Coffee Sensitivity: Caffeine Jitters, Stomach Upset and Your Genes

One cup of coffee leaves some people shaky, anxious or wide awake at 2 a.m., while others drink espresso after dinner and sleep fine. Some people feel fine on caffeine but get heartburn or a rushed trip to the bathroom from coffee itself. Both get called "coffee sensitivity," but they are usually two different problems. Caffeine sensitivity shows up as jitters, anxiety, a racing heart or poor sleep. A reaction to other compounds in coffee mostly shows up in the stomach and gut.

This guide covers how much caffeine is really in your cup, why some people clear it far more slowly than others, what the "slow metabolizer" gene tests can and cannot tell you, what the newest heart trials show about palpitations, why decaf can still upset your stomach, and whether L-theanine or other supplements take the edge off.

Caffeine sensitivity or coffee intolerance?

Caffeine works mainly by blocking adenosine, a brain chemical that builds up during the day and makes you feel sleepy. That is why its effects are mostly about alertness, nerves, heart rate and sleep. Coffee also contains hundreds of other compounds, including acids, chlorogenic acids and compounds formed during roasting, that act on the stomach and bowel whether or not the caffeine has been removed.

The FDA lists the classic signs of too much caffeine as a faster heart rate, palpitations, higher blood pressure, insomnia, anxiety, jitters, upset stomach, nausea and headache. Here is a rough guide to which part of coffee is the more likely cause:

A simple way to tell them apart: switch to decaf for a week. If the jitters stop but the stomach symptoms continue, caffeine was behind the jitters and something else in coffee is behind the stomach trouble. To double-check, try tea or another caffeinated drink that is not coffee: if the jitters come back but your stomach stays calm, that points the same way.

How much caffeine is in your cup

Many people who call themselves sensitive are simply getting more caffeine than they think. Agency figures give a sense of the range:

Even those ranges understate the variation. In a University of Florida study, specialty coffees ranged from 58 to 259 mg per serving, and the same coffee bought from the same outlet on six days in a row ranged from 259 to 564 mg. A study of 104 espressos from cafes in Scotland, Italy and Spain, plus some cappuccinos and instant coffees, found 48 to 317 mg per serving, and its authors concluded that "one cup of coffee" is not a reproducible measurement.

For most healthy adults, the FDA cites 400 mg a day, about two to three 12 ounce cups of coffee, as "an amount not generally associated with negative effects," while adding that people vary widely in how sensitive they are and how fast they clear it. Europe's food safety agency, EFSA, says single doses up to 200 mg raise no safety concern for healthy adults, but that even 100 mg close to bedtime may affect sleep in some people. (The 2017 review the FDA cites as confirming the 400 mg figure was sponsored by an industry-supported food science group that also received grants from beverage and coffee trade associations; the figure itself comes from a 2003 Health Canada review, and EFSA reached the same number in 2015.)

Why caffeine hits some people harder

Almost all caffeine is broken down in the liver by one enzyme, CYP1A2, which handles about 95 percent of it. How fast that enzyme works varies a lot from person to person, and so does caffeine's half-life, the time it takes your body to clear half of the caffeine you drank:

• Healthy adults: averages of roughly 3.5 to 6 hours in small studies; in the largest of them, 25 adults ranged from 2 to 5 hours. That means a coffee at 3 p.m. can still leave half its caffeine in your blood at 8 or 9 p.m.
• Pregnancy: an average of 8.3 hours in one study of pregnant women, up to 16 hours in some, in most women returning to normal within a month after delivery.
• Birth control pills: in a small study, women taking oral contraceptives had an average half-life of 7.9 hours versus 5.4 hours in women who were not.
• Quitting smoking: smoking speeds up CYP1A2. When heavy smokers quit, caffeine clearance fell by about a third, so the usual coffee can suddenly feel much stronger.
• Medicines: the antidepressant fluvoxamine is a strong CYP1A2 blocker. In one trial it stretched caffeine's half-life from 4.9 to 56 hours. At the upper end of its usual doses, the antibiotic ciprofloxacin made caffeine leave the body about 1.5 times more slowly, and the older heartburn drug cimetidine also slows it. If a new prescription coincides with new jitters, ask your pharmacist.
• Liver disease: two patients with alcoholic liver disease had half-lives of 60 and 168 hours in an early study.

