Does caffeine at bedtime reduce deep sleep?
Yes, but the effect in this study was confined to the early part of the night. Eight young men took 100 mg of caffeine at bedtime in a placebo-controlled sleep EEG study. Caffeine lengthened the time it took to fall asleep, lowered sleep efficiency, reduced stage 4 NREM sleep, and cut slow wave activity, an EEG measure of sleep depth.
The important qualifier is that the suppression of slow wave activity was limited to the first NREM episode, not spread across the whole night.
What the study found:
- Participants: 8 young men
- Dose and timing: 100 mg caffeine, taken at bedtime
- Sleep latency: prolonged versus placebo
- Sleep efficiency and stage 4 NREM sleep: both reduced
- Slow wave activity: decreased, mainly in the lowest delta band, and only in the first NREM episode
- Spindle frequency power: slightly increased
- The following night: stage 4 sleep returned to normal, but sleep latency was still elevated
What this study does not show
This paper is frequently cited for the claim that caffeine taken six hours before bed damages sleep. That claim comes from a different experiment. This study administered caffeine at bedtime only. It did not test a 3 hour or 6 hour interval, so it cannot say anything about afternoon coffee.
The six hour finding people are thinking of comes from separate work by Drake and colleagues, which did compare 0, 3, and 6 hour conditions. If that is the question you came here with, that is the study to look for. This one answers a narrower question: what does caffeine do if you take it right as you go to bed.
Dr. Kumar’s Take
The detail I find most interesting is the one that gets dropped in summaries: the slow wave suppression was limited to the first NREM episode.
That matters because the first NREM cycle carries the largest share of the night’s slow wave activity. Sleep pressure is highest at lights out and discharges steeply in the first couple of hours. So a hit concentrated there is not trivial, even though it is not spread across the night. But it is also not the same claim as “caffeine wrecks your deep sleep all night,” and the difference is worth preserving.
The carryover finding is the other one I would not skip. On the night after caffeine, stage 4 sleep had recovered but sleep latency was still elevated. The depth of sleep bounced back faster than the ability to fall asleep did.
I would hold all of this loosely. Eight young men is a very small sample and tells you nothing about older adults, women, or people who metabolize caffeine slowly. Caffeine clearance varies substantially between individuals, largely through CYP1A2, which is exactly the variable that would determine how long a given dose keeps affecting you. This study was not designed to capture that.
Study Limitations
Eight participants, all young men, so nothing here generalizes confidently to women, older adults, or habitual heavy consumers. A single 100 mg dose at a single time point means no dose response and no timing comparison. Individual differences in caffeine metabolism were not examined.
Related Studies and Research
- How Caffeine Works: Blocking Adenosine Receptors to Prevent Sleepiness
- Minimum Effective Caffeine Dose for Strength Training
- Coffee Hydration Study: Does Coffee Dehydrate You?
- Adenosine: The Sleep Chemical That Makes You Tired
- Sleep and Temperature Regulation
FAQs
How much caffeine did this study use?
100 mg, roughly the amount in a standard cup of coffee, taken at bedtime. It was a single dose, so the study says nothing about larger amounts or repeated intake.
Does this prove I should stop caffeine six hours before bed?
No. This study only tested caffeine taken at bedtime. The six hour comparison comes from separate research and should not be attributed to this paper.
Was deep sleep reduced all night?
No. Slow wave activity was suppressed in the first NREM episode only. Stage 4 NREM sleep across the night was reduced, but the slow wave effect was concentrated early.
Did the effects last into the next night?
Partly. Stage 4 sleep normalized by the following night, but sleep latency remained longer than placebo.
Bottom Line
A small EEG study in eight young men found that 100 mg of caffeine at bedtime delayed sleep onset, lowered sleep efficiency and stage 4 sleep, and suppressed slow wave activity during the first NREM episode. It does not establish anything about caffeine consumed hours earlier in the day.
Read the study
Landolt HP, Dijk DJ, Gaus SE, Borbély AA. Caffeine reduces low-frequency delta activity in the human sleep EEG. Neuropsychopharmacology. 1995;12(3):229-238.
