Abstract 15690: Sex- and Age-Specific Indexation of Left Ventricular Mass

Abstract

Background: Left ventricular mass (LVM) is established as a powerful predictor of independent risk for cardiovascular disease and mortality. Despite a large body of literature, no gender-and age-specific data are available in guidelines for echocardiography yet. We investigated an optimized indexation and standardization in a large non-institutionalized sample.

Material: A population-based sample of 5000 Mid-Europeans from the Gutenberg Health Study was analyzed sex-specifically. The relationship between echocardiographically assessed LVM, body size, age, and cardiovascular risk factors (CVRF) was determined in nonlinear, allometric models. Using a representative subsample of presumably CV healthy individuals, reference limits and categories indicating the grade of deviation from the reference were calculated and nomograms were created by quantile regression.

Results: LV Mass was significantly associated with age and height in both men (age: beta 0.78, height: beta 76.2, both p<0.0001) and women (age: beta 0.83, p<0.0001; height: beta 8.1, p<0.0066). Multivariate linear regression confirmed an independent association of LVM with age, diabetes, hypertension and obesity in both sexes (men: R2=0.11, women R2=0.24). After indexing LVM for height, the association with age remained significantly with 0.415 g/m body height per year in men and 0.600 g/m body height per year in women (both p<0.0001). Therefore, gender-specific nomograms were generated by indexation for height showing the association with age. Nonlinear, allometric regression modeling (LVM=β0 (height)β1+E) revealed β0=120, β1=0.74 for men and β0=111, β1=0.36 for women. There was no residual correlation of gender-specifically indexed LVM to height and the association to cardiovascular risk factors was not altered. However, this holds not true, when indexing for height2,7 or BSA.

Conclusion: LVM shows strong associations with age and height. For the first time, gender-specific reference values for LVM which account for age and the allometric relation with height are presented. This indexation was proven to not alter the association to the cardiovascular risk profile.