In one pass Coffee wakes you up because caffeine blocks the signal in your brain that makes you feel sleepy; the tiredness itself has not gone anywhere.
Educational content, not medical advice — consult a clinician.
Coffee wakes you up because caffeine blocks the signal in your brain that makes you feel sleepy; the tiredness itself has not gone anywhere. The chapter on how caffeine blocks sleepiness walks through that step.
Your cup starts as a roasted seed, and hot water carries whatever dissolves out of it into the cup. Green beans have almost none of the coffee aroma; the smell and the dark color come from roasting, when sugars and amino acids go through the Maillard reaction.
What really changes the health picture is how you brew it: whether a paper filter catches two cholesterol-raising molecules in the bean's oils.
A CLOSER LOOK
The cup begins with a roasted seed
An unsweetened black-coffee example is shown; neither roast darkness nor brew colour directly reports caffeine content.
1The bean is a seed
A coffee bean is the seed inside a coffee cherry, roasted and ground for brewing.
2Hot water carries it into the cup
Brewing carries extractable molecules into the cup; sugar and milk are additional ingredients.
Illustration for understanding; not to scale. Saved figures include explanations and sources.
Mechanism · Brewing decides what reaches the blood
Coffee comes mainly from two species, arabica and robusta. The seed inside the coffee cherry is the bean, and what you drink is an infusion: roasted seeds, ground and extracted with hot water.
Three points that often get mixed up:
Arabica is milder and more acidic; robusta has more caffeine and a harsher taste, and it is common in instant coffee and blendsRoast level changes flavor, not caffeine content much, so a dark roast does not necessarily have more caffeineDrip, pour-over, French press, espresso, and boiled coffee differ in more than taste. A paper filter traps oily molecules such as cafestol; a French press and boiled coffee do not. The chapter on why filtered coffee is kinder to cholesterol explains that switch Coffee is among the most-studied parts of the diet, and among the most misunderstood: on one side the old idea that "coffee is harmful", on the other the new hype that "coffee is a superfood". Both need pulling back to the middle.
Chapter 2
How caffeine blocks sleepiness
Caffeine does not give you extra energy; it blocks the signal that makes you sleepy.
With every waking hour, adenosine builds up in the brain and acts like a brake that slows nerve activity. Caffeine looks enough like adenosine to sit in the same receptor, but it sends no signal; it just occupies the seat. With sleepiness shut out, you feel alert.
It does not remove tiredness; it only delays when you feel it. How long caffeine stays in your body decides whether an afternoon cup cuts into your sleep.
Mechanism · It takes the seat but does not push
Adenosine binds to adenosine receptors on nerve cells, damps their activity, and makes you steadily drowsier. It is one of the ways the body keeps track of how long it has been since you slept.
Caffeine can sit in the same receptor without activating it, which keeps the real adenosine out; that is what an "adenosine-receptor antagonist" means (Jacobson 2022). With the sleepiness brake occupied, you feel awake and alert. For how caffeine pairs with theanine, see Caffeine + L-Theanine; for how it hides sleep debt during an All-Nighter, see All-Nighter.
The key point: adenosine keeps building up. Once the caffeine has been broken down, the sleepiness it was holding back returns all at once.
Numbers · A half-life of about 4 to 5 hours
In healthy adults the elimination half-life of caffeine is usually 4 to 5 hours (IOM 2001), meaning the amount in your body halves over that span. A cup at noon still has about half its caffeine working by early evening. That is why the time of your last cup affects sleep more easily than how many cups you drink; the cutoff itself is in the chapter on how much to drink and when.
Besides caffeine, coffee contains chlorogenic acids and other polyphenols. They are the active candidates most often discussed in antioxidant and metabolic research on coffee, and they may be involved in the diabetes association described later.
Coffee's health reputation has flipped once. It was once suspected of causing cancer and harming the heart; after separating out such as smoking, large observational studies found that people who drink moderate amounts have the same or slightly lower risk of death, which is a long way from a superfood.
Nearly all of this is association, and it cannot show that coffee is the cause. It is enough to say that moderate coffee is safe for most healthy adults, not enough to say it is worth drinking for your health. If you do not drink coffee, this association is no reason to start.
Numbers · Around 3 to 4 cups, not more is better
Poole 2017 pooled many into an umbrella review (a review of reviews): compared with none, about 3 to 4 cups a day was associated with lower all-cause mortality ( about 0.83) and lower cardiovascular mortality. The curve is not a straight line; the benefit is clearest around 3 to 4 cups, and more is not better.
Ding 2014 pooled 28 prospective cohorts with more than 1.1 million people: the more coffee people drank, the lower their risk of type 2 diabetes, with a relative risk of about 0.91 for each extra cup a day. Decaf showed the same association, which suggests caffeine is not the only player and chlorogenic acids may be involved. Both findings are observed associations and cannot prove that coffee is the cause. For the blood-sugar side of the mechanism, see Type 2 Diabetes & Prediabetes.
