These results suggest that the respiratory effects of cannabis can decrease with the use of a vaporizer. The data reveal that respiratory symptoms like cough, phlegm, and tightness in the chest increase with cigarette use and cannabis use, but are less severe among users of a vaporizer. Because a sample this large can produce statistically significant effects that might not be clinically meaningful, a focus on odds ratios could prove fruitful. The odds ratio suggests that vaporizer users are only 40% as likely to report respiratory symptoms as users who do not vaporize, even when age, sex, cigarette use, and amount of cannabis consumed are controlled. The use of cigarettes in conjunction with cannabis exacerbated symptoms, as found in previous work [6]. The interaction between vaporizer use and cannabis consumption also appeared, suggesting that a vaporizer should have more impact on respiratory symptoms in people who use more marijuana. Odds ratios suggest that these effects are relatively small, but interactions often prove difficult to detect at all [9].

Several important limitations of these data deserve mention, particularly those related to sampling, Internet reporting, limitations of our measures, and the lack of random assignment to vaporizer use. In an effort to find regular users of cannabis, we targeted people with a potential interest in changing cannabis policy. These individuals might have consciously or unconsciously minimized their reports of symptoms that might cast cannabis in a negative light. Nevertheless, literally thousands of participants admitted to experiencing respiratory symptoms. The symptoms covaried with cannabis use, cigarette use, and the interaction of the two, as work with samples gathered in other ways has revealed [6]. These results suggest that reports among these participants are comparable to those found in other work. Any bias in reporting remains a problem, and only further work can help address this issue.

Vaporizer users might be more inclined to minimize respiratory symptoms than people who smoke cannabis in other ways. The price of a vaporizer can range as high as hundreds of dollars. Vaporizers also lack some of the convenience of other methods of marijuana use. Users who have spent this much money and effort might minimize reports of their respiratory symptoms, consciously or inadvertently, in an effort to justify their actions. Only a more objective measure of respiratory function that does not rely on self-report can sidestep this potential problem. Laboratory measures of lung function would make a nice addition to further work on this topic. Nevertheless, roughly 1/3 of the participants who used a vaporizer (52/152) did report symptoms, suggesting that self-report biases on symptom reports likely does not account for the entire phenomenon.

The use of the Internet for this type of work has advantages and disadvantages as well. This approach might lead individuals who are unwilling to travel to the laboratory to participate, potentially increasing generalizability. Heavy users with severe symptoms might be particularly disinclined to participate without the convenience of the Internet.

Recent work also suggests that people report more drug use while using the Internet than they do on standard paper-and-pencil measures [10]. Nevertheless, because Internet access was required for participation, these data might not generalize to meaningful subsets of the population without such access.

Our measures of vaporizer use and respiratory symptoms could also have been more detailed. A single question about the primary technique used for administering cannabis neglects potentially meaningful variation in vaporizer use. Some participants might use a vaporizer primarily but also smoke cannabis. In contrast, other participants might use a vaporizer exclusively. Both of these groups of participants would end up in the group who uses a vaporizer primarily. Vaporizers come in several forms, including conduction-style machines that employ a hot plate as well as convection-style devices that use warmed air. The efficacy of these different machines could vary substantially, but we could not address the question with the current data. These limitations, however, should decrease power rather than create a spurious result. By lumping participants who occasionally smoke cannabis into the same group with those who vaporize exclusively, we actually weaken the ability to detect effects. Including any type of vaporizer, no matter how effective, also has the potential to weaken effects. In a sense, the current study's estimate of the effect of a vaporizer on respiratory symptoms might be an underestimate of the improvement that could arise from a good vaporizer used as the exclusive method for ingesting cannabis. We also did not assess the length of time each participant had used a vaporizer. A vaporizer's impact might grow more dramatic with longer use. The assessment of respiratory symptoms was also not particularly elaborate, but the same measures revealed a significant impact of cannabis and cigarettes in this sample and in other work [6].

Finally, the use of a vaporizer was not randomly assigned. The possibility exists that cannabis users who choose a vaporizer might engage in a host of other behaviors designed to minimize respiratory symptoms, or simply be more health conscious in general. Like any correlational study, this one cannot address the role of causality. The current data are consistent, however, with the idea that cannabis vaporizers can decrease respiratory symptoms in regular users of the plant. A better test of a vaporizer's potential for minimizing problems would require recruiting cannabis smokers who report respiratory troubles, randomly assigning a group to use a vaporizer, and assessing any decrease in symptoms. The current data suggest that such an intervention could prove helpful.

Although the use of a vaporizer has the potential to increase the safety of cannabis as far as respiratory symptoms are concerned, pulmonary problems are not the only potential negative consequences of the plant. Reviews suggest that 9–12% of cannabis users develop symptoms of dependence [11]. Cannabis can lead to impaired driving skills [12], and heavy use in adolescence might create deviant brain structure [13] as well as decreases in intelligence [14]. A vaporizer offers no protection against these negative consequences. Nevertheless, a vaporizer has considerable potential for increasing cannabis drug safety by minimizing pulmonary troubles.