Split household and smoking behaviour of rural migrants in China.

This study aimed to address the direct impact of split households on smoking behaviour. Three types of spilt households (sole migration, couple migration, and family migration) were incorporated to explore the concise effect of different split household forms. This study also examined factors that contributed to the narrowing gap in smoking prevalence between 2008 and 2018. Retrospective observational study. Data were obtained from the Rural Urban Migration (RUMiC-2009) and China Household Income Project (CHIP2018). Analyses were conducted using chi-squared tests/analysis of variance tests and multiple logit regression. The order probit model with sample selection was conducted to correct for selection bias. Fairlie decomposition was used to quantify the contribution of individual variables to the observed differences in smoking prevalence. Smoking prevalence for all migrants decreased by 5.79% between 2008 and 2018. The results in 2008 reveal the positive and significant contribution of couple migration (coefficient = −0.4608; 95% CI = [−0.6453, −0.2762]) and family migration (coefficient = −0.3705; 95% CI = [−0.5959, −0.1452]) on the reduction of smoking; the finding for family migration remained robust in 2018. Measurable factors partially explain the decline in smoking disparity. The migration of families, working in the construction and manufacturing industries, and educational attainment were the largest contributing factors to the declining gap in smoking prevalence. The increase in family migration and education, and decrease in the number of workers in the construction and manufacturing industries, contributed to a decrease in smoking prevalence. Public policies should target sole migrants, couple migrants, individuals with lower education levels, and those working in the construction and manufacturing industries. [ABSTRACT FROM AUTHOR]