Cross-validation of pedometer-determined cut-points for healthy weight in British children from white and South Asian backgrounds

Background: Evidence-based pedometer cut-points for health have not been sufficiently examined in the context of ethnicity. Objective: To (1) evaluate previously described steps/day cut-points in a sample of White and South Asian British primary school children and (2) use ROC analysis to generate alternative, ethnic specific, steps/day cut-offs for children. Methods and procedures: Height, body mass and pedometer determined physical activity were assessed in 763 British children (357 boys and 406 girls) from White (n = 593) and South Asian (n = 170) ethnic groups, aged 8-11 years. Results: The Vincent and Pangrazi cut-points significantly predicted BMI in white (p = 0.006, Adjusted R(2 )= 0.08) and South Asian children (p = 0.039, Adjusted R(2 )= 0.078). The Tudor-Locke et al. cut-points significantly predicted BMI in White children (p = 0.0001, Adjusted R(2 )= 0.079) but not South Asian children (p < 0.05). ROC analysis indicated significant alternative cut-points in White and South Asian boys and girls (all p = 0.04 or better, Adjusted R(2 )= 0.091 for White and 0.09 for South Asian children). Subsequent cut-points associated with healthy weight, when translated to steps/day were 13 625 for White boys, 13 135 for White girls, 10 897 for South Asian boys and 10 161 for South Asian girls. Conclusions: Previously published steps/day cut-points for healthy weight may not account for known ethnic variation in physical activity between White and South Asian children in the UK. Alternative, ethnic-specific, cut-points may be better placed to distinguish British children based on pedometer-determined physical activity.