© 2003 Diabetes UK. Diabetic Medicine, 20, 99–104 99 Introduction Diabetes mellitus (DM) and coronary heart disease share many risk factors [1]. The metabolic syndrome is the combination of multiple metabolic and physiological risk factors for both Type 2 DM and atherosclerotic cardiovascular disease. In general, metabolic syndrome consists of hypertension, glucose intolerance, obesity, and dyslipidaemia [2,3]. An important feature is insulin resistance, characterized by increased serum insulin among non-diabetic individuals. It has been suggested that increased insulin itself may be atherogenic [4]. However, whether insulin resistance or hyperinsulinaemia plays a central role in the metabolic syndrome is still debated [5,6]. Recently, factor analysis using the principal component method has been adopted to identify major factors of the metabolic syndrome [7]. This statistical tool is particularly useful when variables are highly intercorrelated. The principal component analysis identifies subsets of correlated variables that group together [8]. These subsets define components: mathematically derived variables that are uncorrelated with each other and that explain the majority of the variance in the original data. Thus, a confusingly large number of apparently Correspondence to: Sei Hyun Baik MD, Department of Internal Medicine, Korea University Guro Hospital, 80 Guro-Dong, Guro-Gu, Seoul 152-050, South Korea. E-mail: 103hyun@korea.ac.kr Abstract Aims To examine the relationship between the components of the metabolic syndrome and to explore whether insulin resistance unifies the clustering of components of the metabolic syndrome among urban elderly Koreans using exploratory factor analysis. Methods We included 1314 non-diabetic subjects over the age of 60 years, selected from a cross-sectional study, which was conducted in 1999 in Seoul, Korea. Factor analysis was carried out using the principle components analysis with Varimax orthogonal rotation of the components of the metabolic syndrome. Results We found four major factors of cardiovascular disease risk variables in our study subjects. Impaired glucose tolerance, dyslipidaemia, hypertension and obesity aggregated as the major domain. Obesity and dyslipidaemia variables were closely related and loaded on the same factor. However, hypertension was not linked closely with other factors of the metabolic syndrome. Conclusions Insulin resistance is not the only contributor to the metabolic syndrome among urban elderly Koreans. Although the components of the metabolic syndrome were closely related, the finding of more than one factor suggests that more than one pathophysiological mechanism underlies full expression of the metabolic syndrome among elderly Koreans. Diabet. Med. 20, 99 –104 (2003) Keywords metabolic syndrome, factor analysis, insulin resistance, diabetes, coronary artery disease, risk factors Abbreviations HOMA, homeostasis model assessment; IR, insulin resistance Blackwell Science, Ltd Oxford, UK DME Diabetic Medicine 0742-3071 Blackwell Science Ltd, 2003 20 Original Article Original article Factor analysis of the metabolic syndrome among elderly Koreans K. M. Choi et al. Factor analysis of the metabolic syndrome among elderly Koreans—the South-west Seoul Study K. M. Choi, J. Lee*, K. B. Kim, D. R. Kim, S. K. Kim, D. H. Shin, N. H. Kim, I. B. Park, D. S. Choi and S. H. Baik Department of Internal Medicine and *Department of Preventive Medicine, College of Medicine, Korea University, Seoul, South Korea Accepted 23 May 2002