Rasayan Ayurveda

21 August 20265 min read

Caffeine, Sleep, and Why the Afternoon Cup Costs More Than You Think

The 4pm coffee is still working at midnight, whether or not you feel it.

Most people know caffeine late at night is a bad idea. Far fewer realise how early 'late' begins, and the evidence on this is unusually clean.

The six-hour finding

In a randomised, double-blind, placebo-controlled study, Drake and colleagues gave participants 400 mg of caffeine — roughly two strong coffees — at three different times: immediately before bed, three hours before, and six hours before. Sleep was measured objectively at home.

Caffeine taken six hours before bedtime reduced total sleep time by more than an hour. The researchers concluded it has important disruptive effects on sleep at that distance from bedtime, supporting the standard advice to avoid substantial caffeine for at least six hours before sleeping.

The detail that makes this worth knowing: the disruption showed up in the objective measurements even where participants' own sense of their sleep did not fully register it. People are poor judges of how badly they slept.

Why it lingers

Caffeine's half-life in a healthy adult is roughly five to six hours — the time to clear half of it. Half. After a 4pm coffee, a quarter of the dose is still circulating around 4am.

That range varies considerably between people. Liver enzyme genetics, whether you smoke, some medications and pregnancy all shift it, sometimes by hours. This is why one person sleeps fine after an evening espresso and another does not, and why 'it doesn't affect me' is unreliable — it may simply mean you have never slept without it in your system.

A practical cut-off

Take your usual bedtime and count back six hours. Anything caffeinated after that is borrowing from tonight's sleep.

Bed at 10pm
Last caffeine by 4pm. That includes green tea, black tea, cola and dark chocolate — not only coffee.
Bed at midnight
Last caffeine by 6pm. More generous than most people assume, but it still rules out a coffee with dinner.
If you sleep badly already
Push it earlier. A trial of caffeine abstinence in habitual users with poor sleep is a reasonable thing to try for two weeks before concluding caffeine is not your problem.

What decaf and herbal actually mean

Decaffeinated coffee and tea have had most of their caffeine removed, not all of it. A cup typically retains a small amount — usually trivial, but not zero, and worth knowing if you are unusually sensitive.

Herbal infusions are a different situation entirely. Chamomile, hibiscus, butterfly pea, lavender and peppermint are not the tea plant and contain no caffeine at all — not a reduced amount, none. If you want something warm at 9pm with no pharmacological cost, that is the category that delivers it.

That is a factual statement about the plants, not a health claim, and it is the honest reason to reach for an infusion in the evening rather than a cup of tea.

What we read

Every study below is linked so you can check it rather than take our word for it — including the ones whose results were negative.

  1. 1
    Drake C et al., 2013 — J Clin Sleep Med

    Randomised, double-blind, placebo-controlled study in the Journal of Clinical Sleep Medicine. 400 mg of caffeine taken 0, 3 or 6 hours before bed; the six-hour dose still measurably reduced total sleep time.

  2. 2
    Doty TJ et al., 2020 — J Clin Sleep Med

    'Buzzed before bedtime: hidden harms of late day caffeine consumption' — a review of how the timing of caffeine intake affects sleep quality.

  3. 3
    Irish LA et al., 2021 — Journal of Sleep Research

    The effect of caffeine abstinence on sleep among habitual caffeine users with poor sleep — the basis for suggesting a trial period rather than assuming caffeine is not the cause.

This journal is written for general interest. It is not medical advice, and nothing we sell is intended to diagnose, treat or prevent any condition. If you are pregnant, taking medication, or managing a health condition, speak to a qualified doctor before changing what you drink.

More from the Journal