burden for several comorbidity types differed between studies or between different subgroups within the same study.

A higher burden of cardiovascular [ 60 , 61 ], dermatologic [ 60 , 61 ], migraine/headaches [ 60 , 61 ] and other [ 57 , 60 ] outcomes was found in non-adherent compared with adherent populations, in two studies each, compared with a higher burden for these same comorbidity types in adherent compared with non-adherent populations, in one study of migraine/headaches [ 60 ] and other outcomes [ 60 ], and no studies of cardiovascular or dermatologic outcomes. A higher burden of COPD/asthma was found in the non-adherent versus adherent population in two studies (see Fig. 5 footnotes [ 60 , 61 ]), while a higher burden of COPD/asthma was found in an adherent versus non-adherent population in one of these studies (see Fig. 5 footnotes [ 60 ]).

A higher burden of hypertension [ 61 ] was found in those who adhered to a Phe-restricted diet compared with those who were non-adherent. No studies reported a higher burden of hypertension in those who did not adhere. Another study [ 32 ] reported the prevalence of hypertension among the overall population of individuals with PKU as 7.9%, but the prevalence in the subgroups with controlled and uncontrolled Phe levels was not reported.

There was one study [ 63 ] with a higher burden of bone-related abnormalities in those with uncontrolled Phe levels (classified as non-adherent in the vote-counting analysis) compared with those with controlled Phe levels (considered as adherent in the vote-counting analysis) and two studies [ 29 , 33 ] with a higher burden in those who were adherent compared with those who were non-adherent.

The direction of the higher burden for overweight/obesity differed between studies. In one study [ 32 ], BMI was significantly higher in the total population and in women with uncontrolled Phe levels than in those with controlled Phe levels; median BMI for the total population was 27.45 kg/m 2 versus 24.36, p  = 0.023; median BMI for the female population was 28.11 versus 22.58, p  = 0.007, but the difference in median BMI between men with controlled Phe levels and those with uncontrolled Phe levels was not statistically significant ( p  = 0.923). It should be noted that 18/90 (20%) of included individuals in the study received BH4 rather than dietary therapy, and eight of these required Phe-free amino acid formula to achieve metabolic control [ 32 ]. Two studies [ 61 , 63 ], reported a higher burden of overweight/obesity in individuals who were diet-adherent [ 61 ] or had controlled Phe levels [ 63 ] compared with the non-adherent [ 61 ] or uncontrolled Phe levels [ 63 ] group: 33.3% versus 16.4% with obesity [ 61 ] and higher bodyweight and BMI (absolute and Z-scores) [ 63 ] but statistical significance of the difference between groups for individual comorbidities was not assessed. In a third study [ 33 ], there were numerical increases in bodyweight, WC, BMI, and total fat mass, and numerical increases in appendicular fat-free mass in the diet-adherent group compared with the group who had discontinued the Phe-restricted diet at 18 years of age (measures of total fat-free mass were similar) but differences between groups were not statistically significant. An additional study [ 50 ] found no effect of metabolic control on BMI classification and bioelectrical impedance parameters (indicators of overweight/obesity) but numerical data were not reported, hence this study could not be included in the vote-counting analysis.

The burden of gastrointestinal symptoms and cancer was higher in adherent compared with non-adherent populations in one study [ 52 ], but also higher in non-adherent compared with adherent populations for these same comorbidities. This study [ 52 ] compared groups with varying levels of adherence: the highest incidence of gastrointestinal symptoms was in those who had never been treated with a Phe-restricted diet, was similar in those who had adhered to dietary treatment throughout life and in those who had discontinued dietary treatment between the ages of 15 and 25 years [ 60 ]. None of the group who had discontinued and restarted a Phe-restricted diet experienced gastrointestinal symptoms [ 60 ]. Conversely, the incidence of cancer was highest in those who had discontinued and restarted a Phe-restricted diet, followed by those who had discontinued their diet between 15 and 25 years of age and then those who had adhered to a Phe-restricted diet throughout life. None of the group who had never been treated with dietary therapy had cancer [ 60 ].

Individuals with PKU on different Phe-restricted diets

A comparison of somatic comorbidities in individuals with PKU on different Phe-restricted diets was reported in four studies (Table  4 ). These studies did not compare a PKU population on a Phe-restricted diet with either healthy controls or no intervention, and therefore did not meet the inclusion criteria of the SLR. The scenario of those on different diets being compared was not anticipated by the PICOS criteria but nevertheless these studies have been included because the comparison is of potential interest, from the perspective of the impact of dietary improvements on the clinical burden of somatic comorbidities.

Two studies [ 64 , 65 ] compared L-amino acid (L-AA) supplements versus glycomacropeptide-based (GMP) protein substitute (GMP-AA) or modified casein GMP amino acid (CGMP-AA) supplements; one study [ 66 ] compared a pre-trial Phe-restricted diet (protein substitute that included significant quantities of added carbohydrate) with a new Phe-restricted diet (low-carbohydrate Phe-free protein substitutes); another study [ 67 ] compared individuals not on a diet (normal food group) versus those on a vegan diet without AAM versus those on a vegan diet with AAM versus a protein reduced diet with AAM supplements.

One study [ 65 ] reported no changes in the prevalence of overweight/obesity (BMI p  = 0.367); another study [ 64 ] reported a tendency for increased body weight ( p  = 0.064) and total body fat ( p  = 0.056) in individuals on CGMP-AA when compared with baseline on L-AA, but these changes were not statistically significant. Two studies [ 64 , 65 ] reported on cardiovascular outcomes, and no differences in blood pressure between those on L-AA versus those on GMP-AA or CGMP-AA were found. One study [ 67 ] reported on nutritional outcomes and found no statistically significant difference was found in trace elements iron, zinc, or selenium between those not on a diet versus a vegan diet without AAM versus a vegan diet with AAM versus a protein reduced diet with AAM (no p value reported).

Individuals on a Phe-restricted diet with more severe PKU versus those with HPA or less severe PKU

Five studies reported on somatic comorbidities experienced by individuals on a Phe-restricted diet with cPKU (more severe form of disease) versus individuals with HPA or less severe PKU (Table  5 ). One study [ 68 ] reported bone-related abnormalities in individuals with cPKU versus those with mild or moderate PKU; one study [ 55 ] reported anthropometric parameters and markers of metabolic syndrome/diabetes in individuals with cPKU, mild or moderate PKU, and mild HPA; one study [ 50 ] reported BMI classifications (relating to underweight, normal weight, overweight, and obese) and bioelectrical impedance parameters (relating to fat mass and fat-free mass) in individuals with cPKU and mild PKU; one study [ 69 ] reported nutritional outcomes in individuals with PKU versus those with HPA; and one study [ 45 ] reported nutritional outcomes in individuals with cPKU, mild or moderate PKU, and HPA (Table  5 ).

In the study comparing cPKU with mild or moderate PKU [ 68 ], the prevalence of osteopenia and osteoporosis was reported to be similar between those in either group on a Phe-restricted diet; statistical significance for the difference between groups was not reported.

One study [ 44 ] reported BMI and WC above the upper limit (indicating overweight/obesity) in a significantly higher proportion of individuals with PKU versus mild HPA ( p  = 0.0062 for overall population; p  = 0.010 for BMI and p  = 0.0011 for WC in adults) while another study [ 50 ] found that type of PKU (cPKU or mild PKU) did not affect BMI classifications or bioelectrical impedance parameters (numerical data were not reported). Fasting insulin levels above the upper limit were reported in a significantly higher proportion of those with PKU versus mild HPA ( p  = 0.035). Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) was significantly increased ( p  = 0.034) and Quick index score was significantly decreased ( p  = 0.019) in those with PKU versus mild HPA, both indicating worse insulin resistance in those with more severe forms of the disease. Quick index score was significantly lower, and HOMA-IR was significantly higher, in patients with cPKU than in those with mild or moderate PKU and those with mild HPA; therefore increasing severity was correlated with increasing BMI, WC, and age [ 44 ].

In one study [ 69 ], no statistically significant differences were found in the concentrations of serum pre-albumin, zinc, and iron between adults with PKU and those with HPA ( p value not provided) [ 69 ]. However, there was a statistically significant lower concentration of selenium in adults with PKU compared with adults with HPA ( p  = 0.006). Another study [ 45 ] found that concentrations of selenium and phosphorous were significantly reduced in those with PKU (mild/moderate and cPKU grouped together) versus mild HPA ( p  = 0.0034 and p  = 0.0056, respectively), although there were only five individuals with phosphorous levels lower than the normal limit. Conversely, serum pre-albumin, ferritin, and folic acid concentrations were significantly reduced in those with mild HPA versus those with mild, moderate, or cPKU ( p  = 0.024, p  = 0.0084, and p  = 0.0147, respectively) [ 45 ]. In the same study, vitamin B12 and zinc were significantly reduced in those with mild HPA and mild or moderate PKU compared with those with cPKU ( p  = 0.0046 and p  = 0.03, respectively) [ 45 ]. However, it should be noted that levels of total protein, calcium, phosphorous, vitamin B12, ferritin, and zinc were within the normal range in the majority of individuals with PKU, and none had a folic acid deficiency [ 45 ].

Main findings

This review has highlighted the breadth of somatic comorbidities experienced by individuals with PKU, and the higher clinical burden versus a non-PKU population. The findings add to the published literature, confirming the comorbidity burden in individuals with PKU treated with a Phe-restricted diet [ 26 , 70 , 71 ]. The most commonly reported somatic comorbidities in studies of individuals with PKU on a Phe-restricted diet with or without pharmacologic therapy compared with healthy controls or reference values were bone-related abnormalities [ 15 , 16 , 22 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 ], followed by overweight/obesity [ 15 , 16 , 22 , 33 , 43 , 44 , 49 , 50 , 51 , 52 , 53 , 54 ], nutritional outcomes [ 22 , 26 , 33 , 45 , 48 , 49 , 54 , 56 , 58 , 59 ], and cardiovascular outcomes [ 15 , 22 , 43 , 46 , 47 ]. It was not possible to draw definitive conclusions from the other three population comparisons due to the limited number of studies included in each comparison and differences in the comorbidity types covered: adherent to a Phe-restricted diet versus non-adherent, twelve studies [ 5 , 29 , 32 , 33 , 45 , 50 , 54 , 60 , 61 , 62 , 63 ]; groups on different Phe-restricted diets, four studies [ 64 , 65 , 66 , 67 ]; and more severe PKU versus HPA or less severe PKU, five studies [ 44 , 45 , 50 , 68 , 69 ].

Relation of main findings to prior research

In a published SLR, Pessoa et al . reported a high prevalence of clinical complications (e.g., overweight/obesity and osteopenia), poor adherence to clinical recommendations, negative socioeconomic impact, and negative impact on caregivers of Latin American patients with PKU of all ages (diagnosed with PKU within the first 3 months of life) [ 70 ]. The study concluded that dietary management alone was not sufficient to prevent the burden of PKU, which concurs with the findings of our SLR, in which individuals with PKU were compared with healthy controls or a reference population in the vote counting analysis. It is important to acknowledge that our SLR did not investigate the negative socioeconomic impact, or the negative impact on caregivers.

Abnormal bone status has been a long-standing concern in individuals with PKU [ 3 , 72 ]; therefore, it is not surprising that the most reported somatic comorbidities in individuals with PKU on a Phe-restricted diet in our SLR were bone-related abnormalities. BMD Z-score was the most reported outcome measure, but a variety of other outcome measures was reported among the included studies and there was a higher burden of at least one bone-related outcome in individuals with PKU compared with healthy (non-PKU) controls in many of these studies. Currently, it is unclear whether low BMD in individuals with PKU is a direct consequence of the disease, a complication of following a Phe-restricted diet or due to reliance on low-Phe amino acid supplementation (medical foods), which can increase urinary calcium and magnesium excretion [ 28 , 38 , 73 ]. Emerging evidence suggests that the PKU population may be at increased risk of metabolic acidosis, which has been linked to low bone mineralization [ 74 ]. This adds to the debate on whether the increased renal acid load from consumption of low-Phe medical foods is related to low BMD and highlights the need to further explore the etiology and impact of bone-related abnormalities in individuals with PKU. A meta-analysis has been conducted to investigate BMD outcome measures in adults with PKU, and to explore the impact of the Phe-restricted diet (including the impact of adherence to diet) on BMD. For further details on the BMD meta-analysis and its findings, refer to the separate meta-analysis publication.

Burton et al. [ 16 ] reported that high blood Phe levels may impact biological mechanisms that are related to increased risk of comorbid conditions such as obesity, renal disease, metabolic dysfunction, and cardiovascular complications, which might explain why nutritional outcomes, cardiovascular outcomes, and overweight/obesity were also commonly reported in the studies included in our SLR.

The Phe-restricted diet limits the intake of natural protein to vegetable sources, and despite the availability of low-Phe medical foods, a significant number of adolescents and adults do not consume adequate amounts of protein substitutes [ 58 ]. As a result, individuals with PKU have been shown to be at risk of deficiencies in nutrients such as carnitine and vitamin B12, which are derived from animal protein sources [ 49 , 54 , 58 , 59 ]. Clinical symptoms of carnitine deficiency include muscle weakness or cardiomyopathy, which may be caused by low intake of dietary carnitine, deficient synthesis of carnitine, or acyl-carnitine production [ 59 ]. Vitamin B12 deficiency may lead to anemia, gastrointestinal, and neurological symptoms [ 58 ].

Folate is an essential vitamin that plays a crucial role in metabolism [ 75 ]. High levels of folate have been attributed to the high folic acid content in protein substitutes [ 76 ]; however, higher and lower concentrations of folate in individuals with PKU compared with controls (as well as levels above or within the normal range) have been reported [ 26 , 33 , 45 , 48 , 54 , 76 ]. For both vitamin B12 and folate, the risk of deficiency was higher in those who were not following a strict low-Phe diet with adequate amino acid and vitamin and mineral supplementation [ 45 , 54 ]. Hochuli et al. [ 77 ] also found that a relaxation of AAM intake resulted in insufficient nutrient supply despite a compensatory increase in consumption of natural protein. The evidence from these studies indicates the need for continual dietary guidance through adulthood, as inadequate intake of nutrients can lead to further comorbidities. One study included in our SLR [ 46 ] reported an association between high Phe levels and arterial stiffness, which impacts the risk of cardiovascular disease [ 16 ]. However, another included study [ 47 ] did not identify any significant difference in arterial stiffness or carotid intima media thickness (a surrogate marker of atherosclerosis) compared with healthy controls. Similar to other comorbidities, there are limited data available to explain whether an increased cardiovascular risk in individuals with PKU is due to the disease itself or factors related to the Phe-restricted diet [ 32 ].

Given the potential for increased risk of obesity with high blood Phe levels noted by Burton et al. [ 16 ], we felt it was important to acknowledge the inconsistency among conclusions of the studies reporting this outcome that were included in our SLR and other previously published SLRs. Of the studies included in our SLR, four studies [ 33 , 50 , 53 , 55 ] found no significant difference in the proportions of individuals with PKU who were overweight or obese (as measured by prevalence, body weight, WC, BMI, body fat percentage, total or appendicular body fat mass, total or appendicular fat-free mass, or central obesity) compared with healthy controls; however, a significantly higher BMI, prevalence ratio or percentage of individuals with overweight/obesity in individuals with PKU compared with matched controls was reported in four studies [ 15 , 16 , 22 , 43 ]. Two studies provided supporting evidence for an increased burden of overweight/obesity in individuals with PKU: in one study [ 44 ], there was a higher proportion of overweight/obesity in those with PKU versus controls, but statistical significance of the difference between groups was not reported; and in another study [ 52 ] there was a higher rate of obesity among females with PKU in four of six centers, but the overall proportion of overweight individuals was lower in five of the six centers studied. The results of the tenth study [ 56 ] were conflicting, with higher percentages of overweight and lower percentages of obese individuals in the PKU group versus controls (statistical significance was not reported). Two published SLRs identified in our review have also reported conflicting results on the prevalence or risk of overweight/obesity in individuals with PKU [ 78 ]: one SLR, published in 2021 [ 71 ], concluded that individuals with PKU (including children, adolescents, and adults) had similar BMI to healthy controls, although BMI was significantly higher than healthy controls in a subgroup of individuals with cPKU; another SLR, published in 2023 [ 78 ], concluded that adults with PKU had a higher BMI and higher prevalence of obesity compared with a matched control population but the proportions of the PKU population with obesity varied between studies from 4.5% to 72% and the findings were inconsistent when compared with the general population. A previously reported SLR and meta-analysis investigating whether a Phe-restricted diet is a risk factor for overweight/obesity in individuals with PKU found that BMI was similar between individuals with PKU and healthy controls [ 71 ]. In the study reporting the frequency of overweight/obesity in individuals with PKU receiving different treatments [ 56 ], the highest percentage of overweight individuals was in the dietary therapy group, followed by the pegvaliase group and sapropterin dihydrochloride group, but the highest percentage of obese individuals was in the pegvaliase group and the lowest was in the dietary therapy group; therefore, the pegvaliase group had the highest rate of overweight/obesity overall and the sapropterin dihydrochloride group had the lowest.

Differences in population characteristics relevant to obesity may have contributed to the different outcomes observed between studies included in our SLR, e.g., PKU cohorts in studies with no significant difference from controls tended to include younger participants (mean age 14.4 [ 55 ], 23.5 [ 33 ], and 26.0 [ 53 ] years, mean age not reported [range 6–25 years] [ 50 ] versus mean age 30.8 [ 43 ], 34.6 [ 16 ], 41.2 [ 22 ], and 50.9 [ 15 ] years) and in three studies [ 16 , 22 , 43 ], were all early-diagnosed/treated, as opposed to Trefz et al. [ 15 ], who included a higher number of late- versus early-diagnosed participants ( n  = 216 versus n  = 161) [ 15 , 53 , 55 ]. A study comparing early- versus late-diagnosed individuals found that the proportion of those with a BMI above the upper limit was almost twice as high in late- versus early-diagnosed participants ( p  = 0.023) [ 44 ]. However, timing of diagnosis was not reported in two studies [ 16 , 43 ] and the proportion of late-diagnosed individuals was relatively low (29.7%) in another study [ 22 ]. Differences in male:female ratio may also play a role, as studies reporting a significant difference in the prevalence or risk of overweight/obesity tended to include a higher proportion of females with PKU than studies finding no significant difference between groups (46% [ 55 ], 48.1% [ 50 ], 50% [ 33 ], and 51% [ 53 ] female versus 56% [ 22 ], 58.1% [ 15 ], and 63.7% [ 16 ] female). These results are supported by a published SLR, which found that overweight/obesity was 2–3 times more frequent in females with PKU than males [ 78 ], and the results of Ozel et al. [ 52 ] included in our SLR. Azabdaftari et al. [ 43 ] was the exception, reporting significantly higher BMI in adults with PKU compared with healthy controls, of whom only 39% were female. Furthermore, Rocha et al. [ 55 ] found no effect of male:female ratio on prevalence of overweight/obesity. Overweight/obesity is a complex comorbidity that is likely to be impacted by the components of the Phe-restricted diet, adherence to diet, and other factors relating to individual patient behaviors. The inconsistency in findings indicates a need for further research.

