Kent_legal_3 May 27, 2008:  The Government’s Response, continued

Direct Examination of Dr. Michael Rutter

Ms. Ricciardella reviews Dr. Rutter’s qualifications as a child psychiatrist and as a researcher in the U.K. 

Q:  Were you were involved in the formulation by the DSM-IV and ICD-10?
A:  Yes; I was involved in an attempt to bring the two sets of criteria closer together.

Q:  Please discuss your clinical experience in autism.
A:  It goes back to the early 1960’s, but I continue to see cases today.  I used to be involved quite heavily in treatment, but in the last decade my work has been more advisory.

Q:  How many autistic children have you diagnosed in your career?
A:  Many hundreds.  I’ve followed them into adolescence. 

Q:  Do you continue a research practice?
A:  Yes, I’m involved in genetic studies, molecular gene-environment interplay and long-term longitudinal studies.

Q:  Your CV indicates you’ve published over 400 peer-reviewed articles?
A:  Something like that.  [review of book chapters; a lot of articles on ASDs; served on lots of editorial boards; continues to teach at a post-graduate level; has taught for fifty years; lectures frequently on a range of topics; he’s received a number of awards; he was previously involved in the U.K. MMR litigation]

[Dr. Rutter describes start of autism research with Dr. Kanner in the 1940’s; discussed qualitatively different development; discusses reciprocity in social interaction; strikingly impaired in ASDs; there are other atypical features; discusses restrictive, circumscribed, stereotyped patterns of behavior; the symptoms tend to develop in the later preschool years]

Q:  What are the diagnostic tools?
A:  ADIR, ADOS tests have been used for research but also for diagnosis. 

Q:  What disorders comprise the autism spectrum?
A:  A range of disorders which vary in their severity.  Aspberger’s syndrome is an example, but it includes a range of less specific syndromes under pervasive developmental disorders not otherwise specified.  Rett’s Syndrome is usually included, but all clinicians would usually see that as something different.  It includes childhood disintegrative syndrome but that has been subjected to much less research.

Q:  Does autism persist?
A:  Yes; a number of long-term follow-up studies show that although there are changes, and the young people may become independent, the qualitative abnormalities do persist.  There is a small proportion that appears to recover entirely, but it’s a small percentage.

Q:  Is autism associated with mental retardation and other mental disabilities?
A:  Yes.  It used to be assumed to be usual, but now it’s broader.  But it’s not a new disorder; there have been retrospective studies from before 1943, and it’s clear that the syndrome existed before 1943.

Q:  You said you did one of the first studies of autism compared to other mental disorders.
A:  [describes study of children with different disorders; the pattern of cognitive skills is quite different than other issues in autism.  The head circumference tends to be normal at birth and then gets bigger in the preschool years.]

Q:  When do parents tend to identify autism?
A:  Usually about 18 months; depends on whether it’s a first child.  Communication problems; lack of social reciprocity are often the first signs to be picked up. 

Q:  You report that signs are often visible at 12 months but that diagnosis doesn’t usually occur then.
A:  There are many studies to look for subtle features evident at an earlier point.  This has been done from home videos.  At 12 months, you can find differences in some children, but not all. 

Q:  Are pediatric records a good measure of autism?
A:  No; the reporters weren’t looking for problems.  If there are abnormalities noted, that’s significant.  If they aren’t noted, that’s not conclusive.

Q:  You noted in your report that there have been cases of autism from herpes encephalitis and other postnatal insults.
A:  They’ve been claimed as such; I include them really out of fairness.  These cases are utterly different and they are rare.  I don’t find those comparable.

Q:  You state that brain abnormalities can give rise to ASD features.
A:  Because we don’t know the precise neural basis for autism, it is difficult to decide where you’re dealing with postnatal causes.  The evidence which is reasonable applies to prenatal causes, but it is possible that early postnatal causes might occur.

Q:  Is there evidence of brain abnormalities?
A:  Yes; the increased brain size is one example.  But we don’t have a good test.

Q:  You mentioned that some autistic individuals get seizures in adolescence.  What percentage?
A:  About 25%.  There was a time when the field thought that there might be a particular part of the brain involved in autism.  That seems not to be the case – it’s more a systems abnormality in the brain where the interconnections are not working the way they should.

