Why caffeine stops working

Learn · Energy systems · Explainer, Typed Supps

Coffee blocks your brain’s tiredness signal. Keep blocking it daily, and the brain builds more receivers.

Stage by stage

The block (Adenosine antagonism): Caffeine parks in the receptors that read your brain’s sleep-pressure signal. The pressure keeps building behind it.
The long tail (~5 hour half-life): Half a dose is still circulating five hours later. Even six hours before bed, 400 mg cut about an hour of measured sleep.
The adjustment (More receptors): Daily use grows extra adenosine receptors, so the same dose blocks a smaller share. Much of the morning lift becomes withdrawal reversal.
The reset (1–2 weeks): Withdrawal peaks at a day or two and fades within about nine. A fortnight off — or a slow taper — restores the original effect.

What the evidence supports

Caffeine improves alertness and exercise performance in non-tolerant users — strong evidence. One of the best-replicated effects in sports nutrition, typically at 3–6 mg per kg.
Caffeine taken up to six hours before bed disrupts objectively measured sleep — moderate evidence. Shown in controlled dose-timing work, with subjective reports consistently underestimating the loss.
Daily use produces tolerance via adenosine-receptor upregulation — moderate evidence. Clear mechanism and animal data, with habitual users consistently showing blunted responses.
Habitual users get a net lift above their true, uncaffeinated baseline — limited evidence. Much of the felt benefit in daily drinkers is withdrawal reversal. Stated deliberately — it is the industry’s least favourite finding.

Timing

The lever that matters is when, not whether. With a roughly five-hour half-life, a cutoff 8–10 hours before bed protects sleep for most people; slow metabolisers need longer.

Time to effect

45 min: Roughly when a dose peaks in blood.
1–2 wks: Daily use to tolerance; about the same off it to reset.

Common myth

“Coffee doesn’t affect me — I can fall asleep right after an espresso.” In reality: Falling asleep isn’t the test. Caffeine six hours before bed cut objectively measured sleep by about an hour in controlled testing, while people’s own estimates barely noticed. The cost lands the next day, and usually gets billed to “needing more coffee”.

In the glossary

Half-life