This study sought to investigate whether the metabolically healthy obese (MHO) phenotype is associated with an increased risk of incident type 2 diabetes in a Korean population and, if so, whether systemic inflammation affects this risk in MHO individuals.

The study population comprised 36 135 Koreans without type 2 diabetes. Participants were stratified by body mass index (cutoff value, 25.0 kg/m²) and metabolic health state (assessed using Adult Treatment Panel-III criteria). High-sensitive C-reactive protein (hsCRP) was used as a surrogate marker of systemic inflammation. Subjects were classified into low (ie, hsCRP < 0.5 mg/L) and high (ie, hsCRP ≥ 0.5 mg/L) systemic inflammation groups.

During a median followup of 36.5 months (range, 4.8–81.7 mo), 635 of the 36 135 individuals (1.8%) developed type 2 diabetes. The MHO group had a significantly higher risk of incident type 2 diabetes (multivariate-adjusted hazard ratio [HR], 1.57; 95% confidence interval [CI], 1.16–2.11) than the metabolically healthy nonobese (MHNO) group. However, the risk of the MHO group varied according to the degree of systemic inflammation. Compared with the MHNO/low systemic inflammation group, the risk of type 2 diabetes in the MHO/low systemic inflammation group was not significantly elevated (multivariate-adjusted HR, 1.61; 95% CI, 0.77–3.34). However, the MHO/high systemic inflammation group had an elevated risk of incident type 2 diabetes (multivariate-adjusted HR, 3.73; 95% CI 2.36–5.88).

MHO subjects show a substantially higher risk of incident type 2 diabetes than MHNO subjects. The level of systemic inflammation partially explains this increased risk.