Q:  You discuss congenital abnormalities in autism in your report; please discuss these.
A:  Yes; these abnormalities relate to prenatal development.  They are of interest in showing developmental perturbation; the way development should have taken place went wrong prenatally.

Q:  You associate autism with a deficit in “theory of mind.”  Please discuss this.
A:  I don’t really like the term, but it refers to cues about what other people are likely to be thinking.  The ability to read into another person’s mind.  I think of a case I wrote about in 1983; a higher functioning man.  He complained that everybody else seemed to have an extra social sense that he lacked.  It’s a cognitive deficit.

Q:  What’s the genetic component of autism?
A:  The twin studies show strong genetic predisposition; it’s about 60% for autism in monozygotic pairs.  In dizygotic pairs it’s found in 5% or less or 10% for a broader phenotype.  The gap between the identical pairs and those that share half their genes show a strong genetic liability.  The estimate is about 90%.  And the pattern in family studies is very similar to that in twin studies.

Q:  Do non-genetic factors play a role in autism?
A:  The evidence from the twin studies and family studies is that it is a multifactorial system.  You must expect that the disorder comes from a combination of multiple genes; probably a modest number, between 3 and 12, and also “non-genetic factors.”  The point of saying non-genetic factors, rather than environmental factors, is that it could include chromosomal abnormalities and others.  Many of the abnormalities are not genetic in the ordinary sense.  What’s not genetic is not necessarily an environmental hazard.

Q:  You say that because autism seems 90% heritable, there’s little room for environmental factors.  What do you mean by that?
A:  Heretability is a population specific variable.  [discussion of height and heritability] 

Q:  If there were an environmental influence, when would it be most likely to occur?
A:  Likely to be in the prenatal period; maybe in the early postnatal period, but more likely in the prenatal period.

Q:  Dr. Kinsbourne concluded, based on concordance studies, that 10-40% of autism is not described by genes; do you agree with that?
A:  No.  We only know that genetic factors tend to be more important.

Q:  Dr. Kinsbourne states that the causal role of gene-environment interaction has become mainstream in autism research?
A:  No. 

Q:  Dr. Kinsbourne says that the incidence of autism has risen spectacularly.  Do you agree with that?
A:  No.  What’s agreed is that the diagnosis of autism has risen spectacularly; at least a part of the rise has come from broadening the diagnostic criteria and better ascertainment.  We do not know if the incidence has or has not risen.

Q:  Why do you state in your expert report that the reported diagnoses have increased?
A:  I think the change is mainly methodological, but it’s difficult to rule out that there could be an increase.

Q:  What do you mean by the broadening of the diagnostic categories?
A:  I think that lots of normal individuals manifest signs of autism.  Doctors early on were reluctant to diagnose autism if people had normal intelligence.  The broadening has good research support.  But there are two difficulties.  It’s not so clear where you draw the line.  The other is that the group with these milder manifestations is large.

Q:  What is regressive autism?
A:  It’s not a term I like to use because it implies a different category.  It’s been used in MMR and thimerosal claims.  It’s not new; it’s been used for many, many years.  It’s probably not an either/or category.  Regression is for real; lots of studies confirm the reality, but it’s not a distinct group that’s quite different.

Q:  At what age does it take place?
A:  Typically in the 18-24 month range. 

Q:  What percentage of children who are autistic suffer regression?
A:  A quarter to a third; it’s reasonably common, but it’s a minority.

Q:  Has the rate increased over time?
A:  As far as we can see, it’s remained the same. 

Q:  Dr. Kinsbourne’s report says that the regressive type is 20-30% of the total.  He says that diagnostic substitution is a non-starter.  This indicates that the rise in cases of regressive autism is real and rising.  Do you agree with Dr. Kinsbourne’s statement?
A:  I don’t really.  I do agree that the proportion has remained at roughly the same level.  I note that the quality of measurement varies.  The problem is with his “jump” from regressive to non-regressive autism; his analysis rests on many assumptions.  His statement is that ‘genetics can’t explain this,’ but there is a lot of evidence to show that this could be true. 

Two examples:  Huntington’s disease only becomes apparent in middle age; it’s fully genetic, but the effects only come on later.  There is a loss of skill in the early 40’s.  Nothing to do with the environment.  Take another example – women go into menarche in early adolescence; the timing is strongly genetic.  This is what genes are doing.  There are lots of examples where genes are influencing things way down the line.  Let’s go back to the evidence of increased brain size.  There’s no evidence that environmental factors bring this on.  There’s no reason to invoke an environmental factor unless there’s evidence for it.

