A Novel Mathematical Model for Correcting the Physiologic Variance of Two-Dimensional Echocardiographic Measurements in Healthy Chinese Adults

Background To facilitate differentiation between normal and abnormal values, it is necessary to correct echocardiographic measurements for physiologic variance induced by age, gender, and body size variables. Methods A total of 34 two-dimensional echocardiographic parameters were measured in 1,224 healthy Chinese adults with body mass index 2 . An optimized multivariate allometric model and scaling equations were first developed in 858 subjects (group A), and their reliability was then verified in the remaining 366 subjects (group B). The traditional single-variable isometric model in which parameters are linearly corrected by a single body size variable (height, weight, body mass index, or body surface area) was used for comparison. The success of correction was defined as the absence of significant correlations ( r  > 0.20, P r  > 0.80) between the corrected and uncorrected values. Results Before correction, all 34 parameters correlated significantly with one or more of the physiologic variables of age and body size and differed significantly between men and women on 29 parameters (85.3%) in both groups. The success rate of correction with the single-variable isometric model was only 11.0% (15 of 136 corrections due to four variable corrections used for each parameter), while use of the optimized multivariate allometric model successfully corrected all 34 parameters (100%) for physiologic variance induced by age and body size variables and eliminated the gender differences in 32 parameters (94.1%). A new set of reference values for corrected echocardiographic measurements independent of age, gender, and body size variables were established. Conclusions The novel optimized multivariate allometric model developed in this study is superior to traditional the single-variable isometric model in the correction of echocardiographic parameters for physiologic effects of age, gender, and body size variables and thus should be encouraged in both scientific research and clinical practice.