Letter to the Editor Clustering trend of components of metabolic syndrome Yoshifusa Aizawa a, ⁎ , Hiroshi Watanabe a , Mahmoud M. Ramadan a , Yoshimaru Usuda b , Tohru Watanabe b , Shigeru Sasaki b a Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan b Niigata Association for Comprehensive Health Promotion and Research, Niigata, Japan Received 28 July 2006; accepted 4 August 2006 Available online 9 November 2006 Abstract Hereby we report our observations derived from the study of a huge cross-sectional sample of the general population that metabolic risk factors (RFs) tend to cluster into metabolic syndrome (MS). The clustering trend was considered to be intensified with 4–5 RFs. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Metabolic syndrome; Risk factors; Clustering The term “Metabolic Syndrome” (MS) refers to a con- stellation of cardiovascular risk factors (RFs), being diag- nosed when 3 or more of the five individual elements constituting the syndrome are found together in the same individual [1]. In the present study, we tried to reveal the possible tendency of RFs to cluster within the MS definition framework in a huge community-based sample of adult men and women. All residents aged 40 years or older and not under medical treatment were recommended to receive the annual health examination which was planned and guided by the Niigata Association for Comprehensive Health Promotion and Research. This study involved 119,412 subjects (41,819 males and 77,593 females), the total number of individuals who underwent the above-mentioned examination in 2003 in Niigata prefecture (Japan). Diagnosis of MS in our study was based on the National Cholesterol Education Panel-Adult Treatment Panel III (ATP3) definition criteria [2] after replacing waist circumference (WC) of ATP3 with body mass index (BMI) to define obesity according to the rec- ommendation of the Japan Society for the Study of Obesity [3]. The MS was defined as the combination of at least 3 of the following criteria: BMI ≥ 25 kg/m 2 , ABP ≥ 130/85 mm Hg, hypertriglyceridaemia (triglyceride [TG] level ≥ 150 mg/dl), low high-density lipoprotein-cholesterol [HDL-C] level (b 40 mg/dl in men and b 50 mg/dl in women), and abnormal glucose metabolism (fasting blood glucose [FBG] ≥ 110 mg/ dl). From these data, we determined the actual prevalence of each combination of RFs (3, 4 and 5 factors) within the MS; and compared it with a predicted hypothetical prevalence of the same combination. The latter was calculated upon the assumption that each RF within the corresponding combina- tion would behave independently from the other factors, giving a null hypothesis of non-tendency of RFs to cluster. MS was found in 14.9% of subjects (17.9% of men, and 13.4% of women); significantly lower than 24–25% for US [4]. The actual prevalence of combinations of 3, 4 and 5 RFs was 10.99%, 3.52% and 0.45% respectively; while the International Journal of Cardiology 121 (2007) 117 – 118 www.elsevier.com/locate/ijcard ⁎ Corresponding author. Division of Cardiology, Niigata University Graduate School of Medical and Dental Sciences, 1-754 Asahi Machi, Niigata 951–8510, Japan. Tel.: +81 25 227 2182; fax: +81 25 228 5611. E-mail address: aizaways@med.niigata-u.ac.jp (Y. Aizawa). Table 1 Comparison of actual and predicted prevalence of different risk factor combinations in subjects with metabolic syndrome RF combinations Actual prevalence Predicted prevalence p-value Actual/predicted odds 3 factors 10.99% 9.26% b 0.0001 1.19 4 factors 3.52% 1.24% b 0.0001 2.84 5 factors 0.45% 0.05% b 0.0001 8.89 0167-5273/$ - see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijcard.2006.08.051