A nationally representative survey of over 735,000 U.S. adults found that people who use marijuana are roughly 35% less likely to be obese than non-users, a pattern that held across demographic groups and years. However, because the study is observational and cross-sectional, it cannot determine whether marijuana itself plays any role in preventing obesity or whether other unmeasured factors explain the association.

U.S. adult obesity now exceeds 40%, making it among the most pressing chronic disease burdens in the country. Simultaneously, cannabis use has surged as states continue to legalize both medical and recreational access. Any credible signal linking these two trends demands scrutiny, because if marijuana use genuinely influences body weight regulation, it could reshape how clinicians think about both cannabis counseling and metabolic health. Understanding whether the association is real, artifactual, or confounded is essential before anyone considers cannabis as part of a weight management conversation.

This study analyzed seven annual waves (2016 through 2022) of the Behavioral Risk Factor Surveillance System, a Centers for Disease Control and Prevention telephone survey, focusing on the optional marijuana use module administered in a subset of U.S. states each year. The researchers pooled 735,921 respondents to examine whether self-reported marijuana use prevalence differed between obese and non-obese adults. The biological rationale draws on preclinical evidence suggesting that cannabinoids interact with the endocannabinoid system in ways that may modulate appetite, lipid metabolism, and insulin sensitivity, though these mechanisms remain poorly characterized in humans.

Current marijuana use prevalence doubled from 7.48% in 2016 to 14.91% in 2022, with use rates 81% higher in states with recreational legalization. After adjusting for age, sex, race, education, employment, smoking, and select comorbidities, obese adults in 2022 had significantly lower odds of both current marijuana use (adjusted OR 0.68) and daily use (adjusted OR 0.69) compared with non-obese adults. A dose-response pattern emerged across BMI categories, with progressively higher BMI associated with progressively lower odds of use. However, all data are self-reported, including height and weight, and the cross-sectional design cannot distinguish whether marijuana use precedes lower weight, whether leaner individuals are simply more inclined to use marijuana, or whether unmeasured factors like physical activity and dietary patterns drive both. The authors themselves acknowledge these constraints but then suggest that expanding legalization “may” reduce population obesity, a conclusion that substantially exceeds what this design can support.

I find the consistency of this association genuinely interesting. A 35% lower prevalence of marijuana use among obese individuals, replicated across seven survey years and dozens of subgroups, is not noise. The dose-response gradient across BMI categories is especially notable. But I have to be direct: this study cannot tell us that cannabis protects against obesity, and the suggestion that legalization might reduce population obesity rates is a leap that the data simply do not warrant. The gap between “intriguing signal” and “clinical recommendation” is enormous here.

In my practice, I see patients across the weight spectrum who use cannabis, and I do not counsel anyone to use marijuana for weight management based on this kind of evidence. What I do take from studies like this is a reason to pay closer attention to metabolic markers in my cannabis patients and to keep asking questions about how different cannabinoid profiles interact with appetite, energy balance, and metabolic health. The endocannabinoid system is deeply involved in metabolism, and we need prospective, controlled research to understand how to work with it thoughtfully rather than speculating from survey data.

For clinicians, this study sits at the very beginning of a research arc. It establishes that the previously reported inverse association between cannabis use and obesity, first documented in smaller datasets and the NESARC surveys of the early 2010s, persists in more recent, large-scale U.S. data and extends into the era of widespread legalization. That durability matters for hypothesis generation but does not move the evidence any closer to actionability. The field needs longitudinal cohort studies that track weight trajectories before and after cannabis initiation, ideally with objective measures of both body composition and cannabinoid exposure, before any clinical inference is justified.

From a pharmacological standpoint, the relationship between cannabinoids and energy balance is more nuanced than popular culture suggests. While THC acutely stimulates appetite, chronic exposure may downregulate CB1 receptor activity in ways that theoretically reduce energy storage, a mechanism supported by the history of the CB1 inverse agonist rimonabant, which produced weight loss but was withdrawn due to psychiatric side effects. Clinicians should be aware that different cannabis products, doses, and cannabinoid ratios may have very different metabolic effects. The one concrete takeaway: do not recommend cannabis for weight loss, but do continue monitoring metabolic parameters in cannabis-using patients as part of routine care.