Degree Name

DrPH (Doctor of Public Health)


Public Health

Date of Award


Committee Chair or Co-Chairs

James Anderson

Committee Members

Shimin Zheng, Deborah Slawson


Each year there are more than 350 000 new cases and nearly 100 000 deaths attributed to colorectal, female breast, and cervical cancer in the United States. Screening tests can reduce morbidity and mortality associated with these cancers. Patient marital status has been associated with health outcomes, but no study has focused on the relationship of marriage with disease stage and survival for the 3 cancers with established screening guidance. It is critical to identify special populations that may be at risk for poor cancer outcomes.

The objective of this study was to examine the relationship of marital status with disease stage at the time of diagnosis and cancer-specific survival among population-based cohorts of patients diagnosed with invasive colorectal, breast, or cervical cancers. Subjects came from states or regions reporting to the Surveillance, Epidemiology, and End Results (SEER) tumor registries. The study included more than 243 500 patients diagnosed between January 1st 2004 and December 31st 2006 with 1 of these 3 cancers and who were followed for a minimum of 3 years. Descriptive statistics were calculated to summarize patient demographic and clinical characteristics. Baseline category logit models were fit to evaluate the association between marital status and disease stage. Kaplan-Meier survival curves and Cox proportional hazards models were developed to evaluate differences in patient survival across 4 marital status categories.

Married adults with colorectal, breast, and cervical cancer were diagnosed at an earlier disease stage than those who were divorced/separated, widowed, or single. After controlling for stage and demographic factors, married patients also experienced superior cancer-specific survival (range: 19-33% better) as compared to those in non-married groups.

Divorced/separated, widowed, and single adults are a subset of the population that may benefit from targeted prevention or care initiatives for cancers than can be detected early. Social support networks, selection effects, or other causal mechanisms likely moderate the protective association observed between marriage and cancer outcomes. These findings characterize a meaningful disparity in health outcomes. Additional person-level data on preventive health behaviors and treatment decisions could help solidify understanding of the issue and improve the ability to design effective research, interventions, and policy.

Document Type

Dissertation - unrestricted


Copyright by the authors.

Included in

Epidemiology Commons