To: David Satcher, M.D., Ph.D. July 5, 2000


Assistant Secretary for Health and Surgeon General

Director of Public Health Service

U.S. Department of Health and Human Services

200 Independence Avenue, SW; Washington, DC 20201

Cc: ACIP - The Advisory Committee on Immunization Practices

AAP - The American Academy of Pediatrics

AAFP - The American Academy of Family Physicians

NIP - National Immunization Program

And - Selected CDC personnel and other individuals.

Fr: Teresa Binstock

Researcher in Developmental & Behavioral Neuroanatomy

Box 1788; Estes Park CO 80517 <>

Re: Scientific errors in Joint Statement by PHS et al

 On June 22, 2000, your organization approved a "JOINT STATEMENT CONCERNING REMOVAL OF THIMEROSAL FROM VACCINES" (2).

As a researcher and concerned citizen, and as a co-author of a research paper linking thimerosal with neurodevelopmental problems in children (1), I believe that the Joint Statement offers conclusions and recommendations based upon scientific errors in the methodology used by several CDC study groups that recently explored HMO databases for evidence of neurotoxicity induced by physician-injected ethylmercury (3).

The scientific errors in Davis et al’s two studies for the CDC, ACIP, et al., derive (i) from ignoring susceptibility factors and their influence upon neurotoxic risks at low levels of exposure, and thereby (ii) from interpreting HMO data in accord with an artificially high and unsafe level for ingested methylmercury (EPA, 1997).

Three addenda included herewith delineate (i) neurotoxicity susceptibility from low levels of exposure, (ii) the methodological errors in Davis et al; and (iii) derivative errors in the PHS-ACIP-AAP-AAFP Joint Statement of June 22.

I realize that these several organizations share my concern for vaccinated children and, therefore, request:

  1. A public retraction of June 22 Joint Statement;
  2. The recent thimerosal-studies be redone by researchers not affiliated with your four organizations;
  3. You ensure that no journal affiliated with the Joint Statement's authoring corporations shall publish articles about Davis et al’s ACIP/CDC studies until those authors’ errors of scientific methodology have been corrected and the HMO data and conclusions therefrom reevaluated; and
  4. The CDC distribute this letter via its Immunization Action Coalition "IAC Express" newsletter so as to correct the misinformation distributed in IAC issue number 171 of June 29, 2000 (5).

The issues and requests presented in this letter and its addenda have a certain urgency. Millions of children in this country have been injected with vaccinal ethylmercury. The numbers of susceptible children likely to have developed neurologic impairments from thimerosal are significant (4). Chelation therapies ought be initiated as soon as possible, because years of delay will exacerbate neurotoxicity in these children. The CDC and its affiliated corporations can and should act to correct the scientific errors in Davis et al and the thimerosal misinformation in IAC Express v171.






Teresa Binstock

Researcher in Developmental & Behavioral Neuroanatomy




  1. Sallie Bernard, Albert Enayati, Heidi Roger, Lyn Redwood, Teresa Binstock. Autism: A Unique Type of Mercury Poisoning. Draft released for review and comment on April 3, 2000; Cure Autism Now Foundation website.
  2. The error-filled Joint Statement of PHS, ACIP, AAP, and AAFP is presented herewith as annotated addendum-3.
  3. Bernard et al (1) have received two study-summaries by Richard L. Davis et al. Each of these summaries contains major errors in the science of neurotoxic susceptibility-risk at low-level exposures. Critiques of Davis et al’s two study-summaries are included herewith as addendum-2a and addendum-2b.
  4. Dale Hattis, PhD, of Clarke University has published a number of papers about the role of susceptibility factors in assessing neurotoxicity risks in accord with low levels of exposure. Several of his abstracts are included herewith as addendum-1.

5. CDC newsletter, IAC express, #171, June 29, 2000


Four addenda are included with this letter:

  1. Neurotoxic Susceptibility: Discussion and Hattis abstracts.
  2. Errors in two Davis et al studies (as addenda 2a, 2b).
  3. Annotated critique of the CDC/ACIP et al Joint Statement.
  4. Future research: some considerations.

Addendum 1:

Susceptibility: Increased Risk from Low-Level Exposure

Articles by Dale Hattis, PhD, of Clarke University shall become central to evaluating thimerosal's neurotoxic effects in susceptible infants and toddlers injected with ethylmercury (eHg). Dr. Hattis’ peer-reviewed articles, published in respected scientific journals, make clear that low-level exposures can be neurotoxic to susceptible individuals.

  1. Extensive research-findings document that Hg-susceptibility factors have been identified and relate to familial autoimmunity and detoxification status (eg, CNS glutathione) (reviewed in 6). To the best of our belief and knowledge, Davis RL et al did not utilize susceptibility in their interpretations of HMO data.
  2. A second and crucial factor is that of a "safe" level. For instance, Davis et al (i) utilized the EPA's theoretical "safe" level for ingested methylmercury (mHg) (mHg-contaminated grain eaten by pregnant women in Iraq) and (ii) did not alter the EPA level so as to account for (a) bolus doses injected during vaccinations and present during vaccination-induced pulses of cytokines, and (b) subgroups of susceptible infants and toddlers within the exposed population (ie, virtually all U.S. children receiving mandatory vaccinations).
  3. Hattis' abstracts and articles delineate the importance of including accurate susceptibility evaluations when determing neurotoxic risk and interpreting data from exposed populations (1-5).
  4. In their CDC/ACIP studies, Davis et al used an erroneously high "safe" level for ethylmercury, by invoking the EPA's estimage of a "safe" level for methylmercury, based upon contaminated grain ingested by pregnant women in Iraq). As a result of dependence upon the EPA’s erroneously high "safe" level, the ACIP/CDC model for interpreting data is not realistic and subserves erroneous conclusions that misrepresent ethylmercury’s actual risks and neurologic sequelae.
  5. As set forth in the Hattis reviews (eg, 1), the closer that a model for interpreting data is to real variables (eg, susceptibility factors in specific individuals or groups of such individuals), then the better will be the ability of researchers to determine a toxic substance's neurologic effects.
  6. Consider a phrase in Hattis’ cite-5: "distributions of human susceptibility". That concept seems to be totally lacking in the Davis et al studies whose summaries I’ve perused in June of 2000; yet these wrongfully designed studies are the primary basis for the erroneous statements in the Joint Statement by the ACIP and its affiliated authoring-organizations.


