Caffeine Intake and Performance Outcomes in Taekwondo Athletes: A Systematic Review and Meta-Analysis

Background: The ergogenic role of caffeine in taekwondo-specific performance has not been comprehensively synthesized. This study aimed to evaluate the effects of caffeine supplementation on performance outcomes in taekwondo athletes. Methods: PubMed, Web of Science, Cochrane Library, and SPORTDiscus were searched up to April 2026. Randomized controlled trials comparing caffeine supplementation with placebo in taekwondo athletes were included, including both acute and longer/repeated intervention protocols where eligible. Effect sizes were pooled as Hedges’ g, a small-sample-corrected standardized mean difference, using random-effects models. Evidence certainty was assessed using the GRADE approach. Results: Caffeine supplementation was associated with favorable pooled effects for explosive performance (Hedges’ g = 0.40; 95% CI 0.07–0.72; moderate-certainty evidence), reaction agility (g = 1.43; 95% CI 0.79–2.08; low-certainty evidence), technical accuracy (g = 1.05; 95% CI 0.32–1.77; low-certainty evidence), FSKT-10s performance (g = 3.31; 95% CI 1.42–5.19; very-low-certainty evidence), and FSKT-mult performance (g = 4.29; 95% CI 1.88–6.70; very-low-certainty evidence). No significant pooled effect was observed for endurance performance (g = 0.10; 95% CI −0.19–0.39; low-certainty evidence). Conclusions: Across the clinically heterogeneous acute and longer/repeated protocols included in this review, caffeine supplementation was associated with favorable pooled effects for explosive performance, reaction agility, and technical accuracy, whereas no significant pooled effect was observed for endurance outcomes. These estimates represent averages across different intervention protocols and should not be interpreted as a uniform effect of either acute or repeated supplementation. FSKT-based effects were large but should be interpreted cautiously given high heterogeneity. Observed dose-related patterns should be considered hypothesis-generating because formal subgroup or dose–response analyses were not conducted; independent replication is needed.