Social influences on the rise in autism prevalence

How much does “awareness” affect the increase in autism prevalence? Awareness has many aspects. News reports. Getting doctors to look and test for autism. One way is when someone in your neighborhood has an autistic kid. Does that have a measurable effect? The answer is yes.

In their paper, Social Influence and the Autism Epidemic, Ka?Yuet Liu, Marissa King, and Peter S. Bearman of Columbia University look into this using data from the California Department of Social Services (CDDS).

You may recall that I recently blogged on a paper where Marissa King and Prof. Bearman recently took a look at how diagnostic changes have affected the autism counts in the CDDS. The paper is Diagnostic change and the increased prevalence of autism.

Luckily, Steven Novella has already discussed this paper in detail in Social Factors in Autism Diagnosis. Also, the Autism Natural Variation blog has discussed this and extrapolated the current trends in The Administrative Prevalence of Autism is a Bass Distribution.

All this intro aside, what did Liu, King and Bearman find?

Despite a plethora of studies, we do not know why autism incidence has increased rapidly over the past two decades. Using California data, this study shows that children living very close to a child previously diagnosed with autism are more likely to be diagnosed with autism. An underlying social influence mechanism involving information diffusion drives this result, contributing to 16% of the increase in prevalence over 2000–2005. We eliminate competing explanations (i.e., residential sorting, environmental toxicants, and viral transmission) through seven tests and show that information diffusion simultaneously contributed to the increased prevalence, spatial clustering, and decreasing age of diagnosis.

In other words, if you live in California and you live near a family with an autistic child, you are more likely to obtain services from the CDDS for autism for your own child and be included in the CDDS count.

I’ve been meaning to blog this for a while. Prof. Bearman discussed this in a talk he gave at NIH, Understanding the Increased Prevalence of Autism. The link will take you to a page where you can watch the lecture. Sorry, I can’t figure out to embed the talk.

The talk was very interesting. Prof. Bearman discusses the autism cluster they found. He also discusses this social awareness effect.

A few notes from the lecture:

Prof. Bearman can account for much of the increase in the CDDS count. Four contributions he found are:

with diagnostic change–about 27% of the increase
advanced parental age–about 11% of the increase
spatial structure–about 2-3% of the increase
social mechanisms–about 16% of the increase

About 40-50% of the increase is still unaccounted for in his team’s analyses.

For those who are wondering:his team has not ignored vaccines. He doesn’t discuss them much,

One thing Prof. Bearman did find: vaccination rates go down in areas where autism organizations exist. For every autism organization in a given zip code, the exemption rate for personal beliefs increases by 24%.

One could ask whether this is just a sign that higher autism rates are the factors? In other words, is the presence of an autism organization a proxy for autism rate? The answer is no. The vaccine exemption rate does *not* depend on autism prevalence, just the presence of autism organizations.

In other words, autism doesn’t scare people out of vaccinating. Autism organizations do that.

Prof. Bearman’s group has two more papers I await: “Autism and the Hispanic Puzzle” and “Autism and Vaccines, revisited”. At the time of the lecture, the vaccine paper was “under review”.

The CDDS data are far from the “Gold Standard” dataset that some people have claimed. However, it is a valuable dataset and it is good to see a group like Prof. Bearman’s take on a serious, detailed look at it. If there is a real increase in the autism prevalence, it will take teams studies like this to find it.