Abstract MP58: Left ventricular hypertrophy among Black adults in a low-income country: the Haiti Cardiovascular Disease Cohort Study

Cardiovascular disease (CVD) is the leading cause of death globally and disproportionately affects low-and-middle income countries like Haiti. Left ventricular hypertrophy (LVH) is an important predictor of CVD incidence and mortality, but rarely measured in resource-constrained settings due to limited availability of echocardiograms (ECHO). Electrocardiograms (EKGs) may be a more accessible method to detect LVH. To address this gap, we determined the prevalence of LVH using echocardiography (LVH-ECHO), using EKGs (LVH-EKG), and compared diagnostic performance of EKGs versus ECHOs in detection of LVH in Haiti. We used enrollment data from the Haiti CVD Cohort study. Adults (≥18 years) living in Port-au-Prince were recruited using multistage random sampling between March 2019 and August 2021. All participants underwent standardized questionnaires to capture CVD risk factors, vital signs, physical exams, and EKGs. ECHOs were performed on participants with signs and symptoms of CVD or hypertension. Left ventricular mass was derived using 2D measurements with Devereux’s formula and indexed against body surface area (LVMI). LVH-ECHO was defined as LVMI >95 g/m 2 (female) and >115 g/m 2 (male), and defined by Sokolow-Lyon, Cornell Criteria and Limb-Lead Voltage criteria on EKGs. We compared risk factors for LVH-ECHO using univariate analysis and diagnostic performance of EKGs vs echocardiograms using sensitivity, specificity, likelihood ratio (LR) and accuracy. Among the 3,005 enrolled participants, 1,040 (34.6%) had both EKG and ECHO. Majority were female (67.0%) with mean age of 48.5 years (SD=15.3). Prevalence of LVH-ECHO was 39.0% (95%CI: 36.6%-41.5%), and associated with older age, having arrhythmia and lower education level attainment on univariate analysis. Prevalence of LVH-EKG ranged from 1.9%-5.0%. In comparing LVH-EKG to LVH-ECHO, all criteria had low sensitivity (<10%), high specificity (>90%) and moderate accuracy (62-64%). Limb-Lead voltage and Cornell criteria performed better in females (LR+ = 10.7) and males (LR+ = 8.2), respectively. Our findings show high prevalence of LVH-ECHO amongst a cohort of young Haitian adults with hypertension or CVD symptoms. All LVH-EKG criteria performed well on “ruling-in” LVH but missed cases of LVH due to poor sensitivity highlighting the need and capacity for ECHOs. Further research is needed to understand the utility of LVH diagnosis as a clinical tool to reduce CVD risk in low-resource settings.