|
|
Mercury and Autism in the United Kingdom
By
John Stone
(download pdf)
Part
III: The PHLS Study Reviewed
By the time the U.K. Public Health Laboratory Service (PHLS) study was published in September 2004, Elizabeth Miller’s department had become the Communicable Disease Surveillance Centre, Immunization Department. The authors disclosed no competing interests. However, independent researchers Mark and David Geier wrote to Pediatrics:
“The authors of the Andrews et al. study failed to disclose their significant conflicts of interests to the readership of Pediatrics: Elizabeth Miller disclosed in her 2001 publication (1) and in 2002 to the Committee on the Safety of Medicines previously disclosed that she has received funding to study vaccines from Aventis Pasteur, Wyeth Vaccines, SmithKline Beecham, Baxter Health Care, North American Vaccine, Wyeth-Lederle Vaccine, and Chiron Biocine; and Nick Andrews, Julia Stowe, and Brent Taylor all disclosed in 2001 that they had received funding to study vaccines from Wyeth Vaccines and SmithKline Beecham. (1) These companies all are or were makers of thimerosal-containing vaccines.” (1)
To which Elizabeth Miller, Nick Andrews and Brent Taylor replied: “The requirements by Pediatrics for the conflict of interest declaration were complied with by the authors and the Health Protection Agency’s policy on the condition under which commercial funding is obtained for studies in [sic] available on our web site. (2)
In other words, the authors did not deny these competing interests, although they had not declared them. They do not claim either to have adhered to the Health Protection Agency code, but merely allude to its existence. It is perhaps even more concerning that officials who bear responsibility for implementing the policy they are reviewing do not declare this as an interest, as if by implication they are above having an interest. This is a tactic that deceives a good many journalists and politicians, but it is by no means the reality. In fact, it makes little sense that they should be in the position of reviewing the policy at all.
The conclusion of the PHLS study as stated in the abstract was succinct:
“With the possible exception of tics, there was no evidence that thimerosal exposure via DPT/DT vaccine causes neurodevelopmental disorders.” (3)
This, in itself, restricts the scope and value of the study, and poses certain questions. As noted in part I, Miller failed to deal with the claim of the Sunday Times that there was “increasing use of thiomersal-containing vaccines in the U.K.” after 1988. Evidence was available that the vaccine schedule was quite commonly not administered in the routine way, and there would have been increased exposure from the combined use of single vaccines. It was also documented that 80 thiomersal-containing vaccines, presumably manufactured on a commercial basis, were licensed for pediatric use between 1993 and 2001. These are important confounders that the study does not even consider.
The study apparently treats all infants as if they had the same load, and had received either DT or DPT. It excludes infants who had hepatitis B or influenza shots “in the first six months because such children are likely to be an atypical sub-group,” not apparently because they would have had greater exposure. But it is also not clear what has happened to infants who received the schedule through combined single vaccines, who would have had higher dosages. These are simply not referred to. Were they included as if they had had the DT or DPT, or were they excluded? A sub-group that had had higher exposure would be of exceptional interest, but this was not studied. Salutary to consider the warning concerning the U.K. data source — the General Practitioners’ Research Database (GPRD) — given by Thomas Verstraeten of the Centers for Disease Control and Prevention, and quoted in part II:
“The quality of the comparison group is maybe even more important if you consider all the criticism we have received of comparing high T exposure to no or low T exposure. I am not sure if the GPRD is that reliable that you can be sure that low exposure is really low exposure and not underascertainment in the database.”
These are quite likely to be the remarks of someone who had been able to access the database and been able to sample its data himself.
