0855 Investigate Factors That Modulate Temporal Risk for Developing New Hypertension Among Patients with Sleep Apnea

Introduction Obstructive sleep apnea (OSA), which is the most common sleep-related breathing disorder, often occurs in the context of multimorbidity. The primary purpose of our study was to investigate the temporality of comorbidity that modulate the risk for developing hypertension (HTN) among patients with OSA without pre-existing HTN at baseline OSA diagnosis. Methods Our cohort consisted of cases with OSA and the 20 most common comorbidities defined using our previously validated EHR-based algorithm. We then constructed a survival model, controlling for age, sex, and BMI. In our survival model, the index date was date of first OSA diagnosis and we estimated time-to-first HTN diagnosis. Because our survival analysis requires accurate dates of diagnosis both for OSA and HTN, we also performed a validation of the dates of diagnosis of OSA and HTN identified from our algorithm by utilizing clinical chart reviews in 400 randomly chosen EHR-defined cases. Results We identified 76,519 OSA cases, among 53,035 cases diagnosed from 2012 to 2021. 31,741 cases (59.8%) without pre-existing HTN diagnosis at the date of OSA diagnosis met our inclusion criteria. Within our survival cohort, 15,830 (49.9%) OSA cases did not develop HTN. Age, Black race, BMI, atrial fibrillation, coronary atherosclerosis, pure hyperlipidemia, hypercholesterolemia, GERD, type 2 diabetes, tobacco use, other anemia, osteoarthrosis increased risk of HTN development. Female sex, allergic rhinitis, malaise and fatigue, joint pain, vitamin D deficiency were observed not to increase risk of HTN development. Moreover, clinical chart review demonstrated that OSA diagnosis and HTN diagnosis were documented in notes a median of 38 days and 738 days (respectively) prior to being coded in the EHR. Conclusion Our study, which used a large sample of patients, demonstrates the power of leveraging the EHR to identify factors that modulate risk for developing HTN among patients with OSA. In addition, we described differences between the true diagnosis dates of OSA and HTN and diagnosis dates based on our EHR-based algorithm. Overall, this study provides new insights into disease heterogeneity and potential personalized medicine approaches. Support (if any)