Response to Method Errors or Unexplained Biological Information

To the Editor: We thank Drs de Simone and Devereux1 for their comments and their continued contributions to the subject of allometric normalization of left ventricular mass (LVM). The authors confirm the important residual relationship between height and LVM/height2.7 in an independent cohort and present population-attributable risk (PAR) estimates of LV hypertrophy (LVH) defined by LVM/height2.7 and by LVM/height.1.7 These data are best interpreted in light of the definition and usefulness of PAR estimates and the overcorrection induced by normalizing LVM to height,2.7 resulting in the residual relationship between height and LVM/height.2.7 PAR is the portion of the incidence of a disease in the population that is due to the exposure, representing the proportional reduction in average disease risk that would be achieved by eliminating the exposure from the population while distributions of other risk factors remain unchanged (assuming the exposure is causal).2 In a multi-factorial disease setting like cardiovascular disease, the focus of interest is the risk associated with the modifiable exposure while other risk factors (particularly nonmodifiable factors such as height) are kept unchanged.2,3 Given the relationship between height and LVM/height,2.7 PAR estimates for LVH defined by LVM/height2.7 will not only reflect LVH-associated risk but also height-related risk, whatever its mechanism may …