Interverbal: Reviews of Autism Statements and Research
A critical look at science in the autism world
Saturday, October 22, 2005
Reflecting on the Real Epidemic
I just took a break from the AWARES online conference http://www.awares.org/conferences/. I was thinking about some of the issues raised on the conference and I began to reflect on the Measles Mumps and Rubella vaccination’s theoretical link to autism. Specifically, I was thinking where we have been and where we are going. A lot has changed since the Wakefield, Murch, Anthony, Linnell, Casson, Malik, Berelowitz, Dhillon, Thomson, Harvey, Valentine, Davies, and Walker-Smith, (1998) research that began this issue.
I began to wonder if the scientific community (or anyone really) has the responsibility to speak up when they see a potential problem. So, do we?
I would answer “yes”. There are at least two reasons. The first is dedication to validity. In the absence of well controlled studies we have difficulty seeing how true a given proof is. And when there is no peer review it is hard for knowledgeable others to point out errors which could compromise our research.
The second reason is ethics. There is a guilt which comes, not from committing some transgression, but for failing to do something right. For scientists, this guilt tends to have a way of being felt when someone gets hurt from pseudo-science. I don’t wish to see someone hurt because I failed to do something.
This means I am free to keep an open mind and give other researchers a fair chance to show the merit of their work, but my mind doesn’t have to be so open that I accept theory in the absence of full experimental research.
Maybe it is especially our duty to speak up in treatments that have already hurt someone. If that is so, how have we fared?
The MMR debate has been heated at times; Wakefield et al. (1998) sparked a robust response from other members of the medical community, to say the least. The Lancet has had 26 replies/rejoinders on the issue since that time. Some of the criticisms were fairly serious and Lancet editors responded by expressing some regret that they were not aware of some of the problems in the study (Horton, 2004). The editors mention problems as diverse as threats to internal validity, ethics, and conflicts of interest.
Eventually, most the authors entered a retraction of their research Murch, Anthony, Casson, Malik, Berelowitz, Dhillon, Thomson, Valentine, Davies, and Walker-Smith (2004).
Since a considerable amount of time has been spent on what went wrong in this study, I would rather mention what went right. Beyond a doubt the fact that this research was submitted for peer review (and was subsequently taken to task by others) showed responsibility on the parts of the authors. The retraction also indicates responsibility on the part of some of the authors.
Anyone can be right, but it takes some integrity to admit to being incorrect. I have been told that part of being a scientist is looking stupid sometimes and that if one can not look stupid, one is not doing science. I don’t like the phrasing, but I agree. Looking stupid is emotionally difficult; I don’t envy Murch et al. (2004), but that is part of being a scientist; you agree that you were incorrect and why, and hope that you didn’t hurt anyone in the process, then you go home and laugh or cry a little about it.
Any time I hear others wonder about the necessity of peer review I simply think about Murch et al. (2004) and I am suddenly grateful for it.
I wouldn’t say this is a happy ending though. The scare produced by this research and subsequent promotion dropped the MMR vaccine usage in the United Kingdom. Measles cases rose, and the proper deployment of this vaccine is only now beginning to recover. This research is still considered valid in some portions of society on both sides of the Atlantic. In Montreal Canada, an autism society president claimed measles never killed anyone (Dawson, 2003). This is quite incorrect. The death rate for measles in developing countries is near 25% and while measles is rare in developed countries, it still kills 1-2 per 1000 cases (http://www.cdc.gov/nip/publications/pink/meas.pdf#search='cdc%20measles%20death%20rate'). There was an undisputed US measles epidemic in 1989-1991, mostly among minority pre-school aged children who were not vaccinated. 123 deaths resulted (with the majority being children).
Dr. Wakefield, who was the lead author of the MMR study and continues to support it as valid, will face an inquiry this summer as to whether he is fit to practice medicine.
I wonder sometimes if future epidemics of infectious disease will be triggered by a current epidemic of irresponsibility.
Notes
The phrase "epidemic of irresponsibility" is taken from the subtitle of Dawson (2003) which is cited just below.
Horton, R., (2004). A statement by the editors of the Lancet. Lancet, 363, 820-1.
Murch, S., Anthony, A., Casson, D., Malik, M., Berelowitz, M., Dhillon, A., Thomson, P., Valentine, A., Davies, S., & Walker-Smith, J. (2004). Retraction of interpretation, 363, 750.
Wakefield, A., Murch, S., Anthony, A., Linnell, J., Casson, D., Malik, M., Berelowitz, M., Dhillon, A., Thomson, M., Harvey, P., Valentine, A., Davies, S., & Walker-Smith, J. (1998). Heallymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children. Lancet, 351, 637-641.
Reviewing the theory of a thimerosal based etiology for the autism spectrum: A partial review of Generation Rescue's 25 myths
I was reluctant to begin this post. The thimerosal debate is not very friendly. Also, bad things can happen when we leave our area of knowledge and trod into someone else’s. I am not a medical student or toxicologist. I have only a very novice understanding of chemistry. All the same I can and do have a knowledge base in epidemiology and in research design. The thimerosal argument can not be effectively made in the absence of the previous two disciplines. I will focus my critique on these points. I thought the best way to focus on this will be to look at certain myths in Generation Rescue’s “25 Mercury Myths” (Generation Rescue, 2005).
