Abstract 094: Sex-specific Effects Of Sleep Apnea And Medication-treated Depression On Left Ventricular Mass Index And 3D Left Ventricular Global Longitudinal Strain In The CARDIA Study

Abstract

Background: Cardiovascular disease (CVD) is the leading cause of death and disability worldwide. CVD risk increases with age; however, sex differences are known to significantly affect progression and survival. We designed this study to assess CVD-related risks associated with depression, sleep apnea, and depression comorbid with sleep apnea, in a biracial cohort of middle-aged men and women. We assessed outcomes such as left ventricular mass index (LVMI)-an index of left ventricular hypertrophy, and 3D left ventricular global longitudinal strain (3D LVGLS), an index of global myocardial function. Methods: Data were obtained from the Coronary Artery Risk Development in Young Adults (CARDIA) Study from black and white men and women (n=2778) who attended visit 9 (year 30 of follow-up) echocardiography scan. Demographics and diagnoses of depression and sleep apnea were obtained from visit 8 (year 25 of follow-up). Binary logistic regression was used to asses factors associated with high blood pressure (BP > 130/80 mmHg) and heart failure with reduced ejection fraction (HFrEF). Linear regression analyses were used to assess the strength of the following predictors: race, age, depression diagnosis and medication use for depression on LVMI and 3D LVGLS. Results: In men, depression alone and comorbid with sleep apnea were associated with higher risk of BP >130/80mmHg (OR = 1.47, p = 0.036 and OR = 2.739, p = 0.006, respectively). In both men and women (median age 55), black race was significantly associated with LVMI (p = 0.008 and p <0.001, respectively). Additional predictors of LVMI found in women were age (p = 0.019) and current medication use for depression (p = 0.006). Black race and current use of medications for depression were significantly associated with 3D LVGLS in both men (p <0.001 and p = 0.029, respectively) and women (p < 0.001 and p = 0.007, respectively). Conclusion: Our results suggest that in middle-aged women, but not men, medication-treated depression is a significant predictor of LVMI. The sex differences were not observed for 3D LVGLS. These results suggest that the severity of depression (i.e. necessitating medication use) may be an important predictor of CVD, especially in women. However, a direct effect of these medications on CVD may occur.