2514 - Analyzing Major Depressive Disorder Risk Factors and Survival in Elderly GI Cancer Patients Using the SEER-MHOS Database
Presenter(s)
T. C. Wu1, T. Grogan2, and A. Raldow1; 1Department of Radiation Oncology, University of California, Los Angeles, Los Angeles, CA, 2Department of Medicine Statistics Core, University of California, Los Angeles, Los Angeles, CA
Purpose/Objective(s): To evaluate the risk of major depressive disorder (MDD) among elderly patients diagnosed with gastrointestinal (GI) cancers and examine its association with patient- and treatment-related factors, as well as survival.
Materials/Methods: Using the SEER-MHOS national database, we identified individuals aged =65 years diagnosed with colorectal, esophageal, pancreatic, or hepatobiliary cancer between 1973 and 2017. MDD risk was assessed using patient responses to screening questions from the SF-36 and VR-12 instruments. Patient characteristics and study variables were summarized by MDD risk status (no vs. yes) using means (SD) or frequency (%) and compared using t-tests or chi-square tests. A multivariable logistic regression model evaluated the association between patient- and treatment-related factors and MDD risk, while a time-dependent Cox proportional hazards model assessed the relationship between MDD risk and survival. The time-dependent approach accounted for variation in MDD risk assessment timing.
Results: A total of 48,859 individuals with colorectal (88%), pancreatic (6%), esophageal (4%), and hepatobiliary (2%) cancer were identified, of whom 32% were at risk for MDD. On multivariable analysis, higher MDD risk was significantly associated with being unmarried, having a low household income, identifying as Asian, Black, or Hispanic, receiving chemotherapy, residing in areas with 10–100% poverty, being on Medicaid or uninsured, and having =3 comorbid conditions (all p<0.01). Patients with hepatobiliary cancer had a significantly greater MDD risk compared to those with colorectal cancer (p<0.0001). Survival modeling (Cox proportional hazards) revealed that patients at risk for MDD had significantly lower survival (HR 1.49, 95% CI: 1.44–1.54, p<0.0001). Factors associated with improved survival included being married (widowed vs. married: HR 1.26 [1.23–1.30], p<0.001; never married vs. married: HR 1.08 [1.02–1.14], p=0.007), having at least a college degree (HRs ranging from 0.95 [0.90–1.01], p=0.078 for >4-year college degree to 1.25 [1.18–1.31], p<0.001 for =8th grade education), being female (HR 0.72 [0.70–0.74], p<0.001), having fewer comorbidities (HR 1.33 [1.30–1.36] for =3 comorbidities vs. 0–1, p<0.001), and being a non-smoker (HR 1.22 [1.18–1.27] for smokers vs. non-smokers, p<0.001).
Conclusion: Nearly one-third of elderly patients with GI cancer are at risk for MDD, which is associated with worse survival outcomes. Identifying and addressing modifiable risk factors for MDD may improve patient outcomes.