Predictors of Depressive Symptomatology in Middle-Aged Married Female Cancer Survivors and Their Husbands Compared to Couples without Cancer

Background and Purpose: As of January 1, 2007, there were approximately 11.7 million cancer survivors in the U.S. (National Cancer Institute, 2010), a number expected to grow exponentially as the population ages. Research on psychosocial quality of life of cancer survivors, most of whom are middle-aged or older, is urgently needed. Research on family members of survivors is even more limited. Using social contextual and social structural theories of psychological distress, this study tested hypotheses about individual- and couple-level predictors of depressive symptomatology in married middle-aged female cancer survivors and their husbands. Both intraindividual and interindividual (i.e., cross-spouse) effects were hypothesized. Couples without a history of cancer were included for comparison.

Methods: A dyadic, matched-comparison design was used with cross-sectional self-report data from a large national survey of adults age 50 and older (Health and Retirement Survey 1992). Couples where the wife reported a history of cancer were matched on age, race, and ethnicity to couples where the wife did not report a history of cancer (n = 250 each group). Depressive symptoms were assessed with an 8-item version of the CES-D. Predictors included individual (age, race, education of each spouse) and couple demographics (household income, composition, length of marriage); physical health for each spouse (number of disease conditions, pain, fatigue, health care utilization, bed days, limitations due to health); and cancer-survivor characteristics (cancer type, time since diagnosis, treatment). Nested dual-intercept multilevel models were employed to test differences in depressive symptomatology by gender and household cancer status (and their interaction) and the relative importance of the predictors.

Results: Average age was 54 for wives and 57 for husbands; 10% were minority and 4% were Hispanic. Cancer had been diagnosed 8.7 years earlier, on average (range = .5 to 20+ years). Without adjustment for covariates, wives who were cancer survivors reported significantly higher depressive symptoms than comparison wives; this difference became non-significant once differences in health were controlled. Fatigue and pain were significant risk factors for depressive symptomatology in wives, whereas higher education (of wives) and more numerous physician visits by their husband were significant protective factors for wives, regardless of cancer status. There was no significant difference in husbands' depressive symptomatology between cancer households and non-cancer households. Risk factors for husbands were minority race, higher fatigue, fewer physician visits, and greater physical limitations. Husbands whose wives reported more fatigue and bed days also reported higher depressive symptom levels. Only one effect varied by household cancer status: Husbands in cancer households with additional people present reported significantly higher depression than husbands in cancer households composed of only the couple.

Conclusions and Implications: Findings suggest that social workers in practice with aging, married female cancer survivors (a) include a physical health assessment in conjunction with any assessment of depression, (b) develop psychosocial interventions targeting fatigue, and (c) consider the interdependence of both spouses' health and depressive symptomatology. More research is sorely needed in order to serve the growing population of aging cancer survivors and their families.