Prenatal and Infant Exposure to Thimerosal From Vaccines and Immunoglobulins and Risk of Autism
Cristofer S. Price, ScM; William W. Thompson, PhD; Barbara Goodson, PhD; Eric S. Weintraub, MPH; Lisa A. Croen, PhD; Virginia L. Hinrichsen, MS, MPH; Michael Marcy, MD; Anne Robertson, PhD; Eileen Eriksen, MPH; Edwin Lewis, MPH; Pilar Bernal, MD; David Shay, MD, MPH; Robert L. Davis, MD, MPH; Frank DeStefano, MD, MPH Corresponding Author
Pediatrics (2010) 126 (4): 656–664. https://doi.org/10.1542/peds.2010-0309
| AAP: Researchers reviewed managed care organization records and conducted interviews with the parents of 256 children with autism spectrum disorder (ASD). Another 752 children without autism, matched to the ASD children by birth year, gender and managed care organization, were also studied. Prenatal and early-life exposure to mercury from thimerosal-containing vaccines and immune globulin (a substance used to boost the immune system in people with certain conditions) was not related to increased risk of ASDs. CDC – “This study compared children with Autism to those without, and looked at prenatal and infant exposure to thimerosal from vaccines. This study found no difference in exposure to thimerosal between children with and without Autism.” |
OBJECTIVE:
Exposure to thimerosal, a mercury-containing preservative that is used in vaccines and immunoglobulin preparations, has been hypothesized to be associated with increased risk of autism spectrum disorder (ASD). This study was designed to examine relationships between prenatal and infant ethylmercury exposure from thimerosal-containing vaccines and/or immunoglobulin preparations and ASD and 2 ASD subcategories: autistic disorder (AD) and ASD with regression.
METHODS:
A case-control study was conducted in 3 managed care organizations (MCOs) of 256 children with ASD and 752 controls matched by birth year, gender, and MCO. ASD diagnoses were validated through standardized in-person evaluations. Exposure to thimerosal in vaccines and immunoglobulin preparations was determined from electronic immunization registries, medical charts, and parent interviews. Information on potential confounding factors was obtained from the interviews and medical charts. We used conditional logistic regression to assess associations between ASD, AD, and ASD with regression and exposure to ethylmercury during prenatal, birth-to-1 month, birth-to-7-month, and birth-to-20-month periods.
RESULTS:
There were no findings of increased risk for any of the 3 ASD outcomes. The adjusted odds ratios (95% confidence intervals) for ASD associated with a 2-SD increase in ethylmercury exposure were 1.12 (0.83–1.51) for prenatal exposure, 0.88 (0.62–1.26) for exposure from birth to 1 month, 0.60 (0.36–0.99) for exposure from birth to 7 months, and 0.60 (0.32–0.97) for exposure from birth to 20 months.
CONCLUSIONS:
In our study of MCO members, prenatal and early-life exposure to ethylmercury from thimerosal-containing vaccines and immunoglobulin preparations was not related to increased risk of ASDs.
| FINANCIAL DISCLOSURE: Dr Marcy received honoraria for speaking for Merck and GlaxoSmithKline within the last 5 years and grant support for studies on Gardasil and ProQuad from Merck within the last 5 years; Mr Lewis received grant support from Medimmune, Sanofi Pasteur, Chiron, Wyeth, Merck, and GlaxoSmithKline; and Dr Bernal received research funding from the CDC, the National Institute of Mental Health, Health Resources and Service Administration, and Autism Speaks. The other authors have no financial relationships relevant to this article to disclose. |
Comments
4 Comments
Mercury poisoning: a diagnosis easier to bear
December 30 2010
Annamartina Franzil
Dear Editor,
We read with great interest the article by Price et al.(1) on exposure to thimerosal, a ethyl-mercury containing preservative used in vaccines, and risk of autism spectrum disorder (ASD). Organic mercury products, such as methyl- mercury and ethyl-mercury, can have severe neurotoxic effects, as shown both in laboratory studies and in cases of accidental exposure in humans (Minamata disease). As a consequence, it has been raised the hypothesis that neurological disorders, including ASD, could depend on exposure to environmental or pharmacological mercury. The study from Price et al. showed that ethylmercury exposure from thimerosal-containing injection administered prenatally or during infancy was not related to increased risk of ASD. In fact, pathogenesis of ASD still remains poorly understood and etiologic hypothesis are often regarded with interest by parents, in the hope to improve the cure of their children. But the work of Price should remember the importance of not diverting the attention from consolidate care of ASD children while following uncertain interpretation of the disease.
One, among unfortunately too much examples, can be the case of a young girl, referred to our hospital at the age of two years because of developmental regression. The neurological examination showed opposite behaviour, non communicative speech limited to stereotyped babbling, limited interaction with the environment, stereotyped hand movements and mild signs of self injury. The chromosomal map and X-fragile test were negative and a MRI of the head failed to show any alterations. At the age of four, stereotypies were dominant, particularly of the hands; comprehension of verbal orders was absent; apraxic gait started to appear, as well as purposeless smiling. EEG showed paroxysmal discharges. Given the clinical features, the girl was discharged with the diagnosis of Rett syndrome. However, the girl’s parents decided to seek other opinions, contacting different physicians also in the field of alternative medicine. One of these practitioners provided an alternative interpretation of the symptoms, diagnosing a neurotoxic effect due to heavy metals and vaccinations. However, mercury poisoning, described as Minamata disease, is characterised by quite different clinical features, including finger, eyelid and tongue tremors, gastroenteritis, stomatitis, lethargy and irritability. Nevertheless, hair and blood mercury levels in hairs and blood were measured. The mercury dosage in hair samples revealed high exposure to methylmercury in the girl and in her family (girl 9.8 mcg/g, father 7.0 mcg/g, mother 4.7 mcg/g, sister 4.1 mcg/g). These concentrations are typical of the “frequent fish consumers” according to WHO references(2). This was explained by the large amount of fish which the family consumed, since the girl’s father was employed in the local harbour. This theory was supported by the fact that the mercury concentration in the local sea was higher than average, inducing some physicians to suggest a possible etiologic link between Rett syndrome and Methylmercury poisoning(3). However, it is important to note that no confirmed cases of mercury poisoning were ever registered in this area. Anyway, on the basis of this new diagnosis, the girl started selenium therapy to decrease the mercury levels and was given melatonin with multivitamin products with the hope of reacquiring the lost skills. During the following 3 years contacts with the patient and her family were impossible. The family returned to the Pediatric Department of our hospital requesting a lumbar puncture and the measurement of mercury levels in liquor. Considering the clinical history and the previous clinical diagnosis of Rett syndrome and as well the doubtful scientific basis of the mercury poisoning hypothesis, the lumbar puncture was not performed. On the other hand, a molecular test for the diagnosis of Rett syndrome (not available at the first diagnosis) was performed, which showed the deletion of 41 bases pairs in the exon 3 of the gene MECP2. The diagnosis of Rett syndrome, clinically formulated 5 years before, was finally confirmed by molecular evidence. The genetic diagnosis allowed us to deny a major role of mercury poisoning in the disease.
