The association between late-adolescent smoking and long-term mortality: a dose-response relationship

Abstract

Abstract Introduction Tobacco smoking has detrimental medical effects. However, whether late-adolescence tobacco smoking increases early mortality risk is unclear. The aim of our study was to determine the association between tobacco smoking among Israeli adolescents and mortality in young adulthood. Methods The Israeli Defense Forces (IDF) health survey enrolls annually random sample of 3–5% of IDF recruits who voluntarily report lifestyle and tobacco consumption habits (up to 10 cigarettes a day were defined as light smokers, 10–20 cigarettes a day and above 20 cigarettes a day as medium and heavy smokers respectively). Medical and sociodemographic variables (education, residential socioeconomic status, country of birth or origin) were included. Included in this study were 62,134 soldiers (56% men; average age 17.3 years) who enrolled in the study between 1981 and 2006. Data from the Israeli Ministry of Interior was obtained by linkage. Follow-up terminated at the earliest of December 31st 2018 or death. Cox proportional hazard models were applied. Results There were 42,013 (67.6%) never smokers, 2,108 (3.4%) past smokers, 6,886 (11.1%) light, 9,880 (15.9%) medium and 1,247 (2%) heavy smokers. During 1,348,517 person-years, 548 deaths were recorded (327 were never smokers, and 18,54,121 and 28 were past, light, medium and heavy smokers, respectively; mean age 31.3 years). The hazard ratios for all-cause mortality were 0.98 (0.61–1.58), 1.02 (0.76–1.36), 1.32 (1.06–1.64) and 1.53 (1.03–2.30) for past smokers, light, medium and heavy smokers, respectively, after adjustment for age, gender, BMI, enrollment year, education and intelligence score. There was no sex-specific association and mortality risk among heavy smokers persisted even when the outcome was set as death by the age of 30 years. Conclusion Late-adolescent smoking is associated with long-term all-cause mortality. A dose-response relationship was demonstrated between number of cigarettes consumed at the age of 17 and the future risk of death. Kaplan-Meier Plot Funding Acknowledgement Type of funding source: None