Two conclusions:

  1. When susceptibility is considered and properly utilized in redoing the studies of Davis et al, then a small percentage of children will be found to have suffered neurologic damage from the levels of ethylmercury in infant and toddler vaccines.
  2. In fact, if susceptibility factors and low-level neurotoxicities are integrated into re-evaluations of the HMO data, then the EPA's "safe" level will be realized to be artificially high for evaluating the neurotoxicity of vaccinal ethylmercury; and the HMO data will reveal thimerosal's neurotoxic effects – as one CDC/ACIP study group has already documented (see Joint Statement).


Teresa Binstock

Researcher in Developmental & Behavioral Neuroanatomy

Box 1788; Estes Park CO 80517



1. Hattis D. The challenge of mechanism-based modeling in risk assessment for neurobehavioral end points. Environ Health Perspect 1996 104 Suppl 2:381-90.

Ab: The mathematical form for a dose-time-response model is ideally not just a convenience for summarizing or fitting a particular data set--it represents a hypothesis. The more this hypothesis reflects a mechanistically sophisticated view of the likely reality, the more it can lead to potentially informative validating or invalidating types of predictions about the results of real experiments and (in the long run) reasonably credible predictions outside the range of direct observations. This paper first reviews some distinctive features of the nervous system and neurotoxic responses and theoretically explores some basic quantitative implications of these features…

2. Hattis D, Glowa J, Tilson H, Ulbrich B. Risk assessment for neurobehavioral toxicity: SGOMSEC joint report. Environ Health Perspect 1996 104 Suppl 2:217-26.

Ab: Behavioral end points for neurotoxicity risk assessment have been developed and examined over the past three decades. They are now ready to move from simple qualitative guidelines, such as exemplified by reference doses, to more quantitative models, such as benchmark doses, based on dose-response information. Risk assessors, confronted by a wider array of methodologies and data than in the past, should be offered guidance in interpretation because now they have to deal with unaccustomed questions and problems. These include reversibility, susceptible populations, multiple end points, and the details of dose-response and dose-effect distributions.

3. Hattis D, Erdreich L, Ballew M. Human variability in susceptibility to toxic chemicals--a preliminary

analysis of pharmacokinetic data from normal volunteers. Risk Anal 1987 Dec;7(4):415-26

ab: The tenfold "uncertainty" factor traditionally used to guard against human interindividual differences in susceptibility to toxicity is not based on human observations. To begin to build a basis for quantifying an important component of overall variability in susceptibility to toxicity, a data base has been constructed of individual measurements of key pharmacokinetic parameters for specific substances (mostly drugs) in groups of at least five healthy adults. 72 of the 101 data sets studied were positively skewed, indicating that the distributions are generally closer to expectations for log-normal distributions

than for normal distributions. Measurements of interindividual variability in elimination half-lives, maximal blood concentrations, and AUC (area under the curve of blood concentration by time) have median values of log10 geometric standard deviations in the range of 0.11-0.145. For the median chemical, therefore, a tenfold difference in these pharmacokinetic parameters would correspond to 7-9 standard deviations in populations of normal healthy adults. For one relatively lipophilic chemical, however, interindividual variability in maximal blood concentration and AUC was 0.4--implying that a tenfold difference would correspond to only about 2.5 standard deviations for those parameters

in the human population. The parameters studied to date are only components of overall susceptibility to toxic agents, and do not include contributions from variability in exposure- and response-determining parameters. The current study also implicitly excludes most human interindividual variability from age and illness. When these other sources of variability are included in an overall analysis of variability in susceptibility, it is likely that a tenfold difference will correspond to fewer standard deviations in the overall population, and correspondingly greater numbers of people at risk of toxicity.

4. Hattis D, Banati P, Goble R, Burmaster DE. Human interindividual variability in parameters related to health risks. Risk Anal 1999 19(4):711-26.

Ab: This paper reviews existing data on the variability in parameters relevant for health risk analyses. We cover both exposure-related parameters and parameters related to individual susceptibility to toxicity. The toxicity/susceptibility data base under construction is part of a longer term research effort to lay the groundwork for quantitative distributional analyses of non-cancer toxic risks. These data are broken down into a variety of parameter types that encompass different portions of the pathway from external exposure to the production of biological responses. The discrete steps in this pathway, as we now conceive them, are [with numbers added for clarity]:

1. -- Contact Rate (Breathing rates per body weight; fish consumption per body weight)

2. -- Uptake or Absorption as a Fraction of Intake or Contact Rate

3. -- General Systemic Availability Net of First Pass Elimination and Dilution via

4. -- Distribution Volume (e.g., initial blood concentration per mg/kg of uptake)

5. -- Systemic Elimination (half life or clearance)

6. -- Active Site Concentration per Systemic Blood or Plasma Concentration

7. -- Physiological Parameter Change per Active Site Concentration (expressed as the dose required to make a given percentage change in different people, or the dose required to achieve some proportion of an individual's maximum response to the drug or toxicant)

8. -- Functional Reserve Capacity-Change in Baseline Physiological Parameter Needed to Produce a Biological Response or Pass a Criterion of Abnormal Function.

Comparison of the amounts of variability observed for the different parameter types suggests that appreciable variability is associated with the final step in the process-differences among people in

"functional reserve capacity."...

5. Hattis D, Banati P, Goble R. Distributions of individual susceptibility among humans for toxic effects. How much protection does the traditional tenfold factor provide for what fraction of which kinds of chemicals and effects? Ann N Y Acad Sci 1999;895:286-316.

