RESUMO
PURPOSE: This study was performed to develop risk prediction model for cardiovascular disease using intima-media thickness (IMT) of carotid artery among the community dwelling elderly. METHODS: Follow-up survey was conducted in 2007 to measure the IMT of carotid artery, which is known as a preclinical marker of atherosclerotic change, among participants at the baseline survey of Hallym Aging Study (HAS) in 2004. They were categorized into two groups; increased IMT > or = 0.9 mm and normal IMT < 0.9 mm. To evaluate which factors are independently related with the increased IMT, multiple logistic regression analysis was done. Based on the regression coefficients of these factors weighed by the magnitude of the effect estimates, we calculated the risk scores for increased IMT for every participants. ROC curve was plotted for the each cutoff point of risk scores and its fitness was tested using Area Under the Curve (AUC). Finally, we calculated risk ratios for the increased IMT according to the level of risk based on the risk scores. RESULTS: Several factors were found as related factors for the increased IMT in the multiple logistic regression: age (beta=0.15), cholesterol (beta=0.01), insulin (beta=0.13), and pulse pressure (beta=0.90) for men, and age (beta=0.08), family history of diabetes mellitus (beta=0.94) and stroke (beta=0.79), alcohol drinking (beta=0.39), and high cholesterol (beta=0.77) for women. We assigned the weighed value for each factors. The average risk scores were 14.48 (range 9.69-18.76) for men and 4.59 (range 2.41-7.48) for women. The Area Under the Curve (AUC) were 0.77 for men and 0.71 for women. We also observed the higher risk of increased IMT as the risk scores increased. CONCLUSION: Based on the results, we expect to predict the level of the risk for the increased IMT, which is preclinical marker for atherosclerotic change, among the elderly.