Prospective longitudinal studies evaluating the relevance of “Metabolically Healthy but Obese” (MHO) phenotype at risk for type 2 diabetes mellitus (T2D) and cardiovascular diseases are few and results are contradictory.

As a representative of the general population, 1051 individuals were evaluated in 1997–1998 and re-evaluated after 6 years and 11 years. Subjects without known T2D were given an oral glucose tolerance test. Anthropometric and biochemical variables were measured. Four sets of criteria were considered to define MHO subjects besides body mass index ≥30 kg/m²: A: Homeostatic Model of Assessment-Insulin Resistance Index (HOMA-IR) <90th percentile; B: HOMA-IR <90th percentile, high-density lipoprotein cholesterol >40 mg/dL in men and high-density lipoprotein cholesterol >50 mg/dL in women, triglycerides <150 mg/dL, fasting glucose <110 mg/dL, and blood pressure ≤140/90 mm Hg; C: HOMA-IR <90th percentile, triglycerides <150 mg/dL, fasting glucose <110 mg/dL, and blood pressure ≤140/90 mm Hg; D: HOMA-IR <90th percentile, triglycerides <150 mg/dL, and fasting glucose <110 mg/dL. Subjects with T2D at baseline were excluded from the calculations of incidence of T2D.

The baseline prevalence of MHO phenotype varied between 3.0% and 16.9%, depending on the set of criteria chosen. Metabolically nonhealthy obese subjects were at highest risk for becoming diabetic after 11 years of follow-up (odds ratio = 8.20; 95% confidence interval = 2.72–24.72; P < .0001). In MHO subjects the risk for becoming diabetic was lower than in metabolically nonhealthy obese subjects, but this risk remained significant (odds ratio = 3.13; 95% confidence interval = 1.07–9.17; P = .02). In subjects who lost weight during the study, the association between MHO phenotype and T2D incidence disappeared, even after adjusting for HOMA-IR.

The results suggest that MHO is a dynamic concept that should be taken into account over time. As a clinical entity, it may be questionable.