Ab: A significant data base has been assembled on human variability in parameters representing a series of steps in the pathway from external exposure to the production of biological responses: contact rate (e.g., breathing rates/body weight, fish consumption/body weight); uptake or absorption (mg/kg)/intake or contact rate; general systemic availability net of first pass elimination

[eg, low glutathione] and dilution; systemic elimination or half-life; active site availability/general systemic availability; physiological parameter change/active site availability; functional reserve capacity--change in baseline physiological parameter needed to pass a criterion of abnormal function or exhibit a response. This paper discusses the current results of analyzing these data to derive estimates

for distributions of human susceptibility to different routes of exposure and types of adverse effects. The degree of protection is tentatively evaluated by projecting the incidences of effects that would be expected for a tenfold lowering of exposure from a 5% incidence level if the population distribution of

susceptibility were truly log-normal out to the extreme tails, and if the populations, chemicals, and responses that gave rise to the underlying data were representative of the cases to which traditional uncertainty factor is applied. The results indicate that, acting by itself, a tenfold reduction in dose

from a 5% effect level is associated with effect incidences ranging from slightly less than one in ten thousand, for a median chemical/response, to a few per thousand, for chemicals and responses that have greater human interindividual variability than 19 out of 20 typical chemicals/responses. In practice, for many of the cases where the traditional tenfold factor is applied, additional protection is provided by other uncertainty factors. Nevertheless, the results generate some reason for concern that current application of traditional safety or uncertainty factor approaches may allow appreciable incidences of responses in some cases.

6. Sallie Bernard, Albert Enayati, Heidi Roger, Lyn Redwood, Teresa Binstock. Autism: A Unique Type of Mercury Poisoning. Draft released for review and comment on April 3, 2000; Cure Autism Now Foundation website.


Addendum 2a:


Critiques of Davis et al’s two error-filled thimerosal-studies

Addenda 2a and 2b present two critiques:

  1. of the abstract presented by Davis et al for the AAP conference in mid-May 2000 in Boston;
  2. of a summary sheet given Bernard et al during our meeting with ACIP officials in CDC offices in Washington DC on June 15, 2000.



Critique of Davis et al Abstract

During the mid-May AAP-etc conference in Boston, Robert L Davis et al presented an abstract and lecture entitled "Infant exposure to thimerosal-containing vaccines and risk for subsequent neurologic and renal disease" (1). Their study utilized data from four HMOs and their 400,000 children in appropriate ages. Inclusion and exclusion criteria were not presented in the abstract. The authors conclude "Among children followed up to 6 years of age, ethylmercury exposure at one and three months of age is not associated with adverse neurodevelopmental or renal outcomes."


Conclusion: The Davis et al abstract appears has a number of flaws. If these flaws in Davis et al’s two studies are corrected, then their data demonstrate that, in fact, ASD and other neurologic syndromes (eg, ADHD, anorexia) are likely to have been caused by physician-injected ethylmercury.


List of observations:

  1. Statement 2 in Design/methods: The EPA’s ‘safe’ level is dangerously high for susceptible individuals. When dosage-sensitivity and susceptibility factors (genetic and non-genetic) are considered, an ethylmercury level much lower than the EPA’s "safe" level ought to have been used (2).
  2. The data-set used in the abstract (VSD; 4 HMOs; 400,000 children) appears to have been the same data-pool used in creating the legal-sized summary given us (3) on June 15th, by Martin Myers, MD, Acting Director of National Vaccine Program Office.
  3. That the Davis et al data-pool contained only "153 children with autism" contradicts large data-pools from California and other states, whose reports describe much higher numbers and rates of ASD children (4).
  4. That the Davis et al data-pool contained only "346 with ADHD" is strongly divergent from and lower than the findings of Kelleher KJ et al, who found the ADHD rate to have grown from 1.4% in 1979 to at least 7.6% in 1996, and Kelleher et al controlled for (and thus largely eliminated) reporting-bias (5). In other words, the Kelleher et al findings appear to be actual increases in the rate of ADHD in children.
  5. I have no solid comparison-figures for Developmental Delay rates, although those figures may be implicit in Kelleher et al’s Table 1.
  6. The disparities between Davis et al’s HMO-tallies and the Kelleher et al data strongly suggest that a selection bias is occurring in HMO reporting systems.
  7. Davis et al interpret their findings their rate-finding data as showing "no increased risk" for neurodevelopmental diagnoses, concluding "Among children followed up to 6 years of age, ethylmercury exposure at one and three months of age is not associated with adverse neurodevelopmental or renal outcomes."



  1. The conclusion in Davis et al is necessarily dependent upon the EPA’s "safe" limit for methylmercury exposure. Whereas I am willing to grant the approximate similarity in neurotoxicity between ethyl- and methyl-mercury (Suzuki et al, 1973, book chapter), for reasons listed herein and elsewhere, the EPA’s "safe" limit is artifically high, much too high. Davis et al’s main statistical-strategy appears to have derived from a comparion of ASD and ADHD and other diagnoses for children (A) whose exposure exceeded the EPA figure for ingested maternal mercury, and (B) whose exposure remained below the EPA figure. Because group-A and group-B had similar numbers of ADHD and ASD children, Davis et al concluded that a thimerosal effect was not seen. However, if – for reasons of dose-response sensitivity and other susceptibility factors – the EPA’s "safe" level ought to have been substantially lower, then the numbers of ADHD and ASD in the exceeded-safe-limits group would be much larger than the numbers in the below-safe-limits group. In other words, had a corrected model been used to interpret the actual data, then Davis et al would have had to conclude that Yes, in fact, thimerosal appears to be a major cause of ASD, ADHD, and related neurodevelopmental disorders.
  2. Davis et al findings state "no increase risks" from ethylmercury (eHg) at 1 and 3 months and – aside from the aforementioned selection bias – are necessarily based upon the EPA’s safe limit for mHg. However, EPA calculation (a) did not include bolus doses (known to have more effect, especially in individuals not previously exposed), and (b) was based upon levels needed to induce neurologic damage in 10% of a normal population, and (c) was not based upon the level needed to induce neurologic damage in the much smaller groups (eg, wherein from .20% to 1% to 7.6% are affected, thus paralleling ASD and ADHD rates) that are susceptible to Hg-induced neurotoxicity.
  3. The organic-mercury level needed to induce neurologic damage in .20% to .25% would be much lower for susceptible infants and toddlers than is the EPA’s level for inducing damage in 10% of normal children prenatally exposed by maternal ingestion.
  4. Davis et al’s rates of ADHD and ASD appear to be artificially low. This appears to represent a selection bias arising from how HMO physicians record diagnostic categories. Several parents affirm that negative treatment-consequences can ensue if an HMO physician enters a diagnostic category such as ASD.
  5. Data and analysis in the Davis et al abstract do not fully match the data and analyses in a second summary (see addendum 2b) that utilized the same initial data-pool.