Myth 1
“There is no evidence to suggest that autism is genetic. No autism gene has ever been found and the search will be endless - how can you have a gene for a mythical condition? Autism is mercury poisoning. What is true is that certain children may have an impaired ability genetically to detoxify heavy metals from their systems. These children are more likely to be affected by mercury exposure. However, all children, and adults, if given too much mercury will manifest symptoms of mercury toxicity, which we call "autistic" symptoms. All children born from 1991 forward who received all recommended vaccines were injected with levels of mercury that dramatically exceeded safety levels set by the Environmental Protection Agency for adults. Mercury has become ubiquitous in our environment: in fish and other foods, water, and air. Exceedingly high doses of mercury exposure can result in death - it is that neuro-toxic and damaging to the human body. Two drops of dimethylmercury spilled onto the gloved hand of a Dartmouth chemistry professor, a leader in a study investigating mercury's causal role in cancer, resulting in the progressive loss of her balance, speech, vision, and hearing, and ultimately lead to her coma and death within a year of the exposure.”
No gene has been found which causes autism, that is correct. If autism is mythical why does there seem to be certain cognitive traits (strengths and weaknesses) that have been identified in the research. Is this a departure from what is known about other heavy metal toxicity? Why do older persons who have been exposed to mercury not receive the diagnostic of autism then, it seems there are important dissimilarities.
“It is impossible to have a "genetic epidemic." Since 1991, there is a very real and dramatic rise in the incidence of autism and other neurodevelopmental disorders. In the 1970s, the incidence of autism was 1 in 10,000 children. In 1986, the rate was 1 in 2,500. Today the rate is 1 in 150. It has been estimated that one in six children have some type of learning disability. Epidemics can happen in 10 years, genetic changes to populations require many generations.”
Where is the increase evidenced? Depending on the data, this may or may not be valid, the definition does not attempt to answer for other factors such as change in criteria, increased knowledge, and more precise diagnostic tools. In the 1970s the mean prevalence for autism was (3.27) per 10,000 by the way; if we include the whole autism spectrum then we have 13.7 per 10,000. Also, the present prevalence is probably best described as 1 in 166. While genetic changes take many generations, perhaps knowledge of the various ways of being human take even longer.
Myth 2
“There is a growing body of evidence that children properly treated for mercury poisoning fully recover normal functioning and are indiscernible from their neurotypical peers. Any toxicologist will tell you that mercury poisoning represents a temporary, treatable state. Thorough removal of mercury will resolve most or all of the symptoms. Autism is only life long if mercury poisoning is never treated”
This would be the appropriate time to cite peer reviewed research clearly demonstrating this, this is absent.
Myth 6
“This myth persists despite being refuted by a wide range of scientists, policy makers, and health care organizations. All the available data points to an epidemic. Between 1992-2002, the Department of Education estimates that there has been a 714% increase in the number of autistic children. In the 1970s, autism was estimated to occur in 1 in 25,000 children. Between 1970 and 1990, that number increased to about 1 in 2,500. Today, the CDC acknowledges the number is about 1 in 166, even Eli Lilly, the maker of Thimerosal, says it's 1 in 150. Many believe it is closer to 1 in 125. The anecdotal evidence that we are experiencing an epidemic is overwhelming. If there is no epidemic, then where are all the autistic adults? Ask any doctor, teacher, or day care worker who has been around children for 20 or more years, and they will tell you that the epidemic is unprecedented. What parent, either now or 20 years ago, does not notice that their child who spoke at one year is no longer speaking, or that their child does not respond to their name or look them in the eyes, or is displaying odd, repetitive behaviors like hand-flapping, spinning, and rocking that no other child is doing? Did those parents 20 years ago not notice these things?”
Where is this refuted via research? I have dedicated other posts to why the education data should never have been used to track prevalence or incidence. A Photon In the Darkness (2005) points out that the plural of anecdote is not data. By the way what was the average age of diagnosis twenty years ago; maybe we are not always as good of observers as we think. Also, twenty years ago Autism was a new diagnosis (placed in DSM-III, 1980).
Myth 8
“The symptoms of autism and other neurodevelopmental disorders are identical to the symptoms of mercury poisoning. The rapid rise in the number of these disorders corresponds to the dramatic increase in the amount of Thimerosal (49.6% ethylmercury by weight) given in recommended vaccines to children under two years of age. With the addition of two new vaccines in the early 1990s, the amount of ethylmercury increased 246% with most of this increase being given within the first six months of life when an infant's neural, detoxification, and immune systems are all undergoing rapid development. Equally damaging, the vaccines were given much earlier in a child's life, when the capacity to detoxify is still developing. Numerous studies demonstrate again and again the causal link between neurodevelopmental disorders and mercury, regardless of the source of exposure. Finally and most definitively of all is that when mercury is removed from these children via chelation or other means of detoxification, their symptoms resolve.”
There seems to be important differences, such as the shaking that is seen in mercury poisoning. Again correlation is not causation, there are other correlates as well. No study demonstrates the causal link between autism and mercury, this is in part because, such would never get past an ethics review board. Where is the research?
