Correlation And Principal Component Analysis Of Children'S Cardiovascular Disease Risk Factors: Implications For Risk Assessment

PURPOSE: To examine the relationships among various CVD risk factors in 5th grade children. METHODS: A cross-sectional analysis of 1525 children (856 girls and 669 boys; age: 9-12 years) who participated in the (S)Partners for Health Project from 2010 to 2018 was performed. Fifteen CVD risk factors variables including blood lipids (low-density lipoprotein, high-density lipoprotein, total cholesterol, and triglycerides), blood pressure (systolic and diastolic), anthropometrics (height, weight, body mass index (BMI), % body fat via bioelectric impedance, waist circumference), acanthosis nigricans (a marker of insulin resistance), and cardiorespiratory fitness were analyzed. Descriptive statistics, bivariate Pearson correlations, and principal component analysis were used to determine the relationships among variables. RESULTS: Parallel analysis indicated two components be extracted. For component 1 waist circumference, % body fat, and BMI loaded highest and explained 34% of the total variance. Additionally, cardiorespiratory fitness, acanthosis nigricans, high-density lipoprotein, and triglycerides loaded moderately on component 1 (0.42 and 0.57) and suggest some non-shared variance with these variables. Systolic and diastolic blood pressure loaded predominantly on component 2 and accounted for 17% of the variance. Low-density lipoprotein had low loadings (< 0.31) on each component. Factor loadings were stable across gender. CONCLUSIONS: Among the various CVD risk factors analyzed, the adiposity measures loaded highest on the component that explained the largest proportion of variability in the data, reinforcing the importance measures and indices of adiposity in CVD risk assessment. In addition, measures of blood pressure loaded highest on the second principal component, suggesting their relative independence to the other CVD risk factors measures. The results on factors loading provide valuable insight and rationale for choosing which CVD risk factors to include if assessment time, cost, and participant burden are a concern. For example, various indicators such as BMI due to the factor loading, ease, and low cost of this measurement.