Adherence to a Phe-restricted diet is often determined by using blood Phe levels as an indicator (where low blood Phe levels indicate a greater level of treatment adherence). However, it is important to note that blood Phe levels are also dependent on the extent of an individual’s functional PAH, which in turn is determined by the mutations in the gene encoding PAH for that individual [ 79 , 80 ]. An individual’s genotype correlates with biochemical phenotype, including the degree of Phe tolerance linked to disease severity [ 80 ], and so elevated blood Phe levels in some treated individuals (e.g., those with mutations rendering PAH inactive), may be due to low Phe tolerance and severe disease rather than non-adherence. A recent study, included in the SLR, comparing maintenance and suspension of dietary treatment in adults with PKU noted that an elevated blood Phe level, as a result of abandoning a Phe-restricted diet, has the potential to affect muscle and bone health, but other factors such as reduced quality of protein intake, reduced intake of vitamins and minerals, and lower physical activity, could also play a part [ 33 ]. Further investigation on the impact of adherence to diet on the clinical burden of somatic comorbidities is needed due to the limited number of studies currently published.

As part of MNT, individuals with PKU can be placed on different types of Phe-restricted diets and/or protein-based supplements. Elevated blood Phe levels can lead to serious nutritional deficiencies, and studies have reported that individuals who did not adhere to a Phe-restricted diet had higher blood Phe levels in comparison with those following a Phe-restricted diet [ 67 ]. Despite this, the studies in our SLR reported that many adults do not adhere to dietary recommendations owing to the taste, smell, and texture of protein substitutes/supplements and the inconvenience of these diets [ 66 , 67 ].

It has also been suggested that the Phe-restricted diet may be linked to certain comorbidities. For example, a recent SLR [ 78 ] found that low-Phe food substitutes tend to be higher in carbohydrates, which is a known causative factor for overweight and obesity. However, a previous study found that BMI only increased in 55% of individuals receiving lower versus higher carbohydrate Phe-free protein substitutes; was unchanged in 5%; and actually decreased in 40%; and, overall, the difference between groups was not statistically significant [ 66 ]. Additionally, studies that compared the prevalence of obesity between individuals receiving amino acids versus GMP substitute reported conflicting results [ 64 , 65 ].

Poor metabolic control has previously been reported in individuals with cPKU receiving dietary treatment compared with individuals with mild PKU or HPA [ 6 ]. Despite the extensive amount of research into the importance of metabolic control, there is limited understanding of the impact of PKU severity on the clinical burden of somatic comorbidities.

In our SLR, a study that assessed individuals with mild, moderate, and cPKU found no relationship between PAH genotype and the development of mineral bone disease (MBD), no difference in the blood Phe levels between individuals with PKU who developed MBD and those who did not, and no relationship between diet compliance and MBD [ 68 ]; however, individuals with osteopenia or osteoporosis in this study had significantly lower natural protein intake compared with those without MBD, and the sapropterin dihydrochloride-treated individuals who were able to relax restrictions to natural protein intake in this study did not develop MBD, suggesting that a decrease in natural protein intake has a role in the development of MBD [ 68 ].

Two studies identified in our SLR investigated the nutritional status of individuals with different PKU phenotypes [ 45 , 69 ]. One study [ 69 ] reported significantly lower concentrations of selenium in those with PKU receiving AAM compared with those with HPA; the authors suggested that this result was affected by adults relaxing restrictions to a low-protein diet and, hence, overall intake of amino acids may be higher than the prescribed AAM. Another study [ 45 ] also found significantly lower selenium and phosphorus concentrations in those with PKU versus those with HPA but concluded that this was related to increased dietary adherence and younger age (< 18 years). Conversely, serum pre-albumin, ferritin, and folic acid concentrations were significantly reduced in those with mild HPA versus those with PKU and vitamin B12 and zinc were significantly reduced in those with mild HPA and mild or moderate PKU compared with those with cPKU [ 45 ]. The positive correlation of vitamin B12 and zinc concentrations with disease severity was surprising but the authors postulated that despite lower intake of these nutrients through natural protein sources in those with cPKU, the cobalamin content of Phe-free amino acid supplements is high and there is a more efficient absorption of zinc salts from supplements than from meals [ 45 ].

The relationship between PKU severity and the risk of overweight/obesity is unclear. Mazzola et al. [ 50 ] found no effect of PKU phenotype on BMI classification or bioelectrical impedance parameters, whereas Couce et al. [ 44 ] reported overweight/obesity and worse insulin resistance (a marker of the metabolic syndrome) in a significantly higher proportion of individuals with PKU versus mild HPA and healthy controls, particularly in those with a late diagnosis. Higher levels of insulin resistance were correlated with increasing BMI, WC, and age [ 44 ]; however, in a linear regression model, age had the most influence on BMI; therefore, the authors concluded that the cause of increased insulin resistance in PKU is likely to be multifactorial [ 44 ].

Due to the small number of studies included in this population comparison and the breadth of outcome measures reported, it is not feasible to draw definitive conclusions on the impact of disease severity on the clinical burden of somatic comorbidities.

Clinical implications

The higher clinical burden of somatic comorbidities in individuals with PKU compared with a general (non-PKU) population, as reported here and in other studies, contributes to the clinical and socioeconomic burden of PKU; higher rates of prescribed treatment use have been reported, and mean healthcare costs are significantly greater ( p  < 0.0001) than for the general population [ 15 , 16 ]. The findings from this review point to an unmet need for optimized approaches to Phe control, to maintain Phe levels within recommended ranges over the long term and potentially avoid somatic comorbidities and the associated clinical and socioeconomic implications.

Adherence to dietary restrictions can be challenging, especially as individuals with PKU progress to adulthood. Limited and unpleasant food choices, as well as the overall inconvenience and time-consuming nature of a restricted diet and effect on socializing, are cited as factors negatively impacting adherence in adults with PKU [ 81 , 82 ]. Patients and their caregivers report a lack of disease awareness in hospitality settings, leading to concern around menu choices outside the home that are incompatible with dietary restrictions [ 83 ]. The challenge of dietary adherence may impact a patient’s quality of life (QoL). Studies using a PKU-specific questionnaire to evaluate the effect of PKU on QoL [ 84 , 85 ], reported the emotional impact of PKU and the management of PKU (anxiety about blood Phe levels; guilt regarding poor adherence to dietary restrictions) as high scoring questionnaire domains, indicating a negative impact on QoL. The unpleasant taste of food supplements was also considered a main issue [ 85 ]. Another study [ 86 ], using preference-based measures to estimate the effect of PKU on health-related QoL, reported dietary restrictions and symptoms of PKU as both having a significant negative impact on health-related QoL.

There is a need for careful monitoring of nutritional intake as a component of nutritional status assessment, independent of treatment modality. Closer medical surveillance may prevent losing individuals to follow-up, particularly as they progress from adolescence to adulthood, and increased awareness of screening for comorbidities may identify individuals in most need of improved Phe control. It is important that individuals with PKU, who already have limited options in food choices owing to their Phe-restricted diet, receive ongoing personalized nutritional counselling, with methodical nutritional status monitoring from a multidisciplinary team specialized in inherited metabolic disorders to prevent overweight, obesity, and its related comorbidities.

Some treatment approaches may improve Phe and natural protein tolerance, reducing the reliance on strict Phe restriction for metabolic control, which may lead to improvements in QoL over time. Although a meta-analysis of outcomes before and after relaxation of a Phe-restricted diet with sapropterin dihydrochloride did not find an improvement in QoL following treatment, the authors of the study noted this finding does not reflect clinical practice [ 87 ]. One explanation offered for this anomaly was that the general QoL questionnaires used in the majority of studies included in the analysis may not be sensitive enough to capture the daily burden of a highly restricted diet [ 87 ]. Indeed, the only study utilizing a PKU-specific QoL questionnaire to investigate the impact of treatment on QoL, reported significant improvements in self-reported impact and satisfaction sub-scores and total QoL score over time in adolescents and adults responding to sapropterin dihydrochloride, and QoL improvements were associated with increased Phe tolerance [ 88 ]. It is important to note that these improvements can only be achieved when an adequate and balanced diet is also maintained. Educating both clinicians and individuals with PKU on the role of balanced nutrition to effectively manage and/or prevent chronic disease is required [ 89 , 90 ], as well as adjusting nutrition according to pharmacologic intervention.

Strengths and limitations of the review methodology

This review involved sensitive searches of the peer-reviewed literature only and was guided by the pre-defined eligibility criteria established in the protocol. Comprehensive, relevant, and accurate data abstraction was ensured throughout the review process.

Strengths of this review include the consideration of any somatic comorbidity in assessing clinical burden, and the focus primarily on an adult population (studies conducted exclusively in children and adolescents were excluded). With the introduction of PKU testing into newborn screening programs more than half a century ago, and the availability of treatment accordingly, the global PKU population is increasing in age. The neuropsychologic burden of PKU is well documented, particularly in children, and this review provides new insight into the somatic comorbidity burden in adults.

Heterogeneity in the studies that were included in this review, in terms of the clinical outcomes used (and their definitions); how each outcome was measured; and study designs, including interventions and comparators, meant meta-analysis of effect estimates was not considered appropriate for the majority of outcomes. Vote counting is considered an acceptable alternative synthesis method when meta-analysis is not feasible [ 21 ], and an assessment of the direction of effect of comorbidity burden was still able to be made, even when several studies did not report statistical significance of the difference between groups. Hence, the results of the vote-counting analysis provided an overview of the somatic comorbidity burden, in terms of both the range of comorbidities and the direction of effect, in individuals with PKU compared with a non-PKU population. However, vote counting does have limitations: the analysis does not account for differences in the relative sizes of the studies or methodological aspects and provides no information on the magnitude of effect; it is also difficult to interpret the results when studies report multiple outcomes and/or measures and the direction of effect differs between them. This is compounded by inclusion of small numbers of studies, as was the case for the comparison between individuals with PKU with different levels of adherence to a Phe-restricted diet. Other alternative synthesis methods to meta-analysis are more powerful than vote counting (e.g., combining p values), but were not appropriate for this review due to incomplete data. Meta-analysis was only feasible for BMD Z-scores from a sub-set of the studies reporting bone-related abnormalities included in the SLR (see Fig. 4 ). The objectives of the meta-analysis were more specific than those of the broader SLR and therefore are reported separately.

It is important to acknowledge that the grouping of individual somatic comorbidities into comorbidity types could have been done differently and may have resulted in different conclusions being drawn from the data. In this study, cardiovascular outcomes were among the most reported somatic comorbidities with a higher burden in individuals with PKU on a Phe-restricted diet with or without pharmacologic therapy ( n  = 5). The grouping of cardiovascular outcomes with other comorbidities that are considered risk factors for cardiovascular disease (e.g., diabetes, overweight/obesity, hypertension) was explored, but not undertaken due to variance in opinion on the most appropriate categorization given the complexities of these individual comorbidities, and to maintain a level of granularity on the comorbidity burden in PKU. Consequently, the overall cardiovascular burden in individuals with PKU in the current treatment landscape may be underestimated in the vote-counting analysis.

This review considered studies published in English and retrievable via the PubMed® interface, and in-built filters for human studies and adult age-groups were employed, which rely on appropriate indexing within the MEDLINE database for retrieval. MEDLINE has relatively broad coverage of the medical literature and it is likely that the majority of relevant studies would have been identified, but it is important to acknowledge that any studies reported outside of MEDLINE and other sources accessible via PubMed®, or outside the filters employed in the search, will have been missed.

Finally, it is important to note that a formal risk of bias assessment of the studies included in this review was not conducted, limiting the certainty of the findings. There was a high degree of heterogeneity across studies in terms of different study designs, outcomes and outcome measures reported, patient characteristics, and components of MNT. As such, there is a clear high-risk of bias rendering a formal assessment as unnecessary. However, despite the potential for a loss of the ‘signal’ relating to a higher burden of somatic comorbidities in individuals with PKU compared with healthy controls, owing to the heterogeneity across studies, the results of the vote-counting analysis did show a consistent direction of effect. It is important to acknowledge that there is a potential for publication bias. A formal assessment of publication bias would be challenging for the spread of outcomes and measures reported in this SLR. However, it is possible that studies that did not report a higher burden of somatic comorbidities in individuals with PKU have not been published, which would mean that the higher burden in individuals with PKU may be overestimated. Despite this, the results of the vote-counting analysis were consistently in favor of a higher burden in individuals with PKU versus healthy controls. The potential for publication bias to overestimate the proportion of studies showing a higher burden in individuals with PKU versus healthy controls was minimized by excluding single-cohort studies in which there was no comparison between individuals with PKU and healthy controls or reference values. We believe that this supports our conclusions of a higher somatic comorbidity burden in individuals with PKU, with the caveat that vote-counting analysis is associated with limitations.

Strengths and limitations of the included evidence

Studies included in this review represented individuals with PKU across Asia, Europe, North America, and South America, although most studies were conducted in Europe or the United States, suggesting findings will be broadly generalizable to these region-specific PKU populations. There was a broad publication date range for the included studies, spanning from 1990 to 2023 as no date restrictions were applied to the search. However, improvements in the taste and palatability of protein substitutes and low protein foods that have occurred over time could have improved patient compliance, complicating the interpretation of results across older and newer studies. Most studies identified for inclusion in the review were observational in design, supporting the need for systematic assessments of data such as this one.

Although one study [ 22 ] compared results for the overall population with PKU and results for a subgroup of individuals with early diagnosed PKU against results for the control population and another study [ 50 ] stated that time of diagnosis did not affect anthropometric or bioelectrical impedance parameters (such that results from all individuals with PKU were analyzed as a single group), only two studies included in the review directly compared the somatic comorbidity burden in early and late-diagnosed adults with PKU [ 15 , 44 ], making it difficult to further evaluate comorbidity burden by time of diagnosis or period of treatment. In the claims-based study that was identified and included [ 15 ], a range of comorbidity types was reported in early- and late-diagnosed adults with PKU and at a higher prevalence than a matched control population for many, including infectious gastroenteritis and colitis, overweight and obesity, hypotension, and disorders of lipid metabolism and other lipidemias. In a cross-sectional study conducted in a Spanish PKU population [ 44 ], there was a higher proportion of those with BMI above the upper limit in late-diagnosed compared with early-diagnosed individuals but this was found to be largely driven by the effect of age.

Seven studies included in the comparison of comorbidity burden in a Phe-restricted diet population versus healthy controls or reference values were conducted in mixed treatment populations [ 15 , 22 , 24 , 25 , 32 , 46 , 52 ]. These studies included some individuals on pharmacologic therapy in addition to a Phe-restricted diet [ 15 , 22 , 25 , 32 , 52 , 57 ], and individuals on and not on a Phe-restricted diet [ 24 , 52 ]. These studies were all included but results were not stratified by treatment type. A large proportion of studies included in the SLR did not stratify results by age-groups such that participants < 16 years of age could have been included; the relevance of this to the findings of the review is unknown. The scarcity of data from studies using similar study designs and patient populations, as well as consistency in outcomes used and the way they are measured for many comorbidities, restricts the synthesis methods that can be used to evaluate the burden of somatic comorbidities in adults with PKU. Although vote counting is considered an acceptable alternative synthesis method when meta-analysis is not feasible [ 21 ], not all studies could be included in the vote-counting analysis reported here due to lack of comparative data. A lack of reporting of statistical significance of differences between PKU populations in individual studies also limited interpretation of the presence or absence of a difference in comorbidity burden.

Future studies

The variety of outcomes reported, and outcome measures used in studies included in this review, limited the synthesis of data across many somatic comorbidity types and PKU populations and, accordingly, the conclusions that can be drawn. There is a need for studies assessing somatic comorbidities using more robust study designs and consistent outcome measures, and including specific PKU population comparisons (e.g., by disease severity, timing of diagnosis and treatment [early or late], adherence to treatment, male:female ratio, and age). Consideration of the aging PKU population will be important, given older adults may have different comorbidities to a younger adult population.

It is unclear whether the somatic comorbidity burden in PKU comes from the disease itself or from adherence to severe dietary restrictions and/or inadequate supplementation of amino acids and micronutrients. There is a need to further evaluate the relationship between effective metabolic control and comorbidity burden to understand whether control of blood Phe levels can reduce the incidence of such complications. Consideration of the multidisciplinary healthcare team structure will be important, as adherence to diet may be lower in those centers where individuals are managed by an incomplete team, particularly for individuals with cPKU.

Studies investigating the burden of illness of PKU, incorporating both patient and caregiver health-related QoL assessed with disease-specific instruments, and the impact of different treatments, will provide insight into the indirect costs of somatic comorbidities in individuals with PKU, furthering understanding of the socioeconomic implications of somatic comorbidities and impact of effective treatment. Pessoa et al. also highlighted the need for burden of illness studies to identify the range of ongoing and significant complications experienced by individuals with PKU in order to inform healthcare providers and public health authorities [ 70 ].

Individuals with PKU have a higher somatic comorbidity burden versus a non-PKU population, highlighting the unmet need for optimized approaches to blood Phe control in this population. Improved access to therapeutic interventions to maintain blood Phe levels within recommended ranges over the long term may potentially avoid the clinical and economic implications of managing comorbidities. To build on the evidence from this review and better understand the relationship between blood Phe control, adherence to diet and comorbidity burden, more robust studies reporting consistent outcome measures are needed, especially in specific PKU populations.

Availability of data and materials

The datasets and study protocol used and/or analyzed during the current study are available from the corresponding author on reasonable request. Additional supporting documents may be available upon request. Investigators will be able to request access to these data and supporting documents via a data sharing portal ( https://www.biomarin.com/our-science/funding-and-support/publication-data-request/ ) beginning 6 months and ending 2 years after publication. Data associated with any ongoing development program will be made available within 6 months after approval of relevant product. Requests must include a research proposal clarifying how the data will be used, including proposed analysis methodology. Research proposals will be evaluated relative to publicly available criteria available at https://www.biomarin.com/our-science/funding-and-support/publication-data-request/ to determine if access will be given, contingent upon execution of a data access agreement with BioMarin Pharmaceutical Inc.

Abbreviations

Amino acid mixture

Assigned value

Bone-specific alkaline phosphatase

Body cell mass

Tetrahydrobiopterin

Bone mineral density

Body mass index

Calcium/creatinine index

Charlson Comorbidity Index

Modified casein glycomacropeptide amino acid

Confidence interval

Carotid intima media thickness

Chronic obstructive pulmonary disease

Classical PKU

Docosahexaenoic acid

Extracellular mass

Eicosapentaenoic acid

Gastroesophageal reflux disease

Glycomacropeptide-based protein

Glycomacropeptide-based protein amino acid

Homeostasis model assessment insulin resistance

Hyperphenylalaninemia

Interquartile range

L-amino acid

L-amino acid mixture

Lean bone mass

Large neutral amino acids

L-type amino acid transporter 1

Mineral bone disease

Medical Subject Headings

Methylmalonic acid

Medical nutrition therapy

National Health and Nutrition Examination Survey

Not reported

Osteoclastogenesis

• Phenylalanine hydroxylase
• Phenylalanine

Population, Intervention, Comparator, Outcome, Study design

• Phenylketonuria

Prevalence ratio

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Protein substitute

Quality of life

Spongy bone mineral density

Standard deviation

Standard error

Systemic inflammatory response syndrome

Systematic literature review

Strength strain index

Synthesis Without Meta-analysis

Total bone mineral density

Thyroid-stimulating hormone

Urinary iodine concentration

Creatinine-normalized urinary iodine concentration

Waist circumference

Williams RA, Mamotte CD, Burnett JR. Phenylketonuria: an inborn error of phenylalanine metabolism. Clin Biochem Rev. 2008;29(1):31–41.

PubMed   PubMed Central   Google Scholar  

Blau N, van Spronsen FJ, Levy HL. Phenylketonuria. Lancet. 2010;376(9750):1417–27.