Tolerance matters too. Regular drinkers become far less sensitive to caffeine's anxiety effect. In one study of 379 adults, frequent consumers developed substantial tolerance even when they carried a gene variant linked to caffeine anxiety, and the researchers found that daily users gained no net alertness: after an overnight break, caffeine mostly just reversed the dip in alertness caused by going without it. People who rarely drink coffee, or who go back to it after a break, tend to notice its effects most.

Slow vs fast caffeine metabolizers and the CYP1A2 test

Twin studies suggest that roughly a third to over half of the differences between people in caffeine-related traits are inherited. Large gene studies of coffee intake point mainly to two spots: the CYP1A2 gene for the liver enzyme, and AHR, a control gene on a different chromosome that regulates CYP1A2. Each gene variant shifts intake by only a fraction of a cup a day. People who clear caffeine slowly tend to drink less of it, which suggests most people adjust their intake to how coffee makes them feel.

The CYP1A2 test. Consumer DNA tests usually report a variant called rs762551. You carry two copies of each gene, one from each parent, so your result (your genotype) reads AA, AC or CC: AA is labeled "fast," and AC or CC "slow." What the research shows is narrower than the marketing:

• The variant mainly changes how much the enzyme speeds up when something switches it on, such as smoking or heavy coffee drinking. In non-smokers, caffeine breakdown did not differ significantly between the three genotypes in one study, and a 2018 analysis of 3,570 people found the effect was much larger in smokers.
• Slower clearance means caffeine lasts longer, not necessarily that each cup feels stronger. Even when fluvoxamine cut clearance about tenfold, a single 250 mg dose of caffeine did not produce stronger effects on alertness or performance. A longer half-life matters more if you drink several cups a day, because caffeine can build up.
• In PubMed searches through November 2026, we found no study showing that your CYP1A2 type predicts how jittery a cup of coffee makes you.

The ADORA2A gene, which codes for the brain's adenosine receptor, has a better link to how caffeine feels. In a double-blind trial (neither the volunteers nor the researchers knew who got caffeine) of 94 light caffeine users, people with one version reported more anxiety after 150 mg, and several later studies found similar links with anxiety and caffeine-related sleep problems. But at a high dose of 450 mg, most people became anxious (the gene links were seen at the moderate 150 mg dose), and one study of 1,023 workers found the variant pointed the opposite way for sleep complaints among people taking in under 300 mg a day. A 2023 systematic review summed it up: CYP1A2 variants were linked mainly to cognitive effects, while ADORA2A variants were linked to anxiety and sleep disturbance.

The heart attack finding that did not hold up. A 2006 Costa Rican study that compared 2,014 heart attack survivors with 2,014 similar people who had not had a heart attack found that slow metabolizers who drank 2 or more cups a day had higher odds of a nonfatal heart attack (64 percent higher at 4 or more cups), while fast metabolizers did not. It was widely reported and is still used to sell gene tests. But it was observational, meaning it compared groups of people rather than testing coffee in a trial, so it could not show cause and effect. Later, much larger analyses did not confirm it: in 347,077 UK Biobank participants, CYP1A2 type showed no interaction with coffee for heart disease, a study of 386,258 people found no interaction for heart rhythm problems, and a 2023 genetic analysis found no strong link between genetically higher caffeine levels and heart disease.

Is a caffeine gene test worth it? The lab will usually read the gene correctly. What the result means for how you feel, or for your heart, is much less certain. A 12 month trial that gave young adults diet advice based on their genes, including a caffeine gene, did not report any change in caffeine intake (its senior author held shares in a nutrition gene testing company). Your own response to a known dose of caffeine is a better guide than your genotype.