Q:  Dr. Kinsbourne describes regression as an unknown encephalopathy; is that true?
A:  No; encephalopathy ordinarily means some kind of inflammatory process.  No, we don’t have evidence of that.

Q:  Dr. Kinsbourne says that there are major distinctions between non-regressive and regressive autism.  Do you agree?
A:  No.  There have not been studies comparing autism from regressive autism; it’s pure speculation.  [Discussion of deaf children’s development of language; evidence of problems are apparent only at about six months; babies sound the same everywhere in the early development] 

Q:  Is it valid to conclude that TCVs cause or contribute to autism?
A:  No.  What we need is positive evidence that (a) thimerosal causes autism and (b) that it’s implicated in regression.

Q:  What would you say is the probability that regressive autism is a distinct biological disorder?
A:  I have no idea.

Q:  What, if any, causal inferences can be drawn from the mercury studies?
A:  I think we need to look at what we know about the effects of mercury.  I’m not a toxicologist.  It’s clear that high doses of mercury are toxic and cause death.  There are studies like those from the Seychelles, the Faroe Islands – we look to see if there are more subtle effects.  My conclusions would be in line with most commentators – there may be some slight sequelae from exposure to mercury – it’s hard to say that there’s a safe level.  It’s notable that none of the those studies identify autism as one of the outcomes.

Q:  Are there differences between the symptoms of mercury poisoning and autism?
A:  Yes; numerous differences.  There are very few similarities and a lot of differences.  I’m completely unpersuaded.

Q:  Is there any reliable evidence that chronic, low-dose exposure to thimerosal causes autism?
A:  No.

Q:  Let’s turn to epidemiological evidence.  Is it an important field of science in answering this question?
A:  Yes.  It has been important to look at environmental causes in epidemiology because there are so many things you can’t do directly in the lab.  Have there been successes in this area?  Yes; it has worked in some examples; not in others.  Smoking would be the best example of success.  Back in the 1950’s, there was a study of a strong association between smoking and lung cancer.  Then another looking at a study of doctors who stopped smoking.  It took quite a long time for the evidence to be seen as pretty decisive, although back in the 1960’s, it pointed to this being a likely cause.  Over time, other evidence came in; studies on animals and the effects of tar.  [further discussion of smoking epidemiology] Epidemiology is a crucial part of studying environmental factors.

Q:  What about the epidemiological studies in Petitioner’s master list; do they support the Petitioners’ theory?
A:  They do not support the theory.  My role is to look at the evidence as a whole; not to take a position for or against.  There is no study that is free of limitations.  You look across studies; when you do that, you’re on much stronger ground.  [Discusses Heron; Andrews; Verstraeten – the most satisfactory:  a large sample; a standard methodology; thoroughly analyzed.  The results do not show an association between thimerosal and autism]

Q:  You also discussed several ecological studies [references].  What do those studies tell us?
A:  The time frames are useful in seeing whether the ups and downs are associated with changes in the rates of autism.  They have manifest strength – they can be based on large numbers.  They have important limitations – they are dealing with it at population level and not at an individual level.  [detailed discussion]  When thimerosal is being withdrawn, you should see a decline.  What Stair/Green showed was the same trend in both countries irrespective of what was happening with thimerosal.  In epidemiology, one pays attention to when a risk factor is added or removed in one population or another.  It’s this trajectory that makes it really rather unlikely that thimerosal played a role in the overall rate of autism.  In terms of overall play, I think the studies are pretty compelling.

Q:  Do you find those studies to be credible studies?
A:  Yes, I do.

Q:  You said that large-scale epidemiological studies can’t analyze a small subgroup susceptible to autism.
A:  There are many examples of idiosyncratic behavior. [discussion of his grandson who gets a severe reaction to two types of nuts; there’s a really good test for this susceptibility].  Although there may be differences in reaction to thimerosal, there are no tests for it.

Q:  In your report, you say that it’s “biologically plausible” that there might be an idiosyncratic response to thimerosal in a subgroup.  Please explain.
A:  It’s possible that it might occur.  It certainly doesn’t mean that’s it’s likely because there’s no evidence that it’s likely.  The question here is what the likelihood is.  It’s a theoretical possibility, but at the moment it’s speculative. 

