AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Distribution and Phenotypic Concordance of the Proposed Lancet Commission Definition of Obesity: A National Health and Nutrition Examination Survey Analysis
By patel D, Corsi DR, Visaria A
Background: BMI is an imperfect surrogate for adiposity and cardiometabolic risk. The Lancet Diabetes & Endocrinology Commission proposed an expanded obesity definition that adds waist circumference (WC), waist-to-height ratio (WHtR), and waist-to-hip ratio (WHR) to BMI, and further distinguished preclinical obesity from clinical obesity based on the presence of organ dysfunction or functional limitation. The population distribution of individual anthropometric components, and how they differ across preclinical and clinical obesity phenotypes, remains unknown. Aim: To characterize the distribution of Lancet components and assess how each combination differs in identifying preclinical versus clinical obesity phenotypes in U.S. adults, using DXA-derived adiposity as a reference. Methods: We analyzed adults aged 20-59 years in NHANES 2017-2018 undergoing whole-body DXA. The reference standard for excess adiposity was DXA (Dual-Energy X-Ray Absorptiometry) total body fat ≥25% (men)/≥35% (women). Secondary outcomes were DXA-estimated visceral obesity (≥130 cm ² men; ≥100 cm ² women) and clinical obesity (Lancet obesity plus a metabolic or end-organ manifestation); participants meeting Lancet anthropometric criteria without these manifestations were classified as preclinical obesity. Participants were cross-classified into mutually exclusive combinations of the Lancet components; for each subset we estimated the survey-weighted proportion with each true outcome (the positive predictive value [PPV]) and the subset’s share of total disease burden. Results: Among 1,900 adults (≈98.7 million weighted), 69% (95% CI 67-71) had true DXA-verified excess adiposity, versus 36% (35-38) classified as obese by BMI alone. PPV for DXA-defined adiposity varied widely across combinations (Table 1). BMI≥40 kg/m ² and BMI plus all anthropometrics carried the highest PPVs, with 100% and 96% for excess adiposity and 88% and 62% for visceral obesity, respectively. BMI+WC+WHtR (no hip metric) performed comparably (95% excess adiposity; 66% clinical obesity). In contrast, isolated WHR positivity captured only 16% with excess adiposity and 7.4% with visceral obesity. BMI plus all anthropometrics also accounted for the largest absolute share of visceral (53%) and clinical (41%) obesity (Table 2). Conclusion: The Lancet’s multi-metric components vary substantially in their distribution across preclinical and clinical obesity phenotypes. While the combination of BMI with all central and hip anthropometrics captures the highest proportion of absolute visceral and clinical obesity in the U.S. population, we found that the hip-derived WHR performs poorly alone. Simpler waist-centric combinations (BMI+WC+WHtR) achieve comparable PPVs here. Thus, implementation of the Lancet framework may have the potential to be streamlined.