The other issue that Miller failed to deal with satisfactorily in her letter to the Sunday Times was that of thiomersal content in the routine program, which, as she said, had not increased “over the past decade” (1991-2001) but had actually increased between 1988, the date she first mentioned in the letter, and 1991. The reply had skirted the major event of the introduction of the accelerated DPT schedule in 1990, bringing forward vaccination at 3, 5 and 10 months to 2, 3 and 4 months, representing an approximately 20 percent increase in mercury content in relation to infant body weight, delivered to significantly more immature infants. This was originally recognized as a critical issue by Miller. She had written to Robert Chen on June 27, 2001:
“The licensing authority has now definitely confirmed that the whole cell vaccine we used prior to 1996 did only contain the 50 ug thiomersal dose. This is really annoying as we checked with them several times. I need to discuss the implications of this with WHO. What is the thiomersal exposure in the Harald Heibel study because I believe they used the U.K. whole cell vaccine for their 2, 4, 6 vs 3, 5, 12 month, so even with most accelerated schedule, the Swedish children would get less exposure than our kids routinely get.”
The significance of this event became masked in the PHLS study. However, contemporaneous with the PHLS study was another study sharing four of the same authors (Miller, Taylor, Andrews and Stowe), which studied incidence of autism in a northeast London population in relation to MMR. Published in Archives of Diseases in Childhood (ADC), and submitted in December 2002, six months after the mercury study was presented to the WHO, it was published a year ahead of it in August 2003. However, neither study references the other.
The ADC paper (4) chronicles the rise in autism — based on patient records — in five northeast London districts between 1979 and 1992, arguing that an exponential trend exists over the period independently of the introduction of the MMR in 1988. Without citing evidence, it concludes that this was due to institutional changes and better diagnosis. The authors further argue that this process tailed off from 1992 and that the trend should remain stable beyond the point at which their data begins to peter out in 1996:
“The prevalence of autism, which was apparently rising from 1979 to 1992, reached a plateau at a rate of 2.6 per 1,000 live births. This levelling off, together with reducing age at diagnosis, suggests that the earlier recorded rise in prevalence was not a real increase, but was likely due to factors such as increased recognition, a greater willingness on the part of educationalists and families to accept the diagnostic label, and better recording systems. The proportion of parents attributing their child’s autism to MMR appears to have increased since August 1997.”
There was, in fact, no evidence presented whatsoever for the conclusion that the trend was a cultural phenomenon (as I pointed out on ADC Online at the time, without response (5)). One confounding factor, which the authors might have taken into account given that they were simultaneously engaged on a thimerosal study, was that diagnosed incidence of childhood and atypical autism (excluding Asperger Syndrome) rose from 22 in the birth cohort of 1989 to 36 in 1990 and 46 in 1991, a rise of 109 percent synchronous with the introduction of the accelerated DPT schedule. And yet right at the outset, reporter Rosie Waterhouse had aggravated Miller by her suggestion in the Sunday Times that:
“A key factor may be that mercury-based vaccines are being given to babies as young two months old, when their bodies are less able to cope with it….”
But the data was available to health officials — the very same health officials — and they failed to acknowledge it.
* * * *
The U.K. General Practitioners’ Research Database used in the PHLS study (8) is an inadequate tool, both quantitatively and qualitatively, to research neurodevelopmental disorders. The authors of the PHLS study, amid their opaque and unverifiable data, include only 104 cases of autism among term infants, and two among pre-term infants, representing an incidence of barely more than 0.1 percent. By comparison, a survey from the Office of National Statistics offers a much more plausible one percent incidence of autistic spectrum disorders (ASD) among children born between 1988 and 1999. (6) The question further arises with the PHLS figure as to how much of it is interpretation. The study of Hershel Jick and James Kaye came up with much lower figures still, albeit rising dramatically from 1.6 boys in 10,000 in 1993 to 9.5 boys in 10,000 in 1999. Supposing four ASD boys to one ASD girl, this would amount to 1 in 10,000 in 1993 to 6 in 10,000 in 1999, so that in 1993, the median year of the PHLS study, they perhaps were only picking up one-tenth as many cases as the PHLS total using the same database. (7) This is a wide and unacceptable margin of error.
