In this cross-sectional population-based study of 379,049 adults, inhaled cannabis use was consistently associated with asthma in the overall population, among those with no lifetime tobacco use, across the age spectrum, and in sensitivity analyses using the outcome of any lifetime history of asthma. The odds of asthma increased with increasing frequency of cannabis use in a dose–response manner. Inhaled cannabis use was associated with COPD in the overall population and among adults < 50 years old. We saw an elevated odds of COPD among those with no lifetime tobacco cigarette use in a dose–response manner although this was not statistically significant. These findings demonstrate clear and consistent associations between cannabis inhalation and asthma and raise concern that cannabis inhalation could be associated with COPD.

The high prevalence of cannabis use among those with chronic lung disease is concerning and appears to be increasing in younger birth cohorts. We found a high prevalence of daily cannabis use among younger adults with asthma (9%) or COPD (15%). This highlights the need to screen those with asthma and/or COPD for cannabis use, as part of the comprehensive health evaluation recommended in clinical guidelines for asthma²⁹ and COPD.²⁷ Assessing cannabis use may be increasingly important in the future as today’s young adults age.

Our results linking cannabis use with asthma are consistent with other work. Analysis of the 2020 National Survey on Drug Use and Health, a population-based survey among those 12 + years in the USA, found associations of nearly identical magnitude between past-30 day cannabis use and asthma.²¹ Past 12-month cannabis use was also associated with asthma medication prescriptions in Norway’s nationalized health records system,³⁰ and urine tests positive for cannabis and/or billing codes for cannabis use disorder were associated with asthma in claims data from an integrated US health system.³¹ The present analysis goes beyond these earlier studies^21,30,31 by finding a similar association among those with no lifetime use of tobacco cigarettes and demonstrates a dose–response relationship between frequency of cannabis use and odds of asthma.

We found an elevated odds of COPD with inhaled cannabis use among adults of all ages and those younger than 50 years. These associations were less precise when we restricted to those with no lifetime tobacco cigarette use to address confounding and were not statistically significant. However, as lack of statistical significance does not provide evidence of no effect,³² these results do not provide reassurance that inhaled cannabis is safe with respect to COPD.

This study adds to a growing body of literature that raises concern regarding the association of cannabis with COPD. Cannabis smoke is associated with bronchial inflammation in animal models³³ and in clinical studies¹² with similar pathologic findings as in those who smoke tobacco cigarettes.^10,34 Cannabis is associated with declines in the ratio of forced expiratory volume in one second (FEV1) to forced vital capacity (FVC), which is necessary to diagnose COPD. However, this decline is primarily due to increase in FVC with no consistent detrimental impact on FEV1, a metric of COPD severity.^{7,12,17,18,35,36} FEV1 was examined in one landmark longitudinal study¹⁸ that included 795 adults who used cannabis only followed over 20 years. The median lifetime cannabis use among participants was relatively light at 0.9 joint-years (1 joint-year being equivalent to 1 joint smoked daily for 1 year); modeling found that cannabis inhalation was associated with declines in FEV1 above > 7 joint-years. In another analysis that included 75 adults who used cannabis only with heavier use (mean 54 joint-years) and 91 adults who used cannabis plus tobacco cigarettes, cannabis use was associated with a decline in FEV1, but results were not reported for those who only used cannabis.⁵ In longitudinal studies, cannabis has been associated with a diagnosis of COPD^18,36 and precursor changes in pulmonary function⁷ though one study—which included only adults who formerly or currently used tobacco cigarettes—found no impact of cannabis on COPD incidence or progression.²²

