Cherry Picking

Cherry Picking Childhood Obesity in Leeds

Late in April, folks in Leeds created a bit of a stir. They told the world they had achieved a “notable decrease” in childhood obesity. Better yet, some of the biggest gains were among the most disadvantaged children entering primary school. Headlines endorsed the program deployed in Leeds. But now, two months later, skeptics are speaking up in the BMJ. They suggest this may be yet another case of cherry picking data on childhood obesity to prove a point.

Picking Data to Prove Your Point

The latest concerns come from Tom Frost, who is a senior public health intelligence analyst in the UK. Of the data on childhood obesity in Leeds, he writes:

The data are easy to cherry pick from the National Child Measurement Programme if you choose a peak and a trough from a longer series of data.

None of this means that HENRY (Health, Exercise, Nutrition for the Really Young) doesn’t work and isn’t part of a “solution” to childhood obesity. But the data don’t prove this, rather showing that Leeds is still struggling along with the rest of the country (urban and rural) to change the direction of travel on childhood obesity.

Along the same lines, Eugene Milne writes that the publication from the folks in Leeds omits more recent data showing a rise childhood obesity. However worthy the HENRY program might be, the study that started all of this does not provide solid evidence that it works. And he went on to say:

It has fuelled the individual, behavioural choice paradigm of childhood obesity, against which there is a mass of evidence, and that strikes me as irresponsible epidemiology.

The concerns from Frost and Milne come in addition to a BMJ analysis by Jacqui Thornton. She notes that many in public health outside of Leeds are urging caution. It’s not at all clear that the HENRY program caused childhood obesity to drop in Leeds, they say. Experts point to the need for a combination of population-wide measures.

An Epidemic of Cherry Picking

Obesity is a frustrating problem for public health advocates. If we wait for evidence for effects on health outcomes, we will have no policies in place, they tell us. In other words, we must do something now. And then later we can figure out if it works.

We don’t disagree with implementing programs and then rigorously analyzing them for effectiveness. But the problem comes with less than objective analyses. Over and over again, we see enthusiastic advocates claiming success based on data that are inadequate. They say childhood obesity is “plummeting” based on a single, spurious data point. Now data from preschoolers on food assistance is supporting claims that obesity is down for all preschoolers. The list of such misleading claims is long.

Cherry picking data to claim success does not serve public health well. It only serves to promote confirmation bias. Worse, it fools us into continuing with policies that will never work. We need more curiosity about what really works, combined with a passion for objectivity.

Click here for more on the Leeds study. For further perspective on whole cities working to reduce childhood obesity, click here.

Cherry Picking, photograph © Jessica Lucia / flickr

Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.


June 24, 2019