Abstract
OBJECTIVE: Examine changes in cigarette smoking prevalence among US adolescents by rural-urban residence, gender, and time. METHODS: We compared trends in smoking prevalence by rural-urban residence and gender using 18 years of cross-sectional data (2002-2019) from 298,530 respondents aged 12-17 in the US National Survey on Drug Use and Health. The two primary outcomes were past-month and daily smoking prevalence. Weighted logistic regression models tested whether smoking prevalence varied by rural-urban residence, gender, and time along with two-way and three-way interactions of these variables. RESULTS: Past-month and daily smoking declined over time for all four rural-urban residence and gender combinations. Regarding past-month smoking, there was a significant rural-urban residence by time interaction (p < .001) with higher prevalence and slower rates of decline among rural (adjusted odds ratio [AOR] = 0.86; 95 %CI = 0.85,0.87) versus urban participants (AOR = 0.83; 95 %CI = 0.82,0.84). A significant gender by time interaction (p < .001) was also observed. The rate of decline was faster among females (AOR = 0.82; 95 %CI = 0.81,0.83) versus males (AOR = 0.85; 95 %CI = 0.84,0.86) with prevalence initially higher (2002-2003) yet ending lower (2018-2019) among females. Regarding daily smoking, there was a significant rural-urban residence by time interaction (p = .005) with higher prevalence and slower rates of decline among rural (AOR = 0.82; 95 %CI = 0.80,0.85) versus urban participants (AOR = 0.79; 95 %CI = 0.78,0.80). CONCLUSIONS: Findings provide novel evidence of shrinking yet persistent rural disparities in both past-month and daily smoking, and less declines in past-month smoking among males over time. Together, findings support continued need for tobacco control and regulatory efforts to reduce adolescent smoking and specific efforts to reduce rural and gender disparities.