Safety · Coffee no longer rated a possible carcinogen
The International Agency for Research on Cancer (IARC) once put coffee in its "possibly carcinogenic" group, then re-evaluated it and moved it to "not classifiable" (IARC Monograph Vol. 116 / Loomis 2016): there was no adequate evidence that coffee itself causes cancer.
The same evaluation found that what is actually linked to esophageal cancer is very hot drinks (above 65°C), which IARC classed as "probably carcinogenic". That is about temperature, not coffee. "Don't drink it scalding" is real advice; "coffee causes cancer" is not.
Chapter 4
Does coffee dehydrate or stunt growth?
Coffee carries several claims that have circulated for decades. Each starts from a real fact that got blown up into "harmful".
Caffeine really is a mild diuretic, but people who drink coffee regularly do not get dehydrated from it. A slight rise in calcium lost in urine is real, but it does not add up to stunted growth. And the old impression that coffee causes cancer came largely from not separating out the effect of smoking.
Evidence · In regular drinkers, coffee hydrates like water
Caffeine really is a mild diuretic, and large doses slightly increase urination. Killer 2014 tested this directly in a crossover trial: regular coffee drinkers spent one period drinking coffee that delivered about 4 mg/kg of caffeine a day and another drinking the same volume of water, and the researchers compared 24-hour urine output and several hydration markers. There was no difference.
By this trial, for regular drinkers a moderate amount of coffee hydrates about as well as water and can count toward daily fluids. Drinking extra water to "make up for" coffee is unnecessary.
Myth · Stunted growth and heart harm
There is no reliable human evidence that coffee stunts children's growth or dissolves bone. The idea may come from caffeine slightly raising the calcium lost in urine, but that loss is tiny, a splash of milk makes it up, and it does not make people shorter or cause osteoporosis. What actually shapes bones and height is overall nutrition, protein, calcium, vitamin D, and sleep. It is reasonable that children do not need coffee, but the reasons are sleep and palpitations, not height.
Heart harm and cancer are exactly the part that reversed, as the chapter on moderate coffee and health describes: moderate coffee is associated with the same or slightly lower all-cause and cardiovascular mortality, and IARC has taken coffee off its "possibly carcinogenic" list. For healthy adults, what deserves attention is individual sensitivity, pregnancy, the dose, and the time of day you drink it.
Chapter 5
Why filtered coffee is kinder to LDL
The same coffee means different things for your blood lipids depending on whether it went through a paper filter.
The oils in coffee beans contain cafestol and its close relative kahweol, and cafestol raises what is often called "bad cholesterol", low-density lipoprotein cholesterol (). A paper filter holds most of them back, so drip and pour-over coffee barely raise LDL; a French press and boiled coffee have no paper, so these molecules go straight into the cup.
People whose LDL runs high and who drink several cups a day have the most to gain from switching to a paper-filtered method.
Mechanism · The paper filter is the switch
Cafestol is among the most potent cholesterol-raising compounds known in the human diet: it interferes with the liver's regulation of cholesterol metabolism and raises in the blood. Kahweol is another diterpene in the same family. Both dissolve in coffee's oils.
A paper filter traps most of them, so:
Drip and pour-over (through paper): very little in the cup and almost no effect on LDLFrench press, Turkish coffee, and Scandinavian boiled coffee: no paper, so they go straight into the cupEspresso: somewhere in between; because a serving is small, total intake is usually modest It is one more example of how changing a food's form changes which molecules reach your body: the same beans, one sheet of paper, and a different bill at the liver.
Numbers · Unfiltered raises lipids; filtered barely does
Cai 2012 pooled 12 with 1,017 participants: drinking coffee (mostly unfiltered coffee containing cafestol) raised total cholesterol by about 8 mg/dL, by about 5 mg/dL, and by about 13 mg/dL on average, while filtered coffee produced essentially no such rise.
If your LDL runs high and you drink several cups a day, choose paper-filtered coffee (drip, pour-over, drip bags) and go easy on French press and boiled coffee. A cup or two from a French press makes no real difference for most people; what is worth changing is a lot of unfiltered coffee every day stacked on an LDL that is already high. For how to read your lipids overall, see the Dyslipidemia story.
Chapter 6
Why some can't sleep after one cup
Some people drink coffee before bed and sleep fine; for others one afternoon cup means a sleepless night. That is not imagination: the liver enzyme that breaks caffeine down runs at different speeds in different people.
Most caffeine is broken down by a liver enzyme called CYP1A2. Common variants in its gene roughly split people into fast and slow metabolizers; smoking speeds it up, while oral contraceptives and pregnancy slow it down. So "a half-life of about 4 to 5 hours" is only an average, and your own response is the better guide.