  1. Davis RL et al. Infant exposure to thimerosal-containing vaccines and risk for subsequent neurologic and renal disease. May 2000. (Robert L Davis, Thomas Verstraeten, David Gu, Frank DeStefano, Robert S. Thompson, Robert T. Chen; Vaccine Safety Datalink Team Immunization Studies program, Group Health Cooperative, Seattle WA; National Immunization Program, Centers for Disease Control, Atlanta GA).
  2. Binstock T. Presentation to FDA Vaccine Safety Committee. May 11, 2000. Bethesda MD.
  3. Attendees representing mercury/autism: in DC: Sallie Bernard, Teresa Binstock, Albert Enayati, Heidi Roger; in Atlanta: Lyn Redwood, Jane El-Dahr, Susan Owens.
  4. Yazbak FE. Autism `99, a national emergency. Internet publication 1999.
  5. Kelly J. Kelleher, Thomas K. McInerny, William P. Gardner, George E. Childs, Richard C. Wasserman. Increasing identification of psychosocial problems: 1979-1996. Pediatrics 2000;105:1313-21.



Addendum 2b:

Critique of Davis et al document with Tables and Graphs

During a June 15th meeting in DC and Atlanta, Martin Myers, MD, acting director of the National Vaccine Program Office distributed (in DC) a legal-sized version (1) of what appears to describe the same Davis RL et al study described on the AAP webpage (2). I write "appears" as same because, for instance, the long version denotes autism n=67, whereas the abstract version has autism n=153. This difference may arise because the longer version appears to reflect a winnowing down, ie, a deselection as indicated in the graph entitled "Creation of Study Cohort".

Aside from these differences, the long-version’s conclusion is virtually identical to that of the short version and states: "Exceeding the EPA standard [for chronic, low-level methylmercury exposure] was not associated with a significantly increased risk of developmental neurologic and renal disorders [from repeated bolus doses of ethylmercury in susceptible individuals]".


List of observations:

  1. As with the Davis et al abstract, the primary flaw in the long version is reliance upon the EPA’s "safe" level, as determined for chronic exposure to methylmercury at a level needed to induce neurologic damage to 10% of normal children. This level is artificially high and needs to be lowered. When a lower level is derived from susceptibility factors related to a .25% or 1% of all children, then the EPA’s safe level would be much lower, and the primary conclusions of Davis RL et al would be different. In fact, when a corrected model is used to interpret the Davis et al findings, then the Davis et al data demonstrate an important thimerosal effect in causing neurodevelopmental disorders in autism and ADHD.
  2. Davis et al section entitled "Assumptions in applying EPA limit".
  3. A. In establishing a model for interpreting the collected data, I accept the principle (as a general guideline) that "Ethylmercury [is approximately] as toxic as methylmercury (Suzuki et al book chapter, 1973).

    B. Flaw: "Single exposure = continuous exposure X exposure time". Comment: Instead, bolus doses are more injurious than chronic low-dose exposures leading to the same total level, especially in individuals not previously adapted to mercury exposure.

    C. Davis et al invoke "No excretion of ethylmercury". Comment: this seems fair enough for the model, because eHg would tend to migrate into the infant and toddler brain and relatively little would be excreted.

    D. Flaw: "Injection exposure = oral exposure". Oral exposure, especially at low doses when mixed with food, leads to simple elimination of a goodly portion of the mercury and also leads to some elimination by hepatic and renal routes. Much less organic Hg enters the non-mesenteric bloodstream from ingestion (Klaassen text).

    E. Flaw: "Toxicity in infants as great as in fetus". Comment: What is extremely different between between fetal and infant/toddler periods is the nature of the synaptogenesis. During the infant/toddler periods, synaptogenesis is occurring for exactly the same traits as are central to the diagnostic criteria of autism and also for synaptic refinements of sensory and motor processing. eHg that enters the brain of infants and toddlers can disrupt these crucial synaptogenic processes and does so by various mechanisms already identified for mHg (eg, microtubule disruption, astrocyte alteration, etc). Furthermore, as implicity presumed in the EPA calculations of its "safe" level (via analysis of a specific, large-scale, oral-ingestion incident), the maternal liver would remove much of the orally ingested mHg; however, in regard to thimerosal injections, a lowering of eHg by the maternal liver would not occur.

