Cannabis Does Not Consistently Change Sleep Architecture, Meta-Analysis Finds

A new systematic review and meta-analysis of polysomnographic evidence shows neither clear benefit nor clear harm on most objective sleep measures, but withdrawal from regular cannabis use reliably disrupts sleep, raising important questions about why so many users report subjective improvement.

Why This Matters

Cannabis is one of the most commonly self-reported sleep aids in North America, yet clinical guidance has largely relied on subjective reports rather than objective measurement. The gap between what patients believe cannabis does for their sleep and what laboratory instruments actually record has significant implications for how clinicians counsel the growing number of patients who use cannabis specifically for insomnia or poor sleep quality. This 2025 review arrives at a moment when regulatory shifts and commercial marketing have far outpaced the evidence base, making a rigorous synthesis of polysomnographic data both timely and necessary.

Clinical Summary

Cannabis interacts with sleep physiology through the endocannabinoid system, which modulates circadian signaling, arousal, and sleep-wake transitions. The expectation that exogenous cannabinoids would reliably alter sleep architecture is biologically plausible, particularly given early laboratory studies from the 1970s suggesting that high-dose THC suppresses REM sleep. A team led by researchers at the University of British Columbia and Western University conducted a systematic review and meta-analysis, published in Sleep Medicine Reviews in 2025, to determine whether polysomnographic evidence, the gold standard for objective sleep measurement, supports the widely held belief that cannabis meaningfully changes sleep structure. They screened studies across Embase, Medline, and Web of Science, ultimately identifying 18 studies that used polysomnography to assess the effects of various cannabinoid formulations on sleep parameters.

Of those 18 studies, only 9 provided data compatible with quantitative meta-analysis. Across these pooled analyses, cannabis administration did not consistently alter total sleep time, sleep onset latency, wake after sleep onset, sleep efficiency, or proportions of REM and slow-wave sleep. The early REM suppression findings that have long shaped clinical assumptions were concentrated in small-sample studies using very high THC doses with substantial methodological limitations; more recent and better-designed studies yielded mixed or null results. The most robust and consistent finding concerned cannabis withdrawal: cessation of regular use was reliably associated with reduced total sleep time, prolonged sleep onset latency, and REM rebound. The authors emphasize that the small number of eligible studies, the enormous heterogeneity across dose, cannabinoid type, delivery method, prior use history, and patient populations, and the inability to address subjective sleep benefits all limit the generalizability of the pooled estimates. They call for adequately powered, placebo-controlled polysomnographic trials before clinical recommendations can be made.

Dr. Caplan’s Take

This review does something genuinely valuable: it forces us to confront the distance between what patients feel cannabis does for their sleep and what objective instruments actually measure. The null findings on sleep architecture are not proof that cannabis has no effect, but they are a corrective to the assumption that subjective improvement equates to measurable sleep enhancement. Every week I hear from patients who are certain cannabis helps them sleep, and the honest answer is that we cannot yet confirm or deny that belief with polysomnographic evidence. What we can say, and what this review makes clear, is that withdrawal reliably disrupts sleep, which means some of the perceived benefit may simply be the avoidance of withdrawal-related insomnia.

In practice, I do not recommend cannabis as a first-line sleep intervention, and I use this kind of evidence to have a candid conversation with patients about the difference between feeling rested and objectively improved sleep. For patients already using cannabis regularly, I emphasize the withdrawal sleep disruption risk and work on gradual, supported reduction strategies when appropriate. I prioritize behavioral interventions for insomnia and reserve pharmacological options for cases where those approaches are insufficient, always grounding the discussion in what the evidence can and cannot support.

Clinical Perspective

This review sits at a relatively early point in the research arc for cannabis and objective sleep outcomes. It confirms that the older literature on REM suppression, which has been disproportionately influential, does not hold up under modern scrutiny with better-designed studies. It also establishes that the subjective-objective discrepancy in cannabis sleep research is real and unresolved, meaning that clinicians should not interpret patient-reported improvement as evidence of architectural benefit. For patient-facing conversations, the evidence currently supports neither recommending cannabis for sleep improvement nor categorically warning against it on sleep-architecture grounds alone. What it does support is informing regular users that cessation may produce meaningful, temporary sleep disruption.

