Relations of left ventricular mass to demographic and hemodynamic variables in American Indians: the Strong Heart Study

Abstract

Background: Previous studies have identified associations of left ventricular (LV) mass with demographic (body habitus and sex) and hemodynamic variables (blood pressure, stroke volume [SV], and myocardial contractility), but the relative strength and independence of these associations remain unknown.

Methods and results: We examined the relations of echocardiographically determined LV mass to demographic variables, blood pressure, Doppler SV, and measures of contractility (end-systolic stress [ESS]/end-systolic volume index and midwall fractional shortening [MFS] as a percentage of predicted for circumferential end-systolic stress [stress-independent shortening]) in 1935 American Indian participants in the Strong Heart Study phase 2 examination without mitral regurgitation or segmental wall motion abnormalities. Weak positive relations of LV mass with systolic and diastolic pressures (r=.22 and r=.20) were exceeded by positive relations with height (r=.30), weight (r=.47), body mass index (r=.31), body surface area (r=.49), and Doppler SV (r=.50) and negative relations with ESS/volume index ratios (r= -.33 and -.29) and stress-independent MFS (r= -.26, all P<.0001). In multivariate analyses that included blood pressure, SV, and a different contractility measure in each model, systolic pressure, stroke volume, and the contractility measure were independent correlates of LV mass (multiple R=.60 to .66, all P<.0001). When demographic variables were added, LV mass was more strongly predicted by higher SV and lower afterload-independent MFS than by greater systolic pressure, height, and body mass index (each P<.00001, multiple R=.71).

Conclusions: Additional characterization of volume load and contractile efficiency improves hemodynamic prediction of LV mass (R(2)=.30 to .44) over the use of systolic blood pressure alone (R(2)=.05), with a further increase in R(2) to .51 when demographic variables are also considered. However, nearly half of the ventricular mass variability remains unexplained.