Environmental factors such as vaccines or some foods, which are produced by human activity, are often perceived as non- natural, and thus are looked at as a potential cause of illness. Large epidemiological studies are very important to demonstrate the safety of these human interventions on public health because sometimes the weight of environmental factors could be overestimated. This is important especially in that cases where is possible to find out a definite diagnosis, allowing in this way parents and physicians to concentrate their efforts on the care of the child.
References:
1. Cristofer S. Price, William W. Thompson, Barbara Goodson, Eric S. Weintraub, Lisa A. Croen, Virginia L. Hinrichsen, Michael Marcy, Anne Robertson, Eileen Eriksen, Edwin Lewis, Pilar Bernal, David Shay, Robert L. Davis, and Frank DeStefano. Prenatal and Infant Exposure to Thimerosal From Vaccines and Immunoglobulins and Risk of Autism Pediatrics 2010; 126: 656-664
2. WHO-IPCS, Environmental Health criteria 101, Methylmercury, WHO, Geneva, 1990
3. Kobal AB, Miklavcic V, Byrne AR, Velickovic Perat M. Methylmercury exposure and Rett’s syndrome? In 8th International Child Neurology Congress, Ljubljana, Slovenia, 13-18 September 1998, Monduzzi; Bologna, Italy
Conflict of Interest:
None declared
Submitted on December 30 2010
Conclusions should be based on appropriate methodologies
November 4 2010
Raymond F Palmer
While the Price study is a well-conducted careful investigation, this case-control design would not detect any true differences between the groups based on what most prominent researchers suspect to be true about ASD – that it is a disease that involves genetic susceptibility to environmental triggers. Epidemiological case-control studies that fail to include study variables about individual susceptibility are not sufficient to understand interactions with ubiquitous exposures.
Their study sample of children born between 1994 and 1999 were among a cohort of children who were exposed to thimerosal from routine childhood vaccines – a ubiquitous population exposure. It is doubtful that exposure to thimerosal would differ between any group of individuals during that time no matter what disease was being investigated.
The finding of no differences in exposure between cases and controls in the Price et al., study is not surprising and certainly does not exonerate thimerosal as a potential trigger in susceptible individuals. I would suspect that many ASD children differ on a genetic susceptibility in their innate ability to detoxify thimerosal and not their differential exposure to it. The design used in the Price et al study cannot possibly determine this distinction because susceptibility was not a study variable. Given a gene/environment interaction (G x E) model of disease, a design that looks only at environmental exposure is not appropriate and null results should not mislead scientists or policy makers into asserting the safety of thimerosal for all. A proper case-control design would test a G x E hypothesis – e.g. among genetically susceptible individuals, thimerosal exposure may influence the development of ASD.
I am disappointed that the conclusions put forth by the authors did not mention the important caveat that the case-control design they used, would not be sufficient to investigate the prevailing consensus of G x E in the development of ASD. My major concern is that the null results of this and other studies will translate into a potentially erroneous public health messages that thimersiol is safe for all, when in fact approximately 10% of the population have hypersentive reactions to thimerisol – which is one reason it was removed from many over-the-counter products in the first place. To date there has never been a proper biological study to determine the safety of thimerisol in the general population.
Conflict of Interest:
None declared
Submitted on November 04 2010
Thimerosal, Autism and Hormesis
October 2 2010
Richard C. Deth
Dear Sirs:
The recently released case control study by Price et al. (1) examined a possible association between autism and ethylmercury derived from the vaccine preservative thimerosal and surprisingly found fewer than expected autism case among subjects who received higher thimerosal doses. This apparent protective effect was a robust finding, reaching significance in all six cohorts receiving >100 µg of thimerosal, representing a reduction in autism risk of 40 to 78% in their covariate-adjusted model. Clearly the dose of thimerosal has a significant impact on the likelihood of developing autism, albeit a puzzling beneficial one, leading the authors to state that they “…are not aware of a biological mechanism that would lead to this result”. However, the findings of this study strongly suggest a hormetic response to thimerosal, in which low doses produce a beneficial response to a toxic substance, reflecting mobilization of adaptive mechanisms which protect against successive exposures. Indeed, the episodic nature of thimerosal administration is particularly well- suited to engendering a hormetic response, with sufficient elapsed time between doses for metabolic and epigenetic mechanisms to augment resistance.
A number of studies have reported hormetic responses following exposure to inorganic or organic forms of mercury (Helmcke et al., 2010; Calabrese, 2008; Calabrese, 2005; Toimela and Tahti, 2004; Prati et al., 2002; Brousseau et al., 2000; Fournier et al., 2000; Contrino et al., 2002; Zdolsek et al., 1994; Contrino et al., 1988; Nordlind and Henze, 1984; Dieter et al., 1983), and an earlier preliminary report similarly described a protective influence of thimerosal vs. autism (Jones, 2003).
Hormesis is a dose-response phenomenon characterized by low-dose stimulation and high-dose inhibition, commonly exhibited by toxic exposures. Many xenobiotics, including thimerosal, promote oxidative stress and cells possess a large number of adaptive mechanisms which augment antioxidant resources in response to an initial exposure, resulting in enhanced tolerance to the deleterious effects of subsequent exposures. Adaptation can involve short-term (e.g. metabolic shifts) or long-term mechanisms (e.g. epigenetic-mediated changes in gene expression). Hormetic responses underlie the oft-described benefit of pre- conditioning or post-conditioning stress (Calabrese et al., 2007), and they offer obvious survival advantages, especially in response to a challenging environment.
However, the capacity to mount an effective hormetic response is not necessarily equal across the population, and it importantly depends upon the relevant metabolic pathways, including their genetic underpinnings. Significant impairment of antioxidant and methylation metabolite levels is well-documented in autism (Adams et al., 2009; Al-Gadani et al., 2009; Pastural et al., 2009; Paºca et al., 2009; Vojdani et al., 2008; Geier and Geier 2006; James et al. 2006; James et al., 2004), along with a higher prevalence of risk-inducing single nucleotide polymorphisms (SNPs) in genes which support these pathways (Paºca et al., 2009; James et al. 2006). Thus exposure to a toxin such as ethylmercury might lead to a protective hormetic response in most individuals, but this response would be less robust or absent in other individuals, increasing their risk.
Notably, thimerosal is utilized as a preservative precisely because of its toxicity at high concentrations, and there is an obvious need to fully characterize its activity, including both positive and negative effects on neurodevelopment, across a complete range of doses. Such studies should also take genetic variations in into account, for example by utilizing animal strains known to harbor SNPs affecting the capacity for a hormetic response.
