TOPLINE

A simple definition of meeting BMI or waist circumference criteria outperformed the new Lancet obesity definition on some metrics in US adults when validated against DEXA body scans. It showed higher specificity and better overall accuracy than the more complex Lancet definition, which requires hip measurements that are rarely taken in clinics.

METHODOLOGY

• BMI has long been the standard tool for diagnosing obesity, but it doesn’t fully capture body composition and metabolic risk. The Lancet Diabetes & Endocrinology Commission on Clinical Obesity proposed a more detailed alternative, incorporating waist circumference, waist-to-height ratio, and waist-to-hip ratio. Whether this added complexity improves diagnostic accuracy remains unclear.
• Researchers conducted a diagnostic study using data from 1900 nonpregnant US adults aged 20-59 years whose whole-body DEXA scan data were available in the 2017-2018 National Health and Nutrition Examination Survey.
• Criterion-standard obesity was defined by DEXA as total body fat ≥ 25% in men and ≥ 35% in women, whereas visceral obesity was defined as a visceral fat area ≥ 130 cm² in men and ≥ 100 cm² in women.
• Five obesity definitions were compared: the full Lancet definition, a simplified Lancet definition excluding the waist-to-hip ratio, a definition meeting the BMI or waist circumference criterion, waist circumference alone, and BMI alone.
• Outcome measures included sensitivity, specificity, the area under the receiver operating characteristic curve (AUROC), along with other metrics for each definition.

TAKEAWAY

• The full Lancet definition achieved a sensitivity of 87.7%, a specificity of 70.7%, and an AUROC of 0.790 for identifying DEXA-based obesity; the BMI or waist circumference approach yielded lower sensitivity (74.0%) but higher specificity (89.5%) and a significantly higher AUROC (0.818; P < .001) than the full Lancet definition.
• When applied to the US adult population aged 20-59 years between 2021 and 2023, the full Lancet obesity definition would produce about 15.5 million false positives — people labeled as having obesity who do not meet the DEXA-based standard — and about 14.3 million false negatives — people who have obesity but are missed — compared with about 5.5 million false positives and about 30.3 million false negatives using the BMI or waist circumference approach.
• Waist circumference alone achieved an AUROC of 0.815, with sensitivity and specificity comparable to those achieved by the BMI or waist circumference approach; BMI alone performed the worst, yielding a sensitivity of 51.3% and an AUROC of 0.730.
• For visceral obesity, the full Lancet definition performed worse than the BMI or waist circumference approach, yielding lower specificity (45.1% vs 65.0%) and a lower AUROC (0.717 vs 0.785) despite achieving higher sensitivity (98.3% vs 92.0%).

IN PRACTICE

“For clinical practice and individual-level decision-making, where specificity, appropriate resource use, and avoidance of diagnostic labeling harms are paramount, we recommend BMI or WC [waist circumference] or WC-alone definitions, which offer the best balance of sensitivity and specificity, with substantially fewer false positives than the full Lancet definition,” the authors of the study wrote.

SOURCE

The study was led by Aayush Visaria, MD, MPH, Department of Medicine, Rutgers Robert Wood Johnson Medical School, New Brunswick, New Jersey. It was published online as a research letter on August 7 in JAMA Network Open.

LIMITATIONS

The study was restricted to adults aged 20-59 years, which limits generalizability to pediatric and older populations. The distinction made by the Lancet definition between preclinical and clinical obesity was not assessed in this analysis.

DISCLOSURES

The study received support from the National Institutes of Health (NIH) National Center for Advancing Translational Sciences in the form of a grant and from the NIH National Institute on Minority Health and Health Disparities Rutgers-New York University Center for Asian Health Promotion and Excellence in the form of a pilot grant. One author disclosed receiving grants from Daiichi Sankyo and Pfizer Japan and personal fees from Merck and Regeneron outside the submitted work.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.