Normalization for body size and population-attributable risk of left ventricular hypertrophy: The Strong Heart Study

Background: Left ventricular hypertrophy (LVH) is identified by left ventricular mass (LVM) normalized by body surface area (BSA) or height (in meters) also raised to allometric powers. The presence of LVH detected by these indices predicts increased cardiovascular (CV) events. Whether different indexations of LVH differ in their ability to predict excess risk is unknown. Methods: A total of 2400 subjects, (1589 women and 811 men), 59 ± 8 years of age and without prevalent CV disease, valve disease or wall motion abnormalities and high prevalence of obesity were followed for an average of 86 months. Reference values (mean ± 1.96 SD) for LVM/BSA, LVM/BSA ^1.5, LVM/m, LVM/m^2.7, and LVM/m^2.13 were obtained in 251 normal participants and population-attributable risk percent (PAR%) for fatal and nonfatal CV events were calculated from prevalence of LVH and hazard ratios (HR). Results: In the entire population or in hypertensive participants, prevalence of LVH was higher for LVM/m^2.7 (20% and 28%) and LVM/m^2.13 (18% and 25%) than for BSA (7% and 11%). Age and sex-adjusted PAR% for LVM/m^2.7 or LVM/m^2.13 were on average 1.8-fold greater than for LVM/BSA in the entire population, and 1.6-fold greater in hypertensive participants, differences that were statistically significant. Conclusions: The presence of LVH identified by LVM normalized for height to allometric powers is associated with a higher proportion of incident CV events than is LVH detected by normalization for BSA and is convenient for identification of individuals at high risk and in need of preventive intervention in populations with high prevalence of obesity. Allometric power methods allow detection of prognostically adverse, obesity-related LVH, which is unidentified using BSA.