BackgroundTo compare the prevalence and metabolic characteristics of metabolically healthy but obese (MHO) individuals according to different criteria.MethodsWe examined 186 MHO middle-aged men (age, 37.2 years; body mass index [BMI], 27.2 kg/m2). The following methods were used to determine MHO: the National Cholesterol Education Program (NCEP) Adult Treatment Panel III criteria, 0-2 cardiometabolic abnormalities; the Wildman criteria, 0-1 cardiometabolic abnormalities; the Karelis criteria, 0-1 cardiometabolic abnormalities; the homeostasis model assessment [HOMA] criteria (lowest quartile of HOMA). After dividing the overall subjects into two age groups, we compared the prevalence and clinical characteristics between MHO and at-risk groups according to four different criteria.ResultsThe prevalence of MHO using the NCEP, Wildman, Kaleris, and HOMA criteria were 70.4%, 59.7%, 28.5%, and 24.2%, respectively. The agreement between the groups according to the NCEP and Wildman criteria was substantial (kappa = 0.8, P < 0.001). Among individuals 35 years or younger, and regardless of method, the MHO subjects had significantly lower weight, waist circumference, BMI, body fat percentage, insulin, HOMA, alanine aminotransferase, triglyceride (TG), and TG/high density lipoprotein cholesterol (HDL-C) ratio than the at-risk subjects (P < 0.05); However, among individuals older than 35 years old, and regardless of method, the MHO subjects had different insulin, HOMA, HDL-C, and TG/HDL-C levels than the at-risk subjects (P < 0.05).ConclusionThe differences in metabolic profile between MHO and at-risk groups varied according to age. MHO prevalence varies considerably according to the criteria employed. Expert consensus is needed in order to define a standardized protocol for determining MHO.