Research dossier
Clinical research on Caffeine
11 trials reviewed across 8 indications.
Strongest evidence
Exercise performance and muscular endurance
Mechanism
Caffeine antagonizes adenosine A1 and A2A receptors centrally, lowering perceived effort, and acts at peripheral neuromuscular sites to enhance motor-unit recruitment. Catecholamine release and fat-oxidation effects contribute marginally; the central perception-of-effort mechanism is the dominant ergogenic pathway.
Caffeine is the most-replicated ergogenic aid in sports science. Pooled meta-analyses across 60+ RCTs show 2–3% endurance improvements and small-to-moderate muscular endurance gains at 3–6 mg/kg taken 30–60 minutes pre-exercise. Strength signal is smaller but consistent.
Effects are largest in non-habitual users; habitual high-caffeine athletes show attenuated responses. 'No-cycle' debates remain unsettled — most performance benefit persists in habitual users, but the magnitude shrinks.
Trials cited
Caffeine and resistance-exercise performance (meta-analysis)
positive · Meta-analysis
Grgic et al., 2020, British Journal of Sports Medicinen=30021 RCTs pooled. Caffeine produced small but consistent improvements in muscular strength (SMD ~0.20) and a larger effect on muscular endurance (SMD ~0.38). Upper-body endurance responded more than lower-body strength. Most-replicated ergogenic aid in exercise science.
Effects shrink in habitual high-caffeine users. Lower-body strength signal is the weakest of the four endpoints pooled.
Caffeine and endurance performance (meta-analysis)
positive · Meta-analysis
Southward, Rutherfurd-Markwick, Ali, 2018, Sports Medicinen=300Pooled across 46 endurance RCTs, caffeine improved performance by ~2–3% — meaningful at the competitive level. Effects were larger in time-to-exhaustion than time-trial protocols and larger in cyclists than runners.
Effect sizes attenuated in habitual high-intake athletes. Optimal ergogenic dose is 3–6 mg/kg; higher doses do not produce additional benefit but raise GI and CNS side-effect risk.
Vigilance and cognitive performance
Mechanism
Adenosine accumulates during wakefulness and signals sleep pressure via A1 and A2A receptors. Caffeine competitively blocks these receptors, restoring vigilance toward rested baseline. Indirect dopaminergic effects support attention and motivation.
Caffeine reliably restores vigilance, reaction time, and sustained attention in sleep-deprived adults (Lieberman 2002). In well-rested non-habitual users, alertness gains are smaller but real; in habitual users, much of the 'sharpening' effect reflects reversal of overnight withdrawal rather than absolute enhancement.
Largest effect in sleep-deprived populations and non-habitual users. Habitual users mostly experience withdrawal-reversal rather than enhancement. CYP1A2 slow metabolizers may experience cognitive jitter at doses that fast metabolizers tolerate cleanly.
Caffeine and cognition under sleep deprivation (Navy SEAL trial)
positive · RCT
Lieberman et al., 2002, Psychopharmacologyn=68In severely sleep-deprived SEAL trainees, 200 mg caffeine restored vigilance, reaction time, and reported alertness toward baseline. The 200 mg and 300 mg doses outperformed 100 mg; benefits persisted at least 8 hours. Establishes the operational case for caffeine in sleep-deprived populations.
Effect size in well-rested adults is substantially smaller. This is a sleep-deprivation cognition trial, not a general-cognition trial.
Net cognitive effect of caffeine in habitual consumers (withdrawal-reversal critique)
mixed · RCT
Rogers et al., 2013, Psychopharmacologyn=379Habitual consumers showed apparent alertness gains from caffeine — but the gain largely reflected reversal of overnight withdrawal, not absolute cognitive enhancement. Non-consumers showed minimal alertness benefit and more anxiety. Reframes the popular 'caffeine makes you sharper' claim as 'caffeine restores you to the baseline you had before you became dependent.'
The withdrawal-reversal model is contested. Other trials show small but real alertness effects in non-deprived non-habitual users — the magnitude is modest, not transformative.
