Objective. After concerns about the possible toxicity of thimerosal-containing vaccines in the United States, this study was designed to investigate whether there is a relationship between the amount of thimerosal that an infant receives via diphtheria-tetanus-whole-cell pertussis (DTP) or diphtheria-tetanus (DT) vaccination at a young age and subsequent neurodevelopmental disorders.
Methods. A retrospective cohort study was performed using 109 863 children who were born from 1988 to 1997 and were registered in general practices in the United Kingdom that contributed to a research database. The disorders investigated were general developmental disorders, language or speech delay, tics, attention-deficit disorder, autism, unspecified developmental delays, behavior problems, encopresis, and enuresis. Exposure was defined according to the number of DTP/DT doses received by 3 and 4 months of age and also the cumulative age-specific DTP/DT exposure by 6 months. Each DTP/DT dose of vaccine contains 50 μg of thimerosal (25 μg of ethyl mercury). Hazard ratios (HRs) for the disorders were calculated per dose of DTP/DT vaccine or per unit of cumulative DTP/DT exposure.
Results. Only in 1 analysis for tics was there some evidence of a higher risk with increasing doses (Cox's HR: 1.50 per dose at 4 months; 95% confidence interval [CI]: 1.02–2.20). Statistically significant negative associations with increasing doses at 4 months were found for general developmental disorders (HR: 0.87; 95% CI: 0.81–0.93), unspecified developmental delay (HR: 0.80; 95% CI: 0.69–0.92), and attention-deficit disorder (HR: 0.79; 95% CI: 0.64–0.98). For the other disorders, there was no evidence of an association with thimerosal exposure.
Conclusions. With the possible exception of tics, there was no evidence that thimerosal exposure via DTP/DT vaccines causes neurodevelopmental disorders.
Comments
Re: Re: Please can we have correct information about dosage?
Elizabeth Miller and Nick Andrews’ September 5th reply of to my P3R of 12 June is noted.
I originally wrote on April 24, 2005 to the United Kingdom Department of Health (DH) to inquire through a Freedom of Information request about the historic levels of thimerosal in DPT vaccine, concerned by evident lacunae in Dr Miller’s remark to the Sunday Times of July 15, 2001:
“Your articles, Autism linked to mercury vaccine (May 27) and Inquiry launched into vaccine ‘link’ with autism (June 17) implied that there has been increasing use of thiomersal-containing vaccines in the UK since 1988. In fact, the thiomersal content of vaccines given in the routine vaccination programme has not increased over the past decade.”
The DH wrote back on May 16 that my request was being transferred because it could “more appropriately be responded to by the Medicines and Health products Regulatory Agency” (MHRA). Since an FOI request is supposed to be answered within 20 working days of application, it was apparent that this one was proving an embarrassment as Dr Philip Bryan of the MHRA - replying to me on a connected query on June 27 -wrote:
“…I pointed out in my previous letter to you that a number of UK licensed vaccines have had levels of thiomersal reduced or removed completely from the manufacture of the component antigens or from final vaccine in accordance with EMEA advice. We are currently considering your request to make documents available to you [a reference to my DH request of 24 April] and my colleagues will write to you on this matter within 10 days.”
It is remarkable - if as now claimed the dosage remained unchanged between 1988 and 2001 - that there should have been any difficulty. As it is the said colleagues did not get in touch and I next heard on the matter from Dr Bryan, himself, who wrote to me on August 3 explaining:
“A subsequent review of the content of thiomersal-containing vaccines revealed that the ‘150µg’ calculation was actually based on an incorrect assumption that the ethylmercury content of this vaccine brand was 50µg per dose when it was in fact 50 µg per 1ml (i.e. 25 µg per 0.5 ml dose). Ethylmercury exposure through 3 doses of DTP vaccination by 4 months of age therefore did not exceed 75µg. I hope this clarifies the confusion.”
and he elaborated on September 5:
“This information relating to Trivax AD was supplied some time ago in response to an urgent query at a time when many products containing thiomersal had to be identified”.
Presently we lack any documents which could give rise to such an ambiguity. Moreover, the references to the higher dose exist in Joint Committee on Vaccination and Immunisation (JCVI) documents of 1999 and 2000, before this became a matter of open controversy in the UK, while Elizabeth Miller still seemed to be covering herself on the topic on July 15, 2001 in the Sunday Times. It must also be pointed out that the JCVI document from 1999 gives the measurement as a percentage of volume as well as weight:
“A solution of 0.01” or 0.02% thiomersal is commonly used in vaccines”
which is inconsistent with the present explanation which claims there was an error of weight to volume.
It is perplexing that the head of the UK Government vaccination programme should apparently have incorrect information concerning the composition of the products which her department has researched and promoted, or that she should be in anyway reliant on another government department for technical information on the content. And in the end, we are still left with the problem that there are historical documents saying one thing and recent undocumented official assurances claiming the opposite.
Conflict of Interest:
Autistic son
Re: Please can we have correct information about dosage?
We write to clarify the content of thiomersal in UK DTP vaccines between 1988 to 1997. In our article we give the dosage as 50 micrograms of thiomersal (25 micrograms of mercury) per dose [1]. We can confirm that this information is correct and there was no change in the dosage during this period. The Medicines and Healthcare products Regulatory Agency (MHRA) have recently written to Mr Stone clarifying this fact and explaining that there was initially some confusion about the thiomersal content of one DTP vaccine that led to the misunderstanding that exposure may have been twice as high for that vaccine. We were very careful when designing and analysing our study to ensure we knew the correct thiomersal content.
