Wrist trackers log more broken sleep in long-term cannabis users who rate their sleep no worse than non-users

What makes the analysis useful is that the same people supplied both a device reading and a self-report in the same week ^[4]. It was led by Northwestern Medicine and published in JAMA Network Open ^[11]. On the wrist, more lifetime exposure went with a smaller share of the night spent asleep, more frequent waking and longer stretches awake after first falling asleep ^[5]. On the questionnaires, users rated their sleep no lower than non-users did ^[6].

"This was one of the most interesting findings," said Thanh-Huyen T. Vu, the corresponding author and a research associate professor of preventive medicine at Northwestern University Feinberg School of Medicine ^[12]. "The wearable device detected more nighttime disruption among participants with long-term cannabis exposure, but the participants themselves generally did not report worse sleep." ^[8]

The exposure side is the study's strength. Cannabis use was recorded repeatedly over the decades ^[3], so lifetime exposure does not depend on what a middle-aged participant remembers of early adulthood. The sleep side is thinner: seven nights at one point in life ^[4]. The release calls the differences significant but does not say how large they were, or how many of the 1,266 were recent users. Without those figures, a reader cannot tell whether the gap is a few minutes of wakefulness a night or something a person would feel the next day.

The design is observational, and the release itself says it was not built to establish direct causation ^[9]. The list of adjustments is long ^[7]. Adjustment does not settle direction. The release notes that millions of adults use cannabis products as sleep aids ^[14], so people who already slept badly would be expected to turn up among users. A single week of measurement in midlife cannot show which came first.

The release's interpretation also goes further than its instrument. It says that while cannabis may sedate users or blunt their awareness of awakenings, "it disrupts the micro-architecture of restorative sleep beneath the surface" ^[13]. The outcomes it actually reports are efficiency, fragmentation, wake time after sleep onset and, for recent users, long sleep duration ^[5][10].

The thing this doesn't tell you is whether a given user sleeps better on the nights they use. The study compares users with non-users. It does not set one person's nights with cannabis against the same person's nights without. That within-person comparison is what someone means when they say cannabis helps them sleep. The questionnaire gap also has more than one reading. Users may not notice their awakenings, as the release suggests ^[13], or a questionnaire and a wrist monitor may be measuring different quantities.

I think the cohort supports a narrower conclusion than "Cannabis Destroys Sleep Quality," the headline Neuroscience News ran ^[1]. Among these middle-aged adults, how users rated their sleep was a poor guide to what the wrist monitor recorded ^[5][6].