The Canary in the Coal Mine Is Coughing: Electronic Cigarettes and Respiratory Symptoms in Adolescents

Since the introduction of electronic cigarettes (e-cigarettes) in the United States a decade ago, use of these devices has increased substantially. From 2010 to 2013, the prevalence of current e-cigarette use increased from 0.3 to 6.8%, with the highest prevalence (14.2%) in young adults aged 18 to 24 years (1). Perhaps most concerning is the increased use among adolescents. E-cigarettes are now the most popular tobacco product among both high school and middle school students, with a prevalence of 16 and 5.3%, respectively (2). Many e-cigarette users and some health care providers believe that e-cigarettes are safer than conventional cigarettes (3, 4), prompting some public health organizations, such as Public Health England, to strongly support the use of e-cigarettes as a harm reduction tool, despite a relative lack of data on their safety.

Notably, although studies of the harms of e-cigarettes are limited, early reports suggest that e-cigarettes are not harmless. Cell-based and animal studies have shown that e-cigarettes are associated with a number of toxicities, including increased inflammation and oxidative stress in the lung (5, 6), endothelial dysfunction (7), and impaired pulmonary immunity (8–10). Human studies have similarly suggested that e-cigarettes may impair immunity (11), negatively affect vascular function (12), and increase peripheral airway resistance (13). Additional human studies, particularly those focusing on clinically relevant outcomes, are needed to better characterize the health effects of e-cigarettes to accurately inform regulatory bodies and public perception of these products.

In this issue of the Journal, McConnell and colleagues (pp. 1043–1049) report their analysis of the relationship between e-cigarette use and respiratory symptoms in the Southern California Children’s Health Study (14). Children in kindergarten or first grade from 12 communities in southern California were initially enrolled in this prospective cohort from 2002 to 2003 and followed with questionnaires yearly until 2008 and every other year thereafter. The questionnaires assessed a variety of topics, including sociodemographic information, tobacco product usage, and clinical symptoms. The 2014 questionnaire was the first to assess e-cigarette use and is the focus of this report, with 2,086 subjects providing information on both e-cigarette use and respiratory symptoms, including wheezing and bronchitis. Bronchitic symptoms were defined as self-report of cough for 3 months in a row, congestion or phlegm other than when accompanied by a cold, or bronchitis in the past 12 months.

The authors found that current and past e-cigarette use were both associated with increased odds of bronchitic symptoms, both in unadjusted analyses and in a multivariable model adjusting for sociodemographic characteristics. The risk of bronchitic symptoms increased with frequency of e-cigarette usage over the prior 30 days. Notably, after additional adjustment for lifetime cigarette usage and secondhand smoke exposure, these associations were attenuated, although past e-cigarette use remained significantly associated with bronchitic symptoms. A sensitivity analysis restricted to never smokers showed similar odds of bronchitic symptoms related to past and current e-cigarette use as the overall cohort, after adjustment for sociodemographic factors and secondhand smoke exposure. These findings echo results from a cross-sectional study of Chinese adolescents, which reported an association between respiratory symptoms and e-cigarette use (15).

This study has several strengths that make it a significant addition to the small but rapidly growing body of literature on the potential harms of e-cigarettes. First, this study assesses a large population of adolescents, in whom e-cigarette use is increasing rapidly. Notably, nearly half of current or past e-cigarette users in this study were never smokers, emphasizing the concern that this age group may be particularly vulnerable to trying these products. Furthermore, although e-cigarette use was only assessed in the 2014 survey, subjects had been followed over many years, allowing for a thorough evaluation of potential confounders in multivariate models. The inclusion of sensitivity analyses adjusting for bronchitic symptoms reported in previous questionnaires makes it more likely that the newly reported symptoms are related to e-cigarette use.

Although this study represents an important contribution to the literature on the potential harms of e-cigarettes, areas of uncertainty remain. First, this study was based on self-report, and the symptoms described are fairly nonspecific. Future human studies will need more objective and detailed quantification of the effects of e-cigarettes on lung physiology, biology, and pathology.

Second, given the heterogeneity in e-cigarette devices and products, it remains unclear whether particular device types, usage patterns, or flavors are more likely to produce harm. Future studies in both animal models and humans need to assess e-cigarette device components, usage patterns, and flavors to address critical gaps in our understanding of these relatively new products, particularly in light of the Food and Drug Administration’s recently asserted regulatory authority over product and device characteristics.

Third, as illustrated by this study, studies of e-cigarettes must account for potential confounding effects of conventional cigarettes, given the high prevalence of dual use. In this study, inclusion of lifetime cigarettes smoked and secondhand smoke exposure in multivariate models significantly attenuated associations between e-cigarette usage and respiratory symptoms. Although the authors did assess for exposure to cigarette smoke, passive smoking by self-report is relatively crude, may miss biologically significant lower levels of exposure, and is prone to recall bias (16). The inclusion of biomarkers of conventional cigarette exposure in future studies may be helpful, particularly in accounting for passive smoke exposure.

Last, although beyond the scope of the study at hand, it will be critical to assess how e-cigarette use compares to and affects the risk of using other tobacco products with known harms. Some studies have suggested that e-cigarettes may increase the risk of conventional cigarette or cigar use in adolescents (17, 18), and initial hopes that they would be highly effective for smoking cessation have not been borne out (19). If e-cigarettes are ultimately found to increase nicotine addiction and/or promote use of combustible tobacco products, they may prove more harmful than isolated studies of their physiologic toxicities suggest. We also need additional studies directly comparing the physiologic and biologic toxicities of e-cigarettes to those of combustible tobacco products, to fully compare the risks of these devices to any potential benefit via harm reduction.

It has taken decades to understand much of the harm caused by conventional cigarettes and years to reshape public perceptions of these harms. With lessons learned from the tobacco epidemic, we should be able to expedite our investigations of e-cigarettes, with a focus on their potential toxicities as well as their impact on addiction and use of combustible tobacco products. Regarding toxicity specifically, the scientific community should prioritize the evaluation of e-cigarettes through human studies as well as laboratory models, using endpoints that reflect both short-term toxicity and potential long-term harms, to ultimately improve our understanding of these increasingly popular devices and to inform regulation.