How Differential Are the Effects of Smoking Cannabis versus Tobacco on Lung Function?

Cannabis use has increased in recent years with the decriminalization of its production and medicinal or recreational use. Because of the similarity in smoke contents between cannabis and tobacco (1), there is a nagging concern that smoking cannabis might have deleterious effects on lung function, similar to the well-known consequences of tobacco smoking. However, the relatively sparse literature involving systematic examinations of the impact of cannabis use on lung function suggests little independent effect of cannabis on FEV₁ (2–9) and an actual increase in FVC and other measures of lung volume (7–9). These results are in contrast to the clearly detrimental effect of tobacco on FEV₁ with little or no effect on FVC in otherwise healthy individuals. Studies of the relationship between cannabis and small airway function have been mixed, one showing no effect on multiple measures of peripheral airways function (3) and another showing a decrease in maximal mid-expiratory flow (forced expiratory flow [FEF]_25–75%) (2). Moreover, the few studies that have examined the effect of cannabis on specific airway conductance (sGaw) have uniformly demonstrated a modest decrease (3–7), implying a detrimental effect on central airways of uncertain clinical significance, but consistent with histologic evidence of pathology similar to that found in tobacco smokers in bronchial biopsies (10). The few reported longitudinal studies of the effect of cannabis on lung function have yielded mixed results, some not showing any impact on lung function decline (7, 11) and others suggesting an acceleration in the trajectory of FEV₁ decline compared with control nonusers over periods of several to 20 years (12, 13), although the latter effects were mainly confined to those with heavy lifetime use of cannabis (20 or more joint-years, i.e., number of joints/day times number of years cannabis was regularly smoked). It is also noteworthy that most published studies to date have mainly included young adults despite recognition that the deleterious effects of tobacco on lung function may not become detectable until later in life. In view of the latter issue and the mixed results in the literature, the study of the effects of cannabis on lung function in mid–adult life reported by Hancox and colleagues in this issue of the Journal (pp. 1179–1185) is of particular interest (14).

Hancox and colleagues report the results of an extensive battery of lung function measurements at age 45 years in the Dunedin (New Zealand) cohort they have followed since birth with a principal focus on the relationship between self-reported cannabis use (adjusted for concomitant tobacco use) and pulmonary function. The importance of this cohort study is accentuated by the fact that most studies of the impact of cannabis use on lung health have been largely limited to younger users. Among 45-year-old participants in the current study, findings in association with cannabis use, adjusted for sex, height, weight, and tobacco use in pack-years, are largely similar to those reported previously in the same birth cohort at age 32: higher FVC (or trend thereto), normal FEV₁ but lower FEV₁/FVC (attributable to the trend toward an increased FVC), higher TLC, FRC, residual volume and Va, and lower sGaw (7).

A major exception was a significantly lower FEF_25–75 at age 45 (n = 881) that was not reported in the same cohort at age 32 (n = 919). This finding accords with the significantly reduced maximal expiratory flow rates at 50% and 75% of forced expired volume reported in association with cannabis in an earlier population-based study (2) but conflicts with the findings of no significant effect of cannabis use on several measures of peripheral airways function in a convenience sample of heavy habitual smokers (3). However, because participants in these latter studies were mainly younger adults, the effect of cannabis smoking on lung function in older individuals could not be adequately assessed.

The finding of a significantly reduced FEF_25–75 in independent association with cannabis in the present study (14) suggests that the differential effects between tobacco and cannabis smoking might be confined mainly to the disparate effects on FVC, which might be attributable not to any intrinsic effect of cannabis smoke but rather to the unique smoking topography characteristic of cannabis smokers, who usually take deeper inhalations with a much longer breath-holding time and larger inspiratory volume compared with the typical profile of tobacco smoking (15). The characteristic smoking profile of cannabis smokers appears to be analogous to the deep inhalations and long breath-holds of competitive swimmers, who have been shown to have higher than normal lung volumes (16). In addition to the effect on FEF_25–75, another similarity between the effects of cannabis and those of tobacco on lung function is the qualitatively similar effects on sGaw, as shown in the current and previous studies (3, 6, 7).

An important finding in the current study is a significantly lower transfer capacity of the lung for the uptake of carbon monoxide (Tl_CO) and Tl_CO/Va in cannabis smokers among current or ex-tobacco smokers with adjustment for tobacco pack-years. Nevertheless, the Tl_CO was not reduced in approximately half of the cohort who never smoked tobacco, possibly implying an additive effect of cannabis and tobacco on this outcome. However, if cannabis were indeed associated with a reduced Dl_CO in mid–adult life, it would be another example of a qualitative similarity to a tobacco effect. A reduced Dl_CO could imply a possible relationship to early emphysema, although no independent relationship of self-reported cannabis use to macroscopic emphysema on thoracic computed tomographyscans has been found (6, 17). On the other hand, several case series have reported an association of cannabis with bullous lung disease (18). Clearly, further studies are needed to confirm the relationship between cannabis and diffusion impairment and examine its clinical implications.

It is further noteworthy that cannabis smokers in the current study were mostly light smokers (fewer than 5 joint-years), consistent with the general observation that cannabis is usually smoked in a much lower quantity compared with tobacco. Because a significant impact of cannabis on FEV₁ has been found previously, mostly in heavy smokers of 20 or more joint-years (9, 12, 13), future studies should include larger numbers of heavy cannabis smokers with lung function assessed at later stages of life.