Autism spectrum disorder and chronic disease: No evidence for vaccines or thimerosal as a contributing factor

INTRODUCTION — The number of cases of autism spectrum disorder (ASD) in the United States and other countries has increased since the 1980s, largely due to changes in diagnostic criteria for ASD and increased awareness of ASD. The increased prevalence of ASD occurred at a time when the number of recommended childhood vaccines also increased, leading to a hypothesis that vaccines (eg, the measles, mumps, and rubella vaccine) or vaccine constituents (eg, thimerosal) contribute to the development of ASD and other chronic diseases. Numerous subsequent studies have demonstrated no scientific linkage between vaccines and ASD.

The proposed association between vaccines or vaccine constituents and ASD and the evidence against the association will be discussed here. The epidemiology, pathogenesis, and clinical features of ASD and communication with vaccine-hesitant caregivers are discussed separately. (See "Autism spectrum disorder (ASD) in children and adolescents: Terminology, epidemiology, and pathogenesis" and "Autism spectrum disorder in children and adolescents: Clinical features" and "Standard childhood vaccines: Caregiver hesitancy or refusal", section on 'Approach to management'.)

APPARENT INCREASED PREVALENCE OF AUTISM SPECTRUM DISORDER — The number of cases of ASD in the United States and other countries has increased since the 1980s [1-9]. Rates of ASD in studies from the late 1990s are consistently greater than 10 per 10,000 compared with 4 to 5 per 10,000 in previous decades [10]. Active surveillance from 2020 in the United States suggests a prevalence of approximately 1 in 36 eight-year-old children. This is an increase from the 2018 estimate of 1 in 44 eight-year-old children [11,12]. (See "Autism spectrum disorder (ASD) in children and adolescents: Terminology, epidemiology, and pathogenesis", section on 'Prevalence'.)

Whether the actual incidence of ASD has increased is unclear. Comparing studies with different case definitions, methods of case finding, and sample populations is problematic unless there is rigorous control for these variables. Systematic reviews of the epidemiologic studies suggest that changes in case definitions and increased awareness account for much of the increased prevalence [10,13,14].

PROPOSED ASSOCIATION BETWEEN VACCINES AND ASD — The real or perceived increase in ASD cases occurred at a time when the number of recommended childhood vaccines also increased (to include Haemophilus influenza type b, hepatitis B, varicella, pneumococcal, and influenza vaccines, as well as a second dose of the measles, mumps, and rubella [MMR] vaccine). (See "Standard immunizations for children and adolescents: Overview", section on 'Routine schedule'.)

In the search for a causal relationship, caregivers of children with ASD and some professionals identified a temporal association between immunizations and the onset of more evident symptoms of ASD in the second year of life [7]. The temporal association led to speculation that certain vaccines (eg, MMR) or vaccine constituents (eg, thimerosal) may play a role in the development of ASD [15-21]. However, prospective studies indicate that findings of ASD often are present during the first year of life, before the first dose of MMR. (See "Autism spectrum disorder in children and adolescents: Surveillance and screening in primary care", section on 'Early indicators'.)

MMR VACCINE AND ASD

Proposed association — The alleged association between the measles, mumps, and rubella (MMR) vaccine, enterocolitis, and ASD was first reported in a 1998 study of 12 children that suggested a link between recent injection of MMR vaccine and the onset of symptoms of ASD and enterocolitis [15]. In 2004, 10 of the 13 authors of the study published a statement retracting its interpretation, and the Lancet fully retracted the paper in 2010 [22,23]. In addition, by comparing the case descriptions in the paper with medical records, an investigative reporter found that the study was fraudulent [24-28]. Three of the children did not have ASD; five had developmental concerns before MMR vaccination; behavioral symptoms developed in some children months (rather than days) after MMR vaccination; and colonoscopy results were altered from unremarkable findings to "nonspecific colitis" after "research review." In addition, patients were recruited through an anti-MMR vaccine organization, and the study was commissioned and funded for planned litigation.