Sleep disruption and half-life
Mechanism
Caffeine's plasma half-life is 4–6 hours in average adults but ranges 2 to 12+ hours by CYP1A2 genotype, smoking status (induces CYP1A2), oral contraceptives (inhibit CYP1A2), and pregnancy (markedly slows clearance). A 200 mg afternoon dose can still be at ~50 mg in plasma at bedtime.
Drake 2013 polysomnography showed that 400 mg caffeine 6 hours before bed reduced total sleep time by ~1 hour and impaired sleep efficiency — even when subjects didn't notice. Caffeine is not just hard on sleep onset; it suppresses slow-wave sleep and shifts sleep architecture across the whole night.
Slow metabolizers, oral contraceptive users, and pregnant women clear caffeine far more slowly. For these groups, even early-afternoon caffeine can compromise sleep architecture. Pragmatic guidance: stop caffeine 8–10 hours before bedtime if sleep matters.
Caffeine 0, 3, and 6 hours before bedtime
positive · RCT
Drake et al., 2013, Journal of Clinical Sleep Medicinen=12Small but rigorous PSG-monitored trial. 400 mg caffeine taken 6 hours before bedtime reduced total sleep time by ~1 hour and disrupted sleep efficiency — even when participants did not perceive an effect subjectively. Confirms caffeine's 5-hour half-life pushes its sleep impact well into the evening for afternoon-coffee drinkers.
Small sample (n=12) but objectively measured. Half-life varies 2–4× by CYP1A2 genotype, smoking status, and oral contraceptive use — sleep impact of late caffeine varies similarly.
Cardiovascular: acute pressor effect with tolerance
Mechanism
Acute caffeine antagonizes adenosine-driven vasodilation, producing a transient rise in systolic and diastolic blood pressure. Chronic exposure induces partial tolerance to the pressor effect. CYP1A2 polymorphisms determine clearance rate and thus the duration of pressor exposure per dose.
Acute caffeine (200–300 mg) raises BP ~6–8 mmHg systolic for several hours. Habitual coffee drinkers develop substantial tolerance; large pooled cohort data on habitual coffee are null or modestly cardioprotective. The Cornelis CYP1A2 *1F MI signal flags slow metabolizers as the exception — they may face real cardiovascular risk at heavy intakes.
Honest read: habitual moderate coffee is not a cardiovascular concern for most adults. CYP1A2 *1F slow metabolizers (~50% of adults) drinking ≥2–3 cups/day are the population with a documented MI signal. Anyone with uncontrolled hypertension, arrhythmias, or recent cardiac events should treat caffeine conservatively.
Caffeine and blood pressure (meta-analysis of RCTs)
mixed · Meta-analysis
Mesas et al., 2011, American Journal of Clinical Nutritionn=1000Acute caffeine (200–300 mg) raised systolic BP by ~8 mmHg and diastolic by ~6 mmHg for several hours. Chronic coffee consumption produced much smaller BP effects (often null) — tolerance develops within days for the pressor response. The honest read: an acute jolt of caffeine raises BP transiently; habitual coffee does not meaningfully raise it.
Tolerance to the BP effect is well-established but incomplete in some hypertensives. Pure caffeine (anhydrous, energy drinks) raises BP more than equivalent coffee — coffee polyphenols partially offset the pressor effect.
CYP1A2 genotype, coffee, and myocardial infarction
positive · Observational
Cornelis et al., 2006, JAMAn=4028Carriers of the CYP1A2 *1F slow-metabolizer allele (~50% of adults) drinking ≥2–3 cups/day had a 36–64% higher MI risk than rapid metabolizers with similar intake. Fast metabolizers showed no excess risk, and possibly a slight protective signal. First clean demonstration that caffeine's cardiovascular risk profile is genotype-dependent.
Observational case-control. Confounding by smoking (a CYP1A2 inducer) is hard to fully eliminate. Replicated in some but not all subsequent cohorts; effect is most evident in younger adults with high intake.
