Chinese metabolic syndrome risk score

Fone Ching Hsiao, Chung Ze Wu, Chang Hsun Hsieh, Chih-Tsueng He, Yi Jen Hung, Dee Pei

Research output: Contribution to journalArticlepeer-review

12 Citations (Scopus)


BACKGROUND: The metabolic syndrome (MetS) was first proposed to predict the occurrence of cardiovascular disease and type 2 diabetes. However, it is difficult to identify subjects with MetS early. No previous studies designed to develop a predictive model for MetS in the Chinese population exists; this study was designed to fill that gap. METHODS: A middle-aged Chinese cohort of 198 men and 154 women were followed for two years. The binary logistic regression and receiver operation characteristic (ROC) curve were used to develop a predictive model for the future development of MetS. RESULTS: Over two years of follow up, 30 of the 352 subjects (8.52%) without MetS at baseline subsequently developed MetS. Triglycerides (TG) had the highest area under the curve (AUC), while diastolic blood pressure had the lowest. In order to increase the prediction power, MetS components were arranged in the ROC model according to their AUC. After adding waist circumference (WC) to TG (model 1), the AUC was significantly higher than for TG alone. Adding other components into the model did not increase the AUC significantly. A risk score cutoff (0.078) was selected for the best predictive power of model 1 (sensitivity of 76.7%, specificity of 63.4%, with AUC of 76.8%). CONCLUSIONS: These results imply that WC and TG are related to the pathophysiologies of MetS, and model 1 could also be used clinically for screening subjects at high risks for MetS.

Original languageEnglish
Pages (from-to)159-164
Number of pages6
JournalSouthern Medical Journal
Issue number2
Publication statusPublished - Feb 2009


  • Metabolic syndrome
  • Risk score
  • Triglycerides
  • Waist circumference

ASJC Scopus subject areas

  • General Medicine


Dive into the research topics of 'Chinese metabolic syndrome risk score'. Together they form a unique fingerprint.

Cite this