Recently in Cancer, Sprehn et al reported powerful differences in overall survival among subcategories of unmarried patients (never married, divorced, separated, and widowed).1 Among 3.79 million patients in the Surveillance, Epidemiology, and End Results (SEER) database, patients in the subcategory of “separated” experienced the lowest 5- and 10-year overall survivals: 72% and 64%, respectively.1 However, according to the National Longitudinal Mortality Study, unmarried patients experience increased mortality across many disease states.2 Consequently, investigations into marital status and cancer should use cancer-specific survival (CSS) parameters to control for noncancer-related causes of death. The potential for a causal relation between marital status and cancer outcomes should be deliberated as well.
Therefore, we assessed CSS by marital status across all malignancies.3 This approach identified 2,845,392 patients in the SEER database aged 20 years and older. These patients were diagnosed between 1973 and 2006 with a single malignancy with known stage and marital status.3 We conducted multivariate Cox regression survival analysis to assess CSS while controlling for patient age, year of diagnosis, sex, race, ethnicity, nationality, tumor stage, grade, histology, primary site, and size.
Of the cohort, 61.1% were currently married, 12.2% never married, 8.2% divorced, 1.3% separated, and 17.1% widowed. Compared with currently married patients, separated patients had a 17.3% higher cancer-specific mortality (hazard ratio [HR], 1.173; 95% confidence interval [CI], 1.129-1.218; P<.001). Each subcategory of unmarried patients experienced greater cancer-specific mortality than married patients but to a lesser degree than separated patients.
Compared with previous studies, these results point to a causal relation between marital status and cancer outcomes.1 Further investigation is needed to probe potential differences in cancer biology by marital status, including the impact of psychology and social support.