Q:  Do you agree with Dr. Kinsbourne that regressive autism is usually striking and dramatic?
A:  No; it’s usually subtle and doesn’t occur dramatically in a very short period.

Q:  Do you agree with Dr. Kinsbourne’s theory of overarousal?
A:  I was surprised to see it here described as novel; it’s been a theory that’s been around for a long time.  Research on this hasn’t been conclusive.  [further discussion of overexcitation research; what’s been known since Kanner; need to link it physiologically; that’s what’s lacking]

Q:  What would have to be shown about the nature of overarousal?
A:  You’d want to have a physiological measure; you’d want to show that the overarousal applies to social situations.  If it doesn’t apply to social situations, then it would be hard to see how it applies to social reciprocity. 

Q:  Has Dr. Kinsbourne explained how overarousal accounts for regressive autism only?
A:  No; it doesn’t explain how the overarousal is related to regressive autism.

Q:  Dr. Kinsbourne discusses how toxins, viruses, heavy metals can all play a role in his theory of regressive autism.  Do you think this lack of specificity supports his hypothesis?
A:  No, it doesn’t.  The lack of specificity doesn’t disprove causation, but it certainly doesn’t support it.  His theory is interesting, but entirely speculative.

Q:  Is it more likely than not?
A:  Evidence suggests it is not.

Q:  Do you hold that opinion to a reasonable degree of scientific certainty?
A:  Yes.

Q:  Why did you fly here today to testify for the U.S. government?
A:  I think the scientific issues are important as are the public health issues.  It’s something I know a good bit about; it seemed to me I had a duty to testify.

Cross Examination of Dr. Rutter [Mr. Williams]

Q:  I want to start with a general question about what underlies autism in the brain.  Do you think all the children who meet the DSM criteria have the same underlying brain pathology?
A:  I think we have no idea.  I don’t assume that there will be one pathology.

Q:  I think you said once or twice that a postnatal insult could contribute to the symptoms of DSM-IV autism.
A:  I follow British rules in filing my report.  I must look at the evidence against and the evidence for a proposition with equal weight.  I included the herpes encephalopathy cases that can cause autism to be fair to the postnatal view. 

Q:  Are you saying there are similarities in some of the features?
A:  I certainly think there are similarities in features; I’m cautious about saying it couldn’t happen.  I think the particular  explanations that people have put forward are not very persuasive.  [discussion of gene-environment susceptibility; heritability; height; PKU syndrome]

Q:  Are you aware that the Dept. for Health and Human Services, the Respondent, published an announcement on May 23 about a meeting on vaccine adverse events;  will you be involved in this discussion?
A:  No; I don’t recall it and it’s not in my book.

Q:  You actually attributed autism to an immunization in one of your papers.  Do you remember that? [shows him the paper] 
A:  We know there are vaccine adverse reactions.  The paper doesn’t say what the vaccine was.  But there was onset of epilepsy that plausibly was connected with the immunization. 

Q:  You don’t know if there was thimerosal in the vaccine?
A:  These are dealing with the twin and family studies in the 1970’s, so it’s before the MMR and thimerosal were used.  It’s not likely that either was involved, but I’m not sure.

Q:  Head circumference and head size.  Is it your opinion that head circumference is a diagnostic tool to determine whether a child has autism or not?
A:  I wouldn’t wish to put it as a diagnostic tool.  Eric Courchesne showed that it is a robust finding that is distinctive of autism, but it doesn’t apply to all autistic individuals. 

Q:  Does it matter when the head circumference is measured?
A:  Usually it’s measured at birth.  In the UK, that would be the standard reference.

Q:  We looked really, really hard to find an analysis of brain growth from birth to two years of age.  Could you just summarize that?
A:  That’s not something I’ve personally done.   

Q:  Isn’t it likely that environmental insults would affect the pruning process in the first two years of life?
A:  It’s possible.

Q:  You discussed a number of brain autopsy studies on children.  You don’t mention any of the studies that find neuroinflammation.  Why did you leave them out?
A:  No particular reason.  I think I only became aware of the Vargas paper after I did the report.  I’m not a neuropathologist.  The detailed findings go beyond my expertise.

