Table 4. Logistic regression analysis of variables independently associated with cardiovascular risk factors.
| Model | χ2 value | P value | Odds ratio (95%CI) |
|---|---|---|---|
| Model 1† | |||
| BMI (kg/m2) | 0.027 | 0.870 | 1.01 (0.91-1.12) |
| PBF (%) | 4.850 | 0.028 | 1.04 (1.00-1.07) |
| BMI*PBF | 0.003 | 0.955 | 1.00 (0.90-1.13) |
| Model 2‡ | |||
| BMI (kg/m2) | 3.140 | 0.076 | 1.09 (0.99-1.20) |
| PBF (%) | 10.862 | 0.001 | 1.03 (1.01-1.04) |
| BMI*PBF | 0.103 | 0.748 | 0.98 (0.89-1.09) |
| Model 3§ | |||
| BMI (kg/m2) | 0.433 | 0.511 | 0.97 (0.87-1.07) |
| PBF (%) | 10.147 | 0.001 | 1.05 (1.02-1.09) |
| BMI*PBF | 1.136 | 0.287 | 0.95 (0.85-1.05) |
| Model 4¶ | |||
| BMI (kg/m2) | 3.008 | 0.083 | 1.09 (0.99-1.20) |
| PBF (%) | 4.305 | 0.038 | 1.03 (1.00-1.07) |
| BMI*PBF | 0.119 | 0.730 | 0.98 (0.89-1.09) |
The inclusion criterion for variables was 0.05 and the exclusion criterion was 0.10.
The probability of having one or more risk factors was set as a dependent variable, and the confounding factors such as gender, age, waist-hip ratio (WHR), body mass index (BMI), percent body fat (PBF), BMI*PBF, smoking, drinking, exercise, and family history were set as independent variables. BMI*PBF represented the interaction between PBF and BMI in terms of association with cardiovascular risk factors.
The probability of having hypertension was set as a dependent variable, and gender, age, WHR, BMI, PBF, BMI*PBF, smoking, drinking, exercise, family history, dyslipidemia, and hyperglycemia were set as independent variables.
The probability of having dyslipidemia was set as a dependent variable, and gender, age, WHR, BMI, PBF, BMI*PBF, smoking, drinking, exercise, family history, hypertension and hyperglycemia were set as independent variables.
The probability of having hyperglycemia was set as a dependent variable, and gender, age, WHR, BMI, PBF, BMI*PBF, smoking, drinking, exercise, family history, hypertension, and dyslipidemia were set as independent variables. 95%CI = 95% confidence interval.