Myth 15
“You could analyze the data the government maintains through its "Vaccine Adverse Events Reporting System" and compare the data they already have on children who received Thimerosal-containing vaccines against children who did not receive Thimerosal in their vaccines. This study has already been done by Mark & David Geier and showed a high correlation between Thimerosal dosing and neurological disorders:”
Correlation is not causation.
“You could compare the symptoms of mercury poisoning and the symptoms of autism and see how similar they are. This study has already been done and demonstrated that the symptoms of autism and the symptoms of mercury poisoning are exactly the same:”
Except for the instances where they are not similar.
“You could administer a chelating agent to remove heavy metals, including mercury, to a group of autistic children and to a group of neurotypical children and measure the amount of mercury coming out of the children to see if there are any differences. This study has already been done by Jeff Bradstreet et.al. and showed that autistic children excrete significantly more mercury than neurotypical children:”
Yes, but that does not mean mercury caused the autism. Also, a more equivalent control sample would have been desirable.
Myth 16 “You could inject a group of mice with Thimerosal in doses that proportionally mimic the timing and amount received according to the recommended vaccination schedule and compare these mice to a control group for neurological development. This study has already been done by Mady Hornig et al. and showed that a subset of mice with genetic detoxification impairments who received Thimerosal injections developed "autistic symptoms":”
To my knowledge we have no protocol for identifying autistic mice.
“You could run a trial of 31 autistic children where you chelated patients over the course of twelve months and had parents videotape their children and test urine and fecal samples for toxic metals every other month. You could then compare the children's progress and symptoms from the beginning to the end of treatment. This study was done by Dr. Rashid Buttar and he made the following statement before Congress.”
The videotape is not sufficient to show causation, this is due to observer drift. This also never made it to a peer reviewed journal.
Myth 16
“Generation Rescue believes autism is an issue of toxicology. Yet, you never hear from a toxicologist saying there is no correlation between autism and mercury. This is because toxicologists know that the link is likely. Hearing a psychiatrist comment on mercury toxicity is like seeking the opinion of a urologist for a new heart procedure. It doesn't make sense to accept the expertise of people who have no experience in the field of heavy metal to.”
Anyone can make a criticism, the criticism has to stand on its own merit. Same goes for the validity of an experts opinion, it has to stand on its own merit.
Myth 19
“In May 2004, the Institute of Medicine released a 216-page report titled Immunization Safety Review: Vaccines and Autism and concluded that there did not appear to be a causal link between Thimerosal and the autism epidemic. This study was paid for by the CDC, a conflict in of itself, and there is growing evidence that the conclusion was pre-ordained before any research was done. Regarding the potential link between mercury and autism, Dr. Marie McCormick, Committee Chair of the IOM study, in a recently released transcript, stated (before any research had been done), "We are not ever going to come down that it is a true side effect." Much of the IOM's conclusion was based on the aforementioned CDC and Danish studies - there was no primary research done. This lack of new, primary research is a critical point: the IOM's conclusion was largely based on the studies discussed in Myths 17 & 18 above that are controversial flawed.”
Its seems both sets are flawed. The Danish studies are mire by conflicts of interest and the others by lack of scientific rigor. This is inconclusive.
I think covering the other myths would involve me repeating earlier points.
The data so far seem inconclusive either way. For my part I don’t think the thimerosal theory can hold up in the absence of the proposed epidemic; I take the position that there is no epidemic. Maybe, ultimately the ball is back in the court of the advocates of this theory. They will have to conclusively and repeatedly show the effectiveness and safety of chelation using research that demonstrates causation, and publish in peer reviewed journals. They will also have to effectively demonstrate an increase in autism exists, and they will have to deal with the ethical questions that arise from attempting to remove autism from children. Until this is done chelation will remain pseudo-scientific.
Chakrabarti, S., & Fombonne, E. (2001). Pervasive developmental disorders in preschool children. Journal of the American Medical Association, 285, 3093-3099.
Reviewing the Autism Prevalence (The Epidemiology: Part 3)
The question as to why deviation and variation exist at all can be partly accounted by chance, but more so by other factors and specifically the changes in diagnostic criteria.
During that time span 1966-2001, we progressed from individual researchers criteria for autism such as Treffort (1970), to DSM-III criteria used by Sugiyama & Abe (1989). Finally, we came to the DSM-IV criteria used by Bertrand et al. (2001). Gernsbacher et al. (2005) also mentions that better awareness of Autism Spectrum Disorders and the fact that new diagnostic categories take time to be fully utilized, account for some of the change over the years and more precise diagnostic instruments are now used as well.
This is could account for the rise we see over time. If only the specific discoverers of Autism Spectrum Disorders and those who were familiar with their work used the term of Autism, the bulk of mental health professionals would not be likely to apply that term to a particular person.
In the DSM-III (1980) their were only two categories, but this set down in stone, what the criteria for diagnosing Autism Spectrum Disorders were. Other mental health professionals could more efficiently learn and use the criteria for diagnosis. The 1980 decade correlates to a jump in the prevalence in the 1980s. In DSM-IV (1994) the 1990 decade correlates to another jump. In the DSM-IV, there are five categories and a more permissive diagnostic structure (Gernsbacher et al., 2005).
Although correlation is not causation, it is possible to conclude that the change in the diagnostic criteria is responsible for the increase in the prevalence of Autism Spectrum Disorders as a whole and also of Autistic Disorder Specifically. Put more basically; Autism is what you define it as. The more permissive the diagnostic structure, the more persons whom the term will apply to.
