Sample Abstract
|
DEVELOPMENT AND VALIDATION
OF A PREDICTION EQUATION FOR BODY FAT PERCENT BASED ON SKINFOLDS FOR QQQ
ADULTS AND ADOLESCENTS 1RRR Division, ABC Board, Singapore; 2XYZ University; 3ZZZ Consultant, Singapore. Objective: The study sought to develop and validate prediction equations for body fat percent from skinfold thickness for use among QQQ. Methodology: Data from DDD adults aged FF to DD years old and UUU adolescents aged 12 to 17 years were collected. Body fat percent in the adults was measured using a 4-compartment model whilst deuterium oxide dilution was used for the adolescents. Biceps, triceps, suprailiac and subscapular skinfold thicknesses were also measured. The adult sample was randomly divided into two groups. In one group, a prediction equation for body fat percent based on the sum of four skinfolds was developed using stepwise multiple regression analyses. Results: Ethnicity did not enter into the model. Age did not enter into the regression equation in females though it made a slight but significant contribution to the prediction equation in males. The formulas were validated in the second adult group, total adult group and adolescent group. In adults, the prediction equations showed good cross-validity for the three ethnic groups (Chinese, Malays and Indians) and the bias was not age-dependent. However, in subjects with higher body fat percent, the formulas tend to underestimate body fat percent. The prediction equation based on the sum of four skinfolds was valid in female adolescents but slightly underestimated body fat percent in adolescent males. The bias was not different across the ethnic groups. Conclusion: It is concluded that the 'adult' prediction equations based on four skinfolds can be used to obtain a valid prediction of body fat percent in adolescents and adults of the three main ethnic groups in Singapore. Correspondence: |