Individual- and Area-Level Socioeconomic Disadvantage and Risk of Death Due to Chronic Health Conditions Among Autologous Blood or Marrow Transplant (autoBMT) Recipients for Lymphoma: A BMT Survivor Study (BMTSS) Report

Introduction: Late-occurring chronic health conditions (CHCs: e.g., second cancers and cardiovascular disease) are a leading cause of mortality after autoBMT for Hodgkin lymphoma (HL) and non-Hodgkin lymphoma (NHL) (Bhatia, JCO 2022). AutoBMT recipients are transitioned to care in their local community soon after transplantation, which imposes an increasing role on neighborhood-level social determinants of health (SDOH) in influencing their long-term health. Living in disadvantaged neighborhoods is associated with premature mortality in the general population (Bundy, Lancet Public Health 2023). Furthermore, individual-level socioeconomic status (SES: e.g., household income, insurance status) is a key determinant of the health and well-being of the general population. Whether neighborhood-level socioeconomic disadvantage adds to individual-level SES in increasing the risk of CHC-related late mortality after autoBMT for lymphoma remains unknown.

Methods: We leveraged resources from BMTSS, a multi-institutional cohort study examining long-term outcomes in patients transplanted between 1974 and 2014 and surviving ≥2y after BMT regardless of vital status after cohort entry. The current study reports on autoBMT survivors transplanted for HL or NHL who completed the BMTSS survey a median of 10.5y (range, 2-32) from BMT. The cohort was followed for a median of 6.1y after survey completion to date of death (among deceased) or May 2022 (for those alive) after linkage with the National Death Index. Participants self-reported individual-level SES and CHCs as diagnosed by a healthcare provider. Clinical details and therapeutic exposures were abstracted from medical records. Deaths due to CHCs were the outcome of interest. Area Deprivation Index (ADI) is a composite measure of neighborhood socioeconomic disadvantage that incorporates area-level income, education, employment, and housing quality; ADI scores for 2019 were obtained from residential addresses provided by participants. Proportional sub-distribution hazard regression models estimated the hazard of CHC-related late mortality associated with annual household income (<$75,000; ≥$75,000) and ADI quintiles (Q1=least disadvantaged; Q4/5=most disadvantaged) adjusting for clinical characteristics and therapeutic exposures.

Results: The analytic sample included 988 participants (HL=27.3%; NHL=72.7%). Median age at autoBMT was 48y (range, 5-78); majority were male (58.7%) and non-Hispanic white (78.4%); 53.2% of autoBMTs were performed between 2005 and 2014. Overall, 41.5% of participants reported annual household income ≥$75,000 and 67.7% reported having private insurance. The proportion of autoBMT survivors living in the least (ADI Q1) and most (ADI Q4/5) disadvantaged neighborhoods was 25.6% and 24.9%, respectively. Model 1 (income-mortality association): after adjusting for pertinent variables (age at BMT, sex, race/ethnicity, year of BMT, primary diagnosis, presence of pre-BMT and post-BMT grade 3-4 CHC, total body irradiation conditioning, and insurance status), low annual household income was associated with greater hazard of CHC-related late mortality (<$75,000: HR=1.76, 95% confidence interval [95%CI]=1.10-2.82, P=0.02; ref=≥$75,000). Model 2 (ADI-mortality association): adjusting for the same variables (but not including income), living in a disadvantaged neighborhood was associated with greater hazard of CHC-related late mortality (ADI Q4/5: HR=1.95, 95%CI=1.08-3.50, P=0.03; ref=Q1). Model 3 (ADI+income-mortality association): Survivors with annual household income <$75,000 and living in the most disadvantaged neighborhoods had a 2.64-fold greater hazard of CHC-related late mortality (95%CI=1.24-5.65, P=0.01) than those with household income ≥$75,000 and living in the least disadvantaged neighborhoods.

Conclusions: The joint effect of individual and neighborhood socioeconomic disadvantage more than doubles the risk of late mortality due to chronic health conditions among HL and NHL patients treated with autoBMT. This vulnerable population could benefit from targeted interventions aimed at alleviating the burden of adverse SDOH, such as connecting patients to community resources to address unmet need and informing changes in current policies.