Mechanism · Over 95 percent goes through CYP1A2
Once caffeine enters the body, more than 95 percent of it is broken down by the liver enzyme CYP1A2 (IOM 2001). The gene that encodes this enzyme has common variants:
Fast metabolizers: high enzyme activity, caffeine is cleared quickly, the half-life is shorter, and a cup has little effectSlow metabolizers: low enzyme activity, caffeine lingers, and an afternoon cup can last into the night Outside factors turn the dial too: smoking markedly speeds up caffeine metabolism (it makes the body produce more CYP1A2), while oral contraceptives and pregnancy slow it down. Clearance slows markedly in pregnancy, which is one reason caffeine limits are stricter then. Slow metabolizers who are also very sensitive may get palpitations, shakiness, anxiety, and trouble falling asleep even from daytime coffee. That is not fussiness; it is pharmacology.
In practice · No genotype needed; watch your own response
There is no need to get genotyped; paying attention to your own body is enough. If one afternoon cup disturbs your sleep that night or makes you jittery, you are probably a slower metabolizer or more sensitive, so keep coffee to the morning and cut the total. If coffee sits well with you, there is nothing here to worry about.
"Moderate coffee is fine, maybe even slightly good for you" is an association seen in population averages; for you as an individual, it always gives way to how your own body responds.
Chapter 7
Drinking it, amounts, and cautions
For most healthy adults, the European Food Safety Authority (EFSA) considers up to 400 mg of caffeine a day free of safety concerns, and in large observational studies three or four cups a day also falls in the neutral-to-favorable range of associations. That is a ceiling, not a target, and non-drinkers have no reason to start.
The most overlooked part is timing: stop caffeine at least 6 hours before bed, and earlier still if you are a slow metabolizer or have insomnia. If your lipids run high, prefer paper-filtered brewing; do not turn your cup into a liquid dessert with syrup and cream; and do not drink it scalding. In pregnancy the ceiling is lower: 200 mg a day.
In practice · Paper, sugar, temperature, and the cutoff
If you are watching your lipids, choose paper-filtered coffee (drip, pour-over) and go easy on French press and boiled coffee. Do not turn coffee into a liquid dessert with lots of syrup and cream: then the risk comes from the added sugar, not the coffee, and the problem with sweetened drinks is covered in Sugary Drinks. Do not drink it scalding: IARC notes that very hot drinks above 65°C are linked to esophageal cancer risk, so let it cool a little.
Caffeine's half-life is about 4 to 5 hours, and longer in slow metabolizers. Caffeine from the afternoon may still be sitting on adenosine receptors when you want to sleep, so you fall asleep later and your deep sleep gets lighter. Count the cutoff back from bedtime: in a small trial of 12 people, Drake 2013 gave 400 mg of caffeine 0, 3, or 6 hours before bed, and all three timings disrupted sleep; the authors therefore advise avoiding substantial caffeine for at least 6 hours before bedtime. Slow metabolizers and people with insomnia should stop even earlier. On sleep, see the Melatonin story and All-Nighter.
Safety · In pregnancy, 200 mg is the total
Pregnancy and breastfeeding: caffeine is cleared more slowly in pregnancy. The American College of Obstetricians and Gynecologists (ACOG) advises keeping caffeine under 200 mg a day in pregnancy (about 1 to 2 cups of coffee), and EFSA's figure for pregnancy is also 200 mg a day. That total counts coffee, tea, chocolate, and energy drinks together.
Anxiety, panic, palpitations, or a known heart-rhythm problem: caffeine can make symptoms worse, so cut back or avoid it and follow your doctor's advice. Insomnia or poor sleep: stick firmly to the bedtime cutoff, and switch to decaf if needed (it keeps the chlorogenic acids, and in Ding 2014 the diabetes association held for decaf too). Some medications (certain antibiotics, ephedrine-type drugs, lithium, and others) may interact with caffeine; if you are unsure while taking medication, ask a doctor or pharmacist. Children and teenagers do not need coffee, and it more easily causes them sleep and mood problems.
This page provides general information only and does not replace a doctor's judgment about your medications and health.
