Observational study finds BMI overestimates obesity prevalence by 4% in adults — Evidence Review

A new Italian study finds that BMI often misclassifies individuals' weight status compared to body fat measured by DXA, with over a third of adults placed in the wrong category. Related research largely supports these results, showing BMI both over- and underestimates true obesity rates, especially when compared to direct measures of adiposity. For more details, see the original source.

• Several studies confirm that BMI has limitations as a proxy for body fat, often underestimating or overestimating obesity, particularly among women and older adults, and may miss those at increased health risk despite "normal" BMI readings 2 3 12 14.
• Direct measurements of body fat (such as DXA) and alternative anthropometric measures (like waist-to-height ratio) provide a more accurate assessment of health risks related to adiposity 2 3 11 12 14.
• While self-reported height and weight can be useful for large-scale studies, measured values improve accuracy, and misclassification is more common in certain subgroups, highlighting the need for improved assessment tools 1 5 6 8.

Study Overview and Key Findings

Concerns about the accuracy of BMI as a standard measure for assessing overweight and obesity have grown in recent years. The Italian study, set to be presented at the European Congress on Obesity and published in Nutrients, addresses these concerns by comparing BMI classifications to those based on DXA body fat percentage—a more precise measure of adiposity. The findings suggest that BMI may not only misclassify millions of individuals but also inflate the estimated prevalence of overweight and obesity in the population, with important implications for public health assessments and clinical decision-making.

Property	Value
Study Year	2026
Organization	University of Modena and Reggio Emilia, University of Verona, Beirut University
Journal Name	Nutrients
Authors	Marwan El Ghoch, Chiara Milanese
Population	Adults aged 18 to 98 years
Sample Size	1351 adults
Methods	Observational Study
Outcome	BMI misclassification rates and body fat percentage
Results	BMI overestimated overweight and obesity prevalence by 4%.

Key Study Highlights:

• The study included 1,351 Italian adults (60% female), all White Caucasian, aged 18 to 98.
• BMI and DXA-based body fat measurements were compared; DXA is considered the gold standard for assessing adiposity.
• Among those classified as obese by BMI, 34% were actually overweight by DXA; among those labeled overweight by BMI, 53% were misclassified, with most actually falling in the normal weight category by DXA.
• Overall, DXA-based analysis showed a combined overweight/obesity rate of 37%, compared to 41% using BMI.
• Researchers recommend supplementing BMI with body composition or circumference-based measures for more accurate health risk stratification.

Literature Review: Related Studies

To contextualize these findings, we searched the Consensus database of over 200 million research papers using targeted queries. The following search queries were used:

1. BMI misclassification prevalence study
2. overweight obesity estimation accuracy
3. BMI limitations health risk assessment

Below, we summarize key topics that emerged from the literature, drawing on evidence from a range of related studies.

Topic	Key Findings
How accurately does BMI classify overweight and obesity compared to direct measures?	- BMI both underestimates and overestimates true obesity prevalence when compared to direct body fat measurements, particularly in women and older adults 2 3 12 14. - Misclassification rates vary by population, with substantial proportions of individuals categorized incorrectly as overweight or obese using BMI alone 2 3 14.
What are the limitations of BMI as a health risk assessment tool?	- BMI does not reliably indicate body fat percentage or fat distribution, which are better predictors of health risks such as cardiometabolic disease 11 12 14. - Individuals with "normal" BMI may still have high body fat and elevated health risks, while some with high BMI may be metabolically healthy 3 11 12 14.
How reliable are self-reported height and weight for BMI calculations?	- Self-reported height tends to be overestimated and weight underestimated, leading to further BMI misclassification; measured values are more accurate, especially for identifying obesity 1 5 6 8. - Correction equations can improve the accuracy of self-reported data, but biases persist, especially among older adults, women, and those with higher BMI 1 5 6 8.
What alternative or supplementary measures improve obesity and health risk assessment?	- Direct measurements of body fat (e.g., DXA), waist circumference, and waist-to-height ratio provide a more precise assessment of adiposity and cardiometabolic risk than BMI alone 2 3 9 11 12 14. - Combining BMI with other anthropometric or performance metrics enhances individual risk stratification and guides more effective interventions 11 12 14 15.

How accurately does BMI classify overweight and obesity compared to direct measures?

