MySleepMate / Field Notes

Field note

Loneliness does not shorten your sleep. It breaks it up

The short answer

Seeing people is on my list of things that improved my sleep, and when I first wrote that list I said the evidence was weak. I was wrong, and I had not looked hard enough. Cacioppo and colleagues at the University of Chicago found in 2002 that lonely people had poorer sleep efficiency and more time awake after falling asleep, while their total sleep duration was unchanged. That is a striking shape: loneliness does not take hours off your night, it perforates them. And there is a mechanism connecting social life to sleep that predates it by fourteen years, called social zeitgeber theory, which says your social routine is one of the clocks your body sets itself by.

The finding, and why its shape matters

The paper is Do Lonely Days Invade the Nights? Potential Social Modulation of Sleep Efficiency, by John Cacioppo, Louise Hawkley, Gary Berntson, John Ernst, Amber Gibbs, Robert Stickgold and Allan Hobson, published in Psychological Science in 2002. Cacioppo was at the University of Chicago and spent his career on the physiology of loneliness; Stickgold and Hobson are Harvard sleep researchers.

What they found:

MeasureLonely participants
Sleep efficiencyPoorer
Time awake after falling asleepGreater
Sleep fragmentationSignificantly higher with higher loneliness scores
Total sleep durationUnchanged
Daytime sleepinessUnchanged

Sit with that last pair. If you had asked these people how long they slept, or asked them whether they felt sleepy, you would have found nothing. The effect was invisible to the two questions everyone asks about sleep.

It is the same lesson as the regularity finding arriving from a completely different direction. Duration is the measure we have all agreed to care about, and it keeps turning out to be the one that misses things. A later paper by Hawkley, Preacher and Cacioppo made the point in its title: loneliness impairs daytime functioning but not sleep duration.

Why social life would affect sleep at all

The mechanism is older than the finding, and it is the piece I was missing when I wrote that the evidence was thin.

Social zeitgeber theory was set out by Ehlers, Frank and Kupfer in Archives of General Psychiatry in 1988. A zeitgeber is a time-giver, an external cue that sets your body clock. Light is the dominant one. The argument here is that social life is another: meals with other people, work, appointments, conversations, obligations. Things that happen at times.

The claim that follows is specific. Regularity of social routine produces stable biological rhythms, and irregularity of social routine produces circadian instability. Disruption to the social schedule propagates into the sleep-wake cycle rather than staying at the level of mood.

Which reframes what "socialising helps me sleep" might actually mean. Not that company is soothing, though it may be. That a life with people in it has a shape, and the shape is what your circadian system reads.

The connection to the strongest thing on this site

Put the two findings on this page next to the one with the most weight behind it and something clicks.

The chain is not proven, and I am not going to pretend three separate studies constitute a demonstrated pathway. But it is coherent, and it explains something my own diary hinted at without being able to say why. Stress was my single largest factor, at a median of 24 minutes to fall asleep on high-stress nights against 9 on normal ones, a bigger gap than caffeine, alcohol or screens. Isolation and rumination are stress with a schedule attached.

Which is why this ended up on my list at all. Not because I read a study, but because a week with people in it is a week with a shape, and I sleep better in a week with a shape.

What this does not say

So this belongs in the middle of the evidence table on my list, not at the bottom where I first put it and not at the top.

What I take from it

A note on getting this wrong

When I first published the list of five changes, I put seeing people last and wrote that I could not find research I would put on the same table as 88,975 accelerometers. I invited readers to send me any.

Then I looked properly, and it was there: a 2002 paper in a major journal, with Stickgold and Hobson on it, and a 1988 theory in Archives of General Psychiatry giving the mechanism. I had searched for "socialising and sleep" and concluded the field was thin. The field files it under loneliness.

That is worth recording, because it is a different failure from the ones this site usually documents. Those are about claims being weaker than advertised. This one was me declaring an absence of evidence after a shallow search, which is the same mistake I made about whether the Brooks and Lack nap study had been replicated. Saying "there is no research on this" is a claim, and it needs the same standard of checking as any other. The list has been updated and the change is in the corrections log.

Questions people actually ask

Does loneliness affect sleep?

Yes, but not in the way most people assume. Cacioppo and colleagues found in 2002 that lonely people had poorer sleep efficiency and more time awake after falling asleep, while total sleep duration and daytime sleepiness were unchanged. The effect is on fragmentation and quality rather than quantity.

Does socialising improve sleep?

There is a plausible mechanism. Social zeitgeber theory, set out in 1988, holds that social routines such as meals, work and appointments act as time cues that entrain circadian rhythms, and that regular social rhythms produce stable biological rhythms. That connects social life to sleep through schedule rather than through mood.

What is a social zeitgeber?

A zeitgeber is an external cue that sets your body clock, with light being the strongest. A social zeitgeber is a recurring social event that does the same job: shared meals, work hours, appointments, regular contact with other people. Regularity in those produces circadian stability; irregularity produces instability.

Does poor sleep make you lonely, or does loneliness cause poor sleep?

Both, on the available evidence. Research on older Americans found reciprocal effects between loneliness and sleep disturbance, meaning each appears to worsen the other rather than the relationship running in one direction.

Is being alone bad for your sleep?

The research measures loneliness, meaning perceived social isolation, rather than time spent alone. People can spend a great deal of time alone without being lonely, and can be lonely in company, so time alone by itself is not what these findings are about.

References

The schedule is the thing

Whether the cue is light, company or an alarm, what your clock reads is regularity. The calculator counts backwards from a wake time you keep.

Open the sleep calculator