Abstract
BACKGROUND: Adult overweight and obesity, in addition to the intake of saturated fat and total serum cholesterol must be monitored as biological risk factors for non-communicable diseases (NCDs). Bioelectric impedance analysis (BIA) provides data on body fat for use in epidemiological settings. However, optimized equations should be used to calculate percentage body fat (%BF). The purpose of this study was to assess the differences between %BF calculated using different published BIA equations and %BF measured by BIA in young South African adults. DESIGN AND METHODS: In this observational study, differences in calculated %BF were assessed, with different BIA equations retrieved from the literature used in 1128 healthy young adults aged 20-30 years. The %BF (measured by BIA) was compared between equations, between Black and White men and women, respectively. RESULTS: The results showed statistically significant differences in the %BF calculated from published BIA equations when used in young South African adults (χ² = 946, χ² = 2528, χ² = 2088, respectively, p < 0.0001). In Black and White men and women, respectively, %BF levels were significantly higher when calculated by equations, than when measured by BIA (p < 0.0001). CONCLUSION: There seem to be large discrepancies in estimating %BF by BIA equations and these values cannot be used interchangeably for young South African adults. A South African age, ethnicity and sex-specific BIA equation needs to be developed to accurately estimate %BF in young South African adults.