  4. Section entitled: "Site: The Vaccine Safety Datalink (VSD)." Comment: the database is very large, began in 1991, and if the model used in data-interpretation is corrected in regard to a lower safe-exposure level (via dose-response and susceptibility factors), the data demonstrate that, indeed, thimerosal is causing ASD and other neurodevelopmental disorders.
  5. Section entitled: "Toxicology of organic mercury". Comment: Authors need consider not only "dose" but also effects of bolus dose, which in non-acclimated infant or toddler will have more impact than a the same dose acquired from daily exposure.
  6. Section entitled: "Statistical analyses". Comment: Davis et al’s one page (legal sized) summary presents conclusions entwined with statistical calculations. Davis et al’s core data are not provided, which prevents re-analysis in accord with a more accurate model based upon a lower EPA limit (3). Using a lower eHg level for interpreting the data is justified by bolus doses, injection, injection in the presence of a vaccination-response, and by genetic and non-genetic susceptibility factors.
  7. Section entitled: "Outcome definitions". Comment: This section’s data does not match the data presented in the Davis RL et al abstract presented to the AAP in mid-May. I assume (guess) that the "Outcome definitions" section shows lower ASD totals than did the Abstract because during the "Creation of Study Cohort", some individuals were eliminated. Some of these eliminations may reflect an additional type of selection bias because some of the eliminated categories of children may represent mothers, infants, or toddlers with increased susceptibility to eHg-neurotoxicity toxicity due to factors that induce lower levels of neuroprotective glutathione (Aukrust P et al, 1995; autism optimality/suboptimality studies).
  8. Section entitled: "Limitations": Analysis of the Davis et al Abstract delineates probable selection-bias due to reporting procedures by HMO physicians. A major limitation in the two Davis et al presentations is that neither addresses the small percentage of children who have increased sensitivity to neurotoxicity from organic mercury. This was clearly demonstrated in acrodynia, when the affected population was approximately 1 in 500, noting that acrodynia (Pink Disease) arose from topical use of mercury-based ointments and not from direct injections whereby more mercury (eg, thimerosal’s eHg) would enter the CNS of infants and toddlers.
  9. The Davis et al study does not address the issue of fetal exposure to thimerosal via RhoGam (or flu shots) given to pregnant women (5).


Conclusion: Davis et al have performed an important service in compiling data from nearly 400,000 children. However, (a) there seem to be several kinds of selection bias implicit in the Davis et al data, (b) and their model for interpreting the data has serious flaws. The combined sum of these two categories of error is that the conclusion offered by Davis et al is erroneous. They collected data, interpreted it with a fatally flawed model, and concluded that no thimerosal-effect was apparent. However, if that same data is interpreted in accord with a substantially improved model (3), then the Davis et al data actually demonstrate that vaccinal ethylmercury is causing neurologic impairment in children. Furthermore, due to several types of selection bias in the HMO data, a more accurate determination of the number of thimerosal-damaged children can found by perusing epidemiology rates for ASD and ADHD (4) in other studies. Sadly, the numbers of affected children are huge, and parents and families are paying a severe price for the medical community’s unanticipated effects of physician-injected ethylmercury.



  1. Risk of neurologic and renal impairment associated with thimerosal-containing vaccines. Davis RL (U of W), Verstraeten T, DeStefano F, Gu D, Pless R, Chen R (CDC); Black S, Shinefield H (Kaiser Perm of N. Calif); Robert Wise, Leslie Ball, M. Braun (FDA). (given us on June 15, 2000, at CDC offices in Washington DC).
  2. Pediatric Academic Societies & American Academy of Pediatrics, Joint Conference: May 12-16, 2000, Boston.;
  3. Gilbert et al. Neurobehavioral Effects of Developmental Methymercury Exposure. Environ Health Perspect 103;s6:135-42 1995. Here is the rationale I presented to FDA Vaccine Safety Committee, May 11th, 2000 in Bethesda:


Present EPA "safe" limit for Hg ought be lowered. Doses not thought to be associated with adverse effects have induced damage in humans (19); infants are especially vulnerable to Hg (89); and the EPA's recently published "safe" guideline is too high (47). Two points are noteworthy: First: Because vaccinations induce immune reactions that include extended cytokines pulses (eg, interferon gamma; 49), vaccinal mercury is more dangerous than injected-mercury studies suggest because interferon gamma increases permeability of tissues such as the blood-brain barrier and gastrointestinal tract (50-51). Thus, when a bolus dose of ethylmercury circulates during a vaccination response, more eHg is likely to enter the CNS. Second: The EPA's determination was based upon the amount of ingested mercury needed to induce adverse neurologic sequelae in 10% of exposed fetuses. But vaccinal eHg is not first filtered by the maternal liver or placenta, as was the HgP incident used in the EPA calculations.

Furthermore, a 10% rate of neurologic sequelae is clearly not acceptable. To induce a 1% rate of neurologic sequelae, the necessary Hg level would be even lower than the EPA's current estimate; and to achieve a .25% rate of neurologic sequelae -- which approximates the rate of autism during the 1990s -- an even lower level of organic mercury would be sufficient. For these reasons and because of Hg's dose-response curve, the EPA's current guideline for mercury toxicity is artificially high and ought be lowered; and, given the distribution of organic Hg’s toxic effects, the amount of injected eHg is likely to have caused neurologic damage in a significant number of children.

4. Kelleher KJ, McInerny TK, Gardner WP, Childs GE, Wasserman RC. Increasing identification of psychosocial problems: 1979-1996. Pediatrics 2000;105:1313-21.

5. Luka RE, Oppenheimer JJ, Miller N, Rossi J et al. Delayed hypersensitivity to thimerosal in RhO(D) immunoglobulin. J Allergy Clin Immunol 1997 100.138-9.



95% in ATSDR: Citing Clarkson, the ATSDR’s blue book (Toxicological Profile for Mercury, Revised Edition) states that 95% of ingested mercury enters the bloodstream. In contrast, Rozman and Klaassen (6) describe an enterohepatic cycle whereby intestinal mercury compounds do not necessarily enter the bloodstream: "The liver is in a very advantegeous position for removing toxic agents from blood after absorption from the GI tract, because blood from the GI tract passes through the liver before reaching the general circulation. (p108). Furthermore, Rozman and Klaassen describe that mercury xenobiotics can enter an enterohepatic feedback loop that is so well documented that it "has been utilized the the treatment of dimethylmercury poisoning…" (p108).


Ramification: We cannot necessarily conclude that 95% of the mercury in the EPA’s model reaches the bloodstream. Injected eHg bypasses initial enterohepatic filtering.

6. Rozman KK, Klaassen CD. Absorption, distribution, and excretion of toxicants. In: Casaret & Doull’s Toxicology: The Basic Science of Poisons. 5th edition, Curtis D. Klaassen, editor; McGraw-Hill, 1996.


C:\nj2\Critique Davis long.doc


Addendum 3:


Annotated Joint Statement


June 22, 2000

Prepared by:


Teresa Binstock

Researcher in Developmental & Behavioral Neuroanatomy

The full Joint Statement by the CDC/ACIP et al follows a critique linked to reference numbers [T1] through [T9] placed within the full text.

