Application of the Lancet Commission’s revised obesity definition to the All of Us Research Program cohort revealed that 69% of approximately 301,026 US adults met criteria for obesity—a ~60% increase from the 43% identified using traditional body mass index (BMI) thresholds. The increase stemmed almost entirely from inclusion of patients with anthropometric-only obesity; only about 0.2% of patients with high BMI had nonelevated anthropometrics and would no longer be labeled as having obesity under the new framework.

Published in JAMA Network Open, the study evaluated participants enrolled between May 31, 2017, and September 30, 2023, with a median follow-up of 4 years. The cohort comprised 61% female patients with a median age of 54 years.

Dual Obesity Phenotypes Emerge

The new framework stratified obesity into two mutually exclusive phenotypes. BMI-plus-anthropometric obesity—defined as BMI above traditional thresholds plus at least one elevated anthropometric measure or BMI greater than 40—encompassed ≈42.6% (n = 128,314) of the total cohort. Anthropometric-only obesity, characterized by at least two elevated anthropometric measures with BMI below traditional thresholds, accounted for 26% (n = 78,047) of the total cohort.

Patients with anthropometric-only obesity were older (median age 60 vs. 54 years), more often male (49% vs. 35%), and reported higher educational attainment (45% vs. 35% with college degrees) compared with those having BMI-plus-anthropometric obesity. Among the patients with anthropometric-only obesity, 22% (n = 17,426) had BMI traditionally classified as normal or underweight, while the remainder fell within the overweight category. All three anthropometric measures—waist circumference, waist-to-hip ratio, and waist-to-height ratio—were elevated in 49% (n = 37,856) of this group.

Clinical Obesity Peaks With Age

Overall 36% (n = 108,650) of participants met criteria for clinical obesity by the new definition—presence of obesity-associated organ dysfunction and/or physical limitation. Clinical obesity prevalence increased markedly with age, affecting 9% of all patients aged 18 to 29 years but 54% of all those 70 years or older.

Among participants 70 years or older, 78% (n = 35,268) had obesity under the revised framework—double the prevalence under traditional BMI criteria. Of these, 70% had clinical obesity, underscoring the concentration of organ dysfunction in older adults. Anthropometric-only obesity also increased progressively with age, affecting 27% of patients aged 18 to 29 years with obesity but 53% of those 70 years or older.

Organ Dysfunction Risk

After adjusting for age, sex, and race, odds ratios for organ dysfunction were 3.3 for BMI-plus-anthropometric obesity and 1.8 for anthropometric-only obesity compared with no obesity. Rates of clinical obesity were higher in the BMI-plus-anthropometric group (56% vs. 48%) with a greater median number of organ-dysfunction manifestations (1 vs. 0) than those with anthropometric-only obesity.

Organ dysfunction frequency increased progressively across BMI categories—from 48% and 44% at normal weight to 65% and 63% in obesity class 3 among men and women, respectively. The most common manifestations were hypertension, physical limitation, and obstructive sleep apnea.

Cardiometabolic Outcomes

Clinical obesity conferred adjusted hazard ratios (aHRs) of 6.1 for incident diabetes, 5.9 for cardiovascular events, and 2.7 for all-cause mortality compared with no obesity or organ dysfunction. Preclinical obesity—obesity without organ dysfunction—carried intermediate risks (aHRs 3.3 for incident diabetes; 1.4 for cardiovascular events) and did not show increased all-cause mortality versus the reference group.

By obesity phenotype, BMI-plus-anthropometric obesity yielded aHR 3.9 for incident diabetes, 1.8 for cardiovascular events, and 1.2 for all-cause mortality. Anthropometric-only obesity produced aHRs of 2.1, 1.5, and 1.2, respectively. Organ dysfunction without obesity was independently associated with aHRs of 2.5 for incident diabetes, 4.7 for cardiovascular events, and 2.8 for all-cause mortality.

Racial Disparities and Treatment Implications

Implementation of the Lancet Commission criteria produced the largest relative increase in obesity prevalence among Asian patients—a 90% rise from 27% to 51%. Clinical obesity was the least prevalent in this group (20% overall; 38% among those with obesity), with metabolic dysfunction predominating.

While nearly 45% of the cohort met current BMI-based eligibility criteria for obesity pharmacotherapy, 15,495 of 69,894 (22%) with clinical obesity under the revised framework did not. (This pharmacotherapy analysis excluded patients with diabetes who would otherwise be candidates for GLP-1 receptor agonists GLP−1RAsGLP-1 RAsGLP−1RAs.) Conversely, 57,068 of 111,467 (51%) patients eligible by BMI-based indications did not have clinical obesity.

Among 76,460 patients with normal BMI, approximately 9,388 (12%) were reclassified as having preclinical obesity and 7,902 (10%) as having clinical obesity. Among 91,644 patients with overweight BMI, 31,403 (34%) met criteria for preclinical obesity and approximately 29,218 (32%) for clinical obesity.

The study followed STROBE reporting guidelines. Funding came from multiple National Institutes of Health grants. Disclosures can be found in the study.

Source: JAMA Network Open