Abstract
OBJECTIVES: To evaluate the effects of concurrent training (CT) on body composition, metabolic health, and inflammatory outcomes in postmenopausal middle-aged and older women, and to examine whether training experience, exercise order, and training dose moderate these effects. METHODS: This review followed PRISMA 2020. Web of Science, PubMed, Scopus, and the Cochrane Library were searched for randomized controlled and randomized crossover trials last searched on July 1, 2025. Effect sizes (Hedges’ g) were synthesized using multilevel mixed-effects models (REML) with robust variance estimation to account for within-study dependence. Subgroup analyses and meta-regression explored potential moderators, and certainty of evidence was assessed using GRADE. RESULTS: Twenty-three studies (n = 1001) involving naturally postmenopausal women were included. CT significantly improved body composition (ES = − 0.11, 95% CI [− 0.20, − 0.02], P = 0.02) and metabolic health (ES = − 0.22, 95% CI [− 0.40, − 0.04], P = 0.02), with moderate-certainty evidence for both outcomes. The pooled effect on inflammatory markers was not statistically significant (P = 0.39; low-certainty evidence). No significant moderators were identified, and meta-regression showed no significant dose–response relationship for training volume or intervention duration. Subgroup analyses suggested reductions in body fat percentage and increases in lean mass, alongside small-to-moderate improvements in blood pressure, blood glucose, triglycerides, and maximal oxygen uptake. Biomarker-specific analyses suggested changes in IL-10, IL-6, and TNF-α (P < 0.05), but these findings were based on limited data and were low certainty. Exploratory within-group analyses indicated significant effects in sedentary participants and in studies prescribing resistance training before endurance training or three sessions per week. CONCLUSION: CT may be associated with small improvements in body composition and modest improvements in metabolic health in postmenopausal women. Given the small pooled effect sizes and the lack of MCID-based interpretation, the clinical relevance of these changes remains uncertain. Because moderator analyses were non-significant, subgroup patterns should be considered hypothesis-generating. Larger, adequately powered trials with standardized reporting are needed to clarify clinically meaningful effects and potential effect modifiers. GRAPHICAL ABSTRACT: [Image: see text] SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12905-026-04359-5.