Correspondence (letter to the editor): Percentiles May Be Confusing

The authors have studied pediatric overweight—one of the most important current topics in health prevention—and have shown the factors that influence this development very convincingly. The influences of mother tongue, exercise, dietary behavior, and media consumption that were analyzed in the study were particularly interesting. To investigate these influential factors in a cross sectional comparison, the chosen definition of overweight and obesity based on the percentile method is acceptable (above the 90th and 97th age and sex specific percentiles), and differences by national origin become obvious.

In a longitudinal study, however, the percentile method may cause confusion. Even if each new starter at school in the study period 2005/6 (theoretically) weighs 10 kg more than their peers in 2003/4, the prevalence of overweight according to the percentile method will not change substantially—in spite of the children’s massive weight gain. The percentile method is problematic for health related risk factors—for example, cardiovascular risk factors. In Germany in persons older than 60 years, the prevalence of hypertension is currently about 60%. On the basis of the threshold value of 140/90 mm Hg, which takes into account cardiovascular risk, arterial hypertension can be assumed above the 40th percentile of blood pressure distribution.

For overweight in children, threshold values should be defined as for blood pressure, which takes into account health risk and not statistical distribution. It is possible that for this risk factor, the risk may be present for notably lower percentiles. In any case, the absolute values for body weight or body mass index should be calculated above which the health risk increases.