Development of population‐specific anthropometric prediction equations for children in Brazil

Abstract

The double burden of obesity and underweight is a serious public health dilemma in many developing countries. While anthropometric prediction equations (PE) for fat mass (FM) in children exist, most have been developed in cohorts from developed populations, not countries experiencing the “double burden”. Therefore, using data from 145 pre‐pubertal children (9.8 ± 1.3 yrs) from São Paulo, Brazil, we developed a PE for FM using body composition measured by dual energy X‐ray absorptiometry and skinfold and anthropometric measurements from six sites. The new PE was developed using backward linear regression analysis with weight, triceps, height, gender, and age as independent variables from 30 randomly selected children (R2=0.77). When cross‐validated in 125 children, the population‐specific PE met the criteria for validity (i.e. no significant difference between the means and slope =1 with an X‐intercept = 0) (R2=0.80), while standard, child‐specific PE for FM by Slaughter (R2=0.51), Goran (R2=0.67), and Huang (R2=0.33 and R2=0.23) did not meet these criteria. In conclusion, the newly developed, population‐specific PE appears to be more valid for use in children from developing countries than the standard equations. The use of a ‐specific PE can enhance the assessment of body composition of children in countries experiencing the double‐burden of disease.