[1] Thimerosal Exposure in Infants and Developmental Disorders: A Retrospective Cohort Study in the United Kingdom Does Not Support a Causal Association. Pediatrics vol.114 No.3 September 2004, pp.584-591.
Conflict of Interest:
None declared
Please can we have correct information about dosage?
I present new evidence obtained under the UK Freedom of Information Act which bears on the credibility of this article [1] and the companion article by Heron and Golding [2].
I have already drawn attention in previous P3Rs to the prejudicial letter of Elizabeth Miller to the London Sunday Times of 15 July 2001 and its reference to “theoretical concerns that the small amount of mercury in thiomersal could be harmful” [3, 4]. I have also made this point in greater detail in relation to the Heron and Golding study and its reference to “low doses” of mercury in the UK vaccination schedule [5]. Both studies give the exposure as 25 micrograms of ethyl mercury in three doses administered at 2, 3 and 4 months, and in the case of the Andrews study 3,5 and 10 months prior to 1990. The Andrews study was based on UK wide data drawn from the General Practiotioners Research Database in the years 1988 to 1997. The Heron study dealt with children delivered in the years 1991-2 in the county of Avon.
However, information made available the UK Department of Health suggests that in many cases the dosage was actually double that stated in these papers. For a normal weight two month old – based on my previous calculations – the weight of mercury could have 132 times the Environmental Protection Agency reference dose cited [5]. These would have been controversially large doses by any standard.
A Joint Committee on Vaccine and Immunisation minute of 9 October 2000 states:
"The estimated potential thiomersal exposure through the UK programme was calculated to range between 0.15 and 0.30 mg (equivalent to 75-150 micrograms of mercury)"
I have also been provided with a Medicines Control Agency (MCA later merged into the present MHRA) briefing for dealing with Sunday Times enquiries dated 7 June 2001 which was perhaps in part preparatory to Elizabeth Miller’s letter to the paper of 15 July 2001:
"Thiomersal-containing vaccines have been in use for over 60 years and evidence does not support a causal link with autism. Indeed, reported rates of autism have been continuing to rise over the past decade as thiomersal content in routine UK childhood programme has fallen."
According to this the thiomersal content fell between 1991 and 2001 but the period covered begins in 1988 and ends in 1997. (Very interesting to read also that “reported rates of autism have been continuing to rise over the past decade”, although this has never been officially admitted in the UK.) Elizabeth Miller covers all this in her letter to the Sunday Times by saying:
"In fact, the thiomersal content of vaccines given in the routine vaccine programme has not increased over the past decade" (1991-2001)
But she does not indicate that it was actually reduced, or when – nor is this mentioned in either of British Government studies published here. This, I believe, casts doubt on the reliability of the information provided in these studies. I believe at the very least that clarification is now essential, and detailed information ought to be made available regarding infant mercury exposure in the UK in the past two decades. There can be no grounds for holding back.
[1] Nick Andrews, Elizabeth Miller, Andrew Grant, Julia Stowe, Velda Osborne, and Brent Taylor, 'Thimerosal Exposure in Infants and Developmental Disorders: A Retrospective Cohort Study in the United Kingdom Does Not Support a Causal Association', PEDIATRICS Vol. 114 No. 3 September 2004, pp. 584-591 (doi:10.1542/peds.2003-1177-L)
[2]Jon Heron, and Jean Golding, and the ALSPAC Study Team,'Thimerosal Exposure in Infants and Developmental Disorders: A Prospective Cohort Study in the United Kingdom Does Not Support a Causal Association', PEDIATRICS Vol. 114 No. 3 September 2004, pp. 577-583 (doi:10.1542/peds.2003-1176-L)
[3]'Puzzling circumstances', 1 May 2005 http://pediatrics.aappublications.org/cgi/eletters/114/3/584#1356
[4] 'Re: Heron and Golding: erroneous premise, anomalous results - an update' 3 May 2005, http://pediatrics.aappublications.org/cgi/eletters/114/3/577#1358
[5]'Heron and Golding: erroneous premise, anomalous results' 19 April 2005, http://pediatrics.aappublications.org/cgi/eletters/114/3/577#1346
Conflict of Interest:
Autistic son
Puzzling Circumstances
I wonder whether Nick Andrews, Elizabeth Miller, Andrew Grant, Julia Stowe, Velda Osborne and Brent Taylor would like to comment on a concatenation of puzzling circumstances. When this study [1] was published in September of last year the UK Department of Health were in the process of phasing out the use of thiomersal/ thimerosal in DPT vaccine. As I recall the announcement in the change of policy was made in August 2004 and implemented from October 2004. The publication date in between these two events is in itself remarkable, not least bearing in mind that it is recorded in the published conclusion:
"The results of the 2 United Kingdom studies were presented to the WHO Global Advisory Committee on Vaccine Safety in June 2002. These studies contributed to the conclusion that there is currently no evidence of mercury toxicity in infants, children, or adults who are exposed to thimerosal in vaccines and that there is no reason to change current immunization practices with thimerosal-containing vaccines on grounds of safety. This conclusion is particularly important for developing countries that administer thimerosal-containing DTP vaccines according to the expanded immunization schedule."
Given the apparent good news this seems like a long time to keep the world in the dark. For instance I recall a BBC Radio programme: 'File on Four' concerning thimerosal and autism (June 24, 2003) [2] when none of this was mentioned, although Elizabeth Miller had mentioned in a letter to the London Sunday Times of July 15, 2001:
"The only vaccines for children used in the routine programme that contain thiomersal are DTP (diphtheria, tetanus, pertussis) and DT. Because of theoretical concerns that the small amounts of mercury in thiomersal could be harmful, both European and United Kingdom regulators have recommended that manufacturers phase out its use wherever possible as a precaution. As a further precautionary measure the Public Health Laboratory Service, on behalf of the World Health Organisation, will be undertaking research into any negative effects of thiomersal-containing vaccines in the near future."
