Bending the child obesity curve

The Robert Wood Johnson Foundation has released a fascinating glimpse into data suggesting that childhood obesity may be declining in several US cities and counties.

2013-07-17 09.35.38 am

The report is preliminary and tracks cohorts over just a few time points across the US, but is consistent with other more population-representative, systematic reports suggesting that the obesity prevalence trajectory may be flattening or even declining in some US populations.

Here’s the RWJ preliminary data:

2013-07-17 09.31.21 am

There are several questions that are intriguing here:

(1) First, if these changes are meaningfully large and sustained, are there clear contrasts in policy we can use as “natural experiments” to analyze what changes might have occurred in these communities to drive the declines as compared to their neighboring cities/counties where the declines didn’t occur, controlling for other confounders? (Or even analyzing within the cohorts to examine why certain populations did/did not experience the declines).

(2) Is there selection bias? In-migration of thinner folks or out-migration of heavier folks may be an artifact of gentrification and other changes to the community that resulted in a different population rather than true declines in prevalence. This has implications for whether these results truly reflect declines in obesity prevalence disparities.

(3) Are the results affected by the hard cut-offs we use for body mass index (BMI), such that marginal reductions in BMI may cause several people to cross from obesity to overweight status, but produce an artifactual large decline in absolute obesity prevalence? In other words, how clinically meaningful are the changes for actual outcomes such as diabetes risk and other co-morbid conditions?

Comments are closed.