From: "Clifford G. Miller"

Dear Professor Leventhal,

The World Autism Pandemic

Thank you for your email (text below).  I am obliged to you for your response and apologise for overlooking your professorial rank.

Proof in science requires facts and reliable evidence.  Conjecture, argument or opinion are not acceptable substitutes.  Consider this exchange a scientific peer review of your claims. 

My challenge to you is to produce credible scientific evidence to prove your assertions.  You produce none.  You are the person making the assertions so it is incumbent on you to produce the evidence.  As I already stated, there is none.  And as I have already pointed out, you have claimed regarding the causes of autism that you "will pursue that goal through the careful study of the data." so you should already have that data.

Regarding your points in your email.

1.   When you have returned from your travels, I trust you will endeavour to send me what you claim is the proof of your assertions, albeit I already know there is none and there never will be any - at least any that is valid.

2.   I am not sure what you are referring to.  If you refer to the authors you cite in your point 3, I am for example already familiar with the work of Fombonne and Chakrabati and my reading has gone well beyond them on this topic.

3.     None of the papers you cite assist in proving any of your assertions.  I am not interested in argument.  I am asking for the hard scientific proof to back up your theories.  If you have not got the proof then you must retract your claims until you have the proof.  And that is something you will never have because it does not exist.

Sadly, also, you are mistaken regarding the fundamental issues.  To test your theories we must establish the hypotheses.  One of your hypotheses is that prevalence has always been 1 in 150.  Another hypothesis is that reported increases in prevalence are due to "clinical blindness". 

Further, your claims that a) "the fundamental question is if incidence has changed" and b) there is "no evidence of that" are erroneous bases from which to start.  They are also incorrect claims in themselves.  I will explain further below.

First let me happily discount Fombonne's work, although his data and results are consistent with vaccinations as the cause of the autism pandemic (see also below for why). 

Discounting Fombonne

Cochrane's 2006 MMR [1] review dismissed Fombonne's 2001 MMR/autism paper [2] as "impossible to interpret" and said (page 21) "The number and possible impact of biases in this study is so high that interpretation of the results is impossible".  Which is the politest possible way of putting the point.  In short, Fombonne is not a reliable source.

In another paper [3] Fombonne pulled a stunt over Canadian data, switching the data between Montreal and Quebec city, two different locations [4].

In yet another paper [5] Fombonne claimed:-

"If the prevalence figure for our new sample had increased from that in the previous study, that might have pointed toward an increase in the incidence of pervasive developmental disorders, consistent with the various claims of an "epidemic of autism" (4, 5). We did not find such an increase, and there was no statistical difference between the two prevalence rates, not even a trend in that direction."

His first claim is nonsense.  You do not need an increase or decrease in incidence to achieve an epidemic or a pandemic.  Any steady non trivial level of incidence for a new chronic lifetime condition will achieve a large increase in prevalence in a national population over time. 

And to find a new chronic lifetime condition satisfying that, all we need to do is to postulate that the increase started in the 1980s, with the new vaccination schedules. 

With that hypothesis, all Fombonne achieves is to demonstrate that with a steady level of annual vaccinations, data from the mid 1990s is consistent with a stable steady level of incidence of autism caused by a stable steady level of vaccinations starting in the mid 1980s. 

And further consistent with vaccination being the cause, we would expect to see a large increase in prevalence rates in national populations.  This is because there is no aged adult population with autism and also consistent with vaccination as the cause, they did not have the vaccinations that children have now been getting since the 1980s.  If there was, adult cases would be dying out in the same numbers as new cases arose and so would balance out the new cases in children. There are just not the numbers of aged people with autism dying because they just do not exist.  Instead whilst we have adults who do not have autism dying, we have large numbers of children coming into the population every year increasing year on year the total number of individuals with autism.  If matters continue as they are, we can predict that prevalence will continue to increase for the next 40 years until the current generations of children reach maturity and old age, and start dying out.

So whilst the data and results from the studies you cite are consistent with an environmental insult like vaccination being the cause of the autism pandemic, there is nothing to prove it is not.  If you want to continue repeating your unproven theories, you need to find that clear proof or else withdraw them.  Again, conjecture and argument are not adequate.  The world needs proof and from someone in your position, they have a right to demand it, and more especially Illinois parents of children on the autistic spectrum.

So to prove your theories, you have to find all of the aged adults and you need to find them right now, because remember, you claimed regarding the causes of autism that you "will pursue that goal through the careful study of the data."  So you have to have already proof that there are just as many adults as children with autism.  So where is the proof and where are they, the aged adults?  Nowhere, in both cases.

"fundamental question is if incidence has changed" and "no evidence" it has
These are fundamentally erroneous assertions of yours.  As with some instances of research bias, your "fundamental" question assumes the proposition you seek to prove is true, that the number of autism cases has always been 1 in 150 since 1943, so when you get the answer you are looking for, you erroneously assume the proposition is true also.  You put the cart before the horse, as I show here. 