It was therefore surprising to read the following last week in a BBC report:
"a spokeswoman for the Medicines and Healthcare products Regulatory Agency (MHRA), the UK drugs regulator, said mercury was being phased out of the UK vaccine programme after advice from the World Health Organisation." [3]
Thus in July 2001 the UK regulator (then called Medicines Contol Agency which was later merged into the MHRA) was trying to phase out thimerosal with the co-operation of the manufacturers, though nothing happened for a further three years during which time in my experience public and parental concern was being high-handedly disregarded in the UK. The BBC 'File on Four' report of June 2003 [2] bears testimony not only to the UK Department of Health's tight-lipped approach during this period but also to the fact that the project of phasing out mercury mentioned in July 2001 by Elizabeth Miller was referred to neither by the BBC reporter, Gerry Northam, or in the Department of Health statement. It seemed simply to be off the agenda, and forgotten.
I note also that already by July 2001 the WHO had asked Dr Miller's Public Health Laboratory to investigate the negative effects of thimerosal, and the present paper is apparently the result. However, Elizabeth Miller's team had according to their information reported to WHO in private as long ago as June 2002 and presumably the WHO accepted their advice. It is therefore astounding to read that the UK agencies have been withdrawing thimerosal from infant vaccine on the advice of the WHO [3]. But then you wonder why also that the very people who advised thimerosal should continue to be used on the children of the developing world are withdrawing it from use in the UK.
As we know the policy operated apparently blindly for many years before the present plethora of re-assuring epidemiological studies were undertaken and published. All this is a long way from the transparency desirable in such matters and a cavalier way to treat people legitimately concerned about the well-being of their children over a period of years. It also looks prejudicial in the UK context that the concern having been identified the substance was not at least removed from vaccine pending further research. In the context any positive findings would have been a massive embarrassment. Might I suggest that if the authors had wanted the present study to be taken on trust that a little more humility and caution would have been in order.
[1]'Thimerosal Exposure in Infants and Developmental Disorders: A Retrospective Cohort Study in the United Kingdom Does Not Support a Causal Association', PEDIATRICS Vol. 114 No. 3 September 2004, pp. 584-591 (doi:10.1542/peds.2003-1177-L)
[2] BBC programme No 03VY3025LHO, Reporter: Gerry Northam.
[3] http://news.bbc.co.uk/1/hi/health/4472485.stm
Conflict of Interest:
Parent of an autistic child
Re: Re: Does weight confound?
Sir
Andrews and Miller say they were “investigating whether the cumulative exposure to ethyl mercury is associated with long-term developmental disorders” yet their study Objective says “after concerns about the possible toxicity of thimerosal-containing vaccines in the US, this study was designed to investigate whether there is a relationship between the amount of thimerosal an infant receives via DTP or DT vaccination at a young age and subsequent neurodevelopment disorders”. The Objective does not specify “long term” but “subsequent” developmental disorders of which one would necessarily include acute or chronic, with sequellae that may involve short or long term morbidity or mortality. The US EPA “stringent” (to quote Andrews et al) maximum exposure limit for mercury is 0.1ug/kg/day; adducing that daily, not cumulative, exposure should be a basic requirement for assessment of toxicity.
The data collected does not appear capable of evidencing the Objective, or the surprisingly obtuse Conclusion that “there is no evidence that thimerosal exposure via DTP/DT vaccines causes neurodevelopmental disorders”, or that “ there is currently no evidence of mercury toxicity in infants, children or adults who are exposed to thimerosal in vaccines”.
Responses to specific points by Andrews and Miller:
Points 1, 8, 16. They say they “do not conclude” that thimerosal can have a true protective effect but ‘protective effect ‘ appears several times as “apparent protective effects from DTP/DT exposure” (‘Risk Estimates’), “showed a protective DTP/DT effect” (‘Risk Estimates’), “an apparent protective effect from increasing thimerosal exposure” (‘Discussion’), “early thimerosal exposure generally showed no association or was protective” (‘Discussion’). These statements might influence the reader to conclude that ethyl mercury is protective to infants and young children. Further, they say “The ‘reducing trend’ is not by age but ‘by exposure at a given age’ so cannot be explained by protective effects of increased body weight” but lack of data in the GPRD on an individual’s weight at vaccination makes it difficult for them to predict with confidence ‘reducing trend by age’; the ‘increasing trend by exposure at a given age’, possibly by weight, shown in the US VSD study that also ignored infants’ weight at vaccination adds to the concern that weight influenced toxicity..
Point 2. They say, “The increased risk of developmental disorders in pre-term children compared to term children is not surprising”. Can they validate that statement through studies that did not fail to take into consideration the possible effects of thimerosal-containing, and other, vaccines given to pre-term and full-term infant subjects? The only studies I have seen which compare pre-term and full-term children with developmental disorders ignore the possible effects of vaccination on the subjects. Further, is the data on pre-termers confounded by the 20% ‘invalidity’, the lack of size (in number and weight) of the pre-term cohort, the small number for some outcomes, a cohort admitted to be “not large enough to have the power to identify small effects” and which has children who might have avoided doses of vaccine due to adverse reactions? Has the study too little power or validity to assess risk of developmental disorders in pre-term children by thimerosal exposure?
