J Korean Acad Adult Nurs.
1998 Apr;10(1):48-60.
Canonical Correlation Analysis among Risk factors of arteriosclerosis
Abstract
-
The purpose of this study was to examine the relationship among variables which are categorized by two groups : the first group consists of obesity, serum lipids, and blood pressure which were measured by physical index or physiological index, and the second group consists of demographic variables and variables related to lifestyle that is known to influence the first group. The canonical correlation analysis was conducted with the data collected from 400 male clients who visited one university hospital located in Inchon in a period, from May 1996 to December 1996, for physical examination. According to the results, blood pressure and HDL of the first group were shown to have strong relationship with age, alcohol, smoking, exercise of the second group. And total cholesterol, triglyceride, systolic, BMI of the first group were shown to have strong relationship with income status, education, exercise of the second group. And all of the serum lipids (total cholesterol, triglyceride, HDL), BMI, and diastolic were shown to have relationship with education, smoking, alcohol. It should be noted that fat rate, which was one of the indices for obesity, was not significant in any of the canonical variates. From the comparative study results on which combination of indices of obesity, serum lipids and blood pressure was related to either alcohol or smoking more, it can be seem that smoking was more highly related to two blood pressure indices such as systolic and diastolic, and that alcohol was highly related to serum lipid indices such as HDL, total cholesterol, triglyceride. Also investigation on what differences exist between the combination of variables showing high relationship to systolic and the combination of variables showing high relationship to diastolic was conducted. Systolic turned out to have relationship with total cholesterol, triglyceride, obesity (fat rate), income status, education, exercise. And diastolic was related to HDL-c, total cholesterol, tryglyceride, obesity (BMI), education, smoking, alcohol. From this results, it could be seen that the combination of variables which have high relationship to systolic and diastolic have different patterns. Lastly, the investigation on what relationships exist between serum lipids and demographic variables/lifestyle variables was conducted. It turned out that HDL-c had high relationship with age, education, income status, alcohol, and that all of serum lipids (total cholesterol, triglyceride, HDL) were shown to have high relationship with income status, education, age, exercise. From these results, it could be seen that the demographic variables were more influential both on HDL and on all of serum lipids than lifestyle variables were.