Current Guidelines on Dietary Protein Intake for Patients with Diabetic Kidney Disease: A Scoping Review

Abstract

Background

International nutritional guidelines vary in their recommendations, as well as in the reporting quality of the underlying evidence. An underreported recommendation in guidelines for people with type II diabetes concerns the optimal protein intake in the case of chronic kidney disease (CKD).

Objectives

In this review, we analyze the protein intake recommendations from worldwide practice guidelines for type II diabetes and CKD.

Methods

We conducted a systematic search in the online databases PubMed, Embase, and Cochrane Database of Systematic Reviews, as well as websites, according to Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines, and up to 31 January, 2023. We included nutritional and clinical practice guidelines on protein intake in patients with type II diabetes and CKD of different stages. We assessed the quality of the guidelines using the instruments Appraisal of Guidelines for Research & Evaluation Instrument (AGREE II) and AGREE Recommendation Excellence. We assessed the quality of the underlying scientific evidence using the evidence pyramid and the level of evidence categorization.

Results

Of 24 included guidelines, 5 met the quality assessment of all instruments. Their evidence was based on an overlapping number of randomized controlled trials (RCTs). Nevertheless, their protein intake recommendations differed, ranging from no restriction to restriction to 0.8 g/kg body weight/d.

Conclusions

The reason for the discrepancy of protein intake recommendations between different guidelines could not be determined conclusively, as the methodology of evidence assessment was often insufficiently reported. More long-term and good-quality RCTs are needed. We recommend the use of rigorous development and quality assessment tools in the development of clinical practice guidelines for patients with CKD and type II diabetes.

Introduction

Diabetes mellitus type II is a chronic disease characterized by high plasma glucose levels mainly due to the resistance of the body to the effects of insulin, insufficient insulin secretion, or both [1,2]. In 2021, the International Diabetes Federation estimated that 537 million adults worldwide suffered from diabetes. It was also predicted that by 2045, this number would increase to 783 million (+46%). In addition, it was estimated that almost 1 in 2 adults worldwide are undiagnosed. Even if there are different types of diabetes, type II diabetes is the most common one, with over 90% of the cases [3]. Lack of proper management of diabetes can lead to serious health consequences such as vascular, renal, and neurologic complications [4]. In 2021, nearly 6.7 million adults died because of diabetes or its complications [3].

Diabetic nephropathy ranks as the most common complication of type II diabetes and is the most common cause leading to end-stage renal failure worldwide [2]. Interestingly, the prevalence of chronic kidney disease (CKD) in patients with diabetes varies widely between countries [5]. According to the latest Standards of Care in Diabetes published by the American Diabetic Association, 20%–40% of people with diabetes have diabetic kidney disease. It may progress to end-stage renal disease, which requires dialysis or transplantation [6]. Due to the decrease in kidney function, there is an accumulation of various degradation products in the blood with an increased risk of hyperkalemia and hypophosphatemia, which further impair organ function [7]. In addition, it was shown that an increased protein intake triggers hyperfiltration in the kidneys, which increases the renal workload [7].

Apart from the control of blood glucose, blood lipids, and blood pressure with drugs, a healthy diet, regular physical activity, smoking cessation, and maintenance of a healthy body weight play important roles in the management of type II diabetes [3]. In fact, there is evidence that medical nutritional therapy is an effective and essential therapy in the management of diabetes [8]. However, former reviews showed that nutritional guidelines can vary in their recommendations, as well as in reporting quality of the underlying evidence [9], especially in the field of type II diabetes [10]. An underreported recommendation in nutritional guidelines for people with type II diabetes concerns the optimal protein intake in the case of diabetic CKD. Managing this balance is particularly challenging because most people affected by both diabetes and CKD are elderly adults. A higher protein intake is generally advised in older populations to help prevent sarcopenia, a term summarizing the geriatric syndrome involving progressive and generalized loss of muscle mass and strength with risk of adverse outcomes such as increased risk of falls, physical disability, and poor quality of life [11]. This recommendation conflicts with the dietary restrictions imposed by CKD. In patients with declining kidney function, high protein intake can exacerbate complications such as hyperkalemia and hypophosphatemia [12].

The aim of this review is to analyze the protein intake recommendations from worldwide practice guidelines for type II diabetes with special focus on adult patients with diabetic nephropathy. A clear and evidence-based recommendation on protein intake in people with diabetes and CKD of different stages is important to help dieticians and practitioners to best instruct their patients.