Point 3. Batch and brand do matter - one should compare like with like. Although each vaccine may contain 25 ug of ethyl mercury it is impossible from the GPRD to ascertain if children received identical or even similar vaccines – the material at the heart of this study - unless batches and brands are identified and sampled The presence of other heavy metals, antibiotics, etc. can enhance the toxicity of thimerosal so synergistic toxicities must be considered. I do not think any expert in chemistry would refute that mercury, when combined with aluminium (used as an adjuvant in many vaccines), becomes much more toxic. Although the GPRD might not hold detail about brand or batch the information is available from other sources and it has long been suggested that certain batches and brands increase potential mortality and morbidity. Difficulty in access to such data should not diminish its desirability.
Point 4. What proof is there for “There is no evidence of an association between developmental disorders and any other vaccines”? There are recorded ADRs to vaccines which though not immediately informative of a neurodevelopmental disorder (which can take years to establish physically and diagnostically) are known to lead to such diagnoses eg. epilepsy, encephalitis, meningitis, apparent behaviour problems, apparent visual or hearing problems especially occurring after vaccination in infants for whom ‘normality’ might be some way off being clinically established. Further, vaccines given along with DTP and DT contain adjuvants that modify the toxicity of mercury so other vaccines can confound.
Point 6. Heron et al should not be used to support Andrews et al, as it is not a like with like comparison. Heron et al is based on a cohort from one specific part of England, the Avon area with it’s own idiosyncrasies and environment. They did not establish clinically the cognitive or behaviour development of subjects but relied on mothers completing questionnaires over a 7-year period – the potential for error must be enormous and such data ought not to be used to validate any study which requires, for children, accurate diagnostics in relation to medical interventions, in this case vaccination. Neurodevelopmental data is difficult enough for professionals to analyse and assess (1-4), let alone mothers. One must suspect that Andrews et al were attracted more by the convenience that the appropriateness of the Avon study data.
Point 7. Acute reactions to mercury might include death (SIDS) or a neurological event with or without sequellae such as single or multiple seizures, meningitis or encephalitis, all of which can lead to short and long term neurological illness. Excluding all children who suffered possible acute reactions, from the main cohort, must impact on the results as their “long term” outcomes are ignored. “The presence of such a condition is likely to affect both vaccination and future neurodevelop- mental outcomes” admits that exclusion leads to bias. The excluded children represent the expectations of the US EPA limit for mercury toxicity. I do not think the statement “None of the recognised acute reactions to DTP are associated with subsequent developmental disorders” is true, for example infantile spasms has long been associated with DTP vaccination and though relatively rare it has been associated with the development of autism, not least in my son Epilepsy, which DTP is believed to cause, can result in neurodevelopmental problems. Further, what evidence do Andrews and Miller have for ”Of those who fail to complete very few are likely to do so due to acute reactions”?
Points 9 and 12. Surely the published “Objective” and “Conclusion” require prior assessment of both “possible long-term effects” and “those occurring in the first 6 months” to be valid?
Point 10. Low birth weight is shown to be a risk factor in Andrews et al when pre-termers fared worse than full termers; smoking is a known risk factor for SIDS, which is historically linked to DTP vaccination, and has implications for the child of low birth weight. Weight during exposure to mercury is a risk says the EPA. Full termers in the study lost risk with age (therefore weight) so it cannot be correct to assume “there is no evidence that thimerosal in vaccines is a risk factor in low birth weight children”.
Point 11. Heron et al also excluded confounders such as mothers’ fish diet and dental status during pregnancy. Breastfeeding confounds both studies, and their response to me.
Point 14. I agree “splitting outcomes into too many small subgroups would lead to very small numbers for the analysis” but that is exactly what has been done with disparate subgroups already used. Should all those described as neurodevelopmental disorders be so described? Yet those descriptions are fundamental to the study. How are those diagnoses of “subsequent neurodevelopmental disorders” validated? The low validity exercise Andrews et al performed and reliance on a Jick et al 1991 are inadequate as explained in more detail at 17 below. Heron et al relied on questionnaires completed by mothers over a 7-year period so is also not supported by solid clinical assessments for its subjects.
Point 17. The validation exercise involved 162 record responses, 10 failed to provide any information (an immediate loss of confidence in the GPRD recording process of about 6%), the 152 records received experienced another 20% failure in validation leaving 122. The 122 relied on confirmations by GP of their own recording on the GPRD – therefore is open to bias. The published validation process says it validated dates not diagnoses, and reference 15 is Jick HJ et al 1991 yet 1991 was long before the GPRD existed. The original database, VAMP of c1987, was much more limited with numerous well-published problems (5). I doubt it can validate the GPRD, which it later became, for Andrews et al Sept. 2004.
In summary, I believe Andrews et al too often chose convenience against appropriateness, has insufficient validity for its diagnoses, has no confidence in making conclusions about pre-termers, introduced unknown bias by excluding acute and short term outcomes from the main cohort, ignored other obvious confounders that introduce bias such as additional toxic loads on children through concurrent vaccines, breastfeeding, and comparatively low birth weights.
Regards
John H.
References
1. ”Exploring the borderlands of autistic disorder and specific language impairment: a study using standardised diagnostic instruments”, Bishop DV, Norbury CF, J Child Psychol Psychitr 2002 Oct; 43(7): 917-29
2. “Autism, Asperger’s Syndrome and semantic pragmatic disorder: Where are the boundaries?”, DVM Bishop, http://www.mugsy.org/bishop.htm
3. “Autism and other PDDs: exploring the dimensional view”, Myhr G, Can J Psych 1998 Aug; 43(6): 589-95
4. “Does DSM-IV Aspergers Disorder exist?”, Mayes et al, J Abnorm Child Psychol 2001 June; 29(3): 263-71
5. “Building a research database from computerised general practice records”, Tyrer et al, Journal of Informatics in Primary Care, 1996 Sept: 8-13
Re: Does weight confound?