Q:  Today in your testimony there wasn’t anything on neuroinflammation.  Do you think it’s irrelevant?
A:  I think we don’t know whether it’s relevant or not.

Q:  What’s the significance of the Pardo study?
A:  When one looks at the Pardo study, they are very careful to point out that the meaning remains uncertain.  I would be in agreement with that.  When one’s got new findings, one needs to be very cautious about what conclusion to draw. 

Q:  Dr Courchesne wrote a paper called “Autism at the Beginning” where he discusses neuroinflammation; you don’t mention that in your report or in direct testimony.
A:  It’s not an area of my expertise.  I’ve noticed some of the key findings.  In my information, the neuroinflammation does not show clearly what changes are happening and when.  So the Kempner and Bowman findings didn’t show that.  Were they wrong and the newer ones right?  I don’t know.  I’m open to persuasion.  But I am aware of the uncertainty of the causal implications you can draw.

Q:  Do you know if Drs. Kempner and Bowman looked for neuroinflammation?
A:  I don’t know.

Q:  You are aware that there seems to be a system abnormality, not a legion in a specific area.  Wouldn’t neuroinflammation be a plausible explanation?
A:  The question I would want to ask is, is it likely?  Could it cause autism?  Well, I suppose so. 

[Mr. Williams shows him an NIH funded study; it’s recruiting participants to look at regressive autism and use of antibiotics to treat it.]
Q:  Does this NIH study surprise you?
A:  It doesn’t surprise me; NIH should fund long shots as well as sure-fire treatments.

Q:  Let’s look at the purpose of the study: there’s a subset of autism that loses skills or regresses and it’s associated with chronic brain inflammation.  They give the participants an anti-inflammatory drug.  I can represent that this is the Pardo group.  We thought we were going to hear from Dr. Pardo, but we’re not now.  I want to ask you if it makes sense.  Do you agree that’s a reasonable study to undertake?
A:  NIH has undertaken a number of studies over the years that were very long shots.  That’s an appropriate thing for them to do.  [gives examples; phenfloramine as a treatment for autism; later withdrawn; secretin; studies done; NIH funded]

Q:  I note you didn’t cite to the Pardo paper in your report or testimony.  Have you read that paper by Pardo?
A:  Yes I have.

Q:  Is it reasonable to think that there might be people who are more susceptible than others?
A:  It’s possible, but not demonstrated.

Q:  Do you think this is a reasonable model of how some children could develop autism?
A:  Well, if you look at the title, it’s a “hypothetical” presentation of what might be.  Some of those arrows are better demonsrated than others.  Unlike all other psychological disorders, there is no consistent response to drugs for neurotransmitters in autism.

Scientists tell stories about how things might happen.  Then they undertake research to see if it is really going on.  They are putting this forward in a very cautious way.  They’re saying these are some ideas that are worth testing.

Q:  But you didn’t include it in your report.
A:  I didn’t come across it when I was preparing it.

Q:  I want to come back to your paper.  What do you mean by “biological coherence”?
A:  It’s important to ask, is there a coherence?  You need to look at the biological evidence as a whole.  When it leads all over the place, it seems less likely.

Q:  I’d like to show you five or six pieces of our experts’ theories.  [Mr. Williams shows him animal studies about infant monkey studies.]  How would you follow up on these animal studies?
A:  I’m not a toxicologist; I don’t want to talk about this strategy rather than that strategy.  These studies are at any early point, indicating the need for further study about how mercury works.  They don’t relate too much to the thimerosal-autism theory.

Q:  You agree there’s wide variation among individuals?
A:  Yes, in all biological models.

Q:  Do you agree that neuroinflammation has been found in all autistic brains?
A:  No, there really haven’t been a great number looked at, and the findings have to be replicated by independent investigators.  I agree it’s an area where more research is needed.

Q:  You seem to suggest that you were aware of autopsy studies; what studies are you talking about?
A:  The Kempner study looked at glial cells.  The study by Bailey and his colleagues similarly looked and found some compatible evidence.  So based on a very small number of brains by different investigators, we don’t as yet end up with a coherent story.  It’s premature to build much of a theory.

Q:  Do you agree with the Courchesne point that neuroinflammation can explain the symptoms of autism?
A:  It’s a speculative point.  All of these points were made in the literature after 2005.

Q:  You were retained in 2004, right?
A:  Yes.