Article   CAS   PubMed   Google Scholar  

van Wegberg AMJ, MacDonald A, Ahring K, et al. The complete European guidelines on phenylketonuria: diagnosis and treatment. Orphanet J Rare Dis. 2017;12(1):162.

Article   PubMed   PubMed Central   Google Scholar  

Vockley J, Andersson HC, Antshel KM, et al. Phenylalanine hydroxylase deficiency: diagnosis and management guideline. Genet Med. 2014;16(2):188–200.

Green B, Browne R, Firman S, et al. Nutritional and metabolic characteristics of UK adult phenylketonuria patients with varying dietary adherence. Nutrients. 2019;11(10):2459.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Kanufre V, Almeida MF, Barbosa CS, et al. Metabolic control of patients with phenylketonuria in a Portuguese metabolic centre comparing three different recommendations. Nutrients. 2021;13(9):3118.

Jurecki ER, Cederbaum S, Kopesky J, et al. Adherence to clinic recommendations among patients with phenylketonuria in the United States. Mol Genet Metab. 2017;120(3):190–7.

Krämer J, Baerwald C, Heimbold C, et al. Two years of pegvaliase in Germany: experiences and best practice recommendations. Mol Genet Metab. 2023;139(1):107564.

Article   PubMed   Google Scholar  

Longo N, Dimmock D, Levy H, et al. Evidence- and consensus-based recommendations for the use of pegvaliase in adults with phenylketonuria. Genet Med. 2019;21(8):1851–67.

Rocha JC, Bausell H, Belanger-Quintana A, et al. Development of a practical dietitian road map for the nutritional management of phenylketonuria (PKU) patients on pegvaliase. Mol Genet Metab Rep. 2021;28:100771.

CAS   PubMed   PubMed Central   Google Scholar  

Feldmann R, Schallert M, Nguyen T, Och U, Rutsch F, Weglage J. Children and adolescents with phenylketonuria display fluctuations in their blood phenylalanine levels. Acta Paediatr. 2019;108(3):541–3.

Pilotto A, Zipser CM, Leks E, et al. Phenylalanine effects on brain function in adult phenylketonuria. Neurology. 2021;96(3):e399–411.

Bilder DA, Noel JK, Baker ER, et al. Systematic review and meta-analysis of neuropsychiatric symptoms and executive functioning in adults with phenylketonuria. Dev Neuropsychol. 2016;41(4):245–60.

Romani C, Olson A, Aitkenhead L, et al. Meta-analyses of cognitive functions in early-treated adults with phenylketonuria. Neurosci Biobehav Rev. 2022;143:104925.

Trefz KF, Muntau AC, Kohlscheen KM, et al. Clinical burden of illness in patients with phenylketonuria (PKU) and associated comorbidities - a retrospective study of German health insurance claims data. Orphanet J Rare Dis. 2019;14(1):181.

Burton BK, Jones KB, Cederbaum S, et al. Prevalence of comorbid conditions among adult patients diagnosed with phenylketonuria. Mol Genet Metab. 2018;125(3):228–34.

Burton BK, Hermida Á, Bélanger-Quintana A, et al. Management of early treated adolescents and young adults with phenylketonuria: development of international consensus recommendations using a modified Delphi approach. Mol Genet Metab. 2022;137(1–2):114–26.

Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.

Campbell M, McKenzie JE, Sowden A, et al. Synthesis without meta-analysis (SWiM) in systematic reviews: reporting guideline. BMJ. 2020;368:l6890.

National Institutes of Health. MEDLINE: Overview. 2022. https://www.nlm.nih.gov/medline/medline_overview.html . Accessed 21 July 2023.

McKenzie JE, Brennan SE, et al. Chapter 12: Synthesizing and presenting findings using other methods. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. version 6.3. 2022. Available from: www.training.cochrane.org/handbook .

Google Scholar  

Charrière S, Maillot F, Bouée S, et al. Health status and comorbidities of adult patients with phenylketonuria (PKU) in France with a focus on early-diagnosed patients - a nationwide study of health insurance claims data. Mol Genet Metab. 2023;139(3):107625.

Bilder DA, Kobori JA, Cohen-Pfeffer JL, Johnson EM, Jurecki ER, Grant ML. Neuropsychiatric comorbidities in adults with phenylketonuria: a retrospective cohort study. Mol Genet Metab. 2017;121(1):1–8.

Choukair D, Kneppo C, Feneberg R, et al. Analysis of the functional muscle-bone unit of the forearm in patients with phenylketonuria by peripheral quantitative computed tomography. J Inherit Metab Dis. 2017;40(2):219–26.

Coakley KE, Douglas TD, Goodman M, Ramakrishnan U, Dobrowolski SF, Singh RH. Modeling correlates of low bone mineral density in patients with phenylalanine hydroxylase deficiency. J Inherit Metab Dis. 2016;39(3):363–72.

Demirdas S, van Spronsen FJ, Hollak CEM, et al. Micronutrients, essential fatty acids and bone health in phenylketonuria. Ann Nutr Metab. 2017;70(2):111–21.

Lage S, Bueno M, Andrade F, et al. Fatty acid profile in patients with phenylketonuria and its relationship with bone mineral density. J Inherit Metab Dis. 2010;33(Suppl 3):S363–71.

Lubout CMA, Arrieta Blanco F, Bartosiewicz K, et al. Bone mineral density is within normal range in most adult phenylketonuria patients. J Inherit Metab Dis. 2020;43(2):251–8.

Modan-Moses D, Vered I, Schwartz G, et al. Peak bone mass in patients with phenylketonuria. J Inherit Metab Dis. 2007;30(2):202–8.

Pérez-Dueñas B, Cambra FJ, Vilaseca MA, Lambruschini N, Campistol J, Camacho JA. New approach to osteopenia in phenylketonuric patients. Acta Paediatr. 2002;91(8):899–904.

Zeman J, Bayer M, Stepán J. Bone mineral density in patients with phenylketonuria. Acta Paediatr. 1999;88(12):1348–51.

Dios-Fuentes E, Gonzalo Marin M, Remón-Ruiz P, et al. Cardiometabolic and nutritional morbidities of a large, adult, PKU cohort from Andalusia. Nutrients. 2022;14(6):1311.

Rojas-Agurto E, Leal-Witt MJ, Arias C, Cabello JF, Bunout D, Cornejo V. Muscle and bone health in young Chilean adults with phenylketonuria and different degrees of compliance with the phenylalanine restricted diet. Nutrients. 2023;15(13):2939.

Carson DJ, Greeves LG, Sweeney LE, Crone MD. Osteopenia and phenylketonuria. Pediatr Radiol. 1990;20(8):598–9.

Greeves LG, Carson DJ, Magee A, Patterson CC. Fractures and phenylketonuria. Acta Paediatr. 1997;86(3):242–4.

Millet P, Vilaseca MA, Valls C, et al. Is deoxypyridinoline a good resorption marker to detect osteopenia in phenylketonuria? Clin Biochem. 2005;38(12):1127–32.

Nagasaka H, Tsukahara H, Takatani T, et al. Cross-sectional study of bone metabolism with nutrition in adult classical phenylketonuric patients diagnosed by neonatal screening. J Bone Miner Metab. 2011;29(6):737–43.

Stroup BM, Sawin EA, Murali SG, Binkley N, Hansen KE, Ney DM. Amino acid medical foods provide a high dietary acid load and increase urinary excretion of renal net acid, calcium, and magnesium compared with glycomacropeptide medical foods in phenylketonuria. J Nutr Metab. 2017;2017:1909101.

Porta F, Roato I, Mussa A, et al. Increased spontaneous osteoclastogenesis from peripheral blood mononuclear cells in phenylketonuria. J Inherit Metab Dis. 2008;31(Suppl 2):S339–42.

Roato I, Porta F, Mussa A, et al. Bone impairment in phenylketonuria is characterized by circulating osteoclast precursors and activated T cell increase. PLoS ONE. 2010;5(11):e14167.

de Groot MJ, Hoeksma M, van Rijn M, Slart RH, van Spronsen FJ. Relationships between lumbar bone mineral density and biochemical parameters in phenylketonuria patients. Mol Genet Metab. 2012;105(4):566–1277 70.

Schwahn B, Mokov E, Scheidhauer K, Lettgen B, Schönau E. Decreased trabecular bone mineral density in patients with phenylketonuria measured by peripheral quantitative computed tomography. Acta Paediatr. 1998;87(1):61–3.

Azabdaftari A, van der Giet M, Schuchardt M, Hennermann JB, Plöckinger U, Querfeld U. The cardiovascular phenotype of adult patients with phenylketonuria. Orphanet J Rare Dis. 2019;14(1):213.

Couce ML, Sánchez-Pintos P, Vitoria I, et al. Carbohydrate status in patients with phenylketonuria. Orphanet J Rare Dis. 2018;13(1):103.

Crujeiras V, Aldámiz-Echevarría L, Dalmau J, et al. Vitamin and mineral status in patients with hyperphenylalaninemia. Mol Genet Metab. 2015;115(4):145–50.

Hermida-Ameijeiras A, Crujeiras V, Roca I, Calvo C, Leis R, Couce ML. Arterial stiffness assessment in patients with phenylketonuria. Medicine (Baltimore). 2017;96(51):e9322.

Htun P, Nee J, Ploeckinger U, et al. Fish-free diet in patients with phenylketonuria is not associated with early atherosclerotic changes and enhanced platelet activation. PLoS ONE. 2015;10(8):e0135930.

Huemer M, Födinger M, Bodamer OA, et al. Total homocysteine, B-vitamins and genetic polymorphisms in patients with classical phenylketonuria. Mol Genet Metab. 2008;94(1):46–51.

Hvas AM, Nexo E, Nielsen JB. Vitamin B12 and vitamin B6 supplementation is needed among adults with phenylketonuria (PKU). J Inherit Metab Dis. 2006;29(1):47–53.

Mazzola PN, Nalin T, Castro K, et al. Analysis of body composition and nutritional status in Brazilian phenylketonuria patients. Mol Genet Metab Rep. 2016;6(NA):16–20.

Mezzomo TR, Messaggi Gomes Dias MR, Pereira RM. Adults with early diagnosis of phenylketonuria have higher resting energy expenditure than adults with late diagnosis. Clin Nutr ESPEN. 2023;56:166–72.

Ozel HG, Ahring K, Belanger-Quintana A, et al. Overweight and obesity in PKU: the results from 8 centres in Europe and Turkey. Mol Genet Metab Rep. 2014;1(NA):483–6.

Robertson LV, McStravick N, Ripley S, et al. Body mass index in adult patients with diet-treated phenylketonuria. J Hum Nutr Diet. 2013;26(Suppl 1):1–6.

Robinson M, White FJ, Cleary MA, Wraith E, Lam WK, Walter JH. Increased risk of vitamin B12 deficiency in patients with phenylketonuria on an unrestricted or relaxed diet. J Pediatr. 2000;136(4):545–7.

Rocha JC, van Spronsen FJ, Almeida MF, et al. Dietary treatment in phenylketonuria does not lead to increased risk of obesity or metabolic syndrome. Mol Genet Metab. 2012;107(4):659–63.

Schoen MS, Ramakrishnan U, Alvarez JA, Ziegler TR, Cui X, Singh RH. Characterization of choline nutriture among adults and children with phenylketonuria. Nutrients. 2022;14(19):4056.

Sumanszki C, Kiss E, Simon E, et al. The association of therapy adherence and thyroid function in adult patients with phenylketonuria. Ann Nutr Metab. 2019;75(1):16–23.

Vugteveen I, Hoeksma M, Monsen AL, et al. Serum vitamin B12 concentrations within reference values do not exclude functional vitamin B12 deficiency in PKU patients of various ages. Mol Genet Metab. 2011;102(1):13–7.

Weigel C, Kiener C, Meier N, et al. Carnitine status in early-treated children, adolescents and young adults with phenylketonuria on low phenylalanine diets. Ann Nutr Metab. 2008;53(2):91–5.

Guest JF, Bai JJ, Taylor RR, Sladkevicius E, Lee PJ, Lachmann RH. Costs and outcomes over 36 years of patients with phenylketonuria who do and do not remain on a phenylalanine-restricted diet. J Intellect Disabil Res. 2013;57(6):567–79.

Koch R, Burton B, Hoganson G, et al. Phenylketonuria in adulthood: a collaborative study. J Inherit Metab Dis. 2002;25(5):333–46.

Schulz B, Bremer HJ. Nutrient intake and food consumption of adolescents and young adults with phenylketonuria. Acta Paediatr. 1995;84(7):743–8.

Adamczyk P, Morawiec-Knysak A, Płudowski P, Banaszak B, Karpe J, Pluskiewicz W. Bone metabolism and the muscle-bone relationship in children, adolescents and young adults with phenylketonuria. J Bone Miner Metab. 2011;29(2):236–44.

Pena MJ, Pinto A, de Almeida MF, et al. Continuous use of glycomacropeptide in the nutritional management of patients with phenylketonuria: a clinical perspective. Orphanet J Rare Dis. 2021;16(1):84.

Pinto A, Almeida MF, Ramos PC, et al. Nutritional status in patients with phenylketonuria using glycomacropeptide as their major protein source. Eur J Clin Nutr. 2017;71(10):1230–4.

MacDonald A, Lilburn M, Cochrane B, et al. A new, low-volume protein substitute for teenagers and adults with phenylketonuria. J Inherit Metab Dis. 2004;27(2):127–35.

Das AM, Goedecke K, Meyer U, et al. Dietary habits and metabolic control in adolescents and young adults with phenylketonuria: self-imposed protein restriction may be harmful. JIMD Rep. 2014;13:149–58.

Mirás A, Bóveda MD, Leis MR, et al. Risk factors for developing mineral bone disease in phenylketonuric patients. Mol Genet Metab. 2013;108(3):149–54.

Procházková D, Jarkovský J, Vinohradská H, Konečná P, Machačová L, Doležel Z. Controlled diet in phenylketonuria and hyperphenylalaninemia may cause serum selenium deficiency in adult patients: the Czech experience. Biol Trace Elem Res. 2013;154(2):178–84.

Pessoa ALS, Martins AM, Ribeiro EM, et al. Burden of phenylketonuria in Latin American patients: a systematic review and meta-analysis of observational studies. Orphanet J Rare Dis. 2022;17(1):302.

Rodrigues C, Pinto A, Faria A, et al. Is the phenylalanine-restricted diet a risk factor for overweight or obesity in patients with phenylketonuria (PKU)? A systematic review and meta-analysis. Nutrients. 2021;13(10):3443.

Demirdas S, Coakley KE, Bisschop PH, Hollak CE, Bosch AM, Singh RH. Bone health in phenylketonuria: a systematic review and meta-analysis. Orphanet J Rare Dis. 2015;10:17.

Ney DM, Etzel MR. Designing medical foods for inherited metabolic disorders: why intact protein is superior to amino acids. Curr Opin Biotechnol. 2017;44:39–45.

Rovelli V, Ercoli V, Dionigi AR, et al. Low bone mineralization in phenylketonuria may be due to undiagnosed metabolic acidosis. Mol Genet Metab Rep. 2023;36:100998.

Lyon P, Strippoli V, Fang B, Cimmino L. B vitamins and one-carbon metabolism: implications in human health and disease. Nutrients. 2020;12(9):2867.

Stølen LH, Lilje R, Jørgensen JV, Bliksrud YT, Almaas R. High dietary folic acid and high plasma folate in children and adults with phenylketonuria. JIMD Rep. 2014;13:83–90.

Hochuli M, Bollhalder S, Thierer C, Refardt J, Gerber P, Baumgartner MR. Effects of inadequate amino acid mixture intake on nutrient supply of adult patients with phenylketonuria. Ann Nutr Metab. 2017;71(3–4):129–35.

Tankeu AT, Pavlidou DC, Superti-Furga A, Gariani K, Tran C. Overweight and obesity in adult patients with phenylketonuria: a systematic review. Orphanet J Rare Dis. 2023;18(1):37.

Blau N, Hennermann JB, Langenbeck U, Lichter-Konecki U. Diagnosis, classification, and genetics of phenylketonuria and tetrahydrobiopterin (BH4) deficiencies. Mol Genet Metab. 2011;104(Suppl):S2–9.

Ashe K, Kelso W, Farrand S, et al. Psychiatric and cognitive aspects of phenylketonuria: the limitations of diet and promise of new treatments. Front Psychiatry. 2019;10:561.

Ford S, O’Driscoll M, MacDonald A. Living with phenylketonuria: lessons from the PKU community. Mol Genet Metab Rep. 2018;17:57–63.

Cazzorla C, Bensi G, Biasucci G, et al. Living with phenylketonuria in adulthood: the PKU ATTITUDE study. Mol Genet Metab Rep. 2018;16:39–45.

Poole G, Pinto A, Evans S, et al. Hungry for change: the experiences of people with PKU, and their caregivers, when eating out. Nutrients. 2022;14(3):626.

Bosch AM, Burlina A, Cunningham A, et al. Assessment of the impact of phenylketonuria and its treatment on quality of life of patients and parents from seven European countries. Orphanet J Rare Dis. 2015;10:80.

Maissen-Abgottspon S, Muri R, Hochuli M, et al. Health-related quality of life in a european sample of adults with early-treated classical PKU. Orphanet J Rare Dis. 2023;18(1):300.

Olofsson S, Gralen K, Hoxer C, Okhuoya P, Persson U. The impact on quality of life of diet restrictions and disease symptoms associated with phenylketonuria: a time trade-off and discrete choice experiment study. Eur J Health Econ. 2022;23(6):993–1005.

Evers RAF, van Wegberg AMJ, MacDonald A, Huijbregts SCJ, Leuzzi V, van Spronsen FJ. Dietary liberalization in tetrahydrobiopterin-treated PKU patients: does it improve outcomes? Nutrients. 2022;14(18):3874.

Douglas TD, Ramakrishnan U, Kable JA, Singh RH. Longitudinal quality of life analysis in a phenylketonuria cohort provided sapropterin dihydrochloride. Health Qual Life Outcomes. 2013;11:218.

Qubty L, Hicks-Roof K. How might enhanced interprofessional collaboration between primary care physicians and registered dietitian nutritionists impact clinical outcomes related to obesity and associated illnesses? A commentary. Public Health Nutr. 2022;26:1–4. https://doi.org/10.1017/S1368980022002518 .

Jones G, Macaninch E, Mellor D, et al. Putting nutrition education on the table: development of a curriculum to meet future doctors’ needs. BMJ Nutr Prev Health. 2022;5(2):208–16.

Download references

Acknowledgements

Technical support, including screening and data extraction, was provided by Sofie Norregaard, MSc and Mamuda Aminu, MD, PhD; medical writing support, including assisting authors with the development of the outline and initial draft and incorporation of comments was provided by Alison Blackburn, PhD; and editorial support, including fact checking, referencing, figure preparation, formatting, proofreading, and submission was provided by Michelle Seddon, Dip Psychol, Rashmi Bharali, MSc, MBA, and Penelope Cervelo Bouzo Mres, of Prime Access (a division of Prime, Knutsford, UK), and Philippine Sauzey, MSc, of Prime Patient (a division of Prime, Knutsford, UK), supported by BioMarin Pharmaceutical Inc., Novato, CA, according to Good Publication Practice guidelines ( Link ).

This work was supported by BioMarin Pharmaceutical Inc., Novato, CA. The Sponsor was involved in the study design, collection, analysis, and interpretation of data, as well as data checking of information provided in the manuscript. However, ultimate responsibility for opinions, conclusions, and data interpretation lies with the authors.

Author information

Authors and affiliations.