On June 22, 2000, the CDC’s Advisory Committee on Immunization Practices (ACIP) and three affiliated corporations with vested interest in thimerosal and physician-induced injuries therefrom released a Joint Statement which is a CDC/ACIP et al position paper and summary about thimerosal and neurologic injuries. It is my sincere belief that conclusions and recommendations within the Joint Statement reflect scientific errors derived from errors of methodology in at least two Davis et al studies sponsored by the CDC/ACIP et al.

Comments T1 through T9 refer to sentences, paragraphs, or ideas in the PHS et al Joint Statement; are presented first hereinbelow; and are indicated in the Joint Statement by reference-like enumeration. As will be apparent, several ideas occur again and again in my comments and identify the Joint Statement’s inclusion of Davis et al’s erroneous methodology.

Toxicity principles that ought to have been included in the ACIP et al research design (but were erroneously excluded) are easily understood:

  1. The EPA’s recommendation for a "safe" level of mercury (1997) is injuriously high when susceptible individuals are considered.
  2. To the best of my belief and knowledge, the Davis et al analyses did not include consideration of susceptible infants and children – despite the fact that the need for susceptibility considerations is well established in neurotoxicity studies (eg, peer-reviewed articles by Dale Hattis, PhD).

As will be clear in the sentences or paragraphs denoted by my comments [T1] through [T9], virtually all of the CDC/ACIP et al’s Joint-Statement conclusions derive (a) from the Davis group’s erroneous reliance upon the EPA’s dangerously high "safe" level and (b) from avoiding consideration of individuals with increased susceptibility to low-level exposure via bolus doses during vaccinations, which alter blood-brain-barrier permeability.

However, despite these shortcomings, better science can prevail. Subgroups susceptible to ethylmercury toxicity can be acknowledged; this will lead to a lower "safe" level for ethylmercury.

When these corrections are utilized, the data available to the ACIP and its affiliates will reveal that, indeed, tens of thousands of young Americans with ADHD, Tourette’s, anorexia, PDD/NOS, and autism are that way because as infants and toddlers, they were injected with thimerosal’s ethylmercury during mandatory vaccinations.

The analysis presented here is accurate to the best of my belief and knowledge. The following comments refer to sentences or paragraphs in the Joint Statement of the ACIP et al, which follows these 9 comments:

T1: In the Joint Statement text, this sentence is totally dependent upon EPA "safe" level or on a similarly high level and also upon avoidance of susceptibility factors whereby a small percentage of individuals will be injured at lower doses.

T2: This paragraph sounds good, but we must keep in mind that damaged kids need chelation, not bad science from the CDC re: susceptibility and "safe" levels. Delaying chelation for susceptible children who have been injected with ethylmercury during vaccinations may preclude treatments that would otherwise have been effective.

T3: A level induced by bolus dose has more impact than does a gradually acquired background level, especially during critical periods. Furthermore, the evaluative technique of exploring infant blood levels subsequent to vaccination is not adjusted for ethylmercury or its derivative Hg++ that has already migrated into the infant’s brain nor that has migrated there from prior exposures (eg, RhoGam or maternal flu shots given during her pregnancy, or any prior vaccinations with vaccines containing ethylmercury). Importantly, mere "blood level" is an inadequate guide unless susceptibility factors in individual infants and in susceptible groups of infants are considered.

Published studies by Dale Hattis, PhD, make clear the need to include susceptibility concerns when evaluating neurotoxicity risks from mercury compounds. To the best of my belief and knowledge, this aspect of thimerosal-data analysis has been deliberately or accidentally avoided by the CDC and its chosen "experts". By avoiding scientifically justified use of susceptibility factors, the CDC/ACIP et al have been able – in their official "Joint Statement" to present a scientifically erroneous conclusion of "no thimerosal effect".

However, if and as the HMO data is analyzed in accord with established principles for neurotoxicity evaluation, then the baseline "safe" mercury levels will be lowered so as to include groups of individuals with susceptibility to mercury toxicity.

A major ramification of including susceptibility factors is that the HMOs’ data then demonstrate that vaccinal thimerosal’s ethylmercury has been causing widespread neurologic damage in infants, toddlers, and children.

T4: Preemies have a variety of developmental problems, thereby interweaving with and masking thimerosal effects. No valid conclusion can be determined by Davis et al at this time re: premature infants.

T5: The thimerosal/autism co-authors -- and other researchers not affiliated with the ACIP et al -- need to evaluate these CDC studies in their fullest written forms.

T6: Apparently, chelation improvement is not yet important among the CDC/ACIP et al, despite their own study group’s findings of significant neurotoxicity related to thimerosal. Dr. Stephen Edelson’s long-term heavy-metal detoxification model and Bernard et al’s vaccinal-thimerosal/autism model necessarily interact in many ASD children affected by vaccinal Hg. The CDC/ACIP et al ought be calling attention to the need for chelation among susceptible children who were injected with ethylmercury as infants and toddlers.

T7: Hopefully, the co-authors of the Hg/autism paper and other independent scientists and statisticians can provide input at the design and data-interpretation stages of recent and future CDC studies.

As of June, 2000, a cynic might suggest that we have CDC-appointed foxes guarding the thimerosal-data hen-house. In fact, when we look at the organizations responsible for the Joint Statement analyzed here, we see a list of the parties responsible for injecting ethylmercury into susceptible infants and children. Because of this conflict of interest, our proverbial cynic would not surprised that the Joint Statement glosses over the flaws in the CDC/ACIP et al’s thimerosal-studies methodology and offers erroneous conclusions which are protective of the CDC and affiliates.

However, I and my co-authors have met with many Directors from the ACIP. Each of them is sincere and caring. I believe that they will do their best to correct the scientific errors in Davis et al and will move to revise the Joint Statement.