Sir,
In our paper we are investigating whether the cumulative exposure to ethylmercury by age is associated with long-term developmental disorders. The study did not aim to answer the question about potential acute effects of individual doses of vaccine. Such acute problems would have a clear temporal relationship and none of the acute reactions caused by DTP vaccination (such as local reactions at the injection site) are associated with developmental disorders. We did address the issue of body weight by performing a separate analysis on pre-term children and also by looking at doses by 3 as well as 4 months of age. We did not have information on individual children’s weight so we could not look at exposure by weight.
Responses to specific points made by Dr Heptonstall are as follows:
Points1, 8 and16. We do not conclude that thiomersal can have a true protective effect; we believe the hazard ratios that are below one are due to confounding that we were unable to adjust for. The reducing trend is not by age but is by exposure at a given age so cannot be explained by protective effects of increased body weight.
Point2. The increased risk of developmental disorders in pre-term children compared to term children is not surprising. The analysis that looked separately at pre-term children found no evidence of an increased risk of developmental disorders by thiomersal exposure in this group. Therefore there is no evidence that the increased risk of developmental disorders in pre-term children could be explained by thiomersal exposure.
Point3. All the DTP vaccines in this study contained the same amount of thiomersal. There is no reason that the specific batch or brand is of any importance for the particular question our study aims to investigate.
Point4. We concentrated on DTP since this is the only routinely used thiomersal containing vaccine. There is no evidence of an association between developmental disorders and any other vaccines.
Point5. Missing some cases of developmental problems will lead to almost no bias for these rare outcomes (we just lose power in the analysis). The validation sample cannot therefore be regarded as tiny. When validating the cases we did identify some lack of specificity, and this is acknowledged in the paper. In the discussion the potential bias arising from a 20% false diagnosis rate is indicated as reducing a true hazard ratio per dose or 1.20 down to 1.15.
Point6. The study by Heron et al shows that effect of confounding on the outcomes examined was not large. This is clearly of relevance when interpreting our study in which we could not adjust for many confounding variables.
Point7. It is true that an acute reaction that led to refusal to complete vaccination could lead to bias in our studies, but only if such an acute reaction is expected to lead to long-term problems. None of the recognised acute reactions to DTP are associated with subsequent developmental disorders and very few individuals fail to complete DTP vaccination (2.8% in our study). Of those who do fail to complete very few are likely to do so due to acute reactions.
Points 9 and 12. We are interested in possible long-term effects, not those occurring in the first 6 months.
Point10. As already mentioned we looked at pre-term children separately and this gave similar results to the term babies. Therefore there is no evidence that thiomersal in vaccines is a risk factor in low birth-weight children.
Point11. There are many possible confounding variables, but as shown in the Heron paper these are not likely to be large enough to hide a large true effect. However, as with all observational studies it is possible that confounding may be a problem.
Point13. The validation mentioned here is that the conditions identified were indeed developmental problems that are known to occur more often in boys. It is not suggested that this validates the negative findings; just that it validates the accuracy of the diagnosis.
Point14. The outcomes chosen were based on the provisional results from the HMO study in the US (reference 6). Splitting outcomes into too many subgroups would lead to very small numbers for the analysis.
Point15. The data were analysed using survival analysis, which allows for censored data. Also see point 7.
Point 17. One of the studies we referenced when citing other validation exercises looked at autism (reference 15). We validated a large number of records (N=152) to ensure that in the majority of cases the condition reported was correct.
We hope this letter clarifies the issues raised in the letter by Dr Heptonstall.
Regards
Nick Andrews and Elizabeth Miller
Does weight confound?
Sir
Mercury, whether in methyl or ethyl form, is a known neuro/nephrotoxin; “toxicity may be similar for ethyl and methyl mercury”; “delayed-type hypersensitivity reactions from thimerosal exposure are well -recognised” Ball et al 2001. “There is an established link between exposure to mercury and impaired childhood cognitive developments and early motor skills” Heron et al 2004.
The US EPA maximum recommended level for exposure to mercury is 0.1 ug/kg/day, the WHO 0. 47ug/kg/day. Each shot of DTP/DT vaccine contains 25ug of ethyl mercury which vaccinators inject into babies and young children despite those maximum safety limits of the EPA and WHO being expressed as weight of toxin, per weight of person, per day. A 2-month-old child weighs on average 3 to 5kg, a 3-month-old 4.3 to 6.8 kg, a 4-month- old 5 to 7.4kg, and a 6-month-old 5.9 to 8.8kg. If heavier body weight reduces toxic potential from mercury, lower weight children should be more susceptible to mercury vaccines than heavier children. Why did Andrews et al 2004 and Heron et al 2004 not assess exposure of babies and young children to ethyl mercury in vaccines according to weight of toxin per weight of child on the vaccination day, but instead assess cumulative exposure to ethyl mercury by age over time? The criterion is not consistent with that of the EPA and WHO.
