Factors associated with developmental trajectories of smoking behaviour change among older people in Taiwan: Findings from the Taiwan Longitudinal Study on Aging

While there is evidence that smoking cessation can improve health even late in life and in those who are already ill, there is very little evidence on smoking behaviour change in the over 50s. This study aims to investigate factors that could influence smoking behaviour change in the over 50s. 3,132 elderly from the Taiwan Longitudinal Study on Aging were interviewed in 1996, 1999, 2003, and 2007.

Group-based trajectory modelling was applied to identify clusters of individuals that have similar underlying aetiologies or risk factors predicting or influencing changes in smoking behaviour over time. Three smoking trajectory groups were identified: non-smoker (75.8%), highly-unlikely to quit (15.7%) and gradually reducing smoking prevalence (8.5%) through time.

Females, those aged 65 and above, and non-literate smokers were significantly less likely to be quitting smoking group. We found that marital status, living arrangement and satisfaction, and income have no significant association with any trajectory groups. Factors including having been diagnosed by doctors with one or more chronic disease, self-rated bad health, having one or more limitation to instrumental activities, or having been admitted to hospital or emergency room in the previous year, were significantly associated with reducing participants' smoking prevalence. The tendency of chronic health conditions to reduce smoking behaviour is particularly interesting. This may be due to increased contact with health professionals. Alternatively, it may be that a direct experience of the health consequences of smoking is the motivating factor for reduction or cessation.