Another report compared the presence of persistent measles virus in the intestinal tissue of 91 children with developmental disorders, including ASD, and 70 controls [29]. Persistent measles virus particles were more prevalent among the children with developmental disorders (82 versus 7 percent). Although the authors of the study concluded that the findings confirm an association between persistent measles virus and gut pathology in children with developmental disorders, the study and its conclusions have been criticized for methodologic flaws [30-34]. Methodologic flaws included not determining whether the viral genome that was detected was vaccine virus or natural measles virus (which was still circulating), not providing information about whether and when cases and controls had received the MMR vaccine, not describing procedures to prevent false-positive results from natural measles virus contamination in the laboratory, and not describing blinding of laboratory personnel.

Lack of evidence for association — No causal association between MMR vaccine and ASD is established [35-37].

Biologic mechanisms — One of the criteria for establishing causality is that there is a coherent explanation that accounts for the findings (ie, a plausible biologic mechanism) [38]. Proposed biologic mechanisms for the purported association between MMR vaccine, bowel disease, and ASD include decreased viral immunity or autoimmunity related to MMR vaccine, persistent gastrointestinal or central nervous system measles virus, and opioid excess [15,16,29,30,35,39]. These proposed mechanisms lack supportive evidence [35].

Patients with ASD do not have characteristic markers for immune injury or inflammation [30,33,40]. Although there are reports of measles virus detected with polymerase-chain reaction from intestinal or blood samples of children with ASD [29,39,41], subsequent studies using highly sensitive and specific assays and enhanced laboratory techniques failed to detect measles virus nucleic acids in the white blood cells of children with ASD who had received MMR vaccine, indicating that the findings in the earlier studies may have been false positives [42-44]. A case-control study found no differences in the excretion of opioid peptides in the urine of children with ASD and controls [45]. Cerebrospinal beta-endorphins in patients with autism are not consistently elevated [46-48], nor do social and stereotypic behaviors in children with ASD improve with administration of opioid antagonists [49-51].

Persistent measles infection or abnormally persistent immune response to MMR vaccine is another mechanism that has been proposed to explain an association between MMR vaccine and ASD. In a case-control study in which measles virus and measles antibody were measured in 98 children (age 10 to 12 years) with ASD, 52 children with special needs without ASD, and 90 typically developing children, measles virus nucleic acid was detected in peripheral blood mononuclear cells of one child with ASD and two typically developing children. Antibody response did not differ between cases and controls, and there was no correlation between antibody levels and ASD symptoms [52].

Epidemiologic studies — Multiple large, well-designed epidemiologic studies [2,3,5,32,53-67] and systematic reviews [68-70] do not support an association between the MMR vaccine and ASD. Most of these studies compared the risk of developing ASD among children who received the MMR vaccine with the risk of developing ASD among children who did not receive the MMR vaccine. Similar methods have been used to detect associations between the swine flu vaccine and Guillain-Barré syndrome [71], the Rotashield vaccine and intussusception [72], and the MMR vaccine and immune thrombocytopenia [73]. No such association has been detected between MMR vaccine and ASD.

A 2020 systematic review of two cohort studies (including nearly two million children) [3,67] and four case control studies (including nearly 9000 children) [62-64,66] found no evidence to support a relationship between MMR vaccination and ASD [74], supporting the findings of earlier systematic reviews of observational studies with their inherent limitations (eg, risk of bias related to lack of randomization and blinding). Epidemiologic studies also have failed to find an association between the age at the time of vaccination, the time since vaccination, or the date of the vaccination and the development of ASD, or between MMR vaccination and development of ASD in siblings of children with ASD or in children with other risk factors for ASD (eg, parental age, low birthweight, mode of delivery) [3,67]. In some studies, the prevalence of ASD increased despite declines in MMR vaccination rates [58,62,75].

THIMEROSAL AND ASD

Background — Thimerosal (also known as sodium ethylmercury and thiomersal) is an organic compound that is used as a preservative.

Thimerosal in vaccines — In the United States, inactivated influenza vaccine is the only routinely recommended childhood vaccine that may contain thimerosal (many formulations are thimerosal-free (table 1)). Measles-mumps-rubella (MMR), varicella, inactivated polio, and pneumococcal conjugate vaccines never contained thimerosal [76]. Before 2001, thimerosal was used as a preservative in some vaccines (eg, hepatitis B, diphtheria-tetanus-acellular-pertussis, H. influenzae type b) and Rh immune globulin [77].