The 0.1ug/kg/day EPA maximum exposure level is exceeded in babies and young children at every 25ug ethyl mercury vaccine shot by 50 to 83 times for a 1 month old baby, 37 to 58 times for a 3 month old, 34 to 50 times for a 4 month old and 28 to 42 times for a 6 month old. This exposes gross failure in authorities tasked to protect the public from toxic exposure. Furthermore,
1. Andrews et al conclude that ethyl mercury in vaccines at existing levels is safe and that it can add protective value against some adverse neurological developments as shown by a ‘reducing trend for developing ASDs with the increasing accumulation of mercury through 3 doses at monthly intervals’ in the study. Could the ‘reducing trend’ with age for developing ASDs be due to the protective effects of growth, therefore increased body weight, at successive exposures to 25ug toxic mercury?
2. Andrews et al say the preterm cohort has an “increased risk of general development disorders compared to term children” (4.2%: 2.0%). As the preterm cohort generally weigh less than term children, and were found to have an increased risk of general development disorders, was this due to lower body weight at vaccination offering less protection against mercury poisoning than term weight children? (The birth date for preterm children is usually given as day of birth, not full gestation date).
3. The UK noted organisation JABS received information from parents that their children’s ‘reactions’ to vaccines – of various kinds including neurological - varied according to brand/batch. Andrews et al and Heron et al ignore this/these possible confounders.
4. The effects of vaccine schedules superimposed on the mercury- containing DTP vaccinations eg. Hib, MMR, MMR1, MMR2, MR, Men C, OPV, DT, BCG (some of which have been implicated by parents in the onset of their childrens’ disorders by JABS) are ignored as confounders by Andrews et al.
5. ‘Validation checks’ carried out on accuracy of GPRD data (diagnoses and codes) were through correspondence with general practices and revealed a dismal 20% failure rate – this is offered by Andrews et al as a limitation of the study. Tiny ‘random subsets’ of the order of a few tens of diagnosed outcomes for analysis amongst an over 100,000 person strong cohort are used to ‘validate records’ of disorders yet errors in the GPRD may outweigh the relevance of such tiny datasets.
6. Heron et al subjects were also analysed in subsets taken from the 14,000 children in the Avon study, yet recorded levels of cognitive and behavioural development were judged from questionnaires sent to mothers at intervals between 6 months and 7 ½ years of age so were totally dependent on mothers conveying information they received, their own opinions and subjectivity – Heron et al state that “children’s cognitive and behaviour development was not assessed directly as this might ‘introduce subjectivity’” but what of mothers’ subjectivity? How can subjective values assessed by Heron et al, from a uniquely different cohort from only one relatively small area of the UK, validate Andrews et al?
7. Heron et al adjusted for birth weight only, not weight at vaccination. They state, “outcome questionnaires were less likely to be returned for children with low mercury exposure…. and whose demographic status was associated with poor developmental outcomes”. If these children suffered an adverse event for which parents refused to consent to further DTP vaccinations there is unresolved confounding.
8. Table 4 shows 95% CI for HR by dose for Tics, GDD, ADD and UDD. Other than Tics figures, the HRs for the disorders reduce at each successive dose, as therefore does the CI range; this is said to show “a reducing risk for those conditions with increasing doses”. Doesn't it actually show a reducing risk for those conditions with increasing weight/age of child?
9. Andrews et al exclude from main analysis a group consisting of those with prenatal, perinatal, postnatal and ‘within 6 months’ outcomes. This will hide any cases of acute ethyl mercury poisoning that occurred in postnatal and ‘outcome within 6 months’ children; then combining them with prenatal and perinatal babies further obfuscates this confounder.
10. In UK/Europe ‘smoking mothers’ are said to relate to 11.7% of all SIDS deaths yet smoking mothers have lower birth weight children. Is it the low birth weight, therefore lower weight at vaccinations, the factor in suspected DTP vaccine-mercury-induced SIDS and not smoking? If so smoking mothers confound this study.
11. Young children and babies might suffer additional exposure to mercury through breast milk if, coincident with vaccination, mother regularly ate fish contaminated with mercury or had dental treatment with mercury amalgam fillings. Each filling releases about10ug methyl mercury into mother’s blood stream daily so a mother’s dietary fish and dental visits coincident with vaccination, therefore breast-feeding, confound the study.
12. The excluded group of children, postnatal and ‘outcomes’, who might have suffered mercury damage from vaccines number about 524 - 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 neurodevelopment outcomes”, aren’t these the very children the study should have focussed on?
13. Andrews et al ‘validate’ their results through “all the neurological development disorders were more common in boys than girls” yet testosterone enhances the neurotoxicity of mercury so boys would be expected to succumb more often to mercury poisoning than girls. This actually increases the probability that Andrews et al showed mercury poisoning in their subjects.
14. Andrews et al combine for analysis and statistical purposes outcomes that suggest nephrotoxic, neurotoxic, psychotic, behavioural, emotional, cognitive deficiency, and other events, possibly even social deficiencies. This cannot make statistical sense, and may obfuscate more than enlighten.
15. The year of birth range from 1988 to 1997, and GPRD records selection to 1999, suggest that many developmental disorders like autism would not be uncovered as they may not have been diagnosed during the data collection period. Until the mid 1990s in the UK it was not uncommon for autism to remain undiagnosed until age 7 or 8 years – my own son an example – so the 103,043 cohort may hide many children with undiagnosed developmental disorders from mercury toxicity and other causes, including other vaccines eg. MMR. When a child suffers a serious reaction to a vaccine (one expects ethyl mercury to cause acute events such as seizures, meningitis, encephalitis, developmental inhibition, speech/language impairment etc.)a parent would probably refuse further consent. If the first dose injures, there will be no second; if the second does injures, there will be no third etc. After each dose there will be a reducing trend for further vaccination in injured children, but the uninjured ‘survivors’ continue to the next dose and attain a ‘less risk’ status, or 'survivor' of the previous dose, yet they also may not be free of outcomes as they were only followed up for 4.7 years – too little time for diagnosis and recording of some developmental disorders so easily hidden that might confound the study.