BioMarin Pharmaceutical Inc., 105 Digital Drive, Novato, CA, 94949, USA

Kaleigh B. Whitehall, Sarah Rose & Gillian E. Clague

Clinic for PKU, Copenhagen University Hospital, Copenhagen, Denmark

Kirsten K. Ahring

Huntsman Mental Health Institute, University of Utah, Salt Lake City, UT, USA

Deborah A. Bilder

Oregon Health & Science University, Portland, OR, USA

Cary O. Harding

University of Santiago de Compostela, Santiago de Compostela, Spain

Álvaro Hermida

Queensland Lifespan Metabolic Medicine Service, Queensland Children’s Hospital, South Brisbane, QLD, Australia

Anita Inwood

University of Utah School of Medicine, Salt Lake City, UT, USA

Nicola Longo

Internal Medicine Department and Reference Center for Inherited Metabolic Disease, and the University of Tours, Tours, France

François Maillot

University Children’s Hospital, University Medical Center Hamburg-Eppendorf, Hamburg, Germany

Ania C. Muntau

Federal University of Ceará - UFC, Fortaleza, Ceará, Brazil

André L. S. Pessoa

Hospital Infantil Albert Sabin, Fortaleza, Ceará, Brazil

NOVA Medical School/Faculdade de Ciências Médicas (NMS/FCM), Universidade NOVA de Lisboa, Lisbon, Portugal

Júlio C. Rocha

Reference Centre of Inherited Metabolic Diseases (RC-IMD), Centro Hospitalar Universitário de Lisboa Central, Lisbon, Portugal

Nutrition & Metabolism, CINTESIS, NOVA Medical School, Faculdade de Ciências Médicas, Universidade NOVA de Lisboa, Lisbon, Portugal

Met Ed Consultants, Boulder, CO, USA

Hacettepe University, Ankara, Turkey

Serap Sivri

Prime, Knutsford, UK

Jack Said & Gillian C. Sibbring

Prime, London, UK

Sheun Oshinbolu

You can also search for this author in PubMed   Google Scholar

Contributions

All authors (KBW, SR, GEC, KKA, DAB, COH, AH, AI, NL, FM, ACM, ALSP, JCR, RF, SS JS, SO, GCS) contributed to the concept and design of the systematic literature review. KBW, SR, GEC, JS, and GCS defined the research question and the scope of the analyses. JS and GCS developed the study protocol with input from all authors on protocol revisions. JS conducted the original literature search, and contributed to article screening, data extraction, and narrative data synthesis. GCS conducted the literature search update and article screening for the update. SO contributed to article screening, data extraction, and narrative data synthesis, and conducted the vote-counting analysis. All authors had access to the data. The study was project-managed by GCS and the manuscript development was project-managed by SO and GCS. SO and GCS drafted the data synthesis sections of the manuscript and all authors contributed to data interpretation and participated in critically reviewing and/or revising the manuscript. All authors read and approved the final manuscript.

Authors’ information

COH is co-chair of the National PKU Alliance (USA) Scientific Advisory Board and the Principal Investigator of the Phenylalanine Families and Researchers Examining Evidence (PHEFREE) Consortium, a member of the National Institutes of Health (NIH) funded Rare Disorders Consortium Research Network (RDCRN).

Corresponding author

Correspondence to Kaleigh B. Whitehall .

Ethics declarations

Ethics approval and consent to participate.

Not applicable.

Consent for publication

Competing interests.

KBW, SR, and GEC are employees and stockholders of BioMarin. KKA has received consulting payments from Arla Foods Ingredients, BioMarin, Homology, and Nutricia. DAB has received consulting payments from BioMarin, Encoded Therapeutics, Synlogic Therapeutics, and Taysha Gene Therapies, and travel support from BioMarin. COH has received consulting and speaker fees/payments from BioMarin and has participated as a clinical trial investigator for BioMarin. AH has received consulting payments from Amicus Therapeutics, BioMarin, Chiesi, Genzyme, Shire, and Ultragenyx; speaker fees/payments from Alexion, Amicus Therapeutics, BioMarin, Genzyme, InMedica, Nutricia, Sobi, Takeda, and Vitaflo; travel support from Amicus Therapeutics, BioMarin, Chiesi, Genzyme, Inmedica, Sobi, and Vitaflo; and has participated as a clinical trial investigator for Ultragenyx. AI has received consulting payments for and travel support to advisory boards from BioMarin. NL has received consulting payments for advisory boards from Alnylam, Amicus Therapeutics, Audentes/Astellas, BioMarin, BridgeBio/CoA Therapeutics, Chiesi/Protalix, Genzyme/Sanofi, HemoShear Therapeutics, Horizon Pharma, Jaguar Gene Therapy, Jnana Therapeutics, Leadiant Biosciences, Moderna, Nestlé Pharma, PTC Therapeutics, Recordati, Reneo, Takeda, and Ultragenyx; has received other consultancy payments from Synlogic and travel support from BioMarin; has participated as a clinical trial investigator for Aeglea, Amicus Therapeutics, Audentes/Astellas, AVROBIO, BioMarin, Chiesi/Protalix, Genzyme/Sanofi, HemoShear Therapeutics, Homology, Horizon Pharma, Moderna, Nestlé Pharma, Pfizer, PTC Therapeutics, Reneo, Synlogic, Takeda, Travere Therapeutics, and Ultragenyx; and has been Data Safety and Monitoring Chair for ACI Clinical. FM has received consulting payments from PTC Therapeutics and travel support from BioMarin. ACM has participated as a clinical trial investigator for Nutricia; has received consulting payments from Atheneum, Nestlé, and PTC Therapeutics; speaker fees/payments from AIM, Applied Pharma Research, and Nutricia; and travel support from Nutricia. ALSP has received speaker fees/payments from BioMarin. JCR has received consulting payments from Applied Pharma Research, BioMarin, Merck Serono, Nutricia, PTC Therapeutics, and Synlogic, and speaker fees/payments from Applied Pharma Research, BioMarin, Cambrooke, LifeDiet, Merck Serono, Nutricia, PIAM, and Vitaflo, as well as travel support from Applied Pharma Research, BioMarin, Glutamine, Merck Serono, PIAM, and research grants from BioMarin. FR is a managing partner of Met Ed who has received educational grants from BioMarin. SS has received speaker fees/payments from BioMarin and Sanofi. GCS is an employee of Prime Access. JS and SO were employees of Prime Access at the time the study was undertaken. Prime Access (a division of Prime, Knutsford, UK) is a company sponsored by BioMarin to conduct this study and prepare the manuscript.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: table s1..

PubMed® search string.

Additional file 2: Figure S1.

Distribution of studies by study design.

Additional file 3: Figure S2.

Distribution of studies by geographic location.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Whitehall, K.B., Rose, S., Clague, G.E. et al. Systematic literature review of the somatic comorbidities experienced by adults with phenylketonuria. Orphanet J Rare Dis 19 , 293 (2024). https://doi.org/10.1186/s13023-024-03203-z

Download citation

Received : 25 April 2023

Accepted : 05 May 2024

Published : 12 August 2024

DOI : https://doi.org/10.1186/s13023-024-03203-z

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Comorbidity
• Disease burden
• Systematic review
• Nutritional status

Orphanet Journal of Rare Diseases

ISSN: 1750-1172

• Submission enquiries: Access here and click Contact Us
• General enquiries: [email protected]

SYSTEMATIC REVIEW article

Efficacy of probiotics/synbiotics supplementation in patients with chronic kidney disease: a systematic review and meta-analysis of randomized controlled trials.

• Department of Nephrology, Institute of Kidney Diseases, West China Hospital of Sichuan University, Chengdu, China

Background: Chronic kidney disease (CKD) is a serious and steadily growing health problem worldwide. Probiotic and synbiotic supplementation are expected to improve kidney function in CKD patients by altering imbalanced intestinal flora, regulating microbiota metabolites, modulating the brain-gut axis, and reducing inflammation.

Objectives: Our aim is to report the latest and largest pooled analyses and evidence updates to explore whether probiotic and synbiotic have beneficial effects on renal function and general conditions in patients with CKD.

Methods: We conducted a systematic literature search using PubMed, Embase, Web of Science, and the Cochrane Central Register of Controlled Trials from inception until 1 December 2023. Eligible literatures were screened according to inclusion and exclusion criteria, data were extracted, and a systematic review and meta-analysis was performed. Measurements included renal function-related markers, inflammatory markers, uremic toxins, lipid metabolism-related markers and electrolytes levels.

Results: Twenty-one studies were included. The results showed that probiotic/synbiotic significantly reduced blood urea nitrogen (BUN) (standardized mean difference (SMD), −0.23, 95% confidence interval (CI) −0.41, −0.04; p  = 0.02, I 2  = 10%) and lowered c-reactive protein level (CRP) (SMD: −0.34; 95% CI: −0.62, −0.07; p  = 0.01, I 2  = 37%) in CKD patients, compared with the control group.

Conclusion: In summary, probiotic/synbiotic supplementation seems to be effective in improving renal function indices and inflammation indices in CKD patients. Subgroup analyses suggested that longer-term supplementation is more favorable for CKD patients, but there is a high degree of heterogeneity in the results of partial subgroup analyses. The efficacy of probiotic/synbiotic in treating CKD needs to be supported by more evidence from large-scale clinical studies.

Systematic review registration: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42024526836 , Unique identifier: CRD42024526836.

1 Introduction

CKD is characterized by abnormalities in kidney structure or function that persist for a period of at least 3 months ( 1 ). The clinical features of CKD include decreased renal function and/or increased urinary albumin excretion (proteinuria) ( 2 ). It is a disease that progresses slowly over time, and a significant number of patients eventually reach end-stage renal disease and require dialysis for treatment ( 3 , 4 ). Currently, the global prevalence of CKD is about 11%, and with increasing age, the prevalence in people over 70 years old is as high as 34% ( 5 , 6 ). In addition, patients with CKD have an increased risk of cardiovascular disease, hypertension, diabetes, and infections ( 7 ). Although CKD has received considerable attention from scientists and clinicians, the care and treatment of the condition are still not up to par ( 8 ). Consequently, there is a pressing need to explore new drugs or therapeutic approaches.

Previous studies have shown that ecosystem imbalances are strongly associated with a number of chronic diseases, such as chronic kidney disease, diabetes and cardiovascular disease ( 9 , 10 ). There is also evidence showed that imbalances in the gut microbiota may contribute to CKD ( 11 ). In addition, worsening CKD can further exacerbate imbalances in gut flora ( 12 ). Researchers proposed the gut-kidney axis and the CKD-colon axis in 2011 ( 13 ) and 2015 ( 14 ) to characterize the interactions between the kidney and the gut. Probiotics, synbiotic supplements are then expected to slow the progression of CKD by regulating the balance of intestinal flora ( 15 ). Probiotics, consisting of active microorganisms, colonize the human intestinal tract to improve the microbiological balance and benefit human health ( 16 ). Recent studies indicated that probiotics could potentially offer advantages to individuals with chronic kidney disease ( 17 – 19 ). The current definition of synbiotic has been updated to “a mixture of live microorganisms and substrates selectively utilized by host microorganisms to provide health benefits to the host” ( 20 ). There have been a number of studies aimed at evaluating the role of synbiotic supplementation in patients with CKD ( 21 – 24 ). Currently, non-food probiotics and synbiotic supplements are becoming increasingly available in the United States ( 25 ).

The use of probiotics has shown potential as a nutritional strategy for the prevention and/or treatment of CKD. In some animal studies, Lactobacillus supplementation has been shown to slow the progression of chronic kidney disease and delay renal failure by altering short-chain fatty acid and nicotinamide metabolism ( 26 ). In addition, an exploratory clinical study found that serum levels of tumor necrosis factor-α (TNF-α), Interleukin (IL)-6, IL-18, and endotoxin were significantly reduced in patients with CKD after probiotics administration ( 27 ). Despite growing interest in the potential role of probiotics in improving chronic kidney disease, there is a lack of extensive cross-sectional studies to comprehensively assess the effect of probiotics/synbiotics on the general condition of CKD patients in the population. In addition, although there have been previous meta-analyses of the relationship between probiotics and CKD, the outcome indicators of these analyses have focused on one of many metrics, such as kidney function or metabolism ( 28 , 29 ). Therefore, a systematic review and meta-analysis incorporated latest RCT studies was designed to comprehensively investigate the effects of probiotics/synbiotics supplementation on renal function, lipid metabolism, inflammation, uremic toxin levels and electrolyte levels in dialysis/non-dialysis CKD patients.

2.1 Search strategy

The review program was established by two investigators (LC) and (WW) prior to the start of the study and registered with the PROSPERO International Prospective Registry of Systematic Reviews (registration number CRD42024526836). This study was conducted according to the Cochrane Manual and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) ( 30 ). Two independent reviewers (LC and YLT) searched PubMed, Embase, Web of Science and Cochrane Library from inception until December 2023. We searched the databases using the following terms: “probiotics,” “probiotic,” “synbiotics,” “synbiotic,” “renal insufficiency, chronic,” “chronic renal insufficiencies,” “renal insufficiencies, chronic,” “chronic renal insufficiency,” “kidney insufficiency, chronic,” “chronic kidney insufficiencies,” “kidney insufficiencies, chronic,” “chronic kidney diseases,” “chronic kidney disease” “disease, chronic kidney,” “diseases, chronic kidney,” “kidney disease, chronic,” “kidney diseases, chronic,” “chronic renal diseases,” “chronic renal disease,” “disease, chronic renal,” “diseases, chronic renal,” “renal disease, chronic,” “renal diseases, chronic” and “chronic kidney insufficiency.” Two researchers independently searched and evaluated the included studies, and any disagreements in the literature search were resolved by conferring with a third researcher (WW). Specific search strategies are shown in Supplementary Table 1 .

2.2 Eligible criteria

The study met all of the following criteria (1) study design: randomized controlled study; (2) study participants: patients with a confirmed diagnosis of chronic kidney disease; (3) intervention: the intervention group should receive any dose of probiotic or synbiotic supplementation; (4) comparison regimen/control group: participants in the control group may receive a placebo or other medication and if other medications are used in the treatment group, they also control group must be used in the same way; (6) language: articles published in English.

Studies were excluded for the following reasons: (1) they were reviews, meta-analyses, case reports, conference abstracts, and guidelines; (2) the study was animal-based; (3) the study was published in a language other than English.

2.3 Research screening

After excluding duplicate records, two researchers independently screened the titles and abstracts of all identified records to remove irrelevant documents. A full-text review was then conducted to determine eligibility for inclusion. Any disagreements regarding study selection could be resolved through discussion with a third researcher (LC, YLT, and WW). The study selection process is shown in Figure 1 .

Figure 1 . Flowchart of the employed literature search.

2.4 Data extraction

The following data were extracted from the included studies: (a) The basic information, including first author, publication year, region, data source, study design, and enrollment period; (b) Characteristics of the participants, including sample size, sex ratio (male), median age, median Body Mass Index (BMI) and hemodialysis time; (c) Interventions: probiotics/synbiotics types, dosage, frequency, intervention time; (d) Disease-related indicators: creatinine, BUN, eGFR, hemoglobin, uric acid, potassium, total cholesterol, HDL-cholesterol, LDL-cholesterol, indoxyl sulfate, p-cresyl sulfate, indole-3-acetic acid, CRP, IL-6, triglycerides, blood sodium, blood calcium and blood phosphorus. When continuous variables in the study were reported as median with range or interquartile range, we calculated the mean ± standard deviation through the validated mathematical method. When data were missing or not reported in the study, we contacted the corresponding authors to obtain completed data if available.

2.5 Quality assessment

Quality assessment of eligible RCTs was performed according to the Cochrane Handbook for Systematic Reviews of Interventions 5.1.0, based on seven terms: randomized sequence generation, allocation concealment, participant and personnel blinding, blinding of outcome assessment, incomplete outcome data, selective reporting and other biases sources ( 31 ). Three outcomes were assessed for each study, including low risk, high risk, and unclear risk. Studies with more “low risk” of bias assessment were considered superior.

2.6 Statistical analysis

Evidence synthesis was performed in Review Manager 5.4 version (Cochrane Collaboration, Oxford, United Kingdom). The SMD was applied for the comparison of continuous variables. All metrics were reported with mean ± SD. The heterogeneity in studies was assessed through the inconsistency index (I 2 ). I 2  > 50% were considered as significant heterogeneity. A random-effect model was used to estimate the combined SMD when significant heterogeneity was detected (I 2  > 50%). Otherwise, the fixed-effect model was applied. In addition, we performed one-way sensitivity analyses to evaluate the effect of included studies on the combined results for outcomes with significant heterogeneity. Subgroup analyses were used to explore sources of heterogeneity. The subgroup analysis has not been conducted for indicators which were few included in literatures. Because the limited number of literatures may lead to a significant discrepancy between subgroups, which could impact the accuracy of the results. Publication bias was evaluated visually by creating funnel plots via Review Manager 5.4 version (Cochrane Collaboration, Oxford, United Kingdom), as well as by conducting Egger’s regression tests using Stata 15.0 version (Stata Corp, College Station, TX, United States) for outcomes with 5 or more included studies. p -value < 0.05 was considered as statistically significant publication bias.

3.1 Literature search

The initial search was completed on 1 December 2023. We have identified 310 potentially relevant publications from PubMed, 434 from Embase, 103 from The Cochrane library, 540 from Web of science and 2 from manual retrieval. Endnote was used to eliminate duplicate publications, resulting in 791 records for review. After excluding publications that did not meet the inclusion criteria, we included 21 studies for systematic review and meta-analysis. A flow diagram illustrating the exclusion of articles with specific reasons is shown in Figure 1 (PRISMA flowchart).

3.2 Study characteristics

We conducted a systematic review and meta-analysis of 869 patients with chronic kidney disease involved in 21 RCT studies ( 17 – 19 , 21 , 22 , 32 – 47 ). The sample size ranged between 11 and 80, and the mean age of the patients was recorded ranged from a minimum of 45 years old to a maximum of 76 years old. The majority of the patients were from Asia. The articles were studied from 2013 to 2023 ( Table 1 ).

Table 1 . Baseline characteristics of include studies and methodological assessment.

3.3 Risk of bias assessment

The risk of bias assessment is presented in Supplementary Figures S1 , S2 . Most of the included studies were considered to have a low or unclear risk of bias. Two papers had significant baseline imbalance ( 17 , 36 ). The main source of bias in two pieces of literatures was the failure to implement double blinding ( 21 , 44 ). Six literatures were at high risk because of poor completeness of data results ( 19 , 32 – 34 , 44 , 46 ).

3.4 Renal function parameters

3.4.1 change in blood urea nitrogen.

Twelve studies ( 17 , 19 , 32 , 33 , 36 , 37 , 40 – 42 , 45 – 47 ) including 527 patients (271 probiotics/synbiotics; 256 controls) were included in the evaluation of blood urea nitrogen (BUN). Pooled analysis showed that the reduction of BUN in patients treated with probiotics/synbiotics was significantly better than in the control group (SMD: −0.23; 95% CI: −0.41, −0.04; p  = 0.02; Figure 2A ). No evidence of significant heterogeneity (I 2  = 10%, p  = 0.34) and statistical (Egger’s test, p  = 0.452) or visual ( Figure 3A ) publication bias was detected.

Figure 2 . Forest plots of kidney function outcomes: (A) BUN, (B) serum creatinine, (C) uric acid, (D) eGFR.

Figure 3 . Funnel plots of (A) BUN, (B) serum creatinine, (C) uric acid, (D) eGFR, (E) CRP, (F) IL-6, (G) indoxyl sulfate (H) p-cresyl sulfate.