T8: In the two Davis et al study-summaries given to the Hg/mercury

co-authors (CDC offices in DC, June 15, 2000), the Davis et al researchers did not appear to utilize susceptibility factors and thereby erroneously used an artificially high organic-mercury "safe" level as the basis for interpreting the HMOs' data. By using this strategy in creating a model for interpreting the HMOs' data, our cynic would suggest the CDC group was able to fictionally nullify (hide and coverup) the effects of thimerosal. As mentioned in comment T7, I doubt this interpretation. Instead, if and as the ACIP et al use a correct (and lowered) "safe" level for interpreting the HMOs' data, then the HMOs' data already document that, indeed, thimerosal’s ethylmercury has been causing neurologic damage and diagnosed syndromes in children.

T9: The Joint Statement specifically mentions Autism. The Davis et al study that examined autism most specifically was first presented to the American Academy of Pediatrics for its mid-May conference in Boston. The abstract by Davis et al delineates their scientifically erroneous methodology of (a) relying upon the EPA’s recommended level for ingested methylmercury and (b) failing to consider thimerosal’s neurotoxic effects upon individuals susceptible to low-dose bolus exposures during vaccinations. If Davis et al use more scientific modeling (see Hattis et al studies), then the conclusion of Davis et al will be that, indeed, thimerosal has been inducing autism-spectrum disorders in many children. Our society perceives these problems as the vast, costly, and tragic increases in autism-spectrum disorders, developmental delays, ADHD, anorexia, et cetera.

*** annotated Joint Statement ***




••AAFP - The American Academy of Family Physicians,

••AAP - The American Academy of Pediatrics,

••ACIP - The Advisory Committee on Immunization Practices,

••PHS - The United States Public Health Service

Approved on June 22, 2000








This statement has been prepared by the American Academy of Family Physicians, the American Academy of Pediatrics, the Advisory Committee on Immunization Practices, and the U.S. Public Health Service in response to 1) the progress being made in achieving the national goal declared in July 1999 to remove thimerosal from vaccines, and 2) the results of studies to better assess any potential relationship between exposure to mercury in thimerosal

containing vaccines and health effects.


A Joint Statement issued by AAP and the PHS in July 1999 and agreed to by the AAFP later in 1999 established the goal of removing the vaccine preservative thimerosal as soon as possible from vaccines routinely recommended for infants. Thimerosal is a derivative of ethylmercury and has been used as an additive to biologics and vaccines since the 1930s because it is effective in killing bacteria and in preventing bacterial contamination, particularly in opened

multi-dose containers. While there was no evidence of any harm caused by low levels of thimerosal in vaccines and the risk was only theoretical [T1], this goal was established as a precautionary measure.

There is public concern about the health effects of mercury exposure of any sort, and the elimination of mercury from vaccines was judged a feasible means of reducing an infant's total exposure to mercury in a world where other environmental sources of exposure are more difficult or impossible to eliminate (e.g. certain foods).


During the year since the original statement, substantial progress has been made in removing thimerosal from vaccines. A hepatitis B vaccine without thimerosal produced by Merck Vaccine Division was released in August 1999, and in March 2000 a hepatitis B vaccine that does not contain thimerosal as a preservative was approved for SmithKline Beecham Biologicals. This SKBB product contains only a trace amount of mercury (less than 0.5mcg/dose), a greater than 96% reduction from the 12.5mcg in the previous SKBB vaccine and an amount considered clinically insignificant. A combination vaccine containing both hepatitis B and Haemophilus influenzae type b (Hib) vaccine produced by Merck Vaccine Division, Inc. has always been free of thimerosal.

Thus, as of March 2000, all U.S children had access to hepatitis B vaccines that are free of thimerosal as a preservative. [T2] In addition, three of the four Hib vaccines currently licensed for use in the United States do not contain thimerosal as a preservative. The fourth vaccine is produced by Wyeth Lederle which has marketed this Hib vaccine in both thimerosal free, single dose formulations and multidose, thimerosal-containing preparations.

As of July 2000, Wyeth Lederle is expected to produce only the single dose, thimerosal-free formulation for the U.S. Thus, the Hib vaccine supply being produced will become entirely free of thimerosal as a preservative beginning in July 2000.For DTaP vaccine, a thimerosal free vaccine produced by SKBB has been licensed and available in the United States since 1997.

There are three other vaccine manufacturers whose DTaP vaccines still contain thimerosal as a preservative. Discussions are underway with these manufacturers and it is hoped that at least one additional DTaP vaccine without thimerosal as a preservative will become available in early 2001.Based on this progress, the most likely maximum amount of ethylmercury that an infant may be exposed to from the routine immunization schedule has been reduced by 60% from 187.5mcg to 75mcg. Measles mumps rubella, varicella, inactivated polio, and pneumococcal conjugate vaccines have never contained thimerosal.


Since July 1999, efforts to remove thimerosal from the US vaccine supply have been accompanied by research investigations to better assess the potential health effects of exposure to thimerosal-containing vaccines.

First, NIH scientists are collaborating with investigators from the University of Rochester and the Bethesda Naval Hospital to determine retrospectively the blood levels of mercury achieved following routine pediatric vaccination. Preliminary data from a very small number of term infants in these studies indicate that the blood levels of mercury produced by exposure to thimerosal preservative containing vaccines are less than 2mcg/L, the level many experts consider as background. [T3]

These findings differ from those recently reported by Stajich and coworkers (J Pediatr 2000;136:679-681) who found blood mercury levels of greater than 2.9 mcg/L in 9 of 15 premature infants who had received a hepatitis B immunization within the first week of life. However, all of these infants were very premature (birth weights < 1000 grams, mean birth weight of 748 grams). Hepatitis B immunization is not recommended for infants < 2000 grams unless their mother is HBsAg positive.