Q:  When you wrote you report in this case, all this information was available.

Mr. Matanoski, Lead Government Attorney:  The inference here is that Dr. Rutter has pointedly not commented on neuroinflammation.  Neuroinflammation wasn’t part of Petitioners’ theory.  This 3-week old theory of neuroflammation was not something Dr. Rutter had any reason to comment on.  I think it’s inappropriate to be faulting him for something that wasn’t even presented by Petitioners.

Q:  Isn’t the neuroinflammation theory a biologically coherent theory?
A:  It’s a highly speculative theory.  It’s vague.

Q:  Let’s talk about regression for a minute.  Would you agree that there have been cases of clear and dramatic regression in children who have developed normally to 18 months?
A:  Yes.

Q:  You’ve said regression is about a quarter of the cases.
A:  Yes.

Q:  Are you aware of the California regressive autism study?  The title is regression in autism; the Charge study.  Have you not seen this paper?
A:  I think I’ve seen it. 

Q:  Do you agree that this epidemiological study probably is the best measure we have right now of the perecent we have of regressive autism?
A:  The children we’re talking about here lost both language and social skills, so they’re part of a 15% subgroup.  [discussion of percentages in epidemiological studies; references to Dr. Rust, Dr. Greenland]  The evidence today suggests that regression isn’t an either/or phenomenon.  My clinical experience over the last half century suggests that there are some cases that are severe and dramatic and some less so and all along the line.  I don’t think we have the faintest idea how to describe subgroups.

Q:  Do you think there’s been an increase in DSM-IV autism over the last 20 years?
A:  I don’t know.  Certainly better ascertainment has played a part.  Broadening of the concept has also played a part.  I know of no evidence that can rule that in or out.  I am involved with one study following 100,000 mothers and infants from pregnancy onwards, so I’m very heavily committed to look at genetic as well as environmental causes.  It is reasonable to think there are non-genetic factors.  I think that the evidence as to whether there is or is not a real rise is not clear.  I am in favor of research about things that might have caused a real rise.  That was done with MMR and thimerosal.  In both cases, I think the evidence is against that having been responsible for the real rise.  Clearly, they needed to be investigated.  One of the most decisive factors is what happens when those factors are removed. 

Q:  You think the prevalence estimates are reasonably accurate?
A:  I’m not a great adherent to classification systems, despite my involvement in it.  But yes, because they use standardized measurements; they look at confounding factors.

Q:  When were the epidemiological studies published that we can rely on?  Can we rely on studies after 1995?
[discussion of assessment of quality of studies]

Q:  This good evidence of the last ten years accurately measures prevalence?
A:  Yes.  Autism rates are somewhere between a half a percent and a percent.

Q:  Are you aware of any epidemiological studies on the association between regressive autism and thimerosal?
A:  No.

Q:  Let’s talk about the Verstraeten study.  You said one of the strengths was its size; it’s about 140,000 kids.  Do you know what he said himself?
A:  Yes; he said it was inconclusive. 

Q:  This is Dr. Verstraeten’s letter to the journal that published the article in April 2004.  He says, ‘surprisingly, the study is being seen as negative.’  He describes it as inconclusive.

Q:  Do you think it’s negative or inconclusive?
A:  Studies can’t be divided up like that.  Is there any evidence that’s in support?  No.

Q:  He says the investigators could neither confirm or reject a causal link.  He says more VSD study should be made.
A:  But since then, there have been many other studies and none show an association.

Q:  Did you know that in 2006 the IOM convened a group to discuss whether further studies should be done in the VSD on an association between thimerosal and autism?
A:  No, I didn’t know that.

Q:  [pulls up highlights on committee recommendations]  It recommends an expansion of the Verstraeten study; I think they meant both in time and geographically.  The Verstraeten study only had 3 HMOs in it, and one was very small.  Would you agree with this expert panel?
A:  Remember that I come from the UK, but with the available funds in the UK, I’d have to say there wouldn’t be justification to do an expanded study.  If the US wishes to expand it, fine.  I wouldn’t want my taxpayer money used in that way.

Q:  Let’s turn to your discussion of these ecological studies.  Wouldn’t the confidence intervals be enormous?  In discussing the ecological studies in general, you say one of the limitations in the cohort studies is that the amounts of thimerosal received are relatively the same.  Why is that a weakness?
A:  Well, you can’t look at the extremes if everybody gets the same dose at the same time. 