Based on a range of subgroup analyses, we did not observe an effect of probiotic/synbiotics supplementation on BUN in American patients (k = 3, SMD: −0.13, 95% CI: −0.59, 0.33, I 2  = 0%, p  = 0.57). However, we found a significant decrease in BUN following probiotic/synbiotics supplementation in Asian individuals (k = 8, SMD: −0.28, 95% CI: −0.48, −0.09, I 2  = 22%, p  = 0.004). In addition, probiotic/synbiotics supplementation significantly reduced BUN in the long-term treatment (≥3 months; k = 10, SMD: −0.23, 95% CI: −0.44, −0.01, I 2  = 21%, p  = 0.04), but not in the short term (<3 months; k = 3, SMD: −0.19, 95% CI: −0.60, 0.22, I 2  = 0%, p  = 0.35). In addition, probiotics/synbiotics did not change BUN level in the HD patients (k = 7, SMD: −0.15, 95% CI: −0.4, 0.1, I 2  = 0%, p  = 0.23). However, probiotics/synbiotics significantly decreased the BUN level in non-HD patients (k = 6, SMD: −0.31, 95% CI: −0.54, −0.07, I 2  = 45%, p  = 0.01). And probiotics/synbiotics did not change BUN level in patients ≥60 years (k = 4, SMD: 0.01, 95% CI: −0.35, 0.38, I 2  = 0%, p  = 0.95) and patients <6 0 y (k = 8, SMD: −0.21, 95% CI: −0.43, 0.00, I 2  = 0%, p  = 0.05; Table 2 ).

Table 2 . Subgroup analyses.

3.4.2 Change in serum creatinine

Eleven articles ( 17 , 19 , 21 , 22 , 32 , 33 , 40 , 42 , 45 – 47 ) were included in the analysis of serum creatinine levels involving 423 patients (218 probiotics/synbiotics; 205 controls). The evidence synthesis showed similar changes of serum creatinine in patients in the probiotic/synbiotics group and the placebo group (SMD: −0.24; 95% CI: −0.52, 0.04; p  = 0.09) without significant heterogeneity (I 2  = 47%, p  = 0.04; Figure 2B ). No publication bias was detected by the funnel plot ( Figure 3B ) or Egger’s test ( p  = 0.097).

Based on a series of subgroup analyses ( Table 2 ), we did not observe any changes in serum creatinine following probiotic/synbiotics supplementation in individuals on hemodialysis (k = 5, SMD: −0.06, 95% CI: −0.37, 0.26, I 2  = 0%, p  = 0.72) and non-hemodialysis individuals (k = 6, SMD: −0.36, 95% CI: −0.8, 0.09, I 2  = 66%, p  = 0.11). Based on geographical location, we observed no significant change of serum creatinine in countries located in America (k = 2; SMD: −0.03, 95% CI: −0.56, 0.5, I 2  = 61%, p  = 0.03), Asia (k = 6, SMD: −0.36, 95% CI: −0.76, 0.04, I 2  = 67%, p  = 0.07) and Europe (k = 2; SMD: 0.15, 95% CI: −0.53, 0.83, I 2  = 0%, p  = 0.67). In addition, probiotics/synbiotics did not change the serum creatinine in the short term (<3 months; k = 4, SMD: −0.05, 95% CI: −0.43, 0.33, I 2  = 0%, p  = 0.79) and the long term (≥3 months; k = 7, SMD: −0.32, 95% CI: −0.7, 0.06, I 2  = 63%, p  = 0.1) or in older patients ≥ 60 years (k = 3, SMD: −0.08, 95% CI: −0.51, 0.34, I 2  = 0%, p  = 0.7) and younger patients < 60 years (k = 7, SMD: −0.28, 95% CI: −0.64, 0.08, I 2  = 60%, p  = 0.13).

3.4.3 Change in uric acid

Six studies ( 17 , 19 , 32 , 37 , 42 , 46 ) including 242 patients (125 probiotics/synbiotics; 117 controls) were included in the analysis of uric acid. Pooled analysis showed similar levels of alteration of uric acid in the probiotic/synbiotics group and the control group (SMD: −0.21; 95% CI: −0.47, 0.04; p  = 0.11; Figure 2C ). No significant heterogeneity (I 2  = 0%, p  = 0.55) and no evidence of statistical (Egger’s test, p  = 0.799) or visual ( Figure 3C ) publication bias was observed.

3.4.4 Change in eGFR

Four articles ( 21 , 32 , 39 , 41 ) reported data on eGFR levels between the two groups of 147 cases (77 probiotics/synbiotics; 70 controls). Evidence synthesis observed similar changes in eGFR levels in patients with probiotics/synbiotics and placebo (SMD: −0.16; 95% CI: −0.61, 0.29; p  = 0.48), with significant heterogeneity (I 2  = 42%, p  = 0.16; Figure 2D ). No evidence of visual publication bias was observed ( Figure 3D ).

3.5 Inflammation indicators and uremic toxins

3.5.1 change in c-reactive protein.

Ten studies ( 17 , 18 , 21 , 22 , 32 , 40 , 41 , 43 , 46 , 47 ) with a total of 356 patients (181 probiotics/synbiotics patients; 175 control patients) were included in the analysis of CRP. Pooled analysis showed that the probiotic/synbiotics group was significantly more effective in reducing CRP than the control group (SMD: −0.34; 95% CI: −0.62, −0.07; p  = 0.01; Figure 4A ). No significant heterogeneity was observed (I 2  = 37%, p  = 0.12) and no evidence of statistical (Egger’s test, p  = 0.288) or visual ( Figure 3E ) publication bias was observed.

Figure 4 . Forest plots of inflammation and uremic toxins outcomes: (A) CRP, (B) IL-6, (C) Indoxyl sulfate, (D) p-cresyl sulfate.

Subgroup analysis based on patient population suggested no changes in CRP following probiotics/synbiotics supplementation in individuals on hemodialysis (k = 5, SMD: −0.16, 95% CI: −0.49, 0.17, I 2  = 29%, p  = 0.34; Table 2 ). Furthermore, probiotics/synbiotics did not change CRP level in American individuals (k = 5, SMD: −0.33, 95% CI: −0.76, 0.11, I 2  = 65%, p  = 0.14), older individuals (k = 5, SMD: −0.36, 95% CI: −0.76,0.03, I 2  = 39%, p  = 0.07), younger individuals (k = 5, SMD: −0.31, 95% CI: −0.75, 0.12, I 2  = 48%, p  = 0.16) and individuals treated for a shorter period of time (k = 4, SMD: −0.02, 95% CI: −0.40, 0.36, I 2  = 0%, p  = 0.92). However, we found a significant decrease in CRP following probiotics/synbiotics supplementation in non-hemodialysis individuals (k = 5, SMD: −0.67, 95% CI: −1.02, −0.33, I 2  = 0%, p  = 0.0001), individuals treated for a longer period of time (k = 4, SMD: −0.51, 95% CI: −0.83, −0.19, I 2  = 33%, p  = 0.002) and European individuals (k = 3, SMD: −0.56, 95% CI: −1.05, −0.07, I 2  = 0%, p  = 0.03).

3.5.2 Change in IL-6

Five studies ( 17 , 32 , 33 , 37 , 41 ) were included in the analysis of IL-6, comprising a total of 159 patients (79 probiotics/synbiotics patients; 80 control patients). Pooled analysis suggested no statistically significant difference in the change of IL-6 levels between the two groups (SMD: −0.01; 95% CI: −0.36, 0.33; p  = 0.93; Figure 4B ). No significant heterogeneity was observed (I 2  = 17%, p  = 0.31). Nor was evidence of publication bias observed statistically Egger’s test, ( p  = 0.626) or visually ( Figure 3F ).

3.5.3 Change in indoxyl sulfate

Three hundred and seventeen patients from 8 studies ( 17 , 22 , 32 , 33 , 38 – 41 ) were included in the analysis of indoxyl sulfate (159 probiotics/synbiotics patients; 158 control patients). No statistically significant differences were found in the results of the pooled analysis between the two groups (SMD: 0.06; 95% CI: −0.16, 0.28; p  = 0.58; Figure 4C ). No significant heterogeneity was observed (I 2  = 0%, p  = 0.54) as well as evidence of statistical (Egger’s test, p  = 0.507) or visual ( Figure 3G ) publication bias.

Subgroup analysis based on patient population revealed no changes in indoxyl sulfate level after evaluation with either hemodialysis patients (k = 4, SMD: 0.03, 95% CI: −0.29, 0.34, I 2  = 0%, p  = 0.87) or non-hemodialysis patients (k = 4, SMD: 0.10, 95% CI: −0.21, 0.42, I 2  = 0%, p  = 0.53; Table 2 ). Furthermore, probiotics did not change the indoxyl sulfate level in the short term (<3 months; k = 3, SMD: 0.20, 95% CI: −0.18, 0.58, I 2  = 0%, p  = 0.30) and the long term (≥3 months; k = 5, SMD: −0.02, 95% CI: −0.33, 0.29, I 2  = 19%, p  = 0.90) or in older patients (≥60 years; k = 4, SMD: 0.11, 95% CI: −0.20, 0.41, I 2  = 10%, p  = 0.50) and younger patients (<60 years; k = 4, SMD: 0.02, 95% CI: −0.31, 0.34, I 2  = 0%, p  = 0.93). Besides, probiotics did not change the indoxyl sulfate level in Asia individuals (k = 4, SMD: 0.03, 95% CI: −0.36, 0.42, I 2  = 31%, p  = 0.86), America individuals (k = 2, SMD: 0.17, 95% CI: −0.37, 0.70, I 2  = 0%, p  = 0.54) and European individuals (k = 2, SMD: −0.32, 95% CI: −0.91, 0.27, I 2  = 0%, p  = 0.29).

3.5.4 Change in p-cresyl sulfate

A total of 211 patients from 6 studies ( 17 , 32 – 35 , 41 ) were included in the analysis of p-cresyl sulfate (109 probiotics/synbiotics patients; 102 control patients). No significant statistical differences were found in the results of the pooled analyses between the probiotics/synbiotics group and the control group (SMD: −0.22; 95% CI: −0.5, 0.06; p  = 0.12; Figure 4D ). No significant heterogeneity (I 2  = 5%, p  = 0.38) was observed as well as statistical (Egger’s test, p  = 0.947) or visual ( Figure 3H ) evidence of publication bias.

3.6 Lipid metabolism-related indicators

3.6.1 change in total cholesterol.

A total of six articles ( 17 , 32 , 39 , 43 , 46 , 47 ) involving 280 patients were included to analyze total cholesterol levels (142 probiotics/synbiotics patients; 138 control patients). Pooled analysis showed that the changes in total cholesterol levels were similar in the probiotics/synbiotics group and the control group (SMD: −0.16; 95% CI: −0.39, 0.08; p  = 0.19) with no significant heterogeneity (I 2  = 0%, p = 0.8; Figure 5A ). Neither the funnel plot ( Figure 6A ) nor the Egger test ( p  = 0.544) revealed publication bias.

Figure 5 . Forest plots of lipid metabolism evaluation outcomes: (A) total cholesterol, (B) HDL, (C) LDL, (D) triglyceride.

Figure 6 . Funnel plots of (A) total cholesterol, (B) HDL, (C) LDL, (D) triglyceride, (E) blood calcium, (F) blood potassium, (G) blood phosphorus.

3.6.2 Change in high density lipoprotein

Five studies ( 32 , 39 , 43 , 46 , 47 ) were included in the analysis of high density lipoprotein (HDL), involving 230 patients (117 probiotics/synbiotics patients; 113 control patients). No significant statistical differences were found in the results of the pooled analyses between the probiotics/synbiotics group and the control group (SMD: 0.28; 95% CI: −0.31, 0.87; p  = 0.35; Figure 5B ). Significant heterogeneity (I 2  = 78%, p  = 0.001) was observed. There was no statistical (Egger’s test, p  = 0.874) or visual ( Figure 6B ) evidence of publication bias.

3.6.3 Change in low density lipoprotein

Six articles ( 32 , 39 , 42 , 43 , 46 , 47 ) were included in the analysis for low density lipoprotein (LDL), involving 280 patients (146 probiotics/synbiotics group patients; 134 control patients). Evidence synthesis showed that probiotics/synbiotics group had a similar change in LDL level with the control group (SMD: −0.04; 95% CI: −0.28, 0.19; p  = 0.71) with no significant heterogeneity (I 2  = 0%, p  = 0.57; Figure 5C ). Both funnel plot ( Figure 6C ) and Egger’s test ( p  = 0.936) did not detect publication bias.

3.6.4 Change in triglyceride

Analysis of triglyceride levels in 308 patients (159 probiotic/synbiotics patients; 149 control patients) from six publications ( 17 , 39 , 42 , 43 , 46 , 47 ) showed that there was no statistically significant difference (SMD: −0.28; 95% CI: −0.6, 0.04; p  = 0.08) or significant heterogeneity (I 2  = 47%, p  = 0.09) in the change of triglyceride levels in the probiotics/synbiotics group compared with the control group ( Figure 5D ). No publication bias was found after both funnel plot ( Figure 6D ) and Egger’s test ( p  = 0.26) evaluation.

3.7 Electrolytes

3.7.1 change in blood calcium.

Six publications ( 19 , 21 , 34 , 37 , 40 , 46 ) involving 227 patients (114 probiotics/synbiotics patients; 113 control patients) were included in the analysis regarding blood calcium levels, and the results suggested that there were no statistically significant differences (SMD: 0.21; 95% CI: −0.05, 0.47; p  = 0.11) and no significant heterogeneity (I 2  = 0%, p  = 0.84) in the change of blood calcium levels in patients in the probiotics/synbiotics group compared with the control group ( Figure 7A ). It is worth noting that funnel plots ( Figure 6E ) and Egger’s test revealed significant publication bias ( p  = 0.049).

Figure 7 . Forest plots of electrolytic outcomes: (A) blood calcium, (B) blood potassium, (C) blood phosphorus.

3.7.2 Change in blood potassium

Two hundred and sixteen patients (110 probiotics/synbiotics patients; 106 control patients) originating from seven publications ( 21 , 32 – 34 , 40 , 41 , 46 ) were included in the analysis regarding blood potassium levels. And the results suggested that there were no statistically significant differences (SMD: −0.1; 95% CI: −0.38, 0.18; p  = 0.48) and no significant heterogeneity (I 2  = 8%, p  = 0.36) in the change of blood potassium levels of the patients in the probiotics/synbiotics group compared with the control group ( Figure 7B ). Funnel plot ( Figure 6F ) and Egger’s test did not reveal significant publication bias ( p  = 0.426).

3.7.3 Change in blood phosphorus

Nine publications ( 17 , 19 , 21 , 34 , 36 , 37 , 40 , 42 , 46 ) analyzed blood phosphorus levels in 411 patients (214 probiotics/synbiotics patients; 197 control patients). After comprehensive analysis, there was no statistically significant difference (SMD: −0.08; 95% CI: −0.28, 0.11; p  = 0.41) or heterogeneity (I 2  = 0%, p  = 0.96) in the change of blood phosphorus levels of the patients in the probiotics/synbiotics group compared with the control group ( Figure 7C ). Funnel plots ( Figure 6G ) and Egger’s test did not reveal significant publication bias ( p  = 0.503).

Based on an array of subgroup analyses, we did not observe an effect of probiotics/ synbiotics supplementation on blood phosphorus in hemodialysis individuals (k = 7, SMD: −0.05, 95% CI: −0.29, 0.2, I 2  = 0%, p  = 0.72) or non-hemodialysis individuals (k = 3, SMD: −0.14, 95% CI: −0.46, 0.18, I 2  = 0%, p  = 0.38). Based on geographical location, we observed no significant change of blood phosphorus in patients from countries located in Asia (k = 7; SMD: −0.08, 95% CI: −0.29, 0.14, I 2  = 0%, p = 0.48). In terms of treatment time, we observed no significant change of blood phosphorus in both long term (k = 6; SMD: −0.16, 95% CI: −0.39, 0.08, I 2  = 0%, p  = 0.19) and short term (k = 4, SMD: 0.09, 95% CI: −0.26, 0.44, I 2  = 0%, p  = 0.62). In addition, no significant changes in blood phosphorus were observed with probiotics/synbiotics supplementation in individuals above and below 60 years of age (≥60 years, k = 2, SMD: −0.04, 95% CI: −0.57, 0.48, I 2  = 7%, p  = 0.87), (<60 years, k = 7, SMD: −0.03, 95% CI: −0.27, 0.22, I 2  = 0%, p  = 0.84; Table 2 ).

3.8 Sensitivity analysis

Because the comprehensive analysis of HDL showed significant heterogeneity, we conducted one-way sensitivity analyses for comparison of HDL to evaluate the influence of each individual study on the combined SMD through removing the individual study one by one. Sensitivity analyses revealed that the new combined SMD remained constant after exclusion of any individual study for HDL ( Figure 8 ).

Figure 8 . Sensitivity analysis of HDL.

4 Discussion

4.1 findings from meta-analysis.

The gut microbiota consists of more than 100 trillion bacteria and plays an important role in normal body functions, particularly in immune and metabolic homeostasis. There is growing evidence that alterations in the gut microbiota can affect multiple organ systems and lead to many chronic diseases such as CKD ( 48 ). CKD is a serious and steadily growing health problem worldwide. As a progressive disease, the majority of CKD patients are referred to dialysis treatment, and effective pharmacological treatments are still being explored ( 49 ). One of the promising drug candidates is the modification of dysbiotic gut flora through probiotics/synbiotics supplementation to reduce levels of gut-derived uremic toxins and reduce chronic microinflammation thereby improving renal function ( 50 ). Due to the small number of published studies, it is still highly controversial whether this probiotics/synbiotics intervention affects renal function, uremic toxicity, and inflammation levels in patients with CKD. In this study, we systematically compiled and analyzed the clinical evidence of RCT on probiotics/synbiotics for the treatment of CKD to provide better guidance for clinical practice.

Our results showed that probiotic/synbiotics supplementation of CKD patients can decreased BUN in CKD patients, and no significant heterogeneity was found in the analysis results, which to a certain extent reflects the effect of probiotic/synbiotics to improve renal function in CKD patients. However, the use of probiotic/synbiotics had no effect on eGFR and serum creatinine, indicators of renal function. And the analysis results of eGFR and serum creatinine did not show apparent heterogeneity. This result may be due to the fact that eGFR and serum creatinine is strongly influenced by ethnicity, gender and age. And the baseline could not be standardized across studies. In addition to this, the use of probiotics/synbiotics can reduce CRP expression levels and improve inflammatory status in CKD patients. However, probiotic/synbiotic supplementation did not significantly alter blood electrolyte levels and lipid metabolism-related markers in CKD patients compared with placebo. No significant publication bias was detected by Egger’s test and funnel plot for all indicators except blood calcium.