Anxiety and stress reactivity
Mechanism
Adenosine antagonism elevates noradrenaline, cortisol, and sympathetic tone. In sensitive individuals and at higher doses, this manifests as somatic anxiety symptoms (palpitations, tremor, restlessness) and amplified subjective stress reactivity. ADORA2A and CYP1A2 polymorphisms moderate sensitivity.
Doses >300–400 mg/day raise anxiety scores in sensitive individuals, particularly in non-habitual users and slow metabolizers. The DSM-5 recognizes Caffeine-Induced Anxiety Disorder as a clinical entity. Below ~200 mg/day, anxiety effects are typically minor and dose-dependent.
Anxiety-prone adults, panic-disorder patients, and ADORA2A C-allele carriers report disproportionate anxious response to caffeine. Reducing intake below 200 mg/day or eliminating it entirely is a first-line intervention before reaching for anxiolytic medication.
Net cognitive effect of caffeine in habitual consumers (withdrawal-reversal critique)
mixed · RCT
Rogers et al., 2013, Psychopharmacologyn=379Habitual consumers showed apparent alertness gains from caffeine — but the gain largely reflected reversal of overnight withdrawal, not absolute cognitive enhancement. Non-consumers showed minimal alertness benefit and more anxiety. Reframes the popular 'caffeine makes you sharper' claim as 'caffeine restores you to the baseline you had before you became dependent.'
The withdrawal-reversal model is contested. Other trials show small but real alertness effects in non-deprived non-habitual users — the magnitude is modest, not transformative.
Headache analgesic adjunct and withdrawal management
Mechanism
65–130 mg caffeine added to acetaminophen, ibuprofen, or aspirin produces a small but real boost in acute headache relief — mechanism involves adenosine antagonism, mild cranial vasoconstriction, and improved analgesic absorption. Separately, abrupt caffeine cessation in habitual users reliably triggers withdrawal headache via rebound vasodilation.
Cochrane 2014 pooled 19 trials: 65–130 mg caffeine added to OTC analgesics raises the proportion reaching meaningful 2-hour pain relief by ~5–10%. Juliano & Griffiths 2004 established that ~50% of habitual users develop withdrawal headache on abrupt cessation, peaking 24–48 hours and resolving in ~9 days. Tapering over 1–2 weeks largely prevents this.
Habitual analgesic use that contains caffeine carries a medication-overuse-headache risk. The honest practice: caffeine is useful as an acute headache adjunct; chronic daily use of caffeine-containing analgesics is its own problem.
Caffeine as an adjunct in OTC analgesics for headache (Cochrane review)
positive · Systematic review
Derry et al., 2014, Cochrane Database of Systematic Reviewsn=7000Cochrane pooled 19 trials. Adding 65–130 mg caffeine to a standard OTC analgesic dose produced a small but real boost in headache relief — approximately 5–10% absolute improvement in the proportion reaching meaningful pain relief. Establishes caffeine's role as an analgesic adjunct, not a standalone treatment.
Meaningful efficacy in acute headache only — separate from rebound and withdrawal headaches caused by habitual high intake, which are a chronic-care problem.
Caffeine withdrawal characterization (systematic review)
positive · Systematic review
Juliano and Griffiths, 2004, Psychopharmacologyn=4000Pooled 57 studies. Headache occurs in ~50% of habitual users on abrupt cessation; fatigue, dysphoria, and impaired concentration are nearly universal. Symptoms can occur after habitual intake as low as 100 mg/day. Established the empirical basis for Caffeine Withdrawal Disorder, now recognized in DSM-5.
Symptom severity correlates with usual daily intake and abruptness of cessation. Tapering over 1–2 weeks largely prevents withdrawal headache.
Bone density and fracture risk
Mechanism
High acute caffeine intake produces a small transient increase in urinary calcium excretion. The 1980s–90s concern that habitual caffeine therefore eroded bone density has been progressively walked back as larger and longer cohorts have failed to find clinically meaningful fracture excess in calcium-replete adults.