In any case, the positive findings of Price et al. may be more significant than the negative findings, and may offer an important clue as to the origin of autism.
References Adams, J.B., Baral, M., Geis, E., Mitchell, J., Ingram, J., Hensley, A., Zappia, I., Newmark, S., Gehn, E., Rubin, R.A., Mitchell, K., Bradstreet, J., and El-Dahr JM. (2009). The severity of autism is associated with toxic metal body burden and red blood cell glutathione levels. J. Toxicol. 2009:532640.
Al-Gadani, Y., El-Ansary, A., Attas, O., and Al-Ayadhi, L. (2009). Metabolic biomarkers related to oxidative stress and antioxidant status in Saudi autistic children. Clin. Biochem. 42:1032-1040.
Brousseau, P., Pellerin, J., Morin, Y., Cyr, D., Blakley, B., Boermans, H., and Fournier, M. (2000). Flow cytometry as a tool to monitor the disturbance of phagocytosis in the clam Mya arenaria hemocytes following in vitro exposure to heavy metals. Toxicol., 142:145-156.
Calabrese, E.J. (2008). Pharmacological enhancement of neuronal survival. Crit. Rev. Toxicol., 38:349-389. Calabrese, E.J., Bachmann, K.A., Bailer, A.J., Bolger, P.M., Borak, J., Cai, L., Cedergreen, N., Cherian, M.G., Chiueh, C.C., Clarkson, T.W., Cook, R.R., Diamond, D.M., Doolittle, D.J., Dorato, M.A., Duke, S.O., Feinendegen, L., Gardner, D.E., Hart, R.W., Hastings, K.L., Hayes, A.W., Hoffmann, G.R., Ives, J.A., Jaworowski, Z., Johnson, T.E., Jonas, W.B., Kaminski, N.E., Keller, J.G., Klaunig, J.E., Knudsen, T.B., Kozumbo, W.J., Lettieri, T., Liu, S-Z., Maisseu, A., Maynard, K.I., Masoro, E.J., McClellan, R.O., Mehendale, H.M., Mothersill, C., Newlin, D.B., Nigg, H.N., Oehme, F.W., Phalen, R.F., Philbert, M.A., Rattan, S.I.S., Riviere, J.E., Rodricks, J., Sapolsky, R.M., Scott, B.R., Seymour, C., Sinclair, D.A., Smith-Sonneborn, J., Snow, E.T., Spear, L., Stevenson, D.E., Thomas, Y., Tubiana, M., Williams, G.M., and Mattson, M.P. (2007). Biological stress response terminology: integrating the concepts of adaptive response and preconditioning stress within a hormetic dose-response framework. Toxicol. Appl. Pharmacol., 222:122-128.
Calabrese, E.J. (2005). Hormetic dose-response relationships in immunology: occurrence, quantitative features of the dose response, mechanistic foundations, and clinical implications. Crit. Rev. Toxicol., 35:89-295.
Chauhan, A., and Chauhan, V. (2006). Oxidative stress in autism. Pathophysiology. 13:171-181.
Contrino, J., Marucha, P., Ribaudo, R., Ference, R., Bigazzi, P.E., and Kreutzer, D.L. (1988). Effects of mercury on human polymorphonuclear leukocyte function in vitro. Amer. J. Pathol., 132:110-118.
Contrino, J., Kosuda, L.L., Marucha, P., Kreutzer, D.L., and Bigazzi, P.E. (1992). The in vitro effects of mercury on peritoneal leukocytes (PMN and macrophages) from inbred Brown Norway and Lewis rats. Int. J. Immunopharm., 14:1051-1059.
Dieter, M.P., Luster, M.I., Boorman, G.A., Jameson, C.W., Dean, J.H., and Cox, J.W. (1983). Immunological and biochemical responses in mice treated with mercuric chloride. Toxicol. Appl. Pharmacol., 68:218-228.
Fournier, M., Cyr, D., Blakley, B., Boermans, H., and Brousseau, P. (2000). Phagocytosis as a biomarker of immunotoxicity in wildlife species exposed to environmental xenobiotics. Am. Zool., 40:412-420.
Geier, D.A., and Geier, M.R. (2006). A clinical and laboratory evaluation of methionine cycle-transsulfuration and androgen pathway markers in children with autistic disorders. Horm. Res. 66:182-188.
Helmcke, K.J., and Aschner, M. (2010). Hormetic effect of methylmercury on Caenorhabditis elegans. Toxicol. Appl. Pharmacol. Aug 5. [Epub ahead of print]
James, S.J., Cutler, P., Melnyk, S., Jernigan, S., Janak, L., Gaylor, D.W., and Neubrander JA. (2004). Metabolic biomarkers of increased oxidative stress and impaired methylation capacity in children with autism. Am. J. Clin. Nutr. 80:1611-1617.
James, S.J., Melnyk, S., Jernigan, S., Cleves, M.A., Halsted, C.H., Wong, D.H., Cutler, P., Bock, K., Boris, M., Bradstreet, J.J., Baker, S.M., and Gaylor, D.W. (2006). Metabolic endophenotype and related genotypes are associated with oxidative stress in children with autism. Am. J. Med. Genet. B. Neuropsychiatr. Genet. 141B:947-956.
Jones, D.E. (2003). Abstract. Experiences in nonlinear dose-response relationships in chemical evaluations. NonLinearity Biol. Toxicol. Med., 1:223.
Nordlind, K., and Henze, A. (1984). Stimulating effect of mercuric chloride and nickel sulfate on DNA synthesis of thymocytes and peripheral blood lymphocytes in children. Int. Arch. Allergy Appl. Immun., 73:162- 165.
Paºca, S.P., Dronca, E., Kaucsár, T., Craciun, E.C., Endreffy, E., Ferencz, B.K., Iftene, F., Benga, I., Cornean, R., Banerjee, R., and Dronca, M. (2009). One carbon metabolism disturbances and the C677T MTHFR gene polymorphism in children with autism spectrum disorders. J. Cell. Mol. Med. 13:4229-4238.
Pastural, E., Ritchie, S., Lu, Y., Jin, W., Kavianpour, A., Khine Su- Myat, K., Heath, D., Wood, P.L., Fisk, M., Goodenowe, D.B. (2009). Novel plasma phospholipid biomarkers of autism: mitochondrial dysfunction as a putative causative mechanism. Prostaglandins Leukot. Essent. Fatty Acids. 81:253-264.
Prati, M., Gornati, R., Boracchi, P., Biganzoli, E., Fortaner, S., Pietra, R., Sabbioni, E., and Bernardini, G. (2002) A comparative study of the toxicity of mercury dichloride and methylmercury, assayed by the Frog Embryo Teratogenesis Assay–Xenopus (FETAX). Altern. Lab. Anim., 30:23-32.