We also stratified patients according to region, duration of treatment, age, and whether or not they were receiving dialysis treatment. Subgroups were analyzed for several important indicators: urea nitrogen, serum creatinine, CRP, indoxyl sulfate, and blood phosphorus. The results of subgroup analyses suggested that probiotics/synbiotics were more effective in reducing BUN in non-hemodialysis CKD patients. This probably due to the fact that CKD in hemodialysis patients usually has progressed to the stage of end-stage renal disease, and the renal units are irreversibly damaged, which is difficult to be improved by drug treatment ( 51 ). The results of subgroup analysis also showed that probiotics/synbiotics had a better effect on CRP reduction in non-hemodialysis CKD patients, suggesting that probiotics/synbiotics had a better effect on the improvement of inflammation in non-hemodialysis patients. In CKD patients, increased inflammation is associated with several negative clinical outcomes such as increased oxidative stress, vascular dysfunction and increased risk of cardiovascular disease, as the ameliorative effect of probiotics/synbiotics on inflammation in non-hemodialysis patients is beneficial to patients ( 52 , 53 ). And one of the goals of treatment for hemodialysis patients is to reduce inflammation, thus effectively improving the survival of these patients ( 54 ). Meanwhile, the results of the subgroup analysis stratified on the basis of the duration of probiotics/synbiotics treatment showed that there is a significant reduction in BUN and CRP levels in CKD patients when probiotics/synbiotics are applied for a longer period of time compared to the control group. This suggested that adherence to probiotics/synbiotics for a longer period of time is more favorable for CKD patients. Interestingly, when subgroup analyses were performed on a regional basis, we found that probiotics/synbiotics supplementation was more effective in reducing BUN in Asian patients and in reducing CRP in patients from the Europe. The number of studies that included patients from the Europe was small, so the related results need to be further verified.

4.2 Possible mechanisms

The relationship between probiotics/synbiotics and CKD has been recognized with the increasing understanding of the health effects of microbial balance on the host. Essentially, probiotics/synbiotics supplementation can modulate the imbalance of the gut microbiota for the biosynthesis of targeted compounds with bioactive properties in CKD patients ( 55 ). At the same time, probiotics improves the integrity of the intestinal epithelial barrier and reduces the production of uremic toxins to some extent ( 56 , 57 ). With the fluctuation of gut bacteria, probiotics can modulate inflammation by establishing a balance between pro-inflammatory and anti-inflammatory cytokines in the body ( 58 ). In addition, metabolites from the gut microbiota also play an important role in maintaining homeostasis in the gut for the benefit of host health through fermentation of amino acids and dietary fiber, production of vitamins and neurotransmitters, and modification of bile acids ( 59 ). For example, Zhu et al. ( 60 ) showed that short-chain fatty acids (SCFAs) from a variety of bacteria reduced the expression of genes for inflammatory cytokines, chemokines, and pro-fibrotic proteins in diabetic kidneys, which, in turn, reduced proteinuria, glomerular hypertrophy, pedunculated cell injury, and interstitial fibrosis in mice with acute kidney injury and CKD. Indeed, probiotic or synbiotics supplementation may also reverse the expansion of harmful gut microbes that produce excess uremic toxins and attenuate the development of CKD ( 32 , 61 ).

4.3 Strengths and limitations

4.3.1 limitations.

Firstly, most of the publications included in this meta-analysis were RCT cohort studies. The sample sizes of RCT studies are small, and potential bias from small samples is unavoidable. Secondly, the main population groups of the studies we included were from Asia, with fewer people from other states, and there may be regional selectivity bias. Whether the results can be generalized to other regions needs to be confirmed by further studies. Thirdly, the heterogeneity of the studies included in the analysis of HDL was large, which may hinder the robustness of the results. In addition, the RCTs involved in the study did not report adverse events in patients, so adverse events were not included in the study.

4.3.2 Strengths

Firstly, this study is the latest meta-analysis of probiotics/synbiotics for CKD with the largest sample size available. Secondly, the original studies included in this article were all RCTs, which were of high quality, with good study design and a balanced baseline. Thirdly, this study confirmed that probiotics/synbiotics had an ameliorative effect on renal function and inflammatory status in patients with CKD, which has been consistent with previous studies. Fourthly, compared with previous meta-analyses, our study included a wider range of outcome indicators and incorporated the most recent RCT studies, thus allowing for the most up-to-date evidence on probiotics/synbiotics supplementation in CKD treatment. What’s more, this study provides more options and guidance notes for clinical CKD treatment.

5 Conclusion

In conclusion, this is the latest systematic review and meta-analysis demonstrating that probiotic/synbiotics interventions reduced BUN and CRP in patients with CKD, although there was insufficient evidence of a positive effect of probiotics/synbiotics on lipids and blood electrolytes. Regarding BUN and CRP, the results of our meta-analysis emphasize the positive effects of probiotic/synbiotics supplementation using longer (≥3 months) treatment durations in Asian patients. This area deserves further research to elucidate the mechanism of probiotics/synbiotics for the possible treatment of CKD and to further assess the safety of different types of probiotics/synbiotics through randomized controlled trials.

Data availability statement

The original contributions presented in the study are included in the article/ Supplementary material , further inquiries can be directed to the corresponding author.

Author contributions

CL: Formal analysis, Writing – original draft. LY: Writing – review & editing. WW: Investigation, Writing – review & editing. PF: Funding acquisition, Supervision, Writing – review & editing.

The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. 1.3.5 project for disciplines of excellence from West China Hospital of Sichuan University (grant number ZYGD23015).

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnut.2024.1434613/full#supplementary-material

1. Eckardt, K-U, Coresh, J, Devuyst, O, Johnson, RJ, Köttgen, A, Levey, AS, et al. Evolving importance of kidney disease: from subspecialty to global health burden. Lancet . (2013) 382:158–69. doi: 10.1016/S0140-6736(13)60439-0

PubMed Abstract | Crossref Full Text | Google Scholar

2. Li, L, Astor, BC, Lewis, J, Hu, B, Appel, LJ, Lipkowitz, MS, et al. Longitudinal progression trajectory of GFR among patients with CKD. Am J Kidney Dis . (2012) 59:504–12. doi: 10.1053/j.ajkd.2011.12.009

3. Liu, J, Bankir, L, Verma, A, Waikar, SS, and Palsson, R. Association of the urine-to-plasma urea ratio with CKD progression. Am J Kidney Dis . (2023) 81:394–405. doi: 10.1053/j.ajkd.2022.09.010

4. Kula, AJ, Prince, DK, Flynn, JT, and Bansal, N. BP in young adults with CKD and associations with cardiovascular events and decline in kidney function. J Am Soc Nephrol . (2021) 32:1200–9. doi: 10.1681/ASN.2020081156

5. Mazidi, M, Kengne, AP, Siervo, M, and Kirwan, R. Association of dietary intakes and genetically determined serum concentrations of mono and poly unsaturated fatty acids on chronic kidney disease: insights from dietary analysis and Mendelian randomization. Nutrients . (2022) 14:1231. doi: 10.3390/nu14061231

6. Liu, X, Gao, W, Yang, J, Mao, G, Lu, H, and Xing, W. Association between probiotic, prebiotic, and yogurt consumption and chronic kidney disease: the NHANES 2010-2020. Front Nutr . (2022) 9:1058238. doi: 10.3389/fnut.2022.1058238

7. Zoccali, C, Vanholder, R, Massy, ZA, Ortiz, A, Sarafidis, P, Dekker, FW, et al. The systemic nature of CKD. Nat Rev Nephrol . (2017) 13:344–58. doi: 10.1038/nrneph.2017.52

Crossref Full Text | Google Scholar

8. Neuen, BL, Heerspink, HJL, Vart, P, Claggett, BL, Fletcher, RA, Arnott, C, et al. Estimated lifetime cardiovascular, kidney and mortality benefits of combination treatment with SGLT2 inhibitors, GLP-1 receptor agonists, and non-steroidal MRA compared with conventional care in patients with type 2 diabetes and albuminuria. Circulation . (2023) 149:450–62. doi: 10.1161/CIRCULATIONAHA.123.067584

9. Durack, J, and Lynch, SV. The gut microbiome: relationships with disease and opportunities for therapy. J Exp Med . (2019) 216:20–40. doi: 10.1084/jem.20180448

10. Martens, EC, Neumann, M, and Desai, MS. Interactions of commensal and pathogenic microorganisms with the intestinal mucosal barrier. Nat Rev Microbiol . (2018) 16:457–70. doi: 10.1038/s41579-018-0036-x

11. Krukowski, H, Valkenburg, S, Madella, A-M, Garssen, J, van Bergenhenegouwen, J, Overbeek, SA, et al. Gut microbiome studies in CKD: opportunities, pitfalls and therapeutic potential. Nat Rev Nephrol . (2023) 19:87–101. doi: 10.1038/s41581-022-00647-z

12. Glorieux, G, Nigam, SK, Vanholder, R, and Verbeke, F. Role of the microbiome in gut-heart-kidney cross talk. Circ Res . (2023) 132:1064–83. doi: 10.1161/CIRCRESAHA.123.321763

13. Meijers, BKI, and Evenepoel, P. The gut-kidney axis: indoxyl sulfate, p-cresyl sulfate and CKD progression. Nephrol Dial Transplant . (2011) 26:759–61. doi: 10.1093/ndt/gfq818

14. Pahl, MV, and Vaziri, ND. The chronic kidney disease—colonic axis. Semin Dial . (2015) 28:459–63. doi: 10.1111/sdi.12381

15. Meijers, B, Evenepoel, P, and Anders, H-J. Intestinal microbiome and fitness in kidney disease. Nat Rev Nephrol . (2019) 15:531–45. doi: 10.1038/s41581-019-0172-1

16. Li, T, Teng, D, Mao, R, Hao, Y, Wang, X, and Wang, J. A critical review of antibiotic resistance in probiotic bacteria. Food Res Int . (2020) 136:109571. doi: 10.1016/j.foodres.2020.109571

17. Lim, PS, Wang, HF, Lee, MC, Chiu, L-S, Wu, M-Y, Chang, W-C, et al. The efficacy of Lactobacillus-containing probiotic supplementation in hemodialysis patients: a randomized, double-blind, placebo-controlled trial. J Ren Nutr . (2021) 31:189–98. doi: 10.1053/j.jrn.2020.07.002

18. Shariaty, Z, Mahmoodi Shan, GR, Farajollahi, M, Amerian, M, and Behnam Pour, N. The effects of probiotic supplement on hemoglobin in chronic renal failure patients under hemodialysis: a randomized clinical trial. J Res Med Sci . (2017) 22:74. doi: 10.4103/jrms.JRMS_614_16

19. Lakshmi, P, Babu, S, Ahil, M, Raadhika, M, Susila, M, and Geetha, M. Effect of lobun probiotics in patients with chronic kidney disease stage III and IV, placebo-controlled open-labeled study. Br J Clin Pharmacol . (2020) 86:1198–9. doi: 10.1111/bcp.14266

20. Swanson, KS, Gibson, GR, Hutkins, R, Reimer, RA, Reid, G, Verbeke, K, et al. The international scientific Association for Probiotics and Prebiotics (ISAPP) consensus statement on the definition and scope of synbiotics. Nat Rev Gastroenterol Hepatol . (2020) 17:687–701. doi: 10.1038/s41575-020-0344-2

21. Cosola, C, Rocchetti, MT, Di, B, Acquaviva, PM, Maranzano, V, Corciulo, S, et al. An innovative Synbiotic formulation decreases free serum Indoxyl sulfate, small intestine permeability and ameliorates gastrointestinal symptoms in a randomized pilot trial in stage IIIb-IV CKD patients. Toxins . (2021) 13:334. doi: 10.3390/toxins13050334

22. Rocchetti, MT, Cosola, C, di Bari, I, Magnani, S, Galleggiante, V, Scandiffio, L, et al. Efficacy of Divinylbenzenic resin in removing Indoxyl sulfate and P-cresol sulfate in hemodialysis patients: results from an in vitro study and an in vivo pilot trial (xuanro4-nature 3.2). Toxins . (2020) 12:170. doi: 10.3390/toxins12030170

23. Nakabayashi, I, Nakamura, M, Kawakami, K, Ohta, T, Kato, I, Uchida, K, et al. Effects of synbiotic treatment on serum level of p-cresol in haemodialysis patients: a preliminary study. Nephrol Dial Transplant . (2011) 26:1094–8. doi: 10.1093/ndt/gfq624

24. Vacca, M, Celano, G, Calabrese, FM, Rocchetti, MT, Iacobellis, I, Serale, N, et al. In vivo evaluation of an innovative synbiotics on stage IIIb-IV chronic kidney disease patients. Front Nutr . (2023) 10:1215836. doi: 10.3389/fnut.2023.1215836

25. Sanders, ME, Shane, AL, and Merenstein, DJ. Advancing probiotic research in humans in the United States: challenges and strategies. Gut Microbes . (2016) 7:97–100. doi: 10.1080/19490976.2016.1138198

26. Barrows, IR, Ramezani, A, and Raj, DS. Gut feeling in AKI: the long arm of short-chain fatty acids: the long arm of short-chain fatty acids. J Am Soc Nephrol . (2015) 26:1755–7. doi: 10.1681/ASN.2014111157

27. De Araújo, ÉMR, Meneses, GC, Carioca, AAF, Martins, AMC, Daher, EDF, and Silva Junior, GB. Use of probiotics in patients with chronic kidney disease on hemodialysis: a randomized clinical trial. J Bras Nefrol . (2023) 45:152–61. doi: 10.1590/2175-8239-JBN-2022-0021en

28. Chen, L, Shi, J, Ma, X, Shi, D, and Qu, H. Effects of microbiota-driven therapy on circulating Indoxyl sulfate and P-Cresyl sulfate in patients with chronic kidney disease: a systematic review and Meta-analysis of randomized controlled trials. Adv Nutr . (2022) 13:1267–78. doi: 10.1093/advances/nmab149

29. Zheng, HJ, Guo, J, Wang, Q, Wang, L, Wang, Y, Zhang, F, et al. Probiotics, prebiotics, and synbiotics for the improvement of metabolic profiles in patients with chronic kidney disease: a systematic review and meta-analysis of randomized controlled trials. Crit Rev Food Sci Nutr . (2021) 61:577–98. doi: 10.1080/10408398.2020.1740645

30. Page, MJ, McKenzie, JE, Bossuyt, PM, Boutron, I, Hoffmann, TC, Mulrow, CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. J Clin Epidemiol . (2021) 134:178–89. doi: 10.1016/j.jclinepi.2021.03.001

31. Cumpston, M, Li, T, Page, MJ, Chandler, J, Welch, VA, Higgins, JPT, et al. Updated guidance for trusted systematic reviews: a new edition of the Cochrane handbook for systematic reviews of interventions. Cochrane Database Syst Rev . (2019) 10:ED000142. doi: 10.1002/14651858.ED000142

32. de Faria Barros, A, Borges, NA, Nakao, LS, Dolenga, CJ, Do Carmo, FL, De Carvalho Ferreira, D, et al. Effects of probiotic supplementation on inflammatory biomarkers and uremic toxins in non-dialysis chronic kidney patients: a double-blind, randomized, placebo-controlled trial. J Funct Foods . (2018) 46:378–83. doi: 10.1016/j.jff.2018.05.018

33. Borges, NA, Carmo, FL, Stockler-Pinto, MB, de Brito, JS, Dolenga, CJ, Ferreira, DC, et al. Probiotic supplementation in chronic kidney disease: a double-blind, randomized, placebo-controlled trial. J Ren Nutr . (2018) 28:28–36. doi: 10.1053/j.jrn.2017.06.010

34. Eidi, F, Poor-Reza Gholi, F, Ostadrahimi, A, Dalili, N, Samadian, F, and Barzegari, A. Effect of Lactobacillus Rhamnosus on serum uremic toxins (phenol and P-cresol) in hemodialysis patients: a double blind randomized clinical trial. Clin Nutr ESPEN . (2018) 28:158–64. doi: 10.1016/j.clnesp.2018.08.010

35. Guida, B, Germanò, R, Trio, R, Russo, D, Memoli, B, Grumetto, L, et al. Effect of short-term synbiotic treatment on plasma p-cresol levels in patients with chronic renal failure: a randomized clinical trial. Nutr Metab Cardiovasc Dis . (2014) 24:1043–9. doi: 10.1016/j.numecd.2014.04.007

36. Haghighat, N, Mohammadshahi, M, Shayanpour, S, and Haghighizadeh, MH. Effect of synbiotic and probiotic supplementation on serum levels of endothelial cell adhesion molecules in hemodialysis patients: a randomized control study. Probio Antimicrob Proteins . (2019) 11:1210–8. doi: 10.1007/s12602-018-9477-9

37. Laffin, MR, Tayebi Khosroshahi, H, Park, H, Laffin, LJ, Madsen, K, Kafil, HS, et al. Amylose resistant starch (HAM-RS2) supplementation increases the proportion of Faecalibacterium bacteria in end-stage renal disease patients: microbial analysis from a randomized placebo-controlled trial. Hemodial Int . (2019) 23:343–7. doi: 10.1111/hdi.12753

38. Aida, L, Tities Anggraeni, I, Aulia, R, and Murdani, A. The effects of synbiotics on indoxyl sulphate level, constipation, and quality of life associated with constipation in chronic haemodialysis patients: a randomized controlled trial. BMC Nephrol . (2022) 23:259. doi: 10.1186/s12882-022-02890-9

39. McFarlane, C, Krishnasamy, R, Stanton, T, Savill, E, Snelson, M, Mihala, G, et al. Synbiotics easing renal failure by improving gut microbiology II (SYNERGY II): a feasibility randomized controlled trial. Nutrients . (2021) 13:481. doi: 10.3390/nu13124481

40. Mirzaeian, S, Saraf-Bank, S., Entezari, MH, Hekmatdoost, A, Feizi, A, and Atapour, A. Effects of synbiotic supplementation on microbiota-derived protein-bound uremic toxins, systemic inflammation, and biochemical parameters in patients on hemodialysis: a double-blind, placebo-controlled, randomized clinical trial. Nutrition . (2020) 73:110713. doi: 10.1016/j.nut.2019.110713

41. Mitrović, M, Stanković-Popović, V, Tolinački, M, Golić, N, Soković Bajić, S, Veljović, K, et al. The impact of synbiotic treatment on the levels of gut-derived uremic toxins, inflammation, and gut microbiome of chronic kidney disease patients-a randomized trial. J Ren Nutr . (2023) 33:278–88. doi: 10.1053/j.jrn.2022.07.008

42. Saxena, A, Srinivasa, S, Veerappan, I, Jacob, C, Mahaldar, A, Gupta, A, et al. Enzobiotics—a novel therapy for the elimination of uremic toxins in patients with CKD (EETOX study): a multicenter double-blind randomized controlled trial. Nutrients . (2022) 14:3804. doi: 10.3390/nu14183804

43. Soleimani, A, Motamedzadeh, A, Zarrati Mojarrad, M, Bahmani, F, Amirani, E, Ostadmohammadi, V, et al. The effects of synbiotic supplementation on metabolic status in diabetic patients undergoing hemodialysis: a randomized, double-blinded, placebo-controlled trial. Probiotics Antimicrob Proteins . (2019) 11:1248–56. doi: 10.1007/s12602-018-9499-3

44. Lopes, RDCSO, Theodoro, JMV, Da Silva, BP, Queiroz, VAV, De Castro Moreira, ME, Mantovani, HC, et al. Synbiotic meal decreases uremic toxins in hemodialysis individuals: a placebo-controlled trial. Food Res Int . (2019) 116:241–8. doi: 10.1016/j.foodres.2018.08.024

45. Tayebi-Khosroshahi, H, Habibzadeh, A, Niknafs, B, Ghotaslou, R, Yeganeh Sefidan, F, Ghojazadeh, M, et al. The effect of lactulose supplementation on fecal microflora of patients with chronic kidney disease; a randomized clinical trial. J Renal Inj Prev . (2016) 5:162–7. doi: 10.15171/jrip.2016.34

