Different normalizations for body size and population attributable risk of left ventricular hypertrophy: The MAVI study

Background: Different methods of normalizing left ventricular (LV) mass for body size identify generally similar relative risks of adverse cardiovascular outcome but with variable prevalences of LV hypertrophy (H). Preliminary results from a population with high prevalence of obesity suggest that the population attributable-risk percent (PAR%) of LVH is substantially higher when LV mass is normalized for allometric power of height. Methods: We calculated the PAR% of LVH by different definitions in the cohort of the MAssa Ventricolare sinistra nell' Ipertensione (MAVI) study (n = 1019, 62% women), a population with low prevalence of obesity (22%, with only 3% and 0.1% in class II and class III obesity, respectively). Composite fatal and nonfatal cardiovascular events occurred in 53 participants (5.2%). Results: Prevalence of LVH was between 28% and 56%, with slight greater values for height-based normalization. Age- and sex-adjusted hazard ratios were comprised between 1.37 and 1.44 for different measures of LV mass index. The PAR% was not meaningfully different among the different methods of normalization (between 47% and 56%), and height-based methods showed in general a performance similar to body surface area-based normalizations. Conclusions: In a large clinical population of hypertensive subjects with low prevalence of obesity, population risk attributable to LV hypertrophy was not meaningfully different in relation to the type of normalization of LV mass for body size. Height-based methods perform as well as body surface area-based ones. We suggest that the prevalence of obesity in hypertensive populations might substantially influence differences in population risk attributable to LVH identified by different methods of normalizing LV mass.