Why coffee can make you anxious

Caffeine's effect on anxiety is clearly dose related. In a double-blind study of 102 light caffeine users given 0, 50, 150 or 450 mg, the lowest dose did not raise anxiety, while 450 mg raised it in most people. At low to moderate doses, caffeine often lifts mood and alertness instead.

People with panic disorder are a special case. A 2022 meta-analysis (a study that pools the results of earlier studies) of blinded caffeine challenge tests, in which patients were given a test dose without knowing whether it was caffeine or a dummy pill, found that 51.1 percent of 237 patients with panic disorder had a panic attack after a large dose of caffeine (400 to 750 mg, most studies using 480 mg, roughly five cups), and none after placebo. Only about 2 percent of healthy volunteers did. But in a 2025 trial from the same research group, 150 mg, about one strong cup, did not raise anxiety more than placebo in people with panic disorder. The participants were light caffeine users, and caffeine still raised arousal and avoidance behavior in both groups; the authors suggested that advice to avoid caffeine in panic disorder should be based on higher doses and ideally on individual assessment.

If coffee makes you anxious, the practical steps are to drink less at a time, avoid it on an empty stomach or when already stressed, and try a lower-caffeine drink such as tea. For broader options, see our guide to supplements for anxiety and stress.

Caffeine and sleep

Caffeine's effect on sleep lasts longer than most people expect. In a home sleep study, 400 mg of caffeine taken 6 hours before bed still disrupted sleep, and the authors recommended avoiding substantial caffeine for at least 6 hours before bedtime. A 2023 meta-analysis of 24 studies found that caffeine cut total sleep by an average of 45 minutes, made it 9 minutes slower to fall asleep and reduced deep sleep. Its authors calculated that a typical cup of coffee (107 mg) should be finished at least 8.8 hours before bed, and a standard pre-workout serving (217.5 mg) at least 13.2 hours before bed, to avoid losing sleep.

Sleep loss from coffee is easy to miss: in the 2023 heart trial described below, participants slept 36 minutes less on coffee days, measured by a wrist tracker. If you sleep poorly, moving your last cup to before noon is a free experiment worth trying before any sleep supplement. See also our guide to the best supplements for sleep.

Does coffee cause heart palpitations?

It depends on what kind of palpitation. The best evidence comes from the 2023 CRAVE trial, published in the New England Journal of Medicine. One hundred healthy adults wore a continuous heart monitor for 14 days, and text messages told them at random each day whether to drink coffee or avoid caffeine. The trial was not blinded.

• Extra beats from the upper chambers (premature atrial contractions), the trial's main measure, did not rise significantly: 58 versus 53 a day.
• Extra beats from the lower chambers (premature ventricular contractions, often felt as a skipped or "flip-flop" beat) rose from 102 to 154 a day. In people who had only one coffee drink a day, the analysis did not suggest a link; the rise was seen with more.
• Episodes of fast heart rhythm from the upper chambers were similar on coffee and no-coffee days.
• On coffee days, people also walked about 1,058 more steps and slept 36 minutes less.

Larger studies are reassuring about the rhythm problems people worry about most. In 386,258 UK Biobank participants, each additional daily cup of coffee was linked to a 3 percent lower risk of developing an arrhythmia, an irregular heart rhythm (an observational finding). In the 2026 DECAF trial of 200 current or recent coffee drinkers with persistent atrial fibrillation (a common irregular rhythm that starts in the upper chambers) who had their normal rhythm restored, those assigned to drink coffee (about a cup a day) had fewer recurrences over six months than those told to abstain (47 versus 64 percent). Everyone in that trial knew which group they were in (it was open label), and it did not include people without the condition. A 2026 American Heart Association scientific statement summarized the trials this way: among regular coffee drinkers, caffeinated coffee "decreases the risk of atrial fibrillation occurrence but increases the frequency of premature ventricular contractions." It also noted that data on very high doses, such as those in energy drinks, are limited and "generally suggest cardiovascular harm."

