These cross-sectional analyses confirm previous reports that younger women are more likely to report depressive symptoms following breast cancer diagnosis than older women. We found a monotonic gradient in the crude odds of depression by age, with those 75 and older at time of diagnosis having 83 % lower odds of depression in unadjusted analyses compared to women under age 45 at time of diagnosis.
Many of the variables significantly associated with depression in crude bivariate analyses are consistent with other studies [5
]. However, only a subset of these remained significant following simultaneous adjustment in the full model: chemotherapy regimen, bodily pain, illness intrusiveness, sense of peace, passive coping, and perceived attractiveness. Contrary to Compas et al. [9
], we found that younger women used more active coping. However, they also used more passive coping (although this was not statistically significant), suggesting that they may use more coping strategies in general. Although other studies have also found that psychosocial factors contribute more to depression than treatment factors [3
] we would not conclude, as did Bardwell et al. [5
], that cancer-related variables are unimportant when considering risk of depression. It should be noted that Bardwell et al. combined surgery and chemotherapy into one category and did not consider specific chemotherapy regimens. Our results suggest that it is important to consider aggressiveness of chemotherapy. As found by Wong-Kim and Bloom [24
], a biopsychosocial approach provides the most comprehensive explanation of depression. However, our primary goal was not to confirm previous findings, but to better understand why younger women experience more depression after a breast cancer diagnosis than do older women. Adjustment for sociodemographic and cancer-related variables ameliorated the strong, crude age-related gradient in depression risk somewhat, but age remained a significant predictor of risk of depression. It was not until symptoms and certain psychosocial variables were included in the model that the inverse age-related gradient disappeared.
Of the psychosocial variables, and indeed of all the variables we examined, illness intrusiveness appears to be a key measure which helps to explain the age-related differences—younger women consistently report significantly higher levels of illness intrusiveness than do older women across all 16 domains covered by the illness intrusiveness measure. Although health, diet, active recreation, and social activities were among the top five areas that were most disrupted by cancer across all participants, there were some differences by age group. For example, sex life was the area most affected for those under age 45, while work around the house was in the top five for those 55 and older.
It is not possible to completely eliminate the intrusive-ness of a breast cancer diagnosis on a younger woman. However, a focus on symptom management and on self-reports of illness intrusiveness in distinct areas of life may suggest possible interventions to lessen pain and intrusiveness, and consequently may lessen risk for depression. Younger women reported more severe pain and this may be an important factor that increases illness intrusiveness. For those under age 45, sex life was rated as the area most affected by cancer (4.5 on a 7-point scale) and other studies have shown that younger women often report significant sexual problems as a result of chemotherapy [41
]. Sexual functioning is an area where remedies to decrease vaginal dryness or interventions to increase feelings of sexual attractiveness may be beneficial.
This study has several limitations. The analyses are observational and cross-sectional only, so we cannot eliminate the possibility that women who are depressed are more likely to perceive their breast cancer as more intrusive. The Beck Depression Inventory is a self-report measure of depressive symptomatology and is not a measure of clinical depression. Although characteristic of many samples of breast cancer patients, this sample is relatively homogeneous (mostly white, educated) which limits the generalizability of our findings.
Overall, our results provide empirical support that age differences in psychological morbidity immediately following a breast cancer diagnosis may be accounted for by the greater impact cancer and its treatment have in the lives of younger women.