A Mirage of Child Health Improvement in Arizona

Fasten your statistical seatbelts & prepare for a "wonky" longer-than-normal post:

New National Survey of Children’s Health (NSCH) data claim an improvement in child obesity in Arizona over the last five years--to the extent that AZ supposedly now has the 2nd lowest child obesity for 10-17 year olds in the USA !?  Per the NSCH-based rankings, this implies that Arizona has less child obesity & much better trends than even perennial health leaders Colorado or Utah--in spite of those states having much less poverty & much lower adult obesity than Arizona.  Few states except perhaps North Dakota(!?) & Nebraska(!?) show as much of a decline over the last five years (2016-2020) among children, according to these NSCH data.

These parent-self-reported statistics contradict the Youth Risk Behavior Survey (YRBS) data, which are self-reported by high school students themselves--and which do not show an improvement in teen obesity in Arizona (or in Nebraska or North Dakota).  

The NSCH data also appear to contradict long-term trends based on NHANES data, whic use actual height & weight measurements conducted by medical professionals.  NHANES has clearly documented both worsening long-term child obesity and adult obesity nationwide. In addition, in 2019 the well-respected New England Journal of Medicine published dire adult obesity projections for the coming decade with "high predictive accuracy", including forecasting rapid increases in Arizona.

Unless one can explain what accounts for the discrepancy between one set of data and a broad range of other contraindicating statistics, one should be very careful with this type of "outlier" data. In this case, I am not aware of any major reason why obesity in Arizona, North Dakota or Nebraska would have declined substantially on a statewide level in recent years relative to other states, in spite of a few possible factors:

       - Fortunately, healthier school meals introduced starting in about 2013 under the Healthy Hunger-Free Kids Act (HHFKA) appear to have reduced obesity nationwide among children in poverty (also using NSCH data, 2003-18)--but this trend should have affected all states, not just AZ, ND & NE.  I am not aware of & highly doubt that Arizona, North Dakota and Nebraska had higher rates of implementation of the healthier post-2012 HHFKA/USDA standards.  We do have a larger low-income population in Arizona than some other states. But many other states had large low-income populations whose child obesity should also have disproportionately improved under HHFKA--yet their child obesity rates did not improve at the state level (even using NSCH data).  In addition, ND and NE do not have large low-income populations compared to other states, so HHFKA implementation in these states would have been unlikely to move the needle on overall child obesity to this extent.  

      - The AZ recess law began implementation in 2018-19 for grades K-3, then at the latest 2019-20 for grades 4-5 (for those schools that complied). This should ultimately increase physical activity by 50% for about 250,000 AZ children. But those effects would only show at the end of the 2016-20 range, and only for younger children--who were mostly below the 10-17 age range of these data.  Also, since that law was focused on activity not nutrition, I would expect it to improve health but not necessarily reduce obesity on its own.  Indeed, so many factors contribute to child obesity that usually many things need to be done in parallel to reduce it--hence my doubts about these NSCH data. This is why Healthy Future US focuses on a range of plan & partner, policy, program, and payment-with-payback improvements.

- We have two excellent fresh produce programs in Arizona, which have helped improve healthy nutrition among lower-income families. Farm Express is a mobile service bringing fruit & veggies to neighborhoods in "food deserts." Double Up Food Bucks provides coupons to people in the SNAP progam ("food stamps"), which double the value of their spending on fruit & veggies at supermarkets, outdoor food markets, and other places. The number of locations and the scale of usage for both programs is growing, thanks to funders such as Mercy Care--though their scale & uptake do not seem large enough yet to move the needle on a statewide basis.

What are some possible explanations for the 2016-20 NSCH data showing a reduction in AZ child obesity?:

• As noted above, these data are based on self-reported child height and weight data from parents.  In addition, parents are selected from a database based on social security numbers, and for which IRS 1040 data provide the “most significant information." Given the high number of undocumented parents in Arizona, with a high prevalence of child obesity among their children, this could unintentionally appear to reduce the incidence of child obesity among NSCH respondents in AZ--children of undocumented parents may not show proportionately in the survey sampling & responses. 

• In addition, undocumented families’ numbers declined dramatically over the last 12 or so years, due to both the Great Recession and anti-immigrant laws and practices in Arizona. So even if some were surveyed, their numbers have been much lower recently than in earlier decades.

• At the same time, there have been a high number of relocations to Arizona during 2016-20.  Families coming to Arizona and responding to NSCH may have relatively low incidence of obesity upon relocation relative to families already living in the state, and very likely have lower obesity than undocumented families who have left the state over the last 12 years.  This would distort the picture on what is happening to the long-term population of people who were already in the state pre-Great-Recession and who are still here.

In brief, undocumented immigrant "invisibility" and emigration at the same time as higher-income white families were immigrating probably artificially inflated apparent child obesity improvements.

Indeed, the full decade data (Youth in Arizona tab) imply that the decline in AZ child obesity captured by NSCH surveys began during the Great Recession & after, not over the last five years. So unfortunately, the recent data do not indicate that Arizona is suddenly successfully reducing the obesity of its long-term population here for decades.

I am concerned that these questionable "improvement" data could get excessive publicity & take some pressure off of the need to dramatically improve children’s physical health--which all of the other data indicate is at an appalling level.  Indeed, the severity of obesity and related early prediabetes is actually worsening among teens & young adults, with extremely serious health consequences ever-earlier in life.