Methods

Protocol and registration

The review protocol was registered at the International Platform of Registered Systematic Review and Meta-analysis Protocols (INPLASY), registration number INPLASY202460004.

Data sources and searches

A systematic literature search was performed in the databases PubMed, Embase, and Cochrane Database of Systematic Reviews published until 31 January, 2023, including terms on disease risk factors, outcomes of interest, and study design. The complete search strategies for each database are reported in Supplemental Material 1. The literature search was complemented by a search of reference lists of included guidelines, and online searches on websites of national and international organizations worldwide publishing diabetes guidelines. We followed the PRISMA recommendations for the reporting of scoping reviews [13].

Eligibility criteria

The guidelines found were assessed by 2 independent reviewers for inclusion or exclusion, using predefined criteria. We included all practice guidelines for the management of type II diabetes that contained specific dietary recommendations for patients with clinically diagnosed CKD. All guidelines written in any language other than English were excluded. In case of a relevant non-English guideline, we searched for a published English translation and included them. Guidelines were excluded if they focused on other diseases rather than on diabetes or if they targeted a healthy population. We also excluded guidelines for pediatric patients or if they were specifically formulated for patients with type I diabetes. Study types other than guidelines (i.e., interventional or observational studies, reviews, editorials, expert opinions) were excluded. Guidelines with no dietary management or with exclusively pharmacological treatment recommendations were also excluded. The latest available version of each guideline was selected. Conflicts between the 2 independent reviewers were solved by discussion.

Quality assessment

Each guideline was assessed by 2 independent reviewers using the latest version of the Appraisal of Guidelines for Research & Evaluation Instrument (AGREE II instrument) [14]. This instrument was judged as being among the best and most comprehensive instruments for the assessment of practitioner guidelines. Before assessing the guidelines, a training phase was conducted.

AGREE II includes 23 items categorized in 6 domains: 1) scope and purpose, 2) stakeholder involvement, 3) rigor of development, 4) clarity and presentation, 5) applicability, and 6) editorial independence. For each guideline, each of the 23 items was rated by 2 independent reviewers on a 7-point scale (from 1 = strongly disagree to 7 = strongly agree). Afterward, an average score was calculated according to the given formula for each of the 6 AGREE II domains. Finally, 2 global rating items (overall quality of the guideline and whether the guideline would be recommended for use in clinical practice) were calculated. The inter-rater reliability was calculated using Cohen’s Kappa [15]. The statistical software STATA version 14 was used [16].

Complementary to the AGREE II instrument, the quality of guideline recommendations was assessed by the tool AGREE Recommendation Excellence (AGREE REX) [17]. This tool focuses specifically on the development and reporting of guideline recommendations. AGREE REX includes 9 items organized in 3 domains: 1) clinical applicability, 2) values and preferences, and 3) implementability. The 2 AGREE tools complement each other.

To perform an assessment with AGREE REX, 2 topics had to be determined in advance. The first topic was “protein intake in patients with diabetes.” We rated the 1) clinical applicability, 2) values and preferences, and 3) implementability of the recommendations in relation to protein intake in patients with diabetes. The second topic was “Alterations in case of chronic kidney disease.” We evaluated all 9 items in relation to this second topic. The recommendations were rated again on a 7-point-scale, and a score was calculated by the given formula as in AGREE II. Again, the inter-rater reliability was calculated using Cohen’s Kappa for each of the 2 questions.

Evidence assessment

Guidelines that scored 60% or more (= better than the average of all guidelines, which was 59%) in domain 3 Rigour of Development in AGREE II, 63% or more in the domain 1 Clinical Applicability in AGREE REX (= better than the average of all guidelines, which was 62%) for the question on CKD, and at the same time scored above 50% (= better than half of the points available) in domain 1 Clinical Applicability in AGREE REX for the question on protein intake in patients with diabetes, were selected and examined more closely for evidence. A table was created with the recommended protein amount, as well as the associated evidence sources that were explicitly mentioned in the respective guidelines (Table 1) [6,8,[18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47]]. Evidence explicitly mentioned in the guidelines was assigned to hierarchy level 1. References that contributed to this evidence were assigned to level 2. Other studies mentioned in level 2 references were assigned to level 3. We limited our search to 3 hierarchy levels. If no specific recommendations were found in the guidelines or access to the references was not possible, the search was terminated earlier.