Sauve, S., Brousseau, P., Pellerin, J., Morin, Y., Senecal, L., Goudreau, P., and Fournier, M. (2002). Phagocytic activity of marine and freshwater bivalves: In vitro exposure to hemocytes to metals (Ag, Cd, Hg and Zn). Aquatic Toxicol., 58:189-200.
Toimela, T., and Tahti, H. (2004). Mitochondrial viability and apoptosis induced by aluminum, mercuric mercury, and methylmercury in cell lines of neural origin. Arch. Toxicol., 78:565-574.
Vojdani, A., Mumper, E., Granpeesheh, D., Mielke, L., Traver, D., Bock, K., Hirani, K., Neubrander, J., Woeller, K.N,, O’Hara, N., Usman, A., Schneider, C., Hebroni, F., Berookhim, J., and McCandless J. (2008). Low natural killer cell cytotoxic activity in autism: the role of glutathione, IL-2 and IL-15. J. Neuroimmunol. 205:148-154.
Zdolsek, J.M., Soder, O., and Hultman, P. (1994). Mercury induces in vivo and in vitro secretion of interleukin-1 in mice. Immunopharmacology, 28:201-208.
Conflict of Interest:
None declared
Submitted on October 02 2010
Can this study design produce a useful result?
September 21 2010
John D Stone
The logic of this study is puzzling and after many years of the mercury/autism controversy will do nothing to quieten it. It is as if Sir Richard Doll had tried to lay the smoking controversy to rest by concluding that as not all smokers get lung cancer smoking does not cause lung cancer. This rather trite point will be evident to many people and it is surprising that a study itself many years in the pipeline should be open to such basic criticisms. To learn anything much we need an unexposed population. Most people will also assume that the alleged protective effect of mercury against autism reported in the study is an artefact of the design.
Conflict of Interest:
Autistic son
The Canary Party calls on reporters to take a second look at the CDC’s biased science.
(Cambridge, MA) “The CDC continued its propaganda campaign on behalf of its bloated vaccine schedule last week,” said Mark Blaxill, Chairman of the Canary Party. “Despite significant scientific evidence showing connections between vaccines and autism, and deep problems with CDC’s vaccine safety science, few reporters dug deeper into the quality of this new study (DeStefano et al. 2013).” As recently noted in a peer-reviewed publication, Drs Catherine DeSoto and Robert Hitlan[1] documented major methodological flaws in the 2010 CDC study (Price et al.) said to disprove any link between a mercury preservative in vaccines and autism. This second paper regarding antigens, by CDC staffer Frank DeStefano and colleagues, uses the exact same flawed data set, again to deny the link between vaccines and autism.
“How deeply flawed was DeStefano’s analysis?” continues Blaxill, “Simply put, the study design could not have been more biased. The number, type and timing of vaccines that US children receive are a function of birth year: recommendations for DTP, Rotavirus and Varicella all changed during the years of the study, depending on the year the child was born.
But the CDC data set used a data sample that matched cases and controls by birth year and then only analyzed the differences within their patched groupings (called “strata” in statistics). Matching on birth year meant nearly all variation associated with how many vaccines were recommended was removed from the study as a starting point. If they had said ‘we are controlling for the vaccine schedule the child followed’ in an analysis of how safe the vaccine schedule was, this would have seemed absurd. But that is exactly what they did.”
“The model they were trying to test in their first study was whether exposure to Thimerosal via vaccination was associated with any increased risk of autism. To do this, they needed to compare persons with different levels of exposure. They could not do so because they matched on birth-year, which itself defines exposure level. This ensures that cases were only compared to controls with the same exposure,” said lead author Dr. Catherine DeSoto. “Children who received high and low exposure were not compared with their methods. It is like testing if smoking causes lung cancer but only comparing persons with and without cancer who smoked exactly the same amount — and then statistically testing
if they smoked different amounts. The design flaw is called overmatching, and it makes
the results of both studies invalid.”
“There are additional valid concerns about the study design that have been raised by others and the press should be reporting them,” states Blaxill. “The Canary Party maintains that the CDC, as a government entity, is obligated to promote the official vaccine schedule, and herefore the media should critically review any CDC statements regarding vaccine safety.”
From the Edited Volume
Recent Advances in Autism Spectrum Disorders – Volume I Empty heading
Edited by Michael Fitzgerald
Chapter 6
Vaccine Safety Study as an
Interesting Case of “Over-Matching”
M. Catherine DeSoto and Robert T. Hitlan
Additional information is available at the end of the chapter
http://dx.doi.org/10.5772/53876
1. Introduction
Increasing levels of diagnosed cases of autism have alarmed parents and health officials, but the cause has not been established. It has been hypothesized that vaccination itself, or some component in vaccines, may be somehow related to the onset of autism in some cases (Delong, 2011; Gallagher & Goodman, 2010). Researchers have sought to alleviate such concerns. Although most studies report null effects, work continues to be published that suggests some reason for concern (Hewiston et al., 2010). Some skepticism of the safety of vaccines still exists, documented by scholars on either side of the issue (Austin,
Schandley & Palombo, 2010, Destafano, 2007). As it is, the topic of vaccine safety and triggering of unintended outcomes is one of the most controversial topics in environmental health and toxicology.
After initial safety studies, case- control designs are often employed to continue to investigate both side effects and efficacy of inoculation. Matching is a technique used to improve signal to noise in research case-control designs. Matching cannot – or should not – be done in a way that artificially increases the chance that within strata exposure is the same. This happens when a matching variable is a strong predictor of exposure and is called overmatching. Here, we report a textbook case of overmatching within a widely – cited article.
Focusing on the overmatching as a statistical concept, suggestions are made to standardize when overmatching may have occurred. It is important for statisticians to note when a study that fails to find an effect related to public health outcome has employed a design that would be expected a priori to result in a lack of effect.
