The study, led by Hossein Miraftabi and colleagues including Craig Pickering, Alvaro Lopez-Samanes and Olivier Girard, pooled data from 25 randomized controlled trials involving 390 participants and 47 separate performance outcomes. The team followed Cochrane Handbook and PRISMA guidelines, searched four major databases up to March 2026, and prospectively registered their protocol with PROSPERO. Their headline finding: compared with placebo, caffeine gum improved performance with a standardized mean difference of 0.195 — a “trivial” effect by conventional statistical thresholds, but one that reached clear significance and, crucially, came with almost no inconsistency. Heterogeneity across studies was a remarkably low 4 percent, meaning the results were unusually reproducible from one laboratory to the next.

The magnitude of the effect matters less than its consistency. In elite and competitive sport, where medals are decided by hundredths of a second and single percentage points, a marginal but dependable gain can be the difference between a podium and a parking spot. The authors are careful to note that the statistical effect is small and that the certainty of the evidence, graded using the GRADE framework, ranged from low to very low. Yet their conclusion is pragmatic: even small improvements “may provide a competitive advantage in certain sporting situations.” That framing reflects a growing consensus in exercise science that effect sizes should be interpreted in the context of the event, not against generic benchmarks developed for clinical medicine.

Caffeine’s ergogenic credentials are built on more than a century of research. The compound works primarily by antagonizing adenosine receptors — particularly the A1 and A2A subtypes — in the central nervous system, which raises alertness, increases arousal and dampens the perception of fatigue and effort. It also stimulates the release of catecholamines such as adrenaline and noradrenaline, engages dopamine pathways tied to motivation, and may enhance calcium handling within muscle fibres and the efficiency of neuromuscular transmission. Responses, however, vary widely between individuals, shaped by genetic polymorphisms such as CYP1A2 and ADORA2A, habitual intake, and expectancy effects. The appeal of gum lies in pharmacokinetics: because a portion of the caffeine is absorbed through the buccal mucosa in the mouth, plasma concentrations begin climbing within roughly five to ten minutes, compared with the 45 to 60 minutes typically needed for capsules to reach peak absorption.

When the researchers broke the data down by outcome type, a clear hierarchy emerged. Aerobic endurance showed the largest benefit, a small but significant effect, spanning tasks such as 5-kilometre runs, cycling time trials, time-to-exhaustion protocols and the Yo-Yo Intermittent Recovery Test. Anaerobic performance — sprints, agility drills, rowing ergometer tests — and vertical jump measures each produced trivial but statistically significant improvements. Muscular strength and endurance, by contrast, showed no meaningful effect, with a pooled estimate that did not reach significance. The pattern aligns with the broader caffeine literature, which has repeatedly suggested that the drug favours endurance-style tasks over pure strength efforts, likely because the central mechanisms reducing perceived exertion pay larger dividends in prolonged exercise.

Training status proved to be one of the most striking moderators. Trained individuals derived a small but significant benefit, while recreationally active participants showed essentially nothing — a slightly negative, non-significant effect. The authors offer a physiological rationale: adenosine receptor density appears to be higher in trained than in untrained muscle and nervous tissue, potentially amplifying caffeine’s antagonistic effects on those receptors and sharpening its influence on central drive and effort perception. But they also urge caution. Only six effect sizes came from recreationally active participants, drawn from just 59 people, so low statistical power could be masking a real effect. The apparent trained-versus-untrained divide may partly reflect sampling and design artefacts rather than clean physiology.

Two findings carry particular practical weight for coaches. First, chewing duration mattered: ten minutes of chewing produced a small, significant performance improvement, whereas five minutes did not. Prolonged chewing plausibly releases more caffeine and extends the exposure of the buccal mucosa — and the bitter taste receptors of the TAS2R family that line the mouth — enhancing both absorption and possible chemosensory contributions to arousal. Second, habitual caffeine intake shaped the response. Participants with low or moderate daily consumption showed significant gains, while the small number of heavy caffeine users showed none, and even a slight negative trend. Chronic intake is thought to blunt acute effects by upregulating adenosine receptors, forcing the nervous system to compensate for a perpetually blocked signal. One cited study found that 28 days of daily caffeine abolished the benefit of a 3 mg/kg acute dose.

Dose itself showed a genuine relationship with outcome. In a meta-regression treating dose as a continuous variable, higher caffeine intakes were associated with larger performance effects, with most studies clustering between 2 and 4 mg per kilogram of body mass — the range the authors highlight as ergogenically effective for gum. Timing, surprisingly, did not matter. Whether the gum was chewed 5, 10 or more minutes before exercise, performance outcomes were statistically indistinguishable, a finding the authors note departs from an earlier 2023 meta-analysis by Barreto and colleagues that suggested gum worked best when consumed within a 15-minute window. That earlier review included barely half the studies now available, and the evidence base has more than doubled since its publication.

The team subjected their findings to a battery of sensitivity analyses, and the core conclusions held firm. Varying the assumed within-subject correlation used in crossover designs from 0.2 to 0.8 left the overall effect significant in every scenario, with pooled estimates ranging from 0.156 to 0.282. Leave-one-out analyses, in which each study is removed in turn, shifted the headline estimate only between 0.182 and 0.215. Egger’s regression test found no evidence of publication bias, and risk-of-bias assessment using the revised Cochrane RoB 2 tool flagged most concerns in the domains of randomization reporting and selective outcome reporting rather than in outcome measurement itself. Still, the authors concede that most included studies carried “some concerns” or a high risk of bias, and that the predominance of male participants — 359 men against 31 women across the dataset — limits generalizability.

Where does this leave the athlete standing in the locker room with ten minutes to spare? The evidence suggests caffeine gum is a genuinely practical option: fast-acting, portable, dosed flexibly, and associated in some research with fewer side effects such as gastrointestinal distress and sleep disturbance than liquid or capsule forms. The optimal recipe emerging from this analysis is roughly 2 to 4 mg/kg of caffeine, chewed for a full ten minutes, taken shortly before exercise, by a trained competitor who is not already a heavy daily caffeine user. For recreational gym-goers, the data offer little encouragement that gum will translate into measurable gains. And for everyone, the authors’ own caveats apply: the underlying trials are imperfect, the evidence certainty is modest, and heavy consumers may be chasing a benefit their nervous system has already learned to ignore. But as a last-minute legal stimulant strategy, the science now says the gum is more than marketing — it is a modest, reproducible, and race-day-relevant advantage.