In our analysis, the lack of association between inhaled cannabis and COPD among the age 50 + subgroup (aOR 1.13, 95% CI 0.96–1.33) merits discussion. Today’s older adults may not have accrued sufficient cannabis exposure—in terms of intensity and/or duration—to lead to a measurable association with odds of COPD,¹⁷ which requires decades to develop fully in contrast to asthma.³⁷ In our data, adults older than 50 years had approximately threefold lower prevalence of daily cannabis use than younger adults. Further, the lag time from cannabis initiation to measurable impact on lung function may last decades, but population-level data indicate that older adults have only recently initiated cannabis use. In 2018, the prevalence of cannabis use among those 65 years and older was 5.7%, despite use being essentially zero (0%) in 2002, according to US national surveys.³⁸ Any past-year cannabis use among those 50 + years increased from 5.3% in 2015 to 12% in 2021.^39,40 In a prospective cohort with cannabis use measured as joint-years since age 17, reductions in gas exchange and small airway dysfunction (which may progress to COPD) among adults who used cannabis were only detected after age 45, up to 28 years after initiating cannabis use.^7,41 In another study, the likelihood of chronic bronchitis increased only after > 10 joint-years of cannabis use.¹⁹ Thus, it may be that today’s young adults—who use cannabis more heavily than those in the past—may be the first generation to provide sufficient insight into the question of whether cannabis impacts COPD risk as they age. Future analyses with reliable measures of cumulative lifetime cannabis exposure are needed to corroborate our observed association between cannabis inhalation and COPD among younger adults and all adults in aggregate.

This study addressed key limitations of prior research to understand cannabis’ impacts on pulmonary health by analyzing data from a large number of individuals at younger and older ages, reporting results separately for those with no lifetime tobacco cigarette use, and examining associations by frequency of cannabis use. Specifically, our analysis had a large number of adults who inhaled cannabis and denied lifetime tobacco cigarette use, a key group to isolate the health effects of cannabis. A systematic review and meta-analysis on the impact of cannabis on pulmonary function included 22 studies, which in aggregate contributed 243 individuals who used cannabis only.¹⁰ In contrast, our study included 7277 adults who used only cannabis among 221,767 lifetime tobacco non-smokers. In addition, we restricted to adults who primarily inhaled cannabis, the most biologically plausible route that cannabis would impact pulmonary health, and our exposure of past-30 days inhaled cannabis use matches that proposed for clinical care to screen for cannabis use.¹⁴

This study has limitations. While cross-sectional studies—including hundreds that rely on BRFSS data⁴²—play an essential role in biomedical research to provide preliminary evidence of associations,⁴³ including the link between age at initiation of tobacco use and lung cancer risk,⁴⁴ they have specific limitations. First, we could not assess cannabis use prior to each diagnosis. However, an association is biologically plausible¹⁷ and our results are corroborated by longitudinal studies on COPD^18,36 or precursor changes in lung function.⁷ Further it is unlikely that the observed dose–response relationships are explained by reverse causality. Second, our outcomes are prevalent diagnoses which tend to be more indolent,⁴⁵ and thus, our results could underestimate associations between inhaled cannabis and aggressive chronic pulmonary disease. Despite these limitations, utilizing cross-sectional data allows us to examine associations that would otherwise require following large prospective cohorts for an extended duration. Answering these questions with a prospective cohort would be resource-intensive, impractical, and unlikely to yield meaningful results quickly. In the meantime, 3.7 million Americans initiate cannabis use each year² and deserve to understand the potential health impacts of cannabis inhalation utilizing the full breadth of information available on the health effects of cannabis use including cross-sectional data. Third, exposure and outcome data were self-reported. However, cannabis is more likely disclosed than other drugs and self-reported use shows good agreement with urine toxicology results.⁴⁶ The majority of our survey data was collected in settings with legal cannabis use which promotes accurate disclosure of use; that is, of 79 state-year combinations in which the module was collected (e.g., Alaska in 2016), cannabis was legal in 49.⁴⁷ Outcomes were self-reported, though COPD⁴⁸ and asthma⁴⁹ may be under-diagnosed. Non-differential misclassification of cannabis use, asthma, or COPD would underestimate associations. Fourth, we do not have an estimate of total lifetime exposure which could introduce exposure misclassification especially for those with heavy cannabis use who recently quit or those who recently initiated use. If those with respiratory conditions were to reduce cannabis inhalation, then this would tend to underestimate associations. Finally, we did not have data on all-known risk factors for chronic lung disease such as breastfeeding, other early childhood exposures, prior pulmonary infections, or occupational exposures,⁵⁰ though the relative contribution of each of these risk factors is overshadowed by that of tobacco smoke^27,51 which we addressed through covariate adjustment and exclusion those with lifetime tobacco cigarette use.