References · 11
U.S. Department of Agriculture, Agricultural Research Service. (2019). FoodData Central — Coffee, brewed, and caffeine component data. Brewed coffee ~95-200 mg caffeine per ~240 mL cup (roughly double brewed tea). fdc.nal.usda.gov
Poole, R., Kennedy, O. J., Roderick, P., Fallowfield, J. A., Hayes, P. C., & Parkes, J. (2017). Coffee consumption and health: umbrella review of meta-analyses of multiple health outcomes. BMJ, 359, j5024. ~3-4 cups/day vs none associated with lowest all-cause mortality (RR ~0.83) and lower cardiovascular mortality; benefit non-linear; observational. 10.1136/bmj.j5024
Institute of Medicine (US) Committee on Military Nutrition Research. (2001). Pharmacology of caffeine. In Caffeine for the Sustainment of Mental Task Performance. National Academies Press. Caffeine elimination half-life 4-5 h in healthy adults; >95% metabolized by hepatic CYP1A2; adenosine-receptor antagonism is the principal CNS mechanism. www.ncbi.nlm.nih.gov/books/NBK223808
Jacobson, K. A., et al. (2022). Adenosine receptor antagonists: from caffeine to selective non-xanthines. British Journal of Pharmacology, 179(15), 3496-3511. Caffeine acts as a pan-antagonist at human adenosine receptors, blocking A2A/A1-mediated sleepiness signaling. 10.1111/bph.15103
Ding, M., Bhupathiraju, S. N., Chen, M., van Dam, R. M., & Hu, F. B. (2014). Caffeinated and decaffeinated coffee consumption and risk of type 2 diabetes: a systematic review and a dose-response meta-analysis. Diabetes Care, 37(2), 569-586. 28 cohorts, >1.1 million participants; inverse dose-response (RR ~0.91 per cup/day); holds for decaffeinated coffee. 10.2337/dc13-1203
Loomis, D., Guyton, K. Z., Grosse, Y., et al. / IARC Monographs Vol. 116. (2016). Carcinogenicity of drinking coffee, mate, and very hot beverages. The Lancet Oncology, 17(7), 877-878. Coffee re-classified to Group 3 (not classifiable as to carcinogenicity); very hot beverages >65°C classified Group 2A (probably carcinogenic to the esophagus). 10.1016/S1470-2045(16)30239-X
Killer, S. C., Blannin, A. K., & Jeukendrup, A. E. (2014). No evidence of dehydration with moderate daily coffee intake: a counterbalanced cross-over study in a free-living population. PLoS ONE, 9(1), e84154. ~4 mg/kg/day caffeine via coffee vs water showed no difference in 24-h urine output or hydration markers in habitual drinkers. 10.1371/journal.pone.0084154
Cai, L., Ma, D., Zhang, Y., Liu, Z., & Wang, P. (2012). The effect of coffee consumption on serum lipids: a meta-analysis of randomized controlled trials. European Journal of Clinical Nutrition, 66(8), 872-877. 12 RCTs, 1,017 subjects; coffee (largely unfiltered, cafestol-containing) raised total cholesterol ~8 mg/dL, LDL ~5 mg/dL, triglycerides ~13 mg/dL; filtered coffee produced essentially no rise. 10.1038/ejcn.2012.68
American College of Obstetricians and Gynecologists Committee on Obstetric Practice. (2010, reaffirmed). Committee Opinion No. 462: Moderate caffeine consumption during pregnancy. Obstetrics & Gynecology, 116(2 Pt 1), 467-468. Recommends limiting caffeine to <200 mg/day in pregnancy. 10.1097/AOG.0b013e3181eeb2a1
EFSA Panel on Dietetic Products, Nutrition and Allergies. (2015). Scientific opinion on the safety of caffeine. EFSA Journal, 13(5), 4102. Abstract: single doses up to 200 mg (about 3 mg/kg bw for a 70-kg adult) and habitual intake up to 400 mg per day raise no safety concerns for non-pregnant adults; up to 200 mg per day in pregnancy raises no concern for the fetus; single doses and habitual intake up to 200 mg by lactating women raise no concern for breastfed infants; for children and adolescents the data were insufficient to derive a safe intake, and the Panel suggests the adult 3 mg/kg bw per day as a basis. Typical energy-drink constituents, and alcohol up to about 0.65 g/kg bw, would not affect the safety of single caffeine doses up to 200 mg (abstract, Wiley page via Wayback snapshot 8 January 2025). 10.2903/j.efsa.2015.4102
Drake, C., Roehrs, T., Shambroom, J., & Roth, T. (2013). Caffeine effects on sleep taken 0, 3 or 6 hours before going to bed. Journal of Clinical Sleep Medicine, 9(11), 1195-1200. Randomized, double-blind, placebo-controlled Latin-square crossover in 12 healthy normal sleepers: one fixed 400 mg dose of caffeine (the authors call it moderate) at 0, 3 or 6 h before habitual bedtime vs placebo, sleep measured at home by diary and a portable monitor. All three timings disrupted sleep vs placebo (P < 0.05). By the objective monitor, even at 6 h caffeine reduced sleep by more than 1 hour; by diary, the 6-h reduction in total sleep time (41 min) only approached significance (p = 0.08). The authors conclude that substantial caffeine should be avoided for a minimum of 6 hours before bedtime, and that afternoon caffeine should at least be restricted to before 17:00 (abstract, PMID 24235903; full text, PMC3805807). 10.5664/jcsm.3170
