分析人體生理變項和疾病之關係，常有下列的困擾；在統計上，這些生理變項彼此間的相關極強，即共線性相當嚴重，同時這些變項彼此間實際生理意義重疊性極高。本研究嘗試使用因素分析的方法，儘可能不損失各個生理變項之解釋力的情況下，解決分析時變項之間相關太強及意義重複的困難， 來觀察某些心血管疾病危險因子特性與心肌缺氧指標(ST 段改變 ) 之年齡趨勢的關係。 本研究使用在竹東朴子進行之「臺灣地區重要體液傳染病及心臟血管疾病危險因子長期追蹤研究」第二循環之資料，進行各因素與 ST 段改變幅度年齡趨勢之關係。依變項為心電圖之 ST 段 (St segment) 幅度改變，自變項為年齡、七項凝血因子、四項血清生化項目、血壓、肥胖指標及血脂前蛋白等 22 個變項，以「主成份分析法」執行因素分析， 男性得到 8 個因子，分別為肥胖因子、凝血因子、膽固醇因子、血壓因子、凝血時間因子、insulin 相關因子、血脂蛋白 -A 因子及中段軀體肥胖因子。女性得到 7 個因子、分別為肥胖因子、 凝血因子、 膽固醇因子、 血壓因子、凝血時間因子、insulin 相關及中段軀體肥胖因子與血脂蛋白 -A 因子。經過因素分析後，各因素之間的相關遠比單獨各生理變項的相關要低，且解釋力仍遠顯著，顯示因素分析解決了變項之間相關太強及意義重複的困難。 ST 段隨著年齡的增加而下降，男性在肥胖、高的三酸甘油酯、血糖、尿酸及低的血脂前蛋白 -A 濃度的特性下，ST 段下降的傾向會增加； 女性則在肥胖、高的三酸甘油酯、血糖、尿酸、血壓、及低的血脂前蛋白 -A 濃度的特性下， ST 段下降的危險會上升。另肥胖及血脂前蛋白兩特性之高、中、低三組在方向性上有很好的解釋力。本研究發現心肌缺氧的老化現象 (ST 段改變量 ) 之年齡趨勢，在不同心臟血管危險因子有類似的情形，也就是已知的心血管疾病危險因子無法完全解釋 ST 段改變量隨年齡下降之情形。

　　　　　Problems exist in simultaneously relatingphysiological variables such as obesity indices, blood pressure and biochemistryindices to disease. Correlation of these variables is very strong. Collinearityis difficult to handle. Physiological meanings of these variables largelyoverlap. We discuss these difficulties in this paper. The purpose of this studyis to describe the age trend of ST segment depression and to examine whethersuch trends differ in subjects with variables cardiovasculav disease (CVD) riskprofiles. Using data from CVD facts, we used factor analysis to organize thecorrelated CVD risk factors. There were 22 CVD risk factors selected for thisstudy including body mass index (BMI), skinfold measure, diastolic and systolicblood pressure, seven hemostatic parameters, and apolipoproteins A-1 and B-100.Eight facotrs were studied in men. These were obesity factor, coagulationfactor, cholesterol factor, blood pressure factor, coagulation time factor,apolipoprotein-A1 factor, central-obesity factor, and insulin-related factor. Infemales seven factors were studied including the obesity factor, coagulationfactor, cholesterol factor, blood pressure factor, coagulation time factor,apolipoprotein-A1 factor, and central- obesity-and-insulin-rclated factor. Thecorrelation coefficients among these factors were lower than those amongoriginal CVD risk variables. Regression analysis revealed that age, sex,obesity, blood pressure, and apolipoprotein-A1 were predictors of ST segmentchange. The magnitude of ST segment depression decreased with age. Thesedecreasing trends showed the same shape between the obese group and the leangroup. But there was not no difference beteen the group with high levels ofapo-A and the group with low levels. These results indicate that the agingefffect can not be completely explained by CVD risk factor studies.