Compounding that error, there is no credible scientific evidence to support the assertion there has been no change in incidence of autistic spectrum disorders since 1943.

Additionally, it is no use looking only at the rate of new cases (incidence) after the new vaccination programmes in the 1980s and early 1990s to then claim that the rate of new cases has always been the same (as Fombonne also does).  That is nonscience and nonsense.

Most of the research literature on this is thin on the ground and dates mostly from the mid 1980s.  We simply have no idea what the incidence was prior to the 1980s save to assume it was around the same level as the early 1980s and that could well be an erroneous assertion also.  And the literature that does exist during and since the 1980s (including Fombonne) shows incidence and prevalence are both different to and substantially higher than the mid 1980s and therefore higher than what they must have been prior to the 1980s.  Again, I cannot fathom how you make the claims you do.  There is no factual or scientific basis for any of them.

If we assume your claim of 1 in 150 since 1940 is correct, then and only then would a change in incidence (new cases occurring) be relevant but not otherwise.  So if your claim was wrong and we still found no change in incidence since the new vaccination schedules of the early 1980s, the failure to find a change of incidence would not prove your assertion.  What it would still do is support the assertion that the changes in vaccinations schedules in the mid 1980s and early 1990s are the cause. Your analysis is therefore deeply flawed I regret to say and parents of autistic spectrum children in Illinois are paying the price.

In summary, the answer you seek assumes the proposition you need to prove is true, that the total number of cases in the population is steady at 1 in 150 since 1943.  For that to be true, incidence must be unchanged since 1943.  But if your proposition is wrong, finding an unchanged incidence in recent years since the 1980s, proves nothing.  So your problem is that even if we look for a change in incidence and not find one, you still fail to prove your hypothesis.  So in the end you prove nothing. 

I apologise when I say I struggle to understand how anyone with a reasonable intelligence can make such fundamental errors.

Additionally, whilst there is no literature on which to base the claim of no change in incidence, on the contrary, there is literature supporting the claim there has been a significant change in incidence.  There are numerous studies where the data reports a change in prevalence and, at the least implicitly, in incidence.  It is only people like you who try to put this in doubt by putting forward unproven hypotheses without any evidence to suggest that it might be due to better diagnosis and greater awareness and therefore prior "clinical blindness".  In the absence of any evidence, the reported data in the literature stands.  Unproven speculation is unscientific.

And it is precisely those assertions which you are supposed to be proving and you have produced no evidence to show they are true, again, because there is and can be none. 


4.  You do not like my use of "pandemic".  What numerical definitions of epidemic and pandemic do you rely on? 

A pandemic is an epidemic that affects a wide geographic area.  And an epidemic is a disease or condition occurring suddenly in numbers clearly in excess of normal expectancy.  Whilst his claims are nonsense, Fombonne's data and results confirm "The rate of pervasive developmental disorders is higher than reported 15 years ago" and for childhood autism we are comparing from 2 to 8 in 10,000 in the mid 1980s with figures like Baird 2006's 38 in 10,000.  An increase in prevalence of between 500 to 1900 percent experienced nationally in a lifetime chronic condition is, I would have thought, sufficient to qualify as a pandemic.  The fact it is experienced internationally as well should put the seal on that.  But please correct me if you think I am wrong.

5.  Your assertion is again wrong.  I am not relying on service provider data.  I am relying on the peer refereed literature.  In fact, I am relying on much the same literature as you.  However, as with all such literature, it is fundamental to rely solely on the data and results, and to ignore comments and opinions of authors, if bias is to be avoided.  Every expert should know that.  And that especially applies to papers by Fombonne.  However, even that approach is also taking on trust the data and results are reliable and accurately captured and reported, which is why repeatability is so important as is post publication peer review.

[1] Demicheli V, Jefferson T, Rivetti A, Price D. Vaccines for measles, mumps and rubella in children. Cochrane Database of Systematic Reviews 2005;4:CD0044007.
Access via:

[2] Fombonne E, Chakrabarti S. No evidence for a new variant of measles-mumps-rubella-induced autism. Pediatrics 2001;108:E58. Available at Accessed Oct 25, 2006

[3]  Fombonne E, Zakarian R, Bennett A, Meng L, McLean-Heywood D. Pervasive developmental disorders in Montreal, Quebec, Canada: prevalence and links with immunizations. Pediatrics. 2006 Jul;118(1):e139-50.

[4] A Tale of Two Cities: Flawed Epidemiology - F. Edward Yazbak, MD, FAAP

[5]  Suniti Chakrabarti, M.D., F.R.C.P.C.H., M.R.C.P., and Eric Fombonne, M.D., F.R.C.Psych. Pervasive Developmental Disorders in Preschool Children: Confirmation of High Prevalence Am J Psychiatry 162:1133-1141, June 2005



From: Clifford G. Miller,
CLIFFORD MILLER, Burnhill Business Centre, 50 Burnhill Road, Beckenham, Kent BR3 3LA, England
Tel: + 44 (0) 20 8663 0044 Fax: + 44 (0) 20 8663 0011