A comparison of body mass index and percent body fat as predictors of cardiovascular risk factors

Background: Cardiovascular disease (CVD) is an important cause of global mortality and morbidity. Body mass index (BMI) is the measure of adiposity that is used most frequently in CVD risk algorithms.

Aims: We aimed to assess the relationship between several CVD risk factors (RFs) and percent body fat (PBF), and to compare the predictive values obtained using PBF for these cardiovascular RFs with the values obtained using BMI. The CVD RFs included, hypertension (HTN), diabetes mellitus (DM) and the presence of dyslipidemia (DLP).

Methods and materials: The data were derived from the MASHAD study, a cohort study of 9704 volunteers, aged 35-65 years and living in the city of Mashhad. Based on BMI and PBF values, subjects were classified into 4 groups; group 1 (low or normal BMI and PBF, N = 1670), group 2 (low or normal BMI but high PBF, N = 992), group 3 (high BMI and low or normal PBF, N = 837), and group 4 (high BMI and PBF, N = 6245). Chi-square, covariance and logistic regression were used to analyze the data at a significance level of 0.05.

Results: There was an increasing trend from group 1 to group 4 for the mean values of all CVD RFs and their prevalence. There were significant differences in the frequency of a low HDL-C, this was substantially higher in Group 3 (38.6% in Group 3 versus 12.2% in Group 2); the frequency of a high serum TG (24% in Group 3 versus 9.9% in Group 2) and the frequency of dyslipidemia overall (56.2% in Group 3 and 28.8% in Group 2) (P-value<0.001 for all comparisons). The frequency of hypertension (22.9% in Group 3 versus 16.2% in Group 2) and IFG (8.5% in Group 3 versus 5.0% in Group 2) were also substantially higher in Group 3 compared to Group 2 (P-value<0.001 for both comparisons). All the mean values for the RFs were higher in group 3 from group 2 except HDL-C. When Group 1 was used as a reference and calculated OR of any RF for any group 2-4 rather than group 1, OR for all RF in group 3 was higher from group 2.

Conclusion: The differences in frequency, means and OR of RFs between Groups 2 and 3 showed a differential impact of a high BMI or high PBF. Compared to PBF, BMI may be a better predictor for several RFs for CVD.