Blood pressure. In people with high blood pressure, 200 to 300 mg of caffeine raised it by about 8 points systolic (the top number) and 6 points diastolic (the bottom number), starting within an hour and lasting at least 3 hours. But studies of regular coffee drinking over two weeks found no increase, so for regular drinkers the long-term effect on blood pressure appears small. For more on skipped beats, see our article on magnesium for heart palpitations.

Why coffee upsets your stomach (and makes you poop)

This is where "coffee sensitivity" often has little to do with caffeine.

Heartburn and reflux. Coffee stimulates stomach acid, and small lab studies show it can trigger reflux, but caffeine is not the main reason. In 16 healthy volunteers, regular coffee caused significant reflux, decaf caused much less, and, in a subgroup of 8, caffeine added to water had no effect. In 17 people with reflux disease, the share of time the esophagus was acidic after breakfast was 17.9 percent with regular coffee versus 3.1 percent with decaf. Yet a classic 1975 study found decaf still stimulated a lot of stomach acid.

Studies that follow large groups of people are mixed: a 2014 meta-analysis of 15 studies found no significant link between coffee and reflux disease, while a 2026 meta-analysis of 40 studies found a small increase of "unclear clinical significance." In a large study of US nurses, reflux symptoms were 34 percent more likely at more than six servings of coffee a day than at none, swapping two daily servings for water was linked to slightly lower risk, and tea and soda showed similar links. The fair conclusion is that coffee triggers heartburn in some people but not most, and switching to decaf, drinking less, or not drinking it on an empty stomach are reasonable experiments. See our guide to supplements for GERD and acid reflux.

The "coffee makes me poop" effect. It is real and not imagined. In a survey of 99 young adults, 29 percent said coffee gave them an urge to have a bowel movement. In 8 such people tested in the lab, both regular and decaf coffee increased activity in the lower bowel within four minutes, an effect that lasted at least 30 minutes and did not happen with hot water. Another study found caffeinated coffee stimulated the colon about as much as a 1,000 calorie meal, 60 percent more than water and 23 percent more than decaf. For people with sensitive bowels or irritable bowel syndrome (IBS), that can mean loose stools, though studies of whether coffee triggers IBS give conflicting results. See probiotic strains for IBS for related evidence.

Low-acid coffee, cold brew and dark roast. These are sold as gentler on the stomach, but the evidence is thin:

• Cold brew: in a lab comparison, cold brew and hot brew had similar acidity as measured by pH (4.85 to 5.13; the lower the number, the more acidic), but hot brew contained more total acid. In PubMed searches through November 2026, we found no human study testing whether cold brew causes less heartburn.
• Dark roast: roasting longer creates a compound that lowered acid production in stomach cells grown in the lab, and one small study in nine volunteers found a dark roast stimulated less stomach acid than a regular blend.
• "Low-acid" labels: we could not find a standard definition of low-acid coffee, so the claim means whatever the seller says it means.

None of these has been shown to reduce symptoms in people with reflux, but they are harmless to try.

Can you be allergic to coffee?

True coffee allergy is real but uncommon, and most of it involves breathing in dust rather than drinking coffee. Workers handling green (unroasted) coffee beans can become sensitized, meaning their immune system starts reacting to the dust: in one Italian plant, 25.8 percent of green-coffee workers were, compared with 2.7 percent of roasted-coffee workers. In another Italian coffee plant, castor bean antigens were found in warehouse dust, and castor bean, not coffee, was the major cause of workers' sensitization.

Allergy to drinking brewed coffee appears to be rare; the reports we found are individual case reports. Flavored coffee drinks and creamers can contain milk, nuts, soy or other allergens, so check the label if you react to some coffee drinks but not to plain black coffee. Hives, lip or throat swelling, or wheezing after coffee calls for an allergist, not a home experiment.