Hallström 2013 (n=61,433, 22-year follow-up) showed no fracture risk increase with high coffee intake in women with adequate calcium. Modest BMD reductions of 2–4% have been reported but do not translate into excess fracture. The clean read: caffeine is not an osteoporosis driver in calcium-sufficient adults.
Adults with low calcium intake (<800 mg/day) plus heavy caffeine intake (>400 mg/day) are the only subgroup with a residual fracture-risk signal in some analyses. Fix the calcium gap first, not the coffee.
Coffee consumption and fracture risk
Null · Observational
Hallström et al., 2013, American Journal of Epidemiologyn=6143361,433 women followed 22 years. High coffee intake (≥4 cups/day) showed a modest BMD reduction (~2–4%) but no increase in fracture risk vs low intake (<1 cup/day). Pooled with other modern cohorts, the 1980s–90s panic that 'coffee causes osteoporosis' has not survived large prospective data in calcium-replete adults.
Effect modification by calcium intake: low-calcium plus high-caffeine groups showed a small fracture-risk signal in some sub-analyses. The clean read is 'fine in calcium-sufficient adults; pay attention if your calcium intake is also low.'
Thermogenesis and fat oxidation
Mechanism
Caffeine acutely raises resting energy expenditure by ~3–11% via sympathomimetic effects and stimulates lipolysis. The thermogenic effect tolerizes substantially with habitual use, leaving only a small residual increase in metabolic rate at habitual doses.
Acute caffeine produces small but measurable thermogenic and fat-oxidation effects. Translation to meaningful weight loss in trials is weak — the metabolic boost is real but small (typically 50–100 kcal/day in habitual users) and easily offset by caloric compensation. Caffeine is not a weight-loss intervention.
Acute responders are non-habitual users and lean adults. Habituated heavy users get essentially no thermogenic benefit at typical daily doses. Treat metabolic claims as ergogenic-aid context, not weight-loss claims.
Caffeine and endurance performance (meta-analysis)
positive · Meta-analysis
Southward, Rutherfurd-Markwick, Ali, 2018, Sports Medicinen=300Pooled across 46 endurance RCTs, caffeine improved performance by ~2–3% — meaningful at the competitive level. Effects were larger in time-to-exhaustion than time-trial protocols and larger in cyclists than runners.
Effect sizes attenuated in habitual high-intake athletes. Optimal ergogenic dose is 3–6 mg/kg; higher doses do not produce additional benefit but raise GI and CNS side-effect risk.
Are you deficient? Symptoms, risk groups, lab tests
Caffeine is not a nutrient and has no deficiency syndrome. Roughly 85% of US adults consume caffeine daily (mean ~165 mg/day, FDA dietary survey data); roughly 40–50% of habitual users meet operational criteria for caffeine dependence. The clinically relevant phenomenon is withdrawal in habitual users, not deficiency.
Common symptoms
- Withdrawal headache (typically frontal/diffuse, vascular character)
- Fatigue and low energy 12–24 hours after last dose
- Dysphoria, irritability, or mild depressed mood
- Difficulty concentrating and reduced alertness
- Drowsiness and increased sleep need
- Flu-like symptoms in severe withdrawal (nausea, muscle aches)
- Reduced motivation and psychomotor slowing
- Cravings for coffee, tea, or caffeinated beverages
Who is at risk
Heavy daily coffee or energy-drink consumers
Habitual intake ≥200 mg/day reliably upregulates adenosine receptors and produces withdrawal on abrupt cessation. Withdrawal severity correlates with usual daily dose and chronicity.
Adults attempting elimination diets or pregnancy reduction
Going from 300+ mg/day to zero overnight produces 24–48h headache and several days of fatigue. Tapering by 25–50 mg per week largely prevents withdrawal.
Hospitalized adults with abrupt access loss
Surgery, hospitalization, or NPO status can produce unexpected post-op headache and dysphoria in habitual users. Recognized often enough to be a named clinical entity (perioperative caffeine withdrawal).