The mental health epidemiology has been conducted in multiple countries; including Sweden (Kadesjo, Gillberg, & Hagberg, 1999), Finland (Kielinen, Linna, & Moilanen, 2000), The United States (Treffort, 1970), Japan (Sugiyama & Abe, 1989), Germany (Steinhausen, Gobel, Breinlinger, & Wohlloben, 1986) France (Ciadella, & Mamelle, 1989), The United Kingdom (Webb, Lobo, Hervas, Scourfield, & Fraser, 1997), and Iceland (Magnusson, & Saemundsen, 2001).
It has been noted that the source used to diagnose effects the prevalence rate. The DSM-IV Autistic Disorder is more permissive than is the DSM-III equivalent Infantile Autism (Gernsbacher et al., 2005). In addition nearly three fourths of the present day Autism Spectrum Disorder diagnoses are not Autistic Disorder (Chakrabarti, & Fombonne, 2001). Only one of these variants was present in the DSM-III. The change in diagnosis patterns in Autistic Disorder and in the PDD’s in general can be traced to changes made from the DSM-III to DSM-IV.
However such statements have been disputed. A good example of the problem is given by the Medical Investigation of Neurodevelopmental Disorder (MIND) Institute which was commissioned by the California Legislature to determine if the changes in Autism prevalence could be caused the loosened diagnostic criteria. The MIND Institute concluded that the increase in Autism could not be explained by the loosening of the criteria for Pervasive Developmental Disorder (p. 5, M.I.N.D. Institute, 2002). The study consisted of studying two cohorts. The first was born between 1983-1985. The second was born between 1993-1995. Both sets were assessed using DSM-IV criteria. The authors found that both groups, former and latter, met criteria. Since both cohorts met criteria they were then compared for prevalence. The cohort born in 1993-1995 had far more. Ergo, the authors concluded that the increase was not due to diagnostic changes (p. 5, M.I.N.D. Institute, 2002).
A problem is noted by (Gernsbacher et al., 2005) in that the earlier cohort would have been diagnosed under DSM-III criteria which were stricter. The DSM-IV criteria were loose enough to encompass both cohorts. That is unlikely to be said for the DSM-III criteria. (Gernsbacher et al., 2005) note the faulty logic used here and provide an excellent additional example based on theoretical male height measurement.
In addition, I will provide an example based on pebbles and a strainer. The DSM-III is a strainer which a number of pebbles fell through. The DSM-IV is a strainer with bigger holes compared to the DSM-III strainer. More pebbles are going to fall through. Furthermore, if you put all of the pebbles that fell through the DSM-III strainer into the DSM-IV strainer, of course they will go through. It is impossible to conclude whether change occurred because of diagnostic changes based on such design logic.
References
Same as in the previous post “Reviewing the Autism Prevalence (The Epidemiology: Part 2)”
Reviewing the Autism Prevalence (The Epidemiology: Part 2)
If upon reviewing the evidence, we take the stance that there is no epidemic of autism, we can be asked the totally reasonable question of why there appears to be an increase. I have already made some answers concerning the use of the IDEA and California DDS data; in this post I will discuss the change in the epidemiology statistics.
The first thing to note is the difference between studies over the years. In fact wide discrepancy can be observed in the Autistic Disorder specific prevalence per 10,000 persons in the whole world. This includes (.7) calculated by Treffort (1970) using criteria established by Dr. Kanner’s (The person who first described Autism) observations. Sugiyama & Abe (1989) calculated (13) using The Diagnostic and Statistic Manuel-III (DSM) criteria. This can be compared to the greater prevalence of (72.6) calculated by Kadesjo, Gillberg, & Hapberg (1999) using The DSM-III and the Revised International Disease Classification-10 (ICD).
The ICD-10 and the DSM-IV have some discrepancies and although close, are not totally comparable. When viewing studies that offer a rate of all the Pervasive Developmental Disorders, a range can be found from (11.2) per 10,000 persons (Fombonne, & du Mazabrun, 1992), through (67.4) (Bertrand et al., 2001). The problem with these data is that they are calculated from different diagnostics, some of which are more stringent and that they used population samples of different sizes.
Fombonne (2003) lists the prevalence rates of 32 Autism Epidemiology studies from (1966-2001). These are listed in tables (I, III) within (Fombonne, 2003). These data are taken on world wide epidemiology. I calculated the means of these studies by decade, because an increase can be observed this way. It should be noted that the specific decade mean, that a study was placed in this paper, was determined by the year it was published and not the year of the actual data collection.
Fombonne (2003) lists the prevalence rates of 32 Autism Epidemiology studies from (1966-2001). Based on this I calculated the mean in the 1970’s at (3.27) per 10,000 (Standard Deviation =2.24). I calculated this in the 1980’s at (6.98) per 10,000 (SD=5.12). I calculated this in the 1990’s at (19.26) per 10,000 (SD=22.5). I calculated this in the year’s post 1999 at (21.06) per 10,000 (SD=11.78).
The standard deviations are very large in comparison to the means. In the 1990s the standard deviation exceeded the mean. This is an indication of the very discrepant results within this decade. In years post 1999 the standard deviation decreases, showing a greater consistency of the studies.