Q:  How would you solve that problem?
A:  Each of the designs has its own strengths and limitations.  There is a firm prediction of what should happen when thimerosal is discontinued.  Its limit is that you can’t look at it on an individual basis.  In cohort studies, you can look at it in terms of what the individual received, but you can’t look at changes over time.  So this comes back to the main point that you’re foolish always to look at one kind of study design; you look at many studies.  If all the different studies come up with a broadly similar answer, then the answer is more likely to be solid.

Q:  Let’s turn to the Geiers’ study.  It’s a much a larger study.  Isn’t that a strength?
A:  Quite frankly, I think it’s a poor study.  It starts off with a cohort design; they have records on individuals.  But they don’t do that; they analyze the first and last cohorts. 

Q:  Do you know if they were allowed to look at individual records?
A:  No, of course I don’t know.  All I’ve got is in the paper.  I think their design and strategy were not satisfactory.  In terms of results, what is striking is that you have a significant statistical effect with really a quite heterogeneous set of disorders.  Let’s take the neuroinflammation hypothesis.  If it’s correct, it is dealing with the most significant effect on tics and emotional disturbances.  If this is seen as supportive, you see neural effects across differences in ages of onset and course.  So that the major lack of specificity would make me immediately skeptical.

Q:  But they did use congenital anomalies?
A:  Yes they did, but why are they there and emotional disturbances not there?  The problem is in the inferences they draw. [discussion of emotional disturbance; it’s not a neurodevelopmental disorder]  They have created two groups, one of neurodevelopmental disorders and one without neurodevelopmental disorders.  Emotional disturbances should not have been in the group with neurodevelopmental disorders. 

Q:  Do you know about the tributelene twin study?
A:  I don’t think so.

Q:  Do you know about follow-up animal studies that show neuroinflammation?
A:  No.

Q:  You mention a study by Ipp et al; you say the mercury findings were in hair.  Are you aware this study has been reanalyzed?
A:  Yes.

Q:  Do you think it now points towards a causal association?
A:  No I don’t.  [discussion of cutoffs vs. confidence intervals; one-tail tests] 

Q:  This reanalysis was published in 2007; you didn’t see this when you were working on your report?
A:  I’ve read it, but not when I was working on this report. [discussion of p value in the report showing results are statistically significant; discussion of lead poisoning]

Q:  [Discussion of Palmer mercury exposure study in Texas where mercury exposure is inferred from EPA data available at a county level; in the results section, Palmer says that the mercury vapor emitted by factories in 1998 was detectible in autism diagnoses in 2002 when the kids were age 4.  He states there is an incidence-risk ratio that for every 1,000 pounds of release in 1998, there was a 2% rise in 2002 autism rates.  Finally, they add distance to the equation, they show that for every 10 miles away from the source, the rate of autism diagnosis goes down.] 

Q:  So if someone were trying to evaluate the plausibility of a biological link between autism and mercury, would you put some weight on the Palmer study?
A:  Yes, but not too much.

Q:  In your report, do you blame the families for bringing these claims?
A:  I’m never blaming the families; when orthodox medicine doesn’t have an answer, people look elsewhere.  I don’t blame the parents at all.

Q:  You started working with the vaccine manufacturers about 4 years ago, and then you say you started working on the MMR litigation in 2003.  In 2005, you published a paper on the MMR withdrawal in a population in Japan with Dr. Honda.  There was no disclosure in the paper that you had been retained in the MMR litigation.  Did you make a disclosure about that?
A:  Presumably not, because it would have been in the paper.

Q:  Did you just testify recently in a hearing against Dr. Wakefield?
A:  I did indeed, but that was in somewhat different circumstances.  He was presenting information on his cases, claiming a particular cause and effect.  Mine is a much more indirect connection, but it’s a reasonable point.  There was certainly no attempt to conceal it.

Q:  You think you are not as tempted by conflicts of interest as other scientists are?
A:  Some are and some aren’t.

Redirect Examination

[revisiting some questions from cross examination]

Q:  What inferences can be drawn from the studies on neuroinflammation?
A:  Well, none.  They are hypothesis generating, but they don’t demonstrate causation.

