Physical exercise decelerates executive function decline in aging: An umbrella review

In this umbrella review, we assessed the relationship between regular physical exercise and the deceleration of executive function decline in healthy older adults, focusing on meta-analyses from randomized controlled trials (RCTs). Analyzing 18 meta-analyses, which included 109 original studies and 11,266 participants, we found that physical exercise has a moderate positive effect on executive function (median Cohen's d = 0.157, 95% CI [0.085, 0.251]). Despite some dispersion (median d = 0.47, 95% CI [0.012, 0.127]; range -0.13–0.29) and heterogeneity (median I² = 0.03%; range 0–0.11%) after recalibration, the outcomes still showcased a positive effect. A multilevel analysis of all original studies served as a benchmark for evaluating the results, revealing an overall effect size of d = 0.16, 95% CI [0.08, 0.30], p < .0001, with considerable heterogeneity I² = 32.12%. Compared to the summary effect of the full model, the initially reported and recalculated effects had an average deviation of 0.15 and 0.08, respectively. A positive correlation existed between the publication year and the number of included original studies (Kendall's τ = 0.418, p < .001), reflecting an exponential growth in the literature. Various methods, including visual inspections of funnel plots, the trim-and-fill procedure, and Egger's test, were employed to assess publication bias, ultimately determining the effect size of physical exercise on improving executive function to be d=0.22, 95% CI [0.07, 0.12], higher than the pre-correction estimate. This finding emphasizes the significant role of physical exercise in delaying the decline of executive function in healthy older individuals, while also highlighting the need for future research to remain vigilant about publication bias and to employ comprehensive methods for evaluating and reporting study outcomes.