Marital Status, Marital Transitions, and Smoking Behavior: Gender and Life Course Considerations

Debra J. Umberson, University of Texas at Austin
Hui Liu, University of Texas at Austin

We analyze national longitudinal data to assess the impact of marital status and marital transitions on subsequent change in smoking behavior. Results from Zero-Inflated Poisson regressions show that, compared to the married, the odds of smoking are significantly greater for the continually never-married and continually divorced/separated men and women. The transition into marriage has no impact on subsequent change in smoking—providing no support for the idea that becoming married promotes health behavior change. Marital duration is associated with a reduction in the odds of smoking and in the amount of smoking over time—suggesting that marriage may have a cumulative positive impact on smoking behavior over the life course. Becoming divorced/separated exacerbates smoking among men but not women, having implications for long-term gender and marital status differences in health. Finally, life course stage is more important than gender in explaining how the transition to widowhood affects smoking behavior.