The guide

Your genes and caffeine

By GyroscopeUpdated November 15, 202615 sources

Two genes get most of the attention. CYP1A2 makes the liver enzyme that clears caffeine, so it changes how fast it leaves you. ADORA2A makes one of the brain receptors caffeine blocks, so it changes how strongly you feel it. Both are real. But in the studies that matter most to daily life, like sleep and long-term health, they made less difference than a DNA report suggests, and the pill, pregnancy and smoking move your half-life far more.

A gene report that says “slow metabolizer” feels like a final answer, and it explains a familiar puzzle. One friend has an espresso after dinner and sleeps fine, while another has a 2pm latte and lies awake. It’s tempting to put that down to one letter in your DNA. Some of it is genes. More of it is things you can name and test.

Two genes, two questions

They’re different questions. Clearing caffeine slowly and being sensitive to it can go together or not. Someone who clears it at an average speed can still sleep badly after a small cup, and someone who clears it slowly can sleep fine.

There’s also a third way your genes show up. A study of more than 120,000 coffee drinkers found several gene regions, including one near CYP1A2, linked to how many cups people drink a day. Each gene variant moved it by only 0.03 to 0.14 cups. The researchers took it as a sign that caffeine itself shapes how much coffee people choose to drink.

What the CYP1A2 studies found

Heart attacks. In 2006 a study in Costa Rica compared 2,014 people who’d had a first heart attack with 2,014 who hadn’t. Among slow metabolizers, four or more cups of coffee a day went with a 64% higher risk. Among fast metabolizers, coffee showed no higher risk at all. The difference was biggest under age 59.

Blood pressure. In 2009 a study followed 553 young adults with mild high blood pressure for about 8 years. Slow metabolizers who drank coffee were more likely to end up needing blood pressure medicine, while fast metabolizers who drank a lot were less likely to.

Death from any cause. This is where the gene story gets much weaker. In the UK Biobank, 498,134 people were followed for 10 years. Coffee went with a lower risk of dying across every genetic caffeine score, slow and fast, including at 6 or more cups a day. The same was true for tea in the same cohort, and a separate UK Biobank study found coffee went with fewer heart rhythm problems whatever people’s CYP1A2 gene said.

The two early studies were smaller and looked at one outcome each. The big later studies looked at hundreds of thousands of people and didn’t find the gene turning coffee into a risk. We trust the big studies more, though the gene may still matter for blood pressure in some people. If yours is high, does coffee raise your blood pressure? has what to watch.

Genes and sleep

The 2025 sleep trial tested both genes. Its 23 young men had 100 or 400 mg of caffeine at 12, 8 or 4 hours before bed, or a placebo, and the researchers checked their CYP1A2 and ADORA2A variants. There was no significant effect of either gene on measured sleep or on how they said they slept. The researchers note the trial wasn’t built to detect gene effects, so a small one could have been missed.

ADORA2A has the stronger sleep evidence. In a 2007 study, people who described themselves as caffeine sensitive had a different mix of ADORA2A variants from people who didn’t, and the variant shaped how much caffeine changed their brain waves during sleep. In a 2008 study of 102 light caffeine users, the same ADORA2A variant went with feeling more anxious after 150 mg.

Genes and training

A 2018 study of 101 male athletes found caffeine cut 10 km cycling time by 4.8% at 2 mg per kg and 6.8% at 4 mg per kg, but only in the AA, or fast, group. The AC group showed no effect, and the few CC athletes got slower.

A 2024 review pooled 13 studies with 440 people. Caffeine helped AA (233 people) and AC (175 people) and hurt performance a little in CC (34 people). When the authors left out studies whose authors reported conflicts of interest, the gene made no difference. Caffeine for training and focus has the doses for everyone.

What a gene test can and can’t tell you

It can tell you which CYP1A2 and ADORA2A variants you carry, and that’s real information about your enzyme and your receptors.

It can’t tell you your half-life. That depends on much more than one variant, and some of the biggest shifts have nothing to do with your DNA:

It also can’t tell you how caffeine affects your sleep. In the one trial that checked, it didn’t predict it.

Test yourself instead

Your own nights are the direct test, and they cover your genes, your hormones and your habits at once.

1. Keep your caffeine the same for a week and note how you sleep, or track it with a ring or watch.
2. Then move your last cup at least 3 hours earlier, or make it smaller, for a week.
3. Compare. If the second week is better, your afternoon cup was costing you, whatever your genes say.

The half-life tool shows how far the pill, pregnancy or smoking moves your cutoff, and what makes yours faster or slower explains each one. The quiz, Is my coffee keeping me up?, sets up the two weeks for you.