Abstract DP097: Relationship between Coronary Artery Disease and Post Aneurysmal Subarachnoid Hemorrhage Depression Development: Preliminary EHR-Based Screening from the MIMIC-IV Database

Background: Depression is a common and functionally limiting complication of aneurysmal subarachnoid hemorrhage (aSAH). Patients with comorbid coronary artery disease (CAD) may be particularly vulnerable due to overlapping neuroendocrine and inflammatory mechanisms. As part of an ongoing biomarker study investigating HPA axis dysregulation, GR signaling, IL-33/ST2 imbalance, and miR-124-3p expression, we used a large electronic health record (EHR) database to conduct a preliminary screening analysis to identify high-risk phenotypes. Objective: To evaluate whether comorbid CAD is associated with increased risk of post-aSAH depression, using structured EHR data as a retrospective phenotyping tool to inform prospective translational biomarker research. Methods: We used MIMIC-IV v3.1, an openly available EHR dataset that includes ICU and emergency department admissions to Beth Israel Deaconess Medical Center from 2008 to 2022. A total of 1,576 admissions with aSAH were identified via ICD-9/10 codes (430, I60.x). CAD was defined using standard ICD codes (ICD-9 411.1/412/414.x; ICD-10 I25.x). Depression was defined by in-hospital SSRI administration or a diagnosis of depression (ICD-9 2962/2963/3004/311; ICD-10 F32.x/F33.x). To focus on new-onset depression, patients with prior SSRI exposure were excluded. Logistic regression adjusted for age and sex was used to assess associations between CAD and depression proxies. A secondary analysis evaluated incident SSRI use, defined as ≥48 hours after admission. Results: After exclusion of pre-SSRI users, 1,463 unique aSAH admissions were included, of which 192 (13.1%) had comorbid CAD. Depression proxies occurred in 21.4% of CAD patients versus 16.6% of those without CAD. CAD was associated with increased odds of depression proxy during the index stay (OR 1.66; 95% CI 1.11–2.48). In the subset with incident SSRI initiation (n = 62), rates were similar (4.3% vs. 3.6%), though numbers were underpowered for stratified inference. Conclusions: Preliminary EHR analysis suggests an association between CAD and increased risk of post-aSAH depression. This supports the biologic rationale for our ongoing prospective biomarker study focused on GR signaling and neuroimmune-cardiac pathways. Publicly available EHR datasets like MIMIC-IV can serve as powerful tools for hypothesis generation and early phenotype stratification in translational cerebrovascular research.