Poor Sleep Is Cutting Years Off American Lives, Study Finds

Think about the last time you genuinely got a full night's rest. For roughly one in three American adults, that memory is hazy at best. Now a landmark study published in SLEEP Advances has done something no research team had attempted before: it mapped sleep insufficiency against life expectancy in every single one of the 3,143 counties across the United States — and the numbers are stark. Counties where more residents sleep fewer than seven hours a night consistently show shorter average lifespans, a pattern so robust it survived seven years of data and a global pandemic.

The Sleep Crisis Nobody Is Talking About

We talk endlessly about smoking bans, sugar taxes, and step counts. But sleep? It tends to get filed under personal habit rather than public health emergency. That framing may be about to change.

Millions of Americans clock fewer than seven hours of sleep on a typical night — the threshold below which the American Academy of Sleep Medicine and the Sleep Research Society define "insufficient sleep." We already knew from large meta-analyses that short sleepers face elevated risk of heart disease, diabetes, and all-cause mortality. What we did not know was how that relationship plays out at the neighbourhood level across an entire country, year after year.

That is the gap researchers at Oregon Health & Science University set out to close. Led by Andrew W. McHill and Kathryn E. McAuliffe, the team pulled six years of CDC Behavioral Risk Factor Surveillance System (BRFSS) data — a monthly telephone survey conducted by state health departments — and linked it to county-level life expectancy figures from 2019 through 2025.

Why County-Level Data Changes Everything

Previous research linking sleep duration to mortality typically followed cohorts of thousands of individuals over time — valuable, but limited in scope. This study took a different approach, treating the county as the unit of analysis across all 3,143 of them. That scale is not just statistically impressive; it is politically and practically significant.

Why? Because public health interventions are funded and deployed at the local level. A mayor, a county commissioner, a school board — these are the people who decide whether communities get sleep health programmes, later school start times, or workplace wellness policies. Showing that insufficient sleep correlates with life expectancy in their county, not some abstract national average, is the kind of evidence that actually moves budgets.

The statistical approach was rigorous. The team used mixed-effects modelling — a technique suited to "nested" data where counties sit inside states and observations pile up across years — to isolate sleep's contribution while controlling for smoking rates, physical inactivity, food insecurity, unemployment, insurance coverage, high school completion, and social connections.

How Does Insufficient Sleep Stack Up Against Smoking?

Here is where the findings get genuinely surprising. When the researchers ran their mixed-effects model across all counties from 2019 to 2025, insufficient sleep emerged as the second strongest predictor of reduced life expectancy among all the behaviours tested. Only smoking showed a stronger negative association.

To put that in perspective: sleep insufficiency outranked physical inactivity, food insecurity, social isolation, and lack of insurance as a predictor of shorter lifespans. The model estimated a standardised coefficient of −0.17 for insufficient sleep, versus −0.31 for smoking. These are not trivial margins.

When the team added obesity and diabetes to the model — both known consequences of poor sleep, and therefore potential mediators rather than simple confounders — sleep insufficiency remained statistically significant (coefficient: −0.08), with only smoking and obesity showing stronger effects. In other words, even if some of sleep's harm flows through weight gain and blood sugar dysregulation, there is still an independent, direct relationship between sleeping too little and dying sooner.

The geographic spread was equally striking. In 2025, all but three U.S. states showed a statistically significant negative correlation between sleep insufficiency rates and life expectancy at the county level. In Oregon, used as an illustrative example, the correlation reached r = −0.58 (p < 0.001) — a relationship the researchers describe as consistent and robust. The trend held every year from 2019 to 2024 as well, with significant associations appearing in 84% of states in 2019 and 100% of states in 2024.

What This Means for Your Community — and You

One of the most quietly powerful aspects of this research is what it says about equity. Sleep insufficiency's relationship with shorter life expectancy held regardless of a county's income level, healthcare access, or whether it was classified as urban or rural. That is unusual. Many major health risks — obesity, smoking, cardiovascular disease — show stark socioeconomic gradients. Sleep, it turns out, does not discriminate.

That cuts both ways. It means that a wealthy, well-insured suburb can suffer from the same sleep-driven mortality risk as a low-income rural county if residents are chronically underslept. But it also means that sleep health is a potential lever for equity: unlike access to a cardiologist or the cost of organic food, better sleep is theoretically achievable across income brackets with the right policies and education.

What would those policies look like? The researchers stop short of prescribing specifics, but the evidence base suggests a few directions. Later school start times — already recommended by the CDC — could help adolescents whose circadian rhythms are biologically shifted toward later sleep. Workplace policies that discourage or limit night-shift schedules could protect chronically sleep-deprived workers. And at the individual level, the message is blunt: sleep is not a luxury you can trade for an extra hour of productivity without paying a biological price.

The Questions That Still Need Answering

No study this ambitious is without limitations, and the researchers are admirably candid about theirs. The BRFSS data groups "long sleepers" — people regularly getting 9 or more hours — together with those getting a healthy 7 to 8 hours. Since excessive sleep also carries its own health risks (a finding well-documented in meta-analyses going back to 2010), this could actually underestimate the harm attributable to insufficient sleep specifically.

The dataset also cannot account for shift workers, people with undiagnosed sleep apnea, or those with psychiatric conditions — all groups where sleep deprivation is both more prevalent and more consequential. The COVID-19 pandemic years add further complexity: lockdowns and remote work demonstrably changed sleep patterns globally, though the study found similar associations throughout those years, which is itself reassuring.

And then there is the puzzle of why some states consistently show weaker associations. Is it climate? Cultural attitudes toward rest? Economic structures that allow for more flexible schedules? Those regional outliers are, the team suggests, a rich target for future investigation.

What is clear is that the field is moving toward treating sleep as a quantifiable, modifiable, place-specific public health target — not unlike the way air quality or water safety are approached. The data to support that shift now exists at an unprecedented scale.

• Sleep is the second-biggest lifestyle killer. — Among all modifiable health behaviours studied, insufficient sleep ranked behind only smoking in its negative association with life expectancy, outpacing physical inactivity and obesity.
• Geography and income don't protect you. — The sleep-longevity relationship was consistent across rich and poor counties, urban and rural areas, making this a universal rather than a niche concern.
• The data exists to act locally. — County-level evidence now gives mayors, school boards, and employers a specific, evidence-based reason to treat sleep health as a public priority rather than a personal habit.

"The breakdown of these findings to the county level highlights that this is a problem that is affecting all communities regardless of income level, access to health care services, or geographical classification." — McAuliffe et al., SLEEP Advances, 2025.