Abstract

Studies investigating smoking risk among children and adolescents have typically focused on social smoke exposure via parents, siblings, and peers. Recently, researchers found that second-hand smoke exposure measured with biomarkers among never-smokers uniquely predicted smoking initiation and greater endorsement of withdrawal sensations. Based on these findings, researchers posited a physiological pathway between second-hand smoke exposure and smoking behaviour may exist. The aim of the present study was to simultaneously investigate whether social smoke exposure and pharmacological exposure to nicotine both uniquely contribute to greater smoking risk among never-smoking youth. Participants included 338 never-smoking youth (53.5% females) aged 11-13 years (M = 12.68, SD = 0.67) attending 6 th or 7th grade in French-speaking schools. Participants completed self-report questionnaires measuring their own smoking behaviours, social smoke exposure (parents, siblings, peers, school), and known risk factors for eventual smoking (smoking expectancies, smoking susceptibility, perceived nicotine dependence). Each participant also provided a saliva sample and an expired breath sample, from which cotinine and carbon monoxide biomarkers were derived, to objectively measure second-hand exposure. Structural equation modeling was used to test the research hypotheses. Pharmacological exposure was not associated with smoking risk. Social smoke exposure of parental and peer smoking were significantly associated with smoking risk. When considered simultaneously, despite having models with acceptable to good fit, pharmacological and social smoke exposure together largely explained only a small proportion of the variance in smoking risk (1.4-4.7%), with the exception of peer smoking which explained considerable variance (58%). These findings do not suggest that pharmacological and social smoke exposure are differential predictors of smoking risk. Further, they do not support the possibility of a physiological pathway from second-hand exposure to smoking behaviour. Rather, the results suggest biomarkers may actually be a good proxy for social smoke exposure. To better evaluate the possibility of a physiological pathway, future studies should aim to recruit participants with a wider range of smoke exposure (i.e., low, moderate, or high exposure) and to more precisely measure longer-term exposure to second-hand smoke (e.g., hair nicotine, DNA encoding for CYP2A6, CYP2B6, and CYP2E1 enzymes which metabolize nicotine).