From a pharmacological standpoint, the enormous variability in cannabinoid composition, dose, and delivery method across the included studies means that no specific formulation can be singled out as beneficial or harmful. Clinicians should be aware that patients using high-THC products may face more pronounced withdrawal sleep effects, and that CBD-dominant formulations remain poorly studied with polysomnography. Drug-interaction considerations are relevant for patients on sedative-hypnotics, as cannabinoids may alter cytochrome P450 metabolism. One actionable recommendation: when a patient reports using cannabis for sleep, ask specifically about what happens when they stop, as this often reveals whether the perceived benefit is pharmacological enhancement or withdrawal avoidance, and that distinction changes the clinical approach.

Study at a Glance

Study Type

Systematic review and meta-analysis

Population

Cannabis-naive individuals, chronic users, and patients with insomnia, PTSD, chronic pain, obstructive sleep apnea, and neurological conditions

Intervention

THC, CBD, dronabinol, nabilone; multiple delivery methods

Comparator

Placebo or baseline (pre-intervention) measurements

Primary Outcomes

Polysomnographic parameters: total sleep time, sleep onset latency, wake after sleep onset, sleep efficiency, REM and slow-wave sleep proportions

Sample Size

18 studies identified; 9 included in quantitative meta-analysis

Journal

Sleep Medicine Reviews

Year

2025

DOI or PMID

Not provided in source data

Funding Source

Not specified; author affiliations include University of British Columbia, Western University, and Tranq Sleep Care

What Kind of Evidence Is This

This is a systematic review and meta-analysis, which occupies a high position in the evidence hierarchy when the underlying studies are numerous and methodologically strong. In this case, however, only 9 of 18 identified studies were suitable for quantitative pooling, and those studies varied enormously in design, dosing, population, and cannabinoid formulation. The single most important inference constraint is that null findings from a small, heterogeneous meta-analysis cannot be interpreted as strong evidence of no effect; they may instead reflect insufficient statistical power to detect real effects that exist in specific subpopulations or dosing contexts.

How This Fits With the Broader Literature

This review directly challenges the longstanding narrative, rooted in early 1970s laboratory work, that THC reliably suppresses REM sleep. By contextualizing those findings within their methodological limitations, it reframes a body of evidence that has been cited for decades without adequate critical appraisal. The withdrawal findings align well with established literature on cannabis discontinuation syndromes, including work by Budney and colleagues documenting sleep disturbance as a core withdrawal symptom. The subjective-objective discrepancy echoes broader patterns seen in insomnia research, where patients frequently overestimate sleep improvement from pharmacological interventions compared to what polysomnography records. This review extends prior qualitative reviews by adding the first formal meta-analytic pooling of polysomnographic cannabis data, though its conclusions remain preliminary given the small analytic sample.

Common Misreadings

The most likely overinterpretation is reading the null meta-analytic findings as evidence that cannabis is “safe for sleep” or “has no effect on sleep.” A null result in an underpowered meta-analysis with substantial heterogeneity does not establish absence of effect; it establishes absence of consistent, detectable effect across a small and varied set of studies. Equally problematic would be interpreting the withdrawal findings as applicable to all cannabis users regardless of frequency, duration, or dose. The withdrawal signal was observed in regular, often heavy users, and should not be generalized to occasional or low-dose use without supporting evidence.

Bottom Line

This first meta-analysis of polysomnographic cannabis data finds no consistent objective evidence that cannabis improves or worsens sleep architecture during active use, while confirming that withdrawal from regular use reliably disrupts sleep. The findings are preliminary, limited by a small analytic sample and enormous study heterogeneity. For clinical practice now, the most defensible position is honest uncertainty about sleep benefits combined with clear counseling about withdrawal-related sleep disruption for regular users.

References

1. Systematic review and meta-analysis of polysomnographic evidence on cannabis and sleep architecture. Sleep Medicine Reviews, 2025. University of British Columbia, Western University, Tranq Sleep Care. DOI not available in source data.
2. Budney AJ, Hughes JR, Moore BA, Vandrey R. Review of the validity and significance of cannabis withdrawal syndrome. American Journal of Psychiatry. 2004;161(11):1967-1977.