According to yesterday’s Daily Mail, the ‘pupil obesity crisis’ has forced school uniform sellers to stock sizes ‘usually bought by obese adults’.
The article does not cite comprehensive, time-series data on school uniform sales but we do learn that:
“Uniform shops say demand for extra-large sizes has risen dramatically, with manufacturers providing bigger sizes every year.”
In fact, today’s Sun went a step further stating that:
“Supersize school uniforms have hit the shelves to cope with Britain’s rising childhood obesity epidemic.”
A growth in the sale of extra large clothing might suggest that childhood obesity is on the rise, but we should be wary about drawing such conclusions on this evidence alone.
But are the reports in the Mail and the Sun supported by official statistics?
The prevalence of obesity
The two main sources of information on the prevalence of obesity in England are the National Child Measurement Programme (NCMP) and the Health Survey for England (HSE).
Both measures define childhood obesity as having a BMI greater than the 95th percentile of the UK90 BMI reference.
The NCMP has recorded the prevalence of obesity among children in Reception and Year Six since 2006/7. The graph below (taken from a National Obesity Observatory report) shows the trends across this time (within 95 per cent confidence limits):
We can see that the pattern across the five year period is much clearer in Year Six than Reception with the “overall prevalence for children of this age show[ing] a statistically significant increase of 0.35% per year over the period”.
It is worth noting, however, that there was a “substantial change in both participation and data quality” between 2006/07 and 2007/8 but, even if we exclude 2006/07, we can observe a “statistically significant annual increase of 0.26% per year”.
The NCMP is, however, of limited use for estimating obesity prevalence among all children as it only offers a snapshot of two age groups.
Obesity rates are, however, also recorded by the National Health Survey and published by the NHS Information Centre. The results presented in the graph below show the percentage of boys and girls (aged 2-15) who are considered to be obese; with data taken from the publication of the results of the HSE 2010:
When the data is presented in this way, we can see that although the trend increased to 2004, the trend since then has been much less consistent.
The Information Centre splits the data by age which highlights a downward trend in the percentage of girls aged 11-15 who are obese since 2004:
As with many statistics, it would appear that the reported trend is, at least in part, determined by which age and sex is focused upon.
So the most important point to take from this analysis is that, although we might expect that supply will match demand in the uniforms market, an increase in the clothing sizes stocked in uniform shops is not a reliable way to gauge the prevalence of childhood obesity.
A brief review of some of the most recent evidence on obesity rates demonstrates that obesity measurement is by no means straightforward and the reported trend can often be determined by the way in which the data is presented.