Abstract Body (Do not enter title and authors here): Background: Evidence from observational cohort studies suggests that lifelong unmarried status is associated with worse outcomes in men but not women. The reasons for these sex differences is poorly understood.

Purpose: Identify sex differences in health-seeking behaviors according to relationship status that correspond to sex differences in clinical outcomes

Methods: Data from study visits reporting multiple social and health-seeking behaviors were obtained from the Atherosclerosis Risk in Communities (ARIC, year 23) study, Multiethnic Study of Atherosclerosis (MESA, year 10) and Framingham Heart Study Offspring (FHS, year 40) were pooled and harmonized. Behaviors including recent doctor visits, medication compliance, activity and caloric intake were summarized by relationship status and sex. Behaviors found to be different between relationship groups were then analyzed for association with all-cause mortality using Cox proportional hazards analysis.

Results: There were 2653 subjects with incident HF, of whom 1479 were married, 208 were divorced, 78 were never married, 33 were separated, and 784 were widowed. Never-married status was associated significantly higher risk of all-cause mortality in men (HR 1.83 [1.11-3.04, p=0.019]), but not women (HR 0.75 [95% CI 0.44-1.27]). Compared with unmarried men, married men had significantly better medication adherence (Morisky score, 44% vs. 25%, p=0.003), emotional support (ESSI5 16.7±3.8 vs. 13.8±4.3, p<0.001), and social support (Berkman Social Network Questionnaire 26.6±6.3 vs. 22.9±6.4, p<0.001), whereas there were no differences in these metrics between married and unmarried women. Of these, the higher ESSI and BSNQ indices were both associated with reduced all-cause mortality, whereas the Morisky score was not.

Conclusion: Married men experience significantly more social and emotional support than unmarried men, whereas women do not. Personal support systems may be an important contributor to the relationship between unmarried status and HF outcomes in men.