Most research on sleep after trauma asks how much people sleep and how badly. A 2026 study started from the observation that this question keeps returning weak answers — that mean circadian levels, as the authors put it, often show limited discriminability. So they asked a different one: not how disturbed the rhythm is on average, but how steady it is from one day to the next.

The Design

The participants were 211 survivors of the Supernova music festival mass-trauma event, and 113 matched comparison participants who had not been there.

Rather than bringing people into a laboratory, the researchers monitored them in real-world settings using wearable sensors that tracked heart rate, physical activity and sleep–wake cycles. Monitoring ran for one month, at two to six months after the event.

PTSD symptoms were assessed before and after that recording period, and then again at a follow-up eight to eleven months post-event.

That last detail is what makes the study prospective rather than a snapshot. The circadian measurements were taken before the outcome they were tested against, which is the arrangement that allows a finding to be described as predictive rather than merely associated.

Why Averages Were the Wrong Measurement

The premise of the study is a methodological point worth understanding, because it recurs throughout this field.

If you average a month of sleep timing, two people can produce nearly identical numbers by entirely different routes — one going to bed at much the same hour every night, the other oscillating wildly and averaging out to the same middle.

An average is a summary that discards exactly the information that distinguishes those two people. Measuring instability — how much each day differs from the last — recovers it.

What They Found

Trauma survivors exhibited reduced circadian stability relative to the comparison participants, characterised by greater sleep interday instability and higher variability.

Put plainly: it was not that survivors slept less, or slept later. It was that the rhythm itself became less reliable — more variation from one day to the next in when sleep started and ended.

The study’s larger question was whether that instability, measured two to six months after the event, was associated with PTSD status and severity at the eight-to-eleven-month follow-up. That is the prospective link the title refers to, and it is the reason the design matters more than any single number in it.

Why a Clock Would Be Affected at All

The circadian system is not only a sleep timer. It coordinates hormone release, core temperature, autonomic tone and alertness across the day, and it takes its cues from light, activity and routine.

A system that keeps time by regular input is a system that becomes less regular when the inputs do — and a period of acute threat disturbs almost all of them at once: sleep opportunity, daily structure, physical activity, and the arousal state that determines when sleep is possible at all.

The plausible reading is that circadian instability is downstream of that disruption rather than a separate problem. But this study measured the association, not the mechanism, and the distinction should be kept.

What the Study Cannot Establish

Four limits, and they are not small.

An association, even a prospective one, is not a cause. Circadian instability measured before a later assessment is consistent with instability contributing to risk, with both reflecting a shared underlying process, and with early symptoms producing the instability. The design narrows the possibilities; it does not close them.

It is one event and one cohort. These are survivors of a specific mass-trauma event, monitored in a specific period. Whether the pattern generalises to other kinds of trauma, or to trauma that is repeated rather than singular, is not answered here.

Wearables measure movement and heart rate, not sleep directly. They are the only practical way to monitor hundreds of people in real life for a month, and they are an inference about sleep rather than a recording of it.

And it says nothing about an individual. No one can look at their own sleep variability and read a risk from it. This is a group-level prospective association in a research cohort, and there is no clinical measure derived from it.

Why It Is Still a Useful Result

Because it changes what would be worth measuring. A field that keeps averaging its way to weak findings and then tries a measure of variability instead is doing something methodologically ordinary and consequential — and if instability turns out to carry signal that averages do not, that reframes both monitoring and, eventually, what might be worth stabilising.

It also fits a pattern that recurs across this research: sleep is rarely only about tiredness. It is an active period with its own work to do, and it is one of the more sensitive indicators that a nervous system is under sustained load.

The related question of what sleep reveals about how traumatic experience is processed is covered separately in what a study of nearly 28,000 women found.