Mold-free coffee. Some brands are marketed as free of mycotoxins, the toxins made by certain molds. Ochratoxin A is often found in coffee at low levels: a 2019 meta-analysis of 3,182 samples worldwide found an average of 3.21 micrograms per kilogram, with large differences between countries. The EU sets a legal limit of 3 micrograms per kilogram for roasted coffee and 5 for instant coffee; the pooled average mixes product types and countries (from 0.35 in Taiwan to 79 in Turkey), so it does not mean typical coffee breaks the EU limit. We found no evidence that "mold-free" coffee reduces jitters, anxiety or stomach symptoms.

L-theanine, magnesium and other supplements

L-theanine, an amino acid from tea, is the supplement most often suggested for "smoother" caffeine. Trials of the two together found small improvements in attention and alertness in the first two hours, mostly in healthy young adults given single doses, and several were run by researchers at Unilever, which then owned Lipton tea. A 2014 meta-analysis of 11 trials of tea compounds found trends suggesting the benefit tracked the caffeine dose more than the theanine dose, and a 2015 trial at tea-like doses (75 mg caffeine with 50 mg theanine) found the combination cancelled caffeine's effects on attention and mood rather than adding to them.

The key question for sensitive people is whether theanine reduces jitters. The most direct test gave 48 adults 250 mg of caffeine, 200 mg of theanine, both or neither: theanine blunted caffeine's rise in blood pressure but "did not significantly affect jitteriness." It is reasonable to try and generally well tolerated, but the evidence that it takes the edge off caffeine is weak. (Branded forms such as Suntheanine and AlphaWave L-Theanine have their own pages.)

Paraxanthine, the main compound the body turns caffeine into, is now sold (for example as enfinity) as a smoother alternative. Its published safety data are mostly rat studies, one by the company that sells enfinity and one paid for by another company. We found no human trial showing it causes fewer jitters than caffeine: a 1995 study in 12 people found it raised blood pressure and adrenaline about as much as caffeine at the higher dose tested, while a small 2026 trial in 14 rowers reported better self-rated sleep than with caffeine. Our article on paraxanthine vs caffeine covers it in detail.

Magnesium. Caffeine slightly increases magnesium in urine for a few hours. In a metabolic ward study, two doses of caffeine raised 24 hour magnesium losses by about 4 mg, a tiny amount next to the 310 to 420 mg adults need daily. In PubMed searches through November 2026, we found no trial testing magnesium supplements for caffeine jitters or sensitivity.

Iron and calcium timing. Coffee reduced iron absorption from a hamburger meal by 39 percent in one study. Coffee drunk an hour before the meal had no effect, but coffee an hour after reduced absorption just as much as coffee with the meal. That study measured iron from food, but the same timing logic is a sensible precaution for iron supplements: take them an hour or more before coffee rather than with or soon after it. Caffeine's effect on calcium absorption is very small; one review estimated that 1 to 2 tablespoons of milk offset it.

Hidden caffeine in supplements. It is an easy cause of "sudden" caffeine sensitivity to miss. Guarana, yerba mate, kola nut and green tea extracts all contain caffeine. A government analysis of 53 supplements containing such ingredients found that taking the maximum label dose would provide anywhere from 1 to 829 mg of caffeine a day, and about half the products did not state a caffeine amount. In a study of sports supplements tested repeatedly over nine months, caffeine content in five of six caffeinated products varied from 7 percent less to 266 percent more than the first measurement. Pre-workouts and fat burners can easily double your daily intake on top of coffee (see do fat burners work?). The FDA warns that one teaspoon of pure powdered caffeine can contain as much caffeine as 28 cups of coffee, and that these products have contributed to at least two deaths in the US.

Caffeine withdrawal: how to cut back without the headache

Caffeine withdrawal is real. A major review found that symptoms typically start 12 to 24 hours after the last dose, peak at 20 to 51 hours and last 2 to 9 days. Headache occurred in half of people in experimental studies, along with fatigue, low energy, drowsiness, low mood, irritability and trouble concentrating, and stopping doses as low as 100 mg a day caused symptoms. In a blinded study of 62 adults switched to placebo, 52 percent had a moderate or severe headache.