Using table (3) in the Fombonne (2003) display of studies that calculate Pervasive Developmental Disorder, I calculate the mean in the 1970’s at (13.7) per 10,000 (SD=10.61). I calculated this in the 1980’s at (8.5) per 10,000 (SD=2.96). I calculated this in the 1990’s at (13.75) per 10,000 (SD=3.61). I calculated this in the year’s post 1999 at (49.8) per 10,000 (SD=20.2). In this case the 1970s research seems to have more variation than the more recent decades.
This analysis does not include the most recent epidemiology, specifically Chakrabarti & Fombonne, (2005). However, this study should not significantly affect the standard deviation and mean. It will raise the mean slightly towards the prevalence predicted in Chakrabarti & Fombonne, (2005).
References
American Psychiatric Association. (1980). Diagnostic and Statistical Manual of Mental Disorders, Third Edition. Washington, DC: American Psychiatric Association; 1980.
American Psychiatric Association. (1987). Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised. Washington, DC: American Psychiatric Association; 1980.
American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington, DC: American Psychiatric Association; 1994.
American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision. Washington, DC: American Psychiatric Association; 1994.
Arvidsson, T., Danielsson, B., Forsberg, P., Gillberg, C., Johansson, M., & Kjellgren, G. (1997). Autism in 3-6 year-old children in a suburb of Goteborg, Sweden. Autism, 2, 163-173.
Autism Diva. Sunday April 27, 2005. http://autismdiva.blogspot.com/ Accessed April 27, 2005
Baird, G., Charman, T., Baron-Cohen, S., et al. (2000). A screening instrument for autism at 18 months of age: a 6 year follow-up study. Journal of American Academy of Child and Adolescent Psychiatry 39, 694-702.
Bertrand, J., Mars, A., Boyle, C., Bove, F., Yeargin-Allsop, M., & Decoufle, P. (2001). Pediatrics, 108, 1155-161.
Boyle CA, Yeargin-Allsop M, Doernberg NS, Holmgreen P. Murphy, CC & Schendel, DE. (1996) Prevalence of selected developmental disabilities in children 3-10 year of age: The Metropolitan Atlanta Developmental Disabilities Surveillance Program. MMWR Morbidity and Mortality Weekly Reports. 45 (SS-2):1-14.
Bryson, S. E., Clark, B. S., & Smith, I. M. (1988). First report of a Canadian epidemiological study of autistic syndromes. Journal of Child Psychology and Psychiatry, 4, 433-445.
Burd, L., Fisher, W., & Kerbeshan, J. (1987). A prevalence study of pervasive developmental disorders in North Dakota. Journal of the American Academy of Child & Adolescent Psychiatry, 26, 5, 700-703.
Chakrabarti, S., & Fombonne, E. (2001). Pervasive developmental disorders in preschool children. Journal of the American Medical Association, 285, 3093-3099.
Chakrabarti, S., Fombonne, E., (2005). Pervasive developmental disorders in preschool children: confirmation of high prevalence. American Journal of Psychiatry, 162(6), 1133-41
Cialdella, P., & Mamelle, N. (1989). An epidemiological study of infantile autism in a French department. Journal of Child Psychology and Psychiatry, 30, 1, 165-175.
Day, J. C. (1996). Population projections of the United States by age, sex, race, and Hispanic origin: 1995 to 2050. US Bureau of the Census, Current Population Reports, P25-1130, U.S. Government Printing Office, Washington, DC.
Department of Developmental Services. Changes in the population of persons with autism and pervasive developmental disorders in California's Developmental Services System: 1987 through 1998. Report to the Legislature March 1, 1999 19 pages. http://www.dds.ca.gov1: 1999
Deykin, E. Y., & MacMahon, B. (1979). The incidence of seizures among children with autistic symptoms. American Journal of Psychiatry, 136, 10, 1310-1312.
Ehlers, S., & Gillberg, C. (1993). The epidemiology of Asperger syndrome. A total population study. Journal of Child Psychology and Psychiatry, 34, 1327-1350.
Fombonne, E. (1996). Is the prevalence of autism increasing? Journal of Autism & Developmental Disorders, 6, 673-676.
Fombonne, E. (1997). The prevalence of autism and other pervasive developmental disorders in the UK. Autism, 1, 227-229.
Fombonne, E. (1999). The epidemiology of autism: a review. Psychological Medicine 29, 769-786.
Fombonne, E. (2001). What is the prevalence of Asperger syndrome? Journal of Autism & Developmental Disorders, 31: 3, 363-364.
Fombonne, E. (200l). Is there an epidemic of autism? Pediatrics 107, 411-413.
Fombonne, E. (2002). Prevalence of childhood disintegrative disorder (CDD). Autism 6, 2, 147-155.
Fombonne, E. (2003). Epidemiological surveys of autism and other pervasive developmental disorders: an update. Journal of Autism and Developmental Disorders. 33, 365-382.
Fombonne, E., & du Mazaubrun, C. (1992). Prevalence of infantile autism in 4 French regions. Social Psychiatry and Psychiatric Epidemiology, 27, 203-210.
Fombonne, E., du Mazaubrun, C., Cans, C., & Grandjean, H. (1997). Autism and associated medical disorders in a large French epidemiological sample. Journal of the American Academy of Child & Adolescent Psychiatry, 36: 11, 1561-1569.