CYP1A2 *1F slow metabolizers
Slow metabolizers carry caffeine longer; abrupt cessation produces a longer-tail withdrawal. They are also the population with the documented MI risk signal at heavy intake (Cornelis 2006).
Adults on oral contraceptives
OC use inhibits CYP1A2 by ~50%, doubling caffeine half-life. Withdrawal symptoms can be delayed and prolonged relative to non-OC users.
Pregnant women in third trimester
Pregnancy slows CYP1A2 markedly — caffeine clearance falls to ~30% of pre-pregnancy. Withdrawal from previously habitual intake can be severe; tapering before conception is the safest path.
Adolescents with energy-drink habits
Energy drinks deliver caffeine at higher concentration and faster than coffee, accelerating dependence. AAP and FDA flag energy drinks as inappropriate for adolescents.
Lab markers
There is no routine clinical lab for caffeine status
Serum caffeine and paraxanthine can be measured but are only used in research, perinatology (preterm-infant apnea protocols), or specific toxicology. CYP1A2 genotyping is available commercially but is rarely indicated clinically for the average adult.
Better:CYP1A2 genotyping (research/elective use), Polysomnography for caffeine-attributable sleep disturbance
Side effects and drug interactions
Side effects
Insomnia and disturbed sleep architecture
Common · Any dose within 8–10 hours of bedtime; particularly above 200 mg
Caffeine's 4–6 hour half-life means afternoon doses still occupy adenosine receptors at bedtime. Reduces total sleep time and suppresses slow-wave sleep even when subjects don't perceive disruption.
Anxiety, jitteriness, and restlessness
Common · Threshold varies; common above 300–400 mg/day in habitual users, lower in non-habitual or sensitive individuals
Adenosine antagonism raises sympathetic tone, noradrenaline, and cortisol. Manifests as tremor, palpitations, and subjective anxiety. ADORA2A C-allele carriers and slow metabolizers are over-represented in sensitive responders.
Gastrointestinal upset and acid reflux
Common · Symptomatic at typical 100–200 mg doses on empty stomach in sensitive adults
Caffeine relaxes the lower esophageal sphincter and stimulates gastric acid secretion. Can trigger reflux, dyspepsia, and loose stools — particularly on empty stomach.
Tachycardia and palpitations
Uncommon · Notable above 200–300 mg acutely; energy-drink bolus doses are the highest-risk format
Sympathomimetic effects raise heart rate transiently. In susceptible individuals, can precipitate or worsen atrial ectopy and supraventricular arrhythmias.
Transient blood pressure elevation
Uncommon · Acute doses ≥200 mg; clinically meaningful BP elevation rare with habitual coffee in normotensives
Acute caffeine raises systolic BP by ~6–8 mmHg and diastolic by ~4–6 mmHg for several hours. Tolerance develops with habitual intake but is incomplete in some hypertensives.
Caffeine toxicity (severe)
Severe · Toxic >1 g acutely; lethal range 5–10 g (adults)
Acute intake ≥1,000 mg can cause vomiting, severe anxiety, arrhythmias, seizures, and rhabdomyolysis. Doses ≥10 g are potentially fatal. Pure caffeine powder products and concentrated tablets are the dominant culprits — FDA has taken enforcement action against bulk powder sellers.
Worse with:caffeine anhydrous powder, concentrated caffeine tablets >200 mg, energy shots
Energy-drink cardiac events in adolescents and young adults
Severe · Acute intake ≥400 mg, especially rapid consumption or combined with exercise/alcohol
Case reports and case-control series document arrhythmia, MI, and sudden death in adolescents and young adults after high-acute-dose energy-drink intake. AAP and FDA both issue advisories against adolescent energy-drink use.
Worse with:energy drinks, concentrated caffeine shots, pre-workout blends
Drug interactions
Combined-effect risk
lithiumCaffeine increases renal lithium clearance. Abrupt change in caffeine intake — particularly cessation — can raise lithium levels and risk toxicity.
Hold caffeine intake steady; do not suddenly cut or surge caffeine without prescriber awareness. Monitor lithium level if intake changes.
