How much caffeine is too much for sleep?

Ultrahuman Ring members' sleep window was shortest after days of 400 mg or more of caffeine. Even with an early last drink, it was 16 minutes shorter (likely between 4 and 27 minutes) than on their own lowest-caffeine days. The three middle bands came with sleep 3.2, 6 and 3 minutes shorter, in rising order of caffeine, and only the last one's range includes no change. When each night was compared with the night before instead, the pattern reversed, which fits short nights leading to heavier caffeine days.

How much shorter was sleep after heavier caffeine days?

Sleep was a few minutes shorter after most heavier caffeine days, and shortest after days of 400 mg or more, as the chart below shows.

All four bars in the chart sit below zero, and the heaviest days sit furthest down. Light and moderate days showed a clear change, while 300 to 400 mg days did not. That result overlaps the 200 to 300 mg result and had fewer members, so it does not mean that amount is safer. In a broader version that counted every night, late drinks included, each change was larger. Among members who usually log 150 mg or more a day, sleep still changed most after the heaviest days. The bands were set in advance, so the analysis cannot pinpoint where the change starts.

Ultrahuman compared heavier and lighter caffeine days within each member, using the rules below for which drinks, days and nights counted.

A drink counted when the member logged it in the Caffeine Window with at least 20 mg of caffeine, so decaf did not count. Every caffeinated drink counted, not only coffee: tea, cola, energy drinks and matcha too. A day's caffeine was the total of the drinks timed between waking from the night before and falling asleep, so a drink just after midnight still belonged to that evening. A member's usual bedtime was the middle of the times they fell asleep over the 14 nights before, from at least 10 valid nights on the same ring and firmware version. It never used the night being measured.

Only nights whose last counted drink came at least 8 hours before that usual bedtime were compared, which kept drinks close to bedtime out of every group. Each of those nights was grouped by the day's total logged caffeine: under 100 mg, 100 to 200, 200 to 300, 300 to 400, or 400 mg or more, each group running up to the start of the next.

The night's sleep was timed from the start to the end of the main sleep as the ring detected them, so brief awakenings in between count toward it. Each member's nights in a higher group were compared with the same member's nights after under 100 mg, so only members with both kinds of night counted in a comparison. Nights with no logged caffeine were left out, because a member may simply not have logged.

Each member counted equally, however many nights they logged; the uncertainty ranges resampled whole members.

Is sleeping heart rate different after heavier caffeine days?

Sleeping heart rate, in beats per minute (bpm), ran slightly lower after heavier caffeine days, against the same members' nights after days under 100 mg. The largest change, 0.6 bpm lower, followed days of 400 mg or more. It was likely between 0.1 and 1.2 bpm lower, and every band's range stayed below zero.

After moderate days, sleeping heart rate was 0.5 bpm lower (likely between 0.3 and 0.7 bpm). After heavy days it was 0.4 bpm lower (likely between 0.1 and 0.8 bpm).

What does other caffeine research find about dose and sleep?

Trials and reviews also link caffeine with shorter sleep, which they measured as total sleep time. Clark and Landolt's systematic review found that caffeine typically reduced total sleep time and sleep efficiency, with dose and timing relationships established [1]. Chang and colleagues' meta-analysis of crossover trials found that high and low doses both reduced total sleep time, with no clear linear link to dose [2]. In Robillard and colleagues' crossover trial, the higher caffeine dose shortened sleep more in middle-aged than young adults [3].

For safety more broadly, one review sets a reference point at the top band studied here. A systematic review of caffeine safety concluded that up to 400 mg a day in healthy adults is not associated with overt adverse cardiovascular, behavioral or acute effects [4]. That review did not set a limit for sleep.

What this analysis can't tell you

Within weeknights alone and weekend nights alone, sleep still changed most after the heaviest days. The analysis had no record of which days were workdays. Measured against the night before instead, the changes reversed direction, so this analysis cannot tell which came first. Only logged caffeine counted, so an unlogged drink would put a day in too low a band. Heavier days also ended their drinks later: 13.1 hours before usual bedtime on the lowest-caffeine days, against 10.5 hours on the heaviest days. Gardiner and colleagues' meta-analysis estimated that a larger caffeine serving needs a longer gap before bedtime to avoid losing total sleep [5]. This analysis therefore cannot fully separate a day's caffeine total from when it ended. Ultrahuman's article on caffeine timing looked at how late the last drink came, so its results and these overlap. Members logged caffeine on at least 20 days, so they likely watch caffeine more closely than people who do not log it.

Frequently asked questions

Does it matter whether the caffeine comes from coffee or tea?

This analysis cannot say whether coffee and tea differ, because it grouped days by total caffeine from any logged drink. On average, coffee was the last logged drink on 74.5% of a member's counted nights (likely between 73.0% and 75.8%). In a stricter version that kept only nights ending in coffee, the change after 400 mg or more was larger, and the changes after lighter days were smaller.

Why was sleeping heart rate lower after heavier caffeine days?

This analysis cannot say why sleeping heart rate ran slightly lower after heavier caffeine days. The changes were small, at most 0.6 bpm on average.

Who does this finding not apply to?

This finding does not describe people who drink caffeine late in the day, because every counted night had its last drink at least 8 hours before usual bedtime. It describes adults who wear the Ultrahuman Ring and log caffeine on many days, each compared with their own lighter days. It says nothing about people a clinician has told to limit caffeine.

References