In 1999, the American Academy of Pediatrics and the United States Public Health Service recommended that thimerosal be reduced or eliminated from routine childhood immunizations [78,79]. This recommendation was taken as a precaution because of the possibility that some infants who received thimerosal-containing vaccines at multiple visits could receive cumulative doses of ethylmercury that exceeded United States Environmental Protection Agency guidelines [80-82].

Since then, multiple studies have demonstrated the differences between ethylmercury and methylmercury and the absence of evidence of harm from thimerosal-containing vaccines [82,83]. The World Health Organization Global Advisory Committee on Vaccine Safety has concluded that the available evidence supports the safety of thimerosal as a preservative in inactivated vaccines; safer and equally efficacious alternatives have not been identified [84,85]. (See 'Lack of evidence for association' below.)

Mercury toxicity — The clinical features and severity of mercury toxicity vary with the form of mercury (eg, methylmercury or ethylmercury), the route of entry (eg, ingested versus injected), the dose, and the age at exposure [86,87]. Thimerosal contains 50 percent ethylmercury by weight and has a toxicologic profile that is thought to be similar to ethylmercury from other sources [88]. Ethylmercury (CH₃CH₂Hg⁺) and methylmercury (CH₃Hg⁺) have similar chemical structures and similar initial distribution in the body; they have biologically distinct behaviors [36,86,88-93].

●Effects of high doses – At toxic doses, ethylmercury and methylmercury cause similar damage to the brain. High-dose exposure to ethylmercury (≥3 mg/kg) can cause severe toxicity, including local necrosis, acute hemolysis, disseminated intravascular coagulation, acute tubular necrosis, and central nervous system injury [94-98]. (See "Mercury toxicity", section on 'Organic mercury toxicity'.)

●Effects of low-dose exposure – There is little information regarding the clinical effects of low-dose exposure to mercury. The only studies describing low-dose exposure to organic mercury are those involving prenatal exposure to methylmercury from fish consumption in the Seychelles and Faroe Islands [99-103]. The results of these studies are inconsistent. Studies from the Faroe Islands report subtle and long-term cognitive deficits at methylmercury levels previously thought to be safe [101-103]. However, children from these islands may have been exposed to additional toxins (eg, polychlorinated biphenyls) [104,105]. Studies of children from the Seychelles, evaluating more global outcomes, failed to find significant cognitive or behavioral effects in offspring of women with high fish consumption when other factors, such as social and environmental developmental modifiers and postnatal mercury exposure, were considered [99,100].

Extrapolating the neurotoxic effects of low-dose methylmercury exposure to low-dose ethylmercury exposure is very problematic because the two compounds have different half-lives (50 versus 7 days, respectively) and biologically distinct behaviors [88-93]. Although both are neurotoxic, methylmercury appears to be more potent with greater access to the central nervous system. Methylmercury is actively transported across the blood-brain barrier, whereas the transport of ethylmercury into the central nervous system is hindered by its relatively larger size and rapid decomposition [106,107].

●Pharmacokinetics – Little is known about the pharmacokinetics of ethylmercury injected into human infants. In observational studies, younger infants with lower body weights had higher blood mercury concentrations after immunization with thimerosal-containing vaccines than older, heavier infants [108-110]. However, the concentrations did not exceed those thought to be safe. Furthermore, ethylmercury is rapidly eliminated from the blood of infants who receive thimerosal-containing vaccines [109,110].

As an example, one study measured blood, urine, and stool mercury concentrations before and 12 hours to 30 days after vaccination of 216 infants with thimerosal-containing vaccines [110]. The study population included 72 newborns, 72 two-month-olds, and 72 six-month-olds. The cumulative dose of mercury in the six-month-olds ranged from 112.5 to 162.5 mcg. Blood mercury levels peaked on the first day after vaccination and returned to prevaccination levels within a few weeks. The maximum mean blood mercury levels were 5 mcg/L (25 nmol/L) in newborns, 3.6 mcg/L (18 nmol/L) in two-month-olds, and 2.8 mcg/L (14 nmol/L) in six-month-olds. For all infants, blood mercury levels were within the normal reference range (0 to 9 mcg/L [0 to 45 nmol/L] [111]). Mercury was virtually undetectable in urine in all samples.

