Are Autism Rates Still Rising Among American Children? It's Hard To Tell


Stephen Luntz

Freelance Writer

clockJan 3 2018, 10:39 UTC
autism diagnosis

Establishing whether the increase in autism diagnosis reflects a real trend or just changes to criteria and increased awareness is not helped by data that leaves open whether the increase is continuing. Shidlovski/Shutterstock

For years diagnosed rates of Autism Spectrum Disorder (ASD) have skyrocketed in the United States and other wealthy countries. Unfortunately, this trend has inspired fear of many blameless scapegoats, most importantly childhood vaccines. If rates stabilize, it may go some way to deflating those panics, but determining trends is proving harder than might be expected.

Between 2000 and 2010 ASD diagnoses in US children more than doubled from 0.67 percent to 1.47 percent, according to the Autism and Developmental Disabilities Monitoring (ADDM). Without question part of this increase was a result of better diagnosis, with a wider range of people on the lookout for ASD symptoms. Some experts in the field argue increased recognition can account for the entire increase, but others have been searching for environmental factors that would account for part of the rise. Many explanations, of varying levels of credibility, have been offered.


In 2012, diagnosis rates were actually lower at 1.46 percent, than in 2010, but the decrease was far too small to necessarily be meaningful. Unfortunately, we're still waiting for subsequent data to tell us if this was the start of a new trend, or just a minor bump in an upward trend.

Dr Wei Bao of the University of Iowa turned instead to the National Health Interview Survey (NHIS), which asks a random sample of approximately 10,000 households each year about the health of their members. The NHIS is not only more current, it has the advantage of being a truly national survey, while the ADDM is done at a small number of selected sites around the country.

The figure produced by the NHIS is higher than the ADDM, at 2.41 percent over the period 2014-2016. However, in the Journal of the American Medical Association, Bao argues this could easily be the consequence of different survey methods. The NHIS also changed the way it asked questions about autism diagnosis in 2014, so comparisons before and after the change are unreliable.

Looking at the three years we do have, the diagnosis rate was 2.24 percent in 2014, 2.41 percent in 2015 and 2.58 in 2016. That looks like a rising trend to most of us, increasing at a hefty 7 percent a year. However, taking into account the relatively small numbers of diagnosed children (234-240 each year), Bao found the year-to-year difference isn't statistically significant, describing the diagnosis rate as "stable" from 2014-2016 in a media release, but if the same pattern is seen over further years, or in larger samples, things could look different.


Diagnosis rates were higher among non-Hispanic whites than black or Hispanic children, raising the possibility that under-diagnosis is still an issue in at least some cases. Boys were almost three times as likely to be diagnosed with ASD as girls.

  • tag
  • Autism Spectrum Disorder,

  • ASD,

  • statistical significance,

  • signal and noise