46. Hevilla, F, Padial, M, Blanca, M, Barril, G, Jiménez-Salcedo, T, Ramirez-Ortiz, M, et al. Effect on nutritional status and biomarkers of inflammation and oxidation of an oral nutritional supplement (with or without probiotics) in malnourished hemodialysis patients. A multicenter randomized clinical trial “Renacare trial”. Front Nutr . (2023) 10:1107869. doi: 10.3389/fnut.2023.1107869

47. Mafi, ARCFB, Namazi, G, Soleimani, A, Bahmani, F, Aghadavod, E, and Asemi, Z. Metabolic and genetic response to probiotics supplementation in patients with diabetic nephropathy: A randomized, double-blind, placebo-controlled trial. Food Funct . (2018) 9:4763–70. doi: 10.1039/c8fo00888d

48. Relman, DA. The human microbiome and the future practice of medicine. JAMA . (2015) 314:1127–8. doi: 10.1001/jama.2015.10700

49. Canaud, B, Collins, A, and Maddux, F. The renal replacement therapy landscape in 2030: reducing the global cardiovascular burden in dialysis patients. Nephrol Dial Transplant . (2020) 35:ii51. doi: 10.1093/ndt/gfaa005

50. Huang, HW, and Chen, M-J. Exploring the preventive and therapeutic mechanisms of probiotics in chronic kidney disease through the gut–kidney axis. J Agric Food Chem . (2024) 72:8347–64. doi: 10.1021/acs.jafc.4c00263

51. Chawla, LS, and Kimmel, PL. Acute kidney injury and chronic kidney disease: an integrated clinical syndrome. Kidney Int . (2012) 82:516–24. doi: 10.1038/ki.2012.208

52. Stenvinkel, P, Chertow, GM, Devarajan, P, Levin, A, Andreoli, SP, Bangalore, S, et al. Chronic inflammation in chronic kidney disease progression: role of Nrf2. Kidney Int Rep . (2021) 6:1775–87. doi: 10.1016/j.ekir.2021.04.023

53. Petreski, T, Piko, N, Ekart, R, Hojs, R, and Bevc, S. Review on inflammation markers in chronic kidney disease. Biomedicines . (2021) 9:182. doi: 10.3390/biomedicines9020182

54. Lousa, I, Belo, L, Valente, MJ, Rocha, S, Preguiça, I, Rocha-Pereira, P, et al. Inflammatory biomarkers in staging of chronic kidney disease: elevated TNFR2 levels accompanies renal function decline. Inflamm Res . (2022) 71:591–602. doi: 10.1007/s00011-022-01574-2

55. Mitrea, L, Medeleanu, M, Pop, CR, Rotar, AM, and Vodnar, DC. Biotics (pre-, pro-, post-) and uremic toxicity: implications, mechanisms, and possible therapies. Toxins . (2023) 15:548. doi: 10.3390/toxins15090548

56. Beker, BM, Colombo, I, Gonzalez-Torres, H, and Musso, CG. Decreasing microbiota-derived uremic toxins to improve CKD outcomes. Clin Kidney J . (2022) 15:2214–9. doi: 10.1093/ckj/sfac154

57. Favero, C, Giordano, L, Mihaila, SM, Masereeuw, R, Ortiz, A, and Sanchez-Niño, MD. Postbiotics and kidney disease. Toxins . (2022) 14:623. doi: 10.3390/toxins14090623

58. Wang, J, Chen, W-D, and Wang, Y-D. The relationship between gut microbiota and inflammatory diseases: the role of macrophages. Front Microbiol . (2020) 11:1065. doi: 10.3389/fmicb.2020.01065

59. Feng, W, Ao, H, Peng, C, and Yan, D. Gut microbiota, a new frontier to understand traditional Chinese medicines. Pharmacol Res . (2019) 142:176–91. doi: 10.1016/j.phrs.2019.02.024

60. Zhu, H, Cao, C, Wu, Z, Zhang, H, Sun, Z, Wang, M, et al. The probiotic L. casei Zhang slows the progression of acute and chronic kidney disease. Cell Metab . (2021) 33:1926–1942.e8. doi: 10.1016/j.cmet.2021.06.014

61. Wang, X, Yang, S, Li, S, Zhao, L, Hao, Y, Qin, J, et al. Aberrant gut microbiota alters host metabolome and impacts renal failure in humans and rodents. Gut . (2020) 69:2131–42. doi: 10.1136/gutjnl-2019-319766

Keywords: probiotic, synbiotic, chronic kidney disease, renal function, metabolism

Citation: Liu C, Yang L, Wei W and Fu P (2024) Efficacy of probiotics/synbiotics supplementation in patients with chronic kidney disease: a systematic review and meta-analysis of randomized controlled trials. Front. Nutr . 11:1434613. doi: 10.3389/fnut.2024.1434613

Received: 18 May 2024; Accepted: 17 July 2024; Published: 06 August 2024.

Reviewed by:

Copyright © 2024 Liu, Yang, Wei and Fu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Ping Fu, [email protected]

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Health & Environmental Research Online (HERO)

• Learn about HERO
• Search HERO
• Projects in HERO
• Risk Assessment
• Transparency & Integrity

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings
• My Bibliography
• Collections
• Citation manager

Save citation to file

Email citation, add to collections.

• Create a new collection
• Add to an existing collection

Add to My Bibliography

Your saved search, create a file for external citation management software, your rss feed.

• Search in PubMed
• Search in NLM Catalog
• Add to Search

Measuring the population burden of chronic kidney disease: a systematic literature review of the estimated prevalence of impaired kidney function

Affiliation.

• 1 NHS Grampian, Aberdeen, UK.
• PMID: 21965592
• DOI: 10.1093/ndt/gfr547

Background: Internationally, there have been substantial efforts to improve the early identification of chronic kidney disease (CKD), with a view to improving survival, reducing progression and minimizing cardiovascular morbidity and mortality. In 2002, a new and globally adopted definition of CKD was introduced. The burden of kidney function impairment in the population is unclear and widely ranging prevalence estimates have been reported.

Methods: We conducted a systematic literature review, searching databases to June 2009. We included all adult population screening studies and studies based on laboratory or clinical datasets where the denominator was clear. Studies reporting prevalence estimates based on at least one eGFR <60 mL/min/1.73m(2) or elevated creatinine above a stated threshold were included. Study design and quality were explored as potential factors leading to heterogeneity.

Results: We identified 43 eligible studies (57 published reports) for inclusion. Substantial heterogeneity was observed with estimated prevalence (0.6-42.6%). The included studies demonstrated significant variation in methodology and quality that impacted on the comparability of their findings. From the higher quality studies, the six studies measuring impaired kidney function (iKF) using estimated glomerular filtration rate in community screening samples reported a prevalence ranging from 1.7% in a Chinese study to 8.1% in a US study, with four reporting an estimated prevalence of 3.2-5.6%. Heterogeneity was driven by the measure used, study design and study population.

Conclusion: In the general population, estimated iKF, particularly eGFR 30-59 mL/min/1.73m(2) was common with prevalence similar to diabetes mellitus. Appropriate care of patients poses a substantial global health care challenge.

PubMed Disclaimer

Similar articles

• Incidence of impaired kidney function among people with HIV: a systematic review and meta-analysis. Shi R, Chen X, Lin H, Ding Y, He N. Shi R, et al. BMC Nephrol. 2022 Mar 17;23(1):107. doi: 10.1186/s12882-022-02721-x. BMC Nephrol. 2022. PMID: 35300612 Free PMC article.
• Development of reduced kidney function in rheumatoid arthritis. Hickson LJ, Crowson CS, Gabriel SE, McCarthy JT, Matteson EL. Hickson LJ, et al. Am J Kidney Dis. 2014 Feb;63(2):206-13. doi: 10.1053/j.ajkd.2013.08.010. Epub 2013 Oct 4. Am J Kidney Dis. 2014. PMID: 24100126 Free PMC article.
• Early referral strategies for management of people with markers of renal disease: a systematic review of the evidence of clinical effectiveness, cost-effectiveness and economic analysis. Black C, Sharma P, Scotland G, McCullough K, McGurn D, Robertson L, Fluck N, MacLeod A, McNamee P, Prescott G, Smith C. Black C, et al. Health Technol Assess. 2010 Apr;14(21):1-184. doi: 10.3310/hta14210. Health Technol Assess. 2010. PMID: 20441712 Review.
• The practical implications of using standardized estimation equations in calculating the prevalence of chronic kidney disease. Quinn MP, Rainey A, Cairns KJ, Marshall AH, Savage G, Kee F, Peter Maxwell A, Reaney E, Fogarty DG. Quinn MP, et al. Nephrol Dial Transplant. 2008 Feb;23(2):542-8. doi: 10.1093/ndt/gfm599. Epub 2007 Sep 22. Nephrol Dial Transplant. 2008. PMID: 17890743
• Methodology used in studies reporting chronic kidney disease prevalence: a systematic literature review. Brück K, Jager KJ, Dounousi E, Kainz A, Nitsch D, Ärnlöv J, Rothenbacher D, Browne G, Capuano V, Ferraro PM, Ferrieres J, Gambaro G, Guessous I, Hallan S, Kastarinen M, Navis G, Gonzalez AO, Palmieri L, Romundstad S, Spoto B, Stengel B, Tomson C, Tripepi G, Völzke H, Wiȩcek A, Gansevoort R, Schöttker B, Wanner C, Vinhas J, Zoccali C, Van Biesen W, Stel VS; European CKD Burden Consortium. Brück K, et al. Nephrol Dial Transplant. 2015 Aug;30 Suppl 4(Suppl 4):iv6-16. doi: 10.1093/ndt/gfv131. Nephrol Dial Transplant. 2015. PMID: 26209739 Free PMC article. Review.
• Evaluation of the safety of retroperitoneal laparoscopic partial nephrectomy by investigating the perioperative indicators. Zhang X, Xing T. Zhang X, et al. Front Oncol. 2023 Apr 19;13:1138210. doi: 10.3389/fonc.2023.1138210. eCollection 2023. Front Oncol. 2023. Retraction in: Front Oncol. 2024 Feb 19;14:1383340. doi: 10.3389/fonc.2024.1383340. PMID: 37152056 Free PMC article. Retracted.
• The association between liver fibrosis scores and chronic kidney disease. Xiong S, Wang P, Yin S, Deng W, Zhao Y, Li W, Li Z, Zhou Y, Yu S, Yang H, Guo X, Sun Y. Xiong S, et al. Front Med (Lausanne). 2023 Jan 30;10:1046825. doi: 10.3389/fmed.2023.1046825. eCollection 2023. Front Med (Lausanne). 2023. PMID: 36793875 Free PMC article.
• Prevalence and incidence of chronic kidney disease stage 3-5 - results from KidDiCo. Kampmann JD, Heaf JG, Mogensen CB, Mickley H, Wolff DL, Brandt F. Kampmann JD, et al. BMC Nephrol. 2023 Jan 19;24(1):17. doi: 10.1186/s12882-023-03056-x. BMC Nephrol. 2023. PMID: 36658506 Free PMC article.
• Are We Moving the Needle for Patients with TP53 -Mutated Acute Myeloid Leukemia? Shallis RM, Bewersdorf JP, Stahl MF, Halene S, Zeidan AM. Shallis RM, et al. Cancers (Basel). 2022 May 14;14(10):2434. doi: 10.3390/cancers14102434. Cancers (Basel). 2022. PMID: 35626039 Free PMC article. Review.
• Accelerated epigenetic aging and inflammatory/immunological profile (ipAGE) in patients with chronic kidney disease. Yusipov I, Kondakova E, Kalyakulina A, Krivonosov M, Lobanova N, Bacalini MG, Franceschi C, Vedunova M, Ivanchenko M. Yusipov I, et al. Geroscience. 2022 Apr;44(2):817-834. doi: 10.1007/s11357-022-00540-4. Epub 2022 Mar 2. Geroscience. 2022. PMID: 35237926 Free PMC article.

Publication types

• Search in MeSH

Related information

• PubChem Compound
• PubChem Substance

Grants and funding

• HSRU1/CSO_/Chief Scientist Office/United Kingdom

LinkOut - more resources

Full text sources.

• Ovid Technologies, Inc.
• Silverchair Information Systems
• Genetic Alliance
• MedlinePlus Health Information

Research Materials

• NCI CPTC Antibody Characterization Program

Miscellaneous

• NCI CPTAC Assay Portal

• Citation Manager

NCBI Literature Resources

MeSH PMC Bookshelf Disclaimer

The PubMed wordmark and PubMed logo are registered trademarks of the U.S. Department of Health and Human Services (HHS). Unauthorized use of these marks is strictly prohibited.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• PMC10227138

Predict, diagnose, and treat chronic kidney disease with machine learning: a systematic literature review

Francesco sanmarchi.

1 Department of Biomedical and Neuromotor Science, Alma Mater Studiorum, University of Bologna, Via San Giacomo 12, 40126 Bologna, Italy

Claudio Fanconi

2 Department of Medicine (Biomedical Informatics), Stanford University, School of Medicine, Stanford, CA USA

3 Department of Electrical Engineering and Information Technology, ETH Zurich, Zurich, Switzerland

Davide Golinelli

Davide gori, tina hernandez-boussard, angelo capodici, associated data.

Data that support the findings of this study are available upon reasonable request from the corresponding author, AC.

In this systematic review we aimed at assessing how artificial intelligence (AI), including machine learning (ML) techniques have been deployed to predict, diagnose, and treat chronic kidney disease (CKD). We systematically reviewed the available evidence on these innovative techniques to improve CKD diagnosis and patient management.

We included English language studies retrieved from PubMed. The review is therefore to be classified as a “rapid review”, since it includes one database only, and has language restrictions; the novelty and importance of the issue make missing relevant papers unlikely. We extracted 16 variables, including: main aim, studied population, data source, sample size, problem type (regression, classification), predictors used, and performance metrics. We followed the Preferred Reporting Items for Systematic Reviews (PRISMA) approach; all main steps were done in duplicate.

From a total of 648 studies initially retrieved, 68 articles met the inclusion criteria.

Models, as reported by authors, performed well, but the reported metrics were not homogeneous across articles and therefore direct comparison was not feasible. The most common aim was prediction of prognosis, followed by diagnosis of CKD. Algorithm generalizability, and testing on diverse populations was rarely taken into account. Furthermore, the clinical evaluation and validation of the models/algorithms was perused; only a fraction of the included studies, 6 out of 68, were performed in a clinical context.

Conclusions

Machine learning is a promising tool for the prediction of risk, diagnosis, and therapy management for CKD patients. Nonetheless, future work is needed to address the interpretability, generalizability, and fairness of the models to ensure the safe application of such technologies in routine clinical practice.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s40620-023-01573-4.

Introduction

Chronic Kidney Disease (CKD) is a state of progressive loss of kidney function ultimately resulting in the need for renal replacement therapy (dialysis or transplantation) [ 1 ]. It is defined as the presence of kidney damage or an estimated glomerular filtration rate less than 60 ml/min per 1.73 m 2 , persisting for 3 months or more [ 2 ]. CKD prevalence is growing worldwide, along with demographic and epidemiological transitions [ 3 ]. The implications of this disease are enormous for our society in terms of quality of life and the overall sustainability of national health systems. Worldwide, CKD accounted for 2,968,600 (1%) disability-adjusted life-years and 2,546,700 (1% to 3%) life-years lost in 2012 [ 4 ]. Therefore, it is of the utmost importance to assess how to promptly and adequately diagnose and treat patients with CKD.

The causes of CKD vary globally. The most common primary diseases causing CKD and ultimately kidney failure are diabetes mellitus, hypertension, and primary glomerulonephritis, representing 70–90% of the total primary causes [ 1 , 2 , 4 ]. Although these three causes are at the top of the CKD etiology charts, other features are involved in CKD pathophysiology (e.g., pollution, infections and autoimmune diseases) [ 5 – 9 ]. Similarly, there are numerous factors that play a role in CKD progression, namely non-modifiable risk factors (e.g., age, gender, ethnicity) and modifiable ones (e.g., systolic and diastolic blood pressure, proteinuria) [ 1 , 2 , 4 – 9 ].

Given how dauntingly vast the number of factors that can play a significant role in the etiology and progression of CKD is, it can be difficult to correctly assess the individual risk of CKD and its progression. Naturally, as with any complex problem, humans seek simplification, and therefore the question shifts to what to take into account when assessing CKD risk. Thanks to new methodological techniques, we now have the ability to improve our diagnostic and predictive capabilities.

Artificial Intelligence (AI) is the capacity of human-built machines to manifest complex decision-making or data analysis in a similar or augmented fashion in comparison to human intelligence [ 10 ]. Machine Learning (ML) is the collection of algorithms that empower models to learn from data, and therefore to undertake complex tasks through complex calculations [ 11 – 15 ]. In recent years AI and ML have offered enticing solutions to clinical problems, such as how to perform a diagnosis from sparse and seemingly contrasting data, or how to predict a prognosis [ 16 ]. Given the enormous potential of ML, and its capacity to learn from data, researchers have tried to apply its capacities to resolve complex problems, such as predicting CKD diagnosis and prognosis, and managing its treatment.

In this complex scenario, we aimed to systematically review the published studies that applied machine learning in the diagnosis and prediction, prognosis, and treatment of CKD patients. In doing so, the primary objective is to describe how ML models and variables have been used to predict, diagnose and treat CKD, as well as what results have been achieved in this field.

Search strategy and selection criteria

We conducted a systematic literature review, following the Preferred Reporting Items for Systematic Reviews (PRISMA) approach [ 17 ], including studies that applied ML algorithms to CKD forecasting, diagnosis, prognosis, and treatment. This systematic review’s outcomes of interest are machine learning models, features used, performances and uses regarding diagnosis, prognosis and treatment of CKD. The review itself and its protocol were not registered.

The initial search was implemented on October 20, 2021. The search query consisted of terms considered pertinent by the authors.

We searched for publications on PubMed using the following search string: “((artificial intelligence[Title/Abstract]) OR (machine learning[Title/Abstract]) OR (computational*[Title/Abstract]) OR (deep learning[Title/Abstract])) AND ((ckd) OR (chronic kidney disease) OR (chronic kidney injury) OR (chronic kidney) OR (chronic renal) OR (end stage renal) OR (end stage kidney) OR (ESKD) OR (ESRD) OR (CKJ) OR (CKI) OR (((renal) OR (kidney)) AND (failure)))” .

We included articles for review if they were in vivo studies (human-based), which applied AI & ML techniques in order to assess the diagnosis, prognosis, or therapy of CKD patients and reported original data. We did not limit our inclusion criteria to any specific study design, nor to any outcome of interest, as our main goal was to be as inclusive as possible, and we wanted to capture all available evidence from any study design and any outcome of interest.

We excluded studies that were not in English, those focusing on animals, reviews, systematic reviews, opinions, editorials, and case reports. We decided to exclude in vitro studies (conducted on cellular substrates) and studies focusing on animals, in order to summarize the current evidence on the application of ML models on humans.

Data extraction

Data were extracted by two independent reviewers (AC and FS). Disagreement on extracted data was discussed with an independent arbiter (DGol).

The following data were extracted from each included article (main text and/or supplementary material): author(s) name, date of publication, first author affiliation (country and region), main study objective, objective category (risk, diagnosis, prognosis, and treatment), prognosis category, study population, data source, sample size, problem type (regression, classification), machine learning algorithms examined in the study, predictor categories, number of predictors used, predictor list, performance metrics, final conclusions, use in clinical context and the 5 most important model features. When more than one model was considered in the study, the one the authors deemed best was extracted. Performance metrics always refer to the models’ performance on test sets.

