Caffeine increases peripheral fatigue in low- but not in high-performing cyclists.

The influence of cyclists' performance levels on caffeine-induced increases in neuromuscular fatigue after a 4-km cycling time trial (TT) was investigated. Nineteen cyclists performed a 4-km cycling TT 1 h after ingesting caffeine (5 mg·kg) or placebo (cellulose). Changes from baseline to after exercise in voluntary activation (VA) and potentiated 1 Hz force twitch () were used as markers of central and peripheral fatigue, respectively. Participants were classified as "high performing" (HP, = 8) or "low performing" (LP, = 8) in accordance with their performance in a placebo trial. Compared with placebo, caffeine increased the power, anaerobic mechanical power, and anaerobic work, reducing the time to complete the trial in both groups ( < 0.05). There was a group versus supplement and a group versus supplement versus trial interaction for , in which the postexercise reduction was greater after caffeine compared with placebo in the LP group ( = -34% ± 17% vs. -21% ± 11%, = 0.02) but not in the HP group ( = -22% ± 8% vs. -23% ± 10%, = 0.64). There was no effect of caffeine on VA, but there was a group versus trial interaction with lower postexercise values in the LP group than in the HP group ( = 0.03). Caffeine-induced improvement in 4-km cycling TT performance seems to come at the expense of greater locomotor muscle fatigue in LP but not in HP cyclists. Caffeine improves exercise performance at the expense of a greater end-exercise peripheral fatigue in low-performing athletes. Caffeine-induced improvement in exercise performance does not affect end-exercise peripheral fatigue in high-performing athletes. High-performing athletes seem to have augmented tolerance to central fatigue during a high-intensity time trial.