Sleeping too much can be just as dangerous for your heart, brain and arteries as sleeping too little. It sounds counterintuitive; for years the message has simply been “sleep more”; but that’s the conclusion of the largest meta-analysis to date on sleep and mortality, published in the journal GeroScience.
The team, led by researchers at the University of Oklahoma and Semmelweis University in Budapest, reviewed 79 cohort studies with at least one year of follow-up, after screening more than 31,000 initial records. The goal was to answer a question that hadn’t been updated in 15 years, since Cappuccio et al.’s pioneering work: exactly how little sleep is too little, and how much is too much?
Sleeping less than 7 hours raises the risk of death by 14%. That’s the pooled result from 46 studies comparing people with short sleep (under 7 hours a night) against those sleeping 7 to 8 hours, the reference range. The calculated hazard ratio was 1.14 (95% confidence interval, 1.10 to 1.18), a statistically significant increase consistent with decades of prior research on sleep deprivation.
What’s more surprising is what happens at the other extreme. Sleeping 9 hours or more is associated with a 34% higher risk of death, according to an analysis of 49 studies (hazard ratio 1.34; confidence interval, 1.26 to 1.42), well above the risk linked to short sleep. Oversleeping isn’t a harmless luxury: in this analysis it’s the more dangerous of the two extremes.
The study also broke the data down by sex, and that’s where the most striking figures appear. In men, short sleep raises the risk by 16% and long sleep by 36%. In women, short sleep raises it by 14%, but sleeping 8 hours or more sends it up to 44% (hazard ratio 1.44), the highest figure in the entire analysis. The authors point to hormonal differences, sleep architecture, and the prevalence of certain health conditions between sexes as possible explanations, though they acknowledge more research is needed to confirm this.
The cause isn’t a single mechanism, but several at once. Sleep deprivation has been linked to more inflammation, more oxidative stress, worse regulation of glucose and cholesterol, and added strain on the cardiovascular system. Deep sleep is also when the brain activates its glymphatic system, the mechanism that clears out proteins like beta-amyloid, linked to Alzheimer’s; sleeping poorly or too little interferes with that cleanup. For excessive sleep, the authors point to a different and less alarming explanation than it first appears: sleeping a lot can, in many cases, be a sign (not the cause) of an illness that’s already present but not yet diagnosed.
That nuance matters. This is a meta-analysis of observational studies, not clinical trials that randomly assign how much each participant sleeps, so it can’t prove causation in either direction. The authors note that heterogeneity between studies was very high in the long-sleep group (I2 of 92%, versus 51% for short sleep), and that the link between sleeping a lot and dying sooner could be explained, at least in part, by reverse causation: it’s not that sleeping 9 hours kills you, but that being sick makes you sleep more. Most of the included studies, moreover, measured sleep through self-reported questionnaires rather than objective sensors, adding another possible source of error.
With those caveats in place, the core message holds: the 7-to-8-hour range remains, based on the evidence gathered so far, the one associated with the lowest mortality risk, and both falling short and overshooting deserve equal attention in public health policy, not just the short-sleep end of the spectrum.
Source: Ungvari Z, Fekete M, Varga P, et al. “Imbalanced sleep increases mortality risk by 14-34%: a meta-analysis.” GeroScience, 2025. Open access (CC BY 4.0). DOI: 10.1007/s11357-025-01592-y. PMID: 40072785.
