Science1 publisher2 min readPublished
European neurologists link chronic insomnia to a 26% higher risk of stroke
European Academy of Neurology researchers link chronic insomnia to a 26% higher stroke risk across studies of more than 1.3 million people. The evidence is observational: strong enough to flag insomnia for closer checks, but it cannot show that treating insomnia prevents strokes.
The Scientist · Science desk
What happened
- People with chronic insomnia were 28% more likely to be admitted to hospital, according to a separate meta-analysis of five studies.
- Seven outcomes were in scope, including mortality and workplace accidents, with new meta-analyses run for stroke, hospital admissions and occupational accidents.
- Earlier systematic reviews it drew on linked insomnia to dementia, especially Alzheimer's disease, to major depression and to suicidal behaviour, with uneven statistical consistency.
- The authors say the associations are observational and do not show that insomnia causes any of these outcomes.
- MedUni Vienna co-led the work, which was published open access in Sleep Medicine Reviews on 22 September 2026.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- decision Clinicians can treat a complaint of chronic insomnia as grounds for a closer look at vascular and cognitive risk, because a warning sign only has to predict and does not have to be proven as a cause.
- constraint Insomnia definitions varied, and sleeping-pill use or cardiometabolic health may confound the result, so the 26% cannot be read as stroke risk a patient would lose by sleeping better.
- contradiction The release's summary says more than 1.3 million people were pooled, but its body puts that total on nine studies and pools only six, so the pooled sample is an unstated share of that figure.
The 1.3 million figure is the combined enrolment of the nine stroke studies the team identified [3]. The 26% estimate pools six of them, leaving three outside [1]. The release's own summary calls it a meta-analysis "pooling data across more than 1.3 million participants" [4]. That description fits the nine studies, not the six. The release does not give the participant count for the six pooled studies, confidence intervals, absolute stroke rates, or the result of the new workplace-accidents meta-analysis [5].
Relative risk also hides how big the effect is. A 26% rise on a rare outcome adds few strokes per thousand people. The same rise on a common outcome adds many.
The authors are direct about the limits. The associations are observational and do not show cause and effect [7]. Studies defined insomnia differently, and prescription hypnotic use or a patient's baseline cardiometabolic status could confound the estimates [8]. If sleeping pills account for part of the excess, some of the risk comes from the treatment. If heart and metabolic disease both disturb sleep and raise stroke risk, insomnia is marking the disease. The 28% hospitalisation figure [2] also has a problem of direction, since people who are already becoming ill may sleep badly before they are admitted.
The brain-health links are less firm again. Four of the seven outcomes (dementia, depression, suicidal behaviour and mortality) got no new pooled analysis [2]. According to the release, statistical consistency varied across the findings drawn from earlier reviews [6].
The authors' conclusion is narrower than it first sounds. They argue that chronic insomnia should no longer be treated as an isolated complaint and should instead be prioritised as an early clinical red flag for declining systemic and cognitive health [9]. A red flag only has to predict. So a patient's report of chronic insomnia can prompt a closer look at vascular and cognitive risk on the strength of the association alone. I think the conclusion fits this evidence, with one condition. It supports screening. It does not support the claim that treating insomnia would lower anyone's stroke risk, because an observational review cannot test that [7].
Stefan Seidel, a co-author, is a neurologist and sleep specialist at MedUni Vienna and started the study as an EAN member [11]. He wants sleep taken beyond the clinic. "Sleep should be part of the discussion on brain health at every stage of life. We should raise awareness of insomnia and its treatment and ensure that sleep is also taken into account in health promotion programmes in communities, schools and the workplace," he said [10].
What to watch
- Randomised trials that assign people to insomnia treatment and track strokes or hospital admissions, the only design that could test whether the link is causal.
- Whether the European Academy of Neurology or national neurology bodies write insomnia screening into stroke or brain-health guidance.
- Follow-up analyses that separate people taking prescription hypnotics from those who are not, to see whether the 26% estimate holds.