Obesity prevalence: comparison of traditional and new classification approaches in a Swiss population-based study (2005–2024)

Obesity has traditionally been defined using body mass index (BMI), but this may overlook central adiposity and related metabolic risks. In 2025, new guidelines recommended adding anthropometric measures. This study analyzed obesity prevalence comparing definitions and their associations with cardiovascular and metabolic diseases. Using a population-based study in Geneva, Switzerland (2005–2024), we measured the prevalence of obesity based on the traditional and new definitions. Reclassification patterns were examined, and associations with diabetes, hypertension, and dyslipidemia were assessed via logistic regression and receiver operating characteristic analyses. Among 14,658 individuals (mean age 48.2 ± 13.7; 51.4% women), obesity prevalence ranged from 10.8% to 39.9% using new classifications, compared to 13.1% with BMI alone (p < 0.001). Reclassifications differed among men and women and with age. New classifications demonstrated superior discriminative performance for the detection of cardiovascular and metabolic outcomes compared to BMI alone. BMI + waist-to-hip ratio showed the strongest associations with diabetes (aOR 4.61; 3.87–5.47), and hypertension (aOR 3.61; 3.18–4.09), while waist-to-hip and waist-to-height ratio showed the strongest association with dyslipidemia (aOR 1.95; 1.75–2.16). Adding anthropometric measures to BMI substantially improves obesity detection. Choosing the combination of anthropometric measures and BMI with awareness of sex and age-related changes better identifies at risk individuals. This new definition can be a powerful low-cost tool in detecting individuals on the obesity spectrum early in primary care settings and prevent complications.