More than one in four women with HIV used cannabis over an 18-month period in a major US cohort study, roughly twice the rate of the general female population. Most used combustion-based methods, and about half shifted how often they used cannabis between visits, suggesting that single-visit screening may miss a significant portion of real-world use.

Women with HIV represent a growing, aging population facing layered health challenges, yet they remain underrepresented in cannabis research. As legalization expands and product diversification accelerates, clinicians need contemporary data on how their patients are actually using cannabis, by what methods, how often, and alongside what other substances. Without this foundational understanding, clinicians cannot effectively screen, counsel, or monitor. This study delivers precisely the descriptive baseline that has been missing.

Women with HIV in the United States face a unique intersection of chronic disease management, accelerated aging, polypharmacy, and social determinants that make cannabis use patterns particularly relevant to their clinical care. Researchers used data from the Women’s Interagency HIV Study (WIHS), a multi-site US cohort active since 1994, to characterize cannabis use among 1,246 women with HIV who attended all three semiannual visits between April 2018 and September 2019. The cohort was predominantly Black (65%), low-income (50% earning below $12,000 annually), and older (median age 52), providing a window into a population that is both clinically high-need and historically underrepresented in cannabis research.

The 18-month period prevalence of any cannabis use was 27%, approximately twice national estimates for US women. Among users, smoking was nearly universal (96%), while 30% consumed edibles, 18% vaped, and 34% reported multimodal use. Fifteen percent of the entire cohort reported daily or greater use. Cannabis users were significantly more likely to co-use alcohol (69% vs. 37%), cigarettes (61% vs. 29%), and illicit drugs (16% vs. 4%), and were less likely to have an undetectable HIV viral load (56% vs. 69%). Importantly, about half of users shifted frequency categories across visits, indicating dynamic rather than stable patterns. The authors note that all associations are unadjusted and cross-sectional, and they call for prospective, adjusted analyses linking cannabis use patterns to HIV clinical outcomes.

One in Four Women With HIV Used Cannabis—What the Data Really Show

Ask a patient once whether they use cannabis and you might get the right answer, or you might miss the whole picture. New data from one of the largest cohorts of women with HIV in the United States show that more than one in four women used cannabis over 18 months, that most of them smoked it, and that roughly half of them changed how often they used it between visits. This last finding is the one I keep returning to. The study uses a Sankey diagram to visualize frequency transitions across three semiannual visits, and it reveals something clinicians have suspected but rarely quantified: cannabis use is not a stable binary. Asking someone once whether they drink coffee tells you less than asking them over six months, because some people are daily espresso drinkers every day, while others go weeks without and then binge. A snapshot misses the whole story. The same principle applies here. This means that our current clinical screening tools, which typically capture a single yes or no at one visit, are systematically undercharacterizing the cannabis exposure of our patients. That is a genuine methodological contribution, and the study deserves credit for making it visible.

Where I urge caution is in the bivariate comparisons between cannabis users and non-users, particularly the association between cannabis use and lower viral suppression rates (56% vs. 69%). This is a statistically significant finding, and it is clinically provocative. But the study is descriptive, the comparison is unadjusted, and the confounding is formidable. Finding that people who carry umbrellas are more likely to get wet does not mean umbrellas cause wetness; it means both are explained by the rain. Similarly, cannabis use and lower viral suppression may both be explained by poverty, housing instability, mental health burden, and disengagement from care. The study’s background framing, which cites cannabis-associated harms in aging adults, can subtly prime readers to interpret this association causally, even though the authors themselves do not claim causation. I also note the absence of any data on why these women are using cannabis. We do not know whether use is recreational, therapeutic, or a form of self-medication for pain, anxiety, or the side effects of antiretroviral therapy. Without that context, the clinical meaning of the 27% prevalence figure remains incomplete.

What I would say to a patient is straightforward: I want to ask you about cannabis, not to judge, but because understanding how you use it helps me provide better care. What I would say to a colleague is equally direct: more than a quarter of our patients are using, mostly smoking, and their use patterns shift over time, so do not assume one negative screen means no use. And what I would say to a policymaker is that this population is high-need, underrepresented in research, and using cannabis at twice the national rate, often with combustion and increasingly with high-potency products in legal states. Describing who uses a substance, how often, and how is genuinely valuable work, but it is the beginning of a scientific question, not the answer. This study establishes the foundation; the hard work of understanding consequences, motivations, and intervention targets lies ahead.

