Factor Analysis of the Beck Depression Inventory-II and Long-Term Hematopoietic Stem Cell Transplantation Survival Using the Research Domain Criteria Framework

The presence of depressive symptoms prior to hematopoietic stem cell transplantation (HSCT) is a common experience, with long-term impacts on survival. Using the National Institute of Mental Health Research Domain Criteria (RDoC) framework, this study sought to characterize depressive symptoms in patients prior to HSCT through exploratory factor analysis and to determine whether depressive factors were significant predictors of long-term survival. Individuals were included in the study if they were preparing to undergo HSCT and endorsed depressive symptoms measured by the Beck Depression Inventory-II (BDI-II). Patients were scheduled to undergo transplantation between 2005 and 2010. Survival analyses were conducted in 2022 to assess long-term outcomes. The primary outcomes were exploring the factor structure of the BDI-II and conducing univariate and multivariate Cox regression analyses of depression symptoms and known demographic and disease characteristics that impact survival. Of the 695 participants included in the study, most were male, middle aged (mean age, 55.08 ± 11.75 years), white, and married. Exploratory factor analysis revealed a 4-factor structure consisting of “negative valence systems: internalizing” (eg, worthlessness, guilt), “arousal and threat” (eg, agitation, irritability), “arousal and regulatory systems” (eg, loss of energy, fatigue), and “negative valence systems: externalizing” (eg, loss of pleasure, loss of interest). Univariate survival analyses identified age, sex, disease type, acute graft-versus-host disease (GVHD), and negative valence systems: externalizing as significant predictors of survival. Transplant type, chronic GVHD, performance status, and the other 3 depression factor structures were not significant in univariate models. In the multivariate model, older age (hazard ratio [HR], 1.031; 95% confidence interval [CI], 1.021 to 1.041; P < .001) and presence of negative valence systems: externalizing symptoms (HR, 1.132; 95% CI, 1.030 to 1.244; P = .010) were significant predictors of shorter survival. Additionally, individuals diagnosed with acute leukemia were significantly more likely to have shorter survival compared to those with other disease types, including amyloidosis (HR, .362; 95% CI, .229 to .575; P < .001) and non-Hodgkin lymphoma (HR, .526; 95% CI, .349 to .793; P = .002). Exploratory factor analysis of depressive symptoms mapped well onto the RDoC constructs. Loss of pleasure and loss of interest, two key components of depression, were predictive of shorter survival. Exploration of key components of depression rather than the total depression score may provide important prognostic information for long-term survivorship and may help inform future and more individualized care.