It has been noted that some children received exposure to mercury significantly in excess of safety standards during the 1990’s, before the level of thimerosal in vaccines was lowered (Geier & Geier, 2006), this has been suggested to increase odds of various developmental disorders (Geier & Geier, 2006). The research by Price et al. (2010) spans the birth cohort years that saw a decline in thimerosal exposure and reports that thimerosal exposure was not associated with risk outcome of autism. Indeed, many studies have been published that find no negative effect of vaccination on developmental outcomes whatsoever (Parker, Schwartz, Todd, Pickering, 2004; see Destafano, 2007 for a review), indicating a lack of cause and effect between vaccination and autism. Here, we suggest that a recent widely cited study was flawed, and urge statisticians to carefully and critically review outcomes research on high stakes topics. It should be noted and understood that a flaw in such a study does not mean that vaccines cause autism, nor does it follow that one would properly assume that the flaw leads to the conclusion that vaccination is not safe. Rather the weight of scientific research as a whole should be deferred to
Conditional logistic regression (CLR) is a statistical technique used when the researchers have matched cases with controls on various parameters (e.g., age, gender). CLR is the of‐ten-used and appropriate way to analyze matched data sets (Rahman, Sakamoto & Fukui, 2003). To be clear, matching means that (as an example) for every ‘case’ that is male and aged 12, there is a control selected from a pool of possible controls that is also male and aged 12. If this were done, the researchers “matched on age and gender.” A variant is to have two or three times the number of controls within each condition, or stratum. (Meaning for every male case who is age 12, there are three controls who are male and age 12.) The matched unit is called a stratum. When analyzing the data, CLR analyses are done within strata. When matching is done, only conditions (strata) that have cases and control pairs that vary on the risk factor contribute to the estimate of the effect of the risk factor (Miettinen, 1968). In other words, if exposure level within strata is the same, CLR cannot estimate the effect. As such, matching is a key design feature.
Matching cannot – or should not – be done in a way that artificially increases the chance that within strata exposure is the same; this happens when a matching variable is a significant predictor of exposure and is called overmatching.
Proper design can have important implications and researchers are appropriately cognizant of the possible perils of failing to take enough care in considering the matching design. If matching is used, researchers are wise to give explicit consideration to ensure that the problem of overmatching is avoided when attempting to accurately estimate risk of an exposure of interest (Sasieni and Castanon, 2009; Al-Taiar et al., 2009; Vidal et al., 2008; Agudo & Gonzalez, 1999; Cullison et al., 2007). And this problem has long been known (see for example, West, Schuman, Lyon, Robison & Allred, 1984). In their consensus paper on outcomes research, the American Thoracic Society noted that, “Overmatching, matching for a variable that is associated with the exposure but not the outcome, will reduce the statistical power of the study,” (p. 364). Improper matching cannot later be undone via analysis and the effect of the matched variables cannot be checked, once matching has been done (Rubenfeld et al., 1999). How could this happen? Usually, this arises when a researcher fails to realize he or she is essentially matching on the exposure variable, and inadvertently the researcher matches the effect out.
To illustrate overmatching, a fictitious example will be briefly discussed, followed by an actual example from the literature. Assume the question is whether radiation exposure in nuclear plant workers contributes to cancer. A hundred cancer cases are found, and a control group of 700 is identified. Then, each case is matched with one from the control group on gender, smoking, job location, and age. The researchers match on these variables to increase efficiency (because they think these variables might independently account for disease risk). We will keep this as one to one matching for simplicity, but a 1:3 matching would essentially work the same.
In this example, overmatching would happen if the researchers are looking for effects of radiation but fail to consider that while which power plant the worker is employed might have some independent influence on disease risk (which is why it is matched), location could also be a major determinant of radiation exposure. For example, imagine Plant L often had radiation leaks, while Plant S had better safety. If one then matches on where one works, all of the variance unique to a particular plant is matched out. In such a case, an effect for radiation – even if huge could be missed. It will be clear if one considers that this would be like testing if radiation was related to cancer in Japanese nuclear power plant
workers after controlling for location with one of the locations being Fukishima (Figure 1). If participants who developed cancer were matched on where they worked – the researchers may not detect any true health effects of the radiation exposure from the nuclear meltdown at Fukushima compared to working at other plants that did not have a meltdown. The researchers would have matched out any effects associated with where they worked.
Figure 1. Overlapping variance: Illustration of Overmatching on Radiation Exposure; In this fictitious example, matching on the nuclear power plant of employment in the design of the study would be overmatching because it would
remove the largely overlapping variance associated with radiation due to the Fukishima leak, obscuring the effect.
A now classic paper by Marsh, Hutton and Binks (2002) refers to a real research example and is entitled, “Removal of radiation dose response effects: an example of over-matching.” It details how a true effect can be missed if the researchers overmatch. According to the authors, “If the exposure itself leads to the confounder or has equal status with it, then stratifying by the confounder will also stratify by the exposure, and the relation of the exposure to the disease will be obscured. This is called over-matching and leads to biased estimates of
risk,” (p. 1235). After previous work had suggested that radiation did predict leukemia, the more recent case-control study failed to indicate any relation between radiation and leukemia. The matched factors in the new study that showed no increased for leukemia as a result of radiation included: date of birth, gender, and “date of entry”. “Date of entry” was a measure of what years the workers worked in the industry. The data was properly analyzed given the matched design by conditional logistic regression, yet failed to find a known effect.
This prompted the study of the statistics used, with a focus on the matching process. It was noted that some things are appropriate to match on, for example, gender. “Because of the underlying difference of the risks of leukemia between the sexes,” being male versus female affects the outcome, and it is important not to accidently have more males in the case group as this would be a confound. On the other hand, Marsh et al. clearly showed that radiation exposure varied by year, that is some years were higher than others and this was indeed a major source of radiation variation (see figure 3, Marsh et al., 2001). “The general decline in median dose shows that dose and time are associated. The situation seems to be one where dose is partially ‘explained’ by date of entry, both being related to time;” in sum, “this seems to have had the effect that workers in the same matched set have broadly similar recorded doses. The apparent over-matching on date of entry has distorted the parameter estimate of
the risk of leukemia on cumulative dose by introducing matching (at least partially) on dose,” (Marsh et al., 2002).
What is the take home message of this classic report on the problem of overmatching? When researchers match on a variable closely associated with the risk factor exposure, then actual effects will not be– and cannot be– detected. This danger is written about by various other authors as well. Richard Monson in his text, “Occupational Epidemiology” notes “over matching is a problem in case control studies.” Monson emphasizes that “there should be no possibility that the factor is part of the causal pathway linking exposure and disease under study.” (p. 41). If this is even remotely possible, Monsoon advis‐
es matching should not be done on that variable. Monson discussed an example where overmatching resulted in underestimating the effect of estrogen use on endometrial cancer. Here the matching was on a correlate of intrauterine bleeding, which in effect controlled for a symptom of the cancer itself.
Price et al. do not mention overmatching as a potential concern. The risk factor of interest is thimerosal exposure via its inclusion in vaccine ingredients. There are two things that have a systematic and predictable effect on how much thimerosal exposure a child would receive: 1) the vaccine schedule a child is born into/national recommendations, and 2) which manufacturer a given provider is using for the vaccines (e.g. for the same years, Smith, Kline and
Beecham were using thimerosal in their HepB vaccine, while Merck did not).