Combined-effect risk
theophyllineCaffeine and theophylline are both methylxanthines and share CYP1A2 metabolism. Co-administration produces additive CNS and cardiovascular stimulation and slows theophylline clearance — narrow-therapeutic-index drug.
Avoid significant caffeine intake in adults on theophylline. Theophylline serum levels should guide dosing if any caffeine exposure is unavoidable.
Combined-effect risk
MAO inhibitors (phenelzine, tranylcypromine, isocarboxazid, moclobemide)MAOIs prevent catecholamine breakdown. Caffeine's sympathomimetic load can precipitate hypertensive crisis or severe anxiety in MAOI-treated patients.
Limit caffeine to low-moderate intake and use cautiously on MAOIs. Avoid energy drinks and high-dose pre-workout caffeine entirely.
Reduces nutrient status
oral contraceptives (estrogen-containing)fluvoxamineciprofloxacincimetidineThese agents inhibit CYP1A2, the primary caffeine-clearance enzyme. Caffeine half-life can double, producing exaggerated stimulant, anxiety, and sleep-disruption effects from previously well-tolerated doses.
Halve typical caffeine intake when starting these medications; reassess tolerance after a week. Avoid late-day caffeine entirely in this combination.
Additive effect
smoking (CYP1A2 inducer)Tobacco smoke induces CYP1A2 and roughly doubles caffeine clearance. Smokers metabolize caffeine fast; on smoking cessation, caffeine half-life can double within a week — producing apparent 'caffeine sensitivity' and insomnia that is actually unchanged caffeine intake on a slower-metabolizing baseline.
Adults quitting smoking should pre-emptively cut caffeine intake by 30–50% during the first 1–2 weeks of cessation to avoid amplified caffeine effects.
Combined-effect risk
clozapineolanzapineduloxetineThese drugs are CYP1A2 substrates. High caffeine intake competes for CYP1A2 metabolism and can raise plasma levels of these drugs — clinically meaningful for clozapine in particular.
Hold caffeine intake steady on these medications. Sudden caffeine cessation can substantially raise plasma clozapine levels and risk toxicity.
Combined-effect risk
adenosine, dipyridamole, regadenoson (cardiac stress agents)Caffeine competitively antagonizes the adenosine receptors that pharmacologic stress agents act on. Recent caffeine intake blunts or invalidates pharmacologic cardiac stress testing.
Hold all caffeine (including decaf, chocolate, tea) for 12–24 hours before adenosine or dipyridamole stress testing.
Other
stimulant ADHD medications (methylphenidate, amphetamines)ephedrine, pseudoephedrineAdditive sympathomimetic and CNS-stimulant effects. Raises heart rate, blood pressure, anxiety, and arrhythmia risk above either agent alone.
Reduce or eliminate caffeine when initiating or titrating stimulant medication. Monitor pulse and BP in combination use.
Other critical caveats
- Pregnancy: ACOG recommends limiting caffeine to less than 200 mg/day. Maternal caffeine clearance falls dramatically in third trimester, and observational data link higher intake to fetal growth restriction, miscarriage, and low birth weight. Discuss intake with your OB — total caffeine from coffee, tea, chocolate, energy drinks, and OTC analgesics all count.
- Children and adolescents: the American Academy of Pediatrics recommends against routine caffeine consumption in this group, with energy drinks specifically flagged as inappropriate. Adolescent cardiac events after acute high-dose energy-drink intake are documented in case reports.
- CYP1A2 *1F slow metabolizers (~50% of adults) drinking ≥2–3 cups/day showed a 36–64% higher MI risk in the Cornelis 2006 case-control. Anyone with personal or family cardiac history and heavy caffeine intake should consider CYP1A2 testing or simply reduce intake.
- Caffeine powder and concentrated caffeine tablets are responsible for the documented deaths from caffeine toxicity. Even one teaspoon of pure powder approaches the toxic threshold. The FDA has taken enforcement action against bulk-powder sellers — avoid concentrated caffeine products entirely; stick to standardized capsules, coffee, or tea.