TABLE 1.

Summary of evidence used in the 5 best guidelines.

Guideline	Recommendation on protein intake	Reference I	Reference II	Reference III	Study design	Level of evidence
ADA [6]	0.8 g/kg body weight per day for people with nondialysis-dependent CKD	[18]	[19]	[[20], [21], [22], [23], [24], [25], [26], [27], [28]]	Meta-analysis RCT RCT RCT RCT RCT Crossover RCT Crossover RCT One-group before–after study	A
KDIGO [29]	0.8g/kg body weight per day for people with nondialysis-dependent CKD	[30]	[31]	Not accessible	Unclear	Unclear
Diabetes Canada [32]	Max 0.8 g/kg/d for people with diabetes who have CKD	[33]	[31]	Not accessible	Unclear	Unclear
SEMDSA [34]	Reducing protein intake to <0.8 g/kg/ideal body weight is not recommended for adults with micro and macroalbuminuria	[8]	[21,23,24,25,[35], [36], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46]]		RCT RCT RCT Cross-sectional RCT RCT RCT Cross-sectional RCT crossover RCT RCT RCT Cross-sectional Cochrane review with 9 RCTs Meta-analysis with 8 RCTs Guideline	A
SIGN 2017 [47]	Dietary protein restrictions (<0.8 g/kg/d) are not recommended in patients with early stages of CKD (stages 1–3)	[21,23,36,38]			RCT RCT RCT RCT	A

The guidelines cited references of level I. The references of level I cited references of level II. The references in level II cited references of level III. The process was stopped when the references were original studies or when a reference could not be accessed.

Abbreviations: ADA, American Diabetes Association; CKD, chronic kidney disease; KDIGO, Kidney Disease Improving Global Outcome; RCT, randomized controlled trial; SEMDSA, Society for Endocrinology, Metabolism and Diabetes in South Africa; SIGN 103, Scottish Intercollegiate Guidelines Network.

The cited references were categorized by study type according to the evidence pyramid [48], as well as by publication date. The quality of the evidence was captured by the study types included, following the level of evidence categorization of the European Society of Cardiology and European Atherosclerosis Society [49] (Table 2).

TABLE 2.

Level of evidence categorization depending on study design [49].

Level A	Data derived from multiple randomized clinical trials or meta-analyses
Level B	Data derived from a single randomized controlled trial, or from large nonrandomized studies
Level C	Data derived from expert opinions and/or small studies, retrospective studies, registries

Results

The searches in PubMed and Embase and Cochrane Database of Systematic Reviews returned 47 references. There were no duplicates. During the title/abstract screening and full-text screening, 23 references were marked as irrelevant. Ten of these references were of incorrect study design. Six of the references were outdated (a newer version of the guideline was available), 4 studies were performed on incorrect study populations, and a further 4 references could not be accessed. Three references were written in a language other than English. Four further guidelines from organizational websites were included, achieving a total of 24 references. Figure 1 [50] shows a flow chart of the selection process.

FIGURE 1.

PRISMA 2020 checklist on the flow of guidelines inclusion and exclusion [50].

Results of AGREE II assessments

Overall, domain 1 “Scope and purpose” was best covered by the 24 included studies, with an average of 92% of possible points. Half of the included guidelines reached a top ranking with a score of 100%. The next best covered domain was “Clarity and Presentation” (domain 4), with an average of 89%, 14 guidelines reached a score of 89% or above in this domain. Domain 5 reported on the clinical applicability of the guidelines and was covered in mean to 76%. This domain was therefore rather well covered.

Domain 3 covering Rigour of Development had a very wide range of scores from 14% to 96%, with an average of 59%. Thirteen guidelines reached a score higher than 59%.

Domain 6 “Editorial independence” was the least covered with an average of 54%. In 6 of the assessed guidelines, neither statement concerning potential bias or competing interests nor funding statements could be found, resulting in a score of 0%.

The highest average overall score with 93% was reached by the guideline Kidney Disease Improving Global Outcome (KDIGO) [29], followed by American Association of Clinical Endocrinology [51] with 91% and by Ministry of Health Malaysia Guidelines [52] and National Health and Medical Research Council, Australia [53], with 90% each. For all results, see Supplemental Material 2. The mean inter-rater reliability of all items was 0.66, which is considered as substantial [15].