The US Food and Drug Administration estimates that the body burden of mercury following annual exposure to thimerosal in inactivated influenza vaccine over the first 4.5 years of life remains below the safety threshold for dietary methylmercury established by the United States Environmental Protection Agency (ie, <0.1 mcg/kg per day) [112,113]. Estimated peak body burdens of mercury did not exceed the corresponding safe body burden for methylmercury, even for low-birthweight infants.

Proposed association — Some authors hypothesize that ASD is an expression of mercury poisoning because of their perception that mercury toxicity and ASD have similar clinical manifestations, their observation that the apparent increase in the number of cases of ASD parallels the increased exposure to thimerosal in vaccines, and their perception of a temporal association between the onset of ASD and immunization [17,18]. These observations are discussed in detail and refuted below. (See 'Lack of evidence for association' below.)

Additional concern regarding an association between thimerosal and ASD stems from anecdotal and unpublished reports of improvement in children with ASD and abnormal blood metal levels following chelation therapy [7,114-116]. However, studies evaluating mercury levels in hair, urine, or blood of children with ASD compared with controls have inconsistent results [117-119]. Evidence that chelation therapy improves ASD is lacking [107]. (See "Autism spectrum disorder in children and adolescents: Complementary and alternative therapies", section on 'Chelation'.)

Other studies that propose an association between thimerosal and ASD used data from the Vaccine Adverse Event Reporting System (VAERS) to suggest that the incidence of ASD was greater in children who had received thimerosal-containing vaccines than in those who did not [19,20,120]. However, the limitations of the VAERS reporting system preclude accurate assessment of both the number of cases and the exposure to thimerosal-containing vaccines [87,121]. VAERS is a passive surveillance system that relies on clinicians to voluntarily report possible vaccine adverse events; VAERS reports do not include the information necessary to calculate thimerosal exposure, and therefore cannot attribute causality.

Lack of evidence for association — No causal association between thimerosal and ASD is established [35-37,114,122].

Comparison of clinical features — The characteristic features of mercury poisoning have little in common with those of ASD (table 2).

Common characteristic motor findings in high-dose mercury poisoning include ataxia, dysarthria, tremor, muscle pain, and weakness [98,123,124]. In contrast, the characteristic motor finding in children with ASD is repetitive behavior, such as flapping, circling, or rocking [107]. Hypotonia and clumsiness have been noted in some children with ASD. However, other motor manifestations are uncommon. The presence of ataxia or dysarthria in a child whose behavior has autistic features should prompt a careful medical evaluation for an alternative or additional diagnosis [107].

Individuals with mercury poisoning typically have dysarthric speech. In contrast, individuals with ASD typically have delayed speech or echolalia.

Sensory findings in mercury poisoning include bilateral visual field constriction, which is highly specific [123-125], and paresthesias. Infants with mercury poisoning may have erythema and pain in the hands and feet from peripheral neuropathy [107]. Sensory findings in ASD include decreased responsiveness to pain and hypersensitivity to other sensory stimuli, including sounds [107].

Other manifestations of mercury poisoning include hypertension [126], skin eruption [127,128], and thrombocytopenia [129]. Patients with mercury poisoning also may have toxic psychosis or, in milder cases, nonspecific depression, anxiety, or irritability [130-132]. Hypertension, rash, and thrombocytopenia are not features of ASD [107].

Children with prenatal or early childhood exposure to mercury typically have decreased head circumference [133]. In contrast, patients with ASD often have macrocephaly [134-137].

Additional clinical features of mercury poisoning and ASD are discussed separately. (See "Mercury toxicity", section on 'Clinical manifestations' and "Autism spectrum disorder in children and adolescents: Clinical features".)

Comparison of neuropathology — Pathologic findings in the brains of individuals who died as a result of mercury poisoning include severe atrophy and gliosis of the calcarine cortex; diffuse neuronal loss and gliosis of the auditory, motor, and sensory cortices; extensive cerebellar atrophy; demyelination of the 9^th and 10^th cranial nerve roots; and atrophy of the cerebellar granule cell layer with relative sparing of Purkinje cells [93,138].

