Caffeine improves exercise performance … if you have the right genes Original paper

In this meta-analysis of randomized controlled trials, caffeine improved exercise performance in participants categorized as having “fast” or “intermediate” caffeine-metabolizing genes but reduced performance in participants with “slow” caffeine-metabolizing genes.

The acute effect of caffeine on exercise performance among people with different genotypes (gene variants) for caffeine metabolism.

A total of 440 adult men and women, including untrained people and athletes.

A meta-analysis of 16 randomized controlled trials was conducted. All studies provided the participants with an acute dose of caffeine prior to exercise and reported at least one of the following outcomes related to exercise performance: total work done, average power output, average velocity, average height jumped, and time to completion (exercise performance) or time to exhaustion (exercise capacity).

Caffeine’s effect on exercise performance was evaluated separately for three single nucleotide polymorphisms (SNPs) of the CYP1A2 gene (specifically segment rs762551) related to caffeine metabolism, namely:

• AA: known as fast caffeine metabolizers
• AC: known as intermediate caffeine metabolizers
• CC: known as slow caffeine metabolizers

The influence of caffeine dose and timing were also investigated.

Overall effects of caffeine: Caffeine improved exercise performance for fast metabolizers (small effect size) and intermediate metabolizers (very small effect size). For slow metabolizers, caffeine reduced performance (small effect size).

Effect of caffeine on exercise performance according to CYP1A2 genotype

Effect of dose: Higher doses of caffeine were better for performance in slow metabolizers (i.e., performance was worse with lower doses). There was no relationship between caffeine dose and performance for fast and intermediate metabolizers.

Effect of timing: A longer time between caffeine ingestion and exercise was better for performance in slow metabolizers (i.e., performance was worse with shorter time between caffeine ingestion and exercise). There was no relationship between caffeine timing and performance for fast and intermediate metabolizers.

The benefits of caffeine on exercise performance are widely known and supported by many studies, including an umbrella review of meta-analyses published in 2020 that found performance-enhancing effects of caffeine for aerobic endurance, muscle strength, muscular endurance, power output, jumping performance, and speed.^[1] Caffeine doesn’t appear to discriminate when it comes to modes of exercise.

Interestingly — and common to many supplements and interventions — studies have shown that not everyone benefits from caffeine’s ergogenic effects. In other words, although some people get a performance boost from a cup of coffee, others don’t, and some might even experience worse performance.

One plausible explanation relates to differences in how individuals metabolize caffeine. In humans, a gene known as cytochrome p450 1A2 (also known as CYP1A2) is responsible for metabolizing about 95% of all ingested caffeine.^[2] There are three alleles of the rs762551 segment of the CYP1A2 gene — commonly known as AA, AC, and CC — each of which affect the rate of caffeine metabolism. An allele is one of two or more versions of a DNA sequence at a given location on a gene. Specifically, people with the AA genotype are said to be “fast” metabolizers, and people with the AC and CC genotypes are said to be “intermediate” and “slow” metabolizers, respectively. Because several metabolites of caffeine (e.g., paraxanthine, theophylline) are thought to mediate caffeine’s ergogenic effects, in theory, fast caffeine metabolism will result in a quicker, more intense ergogenic effect, and slow(er) metabolism results in the opposite. Interestingly, caffeine metabolism genes don’t seem to modify the beneficial association between coffee consumption and the risk for all-cause mortality^[3] and cardiovascular disease.^[4]

Several randomized controlled studies have supported this hypothesis. In one study, ingesting 6 mg of caffeine per kilogram of body weight (mg/kg) improved time trial performance more in AA carriers (who experienced a 4.9% improvement) compared to C allele carriers (AC/CC genotypes) who only improved by 1.8%.^[5] Another study found a 3% improvement in cycling performance when male athletes ingested 4 mg/kg of caffeine prior to a 10-kilometer time trial. AA carriers improved by 4.8% with 2 mg/kg and by 6.8% with 4 mg/kg, whereas those with the CC genotype experienced a 13.7% decline in performance.^[6] Finally, CC carriers experienced a 12.8% reduction in handgrip strength after ingesting 4 mg/kg of caffeine compared to a placebo, but no difference in strength was observed for AA and AC carriers.^[7]

In contrast, some other studies have shown that caffeine improves performance irrespective of genotype.^{[8][9][10][8]}

Given the somewhat conflicting results, a comprehensive evaluation of caffeine’s effects among different genotype carriers was needed. Although more research is certainly required, the results of the current meta-analysis suggest that caffeine is ergogenic in AA carriers and ergolytic (impairing performance) in CC carriers, with variable effects seen in AC carriers.

Unfortunately, the effects of caffeine on different modes of exercise weren’t evaluated, so it can’t be concluded that caffeine is best for runners, cyclists, weight lifters, or soccer players. However, the included studies used a wide range of exercise tests: cycling time trials, 30-second Wingate power tests, handgrip strength, jumping ability, ball throwing, agility, sprinting, and general muscular strength and power. Thus, the results are generalizable to exercise performance per se. However, even among the fast metabolizers, the benefits of caffeine seem to be modest based on reported effect sizes. In other words, caffeine won’t make an Olympian out of a weekend warrior.

Unique to this meta-analysis was an evaluation of caffeine’s effect according to studies that reported a conflict of interest (COI). In general, studies funded in part or completely by private companies tend to have a higher chance of showing positive results and proceeding to publication. The authors of many of the included studies worked for or invested in genetic testing companies or the supplement industry. Do COIs influence the results for caffeine studies?

When factoring COI into the meta-analysis results, CC genotypes no longer experienced a performance decrease after caffeine ingestion — studies with COIs actually reported worse outcomes, on average, for CC genotypes. Excluding these studies abolished the observed caffeine-genotype interaction on exercise performance and the influence of dose and timing. In other words, the interaction between genes, caffeine, and exercise performance seemed to be influenced by results of studies in which a COI was reported.

The conclusions of this meta-analysis should not be thrown out due to the potential biases, but this finding does suggest that COIs have a substantial effect on the literature. A COI doesn’t automatically imply a bias, but may call for greater scrutiny and transparency.

It’s probably too soon to declare that all athletes should run to the closest genetic profiling center to get their CYP1A2 genotype determined. Nor does the evidence suggest that athletes who do know their genotype should modify their caffeine consumption habits based on their carrier status. However, there’s probably no harm in knowing whether you’re an AA, AC, or CC, and doing some self-experimentation with caffeine dosing and timing.

Until the field of nutrigenomics advances, the best advice is to enjoy your coffee or preferred caffeine-containing beverage whenever and however you like.

One of the main limitations of this meta-analysis — and really all studies related to the genetics of caffeine metabolism and performance — is that slow metabolizers/CC genotypes are underrepresented. Only 32 of the 440 participants in this meta-analysis were CC genotypes, which may reflect their small makeup of the population (some studies suggest around 10%). More studies including a larger number of participants with a CC genotype are needed to provide a more balanced analysis of the interaction between caffeine, genetics, and exercise performance.