Yes, there is an autism epidemic.

I am writing in response to an editorial by Lawrence Scahill, Ph.D., about autism ("Autism Is Not an Epidemic," Guest Editorial, April 2008, p. 24). I am not an epidemiologist or a neuropsychologist and have no formal training in autism statistics, diagnosis, or therapy. I am a physician, a neuroradiologist, and the father of a little girl with autism. My daughter was diagnosed with autism at the age of 2 and is now 6 years old. In spite of my education I had little knowledge of autism until my daughter's diagnosis. Since then I have learned more than I care to know about autism and have met hundreds of families struggling to cope with this illness. Occasionally, I come across articles explaining the explosion in new cases of autism as not being real but rather due to changes in diagnostic criteria or greater diagnostic precision. Because they cannot explain it, they believe it simply cannot be so.

In this very publication in April 2005, a report from Craig Newschaffer, Ph.D., and colleagues of Johns Hopkins University using data from the U.S. Department of Education stated, "The apparent increase in autism disorders reflects an actual increase in prevalence rather than a reclassification of other developmental disorders as autism." They went on to state, "clear, significant increases in the prevalence of autism among those born between 1987 and 1992. The prevalence of autism rose by about 50% every 2 years." Additional statistics from the Department of Education show an 870% increase in autism over the 9 years between 1992-1993 and 2002-2003.

Dr. Scahill states, "It would be very difficult to argue that there is a genetic explanation for the apparent rise in prevalence." However, numerous articles have reported genetic links to autism. Most recently, Dr. Aravinda Chakravarti from the Johns Hopkins University reported in the American Journal of Human Genetics (Am. J. Hum. Genet. 2008;82:160-4) on a gene located on the long arm of chromosome 7 as having a significant association with autism. Dr. Thomas Insel, director of the National Institute of Mental Health, also seems to contradict Dr. Scahill when he says, "autism is highly heritable."

In the United States, children are diagnosed with autism at a rate of 1 in 150. The Government Accountability Office reported the cost of educating children with autism is almost three times that of normal kids. And the number of children with autism receiving special education services has increased by more than 500% in the last decade. Every day 67 children are newly diagnosed with autism, nearly 1 every 20 minutes. More children will be diagnosed with autism this year than with AIDS, diabetes, and cancer combined. Autism costs this nation over $90 billion every year. Unfortunately, this number is expected to double in the next decade. It is about time for health care professionals get their heads out of the sand and understand that autism is a major health care crisis and yes, an epidemic.

Alan Holz, M.D.

Hollywood, Fla.

Dr. Scahill replies:

I completely agree with Dr. Holz that autism is a major public health problem deserving of full attention and commitment of resources. These resources should be directed at the causes and cures of autism and related conditions.

Second, I also agree that genetics plays a central role in the etiology of autism. I apologize if the article gave any other impression. The point that I was trying to make is that genetics is unlikely to explain the apparent rise in autism. Genetic explanations for such a rise in prevalence in a complex disorder such as autism would likely require several generations and could not occur over a single decade.

Because the apparent rise in prevalence cannot be explained by genetics, investigators have looked for an environmental explanation.

Claims that the [measles, mumps, and rubella] vaccine or thimerosal-containing vaccines are the cause of the apparent rise in prevalence have not been supported upon close inspection. For example, two recently published studies show that the rise in autism has continued long after the use of thimerosal was stopped (see citations below). I hasten to add here that these studies were published after the report on the work of Dr. Newshaffer and associates in 2005 mentioned by Dr. Holz. Given the public health importance of autism, it seemed reasonable to step back and consider the possibility that prior estimates of prevalence were simply inaccurate and vulnerable to systematic underestimation. It is this systematic underestimate that I hoped to illustrate in my editorial.

(1.) Fombonne E et al. Pervasive developmental disorders in Montreal, Quebec, Canada: Prevalence and links with immunizations. Pediatrics 2006; 118;139-50.

(2.) Schechter R, Grether JK. Continuing increase in autism reported to California's developmental services system. Arch. Gen. Psychiatry 2008;65:19-24.