Bmi system could ‘misclassify’ patients — scans show 34% of obese labeled by BMI were overweight
A scan-based comparison of body composition has found the bmi tool may misclassify a sizable fraction of patients referred for assessment. The study used dual-energy X-ray absorptiometry (DXA) scans to re-evaluate 1, 351 adults who had been categorised by BMI, and found that 34% of people classed as obese by BMI were actually in the overweight category on scan measurements, while 53% of those labelled overweight were assigned a different category after scanning.
Why this matters right now
The contrast between BMI categories and DXA measurements matters because population figures show high levels of excess weight: NHS data show 30% of adults in England were obese in 2024, and 66% were overweight or obese. Clinical decision-making, public-health messaging and individual counselling often rely on BMI thresholds. If the tool assigns the wrong category to a large share of people, there are immediate implications for who is prioritised for care and how health risks are communicated.
Bmi reclassification: what lies beneath the headline
The Italian research team compared standard BMI assignments — produced by dividing weight in kilograms by height in metres squared — with DXA scans that analyse fat, muscle and bone. In the study sample of 1, 351 adults, initial BMI measurement placed 19 people in the underweight category, 787 in the normal-weight category, 354 as overweight and 191 as obese. After DXA scanning, more than a third (34%) of those recorded as obese by BMI were reclassified into the overweight category, and just over half (53%) of those recorded as overweight by BMI were shown to be in a different category when body composition was measured directly.
Importantly, three-quarters of the misclassified overweight patients were found to be a normal weight on DXA scans, while the remaining misclassified overweight individuals should have been in the obese category based on scan results. The use of DXA allowed the researchers to separate fat, muscle and bone compartments, exposing discrepancies between a height–weight index and direct measures of body composition.
Expert perspectives and regional impact
Researchers at the Department of Neurosciences, Biomedicine and Movement Sciences at the University of Verona conducted the reclassification study on adults referred to that department; the sample therefore reflects referred patients rather than a general population survey. The study team highlighted the magnitude of misclassification when BMI is compared with DXA-based body composition analysis and noted the specific counts and percentages observed across BMI categories.
At a regional level, the findings intersect with NHS statistics that place obesity prevalence in England at 30% and combined overweight-or-obesity prevalence at 66% for 2024. Those population figures underscore why a mismatch between BMI and direct body composition measures could alter assessments at scale: if large groups are categorised differently by measurement method, estimates of who is overweight or obese will differ depending on the tool used.
What this means going forward
The study shows clear numerical mismatches between BMI categories and DXA scan results within a clinical referral sample. The differences — 34% of BMI-classed obese reclassified as overweight, 53% of BMI-classed overweight reclassified — point to limits of a height-and-weight index when compared with direct composition analysis. Given the NHS figures on overweight and obesity prevalence, the choice of measurement method will influence both individual clinical pathways and broader surveillance figures.
How should clinicians and health systems reconcile BMI with direct measures like DXA when the two lead to different classifications, and what balance should be struck between practicality and precision?