Quality and risk assessment

Evaluation of the included studies was performed using both PROBAST [ 18 ] and the Guidelines for developing and reporting machine learning predictive models in biomedical research developed by Luo and colleagues [ 19 ].

Included studies

Of the 648 articles retrieved from PubMed, 421 were ruled out after title screening, and 140 were excluded after abstract screening; a total of 87 articles were selected for full-text screening (Fig.  1 ). Of these 87 studies, 68 were included in the final set of articles (Table ​ (Table1) 1 ) [ 20 – 87 ].

PRISMA flow-chart

Extracts of the main findings

Most of the included articles ( n  = 51) were published from 2019 to 2021. Among the 68 articles selected for data extraction, the majority were published by authors from organizations based in Asia ( n  = 33; 48.5%). The remaining articles were published by authors from Europe ( n  = 17; 25%), North America ( n  = 12; 17.6%), Africa ( n  = 5; 7.35%) and South America ( n  = 1; 1.47%). The analyzed studies were classified as observational.

A total of 28 studies focused on the use of ML algorithms in disease prognosis analysis, 21 investigated the use of ML techniques on diagnosis (4 evaluated both), 12 evaluated the risk of developing the disease, and 3 investigated the use of ML in CKD treatment. Among the articles focusing on prognosis, the majority studied the application of ML in evaluating CKD progression ( n  = 13) and mortality ( n  = 8).

Study populations and sample size

The most commonly investigated study population consisted of patients with CKD and healthy subjects ( n  = 26; 38.2%), followed by patients with CKD only ( n  = 16; 23.5%) and patients with CKD treated with hemodialysis ( n  = 12; 17.6%). The sample size investigated in the selected articles varied from a minimum of 30 individuals to a maximum of 550,000 (median = 776; IQR 400–12,020).

Data sources

The majority of the included articles analyzed data obtained from single-hospital registries ( n  = 33; 48.5%), datasets provided by universities ( n  = 15; 22.1%), and datasets collected in multi-center studies ( n  = 12, 17.6%). Five studies analyzed health insurance data (7.35%) and 3 studies used data provided by national health services (4.41%).

The most commonly used data were various combinations of demographic data along with individual clinical characteristics and laboratory data ( n  = 60; 82.24%), followed by data obtained by medical imaging technologies ( n  = 5; 7.35%) and genomic data ( n  = 3; 4.41%).

The number of models tested and reported in each article varied from a minimum of 1 model to a maximum of 10 (mean = 3). The most frequently tested model class was tree algorithms ( n  = 58, 33.53%), such as random forest ( n  = 27, 15.61%), decision trees ( n  = 10, 5.78%) and extreme gradient boosting ( n  = 9, 5.20). Subsequently, neural networks (NNs) were often inspected ( n  = 44, 16.18%), especially the multilayer perceptron (MLP) ( n  = 28, 16.18%). Another popular choice of machine learning model class was Support Vector Machines ( n  = 25, 14.45%) and logistic regression ( n  = 18, 10.45%) with various regularizations. Another popular method that we did not classify into a larger model class was the non-parametric k-Nearest Neighbors algorithm ( n  = 8, 2.31%). The complete list of models can be found in Table ​ Table2 2 .

List of machine learning models used in the selected papers

The models were also classified in larger model families to present a general overview. Some models that we were not able to classify in larger model families were classified as “Others”

All the articles implemented supervised learning algorithms, 57 (83.8%) of them addressed classification tasks and 11 (16.2%) regression tasks.

The majority of the included articles ( n  = 52) specified the total number of features used to train the models. These models used a highly variable number of features, ranging from 4 to 6624 (median = 24; IQR = 17—46). Of the 68 included studies, 55 specified the variables used in the models ( n  = 130). The most frequently used features are reported in Fig.  2 .

Occurrence of variables in the selected articles, divided per aim

Performance metrics

The most common performance metrics were accuracy ( n  = 30, 17.05%) and the area under the receiver operating characteristic curve (often also referred to as ROC-AUC, AUROC, AUC, or C-statistic) ( n  = 30, 17.05%). Subsequently, other classification metrics, such as sensitivity ( n  = 29, 16.48%), specificity ( n  = 24, 13.64%), precision ( n  = 16, 9.09%), and F1-score ( n  = 14, 7.95%) were often used to compare the machine learning models. Note that all the aforementioned metrics, except ROC AUC, were used for classification and required establishing a risk threshold as a decision boundary. ROC AUC conversely did not require setting a decision threshold as it was calculated by iterating over all the decision thresholds. In terms of regression, the most used metrics for comparison were mean absolute error ( n  = 6, 3.41%) and root mean squared error ( n  = 5, 2.84%). The full list of the metrics and how often they occurred can be found in Table ​ Table3 3 .

List of metrics and their occurrence in number and percentages of the selected papers

Furthermore, the last column specifies for which task the metric is used

Best performing models, and their performances

In the included articles, neural networks were the models that commonly performed best ( n  = 28, 41.18%) compared to the median performance of other models, such as MLP ( n  = 18, 26.47%) and convolutional neural networks ( n  = 7, 24.53%). Tree-based algorithms performed best ( n  = 24, 35.29%); these algorithms included Random Forest ( n  = 16, 23.53%) and Extreme Gradient Boosting ( n  = 5, 7.35%). The results for Support Vector Machines ( n  = 5, 7.35%) were also noteworthy. A complete list of the best performing models in the selected papers can be found in Table ​ Table4 4 .

List of the best performing models throughout the selected papers, classified by model family

In terms of performance, we compared the metrics of prediction models, diagnostic models and risk prediction models separately. Of the 25 (36.76%) machine learning models for diagnosis, 19 papers reported accuracy. Three models reported the highest accuracy of 1.00 while the lowest reported accuracy is 0.80 (mean = 0.95, median = 0.98). Sensitivity was reported 15 times, with a maximum of 1.00, a minimum of 0.56, a mean of 0.95 and a median of 0.99. In addition, specificity was reported in 13 cases (max = 1.00, min = 0.79, mean = 0.96, median = 0.99). The ROC-AUC was reported in 6 papers (max = 0.99, min = 0.91, mean = 0.941, median = 0.94).

For the prediction models ( n  = 32, 47.06%), 15 papers reported the ROC-AUC with a maximum of 0.96 and a minimum of 0.69 (mean = 0.82, median = 0.82). Ten papers reported accuracy, ranging from 0.54 to 0.99, with a mean of 0.85 and a median of 0.87. Sensitivity was reported 8 times, ranging from 0.54 to 0.93 (mean = 0.765, median = 0.76), and specificity was reported 5 times (max = 0.99, min = 0.78, mean = 0.917, median = 0.96).

Next, the risk prediction models ( n  = 12, 17.65%) showed ROC-AUC 9 times (max = 0.96, min = 0.76, mean = 0.864, median = 0.86) and accuracy 4 times (max = 0.99, min = 0.82, mean = 0.901, median = 0.91).

Finally, 3 (4.41%) papers focused on therapy, one of which reported an accuracy of 0.95, while the other two focused on outcome differences ( p -values).

Most common variables and most important ones

The total number of variables used in the included studies was 813. The five most common ones were: Blood Pressure ( n  = 62, 7.63%), Age ( n  = 45, 5.54%), Hemoglobin ( n  = 37, 4.55%), Creatinine (serum) ( n  = 31, 3.81%) and Sex ( n  = 31, 3.81%).

Nonetheless, to better capture how variables were used in the selected papers, we classified the variables into 4 subsets (CKD Prognosis, CKD Diagnosis, Risk of Developing CKD, CKD Treatment) based on the primary aim the authors stated their model would have attempted to achieve.

Regarding CKD Prognosis, 342 variables were used out of 813 total (42%). The most common ones were: Blood Pressure ( n  = 24, 7%), Age ( n  = 19, 5,56%), Cholesterol (serum) ( n  = 18, 5.26%), Sex ( n  = 14, 4%) and Hemoglobin (blood) ( n  = 13, 3.8%), with the most important variables being: Age, Hemoglobin and Proteinuria.

Concerning CKD Diagnosis, 311 variables were used out of 813 total (38.25%). The most common ones were: Blood Pressure ( n  = 22, 7%), Hemoglobin (blood) ( n  = 19, 6.1%), Pus Cell General—used to indicate the number of dead white cells in urine—( n  = 18, 5.79%), Age ( n  = 14, 4.50%) and Glucose (serum) ( n  = 14, 4.50%). The most important variables in this case were Albumin, Creatinine, and Hemoglobin.

With regard to Risk of Developing CKD, 137 variables were used out of 813 total (16.85%). The most common ones were: Blood Pressure ( n  = 12, 8.75%), Age ( n  = 9, 6.57%), Sex ( n  = 7, 5.11%), History of Cardiovascular Disease ( n  = 6, 4.38%) and estimated Glomerular Filtration Rate (eGFR) ( n  = 6, 4.38%). The most important variables were Age, GFR and Blood Pressure.

Finally, regarding CKD Treatment, 23 variables were used out of 813 total (2.83%). The most common ones were: Blood Iron ( n  = 5, 21.74%), Hemoglobin ( n  = 3, 13%), Drugs Used ( n  = 2, 8.70%), MCV ( n  = 2, 8.70%) and White Blood Cells (blood) ( n  = 2, 8.70%). Regarding this aim, no weights were listed in the examined articles.

The complete spreadsheet with all variables and percentages can be found in Supplemental Material, together with the most important variables, divided per aim.

Other than using PROBAST to assess risk of bias, we also assessed fairness based on how the authors explicitly used variables. In some studies, variables were not fully listed, and in such cases, if the variable (sex, or race/ethnicity) was not indexed, we considered the feature as not included in the general model.

Out of 68 studies, 43 included gender in the model and 12 included race/ethnicity. When Non-Hispanic Whites were part of the assessed cohort, they were the majority group, ranging from 87 to 31%. Ten out of 68 studies addressed both gender and race/ethnicity, and included these variables in the model.

Race/ethnicity was included in 4 out of 12 studies predicting risk, in 5 out of 28 studies predicting prognosis, and in 3 out of 21 studies classifying diagnosis. It was never included in models investigating prognosis and diagnosis combined, and therapeutics.

Clinical Deployment

Regarding Diagnosis, just one model was actually deployed in a clinical environment [ 60 ]. The authors applied a lasso regression with metabolites as features, achieving an accuracy of 99%; the authors used data from a real clinical context, and therefore they deployed and evaluated their model performance on a clinical context, nevertheless, they did not validate their model. Regarding Prognosis, just 3 studies were conducted in a clinical setting [ 49 , 50 , 62 ]. Komaru et al. [ 49 ] predicted 1-year mortality following the start of hemodialysis through hierarchical clustering and achieved an AUC of 0.8; the authors used data from a clinical prospective study to deploy and evaluate their model. Furthermore, they validated the used clusters. Kanda et al. [ 50 ] applied a support vector machine model onto a real population in an observational study to deploy and evaluate their model. The authors achieved an accuracy of 89% through 13 variables; unfortunately, they did not disclose the weights of the variables nor did they validate the model, and therefore we do not know which variables were the most important. Akbilgic et al. [ 62 ] used a model based on a Random Forest algorithm, and achieved an AUC of 0.69; the most important features were eGFR, Spontaneous Bacterial Peritonitis, Age, Diastolic Blood Pressure and BUN. The authors used data from a real clinical context to deploy and evaluate their model; furthermore, they validated their results and model internally. Regarding Risk of developing CKD, one study’s model was used in a clinical context [ 42 ]. The authors used a NN, achieving an AUC of 0.89, using retinal images as features from a clinical context to deploy, evaluate and validate their model. Finally, regarding CKD Treatment, one study’s model was used in a clinical environment [ 26 ]; they presented their results through differences in achieved values by their algorithms, and the best performance was achieved by a NN. They evaluated the model with clinical data, but did not validate it.

Quality assessment

According to the PROBAST assessment tool [ 18 ], most of the included articles showed an overall low risk of bias ( n  = 48; 67.6%), and 65 (91.5%) of the included articles showed low applicability. Moreover, only 8.5% of the included studies scored less than 70% in the reporting guidelines for machine learning predictive models in biomedical research developed by Luo and colleagues [ 19 ]. The complete quality assessment can be found in Supplemental Material.

This systematic review describes how machine learning has been used for CKD. Six overarching themes were found, each of which underlines the need for further consideration by the scientific community.

First, despite the ever-growing number of studies focusing on the topic, a staggeringly low amount are being considered for actual clinical implementation. In this review, just 5 out of 68 articles tried to deploy their model in a real clinical setting. This might indicate either that the technology is not ready yet, or, considering 4 of these 5 articles were published in the last 3 years, that the technology is just starting to creep into real clinical settings. Recent evidence suggests that it is paramount to test newly developed algorithms in clinical settings before trying to deploy them [ 88 ]. Despite promising laboratory results, clinical translation is not always guaranteed. As an example, when studying the feasibility of providing an automated electronic alarm for acute kidney injury in different clinical settings, substantial heterogeneity in the findings among hospitals was described, with the worrying result of a significantly increased risk of death for some hospitals [ 89 ].

Second, as expected, the most important features were profoundly related to the main aim the authors were pursuing. In this regard, there were no surprises in the studied topics as the most important features were related to conditions known to lead to CKD diagnosis, worsening of prognosis and risk of developing CKD (e.g., age, comorbidities, systolic and diastolic blood pressure and eGFR values).

Third, a lack of consistency in reporting results was found. Most of the studies chose to report accuracy, but this was not the norm. Furthermore, while accuracy provides information on model performance, it fails to consider class imbalance and data representation. This is extremely important as accuracy in highly unbalanced datasets can be very high by always predicting the same binary outcome because of a flawed model. For instance, considering a low prevalence disease, if the algorithm is flawed for it always predicts a negative event, the accuracy will be high, but the veracity of the model will not [ 90 ]. As a result, AUCs and ROCs better measure the model precision without requiring the definition of a risk threshold. Twenty-nine authors chose to express their results including AUCs and ROCs: the minimum value was 0.69 and the maximum was 0.99 (mean: 0.83, median: 0.84). These results best express how precise the algorithms were and confirm the overall high performance of the assessed models.

Fourth, a common conundrum regarding feature selection and output was found in studies assessing CKD diagnosis. The definition of CKD requires certain variables to be present in order to make a diagnosis, thus including those variables in the model might be considered mandatory. Nonetheless, including those variables forces the model to streamline its decision process to a simple match in altered values, effectively transforming a complex machine learning model into a linear decision flow-chart, the performance of which will always be stellar.

This phenomenon is especially clear in four of the studies this systematic review assessed [ 36 , 39 , 46 , 47 ]. In these studies, the same database [ 91 ] is used, and accuracy, sensitivity, specificity, and ROC-AUC are never below 98%. We believe researchers should carefully assess the variables used in their machine learning models to make sure that no data leakage is present between features and results.

Fifth, model bias and fairness were almost never considered. This is critical, as both biased and unfair models will not achieve the same results in different demographics, and their societal impact could exasperate disparities in certain populations. These issues need to be further explored before any model can be implemented at point of care.

Finally, among the included studies, only 6 evaluated their models in a clinical setting [ 26 , 42 , 49 , 50 , 60 , 62 ], and only 3 were validated [ 42 , 49 , 62 ]. These studies showed promising results and did not report any unintended consequences after evaluation and/or validation. Notwithstanding the robust results described by the authors, as discussed before, recent evidence suggests that it is paramount to test newly developed algorithms in clinical settings to avoid adverse or unintended consequences [ 88 , 89 ]. Taking into account the pinnacle of importance of validating ones’ results in real clinical contexts and not just “in lab”, in reading their results, their generalizability has to be questioned, especially since no multi-center validations were described among the validated models.

This systematic review presents a few limitations: first, only one database (PubMed) was used to collect studies of interest. It should be noted that systematic reviews are usually exhorted to use at least two databases as stated by the PRISMA statement. Nonetheless, as PubMed has grown to be one of the most used search engines for medical sciences this limitation should be self-amending. Secondly, this systematic review assessed only papers written in English since English is the most widely adopted and commonly used language for the publication of medical papers.

In addition to these limitations, due to this review’s design, all in vitro studies (on cellular substrates) were excluded. Consequently, the evidence presented in this review is not to be interpreted as definitive for all things concerning CKD, since in vitro studies (on cellular substrates), the insight of which is critical in understanding pathogenetic as well as therapeutic mechanisms, were not assessed.

Lastly, the majority of included studies did not evaluate the integration of ML models in daily clinical practice, therefore the results and discussion have to be considered largely from an academic standpoint. Despite these limitations, we feel this review advances the knowledge on the current state of data-driven algorithms to advance CKD diagnosis, prognosis and treatment.

Despite the potential benefits, the application of machine learning for CKD diagnosis, prognosis, and treatment presents several issues, namely fairness, model and result interpretability [ 90 ], and the lack of validated models. Result interpretability concerns reflect the inability to explain which aspects of the dataset used in the training phase led to a predicted result in a particular case [ 92 , 93 ]. Therefore, as the trend in machine learning techniques moves from traditional algorithms (e.g., lasso regressions, support vector machine, and decision trees), to more complex ones (e.g., ensemble algorithms and deep learning), the interpretability concerns become more pronounced [ 90 ]. Notably, researchers highlighted the need for explainability and for models that could have a significant impact on patients' health [ 94 , 95 ]. These models should be reported using best practice reporting guidelines such as the Transparent Reporting of a Multivariate Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) [ 94 ] or MINimum Information for Medical AI Reporting (MINIMAR) [ 97 ]. Transparent and accurate reports are also fundamental in advancing multi-center validations of the applied models, which in turn is an essential step to ensure that only safe and sound models are applied on a large scale.

Most of the studies failed to report on the ethical issues revolving around their model development; the impact on the patient's well-being can also be affected by algorithmic bias [ 98 , 99 ] and this can be worse in certain underrepresented populations. This concern is closely related to the generalizability of the developed model [ 100 – 102 ]. Specifically, retrospective data that are usually used during the training phase often have significant biases towards subgroups of individuals that have been defined by factors such as age, gender, educational level, socioeconomic status, and location [ 98 ]. The issues of fairness and bias in algorithms should be evaluated by investigating the models’ performance within population subgroups.

This systematic review underlines the potential benefits and pitfalls of ML in the diagnosis, prognosis, and management of CKD. We found that most of the studies included in this systematic review reported that ML offers invaluable help to clinicians allowing them to make informed decisions and provide better care to their patients; nonetheless most of those articles were not actually piloted in real life settings, and therefore, notwithstanding the excellent model performance results reported by authors, the technology might not be ready for mass real-time adoption or implementation.

Although future work is needed to address the viability, interpretability, generalizability, and fairness issues, to allow a safer translation of these models for use in daily clinical practice, the implementation of these techniques could further enhance the effective management of hospital resources in a timely and efficient manner by potentially identifying patients at high risk for adverse events and the need for additional resources.

We hope the summarized evidence from this article will facilitate implementation of ML approaches in the clinical practice.

Below is the link to the electronic supplementary material.

Acknowledgements

Authors contribution.

FS and AC had the idea, extracted, and analyzed the data and wrote the manuscript. CF analyzed the data and wrote the manuscript. DGol, DGor, helped in results interpretation. THB revised the manuscript and helped in results interpretation. AC supervised the entire process.

Open access funding provided by Alma Mater Studiorum - Università di Bologna within the CRUI-CARE Agreement.

Data availability Statement

Declarations.

The authors did not receive support from any organization for the submitted work. The authors do not have any conflicts of interest to report.

Not Applicable.

The original article has been updated: Due to Abstract changes.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Change history

A Correction to this paper has been published: 10.1007/s40620-023-01609-9