Another British government-sponsored study into autism and MMR testifies to the ramshackle nature of the data included in the GPRD:
“Medical notes for 318 subjects were obtained. They varied in quality and exhaustiveness. For some children, GP records included several consultant reports, speech and language assessments, and educational psychology reports. For other children, the information available was scanty, with sometimes the only available data consisting of one, or a few, letters between the GPs and consultants. A high proportion of records had missing data on parental age, socio-economic status, and detailed psychometric assessment of the child and, therefore, the frequencies of these variables are not described here. Of the 318 children whose medical forms were obtained, the raters confirmed a diagnosis of PDD (pervasive developmental disorders) in 294 children (92.5 percent). Compared to children with a confirmed PDD diagnosis, children for whom the diagnosis was not confirmed (n = 24) had significantly fewer PDD symptoms (2.1 vs 6.2; p < .001), higher language level (phrase speech: 80% vs 45%; p=.051), and more frequent parental concern arising for the first time after the age of 3 years (20% vs 2.9%; p=.024).” (8)
It might be considered after initial assessment that drawing any significant conclusions from a database of this quality in this particular field would be rash. This, however, failed to stop a succession of studies related to the vaccine/autism issue from proceeding. Apart from the PHLS study, these included three by the Boston Collaborative Drug Surveillance Program, (9, 10, 7) and four studies by Smeeth et al backed by the U.K. Medical Research Council. (8, 11, 12, 14)
Serious questions must arise as to the wisdom and motivation behind these enterprises, which were extensively reviewed by John Heptonstall in BMJ Rapid Responses and online Pediatrics. (14, 15) There are, in fact, too many problems with the database even before starting to consider issues of method and transparency: for instance, there is the likelihood that the vast bulk of the cases are lost in the control group and adversely weighting the sample. Moreover, the PHLS study takes the precaution of removing all the most vulnerable clinical cases as stated under the heading “exclusion criteria:”
“Children with Read and OXMIS codes relating to a variety of prenatal, perinatal and postnatal conditions that occurred before 6 months of age were excluded as were children who were recorded as having an outcome event in the first 6 months of life. These children were excluded from the main analysis because the presence of such a condition is likely to affect both vaccination and future neurodevelopmental outcomes. Examples of exclusions were birth asphyxia, cerebral palsy, meningitis, encephalitis and head injury. Children were also excluded when they received either hepatitis B or influenza vaccination in the first 6 months of life because such children are likely to be an atypical subgroup. Children who were born preterm (<37 weeks’ gestation) are likely to be of low birth weight and many stay small. Such infants might be more susceptible to standard doses of thimerosal. Preterm infants were therefore analyzed separately.”(3)
Since the listed exclusions are on their own testimony only “examples,” this is the epidemiological equivalent of a blank cheque. Significantly, perhaps, when authors Andrews and Miller attempted to answer criticism by Heptonstall, they did not respond to point 12 in his online letter to Pediatrics of Oct. 30, 2004: (15)
“12. The excluded group of children, postnatal and ‘outcomes,’ who might have suffered mercury damage from vaccines — a tiny fraction of the 103,043 cohort but a fair proportion of approximately 5,000 cited as having outcome conditions. Andrews et al state the exclusions were made “because the presence of such a condition is likely to affect both vaccination and future developmental outcomes.” Aren’t these the very children the study should have focussed on?”
* * * *
The most illuminating response I received from a Freedom of Information inquiry came in reply to questions fired off in irritation to the U.K. Department of Health after having an earlier query passed on to the licensing authority, the MHRA. Aside from the collection of interesting material supplied were these two astounding answers (June 1, 2005):
“Q: At what point was it that
the Department of Health became aware that
there might be a problem with thiomersal/thimerosal?
“A: The Department of Health is not aware of problems with Thiomersal, only that concerns had been raised in the U.S. based on the U.S. immunization schedule and the EPA recommendations. The Department's interest followed the U.S. information that was in the public domain.
