Coffee timing: morning, afternoon, and the adenosine question

Caffeine is one of the most-studied substances in pharmacology. Its effects on alertness, attention, and cognition are well-established. Timing (when in the day you have it) has spawned a small industry of opinion online, much of it more confident than the evidence supports.

The pharmacokinetics matter most#

Caffeine has a mean half-life in healthy adults of around 5 hours, but the range is wide, roughly 2 to 10 hours depending on genetics (CYP1A2 polymorphism), pregnancy, oral contraceptives, smoking, and liver function.

This is the most important fact about coffee timing. A 200 mg dose at noon:

• In a fast metaboliser (half-life 2 hours): roughly 12 mg left by 10 pm
• In a typical metaboliser (half-life 5 hours): roughly 50 mg left by 10 pm
• In a slow metaboliser (half-life 10 hours): roughly 100 mg left by 10 pm — clinically significant for sleep

The "stop caffeine 8 hours before bed" guidance is the average-population rule. For slow metabolisers it's more like 10–12 hours; for fast metabolisers it's overkill.

The morning question#

A popular online claim suggests delaying coffee for 90 minutes after waking, to let cortisol peak first and avoid blunting the natural awakening response.

What the evidence actually shows:

• Cortisol does peak shortly after waking (cortisol awakening response, CAR).
• Caffeine can modestly affect cortisol secretion, but in habitual coffee drinkers this effect is small (tolerance).
• There is no controlled trial demonstrating that "delaying coffee 90 minutes" produces better alertness across the day in habitual drinkers.
• The "you'll get an afternoon crash if you don't delay" claim is not supported by trial data; afternoon dips are largely circadian, not caffeine-related.

For people who genuinely enjoy their coffee on waking, delaying it 90 minutes is unlikely to add measurable benefit. For people who get jittery or anxious from caffeine first thing, eating first or delaying is reasonable on common-sense grounds, not because of a cortisol theory.

The afternoon question#

This is where the evidence is more consistent. A 2013 Drake et al. trial (Journal of Clinical Sleep Medicine) gave 400 mg caffeine at three time points (0, 3, and 6 hours before bed) and measured sleep:

• All three doses significantly disrupted sleep compared to placebo
• The 6-hours-before-bed dose still measurably reduced sleep efficiency

The practical takeaway: in average-metabolism adults, caffeine taken less than 6 hours before intended sleep onset measurably impairs sleep. The "evening espresso is fine because I don't feel it" position is unsupported.

Many habitual drinkers can fall asleep with elevated caffeine on board. That's not the same as sleeping well. Polysomnography and HRV tracking consistently show degraded sleep quality in adults with late caffeine intake.

Tolerance#

Daily caffeine produces tolerance to many of its acute effects within days to weeks. Tolerance is partial and reverses with abstinence over 1–2 weeks.

This matters because:

• The 200 mg dose that produced a measurable cognitive boost in a non-tolerant participant produces a smaller effect in a habitual drinker.
• Caffeine in habitual drinkers may function mostly as withdrawal prevention; the "I haven't had coffee yet" headache is a withdrawal symptom.
• Strategic dosing (using caffeine on occasions when you specifically want the alerting effect) produces larger acute boosts than habitual high-dose use.

A practical pattern#

For most adults who care about both cognition and sleep:

• A first cup any time in the morning is fine. Delay it if it makes you feel better; don't delay it on a cortisol theory.
• Stop caffeine 8+ hours before intended sleep. For slow metabolisers, longer.
• Total daily dose under 400 mg (EU EFSA suggested limit) for non-pregnant adults. Pregnancy guidance is 200 mg/day.
• Caffeine in tea, dark chocolate, and pre-workout supplements counts: read labels.

Decaf, half-caf, and the placebo#

Decaf has trace caffeine (typically 2–7 mg per cup); for most adults this is negligible. Mixing half-caf is a reasonable way to taper or maintain ritual without the dose. The behavioural ritual of coffee (warm drink, break, social context) produces its own subjective alerting effect that some studies have separated from the pharmacology.

What this isn't#

Not dosing advice. Individual responses vary considerably. People with anxiety, arrhythmias, panic disorder, or those on stimulants should be cautious; caffeine interactions with many medications are real.

For most healthy adults, the cleanest summary is: morning coffee is fine when you want it; afternoon coffee is less harmless than it feels; and slow metabolisers (perhaps 15–20% of adults) need a wider buffer than average advice assumes.