ADA 2026 obesity criteria raise prevalence across 45 population surveys

The study

This paper in BMJ Open Diabetes Research & Care compared two ways of defining obesity in adults aged 20 to 59 years. One is the WHO body mass index (BMI) criterion. The other is the 2026 American Diabetes Association (ADA) criteria. The authors also asked how reclassification relates to elevated measured blood pressure.

The analysis used 45 population-based surveys with harmonised measurements of body size, blood pressure and sampling weights. It included 137 600 participants. The authors used modified Poisson models adjusted for age, sex and survey. Sensitivity analyses varied the survey weighting, the way Asian thresholds were assigned, and retention. The abstract does not report which countries the surveys came from, how elevated blood pressure was defined, or the funding.

What they found

Mean survey-specific obesity prevalence was 28.2% with the WHO criterion and 53.6% with the ADA criteria. The mean increase was 25.4 percentage points (median 25.9, range 4.4 to 41.2). Surveys differed a great deal (I² 99.1%; 95% prediction interval 3.3 to 47.4).

For elevated measured blood pressure, the ADA criteria had higher sensitivity than WHO (69.8% vs 42.5%). They had lower specificity (51.4% vs 76.2%) and lower positive predictive value (31.3% vs 37.3%). Balanced accuracy was 60.6% with ADA and 59.3% with WHO. The abstract does not say whether this difference was tested.

People reclassified as having obesity through central adiposity had a higher prevalence of elevated measured blood pressure than people with overweight who were not reclassified. The adjusted prevalence ratio was 1.34 (95% CI 1.24 to 1.46). The authors report similar findings across the sensitivity analyses.

How much weight to give it

This is an observational analysis of survey data. It shows association, not cause. Blood pressure was a single type of outcome, and the abstract gives no hard clinical outcomes such as cardiovascular events.

The number of surveys and participants is large. But the differences between surveys were very large, so the pooled mean may not describe any one population well. The prediction interval is wide, and the 25.4 percentage point mean increase should be read with that in mind.

The trade-off in screening performance is clear. The ADA criteria had higher sensitivity for elevated blood pressure but lower specificity. Balanced accuracy was close between the two approaches. The abstract does not describe the ADA criteria or the Asian thresholds in detail, so the full paper should be read before drawing conclusions for a particular population.

For your clinic

In these surveys, adults reclassified as having obesity through central adiposity had a higher prevalence of elevated measured blood pressure than adults with overweight who were not reclassified (adjusted prevalence ratio 1.34; 95% CI 1.24 to 1.46). The authors suggest reporting waist measures alongside BMI to strengthen cardiometabolic surveillance and to inform locally proportionate prevention.

This is population-level evidence from surveys and does not guide the care of an individual patient. Before changing how patients are classified or counselled, read the full paper and the current ADA and national guidance. Check especially how the criteria and thresholds are defined and whether the surveys resemble your patient population.