16. Andrews et al conclude differently to the US VSD study, which found a risk from mercury in vaccines. They say that, other than for Tics, the study does not confirm the US findings. The UK cohort had similar thimerosal exposure to 4 months of age but the US exposure increased from 4 to 7 months and Andrews et al state “if the increased risk in the US study were attributed only to the additional thimerosal exposure after 4 months, it is possible the UK study is not able to detect the risks found in the US study which had a longer follow up time…..preliminary results from the US study were probably attributed to confounding or chance”. They ignore the probability that the continuation of harm to American children was due to their increasing weight after 4 months of age being insufficient to outweigh the increasing weight of ethyl mercury per dose the defenceless children were exposed to.
17. Andrews et al try to validate GPRD records referring to “other validation exercises found the GPRD accurate” omitting the fact that those studies had relatively easy conditions to diagnose and record whereas ASD/PDDs etc. are far more difficult to process accurately and no study has successfully validated the GPRD in that regard so it remains an unknown quantity on which childrens’ lives cannot rely.
18. Andrews et al attempt validation using Danish study Hvlid et al JAMA 2003 yet when Madsen et al NEJM 2002 is compared to Hvlid et al one finds their totals for ASDs for Denmark for not dissimilar periods calculated from the same National Database that are 100% out of sync. If they cannot validate each other how can either validate Andrews et al?
We must consider that Andrews et al may have got it totally wrong and if one considers their results with respect to mercury toxicity, in terms of EPA exposure safety criteria, Andrews et al have probably shown that mercury becomes less toxic with growth/weight, therefore the increasing age of the child.
Perhaps the most telling statement Andrews et al make about their attempts at validation is “lack of specificity limits the study as it biases against finding an association”. I must agree.
Regards
John H.
Re: THIMEROSAL DOES NOT BELONG IN VACCINES
We write to correct inaccuracies and misunderstandings in the letter by Dr. Geier in response to our recent paper. First, in the fifth paragraph of his letter, Dr. Geier misunderstands what data are shown in Table 3 of our paper. It is clearly stated in the statistical methods that the main analysis included all children whether they received 0,1,2,or 3 doses at any age. The additional analysis, with the result only shown for tics, only included those receiving 3 doses by the age of one year. For all other outcomes the results were similar when restricting to those with 3 doses by the age one to those including all children. Dr. Geier has somehow missunderstood that the results shown are for the analysis that did include children receiving no doses during the first year of life. The tics result is discussed in detail in our paper.
We avoided the bias that sicker children delay vaccination by excluding such children where possible (see exclusion criteria).
His charge that our study is disingenuous because in the UK the "vaccine was scrapped over autism fear" is based on a misleading media report that bears no relationship to the truth. The recent decision in the UK to change the vaccine used for primary immunisation was driven by the need to change from oral to inactivated polio vaccine along with the availabilty of an acellular vaccine with high demonstrated efficacy. The vaccine happens to be thiomersal free which is consistent with the overall international aim of reducing the exposure of chidren to mercury from avoidable sources (www.dh.gov.uk).
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 available on our Website (www.hpa.org.uk/infections/about/dir/psp.pdf)
THIMEROSAL DOES NOT BELONG IN VACCINES
The authors of the Andrew 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 received funding to study vaccines from Wyeth Vaccines and SmithKline Beecham (1). These companies all are or were makers of thimerosal-containing vaccines.
This study seems disingenuous, since the British Government has announced the removal of thimerosal from their routine childhood vaccines effective the end of September 2004. The BBC wrote on 7 August 2004, “Vaccine scrapped over autism fear. A vaccine containing mercury given to babies when they are eight weeks old is to be scrapped amid fears of a link with autism. The move follows recent research in America that suggests a connection between mercury used to preserve whooping cough vaccine, and autism.” Health Minister John Hutton confirmed the changes stating, ‘Childhood immunization has been extremely effective in protecting children from serious-life threatening diseases, We are continually looking at ways to improve this program as new, more effective products become available.’ We congratulate the British Government for doing the right thing, and predict that the new vaccine will soon result in a concomitant drop in the currently devastating rate of neurodevelopmental disorders in England as has already potentially been seen in California which has reported a three-consecutive quarter decrease in the number of new cases of autism for the first time in approximately 20 years among children receiving childhood vaccines with reduced thimerosal content.
Unfortunately, thimerosal continues to remain in many vaccines in the US. Influenza vaccine has been added to the routine childhood immunization schedule, and the CDC has refused to state a preference for children to receive thimerosal-free influenza vaccine despite their position of encouraging the removal of thimerosal from childhood vaccines,
The current study was extremely underpowered in its ability to discern the effects of thimerosal on neurodevelopmental disorders because it only examined a maximum exposure of 75 micrograms of mercury in the first year of life and further limited potential thimerosal exposure differences by excluding children that had not received 75 micrograms of mercury from childhood vaccines by age one. The study did contain analyses for tics based upon exposure at 3, 4, and all mercury exposure, when using a reference group of children not receiving any mercury from thimerosal-containing childhood vaccines during the first year of life. The results of these analyses showed that there were statistically significantly increased hazard ratios for tics at 3 months. By comparison, there was no statistically significant correlation between mercury exposure from thimerosal-containing vaccines and tics by excluding children not receiving thimerosal-containing vaccines during the first year of life. The authors provide no other complete data for any other outcomes when employing children receiving no mercury during the first year of life as the reference population. The tic result in the Andrews et al. study is similar to a previous study, from the Vaccine Safety Datalink (VSD) database (2), indicating an apparent causal relationship between mercury containing childhood vaccines and the development of tics.
