Meal Timing and Anthropometric and Metabolic Outcomes: A Systematic Review and Meta-Analysis

Abstract

Importance: Meal timing strategies, such as time-restricted eating (TRE), reducing meal frequency, or altering calorie distribution across the day, have gained interest for their potential to enhance weight loss and metabolic health, particularly in managing chronic diseases, yet their long-term benefits are not known.

Objective: To evaluate the association between meal timing strategies (≥12 weeks) and anthropometric and metabolic indicators.

Data sources: Medline, Embase, CINAHL, and Cochrane CENTRAL were searched from inception to October 17, 2023.

Study selection: Randomized clinical trials, regardless of language and publication date, involving adults 18 years and older, evaluating within-day meal timing patterns for 12 or more weeks, and reporting anthropometric measures were included. Studies were excluded if participants had eating disorders, prior significant weight change, underwent bariatric surgery, were pregnant, or if controlled variables differed between groups.

Data extraction and synthesis: Study quality was determined via Risk of Bias 2.0 tool. Data were extracted independently by multiple reviewers. Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines were used. Meta-analysis was performed using random-effects model on pooled continuous outcomes with 2 or more studies.

Main outcome and measures: Weight change in kilograms, reported as between-group mean difference with 95% CIs.

Results: Sixty-nine reports of 29 randomized clinical trials including 2485 individuals (1703 [69%] female; mean [SD] age, 44 [9.5] years; and mean [SD] body mass index, 33 [3.5]) were included. Study interventions included TRE (17 studies), meal frequency (8 studies), and calorie distribution (4 studies). There were some concerns of risk of bias for 7 studies and high concerns for 22 studies. Statistically significant weight change was observed in TRE when compared with control (-1.37 kg; 95% CI, -1.99 to -0.75 kg). Lower meal frequency and earlier caloric distribution were also both associated with greater change (-1.85 kg; 95% CI, -3.55 to -0.13 kg; and -1.75 kg; 95% CI, -2.37 to -1.13 kg, respectively).

Conclusions and relevance: The findings of this meta-analysis suggest that TRE, lower meal frequency, and earlier caloric distribution in the day may reduce weight compared with standard care and/or nutritional advice; however, the effect sizes found were small and of uncertain clinical importance. High heterogeneity and risk of bias among included studies led to concerns about the certainty of the underpinning evidence. Further research, including trials with larger sample sizes, standardized interventions with prescribed or matched energy intake, and longer follow-up, are needed.

Figure 1.. Flow Diagram of Included Randomized Clinical Trials

Figure 2.. Meta-Analysis of Difference in Mean Difference (95% CIs) for the Effect of Meal Timing Interventions on Weight, Grouped by the Nature of the Meal Timing Intervention

The forest plot shows effect estimates (squares) and 95% CIs (horizontal lines) for each randomized clinical trial (RCT). Larger squares indicate a larger weight has been assigned to that RCT. Left of the 0 line shows a finding in favor of interventions, whereas right of the 0 line shows a finding in favor of control. The diamond at the base of each plot demonstrates the pooled effect estimates and confidence intervals from all RCTs included in the meta-analysis. 2M/3M/6M, 2, 3, or 6 meals; 3M+3S, 3 meals and 3 snacks; BL, back loading (eating the most substantial/calorie-dense meal toward the end of the day); BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); FT, feeding time; ED, early dinner; EED, equal energy distribution (spreading calorie intake evenly throughout the day’s meals); FL, front loading (consuming the largest or most calorie-dense meal early in the day, typically at breakfast or breakfast and lunch); HCB, high-calorie breakfast; HCD, high-calorie dinner; LD, late dinner; ML, middle loading (eating the most substantial/calorie-dense meal in the middle of the day, usually at lunch).

Figure 3.. Meta-Analysis of Difference in Mean Difference (95% CIs) for the Effect of Meal Timing Interventions on HbA_1c (%), Grouped by the Nature of the Meal Timing Intervention

The forest plot shows effect estimates (squares) and 95% CIs (horizontal lines) for each randomized clinical trial (RCT). Larger squares indicate a larger weight has been assigned to that RCT. Left of the 0 line shows a finding in favor of interventions, whereas right of the 0 line shows a finding in favor of control. The diamond at the base of each plot demonstrates the pooled effect estimates and confidence intervals from all RCTs included in the meta-analysis. 2M/3M/6M indicates 2, 3, or 6 meals; BL, back loading (eating the heaviest/most calorie-dense meal toward the end of the day); BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); ED, early dinner; FT, feeding time; duration, follow-up duration in weeks; LD, late dinner; ML, middle loading (having the most substantial/calorie-dense meal in the middle of the day, usually at lunch).