The new Italian study aligns with a growing body of research indicating that BMI is an imperfect proxy for body fat percentage and true obesity status. Studies using gold-standard body composition techniques such as DXA consistently show that BMI can both underestimate and overestimate obesity rates, with women and older adults particularly affected by misclassification 2 3 12 14.

• BMI misclassification rates are substantial, especially for overweight and obese categories, when compared to DXA or plethysmography 2 3 14.
• In one study, 48% of women classified as non-obese by BMI were obese by DXA 2.
• Older adults and women are more likely to be misclassified, with BMI underestimating obesity prevalence in these groups 2 3 14.
• The Italian study's finding of a 4% overestimation of overweight/obesity is consistent with reports that BMI both over- and underestimates prevalence, depending on population and cutoff values 2 3 12 14.

What are the limitations of BMI as a health risk assessment tool?

Multiple studies highlight that BMI, while practical for population surveillance, is limited in individual risk prediction because it does not capture fat distribution or percentage body fat—factors more strongly linked to cardiometabolic risk 11 12 14. The new study's recommendation to use additional measures echoes these concerns.

• Central obesity (measured by waist circumference or waist-to-height ratio) is more predictive of early health risks than BMI alone 11.
• Individuals with normal BMI but high body fat may have elevated metabolic and cardiovascular risk—a group often missed by BMI screening 3 11 12 14.
• BMI does not account for age, sex, or ethnic differences in body composition that affect health risk profiles 12 14.
• Evidence shows that BMI cut-offs may need adjustment for different populations to improve risk classification 14.

How reliable are self-reported height and weight for BMI calculations?

Large cohort studies often rely on self-reported anthropometrics, but this introduces systematic biases—height is overreported, weight underreported, and BMI subsequently underestimated. While self-reported data are useful for large-scale analyses, measured values are more accurate, particularly for identifying obesity 1 5 6 8.

• Self-reported BMI underestimates obesity prevalence, especially among women, older adults, and those with higher BMI 1 6.
• Misclassification can be reduced by using correction equations, but residual bias remains 1 5 8.
• In the U.S., approximately 13% of men and 7% of women were classified into a lower BMI category based on self-report versus measured data, with obese individuals most affected 5.
• Self-reported data are more reliable for identifying trends and associations than for precise prevalence estimates 1 5 6 8.

What alternative or supplementary measures improve obesity and health risk assessment?

The literature consistently supports supplementing BMI with additional anthropometric or direct adiposity measures to improve health risk assessment. Waist circumference, waist-to-height ratio, and direct body composition techniques (DXA, plethysmography) provide better estimates of adiposity and associated health risks 2 3 9 11 12 14.

• Waist-to-height ratio (cutoff 0.5) is a simple, predictive indicator of early health risk, outperforming BMI and waist circumference matrices 11.
• Combining BMI with body composition and functional fitness measures yields a more complete picture of individual health risk 12 14.
• Direct assessment of body fat is recommended where possible, particularly for clinical decision-making and in populations at higher risk of misclassification 2 3 14.
• For energy expenditure estimation in obesity, no single anthropometric formula is ideal, highlighting the complexity of translating body size to metabolic health 9.

Future Research Questions

Further research is needed to refine obesity assessment tools, validate findings across diverse populations, and improve individual risk stratification. The following questions address gaps and next steps:

Research Question	Relevance
How do BMI misclassification rates vary across different ethnic groups?	The current study focused on White Caucasian adults; research is needed to determine if misclassification patterns are similar in other ethnicities, as body composition and BMI-health risk relationships differ 12 14.
What is the impact of using waist-to-height ratio or other anthropometric measures alongside BMI in clinical practice?	Supplementing BMI with additional metrics may improve risk stratification and clinical outcomes; further studies could clarify implementation and effectiveness in practice 11 12 14.
How do BMI misclassification rates affect public health policy and resource allocation?	Over- or underestimating obesity prevalence influences policy decisions, funding, and intervention targeting; understanding these effects can optimize population health strategies 7 14.
Can body composition assessment methods be made more accessible and cost-effective for routine use?	The wider adoption of DXA or similar techniques is limited by cost and feasibility; research into simpler, affordable alternatives could enhance routine obesity assessment 2 3 11 12 14.
What are the long-term health outcomes for individuals misclassified by BMI?	Understanding whether misclassified individuals (e.g., those with normal BMI but high body fat) face higher long-term cardiometabolic risks would inform screening and intervention priorities 3 11 12 14.