By comparison, the brains of autistic children examined at autopsy or with magnetic resonance imaging are typically enlarged in weight and volume compared with those of controls [139,140]. Additional findings include unusually small, closely packed neurons and increased cell-packing density in portions of the limbic system, consistent with curtailment of development of this circuitry; reduction in the number of Purkinje cells in the cerebellum, primarily in the posterior inferior hemispheres; rare involvement of granule cells; and focal disruption of cortical laminar architecture in the cortexes of a majority of young children with ASD, with data supporting a probable dysregulation of layer formation and layer-specific neuronal differentiation at prenatal developmental states [139,141-143]. (See "Autism spectrum disorder (ASD) in children and adolescents: Terminology, epidemiology, and pathogenesis", section on 'Neurobiologic factors'.)

Biologic mechanisms — One of the criteria for establishing causality is that there is a coherent explanation that accounts for the findings (ie, a plausible biologic mechanism) [38]. Although biologic mechanisms for an association between thimerosal and ASD have been proposed, the information upon which the hypotheses are based is indirect and incomplete [114]. Low-dose thimerosal exposure in humans has not been demonstrated to be associated with effects on the nervous system [114]. Neurodevelopmental effects have been demonstrated in some populations for prenatal exposure to low doses of methylmercury but not for postnatal exposure [101]. There is no evidence that ethylmercury causes any of the pathophysiologic changes that are known to be associated with ASD. Thimerosal exposure from vaccines has not been proven to result in mercury levels associated with toxic responses. (See 'Mercury toxicity' above.)

Epidemiologic studies — Multiple large, well-designed epidemiologic studies [4,144-148] and systematic reviews [35,84,85,107,114,121] do not support an association between thimerosal-containing vaccines and ASD [4,58,144-146,149] or other developmental disorders, with the possible exception of tics [145,147,148,150]. Most of these studies compare the risk of developing ASD among children who received thimerosal-containing vaccines with the risk of developing ASD among children who did not receive thimerosal-containing vaccines. Similar methods have been used to detect associations between the swine flu vaccine and Guillain-Barré syndrome [71], the Rotashield vaccine and intussusception [72], and the MMR vaccine and immune thrombocytopenia [73]. No such association has been detected between thimerosal-containing vaccines and ASD.

A 2014 systematic review of three cohort studies (including 718,200 children) [145,146,148] and one case-control study (including 1008 children) [149] found no relationship between vaccine-related mercury or thimerosal exposure and increased risk of ASD [151]. In the United States and other countries, the incidence of ASD continued to increase after the use of thimerosal-containing vaccines was discontinued [4,58,152], an observation inconsistent with the hypothesis that increased exposure to thimerosal accounts for the apparent increase in the rates of ASD. The World Health Organization Global Advisory Committee on Vaccine Safety has concluded that the available evidence strongly supports the safety of thimerosal as a preservative for inactivated vaccines and that no additional studies are necessary [84].

NUMBER OF VACCINE ANTIGENS AND ASD — The lack of evidence for an association between ASD and the number of vaccine antigens administered to children in the first two years of life is discussed separately. (See "Standard childhood vaccines: Caregiver hesitancy or refusal", section on 'Target education'.)

VACCINES AND OTHER CHRONIC DISEASES

Multiple sclerosis — There has been concern that onset or relapse of multiple sclerosis may be precipitated by hepatitis B or other vaccines. This is discussed separately. (See "Pathogenesis and epidemiology of multiple sclerosis", section on 'Vaccinations'.)

Type 1 diabetes mellitus — Although there have been reports describing a temporal association between receipt of vaccines or the timing of the first dose of vaccine and development of type 1 diabetes mellitus [153,154], no causal association is established [155].

The report linking type 1 diabetes mellitus compared the rates of diabetes and vaccination schedules among various countries [153]. Such ecologic studies may provide the basis for a hypothesis that a vaccine is associated with a particular disease but do not provide evidence for the association. Many factors may affect the rates of diabetes in specific countries (eg, genetic predisposition, environmental exposures, breastfeeding). (See "Pathogenesis of type 1 diabetes mellitus".)

