By Dan Hollenbeck
The CDC has two ongoing projects to study autism prevalence: ADDM Network (1) and the MADDSP (2).
In 2003 the MADDSP program published a study pegging autism prevalence at 1 in 250, based on data recorded in 1996 (3). A short four years later, on Feb 7, 2007 the ADDM Network published two studies indicating autism prevalence at 1 in 150, using data from 2000 and 2002 (4, 5). The second of these studies showed that several states showed a statistically significant growth of autism prevalence between 2000 and 2002 (5). Therefore, the current CDC autism prevalence of 1 in 150 is based on six-year old data.
|1 in 250||2003||1996||MADDSP|
|1 in 150||2007||2000 and 2002||ADDM Network and MADDSP|
In addition to publishing the two studies on Feb 7, 2007 the CDC updated their website in order to document ongoing CDC Autism Activities (6). There is a paragraph in this amendment that details future autism prevalence studies the CDC will publish. It states:
"CDC will soon publish the following:
An update from MADDSP on the prevalence of ASDs over time in metropolitan Atlanta.
Updated reports from the ADDM Network on the prevalence of ASDs in multiple areas of the United States in 2004 and 2006."
This begs the question, why has the CDC not published these reports and data yet?
Allow me to comment on these as-yet unpublished reports separately.
1) An update from MADDSP on the prevalence of ASDs over time in metropolitan Atlanta:
To clarify, the MADDSP program is the CDC. The data used in the program are recorded by the CDC itself. Therefore, we rightfully expect the data to be of excellent quality and consistent methodology. The MADDSP has autism prevalence data recorded in 1996, 2000, 2002, (possibly) 2004, and 2006, which is a ten-year time-frame to determine any changes in autism prevalence; a wealth of information by any measure.
As of April, 2008, MADDSP has publicly released data only from 1996, 2000 and 2002 via the published CDC studies. Using data points taken from these studies, we can track the autism prevalence in Atlanta over 1996, 2000 and 2002, a six-year window; from this, we calculate the percentage change of autism prevalence in Atlanta to be +62%.
for 8 year olds
|2000||6.5||5.8 - 7.3|
Is it possible that the reason the CDC has not published this study is that the CDC’s own data clearly shows autism prevalence increased in Atlanta by 62% from 1996 to 2002?
What about the degree of cognitive impairment by IQ score (IQ ≤ 70)? Three CDC studies provide us with cognitive impairment data from the Atlanta autism population:
|Males with autism|
IQ ≤ 70
The above data from the three CDC studies clearly show that the cognitive impairment of children with autism is not becoming less severe over time. The data indicate that the level of impairment for girls is getting worse, and is relatively stable for boys. This contradicts the CDC’s claim that less impaired children are receiving a diagnosis of autism recently than in previous years. While the 1996 data cover a wider age-range, the CDC has the complete 1996 data-set, and could separate it by age to indicate particular ages for the three study years.
It is not unreasonable to hypothesize that the CDC has not published this study because their own data show the cognitive impairment of girls with autism is getting worse, and is stable for boys, from 1996 to 2002.
2) Updated reports from the ADDM Network on the prevalence of ASDs in multiple areas of the United States in 2004 and 2006:
It is unclear if either the ADDM Network or the MADDSP program have collected autism prevalence data for 2004. While part of the CDC’s website indicates that data was recorded in 2004, the ADDM Network and MADDSP pages make no mention of any for that year.
The table below summarizes the autism prevalence data the CDC has amassed, and their reference by the CDC projects:
|2010||to be done||10 sites|
Putting the issue aside about whether the ADDM Network sites recorded autism prevalence data in 2004, the project most certainly recorded data in 2006. Therefore, the ADDM Network has six years of autism prevalence data to analyze and publish.
The bottom line is that the MADDSP Project (i.e., the CDC) has data recorded over a ten-year time period in Atlanta and the ADDM Network sites have data recorded over a six-year window. As things stand, the Federal Government holds itself captive to the fantasy that there is such a thing as a galloping genetic epidemic called autism-- at the same time that it is seemingly content to sit on information that might give great insight into whether or not the cause of autism has an environmental component.
The answer to that question would be priceless.
I hope the CDC will take note, and utilize their considerable resources to publish their own analysis of this data. Whatever the cost in present embarrassment, it will be a blessing for all future Federal agencies.
Dan Hollenbeck is the author of the "Public Schools Autism Prevalence Reports Series" that is used by autism advocates, researchers and educators across the country. (7) He is the Director of Information Technology for Thoughtful House Center for Children and Director of FightingAutism. (8) He has previously served on the Board of Directors, as well as the Research Committee, for SafeMinds, an influential not-for profit autism organization. (9) As an autism advocate and “tech geek”, he is very interested in applying technology in support of autism research, education, treatment, and advocacy.