Results of AGREE REX assessments

For the first question on CKD, 13 guidelines met the quality criteria in domain 1. Many, but not all, overlapped with the results of the AGREE II assessment. Three guidelines reached 100% of points in this question: KDIGO [29], Scottish Intercollegiate Guidelines Network (SIGN) 103 [54], and Society for Endocrinology, Metabolism and Diabetes in South Africa (SEMDSA) [34].

For the second question on protein intake of patients with CKD diabetes, 7 guidelines met the quality criteria in domain 1. Two guidelines reached 100% of points in this question, KDIGO [29] and SIGN 103 [54]. For all results, see Supplemental Material 3. The mean inter-rater reliability of all items was 0.76 for question 1 (substantial), and 0.82 for question 2 (almost perfect) [15].

Results of the recommendations for protein intake

When both instruments were taken together, 5 guidelines met all quality criteria [6,29,34,54,55]. From the 5 guidelines that showed the highest ranking in AGREE II and AGREE REX, 3 recommended a maximal intake of 0.8 g protein per kg body weight and per day for patients with type II diabetes and with nondialysis-dependent CKD [6,29,55]. This amount corresponds to the recommended minimal daily allowance for healthy adults [30]. The recommendations of the American Diabetes Association [6] are based on multiple RCTs and show, therefore, a high level of evidence (A). However, in the other 2 guidelines (KDIGO [29], Diabetes Canada [55]), the level of evidence is unclear, as the source of evidence seems to go back to a book chapter [31] that is not available online. Two further guidelines do not recommend a protein restriction in early-stage CKD with micro and macroalbuminuria (SEMDSA [34], SIGN 103 [54]). These guidelines are based on multiple RCTs and have therefore a high level of evidence (A). For detailed results, see Table 1.

Discussion

Most nutritional guidelines for people with CKD and diabetes II included in our review did not show a high quality in all assessed domains, in the instrument AGREE II, as well as in AGREE REX. Previous reviews on the quality of dietary guidelines on nutrient reference values for the healthy adult population and for pregnant or lactating women [9] and on recommendations on the delivery of nutrition therapy [56] found similar results. The optimal macronutrient composition of diet remains an area of uncertainty in guidelines for the management of diabetes [10,57].

The 5 highest ranking guidelines with a quantitative protein intake statement showed a divergent agreement on the maximal protein intake per day for patients with type II diabetes and early-stage CKD. Although 3 guidelines recommended limiting protein to a daily maximum of 0.8 g/kg body weight, 2 other guidelines concluded that the available evidence was not sufficient to limit protein intake for this patient group. Interestingly, all 5 guidelines partially included the same source references, which were mainly RCTs. The 2 guidelines that did not recommend a protein intake restriction judged the available evidence as insufficient, as they judged the included studies as being too small and of too short duration. It remains unclear how the 5 guidelines came to the sources of evidence cited, how these sources were assessed, and why the guidelines came to different interpretations of the evidence.

The Kidney Disease Outcomes Quality Initiative (KDOQI) published an update on their clinical practice guideline for nutrition in CKD in 2020 [7]. They recommend a dietary protein intake of 0.6–0.8 g/kg body weight per day in persons with CKD 3–5 and who have diabetes. In persons with CKD 5D who have diabetes, they recommend a dietary protein intake of 1.0–1.2 g/kg body weight per day to maintain a stable nutritional status. Furthermore, for patients at risk of hyper- and/or hypoglycemia, they recommend higher levels of dietary protein intake to maintain glycemic control. These recommendations were formulated as opinions, as they judged the available evidence as not sufficient or of too low quality to formulate a graded recommendation, similar to our findings.

Apart from protein quantity, protein type (animal compared with vegetal) was described as to be of relevance for risk and progression of CKD [58]. However, regarding protein type, similarly to other guidelines and reviews [59], KDOQI did not find sufficient evidence to recommend a particular protein type, although they acknowledge, that a vegetarian diet might have health benefits in people with CKD by reducing serum phosphate and fibroblast growth factor 23 in patients not receiving dialysis, and therefore delaying the development and progression of CKD [7].

