Correlates, predictors, reference ranges and agreement between percent body fat measured using bioelectric impedance analysis and skinfold-thickness measurements in young nigerian adults

Background: Establishment of population specific normative databases for body composition is important in determining clinically useful cut-points for definition of excessive body fat or obesity for preventive or treatment programmes. This study provides the correlates, predictors, reference ranges and agreement between Percent Body Fat (PBF) measured using Bioelectric Impedance Analysis (BIA) and Skinfold- Thickness Measurements (STM) in young Nigerian adults. Materials and Methods: A total of 531 apparently healthy individuals who were within ages 18 and 30 years were consecutively recruited in this study. BIA and STM (from three anatomical sites) were used to estimate PBF while anthropometric parameters were assessed following standard procedures. PBF computation was based on standard formulae. Socio-demographic characteristics were obtained using a proforma. Data were analysed using descriptive and inferential statistics at 0.05 alpha level. Results: There was significant difference in PBF measured by BIA and STM (20.69±9.97 vs. 13.4±9.37; t=12.28, p=0.001). There was significant correlation between PBF assessed by BIA and STM (r=0.912; p=0.001). The socio demographic and anthropometric variables were significantly correlated with PBF (p <0.05) measured by BIA and STM respectively. The variability of the predictive of models for PBF from age, weight and height was 34.0% (R2 =34.0) and 50.9% (R2 =50.9) for males and females using BIA, while, the variability for PBF assessed using STM was 49.1.0% (R2 =49.1) and 40.4% (R2 =40.4) for males and females. Conclusion: This study provides reference values for PBF using BIA and STM in young Nigerian adults. BIA method compared with STM overestimated PBF, although these two methods correlates strongly in the acceptable and moderately overweight range and not among the lean and obese participants. Age, gender and anthropometric factors were significantly determinants of the PBF. Age, height and weight seems to be moderate predictors of PBF but not without significant errors.