Abstract 712: Left Atrial Volume and Mortality Prediction: Does the Method of Indexing Matter?
Echocardiographic left atrial (LA) enlargement (LAE) is routinely measured as LA volume (LAV) indexed to body surface area (BSA) and is significant predictor of CV outcomes. However, such indexing could underestimate the prevalence of LAE in some populations, such as obese. In contrast, LAV indexed to height (Ht) or Ht with allometric power (Ht2.0 or Ht2.7) may be better to identify LAE, given the known relation of height with lean body mass. We evaluated 47,865 patients (age: 61.6±15.4 yr; 54% female) with preserved EF to compare various indexing methods for LAV and to examine their impact on mortality during an average follow-up of 1.7±1.0 yrs. All LAV indexing methods were highly correlated (r=0.95 – 0.99, P<0.0001). Prevalence of LAE was calculated using the upper limit of mean plus two SD as a cut point for each LAV indexed to BSA, Ht, Ht2.0 and Ht2.7. Prevalence of LAE determined by LAV indexed to all Ht parameters increased significantly (P<0.0001) with increase in BMI. In contrast, prevalence of LAE by LAV/BSA decreased significantly (P<0.0001) with increasing BMI (Figure⇓). LAV indexed to Ht2.7 was the strongest independent predictor of mortality compared to Ht2.0, BSA, and Ht (HR: 1.025 (1.020–1.030) vs 1.017 (1.014–1.021) vs 1.011 (1.009–1.013) vs 1.010 (1.008–1.012), P<0.0001, respectively), although all methods predicted mortality. In conclusion, LAV indexed to BSA may over- or underestimate the prevalence of LAE compared to other methods in patients depending upon their obesity status. However, LAV indexed to Ht or Ht with allometric powers were unaffected by the level of obesity and appears to be preferable to indexing by BSA, which will need to be assessed in other populations.