When you last had caffeine, and how you sleep
Evidence last assessed 2026-08-24.
In short
- What the research shows
- Across 24 controlled studies, caffeine shortened sleep by about 45 minutes.
- What it doesn't show
- Every study gave a set dose; none measured the time of your last caffeine.
- Sleep and energy
See studies and method
What the evidence says
This is the best-measured relationship on this list and the one that carries over to you least well: a meta-analysis of 24 controlled studies and two crossover experiments, not one of which measured what Neta actually records.
Pooled across 24 controlled studies, caffeine cut total sleep time by about 45 minutes and sleep efficiency by 7%, added about 9 minutes to falling asleep and about 12 minutes awake during the night. The same data yields a timing rule: a 107 mg coffee should finish at least 8.8 hours before bed. A crossover experiment found that 400 mg taken six hours before bed still disturbed sleep, on diaries and on instruments alike. A third found evening caffeine delayed the body clock's melatonin rhythm by about 40 minutes, so late caffeine does not only lengthen falling asleep, it moves the clock itself.
What this does not show
No study measured what Neta records. Every one gave a known dose at a set time before bed; what you log is the clock time of your last caffeine, with no amount attached. A nine o'clock sip and a nine o'clock triple shot look identical in your data. The amount matters, and the research's own arithmetic is dose-dependent. No study used a five-point morning rating; they measured sleep with instruments, and the step from those to how you would rate a night is plausible and unmeasured. No source studied women with this condition: the condition-specific caffeine research so far is in animals. And the doses were fixed and given to mostly well-rested volunteers, while your caffeine is habitual and self-chosen and regular use builds tolerance, so the direction is expected to carry over to you and the size is not.
What Neta records, in its own terms
- What Neta watches
- the time of your last caffeine
- What it is read against
- Sleep quality
Where the research and this product do not line up
- The studies measured a different thing from the one you log.
- The studies measured a different outcome from the one Neta reads.
- The studies were not done in women with this condition.
Sources, as Neta holds them
Gardiner C, Weakley J, Burke LM, Roach GD, Sargent C, Maniar N, Townshend A, Halson SL. The effect of caffeine on subsequent sleep: a systematic review and meta-analysis. Sleep Medicine Reviews 69 (2023) 101764, doi 10.1016/j.smrv.2023.101764. 24 controlled studies.
Systematic review
Drake C, Roehrs T, Shambroom J, Roth T. Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed. Journal of Clinical Sleep Medicine 9(11):1195–1200 (2013), doi 10.5664/jcsm.3170.
Crossover experiment, each person compared with herself
Burke TM, Markwald RR, McHill AW, Chinoy ED, Snider JA, Bessman SC, Jung CM, O'Neill JS, Wright KP Jr. Effects of caffeine on the human circadian clock in vivo and in vitro. Science Translational Medicine 7(305):305ra146 (2015), doi 10.1126/scitranslmed.aac5125.
Crossover experiment, each person compared with herself · 5 participants · 49 days observed
Rétey JV, Adam M, Khatami R, Luhmann UFO, Jung HH, Berger W, et al. A genetic variation in the adenosine A2A receptor gene (ADORA2A) contributes to individual sensitivity to caffeine effects on sleep. Clin Pharmacol Ther. 2007;81(5):692-8. doi:10.1038/sj.clpt.6100102
Crossover experiment, each person compared with herself