Fombonne, E., Simmons, H., Ford, T., et al. (2001). Prevalence of pervasive developmental disorders in the British national survey of child mental health. Journal of the American Academy of Child & Adolescent Psychiatry 40, 820-827.
Gernsbacher MA, Dawson M, & Goldsmith HH. (2005). Three reasons not to believe in an autism epidemic. Current directions in psychological science. In press.
Gillberg, C. (1984). Infantile autism and other childhood psychoses in a Swedish region: epidemiological aspects. Journal of Child Psychology and Psychiatry, 25, 35-43.
Gillberg, C., Steffenburg, S., & Schaumann, H. (1991). Is autism more common now than ten year ago? British Journal of Psychiatry, 158, 403-409.
Individuals with Disabilities Education Act. (1990). Public Law 101-476, U.S.C.
Individuals with Disabilities Education Act Amendments of 1997, P.L. 105-17, 20 U.S.C. 1400 et seq.
Kadesjö, B., Gillberg, C., & Hagberg, B. (1999). Autism and Asperger syndrome in seven-year old children. A total population study. Journal of Autism & Developmental Disorders, Vol 29(4), 327-331.
Kielinen, M., Linna, S. L., & Moilanen, I. (2000). Autism in Northern Finland. European Child and Adolescent Psychiatry, 9, 162-167.
Lotter, V. (1966). Epidemiology of autistic conditions in young children: I. Prevalence. Social Psychiatry, 1, 124-137.
Lotter, V. (1967). Epidemiology of autistic conditions in young children: II. Some characteristics of the parents and children. Social Psychiatry, 1, 4, 163-173.
Magnusson, P., & Saemundsen, E. (2001). Prevalence of autism in Iceland. Journal of Autism & Developmental Disorders 31: 153-163.
Mason-Brothers, A., Ritvo, E. R., Pingree, C., Petersen, P. B., Jenson, W. R., McMahon, W. M., Freeman, B. J., Jorde, L. B., Spencer, M. J., Mo, A., & Ritvo, A. (1990). The UCLA-University of Utah epidemiologie survey of autism: prenatal, perinatal, and postnatal factors. Pediatrics, 86,514-519.
Matsuishi, T., Shiotsuki, M., Yoshimura, K., Shoji, H., Imuta, F., & Yamashita, F. (1987). High prevalence of infantile autism in Kurume City, Japan. Journal of Child Neurology, 2, 268-271.
Ritvo, E. R., Freeman, B. J., Pingree, C., Mason-Brothers, A., Jorde, L., Jenson, W. R., McMahon, W. M., Petersen, P. B., Mo, A., & Ritvo, A. (1989). The UCLA-University of Utah epidemiologic survey of autism: prevalence. American Journal of Psychiatry, 146, 2, 194-199.
Rutter, M. (1970). Autistic children: infancy to adulthood. Seminars in Psychiatry, 2, 435-450.
Sponheim, E., & Skjeldal, O. (1998). Autism and related disorders: epidemiological findings in a Norwegian study using ICD-10 diagnostic criteria. Journal of Autism & Developmental Disorders, 28, 217-227.
Steffenburg, S., & Gillberg, C. (1986). Autism and autistic-like conditions in Swedish rural and urban areas: a population study. British Journal of Psychiatry, 149, 81-87.
Steinhausen, H.-C., Göbel, D., Breinlinger, M., & Wohlloben, B. (1986). A community survey of infantile autism. Journal of the American Academy of Child Psychiatry, 25, 2, 186-189.
Sugiyama, T., & Abe, T. (1989). The prevalence of autism in Nagoya, Japan: a total population study. Journal of Autism & Developmental Disorders, 19, 1, 87-96.
Tanoue, Y., Oda, S., Asano, F., & Kawashima, K. (1988). Epidemiology of infantile autism in Southern Ibaraki, Japan: Differences in prevalence in birth cohorts. Journal of Autism & Developmental Disorders, 18, 155-166. 001
Treffert, D. A. (1970). Epidemiology of infantile autism. Archives of General Psychiatry, 22, 431-438.
Volkmar, F. R. (1992). Childhood disintegrative disorder: issues for DSM-IV. Journal of Autism & Developmental Disorders, 22, 625-642.
Volkmar, F. R., & Nelson, D. S. (1990). Seizure disorders in autism. Journal of the American Academy of Child & Adolescent Psychiatry, 1, 127-129.
Volkmar, F. R., Klin, A., Siegel, B., Szatmark, I., Lord, C., Campbell, M., Freeman, B. J., Cicchetti, D. V., Rutter, M., Kline, W., Buitelaar, J., Hattab, Y., Fomhonne, E., Fuentes, J., Werry, J., Stone, W., Kerbeshian, J., Hoshino, Y., Bergman, J., Loveland, K., Szymanski, L., & Towbin. (1994). Field trial for autistic disorder in DSM-IV, American Journal of Psychiatry, 151, 9, 1361-1367.
Webb, E. V. J., Lobo, S., Hervas, A., Scourfield, J., & Fraser, W. I. (1997). The changing prevalence of autistic disorder in a Welsh health district. Developmental Medicine & Child Neurology, 39, 150-152.