To provide evidence of an association between a particular vaccine and diabetes mellitus, it is necessary to compare the relative risk of developing type 1 diabetes among children who did and did not receive the particular vaccine. A meta-analysis of 23 such studies found no association between routine childhood vaccination and development of type 1 diabetes [155]. Similarly, no association was found in a meta-analysis of observational studies of the safety of vaccines used for routine immunization in the United States or in data from Vaccine Safety Datalink [156,157].

Vaccine-related stimulation of beta-cell autoimmunity has been suggested as a mechanism for the proposed association between vaccines and development of diabetes. Observational studies do not support this hypothesis [158,159].

PROVEN BENEFITS OF VACCINES — Evidence for an association between vaccines and ASD or chronic disease is lacking [160]. On the other hand, the benefits of vaccines are clear (figure 1). Several infectious diseases that were once associated with significant morbidity and mortality have been almost completely eliminated through the development, distribution, and almost universal administration of protective vaccines (figure 2 and figure 3). (See "Standard childhood vaccines: Caregiver hesitancy or refusal", section on 'Target education'.)

CONSEQUENCES OF VACCINE REFUSAL — With the declining incidence of once-common vaccine-preventable diseases, caregivers of young children may not appreciate the potential severity or dire consequences of the illnesses. Caregivers who lack such appreciation may be willing to forego immunizations for their children, particularly if unproven risks are highly publicized [161]. When this occurs, immunization rates decline, and outbreaks of infectious diseases, such as measles and pertussis, may occur with significant morbidity and mortality.

Consequences of vaccine refusal are discussed in greater detail separately. (See "Standard childhood vaccines: Caregiver hesitancy or refusal", section on 'Consequences of vaccine refusal'.)

RESOURCES — The following websites provide additional information about vaccines and ASD, diabetes, or multiple sclerosis. They include sections on frequently asked questions that may be helpful when discussing these issues with caregivers.

●The American Academy of Pediatrics provides a representative list of studies that found no association between vaccines or thimerosal and ASD but cautions that the list is not exhaustive; links to the original publications are included.

●The Immunization Action Coalition.

●The United States Centers for Disease Control and Prevention.

●The United Kingdom's National Health Service.

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or email these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient education" and the keyword[s] of interest.)

●Beyond the Basics topics (see "Patient education: Why does my child need vaccines? (Beyond the Basics)" and "Patient education: Autism spectrum disorder (Beyond the Basics)")

SUMMARY

●Apparent increased in ASD – The prevalence of autism spectrum disorder (ASD) appears to have increased since the 1980s. Much of this trend is accounted for by changes in the case definition and increased awareness of ASD. Whether the actual incidence of autism has increased is unclear. (See 'Apparent increased prevalence of autism spectrum disorder' above.)

●Proposed association between vaccines and ASD – The real or perceived increase in ASD cases occurred at a time when the number of recommended childhood vaccines also increased. Caregivers of children with ASD and some professionals identified a temporal association between immunizations and the onset of more evident symptoms of ASD in the second year of life, leading to speculation that certain vaccines or vaccine constituents may play a role in the development of ASD. (See 'Proposed association between vaccines and ASD' above.)

●Lack of evidence of an association – Multiple large, well-designed epidemiologic studies and systematic reviews do not support an association between the measles, mumps, and rubella vaccine and ASD. (See 'Lack of evidence for association' above.)

Similarly, multiple large, well-designed epidemiologic studies and systematic reviews do not support an association between thimerosal and ASD. (See 'Lack of evidence for association' above.)

●Proven benefits of vaccines and consequences of vaccine refusal – The administration of childhood vaccines has led to a decline in the incidence of childhood diseases that can have severe sequelae (figure 1). (See 'Proven benefits of vaccines' above.)

Withholding vaccines from a child because of a hypothetical risk places the child at risk for real infection that may have real sequelae. (See "Standard childhood vaccines: Caregiver hesitancy or refusal", section on 'Consequences of vaccine refusal'.)