BACKGROUND: Long-term tobacco and alcohol consumption is a strong mortality predictor. Previous studies, due to their cross-sectional design, were unable to identify the change trajectories of tobacco and alcohol consumption and their representative demographic characteristics. Therefore, this study analyzes consumption trajectories, identifies demographic characteristics, and determines high-risk mortality trajectories.
METHODS: The analysis included 174968 participants (99273 males and 75695 females) aged 40-79 from the National Health Screening Cohort (2002-2019). Smoking (tobacco/day) and alcohol consumption (g/day) trajectories from 2002 to 2009 were assessed by gender using Multiple-Group Latent Class Analysis. Cluster analysis identified the demographic characteristics of each trajectory. Cox proportional-hazards models estimated HR for all-cause, cardiovascular disease (CVD), and cancer mortality over a 10.7-year follow-up (2008/2009–2019).
RESULTS: In males, eight-trajectories from 2002 to 2009 were identified: Type A (non-smokers/non-drinkers, Trajectory 2-(16.6%)), Type B (non-smokers/drinkers, Trajectory 1-(8.5%); 3-(26.3%); 5-(3.5%)), Type C (smokers/non-drinkers, Trajectory 6-(7.7%)), Type D (reduced smoking/reduced drinking, Trajectory 4-(8.2%)), and Type E (smokers/drinkers, Trajectories 7-(23.9%); 8-(5.3%)).
Type D consisted of younger, metropolitan, NHI, normal weight, exercised 3–4 times/week. Type E was younger, rural, medical aid, obese, inactive individuals. In addition, Types C and E, characterized by smoking, had higher HR for all-cause and cause-specific mortality than Type A during the follow-up.
In females, five-trajectories were identified: Type α (non-smokers/non-drinkers, Trajectory 3-(68.7%)), Type β (non-smokers/drinkers, Trajectories 1-(13.1%); 2-(6.0%); 4-(10.2%)), and Type γ (smokers/drinkers, Trajectory 5-(2.0%)).
Type γ consisted of older, urban, medical aid, underweight, inactive individuals. Compared to Type α, Type γ was associated with all-cause and cause-specific mortality.
CONCLUSIONS: The long-term analysis showed that non-smokers or non-drinkers did not transition into smokers or drinkers. Among smokers or drinkers, consumption decreased with age. However, in certain male and female groups, consumption stayed high, increasing mortality risk. Co-occurring smoking and alcohol further raised mortality risk, with smoking having a greater impact than alcohol.