Validation of evaluating left ventricular diastolic function with estimated left atrial volume from anteroposterior diameter

Background Left atrial (LA) volume (LAV) is one of the recommended key variables for evaluating left ventricular (LV) diastolic function. However, only LA anteroposterior diameter (LAAP) is available in numerous large-scale existing databases. Therefore, this study aimed to validate whether LV diastolic function could be evaluated with estimated LAV from LAAP. Methods A total of 552 inpatients with sinus rhythm were consecutively enrolled. LAV was measured by biplane Simpson's disk summation method. LV diastolic function was evaluated according to the 2016 proposed recommendations. Best-fitting regression models of LAAP index (LAAPI)-LAV index (LAVI) were developed and equations with the highest F-value were chosen in the first 276 subjects (derivation set), and concordance for evaluating LV diastolic function between using estimated and observed LAVI was verified in the remaining 276 subjects (validation set). Results In the derivation set, the linear model has the highest F-value in all subjects and in the subjects with normal or depressed LV ejection fraction. In the validation set, using the linear equation (LAVI = 2.05 x LAAPI - 13.86), the higher area under curve and narrower range of difference were shown between estimated LAVI and observed LAVI, respectively. Further, concordance for diagnosis (overall proportion of agreement, 88.4%; kappa = 0.79) and grading (overall proportion of agreement, 84.8%; kappa = 0.74) of LV diastolic dysfunction was substantial between using estimated and observed LAVI. Conclusions LV diastolic function can be evaluated with estimated LAVI from LAAPI, which might provide a surrogate method when the direct measurement of LAV is not available.