Effect of different dietary patterns on patients with type 2 diabetes: a systematic review and meta-analysis of randomized controlled trials.

Abstract

Background: Type 2 diabetes mellitus (T2DM) affects over 500 million people worldwide, and dietary management is a cornerstone of treatment. This meta-analysis descriptively evaluated the comparative efficacy and safety of four common dietary patterns-Mediterranean diet (MD), intermittent fasting (IF), energy-restricted diet (ERD), and ketogenic diet (KD)-in adults with T2DM.

Methods: \We searched the Cochrane Library, Embase, PubMed, and Web of Science from inception to March 13, 2025, for randomized controlled trials (RCTs). Outcomes included glycemic control (FBG, HbA1c), anthropometric parameters (weight, BMI, waist circumference), lipid profiles (TC, LDL-C, TG), and safety data. Random-effects models were used, with heterogeneity assessed by the I² statistic, and sensitivity analyses via leave-one-out. Risk of bias was evaluated using the Cochrane RoB 2.0 tool.

Results: A total of 25 RCTs were included. Reductions in FBG (MD = -0.63; I² = 62.1%) and HbA1c (MD = -0.50; I² = 83.8%) were observed, but the pooled estimates were limited by substantial heterogeneity and prediction intervals crossing zero. The MD showed excellent cross-study consistency for FBG (I² = 0%) with its prediction interval entirely within the beneficial range, whereas the KD showed limited clinical relevance for FBG due to extremely high heterogeneity (I² = 79.7%). For weight loss, the overall MD was -3.00 kg, but with substantially high heterogeneity (I² = 93.9%) across subgroups. BMI outcomes were limited by extremely high heterogeneity (I² = 97.5%) and all prediction intervals crossing zero, while waist circumference showed no statistically significant effect. None of the diets demonstrated a statistically significant advantage in lipid modification (TC, LDL-C, TG), with all confidence intervals crossing the line of no effect. Safety analyses showed KD was associated with a higher incidence of adverse events (gastrointestinal symptoms, transient hypoglycemia), while the other three diets exhibited favorable safety profiles.

Conclusion: This meta-analysis descriptively supports the potential benefits of all four dietary patterns for glycemic control and weight reduction. However, due to substantial inter-study heterogeneity and prediction intervals crossing the line of no effect, the pooled estimates preclude definitive claims of relative superiority among diets. Consequently, personalized dietary strategies balancing efficacy, safety, and patient preferences are recommended, with cautious interpretation of indirect cross-diet comparisons.

Systematic review registration: https://www.crd.york.ac.uk/PROSPERO/view/CRD420261478257, identifier PROSPERO (CRD420261478257).