• Taper gradually. The FDA advises cutting back gradually, and a trial of a self-guided program that cut caffeine step by step over six weeks reduced intake and caffeine-related distress, with results lasting to 20 weeks. We found no trial comparing tapering with quitting abruptly.
• Try half-caf. Mixing regular and decaf beans lets you drop the dose without changing the ritual.
• Swap the source. A 12 ounce green tea has about a third of the caffeine of the lowest estimate for a 12 ounce coffee. Chicory root and grain "coffees" on their own contain no caffeine, but many mushroom coffees are blended with real coffee, so check the label. For ideas, see supplements for energy without caffeine.
• Decaf is not zero. Decaf typically has 2 to 15 mg per 8 ounce cup; one study found up to 15.8 mg in a decaf espresso shot. That is fine for most people but adds up if you are very sensitive.

A note on decaf solvents. Some decaf is made with methylene chloride, which US rules allow at up to 10 parts per million in the finished coffee. Advocacy groups petitioned the FDA in December 2023 to revoke that use; according to trade press, the FDA reopened public comments, which closed in June 2026, and we found no final decision as of November 2026. A 2024 EPA ban on most uses of methylene chloride did not cover food. Decaf made by water or carbon dioxide methods avoids the issue if it concerns you.

Who should limit caffeine

• Pregnancy: caffeine lasts longer during pregnancy. ACOG considers under 200 mg a day moderate and not a major factor in miscarriage or preterm birth, and EFSA says up to 200 mg a day raises no safety concern for the fetus. See supplements during pregnancy for hidden caffeine sources.
• Panic disorder or anxiety: large doses trigger panic in about half of people with panic disorder; smaller amounts may be tolerated.
• Insomnia: even 100 mg near bedtime may affect sleep in some people.
• Heart rhythm problems or high blood pressure: individual advice from your doctor matters more than averages.
• Children and teens: the American Academy of Pediatrics says energy drinks have no place in their diet. EFSA suggests no more than 3 mg per kilogram of body weight a day for children and teens; Health Canada suggests 2.5 mg per kilogram for children.
• People taking interacting medicines such as fluvoxamine, ciprofloxacin or estrogen-containing birth control, or who recently quit smoking. See our guide to supplement and drug interactions.

Frequently asked questions

What are the symptoms of caffeine sensitivity?

Common signs are jitters or shaky hands, restlessness, anxiety, a fast or pounding heartbeat, skipped beats, trouble sleeping, upset stomach, nausea and headache, often after an amount of caffeine that other people tolerate. The FDA lists the same symptoms as signs of consuming too much caffeine.

Why does coffee make me jittery?

Usually because of the caffeine dose, how quickly your body clears caffeine, and how used to it you are. A 12 ounce coffee can contain about 113 to 247 mg of caffeine, and people who rarely drink it, are pregnant, take birth control pills or certain medicines, or have recently quit smoking clear caffeine more slowly or feel it more.

Why does coffee make you anxious?

Mostly because of the caffeine dose. In a double-blind study of light caffeine users, the lowest dose (50 mg) did not raise anxiety, while 450 mg raised it in most people. People with panic disorder react more to large doses: in blinded studies, about half had a panic attack after 400 to 750 mg, though 150 mg did not raise anxiety more than placebo in a 2025 trial. Drinking less at a time is the simplest fix.

Why does coffee make you poop?

Coffee stimulates the lower bowel, partly through compounds other than caffeine. In a survey of 99 young adults, 29 percent said coffee gave them an urge to have a bowel movement, and when 8 of them were tested in the lab, both regular and decaf coffee increased lower bowel activity within four minutes. In another study, caffeinated coffee stimulated the colon 23 percent more than decaf.

Is coffee sensitivity the same as a coffee allergy?

