Longitudinal Examination of Loneliness and Smoking in Adolescents and Young Adults: A Multiple Dataset Study

Loneliness might be associated with smoking, especially in youth. We examined temporal associations from loneliness to smoking behaviours among adolescents and young adults using a multi-dataset approach. For adolescents (sample 1: N = 600, M age = 12.6, 52.7% female; sample 2: N = 551, M age = 12.8, 54.3% female; sample 3: N = 261, M age = 13.8, 46.7% female), discrete-time survival analyses revealed that loneliness did not predict smoking initiation over time. For young adults (sample 4: N = 2,636, M age = 21.7, 70.5% female; sample 5: N = 404, M age = 22.6, 65.6% female), structural equation models revealed that baseline loneliness did not predict smoking status (non-smoker/occasional smoker/regular smoker) after 6 and 18 months (except in sample 4 after 18 months; β = 0.07, p = 0.035). Complementary analyses showed that baseline loneliness preceded regular but not occasional smoking. Utilizing multiple datasets enabled focusing on overall patterns and obtaining robust insights. Future studies could explore at-risk subgroups. With complex behaviours like smoking, many risk factors play a role, each having small effects.