Caffeine and Endurance Performance: What the Evidence Supports

Caffeine occupies an unusual position in sports nutrition. Unlike most supplements marketed to endurance athletes, its ergogenic effect is not contested. What remains genuinely worth examining is dose, timing, and a source of individual variation most general guidance ignores. A specific genetic difference in how the body metabolises caffeine can turn an effective protocol into an ineffective, or even counterproductive, one for a meaningful share of athletes.

This article covers the evidence for caffeine’s endurance benefit and the dose range that supports it. It also covers the genetic variation that changes the picture for a substantial minority of athletes.

---

The Baseline Evidence

The ergogenic effect of caffeine on endurance performance is well established across decades of research. An umbrella review covering 21 published meta-analyses found caffeine ingestion produces an acute ergogenic effect. This spans aerobic endurance, muscular strength, muscular endurance, power, jumping performance, and exercise speed. Few supplements in sports nutrition carry this breadth of consistent supporting evidence.

For cycling time trials specifically, a meta-analysis found caffeine intake produced a statistically significant improvement in performance across pooled trials. A separate controlled study on a 10-kilometre cycling time trial found 4 milligrams per kilogram reduced completion time by 3 percent relative to placebo. That is a meaningful margin in a competitive endurance context.

---

---

Where the Story Complicates: CYP1A2

The consistent average effect across a population conceals substantial individual variation. The leading explanation is a single genetic polymorphism in the CYP1A2 gene, which governs how quickly the liver metabolises caffeine. Athletes carry one of three genotype variants. AA is associated with fast metabolism, AC with intermediate metabolism, and CC with slow metabolism.

A meta-analysis pooling 13 studies and 440 participants examined how caffeine’s ergogenic effect differed across these genotypes. Caffeine improved performance meaningfully in the AA group and to a smaller degree in the AC group. In the CC group, caffeine worsened performance. This is a striking result: for a meaningful minority of athletes, a standard caffeine protocol may not simply fail to help. It may actively hinder the performance it was intended to enhance.

The controlled cycling time trial study referenced above found the same pattern in its own data. The overall 3 percent improvement at 4 mg/kg was driven almost entirely by the AA genotype group. They improved by nearly 7 percent at that dose. Athletes with the AC genotype showed no meaningful effect. Athletes with the CC genotype saw performance decline at the higher dose tested.

---

---

Why This Matters Practically

Most caffeine guidance given to endurance athletes assumes a uniform population response. This assumption is where general advice tends to break down for a specific subset of athletes. An athlete who has tried caffeine before competition and reported feeling flat, jittery without benefit, or simply unaffected, is not necessarily doing something wrong in dose or timing. They may fall within a genotype group for whom the standard protocol does not apply, or applies in reverse.

Genetic testing for CYP1A2 status is not yet standard practice for most athletes and coaches. It remains a reasonable long-term consideration for anyone whose caffeine response has been inconsistent or counterproductive despite conventional dosing. In the absence of genetic testing, the more accessible approach is systematic self-experimentation. Test a standard protocol in low-stakes training sessions, track subjective and objective response, and treat an unhelpful or negative result as genuine data, not a dosing error to fix by taking more.

---

Timing and Metabolism Speed

A separate line of research examined how CYP1A2 status interacts with timing, independent of the genotype-and-dose interaction described above. This research found fast metabolisers achieved their strongest endurance benefit when caffeine was consumed roughly one hour before exercise. Slow metabolisers, by contrast, showed a stronger benefit when caffeine was consumed roughly two hours beforehand.

This finding is consistent with the underlying metabolic mechanism. A slow metaboliser clears caffeine from the bloodstream more gradually. Peak blood concentration, and the performance window tied to it, arrives later relative to ingestion than it does for a fast metaboliser. A single fixed timing recommendation applied uniformly across a roster will be well-matched for some athletes and poorly matched for others. This holds independent of whether the dose itself was appropriate.

---

Practical Framework

For an athlete without access to genetic testing, the standard evidence-supported range remains a reasonable default. 3 to 6 milligrams per kilogram of body weight, consumed 60 to 90 minutes before competition. This range and timing window reflects where most positive findings across the literature cluster. Most athletes carry the AA or AC genotype rather than the less common CC variant, and will see a genuine benefit here.

What the CYP1A2 research adds is a diagnostic question. Has this athlete tried the standard protocol and found it ineffective or counterproductive? Rather than assuming the athlete needs a higher dose or different timing, the more useful first step is treating their reported experience as informative. A consistent pattern of no benefit, or a negative response, across multiple honest attempts is meaningful data. It is not noise to be dosed around. For that subset, the evidence-supported answer may be a lower dose, a different timing window, or, in a genuine minority of cases, no caffeine at all.

---

References

1. Umbrella review of 21 meta-analyses on caffeine ingestion and acute ergogenic effects across sport performance domains. J Int Soc Sports Nutr. 2020. DOI: 10.1186/s12970-020-00349-6
2. Guest N, Corey P, Vescovi J, El-Sohemy A. Caffeine, CYP1A2 genotype, and endurance performance in athletes. Med Sci Sports Exerc. 2018;50(8):1570-1578. DOI: 10.1249/MSS.0000000000001596
3. Caffeine, CYP1A2 genotype, and exercise performance: a systematic review and meta-analysis. Med Sci Sports Exerc. 2024. DOI: 10.1249/mss.0000000000003313
4. Does ergogenic effect of caffeine supplementation depend on CYP1A2 genotype? A systematic review with meta-analysis. J Sport Health Sci. 2023. DOI: 10.1016/j.jshs.2023.11.001
5. Aji YG, Melita S, Dijaya R, Subali D, Kartawidjajaputra F, Suwanto A. Evaluation of caffeine ingested timing on endurance performance based on CYP1A2 rs762551 profiling in healthy sedentary young adults. Rep Biochem Mol Biol. 2023;11(4):663-671.