Over 90% of smokers start smoking before age 18.1 Smoking initiation in adolescence increases risk of tobacco-related mortality and morbidity, regardless of smoking status in adulthood.2 Secondhand smoke (SHS) exposure also causes premature death and illness; large proportions of the population are still exposed to SHS, including nearly 50% of children.3,4 Peer smoking influences both tobacco use and SHS exposure among youth.5 Yet even though youth with close friends who smoke are more likely to be exposed to SHS and to use tobacco than youth with no close friends who smoke,6,7 caregiver or parental smoking can have a sizable influence on youth tobacco use and SHS exposure.8-10 As the American Academy of Pediatrics Section on Tobacco Control notes, it is a public health priority to address caregiver and parental tobacco use in order to reduce tobacco use and SHS exposure, as parental tobacco use can impact youth behavior as well as mental and physical health across the life course.8 It is difficult for many adult smokers to reach sustained tobacco cessation, and therefore, implementing voluntary, smoke-free rules in homes and cars is a desirable step for reducing tobaccorelated harm among youth.8,11 Smoke-free public policies are becoming more common but they rarely address private spaces,12 which are common locations for SHS exposure for children.13,14 Prevalence of voluntary smoke-free rules in the home has increased from 43% in 1992 to 1993 to 83% in 2010 to 2011, and these rules tend to be more common in households with children under the age of 18.15,16 However, recent evidence from Minnesota demonstrates that over 80% of smokers do not restrict smoking in both the home and car.14 This study also found that adults living with a child in the home were not more likely to implement smoke-free rules in cars,14 corroborating previous findings on the lack of smoke-free car rules among caregivers and parents who smoke.17 Implementing comprehensive smoke-free household rules (ie, smoke-free rules in the home and car14) could potentially help protect youth from the risk of early initiation of tobacco use as well as harms of SHS exposure.18-20 Yet there is a dearth of population-based research on how voluntary implementation of comprehensive smoke-free rules impacts tobacco use and SHS exposure among children and adolescents, particularly in the context of the changing tobacco landscape.21 In this study, we examine how household smoking restrictions relate to tobacco-related behavior and harm among adolescents who live with a smoker. We categorize household smoke-free rules according to comprehensive (home and car), partial (home or car only), and no (neither home nor car) smoke-free rules. We examine youth tobacco use—both combustible cigarette and other tobacco products in light of recent youth smoking trends22-26—as well as SHS exposure using data from the 2014 Minnesota Youth Tobacco Survey (MYTS).

Among a host of physiological and psychological factors that contribute to the maintenance of smoking behavior, using smoking as a regulation strategy in order to reduce negative affect figures prominently (Tate & Stanton, 1990; Tomkins, 1966; Torres & O’Dell, 2016). For example, smokers report that stress motivates them to smoke and stress-induced smoking has also been demonstrated experimentally (Kassel, Stroud, & Paronis, 2003; Marks, Murray, Evans, & Vida Estacio, 2011). Eating is another way for many individuals to cope with stress (Greeno & Wing, 1994). In contrast to smoking, which is increased by stress, however, stress may increase food intake in some individuals, but can also lead to reduced food intake in others (Oliver & Wardle, 1999). Smoking appears to be a primary reinforcer while reinforcing properties of natural rewards such as food are reduced in current smokers. For example, current smokers had reduced activation of reward-related brain areas in response to food cues (Jastreboff et al., 2015) and—in contrast to nonsmokers—showed an approach bias towards smoking-related cues but not towards food cues (Machulska, Zlomuzica, Adolph, Rinck, & Margraf, 2015). In line with these findings, current smokers tend to have a lower body mass index (BMI) than nonsmokers, which may be due to both nicotine-induced increases in energy expenditure and decreases in appetite (Chiolero, Faeh, Paccaud, & Cornuz, 2008). In this study, relationships between smoking status, stress, stress eating, and BMI were examined. Based on the above-mentioned findings, it was expected that current smokers would have a lower BMI than nonsmokers. Furthermore, current smokers were expected to report a tendency to eat less when stressed, due to their preference for smoking in response to stress. In addition, if such a mutual exclusiveness of either smoking or eating in response to stress represents a central mechanism of reduced BMI in smokers, then this stress-induced reduction in food intake would mediate the effect of smoking status on BMI. Finally, it was expected that these effects would be particularly pronounced in individuals who actually report experiencing high levels of stress and, thus, that perceived stress would moderate the indirect effect of smoking status on BMI. Therefore, a moderated mediation model was proposed, in which being a smoker was associated with a lower BMI through less stress eating, particularly in stressed individuals (Figure 1).