“Q: Who knew, and when?
“A: This was public knowledge amongst numerous Department of Health officials from the time of publications from the U.S.”
In other words, officially speaking, the hypothesis that mercury vaccine could cause autism was never, ever considered. It was not considered in 1999, in 2001, 2002, 2004 or 2005. It apparently never will be.
References:
| 1. | Geier M, Geier D. Thimerosal does not belong in vaccines. Pediatrics, Sept. 8, 2004. Available at: http://pediatrics.aappublications.org/cgi/eletters/114/3/584 Accessed Dec. 11, 2005 | |
| 2. | Miller E, Andrews N, Taylor B. Re: Thimerosal does not belong in vaccines. Pediatrics, Sept. 10, 2004. http://pediatrics.aappublications.org/cgi/eletters/114/3/584 | |
| 3. | Andrews N, Miller E, Grant A et al. Thimerosal Exposure in Infants and Developmental Disorders: a retrospective cohort study in the United Kingdom does not support a causal association. Pediatrics 2004;114:584-591. http://pediatrics.aappublications.org/cgi/content/full/114/3/584 | |
| 4. |
Lingam R, Simmons A, Andrews N,
Miller E, Stowe J, Taylor B. Prevalence of autism and parentally reported trigger in a northeast London population. ADC 88;2003:666-70. http://adc.bmjjournals.com/cgi/content/full/88/8/666 |
|
| 5. | Stone J. Prevalence of autism; no evidence for the conclusion. ADC Online, Aug.8, 2003. http://adc.bmjjournals.com/cgi/eletters/88/8/666#529 | |
| 6. | National Statistics; U.K. Snapshot: prevalence of children’s mental disorders 2004, GB: http://www.statistics.gov.uk/cci/nugget.asp?id=1229 | |
| 7. | Jick H, Kaye J. Epidemiology and possible causes of autism. Pharmocotherapy: 2003;23:1526-1530. | |
| 8. | Fombonne E, Heavey L, Smeeth L et al. Validation of the diagnosis of autism in general practitioners records. BMC Public Health 2004;4:5. http://www.biomedcentral.com/1471-2458/4/5 | |
| 9. | Kaye JA, Melero-Montes M del M, Jick H. Mumps, measles and rubella vaccine and the incidence of autism recorded by general practitioners: a time trend analysis. BMJ 2001;322:460-3.http://bmj.bmjjournals.com/cgi/content/full/322/7284/460 | |
| 10. | Black C, Kaye JA, Jick H. Relation of childhood gastrointestinal disorders to autism: nested case-control study using data from the U.K. General Practice Database. BMJ 2002;325:419-21. http://bmj.bmjjournals.com/cgi/content/full/325/7361/419 | |
| 11. |
Smeeth L, Hall AJ, Fombonne E et
al. A case control study of autism and mumps measles rubella vaccination using
the general practice research database. BMC Public Health 2001;1:2.
 http://www.biomedcentral.com/1471-2458/1/2 |
|
| 12. | Smeeth, L, Cook C, Fombonne E et al. MMR vaccination and pervasive development disorders: a case-control study. The Lancet 2004;364: 963-69. | |
| 13. | Smeeth L, Cook C, Fombonne et al. Rate at first recorded diagnosis of pervasive developmental disorders in United Kingdom general practice, 1988-2001. BMC Medicine 2004;2:39. http://www.biomedcentral.com/1741-7015/2/39 Accessed Sept. 1, 2005. | |
| 14. | Rapid Responses: Mayor S. Study shows no link between MMR vaccination and autism. BMJ:2004;329. http://bmj.bmjjournals.com/cgi/eletters/329/7467/642 | |
| 15. | P3Rs: Andrews et al. Thimerosal exposure in infants etc. Pediatrics : 2004;114. :http://pediatrics.aappublications.org/cgi/eletters/114/3/584 |
First published on Red Flags