Figure 2. Controlling for Birth Year is overmatching due to the overlap with Amount of Exposure; similar to the radiation risk for leukemia written about by Marsh, controlling for time is (at least partly) controlling for exposure, which
varies with birth year. The matching on birth year is matching on the exposure. This seems to have had the effect that children in the same matched set have similar recorded exposures to thimerosal, removing much of the variance
Price et al. matched out both of these variations in exposure. This has the effect of ensuring that the control group is nearly identical with the case group on the risk factor, which prevents its effect from being accurately measured. Considering cumulative exposure for thefirst 7 months of life, the overall mean for the full data set is 102.88 micrograms/Hg and a standard deviation of 42.2. The means for the cases and matched controls is 100.0 and 103.2 micrograms of Hg: this similarity (less than one tenth of the standard deviation) is forced by
the matching on the variables that define exposure. Birth year dictates which vaccine schedule a child is born under as well as which batch brands and formulations are available on the market at a given time. Doctors within a practice will be using the same manufacturer across children (vaccines are ordered in large batches room a given manufacturer; the Vaccine Data Set used by Price et al. documents that the same providers use the same manufacture. Thus, this is a text book case of overmatching: variables were matched on that essentially define exposure. It is well known that matching on a variable that is associated only with exposure, not with disease, reduces statistical efficiency (Zondervan et al, 2002; Rubenfeld et al., 1999; Day, Byar, & Green, 1980) and that care needs to be taken to avoid this in a case-control research design.
Across the different years, the average cumulative exposure varies from 42.3 micrograms to 125.46 micrograms; while within the birth year stratas, the mean exposures do not vary by more than 15 micrograms. Birth year is a variable that defines exposure due to changes in recommendations regarding the vaccine schedule and changes in vaccine formulas that occurred at different times. The above panels suggest that variance within the matched variable (year) is small compared to the variance between birth years: birth year is accounting for
much variance in thimerosal exposure.
Figure 3. The difference across birth years on the risk factor of interest
During the past decades, there have been three main exposure sources of thimerosal: DPT/DTaP, then Hepatitis B and Hib vaccines, while flu shots are currently the primary source in the USA today. The Hib/Hep B introductions came in during the late 1980s and early 1990s. The recognition that the cumulative mercury burden may have been too high came in 1999, and mercury levels dropped for most vaccines given to children in the USA. Some people have raised concerns that the increase in autism is associated with the changes in thimerosal exposure; that is, the increase in autism is thought to be a function of the increases in the number and amount of mercury containing vaccines. Whether or not one finds this model persuading, matching on birth year is questionable if the goal is to test the model that differences in thimerosal exposure via vaccine schedule increase ASD risk since — as most people
are aware — birth year essentially dictates which vaccine guidelines a child is born into. It could be that the authors intended to control for hypothesized changes in diagnostic criteria trends across the six birth years. The problem is that diagnostic effects on risk is not measured while birth year effects on exposure are clear.
Moreover, HMO is not known to be a significant predictor of the outcome of autism diagnosis, so potential reasons to match on this variable are less clear. As Hansson and Khamis (2008) write in their paper on matched-sample logistic regression, “Generally, matching will increase the efficiency of the study when the matching variable is a strong outcome determinant, but will actually reduce it when the matching variable is strongly related to the exposure variable (over-matching),” (p.595-596). Meittinem (1969) states that, “matching reflects the notion that the probability P of response is related to M,” (p. 340) meaning that when one matches, one infers that the matching variable effects the probability of risk (here for autism). HMO / health care provider was a major determinant of thimerosal exposure, but we are not aware of papers that identify HMO is an independent risk for autism. Thus, it should not have been matched. What was needed was a design that compared persons with different exposures. “Studies with uniform developmental assessments of children with a range of cumulative thimerosal exposures are needed,” (Vertraeten et al., 2003). Here Price
et al., began with such a data set, but then matched on birth year and HMO, matching out exposure differences and negating comparisons of different exposures (see Miettinen, 1969 for a mathematical discussion).
Figure 4. The apparent over-matching on HMO distorts the estimate of the risk of autism on thimerosol by introducing matching on exposure. If one matches on provider, one is matching on the vaccine manufacturer. There are different manufacturers available, but a given provider will be using one or the other. This seems to have had the effect that children in the same matched set have similar recorded exposures to thimerosal. Again, this removes this variance and
obscures the effect
The model Price et al. were trying to test was whether thimerosal exposure via the US vaccination schedule was associated with any increased risk of autism. To do this, they needed to compare persons with and without high levels of exposure. They did not do this because due to the conditional logistic regression matched on both birth year and HMO they have inadvertently made sure that cases were only compared to controls with the same exposure. Because Price et al. did not mention the possibility of overmatching, we assume this did not occur to the research team. We assume this was accidental, but it does underscore the need to have a balanced research team that does not start with assumptions that might flaw the design. For example, assuming that the increase in autism
is only due to diagnostic changes would lead to controlling for birth year, which might have been flagged by someone who does not share this assumption. It is harder to understand why HMO would be matched. Overall, this is unfortunate because the question of vaccine safety is high stakes. There are concerns that a proper test of the full vaccine schedule has not been properly tested, and that the safety tests that exist have been designed by the vaccine industry itself. Such concerns about conflicts of interest may be preventing otherwise willing parents to adhere to the full vaccine schedule. Vaccines have been and will continue to be a huge benefit to humanity. But this paper is flawed. Unfortunately, there is not an analytic fix for overmatching: it is design flaw.
Figure 5. Which manufacturer a given provider used for the vaccines varied by HMO. Manufacturers differed in their thimerosal use. For example, in 2002, Smith, Kline and Beecham were using thimerosal in their HepB vaccine, while Merck did not. While the data set is careful to note manufacturer and Hg in the associated batch and manufacturer,
but CLR matching on HMO results in comparing cases to controls who had the same levels of exposure
The Price et al. research is an interesting case of overmatching that we think is of general
interest in the field of epidemiology. To avoid misunderstanding, we wish to state that this
research does not support the argument that vaccines or thimerosal in vaccines cause au‐
tism. It is however, uninformative to the question.
2. Suggestions for avoiding the problem of publishing overmatched
results
One way to conceptualize the problem of overmatching in conditional logistic regression is
the preemptive removal of variance that should stay available for the hypothesized predic‐
tor variable to attempt to account for. The total variance in a data set can be defined using
the average squared distance from the mean score for each participant: s2= SS/df. A question related to overmatching concerns how much of the total variance is taken out beforehand (matched out)? How much is too much? 10%? 90%? 50%? We would propose that the percent removed before testing should normally be small compared to the total. Further, the removal of this variance should only occur when there is authentic need: when the potential matching variable is likely related to the outcome of interest via a path that is distinct from the risk variable of interest in a case-control design.
As elaborated above, matching is appropriate only if the matching variable is a strong pre‐
dictor of the outcome of interest, but it is not appropriate when the matching variable is
strongly related to the exposure risk variable. We offer three suggestions to help objectively
identify, and thus avoid, the problem of overmatching.