- Withdrawal is an expected feature of habitual use, not a sign of harm. ~50% of habitual users develop withdrawal headache on abrupt cessation. Taper by 25–50 mg per week to prevent withdrawal — particularly important before pregnancy, surgery, or major caffeine restriction.
Frequently asked
How much caffeine is safe for healthy adults?
The FDA cites 400 mg/day as the typical upper threshold for healthy adults without adverse effects — roughly 4 cups of coffee or 10 cans of cola. Pregnancy: ≤200 mg/day per ACOG. Adolescents: AAP recommends avoidance. Sensitive individuals, slow CYP1A2 metabolizers, oral-contraceptive users, and adults with anxiety, arrhythmias, or hypertension may need substantially less.What's the best dose for exercise performance?
Pooled meta-analyses converge on 3–6 mg/kg body weight, taken 30–60 minutes pre-exercise. For a 75 kg adult that's 225–450 mg. Higher doses don't produce additional ergogenic benefit and substantially raise GI and CNS side-effect risk. Effect size is largest in non-habitual users; habitual heavy users see attenuated but still real benefit.When should I stop drinking caffeine before bed?
Caffeine's half-life is 4–6 hours for average metabolizers, longer for slow CYP1A2 carriers, oral contraceptive users, and pregnant women. Drake 2013 showed that 400 mg taken 6 hours before bedtime reduced total sleep time by ~1 hour even when subjects didn't notice. Pragmatic guidance: stop caffeine 8–10 hours before bedtime. If you sleep poorly, try cutting after noon entirely.How do I taper off caffeine without getting a headache?
Roughly 50% of habitual users develop withdrawal headache on abrupt cessation. The reliable approach is tapering by 25–50 mg per week — e.g. drop one cup, or mix half-decaf, then half-half, then full decaf. Total taper time of 2–4 weeks largely prevents withdrawal. Hydration, ibuprofen for breakthrough headache, and avoiding scheduled-medication conflicts (especially lithium) all help.Is caffeine bad for my bones?
Modern data say no, with one caveat. The 1980s–90s concern that habitual coffee erodes bone density has not held up in large prospective cohorts. Hallström 2013 followed 61,433 women for 22 years and found no fracture-risk increase with high coffee intake in adults with adequate calcium. The exception: adults consuming <800 mg/day calcium plus >400 mg/day caffeine show a small residual fracture-risk signal. Fix the calcium intake first.Why does caffeine seem to give me more anxiety than other people?
Two genes do most of the explaining. CYP1A2 *1F slow metabolizers carry caffeine 2–4× longer than fast metabolizers — same cup, more exposure. ADORA2A C-allele carriers have adenosine receptors that respond more strongly to blockade — same exposure, more anxious effect. Oral contraceptives slow CYP1A2 by ~50%. None of this is fixable except by lowering the dose. If 200 mg makes you anxious where peers handle 400 mg, your biology is real — drink less.Is decaf actually caffeine-free?
No. Decaf coffee typically contains 2–15 mg per cup — about 5–10% of regular coffee. For an average drinker this is negligible. For pregnant women near the 200 mg ceiling, slow metabolizers, or adults preparing for cardiac stress testing, decaf still counts. For stress testing specifically, eliminate all caffeine — including decaf, chocolate, and tea — for 12–24 hours.
References
- 01FDA — Spilling the Beans: How Much Caffeine is Too Much?
- 02ACOG Committee Opinion 462 — Moderate Caffeine Consumption During Pregnancy
- 03American Academy of Pediatrics — Sports drinks and energy drinks for children and adolescents
- 04StatPearls — Caffeine (NCBI Bookshelf)
- 05Juliano LM, Griffiths RR, 2004, Psychopharmacology — A critical review of caffeine withdrawal
- 06Cornelis et al., 2006, JAMA — Coffee, CYP1A2 genotype, and risk of myocardial infarction
- 07Grgic et al., 2020, British Journal of Sports Medicine — Wake up and smell the coffee: caffeine supplementation and exercise performance
Last reviewed2026-05-22