Wing, L. (1980). Childhood autism and social class: a question of selection? Britishjournal of Psychiatry, 137, 410-417.
Wing, L. (1993). The definition and prevalence of autism: a review. European Child and Adolescent Psychiatry, 2, 61-74.
Wing, L., Yeates, S. R., Brierly, L. M., & Gould, J. (1976). The prevalence of early childhood autism: comparison of administrative and epidemiological studies. Psychological Medicine, 6, 89-100.
World Health Organization. (1992). International Classification of Diseases, 10th Revision (ICD-10). Geneva, Switzerland: World Health Organization; 1992.
Yeargin-Allsopp, M., Rice, C., Karapurka, T., Doernberg, N., Boyle, C., Murphy, C. (2003). Prevalence of autism in a US metropolitan area. Journal of the American Medical Association, 289, 49-89.
Reviewing the Autism Prevalence (The Epidemiology: Part 1)
The epidemiology establishes the prevalence for the autism spectrum. It can be the most precise method we posses. However it is not without problems or complications. While, these data seem to be the most correct, they still deserve a careful and critical look
The (1) in 166 children, which converts to (60) in 10,000, was a rough estimate given by the Center for Disease Control based on several epidemiological studies they helped conduct (Center for Disease Control, 2005). The Metropolitan Atlanta Developmental Disabilities Surveillance Program (MADDSP) found (20-30) per 10,000 children had an Autism Spectrum Disorder (Boyle, Yeargin-Allsop, Doernberg, Holmgreen, Murphy, & Schendel, 1996). The CDC website also discusses (Bertrand et al., 2001) which found a prevalence of (67) per 10,000 for the autism spectrum and (40) per 10,000 for Autistic Disorder. Only (Bertrand, Mars, Boyle, Bove, Yeargin-Allsop, & Decoufle, 2001) found near (60) per 10,000. The other studies the CDC helped conduct were lower.
Bertrand et al., (2001), like the 3 other studies that produced high prevalence rates (Kadesjo et al., 1999; Baird et al., 2000; Chakrabarti, & Fombonne, 2001) were noted to have small sampling populations (Fombonne, 2003).
I have concerns about the prevalence given by Bertrand et al., (2001). The CDC was aware of a potentially higher incidence in this township as they were going into this city (CDC, 2005; Fombonne, 2003). Originally I had felt that while the data taken from this township might be appropriate for setting the upper limit of autism spectrum prevalence, it might be harmful if applied universally.
Yeargin-Allsopp, Rice, Karapurka, Doernberg, Boyle, Murphy (2003) provides another recent US study to compare to for Autistic Disorder. They found a prevalence of (34) per 10,000. This is lower than the results found by (Bertrand et al., 2001).
However, a recent developments alleviates my concerns and seem to indicate that the (60) per 10,000 is an appropriate prevalence rate. Chakrabarti & Fombonne, (2005) compare a recent cohort to a previous study (measured from 1992-1995) which they conducted in the identical geographical area in the United Kingdom. They conclude that the prevalence rate is high, but stable at (62) per 10,000. And the new study with a sampling population of (10,903) is notably high and resolves my concern about other studies small sampling populations.
The difference between the prevalence rates calculated in the United Kingdom based Chakrabarti & Fombonne, (2005) and United States Bertrand et al., (2001) are insignificant. Dr.s Chakrabarti and Fombonne have convincingly shown that the autism spectrum prevalence rate is high, stable across time, and international. This seriously calls into question the validity of the concept of an epidemic of autism.
References
Baird, G., Charman, T., Baron-Cohen, S., et al. (2000). A screening instrument for autism at 18 months of age: a 6 year follow-up study. Journal of American Academy of Child and Adolescent Psychiatry 39, 694-702.
Bertrand, J., Mars, A., Boyle, C., Bove, F., Yeargin-Allsop, M., & Decoufle, P. (2001). Pediatrics, 108, 1155-161.
Boyle CA, Yeargin-Allsop M, Doernberg NS, Holmgreen P. Murphy, CC & Schendel, DE. (1996) Prevalence of selected developmental disabilities in children 3-10 year of age: The Metropolitan Atlanta Developmental Disabilities Surveillance Program. MMWR Morbidity and Mortality Weekly Reports. 45 (SS-2):1-14.
Chakrabarti, S., & Fombonne, E. (2001). Pervasive developmental disorders in preschool children. Journal of the American Medical Association, 285, 3093-3099.
Chakrabarti, S., Fombonne, E., (2005). Pervasive developmental disorders in preschool children: confirmation of high prevalence. American Journal of Psychiatry, 162(6), 1133-41
Kadesjö, B., Gillberg, C., & Hagberg, B. (1999). Autism and Asperger syndrome in seven-year old children. A total population study. Journal of Autism & Developmental Disorders, Vol 29(4), 327-331
Yeargin-Allsopp, M., Rice, C., Karapurka, T., Doernberg, N., Boyle, C., Murphy, C. (2003). Prevalence of autism in a US metropolitan area. Journal of the American Medical Association, 289, 49-89.
Autism Diva (2005) wrote to the an official in the California Department of Disability Services (DDS) and displays the email from that person. The email indicates in no uncertain terms that the DDS data should not be used to track autism prevalence (Autism Diva, 2005).