Q:  Have we seen any evidence of hyper susceptibility to mercury?
A:  No.  The studies have not been sufficient to rule out such a group, but they don’t point to it, either.

Q:  You were also asked if further resources should be used to follow up the Verstraeten study.  Is that just an economic issue?
A:  No; one wants to put resources where they are likely to pay off.  We’ve talked about thimerosal being a causative factor.  The evidence that it’s worth doing further research is not persuasive.  I’m much more positive about the effects of mercury coming from other sources, such as fish. 

Q:  You were asked a series of questions about the Palmer study.  Does that have any relevance to the question of whether mercury in vaccines causes autism?
A:  I do see the research as worthwhile, but the connection with thimerosal is indirect, putting it mildly.  It’s an interesting hypothesis generating study, but in itself it doesn’t take us very far on mercury generally, and it doesn’t take us anywhere on thimerosal.

Q:  Mr. Williams asked you about your role in the Honda paper on thimerosal in Japan.  And he asked you about your testimony in the Wakefield matter.  Is there anything else you’d like to say about that?
A:  With the benefit of hindsight, I can see it might have been prudent to disclose my involvement in the litigation.  In the British situation, it’s not like a conflict here; I was testifying for the court, not for a party.  For Wakefield, he was funded to do the study; many of the cases he was studying were in the litigation, so the interests were direct and they were concealed.  In my case, I certainly never intended to conceal it.  It is different from the Wakefield situation where there were direct financial and case connections.


Q:  Who was paying you?
A:  The drug company was paying me; but you’re actually not responsible to the lawyers who call you, you’re responsible to the court.

Special Master Hastings

Q:  I wondered if you had any further comments on the Young, Geier and Geier and study.  Did you have the chance to read it?
A:  I think it’s a poor study; they put cases in inappropriate groups; it was a time trend study.  I do not see it as a study worth very much.  Dr. Fombonne will surely comment on it.

Q:  You were asked about Petitioners’ theory of neuroinflammation being caused by inorganic mercury.  As I look at that theory, there are two parts to it.  Inorganic mercury can cause neuroinflammation, and neuroinflammation can cause autism.  Do you see either part as more meritorious than the other, or are they both equally unmeritorious?
A:  Inflammation is a very non-specific process; it’s a bit like a fever.  The number of causes is enormous.  It is indicating a non-specific response to something going wrong.  The question is, it is more than that? 

If one takes fevers and infections as an example, you need the inflammation to gear up the body to fight the infection.  It’s a good process because it’s part of the body’s defenses.  But if you ask whether it’s thimerosal that’s causing this, you’d have to figure out that it comes from the thimerosal and not the amalgams or the factory up the road.  You’ve got the further problem that if it’s occurring through the brain, why does it lead to the pattern you have with autism – not just the symptoms, but the increase of head size, the particular kind of abnormalities encapsulated by “theory of mind;” the idea that neuroinflammation or oxidative stress plays a role.  The ideas make sense, but you have to ask if it’s at work here.

Mary Holland teaches at New York University Law School and may be contacted HERE. Thank you to A-CHAMP for providing this important service.


Feed You can follow this conversation by subscribing to the comment feed for this post.


My favorite bit is the combination of two contradictory opinions:
1)We can't figure out if autism is increasing or not
2)We can do epidemiological studies of vaccines and autism and see that autism is still increasing and therefore couldn't be caused by vaccines.
If you can't tell if it is increasing, then epidemiological studies which are based on the numbers going up aren't valid.

David Taylor

As always, Mary, thank you very much. This man was publically undressed and humiliated during cross and re-cross.


Q: You were also asked if further resources should be used to follow up the Verstraeten study. Is that just an economic issue?
A: No; one wants to put resources where they are likely to pay off. We’ve talked about thimerosal being a causative factor. The evidence that it’s worth doing further research is not persuasive. I’m much more positive about the effects of mercury coming from other sources, such as fish.

We should not study the effects of injecting mercury directly into babies and children but we should study the effects of mercury in fish?!?! I guess he knows on which side his bread is buttered.

And it is o.k. for Dr, Rutter not to disclose conflicts of interest, but it is a whole 'nother matter in the case of Dr. Wakefield?

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been saved. Comments are moderated and will not appear until approved by the author. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Comments are moderated, and will not appear until the author has approved them.

Your Information

(Name and email address are required. Email address will not be displayed with the comment.)