This study has virtually no applicability to the US experience with thimerosal. Despite incorrect statements to the contrary by Andrews et al., by 4 months of age US children received approximately 2-fold higher doses of mercury from vaccines (125 micrograms) compared to those England (75 micrograms). This study contains significant biases because sicker children were the ones that tended to have vaccinations delayed resulting in an apparent preventive effective for mercury on the risks of neurodevelopmental disorders.
It has become apparent from recently emerging clinical, animal model, and molecular evidence that thimerosal is indeed responsible for neurodevelopmental disorders in a substantial number of children, regardless of the findings of large population-based epidemiological studies. Independent investigators have shown children with autistic spectrum disorders have significantly higher body-burdens of mercury than those of neurotypical children (3-5), a genetically susceptible mouse strain develops autistic features, including: growth delay, reduced locomotion, exaggerated response to novelty, increased brain size, decreased numbers of Purkinje cells, significant abnormalities in brain architecture, affecting areas sub-serving emotion and cognition, and densely packed hyperchromic hippocampal neurons with altered glutamate receptors and transporters following administration of thimerosal mimicking the US childhood immunization schedule (6), and molecular studies in vitro have demonstrated that acute thimerosal exposure at extremely low concentrations (i.e. at parts-per-million or lower) (7-9), that are comparable to the expected body distribution of mercury resulting from thimerosal-containing vaccines that were administered in the US, can kill or significantly adversely effect neuronal growth and development. Pharmacokinetic studies on infant primates exposed to solutions containing similar concentrations of thimerosal, as thimerosal-containing vaccine childhood vaccines, have shown that the half-life of mercury in the brain of the infant primates was approximately 28 days (10). Male mice were at considerably more sensitive than females to the neurotoxic effects of low dose alkyl mercury exposure (11). These results were consistent with some human fetal/infant population exposures to low doses of alkyl mercury where it has been observed that males were more sensitive than females to psychomotor retardation. Autistic spectrum disorders, of course, are significantly more prevalent in males than females (12).
In conclusion, we are, and always have been, strong supporters of the US vaccine program and of pediatricians that administer vaccines, but given the fact that many US states now have either banned (Iowa), or are in the process of banning thimerosal, (California, Missouri, Nebraska, and New York, among many others) and given that fact that there is now a bipartisan national bill introduced in the US House of Representatives (Weldon/Maloney bill) to ban it nationally, we now strongly suggest that the United States pediatricians should insist on giving only thimerosal- free vaccines, lest they become involved in the terrible morass of lawsuits that are already beginning on this issue.
References
1. Miller E, Waight P, Farrington CP, Andrews N, Stowe J, Taylor B. Idiopathic thrombocytopenic purpura and MMR vaccine. Arch Dis Child 2001;84:227-9.
2. Verstraeten T, Davis RL, DeStefano F, Lieu TA, Rhodes PH, Black SB, Shinefield H, Chen RT; Vaccine Safety Datalink Team. Safety of thimerosal-containing vaccines: a two-phased study of computerized health maintenance organization databases. Pediatrics 2003;112:1039-48.
3. Bradstreet J, Geier DA, Kartzinel JJ, Adams JB, Geier MR. A case- control study of mercury body-burden in children with autistic spectrum disorders. J Am Phys Surg 2003;8:76-9.
4. Holmes AS, Blaxill MF, Haley BE. Reduced levels of mercury in first baby haircuts of autistic children. Int J Toxicol 2003;22:277-85.
5. Hu LW, Bernard JA, Che J. Neutron activation analysis of hair samples for the identification of autism. Trans Am Nucl Soc 2003;89.
6. Hornig M, Chian D, Lipkin WI. Neurotoxic effects of postnatal thimerosal are mouse strain dependent. Mol Psychiatry 2004;9:833-45.
7. Baskin DS, Ngo H, Didenko VV. Thimerosal induces DNA breaks, caspase-3 activation, membrane damage, and cell death in cultured human neurons and fibroblasts. Toxicol Sci 2003;74:361-8.
8. Waly M, Olteanu H, Banerjee R, Choi SW, Mason JB, Parker BS et al. Activation of methionine synthase by insulin-like growth factor-1 and dopamine: a target for neurodevelopmental toxins and thimerosal. Mol Psychiatry 2004;9:358-70.
9. Brunner M, Albertini S, Wurgler FE. Effects of 10 known or suspected spindle poisons in the in vitro porcine brain tubulin assembly assay. Mutagenesis 1991;6:65-70.
10. Institute of Medicine (US). Immunization Safety Review: Vaccines and Autism. Washington, DC: National Academy Press, 2004.
11. Clarkson TW, Nordberg FJ, Sager PR. Reproductive and developmental toxicity of metals. Scand J Work Environ Health 1985;11:145- 54.
12. Bertrand J, Mars A, Boyle C, Bove F, Yeargin-Allsopp M, Decoufle P. Prevalence of autism in a United States population: the Brick Township, New Jersey, investigation. Pediatrics 2001;108:1155-61.
Dr. Mark R. Geier has been a consultant and expert witness in cases involving vaccines before the National Vaccine Injury Compensation Program and in civil litigation. David A. Geier has been a consultant in cases involving vaccines before the National Vaccine Injury Compensation Program and in civil litigation.