Sleep & recovery

Coffee: what is in the cup, caffeine amounts, filtering, sleep, and the health evidence

9 sources · Latest source check 2026-10-20 UTC

On this page

Coffee is a drink made by brewing the dried, roasted, ground seeds of coffee cherries. Its familiar name covers many different drinks: a small espresso, a large brewed coffee, and a sweetened latte can differ in preparation, size, and ingredients. Looking at the actual cup is more useful than treating every coffee as the same exposure.[1]

Coffee is more than caffeine

The beans come from a plant, then undergo harvesting, processing, drying, roasting, and grinding before brewing. Arabica and robusta are different coffee varieties. Roasting is processing, and brewing extracts material into water; the mere fact that coffee is processed does not determine its health effects.[1]

Coffee contains caffeine, chlorogenic acids, and the oily compounds cafestol and kahweol, among many other constituents. Research on the whole beverage therefore cannot automatically be applied to caffeine tablets, energy drinks, or an isolated coffee extract.[2]

A cup is not a reliable caffeine unit

A 2024 laboratory study tested 299 coffee samples and found substantial differences among servings and preparation methods. In that sample, 4% of servings contained more than 200 mg caffeine. Espresso was concentrated, but its smaller serving had less caffeine on average than some larger drinks. Bean choice, brewing, serving volume, and extra shots complicate comparisons. The study measured products, rather than what consumers actually drank each day.[3]

Questions that make coffee comparisons clearer

Questions that make coffee comparisons clearer
Drink	What to check
Espresso	Concentration per milliliter and caffeine in the entire shot are different measurements.[3]
Americano, cappuccino, or latte	The drink name does not specify one universal serving size or caffeine amount.[3]
Sweetened coffee	Syrups, sugar, and flavored creamers change the nutritional comparison.[8]

Why coffee looks beneficial in some studies

The 2017 umbrella review combined existing reviews across many outcomes, including mortality and cardiovascular disease. Much of that evidence compared people who chose different amounts of coffee. Smoking, other habits, illness-related changes in drinking, and imperfect intake measurement can influence such comparisons. An association is useful evidence, but it is not a prescription to drink a particular number of cups.[2]

Filtering can matter for cholesterol

Paper filters retain much of coffee’s oily fraction. In a randomized study of 64 healthy adults, unfiltered boiled coffee raised total and LDL cholesterol compared with the same coffee passed through paper filters. This was a relatively high-intake experiment, and it measured blood lipids rather than heart attacks. Paper filtration and simply separating visible grounds are not interchangeable descriptions of a brew.[4]

A drink can fit a diet and still disturb sleep

A meta-analysis searched through March 2025 and included 22 controlled crossover trials in healthy adults. Caffeine reduced sleep duration and efficiency on average and delayed falling asleep. The pooled average was about 35 minutes less sleep, rather than a prediction for every drinker. These studies concern acute exposure; they do not produce one timing rule that guarantees good sleep. Feeling less sleepy after coffee differs from the quality of the next night’s sleep.[5]

Coffee does not automatically cancel its own water

In a short crossover study, 50 men who regularly drank coffee received coffee or an equivalent volume of water under controlled conditions. Several hydration measures did not differ meaningfully. That supports including ordinary coffee in fluid intake for similar habitual drinkers. It does not settle large caffeine doses, unfamiliar users, illness, strenuous exercise, or extreme heat.[6]

Very hot is a separate issue

IARC’s 2016 assessment did not classify coffee drinking as carcinogenic on the available evidence. It classified drinking very hot beverages, above 65°C, as a probable carcinogenic hazard to the oesophagus. The concern applies to drinking temperature across beverage types. A hazard classification describes evidence for a possible cause; it does not state the chance that one person’s cup will cause cancer.[7]

Consider the rest of the drink

CDC includes flavored coffees and coffee with added sugar or flavored creamer among sugary drinks. A coffee-based dessert drink should not inherit the health interpretation of plain coffee. Comparing the label or menu information for added sugars, calories, and portion size answers a different question from comparing caffeine.[8]

Cold brew describes extraction, not merely serving temperature

Cold brew extracts coffee at a lower temperature over longer contact; hot-brewed coffee can also be chilled and served over ice. Ask how it was prepared rather than inferring the method from a cold cup.[9]

Evidence & uncertainty

The evidence is reassuring for many ordinary adult coffee drinkers, but disease-prevention headlines mostly come from observational research. A major umbrella review found many favorable associations alongside context-specific harms. These findings do not establish that starting coffee will prevent disease, or that more is always better.[2]

Selected-source overview · Independent review pending

Published questions

Is coffee good or bad for you?

Selected-source overview · Independent review pending. Read the evidence context.

For many adults, ordinary coffee consumption is compatible with a healthy diet. Favorable long-term associations are mostly observational, and individual symptoms and preparation matter. The evidence does not justify presenting coffee as a disease-prevention treatment.[2]

Evidence behind this answer

Coffee consumption and health: umbrella review of meta-analyses of multiple health outcomes
systematic review · 2017
Original indexed review summary and selected methods/results inspected; no independent full-text appraisal · Original indexed abstract and selected BMJ introduction/results; 2017 evidence synthesis