REF-909 · Gonçalves et al., 2017 · Journal of Applied Physiology, Vol. · Paper · PubMed

Dispelling the myth that habitual caffeine consumption influences the performance response to acute caffeine supplementation

B+moderate evidenceMeasured on trained athletesWritten before October 2026, being re-checked

Forty male endurance-trained cyclists were split into tertiles by daily caffeine intake — low, moderate and high consumers — and each rode simulated time trials after 6 mg/kg caffeine, a placebo and no supplement. Caffeine improved time-trial performance over both placebo and control.

Habitual caffeine intake had no influence as a covariate, performance did not differ among tertiles, and habitual intake did not correlate with the caffeine-induced change. Similar numbers of riders in each tertile improved beyond the test's variation. The authors conclude that high habitual caffeine intake does not negate the benefit of acute caffeine.

Participants

n=40 male endurance-trained cyclists, allocated into tertiles by habitual daily caffeine intake:

low: 58 ± 29 mg/d
moderate: 143 ± 25 mg/d
high: 351 ± 139 mg/d

Protocol

Double-blind, crossover, counterbalanced; three trials: caffeine (CAF, 6 mg/kg body mass), placebo (PLA), no supplement (CON).
Simulated cycling time trial (TT) in the fastest time possible (~30 min).
Mixed-model analysis; habitual intake tested as a covariate (ANCOVA) and by correlation with the change CAF - CON.

Key Results

TT time: 29.92 ± 2.18 min with CAF vs 30.81 ± 2.67 (PLA) and 31.14 ± 2.71 min (CON); P = 0.0002.
Habitual intake as covariate: no influence on performance (P = 0.47).
Between tertiles: TT performance not different (P = 0.75).
Correlation between habitual intake and absolute change (CAF - CON): none (P = 0.524).
Individual responders (improvement beyond test variation): eight, seven and five in the low, moderate and high tertiles; not different by Fisher's exact test (P > 0.05).
Blood lactate and RPE did not differ between trials or tertiles (P > 0.05).

Practical Application

"I drink coffee every day — will a caffeine dose still help me race?" In 40 trained male cyclists, 6 mg/kg caffeine sped up a ~30-min time trial (29.92 vs 30.81 min on placebo), and the gain was the same in low (58 mg/d), moderate (143 mg/d) and high (351 mg/d) habitual consumers. Habitual intake is not a reason to skip race-day caffeine.
Not from this paper: a trial that gave daily low-dose caffeine for several weeks to low consumers did see the ergogenic effect shrink (REF-006, B), so the two results answer different questions — a new daily habit vs an existing one; neither tests pre-race withdrawal.
Not from this paper; the base's own practice: no caffeine withdrawal before a race; standard dosing and timing come from REF-350 and REF-249; the decision framework is SYNTH-67.

Limitations

Men only; one dose (6 mg/kg) and one ~30-min time trial.
Habitual intake was estimated per rider and compared across tertiles, so each comparison group is a third of a small sample (exact group sizes not stated in the abstract).
Did not test abstaining before the trial as an intervention.
Written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

controlled crossover, n=40; the habitual-intake comparison is between tertiles, so it is observational within the trial