Furthermore, more long-term and larger RCTs on the maximal protein intake per day for patients with type II diabetes and CKD and over the age of 65 are needed. As the global population is increasingly aging, with many individuals facing diabetes, the need for adequate dietary protein intake becomes increasingly important to prevent muscle-wasting conditions such as sarcopenia, especially in individuals over the age of 65. Aging is often associated with reduced appetite, further compounding risk of insufficient protein intake. However, in this population, CKD is also common, and CKD limits the ability to consume high amounts of protein due to the increased risk of complications such as hyperkalemia and hypophosphatemia. As a result, individuals with CKD are particularly vulnerable to sarcopenia, driven by protein-energy wasting and malnutrition. Anti-inflammatory diets, such as the Dietary Approaches to Stop Hypertension (DASH) diet and plant-based diets, are therefore recommended in elderly people with CKD [12].

Despite the fact that there are guidelines on how to write clinical practice guidelines [60,61], the problem of unclear assessment and interpretation of evidence in nutritional guidelines seems to persist [[62], [63], [64]], and tools to assess the quality of dietary guidelines were proposed, such as the Reporting Items for practice Guidelines in Healthcare checklist that includes 22 items that are considered essential for good reporting of practice guidelines [17,65,66]. Such quality assessment tools were reviewed for their performance [67,68]. As we still encountered unclear descriptions of assessment methodology and unclear deduction of recommendations from evidence, a more rigorous adherence to quality criteria in the formulation of clinical practice guidelines is recommended. The fact that the clinical applicability of the assessed guidelines showed a coverage of 76% is encouraging, but leaves room for improvement. The applicability of guidelines is of major importance, as this is their main aim.

Strengths and limitations

This review is based on a comprehensive literature search and on standardized methods and tools for the performance of systematic reviews. In addition to the data search on PubMed, Embase, and Cochrane, a web search was conducted to include any relevant guidelines that may have been missed. The AGREE II instrument was used to assess the quality of the guidelines. Additionally, AGREE REX was used to assess the evidence. The inter-rater reliabilities of these assessments were substantial to almost perfect. The evidence pyramid and a published table of levels of evidence were used to assess the quality of the source references mentioned in each guideline.

A limitation of the present guidelines is that languages other than English were not included. As a result, some data might have been lost, which also means that geographical regions such as South America or some Asian countries are not represented. A thorough quality assessment from the source references, for instance, using the tool Grading of Recommendations, Assessment, Development, and Evaluations [69], could not be performed because of feasibility.

The low coverage of the domain 6 “Editorial independence” with an mean of 54% among all guidelines in AGREE REX II is worrisome. In 6 of the assessed guidelines, no statement concerning potential bias or competing interests, nor funding statements, could be found, making a judgment of the reliability of the recommendations difficult. However, the 5 guidelines that were included in the quality of evidence assessment had a coverage in domain 6 of 63%–100%, showing a higher level of reliability of the recommendations given.

In conclusion, there is still no consensus on the quantity and quality of daily protein intake for patients with CKD and type II diabetes. The discrepancy of recommendations between different clinical practice guidelines could not be determined conclusively, as the reporting on the methodology of evidence assessment was often insufficient. We recommend the consistent use of rigorous development tools, as well as quality assessment tools, in the development of clinical practice guidelines, including nutritional guidelines for patients with CKD and type II diabetes. Furthermore, more long-term and larger RCTs on the maximal protein intake per day for patients with type II diabetes and CKD are needed. Our results shall inform us on which specific aspects of such guidelines still need improvement in the future.

Author contributions

The authors’ responsibilities were as follows – SR: contributed significantly to the work’s conception, design, data collection, and data interpretation and analysis, participated in the writing or critical revision of the article in a manner sufficient to establish ownership of the intellectual content, and read and approved the version of the manuscript being submitted; AF: contributed significantly to the data collection, and data interpretation and analysis, participated in the writing or critical revision of the article in a manner sufficient to establish ownership of the intellectual content, and read and approved the version of the manuscript being submitted; and FR, NB: contributed significantly to the work’s conception, design, and data interpretation and analysis, participated in the writing or critical revision of the article in a manner sufficient to establish ownership of the intellectual content, and read and approved the version of the manuscript being submitted; and all authors: read and approved the final manuscript.

Data availability

Original data can be requested at the corresponding author.

Funding

This study was funded by Mäxi Foundation Zurich.

Conflict of interest

The authors report no conflicts of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article: