Poor quality of dietary assessment in randomized controlled trials of nutritional interventions for type 2 diabetes may affect outcome conclusions: A systematic review and meta-analysis.

Objectives: The aim of this study was to evaluate the quality of dietary assessment methods in randomized controlled trials focusing on individuals with type 2 diabetes (T2DM), and its impact on the favorability of conclusions.

Methods: MEDLINE, EMBASE, CINAHL, and CENTRAL were searched, from inception until September 2019 for RCTs of dietary interventions in individuals with T2DM. Investigators assessed risk of bias and quality of the dietary measurements using the Cochrane Risk of Bias Assessment Tool 2.0 and the redeveloped EURICA tool, respectively. Random-effects meta-analysis assessed mean changes in hemoglobin (Hb)A1c. The study was conducted in accordance with the Preferred Reporting in Systematic Reviews and Meta-analyses. PROSPERO registration number: CRD42019146471.

Results: Of 2552 records retrieved, 23 studies met the inclusion criteria. Two studies were rated as good, 6 as medium, and 15 as poor in the quality assessment of the dietary measurement tool. All eight studies with higher quality of dietary assessment were associated with favorable conclusions. Among the 15 studies with poor quality, 5 failed to draw favorable conclusions. Among studies that sought to produce a reduction in HbA1c, 3 of 6 with better dietary assessment quality produced a significant difference of -0.38% (-0.67% to -0.08%), and 4 of 12 of poorer quality produced a significant difference of -0.26% (-0.37% to -0.14%).

Conclusions: The poor quality of dietary assessment in clinical trials casts uncertainty on the dietary outcomes and the validity of possible causal mechanisms. Attention to the accuracy and reliability of dietary assessment methods is indicated.