A longitudinal study of childhood physical activity at the Government-recommended level and obesity-related health outcomes questions the value of UK and US government guidelines and the use of BMI as the outcome measure.

The study carried out by the EarlyBird Diabetes Research Centre at the Peninsula Medical School in Plymouth aimed to determine the extent to which physical activity at the Government-recommended intensity is associated with change in body mass/fat and metabolic health in pre-pubertal children.

In this non-intervention longitudinal study of 113 boys and 99 girls (born 1995-96) recruited from 54 local schools, physical activity (MTI accelerometers), changes in body mass (raw and age/gender-standardized BMI), fatness (skin-fold thickness and waist circumference) and metabolic status (insulin resistance, triglycerides, cholesterol/HDL ratio and blood pressure - separately and combined as a composite metabolic score) were measured on four annual occasions (5, 6, 7 and 8y).

Results showed that mean physical activity did not change over time in either sex. Averaging the seven-day recordnigs from four time-points rather than one increased the reliability of characterising a child's activity from 71% to 90%. Some 42% of boys and 11% of girls met the guideline.

There were no associations between physical activity and changes in any measure of body mass or fatness over time in either sex (e.g. BMI-SDS: r=-0.02, p=0.76). However, there was a small-to-moderate inverse association between physical activity and change in composite metabolic score (r=-0.19, p