Abstract 15901: Body Mass Index Threshold for Obesity Should Be Personalized Based on Polygenic Score
Min Seo Kim, Injeong Shim, Beomsu Kim, Pradeep Natarajan, Ron Do, Woong-Yang Park, Amit V Khera, Patrick T Ellinor, AKL FAHED, Hong-Hee Won- Physiology (medical)
- Cardiology and Cardiovascular Medicine
Background: Current guidelines use body mass index (BMI) ≥30 kg/m 2 to define obesity and indicate obesity management, including lifestyle modification and anti-obesity pharmacotherapy.
Hypothesis: The conventional BMI threshold (≥30 kg/m 2 ) may have different clinical significance by the individual’s inherent genetic risk for obesity.
Methods: Polygenic score (PGS) for BMI was calculated to quantify inherited susceptibility to obesity for 335,835 European-ancestry UK Biobank participants. We calculated odds ratio (OR) of a composite outcome (type 2 diabetes and cardiovascular outcomes [coronary artery disease, heart failure, atrial fibrillation, pulmonary embolism, venous thromboembolism, hypertension, and aortic valve stenosis]) for people living with obesity (PLwO, BMI ≥30 kg/m 2 ), compared to a reference group (BMI 18.5-24.9 kg/m 2 ). Using a regression model, we assessed the BMI thresholds demonstrating the equivalent OR of composite outcome in each genetic risk stratum, adjusting for clinical, lifestyle, and socio-behavioral covariates. We quantified the extent of reclassification among PLwO by using genetically-inferred BMI.
Results: For 335,835 participants (mean age 57.37±7.99 years, 53.6% female), the genetically-inferred BMI thresholds were 29.44 kg/m 2 in the high PGS group (top 20%) and 30.62 kg/m 2 in the low PGS group (bottom 20%) for risk of composite outcome equivalent to the conventional BMI threshold. The genetically-inferred BMI thresholds diverged wider for more extreme genetic groups (i.e., top 5% PGS). With the genetically-inferred BMI threshold, 6.9% of PLwO based on conventional BMI threshold were recalibrated (either excluded or newly ascertained to PLwO).
Conclusions: The current one-size-fits-all BMI threshold could benefit from incorporating genetic susceptibility to better define obesity health risks. Accounting for genetic risk to define the obesity BMI threshold could reclassify up to 6.9% of the PLwO.