Relationship transitions and change in health behavior : A four-phase, twelve-year longitudinal study

Rationale: Extensive scientific evidence shows an association between involvement in social relationships and healthy lifestyle. Prospective studies with many participants and long follow-ups are needed to study the dynamics and change in social factors within individuals over time. Objective: Our aim was to determine whether a change in relationship status (single, married, divorced, widow, cohabiting) is followed by a change in health behavior (smoking, alcohol consumption, physical activity, and body mass index). Methods: We used data from 81,925 healthy adults participating in the prospective longitudinal Finnish Public Sector Study in the period 2000-2013. We analyzed 327,700 person-observations from four data collection phases. Missing data were multiply imputed. A within-individual methodology was used to minimize the possibility of selection effects affecting the interpretation. Results: All four health behaviors showed associations with relationship status. The effects were very similar and in the same direction in women and men, although there were gender differences in the magnitudes of the effects. The end of a relationship was followed by a decrease in body mass index, increased odds of being a smoker, increase in physical activity, and increase in alcohol consumption (widowed men). The effects were reverse when forming a new relationship. Conclusion: A change in relationship status is associated with a change in health behavior. The association is not explained by socioeconomic status, subjective health status, or anxiety level. People leaving or losing a relationship are at increased risk of unhealthy behavior (smoking and alcohol consumption), but at the same time they have a lower BMI and show higher physical activity compared to the time they were in a relationship. It is not clear if the cumulative health effect of these health behavior changes is positive or negative.Peer reviewe