Empirical Support. Before matching, first and foremost, researchers should locate studies
that suggest the potential match is likely correlated to the probability of the outcome occur‐
ing. These should be cited to support the need to match on that variable. If there is no reason to think the matching variable relates to the outcome, there is no reason to match it.
Remaining Variance. Next, once the participants have been selected as a matched data set, researchers can check to get an idea how much variance in the exposure variable is actually ccounted for by the matching variable M. If only a small amount of the variance is left after the various matching, matching on the variable(s) cannot be justified and an unmatched or lesser matched set of participants is called for. Specifically, a check to see if too much of the total variance in the outcome of interest is matched out could be done by requesting Partial Eta Squared. Partial Eta Squared represents the proportion of the total variance that is explained by the between factor when an ANOVA is performed. Specifically, one can take the extra step of analyzing the variance in the risk factor of interest (e.g., thimerosal exposure) as a function of the matched variable (e.g., HMO or BirthYear). In this example, using thimerosal exposure as the dependent variable, the total SS is 23507522. The SS associated with the Birth Year is 1485471. This gives Partial Eta Squared =.456, meaning that about 46% of the total variance in thimerosal exposure is fully explainable based on Year of Birth. When one matches on this, only about half (54%) of the variance is left.
HMO, the other variable matched on, removed about 30% of the variance.
The percent that should be left would depend on the research question and causal assump‐
tions, but we suggest that if a matched variable is removing more than a fourth (25%) of the
variance (corresponding to a large effect size, Cohen, 1977), matching is unlikely to be war‐
ranted for this reason alone and welcome commentary on this benchmark proposal.
Relative relations. Finally, there are times when it could be proper to match on a variable
that accounts for variance in the risk factor being tested. A recent case coincidentally also
related to vaccines helps to illustrate this more. It had been pointed out that the enormous
benefits of the flu vaccine among the elderly appeared to far surpass even the effect that a
total eradicating of flu from the vaccinated population could account for (Jefferson, 2006).
After additional investigation, much of the original effect appears to be due to the tendency
for seriously ill and/or less healthy elderly persons not to have the flu shot. To be clear, most
of the flu vaccine effect on mortality was found to be due to health of the participants inde‐
pendent of the flu shot (Jackson et al., 2006). In this case, if this had been a case control design, the risk factor would be flu vaccine and the probability outcome of interest would
hospitalization or death. In such a case control study, it would be proper to match on preex‐
isting health, even though one would find that health accounts for some of the variability in
getting or not getting the risk factor (flu vaccine). BUT: health would also relate to the mor‐
tality outcome, and even more strongly. It is this strong relationship that is key. If the varia‐
ble is more strongly related to the outcome – this serves to justify matching.
To objectively quantify this, one needs to know how strongly M is related to the Risk Factor
R; and then how strongly M is related to the probability of response P. A problem is that
different types of data can make precise comparisons of effect size hard to judge.
Assume that M would be HMO, R would be mcg Thimerosal exposure, and P would be
ASD diagnosis. It would be desirable to compare the size of this relationship M to R
with the relationship of M to P. It would be ideal if one could simple compute correla‐
tions for M and R and for M and P. However, in most cases this would not work: the
scales are not all continual, and even if one were to employ a Spearman correlation, it
would not be apparent how to code something like HMO to insure a linear relationship.
What if HMO 2 was associated with an increase in thimerosal, and HMO 1 and 3 both
had low levels? This would result in a low correlation due to the curvilinear relation‐
ship, even IF much of the variance were in fact associated with HMO. On the other
hand, the relationship between HMO and thimerosal (M and R) can be checked via AN‐
OVA easily enough since R is continual and M is categorical. ANOVA would not work
for testing association between M and P because both M and P are both categorical in
this case. Chi – Square would be appropriate. However, regardless of the correct hypoth‐
esis test, all hypothesis tests are in fact unified by the p value.
The p value is a function of the size of the effect and the sample size. Different types of stat‐
istical tests have different probability distributions, but the total area under the curve has a
constant meaning across tests. The percent of area covered means the same thing in any test, regardless of the precise shape of the curve associated with a particular statistical test (correlation, ANOVA, Chi-square). A small p value could be due to a large effect, or it could be due to a very small effect and a very large sample. It should be stated that when sample
sizes are similar, it will not be unduly affected by sample size differences. Since the sample
will be the same for testing M to P or testing M to R, we propose the p values are the most
readily available means to index the comparison.
Compute a measure for the relationship of M and P and the associated p value. (e.g., HMO
and ASD: X2 (2) = 1.59, p =.45 )
Compute a measure for the relationship of M and R and the associated p value. (e.g., HMO
and Thimerosal exposure: F (2,1090) = 237, p <.0001).
The p value in all cases should be smaller for the M to R relationship, compared to the M to
P relationship test. This will serve to demonstrate that even if the Matching variable does
bear some relationship to the risk factor for the outcome probability, there is clearly a stron‐
ger relationship to the outcome itself, thus objectively justifying the matching. (e.g., the p
value of.45 indicates no relationship exists between the matched variable and the outcome
of interest, while the p value <.0001 indicates that matched variable is related to the exposure variable being tested. It is well known that matching on a variable that is associated only with exposure, not with disease, reduces statistical efficiency in a case – control design (Zondervan et al, 2002; Rubenfeld et al., 1999; Day, Byar, & Green, 1980), and this in essence, defines the problem of overmatching).
To sum, variables such as birth year, HMO, age, gender, address should first and foremost
be matched if and only if there is a truly justifiable rationale to expect they have an inde‐
pendent causal pathway to the outcome; “matching will increase he efficiency of the study
when the matching variable is a strong outcome determinant, but will actually reduce it
when the matching variable is strongly related to the exposure variable (over-matching),”
(Hanson & Khamis, 2008, p.595-596). Second, if the majority of the variance in the risk factor being tested is removed by matching, before the hypothesis is tested, extreme caution in reporting a lack of effect is warranted. Finally, recalling that sample size will be held constant, testing the relationships of M to R and P and comparing the p values can be used to justify matching in the context of the matching variable removing variance relating to the risk factor. We would propose that overmatching has and will continue to be a problem in matched case control designs, but suggest that employing the three checks above will serve to lessen deleterious effects associated with publishing overmatched results.
We welcome comments on these proposals.
Acknowledgement
This work was partially funded by a small grant awarded to the second author, Robert T.
Hitlan from Safeminds. We thank Safeminds for their support.
Author details
M. Catherine DeSoto and Robert T. Hitlan
University of Northern Iowa, Cedar Falls, USA
References
[1] Austin D. W., Shandley K. A., Palombo E. A. 2010. Mercury in vaccines from the
Australian childhood immunization program schedule. J. Toxicol. Environ Health A,
73: 637-40.
