(P190) DEBUNKING THE OBESITY PARADOX IN TAVR USING VOLUMETRIC BODY COMPOSITION ANALYSIS

Background: Elevated body mass index (BMI), a measure of obesity, is paradoxically associated with lower mortality in patients undergoing transcatheter aortic valve replacement (TAVR). Given the limitations of BMI, it remains unclear whether this association is mediated by adiposity as opposed to other tissue types or confounders.

METHODS AND RESULTS: We assembled a cohort of 445 patients 70 years and above who underwent TAVR at 14 hospitals in Canada, the United States, France, and Ireland. We measured body composition using a 3-compartment model (skeletal muscle, subcutaneous fat, visceral fat) from the pre-TAVR computed tomography images. Specifically, we used deep learning software (coreslicer.com) to segment the tissue types on consecutive axial slices from the L1 vertebrae to the sacrum and sum them volumetrically. Then, we used a logistic regression model to determine the association of each tissue type with 1-year mortality adjusting for age, sex, height, and Charlson comorbidity index.

The mean age was 83 years (IQR 79-86) with 38% women. In men and women, respectively, the mean BMI was 27 kg/m2 (IQR 24-30) and 26 kg/m2 (IQR 22-30), the mean skeletal muscle volume was 4920 cm3 (IQR 4329-5444) and 3368 cm3 (IQR 2971-3672), the mean subcutaneous fat volume was 4882 cm3 (IQR 3720-6206) and 5365 cm3 (IQR 3628-7681), the mean visceral fat volume was 4395 cm3 (IQR 3112-5982) and 2497 cm3 (IQR 1472-3549). The inverse association between BMI and mortality was confirmed (OR 0.92, CI 0.87 to 0.98, P=0.01), and high subcutaneous fat volume was found to be predictive of lower mortality (OR 0.80 per 1000 cm3, CI 0.68 to 0.94, P=0.006) whereas other tissue types were not.

Conclusion: The so-called obesity paradox appears to be mediated by protective effects of subcutaneous fat. Further research is needed to confirm whether this is related to nutritional reserves, anti-inflammatory properties, or reductions in natriuretic peptides.