Exercise- and dietary pattern-based lifestyle interventions and neurocognitive function in adults: a Bayesian...

To systematically review the effects of lifestyle interventions focused on exercise and dietary patterns on neurocognitive function in adults, and to examine their relative effects and dose-response relationships.

PubMed, Embase, Web of Science, and the Cochrane Library were searched from database inception to March 20, 2026. A total of 23 randomized controlled trials (RCTs) involving 12,286 participants were included. Bayesian multilevel pairwise, network, and dose-response meta-analyses were conducted to evaluate overall cognition, memory, and executive function. RoB 2 and CINeMA were used to assess the credibility of the evidence.

Combined interventions yielded small, but statistically significant cognitive benefits (overall cognition: SMD = 0.18; memory: SMD = 0.10; executive function: SMD = 0.09). These benefits were driven primarily by differences between the combined intervention and control or usual care, as well as exercise alone; compared with diet alone, the combined intervention still showed a small advantage for executive function. Network meta-analysis suggested that AE_MIND and AE_RT_MED were more beneficial for overall cognition, AE_RT_MIND and AE_RT_MED for memory, and AE_RT_CR and AE_MIND for executive function. Exploratory dose-response analyses suggested possible curvature in some spline-based models; however, comparison with simpler monotonic specifications did not provide robust evidence for a clinically interpretable non-linear or monotonic dose-response relationship.

Combined lifestyle interventions may produce statistically significant but modest improvements in cognitive function in adults, with benefits varying across cognitive domains. However, current evidence remains insufficient to support a standardized precision prescription, and high-quality RCTs with long-term follow-up are needed to determine the optimal combinations and dosages.