RESUMO
OBJECTIVE: WHO BMI-for-age z score (BAZ) is widely used in epidemiology, yet it does not distinguish body fat-free mass and fat mass which are better indicators of obesity and related risks. The stable isotope dilution techniques (SIDT) are gold standard methods of assessing body composition. Main objective was to assess significant differences in measurement and validity of WHO BMI-for-age classification for defining childhood obesity by comparing with body fatness using SIDT among schoolchildren. DESIGN: A cross-sectional analytical study. A questionnaire, anthropometry and body composition data were used. SPSS was used to analyse data at P < 0·05 at 95 % CI. SETTING: Primary schools in Nairobi City County, Kenya. PARTICIPANTS: One hundred seventy-nine schoolchildren aged 8-11 years were randomly sampled. RESULTS: Prevalence of adiposity by reference SIDT (24·0 %) was significantly higher than that of obesity by BAZ > 2 sd (2·8 %) (Wilcoxon test, P < 0·05). Concordance coefficient between SIDT and BAZ > 2 sd in diagnosing obesity was poor (κ = 0·167). Only 11·6 % of children with excess body fat were correctly diagnosed as obese by BAZ > 2 sd. The use of BAZ > 1 sd for overweight and obesity showed fair concordance coefficient (κ = 0·409, P < 0·001) with 32·5% of children with excess fat positively identified as overweight and obese. CONCLUSION: WHO BMI-for-age cut-off points severely underestimate the prevalence of overweight and obesity compared with body composition assessment by stable isotope dilution techniques. Evidence-informed interventions should be based on more accurate estimates of overweight and obesity than that can be provided by BAZ.