The letter indicates that it is inaccurate to label the change between given quarters as “new intakes”. This is due to change in service categorization (Autism Diva, 2005). This is comparable to what occurs in the IDEA data sets; we are dealing with a assignment to service categories rather than strict descriptive label.
Based on these data, Autism Diva (2005) went on to calculate the prevalence of autism via the DDS system at 1 per 321, which like the IDEA data is discrepant from what is established in the epidemiology.
For additional interesting and problematic use of the DDS data please see Gernsbacher, Dawson, & Goldsmith (2005). That article can accessed here http://www.psychologicalscience.org/pdf/cd/autism_epidemic.pdf
Gernsbacher MA, Dawson M, & Goldsmith HH. (2005).Three reasons not to believe in an autism epidemic.Current directions in psychological science, 14 (2), 55-58.
Reviewing the Autism Prevalence (The Use of the IDEA Data: Part 3)
The final reason I will give not to accept the IDEA autism data as a good predictor of prevalence is the discrepancy to the epidemiology prevalence rates. I consider the two main reasons already given along with this reason, to be adequate to explain why the IDEA data should not be used to calculate autism prevalence.
Based on the data in the annual report to congress, provided by (IDEAdata.org) there are 77,453,872 students ages 3-21 in the US in 2003. There are 163,746 students meeting criteria for Autism ages 3-21 in the US in 2003 (IDEAdata.org). Based on this (Autism Diva, 2005) calculated a prevalence of 1 per 473 students age 3-21; this converts to 21 per 10,000.
This discrepant from Fombonne (2003) and the Center for Disease Controls (2005) upper estimation for pervasive developmental disorders at 1 per 166 or 60 per 10,000. References
Autism Diva. Sunday April 27, 2005. http://autismdiva.blogspot.com/ Accessed April 27, 2005
Fombonne, E. (2003). Epidemiological surveys of autism and other pervasive developmental disorders: an update. Journal of Autism and Developmental Disorders. 33, 365-382.
Reviewing the Autism Prevalence (The Use of the IDEA Data: Part 2)
A major concern with the IDEA autism numbers is that they are different from the epidemiology that exists for autism. I mentioned some of these differences in the last post and will add to this in the current post. Such differences are important as they show that the IDEA based autism numbers are not meant to calculate prevalence.
I want to point out that the education numbers are important. They can show emerging trends (if not actual incidence) and are useful if we need to decide to allocate more funding to this or that area, or urge more Universities to start a autism certification program for their student special education teachers. As they are different from the mental health based DSM-IV diagnostic system, they inevitably measure different things.
I have mentioned before that some groups persist in using these numbers to show prevalence, and that this is not a good idea. However, if we insist on using the IDEA numbers and that they show an epidemic, we could make a number of predictions. We can base these on other times a given population increased. Take fore example the US population under the baby boom (The great increase of babies being born just following World War II). If we were to graph all ages from 1947-1980, we would see a curve or a wave, slowly travel across the years as the baby boomers advanced in age. We could expect to see something similar in IDEA numbers.
I included some graphs I made that show the number of students receiving services in the autism service category by age from 1996-2003.
The traveling wave, is absent. The most notable trend, is an increase year to year at nearly every data point as noted by Laidler (2005). I included a second chart with a red line to show the increase as six year olds become seven year olds and so on.
Analyzing the trend is somewhat difficult because there seems to be no one receiving services for autism below age 6 in the years prior to 2001. I know several children who were below age 6 and were diagnosed in the years prior to 2001, I would assume an awful lot of people do.
By why the increase from corresponding data points? We would assume that there would be nearly the same number of 7 year old in 2001 as we had 6 year olds in 2000. That is not what we see. This brings us into the second major trend in the graph; the lines follow a very similar pattern of increase, plateau, and then decrease. This indicates that the line form is a function of another factor besides actual prevalence.
We know that students do not first demonstrate the characteristics of autism at age 13, so why the increase from the 12 year olds the previous year?
Two things come seem possible, either there is a progression to autism that can occur much later than indicated in the DSM-IV or we missed them earlier and are playing catch up by diagnosing now. But remember the IDEA categories do not constitute a diagnosis, simply an assignment to a service category.
And that is the answer I think. We simply assign to the category as necessary for services, and we have been increasingly doing so since 1990 as the service category becomes more accepted and used (Gernsbacher, Dawson, Goldsmith, 2005). As the students leave school, or enter private school, or are reassigned to some other service category as they grow older, they appear to be the lower prevalence of the older children.
References
American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington, DC: American Psychiatric Association; 1994.
Gernsbacher MA, Dawson M, & Goldsmith HH. (2005).Three reasons not to believe in an autism epidemic.Current directions in psychological science, 14 (2), 55-58.Individuals with Disabilities Education Act. (1990). Public Law 101-476, U.S.C.
Laidler, J. (2005). US Department of Education data on "autism" are not reliable for tracking autism prevalence. Pediatrics, 116 (1), 120-124. Notes
All visitors to this blog are welcome to comment in a manner they see fit. Vehement disagreement is permissible; however comments must be on topic, and must involve actual argumentation. No drive by character assassination or trolling please. Also, no services, events, or research recruitments may be advertised on this blog…period.
Violating posts will be removed.
