Abstract
This study aimed to determine the effects of acute caffeine ingestion on cycling performance through a systematic review and meta-analysis, while also exploring the moderating roles of caffeine dosage, training status, and athlete age. A comprehensive search was conducted across five databases, yielding 20 eligible studies with a total of 226 participants. A three-level mixed-effects model was applied to pool main effects on cycling time trial performance, mean power output, mean heart rate, and ratings of perceived exertion (RPE). Subgroup analyses and meta-regression were performed to examine potential moderators. Caffeine intake significantly reduced cycling completion time (SMD = -0.36, 95% CI: -0.57 to -0.15, p = 0.0017) and increased mean power output (SMD = 0.29, 95% CI: 0.05 to 0.52, p = 0.02), but had no significant effect on heart rate or RPE. Subgroup analysis indicated that a low dose of caffeine (≤3 mg/kg; SMD = -0.42) was more effective in reducing completion time compared with a higher dose (4-6 mg/kg; SMD = -0.34). Meta-regression further revealed a significant moderating effect of age on time trial performance (β = -0.0501; p = 0.02). Taken together, these findings suggest that ingesting caffeine approximately one hour before exercise can effectively enhance cycling performance, with low doses achieving improvements comparable to higher doses.