Cannabis is widely used, and widely marketed, as a sleep aid. A study tracking more than a thousand adults for 35 years — from young adulthood into middle age — complicates that reputation considerably. Using data from the long-running Coronary Artery Risk Development in Young Adults (CARDIA) study, researchers led by Northwestern University found that both long-term cumulative cannabis use and recent use were associated with measurably worse objective sleep, even though the same users reported no difference in how well they thought they slept.
Three and a Half Decades of Follow-Up
The study, published October 9 in JAMA Network Open, drew on the CARDIA cohort, which has followed a racially diverse group of Black and white adults since young adulthood. Researchers analyzed cannabis use data spanning roughly 35 years of follow-up and linked it to sleep measured during the study's Year-35 exam (conducted 2020–2022) through a dedicated sleep ancillary study.
Sleep was assessed two ways: objectively, using wrist-worn actigraphy devices that participants wore continuously for seven days (891 participants), and subjectively, through standard sleep-quality questionnaires like the Pittsburgh Sleep Quality Index (1,128 participants). That dual measurement is what makes the study's central finding possible — and striking.
What the Watches Showed
Both cumulative lifetime cannabis use and more recent use were associated with worse objectively measured sleep. Specifically, cannabis use tracked with a lower percentage of time actually spent asleep while in bed, greater fragmentation of sleep (more frequent brief awakenings disrupting continuity), and more time spent lying awake after initially falling asleep. In adjusted models, the sleep-percentage estimate for cannabis use was -0.18 and the fragmentation estimate was 0.20, with confidence intervals excluding zero for both — a statistically meaningful signal rather than noise. Recent cannabis use was also associated with longer total sleep duration, suggesting users may spend more time in bed to compensate for less efficient, more broken-up sleep.
These associations held after the researchers adjusted for age, sex, race, cigarette smoking, alcohol use, physical activity, depression, chronic health conditions, and other substance use — ruling out several of the most obvious alternative explanations.
The Gap Between Feeling and Fact
The second half of the finding is arguably the more important one for anyone using cannabis specifically to sleep better: self-reported sleep quality showed essentially no association with cannabis use at all (an adjusted estimate of -0.003, effectively zero). Users were not reporting worse sleep — their perception of their own sleep looked no different from non-users. The disconnect only showed up when sleep was measured directly.
That gap matters because it means cannabis users relying on how rested they feel to judge their sleep may be missing a real, measurable pattern of fragmented, less efficient sleep that subjective reporting simply doesn't capture.
Limitations Worth Weighing
The study's authors are explicit that this design cannot establish that cannabis use causes worse sleep. Cannabis use and sleep were both measured using reasonably rigorous methods, but the actigraphy data reflect only one week in middle age, cannabis exposure relied on self-report across decades, and the study cannot rule out that people with worse sleep architecture for other reasons were more likely to use cannabis, rather than the reverse. The research was funded by the National Institutes of Health's National Heart, Lung, and Blood Institute and the Kaiser Foundation Research Institute.
What This Means for Patients
For the substantial share of adults who use cannabis believing it helps them sleep, this study is a useful corrective: feeling like cannabis helps and actually sleeping better are not the same thing, and the two can diverge in exactly the way this data suggests. Fragmented sleep and more time spent awake overnight — even when unnoticed — are patterns linked elsewhere in the sleep literature to worse next-day cognitive and cardiovascular outcomes over time.
That doesn't mean occasional cannabis use is dangerous for sleep, and this study wasn't designed to parse dose, frequency, or specific cannabis products. But for people with chronic sleep complaints who use cannabis regularly and feel it's working, these findings are a reasonable prompt to ask a sleep specialist about objective testing — a home sleep tracker or a clinical sleep study — rather than relying solely on how rested they feel in the morning.
