Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Social Work

Committee Chair

Faul, Anna

Author's Keywords

Distress; Older adults; Oncology; Multilevel modeling; Cancer; Longitudinal


Cancer--Patients--Psychology; Aged; Stress (Psychology)--Aged


Data from the Health and Retirement Study (HRS) were used to examine factors that influence emotional distress in older adults with cancer, arid to learn whether these distress trajectories are different from older adults without cancer. The model investigated the influence of life course factors, internal, external and health-related resources on distress over eight years among a nationally representative sample of older adults ages 50-91. Methods: This longitudinal panel study tested a multilevel random intercept multinomial ordered categorical model with a two-nested-Ievel structure using empirical Bayes Markov chain Monte Carlo (MCMC) estimation. The model examined individual differences in distress over eight years (2000-2008), testing multiple potential predictors which were either time-variant within-person or time-invariant between-person data measured at five consecutive time periods. All those with a first diagnosis of cancer in 1998-2000 were included in study (n=200) together with a representative subsample of all non-cancer cases (n=1190), for a total of 1390. Results: In each step, the model fit improved significantly using the DIG statistic. Overall, 33 percent of sample reported one or more depression symptoms in 2000, and by 2008 the number decreased to 30 percent. Keeping all main effects constant, those with cancer had one percent lower probability of depression symptoms over time than those without cancer. Significant two-way interaction effects were detected between having cancer and absence of spouse/partner in the home, and cancer and lower life expectancy; each resulted in higher probabilities of depression symptoms. A significant three-way interaction effect was detected between cancer, gender, and social support with women with a cancer history and low social support having the highest probability of depression symptoms. Conclusions: Those with four or five depression symptoms at the beginning of the study tended to stay the same over time, while those with one to three symptoms tended to improve slightly. A small significant protective factor of cancer on distress over time was detected. A large protective factor of cancer existed for older adults with a partner present in the home, but cancer became a significant risk factor for distress with no partner presence. The protective factor remained for older adults with cancer and longer life expectancy, but cancer became a significant risk factor for distress with low life expectancy. While there was a large protective factor of cancer among women with high social support, cancer history became a risk factor for women with low social support.