Second, CDC is using large automated databases that link vaccination and International Classification of Disease codes (ICD-9) stored in medical records in managed care organizations (the Vaccine Safety Datalink project, VSD) to rapidly screen for any possible association between exposure to thimerosal containing vaccines and a variety of neurologic, developmental, and renal outcomes. [T1] In the preliminary screening phase of this investigation, CDC and VSD investigators observed no association between exposure to thimerosal containing vaccines and 12 of the 17 renal and neurologic ICD-9 codes examined from two of the managed care organizations studied. [T1]

These 12 ICD-9 codes examined were extrapyradimal disease, autism [T9], childhood psychosis, stammering, sleep, eating, misery disorders, mixed emotional conditions, infantile cerebral palsy, epilepsy, migraines, and unspecified renal conditions [T8]. From these preliminary data, an inconclusive correlation (i.e., one that is inadequate to support or refute a causal link) was observed between exposure to thimerosal containing vaccines and five of the 17 ICD-9 codes, including language delays, speech delays, attention deficit disorder, unspecified developmental delays and tics. [T1] There was no evidence of any increased risk for these codes among premature infants. [T4;T5]

Reviews of these preliminary observations by expert consultants first at CDC and then from outside PHS identified many important shortcomings in the dataset and in this type of analytic approach. [T1] These consultants concluded that the correlation is very weak and insufficient to support a causal relationship. [T5] This inconclusive information does not provide a sound scientific basis for making new policy decisions or changes in current policies. Nevertheless, because of the potential implications of this screening-phase observation, consultants urged further investigation.

In pursuit of these further studies, CDC investigators have obtained preliminary data from a third managed care organization. Analyses of these data using the same methods and having similar limitations as in the two earlier managed care organizations showed no association for two specific conditions, namely, speech delay, which in this dataset included language delay, and attention deficit disorder. [T1;T5] The number of events was too small to examine the association with tics and the category of unspecified developmental delays was not defined clearly enough to permit reanalysis. Additional review of this dataset is planned and new studies which can test the hypotheses of interest more directly and definitively also are being considered by CDC [T1;T5].


The AAFP, AAP, and the PHS in consultation with the ACIP reaffirm the goal set in July 1999 to remove or greatly reduce thimerosal from vaccines as soon as possible for the following reasons: 1) the removal or substantial reduction of thimerosal from vaccines is feasible, 2) the progress in removal which has been made to date is substantial, 3) the discussions between the Food and Drug Administration and the vaccine manufacturers in removing thimerosal are ongoing, and 4) the public concern about the use of mercury of any sort remains high. Based on information from the FDA and manufacturers, the PHS projects that the United States will complete its transition to a secure routine pediatric vaccine supply free of thimerosal as a preservative (i.e. at least two vaccine products each for Hep B, Hib, and DTaP) by the first quarter of 2001. The use of any Hib or DTaP vaccine should continue according to the currently recommended schedule.

The risk of not vaccinating children on time with DTaP to protect them against pertussis or with any remaining Hib vaccine is believed to far outweigh the risk, if any, of exposure to thimerosal containing DTaP and Hib vaccines which are still available or still being produced. Any new information from ongoing investigations will be monitored carefully by the PHS to determine if any change in this assessment and in existing recommendations is warranted. [T6]

Other vaccines such as diphtheria-tetanus, meningococcal, and influenza vaccines will still contain thimerosal after the first quarter of 2001. Diphtheria-tetanus (DT) and meningococcal vaccines are not recommended for children as part of the recommended childhood immunization schedule. Influenza vaccine is not recommended routinely for infants under 6 months of age, but should be given to infants and children 6 months of age and older who are at high risk of morbidity and mortality from the influenza virus. Continued use of these products as indicated is recommended until thimerosal is removed or until new products without thimerosal are licensed.

The vaccination of children in much of the world will continue to require the use of multi-dose vials for reasons of cost, production, and storage capacity. Multi-dose vials require a preservative to prevent microbial contamination after the vial is opened. While thimerosal is currently the preferred preservative, manufacturers are encouraged to seek alternatives.


In 1999, family physicians, pediatricians, federal health officials, and vaccine manufacturers stated that because any potential risk from mercury is of concern, and the elimination of exposure to mercury in the form of thimerosal from vaccines is feasible, thimerosal should be removed from vaccines as soon as possible. However, there remains no convincing evidence of harm caused by low levels of thimerosal in vaccines.Since mid-1999, two new hepatitis B vaccine products have been introduced and one new Hib product will be produced next month to make the new supply of both hepatitis b and Hib vaccines for infants entirely free of thimerosal as a preservative. One of the four licensed DTaP vaccines is already thimerosal free, and at least one other thimerosal free DTaP vaccine is anticipated to be licensed by early 2001.

Thus, [in the future] the likely maximum number of micrograms of ethylmercury that an infant may be exposed to from the routine immunization schedule will have been reduced by 60%. This amount will be reduced even further in early 2001 when at least two vaccine products for hepatitis B, Hib, and DTaP are expected to be available. Meanwhile, research on the potential health effects of exposure to thimerosal is continuing [T7], and information will be monitored closely by the PHS to determine if any changes in policy are needed. The AAFP, AAP, ACIP, and the PHS recommend continuation of the current policy of moving rapidly to vaccines which are free of thimerosal as a preservative. Until an adequate supply of each vaccine is available, use of vaccines which contain thimerosal as a preservative is acceptable.

*** End of "Joint statement" ***



Addendum 4.


Suggestions for additional research

1. Future studies (a) should avoid using an assumption that bolus doses can be converted into daily average doses, and (b) should utilize a model wherein exposures and neurotoxic effects are cumulative.

2. Data evaluation should be carried out to 2 years of age when all immunizations, with the exception of boosters, are complete. This extension of time would add another 50 mcg or more of exposure to ethyl mercury from thimerosal since the DTaP and Hib -- which are recommended during this time – were not included in the Davis et al data which only tallied exposures within an infant’s first 6 months.

3. Consideration should be given to RhoGam. Since 15% of the population is RH- and would receive 30 mcg of ethyl mercury at 28 weeks gestation, these additional (and fetal) exposures need either (i) be calculated in the data or (ii) be addressed as a limitation to the study as a variable Davis et al were not able to control.

4. Similarly, a substantial percentage of pregnant U.S. women receive flu shots that contain thimerosal. These exposures would add to the total ethylmercury in many individual children and ought be included in reanalyses of the HMO data.