No. Most coffee sensitivity is a reaction to caffeine or to coffee's effects on the stomach and bowel, not an immune reaction. True allergy to drinking coffee appears to be rare and is described mostly in individual case reports. Hives, swelling or wheezing after coffee should be checked by an allergist.

Are you a slow caffeine metabolizer, and is the CYP1A2 gene test worth it?

A CYP1A2 gene test can tell you whether you carry the variant linked to slower caffeine clearance, but the variant mainly affects how much caffeine clearance speeds up in smokers and heavy coffee drinkers, not how jittery you feel. Large studies did not confirm that slow metabolizers get more heart disease from coffee. How you respond to a known dose is a better guide.

Why can coffee upset your stomach, even decaf?

Coffee contains acids and other compounds that stimulate stomach acid and bowel movement regardless of caffeine. In small studies, decaf caused less reflux than regular coffee but still stimulated acid, and both regular and decaf coffee stimulated the lower bowel in people prone to that effect.

Is low-acid or cold brew coffee better for a sensitive stomach?

Possibly for some people, but it is unproven. In a lab study, cold brew had a similar pH to hot brew and less total acid. We found no human study testing whether cold brew causes less heartburn, and we could not find a standard definition of low-acid coffee.

Does coffee cause heart palpitations?

It can increase one kind. In a 2023 randomized trial of 100 healthy adults, coffee days brought more extra beats from the lower chambers of the heart (154 versus 102 a day) but no significant rise in extra beats from the upper chambers or episodes of fast rhythm. Large studies have not linked moderate coffee to more arrhythmias. Chest pain, fainting or palpitations that do not settle need medical care.

Does L-theanine stop caffeine jitters?

The evidence is weak. L-theanine with caffeine produced small, short-lived improvements in attention in lab studies, several run by researchers at Unilever, which then owned Lipton tea. In the most direct test of jitteriness, 200 mg of L-theanine blunted the blood pressure rise from 250 mg of caffeine but did not significantly reduce jitteriness.

How long does caffeine stay in your system?

Caffeine's half-life averages roughly 3.5 to 6 hours in healthy adults, so half of an afternoon coffee can still be in your blood in the evening. It is longer in pregnancy (about 8 hours on average in one study), with birth control pills, liver disease, some medicines, and after quitting smoking. A 2023 meta-analysis suggested finishing a typical coffee at least 8.8 hours before bed.

How can you cut back on caffeine without withdrawal headaches?

Reduce gradually over a few weeks, for example by mixing regular and decaf beans or swapping some cups for tea. Withdrawal symptoms typically start 12 to 24 hours after the last dose and last 2 to 9 days. The FDA recommends cutting back gradually, and a six week step-down program reduced intake in a trial.

How much caffeine is safe if you are sensitive?

There is no single safe number for sensitive people. The FDA cites 400 mg a day as generally not associated with negative effects in most healthy adults, EFSA says single doses up to 200 mg raise no concern, and ACOG considers under 200 mg a day moderate in pregnancy. Sensitive people often do best well below these amounts, with the last cup early in the day.

The bottom line

"Coffee sensitivity" is usually two separate things. Jitters, anxiety, palpitations and poor sleep come mainly from caffeine, and the biggest factors are dose, timing, tolerance and anything that slows caffeine clearance, such as pregnancy, birth control pills, quitting smoking or certain medicines. Stomach and bowel symptoms come partly from other compounds in coffee, which is why decaf can still cause them.

Gene tests read real genes but predict less than they promise, and the heart risk claims based on them were not confirmed in large studies. L-theanine has small effects on attention but did not reduce jitters in the most direct test, and we found no trial supporting magnesium for caffeine sensitivity. The most useful steps cost nothing: measure your actual dose, check supplements for hidden caffeine, finish your last cup at least 6 hours before bed, and ideally about 9 for a typical cup, taper slowly rather than quitting abruptly, and use decaf to work out whether the problem is the caffeine or the coffee.