Caffeine and sleep — half-life and timing

How caffeine works

Adenosine is a neurotransmitter that accumulates throughout the day. As it builds up, it binds to adenosine receptors in the brain — particularly in the basal forebrain — and reduces neural activity in arousal-promoting regions. This creates the subjective feeling of sleepiness.

Caffeine is structurally similar to adenosine and binds to the same receptors as an antagonist: it occupies the site without activating it, blocking adenosine from doing so. The adenosine is still there; you just stop feeling it.

This is a critical distinction. Caffeine doesn't add energy. It masks the depletion signal. When the caffeine clears, the accumulated adenosine (still waiting) hits you all at once — the “caffeine crash.”

The half-life calculation

Caffeine follows first-order elimination kinetics: a constant fraction is cleared per unit time. The math is straightforward: C(t) = C₀ × 0.5^(t / half-life).

For an average 5-hour half-life, a 200mg dose decays roughly:

• 0 hours: 200 mg
• 5 hours: 100 mg (still strong)
• 10 hours: 50 mg (still meaningful)
• 15 hours: 25 mg (background level)

Use the half-life calculator for your specific dose and metabolism.

The CYP1A2 lottery

Caffeine clearance is dominated by the liver enzyme CYP1A2, encoded by the gene of the same name. A common polymorphism produces two functional variants:

• Fast metabolisers (~50% of population) — clear caffeine in ~3 hours.
• Slow metabolisers — half-life closer to 7–8 hours.

This is why two people drinking the same coffee feel wildly different effects. It's also why blanket recommendations like “no coffee after 2pm” work for some people and are uselessly conservative or catastrophically generous for others.

The sleep architecture cost — even when you fall asleep

Drake et al (2013) demonstrated the most surprising finding in recent caffeine research: caffeine measurably degrades sleep even when you fall asleep on time. The study gave participants 400mg of caffeine at 6 hours, 3 hours, or 0 hours before bedtime. All three groups showed reduced slow-wave sleep, and the 6-hour group showed reductions roughly equivalent to the 3-hour group.

Implication: subjective sleep onset isn't a reliable indicator of whether caffeine is affecting sleep. If you have afternoon coffee and sleep “fine,” your deep sleep is probably still being suppressed.

Tolerance — partial and unreliable

Regular caffeine users develop tolerance to the alertness effect within a few days. This is well-documented in chronic-coffee-drinker studies — the “wake-up” effect on a morning cup is largely just relief from overnight withdrawal, not stimulation.

Tolerance to the sleep-disruption effect is less complete. Habitual users still show reduced slow-wave sleep from late caffeine, though the effect may be modestly attenuated.

Practical takeaways

• Most people benefit from a caffeine cut-off 8–10 hours before bed — earlier than the standard “2pm” rule suggests.
• If you suspect you're a slow metaboliser (you feel coffee's effects for hours), cut earlier.
• You can't feel whether late caffeine is hurting your sleep architecture — only changes in next-day function reveal it.
• A 1–2 week wash-out is a useful experiment to see your baseline alertness without caffeine.