Long-Term Cannabis Users Said They Slept Fine. Wrist Monitors Logged 18 More Minutes Awake Each Night.

The wrist monitor and the person wearing it disagreed. In a 35-year Northwestern Medicine study, middle-aged adults with decades of cannabis use rated their sleep about like adults who never used it. The monitor, worn for a week, told a different story.

The study appeared October 9 in JAMA Network Open. It draws on 1,266 people from CARDIA, a cohort that enrolled Black and white adults aged 18 to 30 in 1985 and 1986 in four US cities. At the last exam, held between 2020 and 2022, the 891 people with usable monitor data averaged 60.7 years old.

What the wrist monitor recorded

The gap is clearest in the raw group averages. The 180 participants with the most lifetime exposure spent an average of 57.3 minutes awake after first falling asleep. That group had more than two cannabis-years, and one cannabis-year equals 365 days of use. The 141 who had never used cannabis averaged 39.3 minutes. The share of the night spent asleep was 87.7 percent against 91.2 percent.

Sorted by commonly used cutoffs, the split gets sharper. In the highest-exposure group, 27.2 percent slept less than 85 percent of the night, compared with 5.7 percent of never users. Long wake periods of 30 minutes or more covered 91.7 percent against 68.1 percent. That cutoff caught 75.1 percent of the whole sample, though, so waking at this age is common. The question is how much more of it the heavy-use group had.

Adjustment shrinks the numbers but keeps the direction. The researchers accounted for age, sex, race, smoking, alcohol, exercise, depression, chronic conditions and other substances. Each added cannabis-year still went with 21 percent higher odds of prolonged waking (odds ratio 1.21, 95% confidence interval 1.10 to 1.34) and 10 percent higher odds of low sleep efficiency.

Longer nights that were not better nights

Recent users, meaning anyone who had used in the past 30 days, were the odder case. Against never users, they had 3.23 times the odds of prolonged waking and 4.60 times the odds of sleeping more than eight hours. The authors suggest people may read a longer night as a better one. The interval on that 4.60 runs from 1.81 to 11.70, and only 108 participants were recent users. Read it as a direction, not a measurement.

Most of those 108 were not casual users. Ninety-two of them sat in the highest lifetime-exposure group, so recent here mostly means long-time and still going.

“The wearable device detected more nighttime disruption among participants with long-term cannabis exposure”

Vu, the corresponding author, called the mismatch one of the most interesting findings. Participants generally did not report sleeping worse, and their Pittsburgh Sleep Quality Index scores showed no link to cannabis use. Senior author Mercedes R. Carnethon, who chairs preventive medicine at Feinberg, put the conclusion bluntly: the data do not support the belief that cannabis helps as a sleep aid.

What the study cannot say

Northwestern's own release says the study was not designed to show whether cannabis directly causes sleep problems. The paper lists the reasons. Cannabis use was self-reported, with no data on potency, how it was taken, or how close to bedtime. Potency and legal status changed a lot over 35 years. Sleep was monitored for a single week at one exam, so the study cannot say which came first.

The comparison groups also differ in other ways. In the highest-exposure group, 56.7 percent had ever smoked cigarettes, against 5.0 percent of never users. Median alcohol intake was 12.2 milliliters a day against zero. The authors adjusted for both and report the link held, but they concede residual confounding is possible. The cohort is also in its sixties, and may not reflect younger users or today's higher-potency products. Coverage so far traces back to the paper and Northwestern's release, and no outside researcher's comment turned up.

Medical cannabis research pulls the other way, with a catch. Meta-analyses cited in the paper report small gains in self-rated sleep. Those trials ran eight weeks or less and mostly enrolled clinic patients with pain, cancer or sleep disorders. The authors' guess is that cannabis may ease symptoms that wake people up, so they feel better rested even when objective measures barely move. Smaller earlier studies lean the new way. A monitor study of 178 young adults and a lab study of 177 healthy adults both tied heavier use to more time awake at night.

Why the feeling is the problem

About 30 percent of US adults reported trouble sleeping between 2017 and 2020, according to national data the paper cites, and cannabis is widely promoted as a fix. If users cannot feel a difference the monitor can see, a self-rating is a poor test of whether it is working. The authors call for assessing both perceived and measured sleep, and for longer trials with objective monitoring.

Co-author Kristen L. Knutson, an associate professor of neurology who specializes in sleep medicine, points to the unglamorous options. A regular bedtime. A cool, dark, quiet bedroom. No screens right before bed. A calming routine such as reading or a shower. She adds that persistent problems, or loud snoring, are worth raising with a health care provider to rule out a sleep disorder.