Cardiology · Medical Literature
ECG Criteria for Left Ventricular Hypertrophy in Hypertensive Black Africans: Insights from the coArtHA Trial.
Abstract
Electrocardiographic (ECG) criteria for detecting left ventricular hypertrophy (LVH) have largely been derived from non-African populations. Their diagnostic performance in Black African adults with untreated hypertension remains uncertain. To evaluate the diagnostic performance of established ECG-LVH criteria compared with echocardiographically assessed left ventricular mass index (LVMI) in Black African adults with untreated, uncomplicated hypertension enrolled in the coArtHA trial. In this subanalysis, we included 1,125 participants from rural Tanzania and Lesotho who underwent baseline 12-lead ECG and focused transthoracic echocardiography (fTTE). LVH was defined as LVMI > 95 g/m² in women, and >115 g/m² in men. Diagnostic performance of multiple ECG-LVH criteria was assessed using correlation analyses, sensitivity and specificity estimates, receiver-operating characteristic (ROC) analysis, including calculation of the area under the ROC curve (AUROC), and precision-recall curves (PRC). Echocardiographic LVH was present in 56 (5%) participants. Across continuous ECG indices, Cornell voltage product adjusted by +0.8 mV in women showed the strongest association with LVMI (Spearman's rho = 0.37, p < 0.001). Among evaluated criteria, Cornell voltage product-based measures demonstrated the most consistent diagnostic performance across evaluated metrics, with an AUROC of approximately 0.76. Using established cut-off values, this criterion yielded a sensitivity of 64% and a specificity of 80%, whereas voltage-only criteria, including Multi-Ethnic Study of Atherosclerosis (MESA)- and Sokolow-Lyon-based indices, showed limited accuracy in this setting. Exploratory analyses of alternative cut-off values demonstrated the expected sensitivity-specificity trade-offs. In this cohort of hypertensive Black African adults, ECG criteria showed a moderate association with LVMI by echocardiography. Among established ECG criteria, Cornell voltage product-based measures demonstrated the most consistent diagnostic performance. Diagnostic characteristics were strongly influenced by cut-off selection and clinical context, underscoring the need for population-specific evaluation and external validation of ECG criteria for LVH in African settings. Clinicaltrials.gov NCT04129840. Registered on October 17, 2019 (https://www.clinicaltrials.gov/).
Clinical relevance
ECG-LVH criteria have been predominantly derived and validated in non-African populations, leaving a critical evidence gap for the growing burden of hypertensive heart disease in sub-Saharan Africa. This subanalysis of the coArtHA trial directly addresses diagnostic uncertainty in Black African adults with untreated, uncomplicated hypertension—a population at high cardiovascular risk yet underrepresented in landmark ECG-LVH validation studies. As resource-limited African settings often rely on ECG rather than echocardiography for LVH screening, accurate characterization of ECG diagnostic performance in this population has immediate clinical and public health relevance.
Key findings
1. Echocardiographic LVH (LVMI >95 g/m² in women, >115 g/m² in men) was present in only 56 of 1,125 participants (5%), indicating a relatively low prevalence of LVH in this cohort of rural hypertensive Black African adults from Tanzania and Lesotho. 2. Among all continuous ECG indices evaluated, Cornell voltage product adjusted by +0.8 mV in women demonstrated the strongest correlation with LVMI (Spearman's rho = 0.37, p < 0.001). 3. Cornell voltage product-based criteria achieved the highest area under the ROC curve (AUROC ≈ 0.76), representing the most consistent diagnostic performance across all evaluated metrics including ROC and precision-recall curve analyses. 4. Using established cut-off values, Cornell voltage product yielded a sensitivity of 64% and a specificity of 80% for echocardiographically confirmed LVH, reflecting a moderate but imperfect diagnostic trade-off. 5. Voltage-only criteria—including MESA-based and Sokolow-Lyon-based indices—demonstrated limited diagnostic accuracy in this population, performing inferiorly to Cornell voltage product across evaluated metrics. 6. Exploratory analyses of alternative cut-off values confirmed the expected inverse sensitivity-specificity trade-off, highlighting that diagnostic performance is strongly contingent on cut-off selection and clinical context.
Practice implications
Clinicians practicing in sub-Saharan African settings or managing Black African hypertensive patients should preferentially apply Cornell voltage product criteria (with the sex-specific +0.8 mV adjustment in women) over voltage-only indices such as Sokolow-Lyon when using ECG to screen for LVH, given its superior and most consistent diagnostic performance in this population. However, the moderate AUROC of ~0.76 and sensitivity of only 64% underscore that a negative ECG does not reliably exclude LVH, and echocardiographic confirmation should be pursued whenever feasible—particularly in high-risk patients or when therapeutic decisions hinge on LVH status. Population-specific validation and recalibration of ECG-LVH cut-offs in African cohorts are urgently needed before these criteria can be confidently deployed as standalone diagnostic tools in resource-limited settings.