Changes in social capital and cigarette smoking behaviour over time: a population-based panel study of temporal relationships
Introduction: Identifying factors that influence individuals’ smoking behaviour remains a huge public health concern. This study aimed to investigate changes in individuals’ cigarette smoking whilst considering well-known smoking determinants, including social capital, its presence being postulated to reduce smoking.
Methods: From British Household Panel Survey data, two baseline smoking cohorts were created (‘smoking’ and ‘not smoking’). The same individuals from this nationally representative sample (NT = 8114, aged 16-91 years) were interviewed on four occasions between years 2000-07 to investigate changes in cigarette smoking behaviour. Logistic regression models with random effects compensated for within- individual behaviour over time. Temporal pathways were investigated by lagging independent variables (t-1) in relation to our cigarette-use outcome at time (t).
Results: Active social participation at (t-1) was positively associated with smoking cessation at (t), (Odds Ratio (OR) = 1.39; 95% CI (1.07-1.82)). Separating from one’s spouse at (t-1) increased risk for smoking relapse/initiation at (t) (OR = 6.63; 95% CI (1.70-28.89)). Conversely, being married protected against smoking cigarettes (OR = 1.87; 95% CI (1.15-3.04)). These associations held in our robustness checks.
Conclusions: Initial marital breakdown predicted a high risk of smoking relapse/initiation. The timing of this life-event provides a critical window where adverse smoking behaviour might occur. Conversely, the positive effects of active social participation on cigarette cessation remained consistent, its absence further predicting smoking relapse/initiation. Robustness of results confirms the important role that active participation has on cigarette smoking behaviour. Group smoking cessation interventions could harness participatory elements to better achieve their goals.
Nicotine and Tobacco Research
Volume and page numbers
18 , 2106 -2114
Not held in Research Library - bibliographic reference only