Introduction: Increased left ventricular mass (LVM) predicts excess cardiovascular morbidity and mortality. LVM is often indexed to height (HT) or body surface area (BSA). HT-indexation ignores weight and thus obesity. The relationship of indexed LVM with weight, obesity, and adiposity is unclear. We investigated the association of LVM with total-body mass (TBM, i.e. weight) and fat-free mass (FFM), hypothesizing that LVM scales better to FFM than TBM.
Methods: From the 1794 members of Framingham Offspring cohort who underwent cardiac magnetic resonance (CMR), we identified 845 (aged 65±9y) without any history of hypertension, myocardial infarction, heart failure, or CMR wall-motion abnormality. LVM was measured from cine bSSFP images and indexed to HT, BSA, TBM and FFM (by DEXA). Participants were stratified by sex and body mass index (BMI) category (NL=BMI 18.5 - 24.9; OW=25.0 - 29.9; OB=30.0+ kg/m2). We used sex-specific ANCOVA to test for linear trend across BMI categories, and determined sex-specific Pearson correlation coefficients of LVM with BMI; both sets of analyses were adjusted for age and systolic blood pressure. Data are summarized as mean ± SD.
Results: The Table shows that LVM and LVM/HT increase with greater BMI-category in both sexes. LVM/BSA was similar across BMI-categories, while LVM/TBM decreased. There was no linear trend across BMI-categories for LVM/FFM in either sex. On a continuous basis, LVM and LVM/HT were positively correlated with BMI in both sexes (r=0.20 to 0.35, p<0.001 for all); LVM/BSA was weakly inversely correlated with BMI in women only (r=-0.14, p=0.001) but not men (r=-0.08, p=0.18); LVM/TBM was inversely correlated with BMI (M: r=-0.39, W: r=-0.54; p<0.0001 both); LVM/FFM was not correlated with BMI in either sex (M: r=-0.05, p=0.38, W: r=0.08, p=0.055).
Conclusion: LVM as a proportion of TBM (weight) actually decreases with greater BMI, while LVM/FFM is not associated with BMI. Whether LVM is better scaled to FFM vs HT, BSA or TBM for risk stratification purposes remains to be determined.