To elucidate the role of visceral fat accumulation in the metabolic syndrome, differences in the pathology of the metabolic syndrome with or without visceral fat accumulation were investigated. A total of 472 prediabetic Japanese men (mean age, 47.5 +/- 7.2 yr) with impaired fasting glycemia (IFG) levels of 110-125 mg/dL were eligible for participation in the study. The study subjects were divided into the following four groups, and intergroup comparisons were made: group I without visceral fat area [VFA] > or = 100 cm2 but presenting with fewer than two other risk factors (i.e., TG > or =150 mg/dL, HDL-C < 40 mg/dL, BP > or = 130/ > or = 85 mmHg, or FPG > or = 110 mg/dL) (n = 231); group II without VFA of > or = 100 cm2 but presenting with three or more other risk factors (n = 57); group III with VFA of > or = 100 cm2 accompanied by FPG 110 mg/dL alone (n = 27); and group IV with VFA > or =100 cm2 and two or more other risk factors (n = 157). The prevalence of patients who had three or more risk factors with or without VFA > or = 100 cm2 was 45.3% (214 out of 472 patients), while that of those with VFA > or = 100 cm2 who had two or more other risk factors was 33% (157 out of 472 patients). Group II had significantly higher VFA values than group I (p < 0.05), and group IV had significantly higher VFA values than group II (p < 0.001). While no significant differences in HOMA-R values were seen between groups I and II, these values were significantly higher in group IV compared to groups I and II (p < 0.001 and p < 0.05, respectively). Furthermore, group IV showed significantly higher 2-h insulin levels after glucose loading compared to group I (p < 0.001). While no significant differences were seen between groups II and IV, insulin levels tended to be higher in group IV. Adiponectin levels showed an incremental fall in VFA from group I through groups II and III to group IV. Groups III and IV showed significantly lower adiponectin levels compared to group I (p < 0.05, p < 0.001, respectively); and group IV showed significantly lower adiponectin levels than group II (p < 0.05). A logistic regression analysis using VFA, TG and HDL-C, and BP as explanatory variables showed that the relative risk for high HOMAR values were 2.65 (p < 0.001) for patients with VFA > or =100 cm2; 1.64 (p < 0.05) for those with TG > or = 150 mg/dL and HDL < 40 mg/dL; and 1.79 (p < 0.01) for those with BP > or = 130/ > or = 85 mmHg. These findings demonstrate that the degree of insulin resistance and the risk of arteriosclerosis vary depending on whether or not the metabolic syndrome accompanied by a clustering of risk factors has visceral fat accumulation as an underlying pathology, strongly suggesting a crucial role for visceral fat accumulation in the metabolic syndrome.