[2] Agudo A. and González C. A. 1999. Secondary matching: A method for selecting
controls in case-control studies on environmental risk factors. Int. J. Epidemiol., 1999:
1130-1133.
Vaccine Safety Study as an Interesting Case of “Over-Matching”
http://dx.doi.org/10.5772/53876
109
[3] Al-Taiara, A., Assabria, A., Al-Haboria, M., Azazya, A., Algabrib, A., Alganadib, M.,
Christopher J. M., Whittyc, and Jaffarc, S. 2009. Socioeconomic and environmental
factors important for acquiring non-severe malaria in children in Yemen: A case–con‐
trol study. Transactions of the Royal Society of Tropical Medicine and Hygiene, 1;
72-78.
[4] Bonner, P. C., Schmidt, W. P., Belmain, S. R., Oshin, B., Baglole, D., and Borchert, M.
2007. Poor housing quality increases risk of rodent infestation and lassa fever in refu‐
gee camps of Sierra Leone. Am. J. Trop. Med. Hyg., 77; 169–175.
[5] Day, N. E., Byar, D. P., and Green, S. B. 1980. Overadjustment in case control studies.
Am. J. Epidemiol., 112: 696–706.
[6] Delong, G. A. 2011. Positive association found between autism prevalence and child‐
hood vaccination uptake across the U.S. population. J. Toxicol. Environ. Health A, 74:
903-16.
[7] DeStefano F. 2007. Vaccines and autism: Evidence does not support a causal associa‐
tion. Clin. Pharmacol. Ther., 82: 756-759.
[8] Gallagher, C. M. and Goodman, M. S. 2010. Hepatitis B vaccination of male neonates
and autism diagnosis, NHIS 1997-2002. J. Toxicol. Environ. Health A, 73: 1665-1677.
[9] Geier D. A., and Geier M. R. 2006. An evaluation of the effects of thimerosal on neu‐
rodevelopmental disorders reported following DTP and Hib vaccines in comparison
to DTPH vaccine in the United States. J. Toxicol. Environ. Health A, 69: 1481-1495.
[10] Hansson, L. and Khamis, H. J. 2008. Matched samples logistic regression in case-con‐
trol studies with missing values: when to break the matches. Statistical Methods in
Medical Research, 17: 595–607
[11] Hewitson, L., Houser, L. A., Stott, C., Sackett, G., Tomko, J. L., Atwood, D., Blue, L.,
White, E. R. 2010. Delayed acquisition of neonatal reflexes in newborn primates re‐
ceiving a thimerosal-containing hepatitis B vaccine: Influence of gestational age and
birth weight. J. Toxicol. Environ. Health A, 73: 1298-1313
[12] Jackson, L., Kackson M. et al. (2006). Evidence of bias in estimates of influenza vaccine ef‐
fectiveness in seniors. Int. J. Epidemiol. 35 (2): 337-344.
[13] Jefferson, T. 2006. Influenza vaccination: policy versus evidence. British Medical Jour‐
nal, 333: 912-915.
[14] Marsh, J. L., Hutton, J. L., and Binks, K. 2002. Removal of radiation dose response ef‐
fects: an example of over-matching. British Medical Journal, 325: 327-330.
[15] Miettinen, O.S. 1969. The matched pairs design in the case of all-or-none responses.
Biometrics, 24: 339–52
[16] Monson, R. R. 1990. Occupational Epidemiology (2nd Ed.). BocaRaton, FL: CRC Press.
Recent Advances in Autism Spectrum Disorders – Volume I110
[17] Parker, S. K., Schwartz, B., Todd, J., and Pickering, L. K. 2004. Thimerosal-containing
vaccines and autistic spectrum disorder: A critical review of published original data.
Pediatrics, 114: 793-804.
[18] Price, C. S., Thompson, W. W., Goodson, B., Weintraub, E. S., Croen, L., A., Hinrich‐
sen, V. L., Marcy, M., Robertson, A., Eriksen, E., Lewis, E., Bernal, P., Shay, D., Davis,
R. L., and DeStefano, F. 2010. Prenatal and infant exposure to thimerosal from vac‐
cines and immunoglobulins and risk of autism. Pediatrics, 126: 655-664.
[19] Rahman. M, Sakamoto, M. D., and Fukui, T. 2003. Conditional versus unconditional
logistic regression in the medical literature J. of Clin. Epidem., 56: 101-102.
[20] Rubernfeld, G. D., Angus, D. C., Pinsky, M. R., Curtis, J. R., Coners, A. F., and Ber‐
nard, G. R. 1999. Outcomes research in critical care. Results of the american thoracic
society critical care assembly workshop on outcomes research. Am. J. Respir. Crit.
Care Med., 160: 358-367
[21] Sasieni, P., Castanon, A., and Cuzick, J. 2009. Effectiveness of cervical screening with
age: Population based case-control study of prospectively recorded data. British Med‐
ical Journal, 339: b2968.
[22] Vertraeten, T., Davis, R. L., DeStefano, F. 2003. Safety of thimerosal containing vac‐
cines: A two-phased study of computerized health maintenance databases. Pediatrics,
112: 1039-1048.
[23] Vidal, J. S., Vidailhet, M., Elbaz, A., Derkinderen, P., Tzourio, C., and Alpérovitch, A.
2008. Risk factors of multiple system atrophy: A case-control study in French pa‐
tients. Movement Disorders, 23: 797-803.
[24] West, D. W., Schuman, K. L., Lyon, J. L., Robison, L. M., and Allred, R. 1984. Differ‐
ences in risk estimation from a hospital and a population-based case-control study.
Int. J. Epidemiol., 13: 235-239.
[25] Krina, T., Zondervan1, L. R., Cardon, L. and Kennedy, S. H. 2002. What makes a
good case–control study? Human Reproduction, 17: 1415-1423.
Vaccine Safety Study as an Interesting Case of “Over-Matching”
http://dx.doi.org/10.5772/53876
111
az, A., Derkinderen, P., Tzourio, C., and Alpérovitch, A.
2008. Risk factors of multiple system atrophy: A case-control study in French pa‐
tients. Movement Disorders, 23: 797-803.
[24] West, D. W., Schuman, K. L., Lyon, J. L., Robison, L. M., and Allred, R. 1984. Differ‐
ences in risk estimation from a hospital and a population-based case-control study.
Int. J. Epidemiol., 13: 235-239.
[25] Krina, T., Zondervan1, L. R., Cardon, L. and Kennedy, S. H. 2002. What makes a
good case–control study? Human Reproduction, 17: 1415-1423.
Vaccine Safety Study as an Interesting Case of “Over-Matching”
http://dx.doi.org/10.5772/53876
111