Psychological outcomes in long-term survivors of childhood leukemia, Hodgkin's disease, and non-Hodgkin's lymphoma: a report from the Childhood Cancer Survivor Study.

OBJECTIVE

To evaluate and compare psychological outcomes in long-term survivors of pediatric leukemia, Hodgkin's disease, and non-Hodgkin's lymphoma and sibling controls.

METHODS

Adult survivors of childhood leukemia, Hodgkin's disease, and non-Hodgkin's lymphoma (N = 5736) and sibling controls (N = 2565) were administered a long-term follow-up questionnaire allowing assessment of symptoms associated with depression and somatic distress.

RESULTS

The majority of respondents in this study did not demonstrate symptomatology indicative of depression or somatic distress. Survivors, however, were significantly more likely than sibling controls to report symptoms of depression and somatic distress. Women were significantly more likely to indicate symptoms of depression and somatic distress than were men; however, this difference did not vary by survivor/sibling status. Similarly, socioeconomic (SES) variables predicted symptomatic levels of depression and somatic distress for both survivors and siblings, and these effects did not vary by survivor/sibling status. Among leukemia, Hodgkin's disease, and non-Hodgkin's lymphoma survivors, in addition to gender and SES, the only treatment variable that predicted scores indicating depressive symptomatology was exposure to intensive chemotherapy. Exposure to intensive chemotherapy also predicted scores indicative of somatic distress symptoms. No other medical variables, including diagnostic category, age at diagnosis, time since diagnosis, and duration of treatment, predicted symptomatic scores for depression and somatic distress.

CONCLUSIONS

This large, sibling-controlled, multisite study of young adult survivors of childhood leukemia, Hodgkin's disease, and non-Hodgkin's lymphoma found that survivors had significant increased risk for reporting symptoms of depression and somatic distress and that intensive chemotherapy added to this risk. However, being a cancer survivor did not compound the effects of gender and SES variables on the 2 outcomes measured. The ability of SES, gender, and treatment-related variables to predict psychological symptoms in this cohort of childhood survivors and sibling controls calls for future research into varied biological and psychosocial pathways by which cancer influences future psychosocial functioning.