This study sits at the early, descriptive stage of the research arc for cannabis use among women with HIV. It provides the contemporary prevalence and pattern data that are prerequisite to designing adjusted longitudinal analyses or intervention trials. In this respect, it fills an important niche: most existing cannabis research either excludes people with HIV or does not disaggregate findings by sex, leaving clinicians with little population-specific data to guide care for a group they see frequently.

From a pharmacological and safety standpoint, the near-universal reliance on combustion (96% of users smoked cannabis) is notable in a population with elevated respiratory risk due to HIV-related immune dysregulation, aging, and frequent tobacco co-use. Potential interactions between cannabis and antiretroviral therapies remain insufficiently studied, though some antiretrovirals are metabolized via CYP3A4 and CYP2C9 pathways that cannabinoids can modulate. Clinicians should incorporate multi-dimensional cannabis screening, capturing mode, frequency, and product type, into routine HIV clinical visits as a concrete, immediately actionable step drawn from this evidence.

This is an original observational cohort study using descriptive and bivariate analytic methods, situated in the lower-to-middle tier of the evidence hierarchy. It can characterize patterns and identify associations but cannot establish causal relationships. The single most important inference constraint is that all comparisons between cannabis users and non-users are unadjusted for confounders, meaning observed differences may reflect pre-existing population characteristics rather than any effect of cannabis use itself.

The 27% period prevalence of cannabis use is consistent with prior WIHS analyses that have documented elevated rates of substance use among women with HIV compared to the general population. This study extends earlier work by capturing mode and frequency across multiple time points, rather than relying on single-visit prevalence. It confirms findings from the National Survey on Drug Use and Health (NSDUH) showing that cannabis use is rising nationally among women, while demonstrating that the increase may be amplified in populations facing intersecting social disadvantage. No prior study of this scale has reported within-person frequency variability among women with HIV, which represents a novel contribution to the broader cannabis epidemiology literature.

The most consequential analytic choice was the restriction to women attending all three semiannual visits, which excluded 42% of eligible participants. If the excluded women had higher rates of cannabis use or worse health outcomes, as is plausible given that disengagement from cohort visits often correlates with social instability, then both the prevalence estimates and the health-characteristic comparisons could differ meaningfully. Additionally, multivariable adjustment for income, housing, mental health, and other substance use in the cannabis user versus non-user comparison would likely attenuate or eliminate the viral suppression association, potentially changing the study’s most clinically provocative finding.

The most likely overinterpretation is that cannabis use causes lower HIV viral suppression among women with HIV. This study found a cross-sectional association between the two, but the comparison is entirely unadjusted. Cannabis use and viral non-suppression share numerous common social determinants, including poverty, unstable housing, mental health burden, and care disengagement, any of which could explain the observed difference without cannabis playing a causal role. Readers should also avoid treating the 27% figure as a point prevalence; it represents any use across 18 months and three visits, and actual use at any single visit would be lower.

This study provides the most detailed contemporary description of cannabis use among women with HIV in the United States. It documents elevated prevalence, near-universal reliance on smoking, significant polysubstance co-use, and dynamic within-person frequency patterns. It does not establish whether cannabis use affects HIV outcomes in any direction. Its principal clinical implication is that routine, multi-dimensional, non-judgmental cannabis screening should become standard practice in HIV care settings.

Does cannabis use make HIV harder to manage?

This study found that women who used cannabis were less likely to have an undetectable viral load, but the finding is observational and not adjusted for other factors. Many of the same social challenges that make viral suppression harder, such as poverty, housing instability, and limited access to care, also correlate with higher cannabis use. We cannot conclude from this study that cannabis itself affects viral control.

Is smoking cannabis especially risky for women with HIV?

The study found that 96% of cannabis users smoked the drug, which is a concern because combustion produces irritants and toxins that can harm lung tissue. Women with HIV already face elevated respiratory risks due to immune changes, aging, and high rates of concurrent tobacco use. While this study did not measure lung outcomes directly, the dominance of smoking as a consumption method warrants clinical attention and conversations about alternative modes.

Should I tell my HIV doctor that I use cannabis?

Yes. Sharing this information allows your clinician to monitor for potential interactions with your HIV medications, keep an eye on respiratory health, and understand your full health picture. This study underscores that cannabis use is common among women with HIV, and the goal of screening is not judgment but better, more personalized care.

Does this study mean cannabis is harmful for people with HIV?

No. This study describes who is using cannabis, how they are using it, and how often. It does not measure health outcomes or test whether cannabis use is helpful or harmful. Separate studies are needed to answer those questions, and the authors call for exactly that kind of follow-up research.

Have thoughts on this? Share it: