Association of left ventricular mass with standard, unattended automated, and ambulatory blood pressure during household-based screening in lesotho

Objective: Left ventricular (LV) geometry and mass index (LVMi) including concentric remodeling (CR) and LV hypertrophy are important determinants of outcome in hypertension. Data concerning the association between standard blood pressure measurement (SBPM) and LVMi in sub-Saharan Africa populations are scarce, and data analysing this association with different BPM methods are even less. Aim of the study was to evaluate the association of LVMi with SBPM, unattended automated BPM (uABPM) and 24hr-ambulatory BPM (24h-ABPM) during a household-based screening program in Lesotho. Design and method: 6108 individuals with measurement of SBP were included into the screening program. For this nested study 639 received focused echocardiography for analysing LV geometry over different SBPM hypertension grades, and 209 additional uABP and 24h-ABPM to compare BPM methods. Pearson's correlation coefficient was used to determine linear correlations of LVMi with SBPM, uABP and 24h-ABPM. Logistic regression models with CR as dependent variable, and the three BPM approaches as primary independent variables were developed. C statistics were calculated for each logistic model as area under the receiver operating characteristic curve, and compared for the capacity of each BP measurement approach to discriminate the odds of CR. Results: Among the 639 participants (mean age 55+/-18 years), 59.6% were females and 53.4% had elevated SBPM values. Mean LVMi was 59.4+/-15.4 g/m2. For SBPM, mean LVMi increased across BP grades, starting at high normal (p<0.0001 compared to optimal/normal) (figure1a). CR, present among 56% of all participants, was evident among 34% participants with optimal BP and increased across BP grades (figure1b). Among participants with all three BPM, correlation between systolic BP and LVMi was highest for 24h-ABPM, followed by uABPM (figure1c). AUROC for prediction of CR was 0.66, 0.70 and 0.70 for SBPM, uABP, and 24h-ABPM respectively. Conclusions: LVMi increased at high normal BP compared to normal and optimal BP. CR was unexpectedly common with 1/3 participants with optimal and half of participants with normal BP values. 24h-ABPM showed higher correlation with LVMi than uABP and SBPM, with a numerically higher AUROC to predict CR than SBPM.