Although the assertion of a link between vaccines and autism has been scientifically rejected, the theory continues to be popular and may influence the attitudes of parents of children with autism spectrum disorders. The authors sought to assess how often parents change or discontinue their child's vaccine schedule after autism spectrum disorder diagnosis and whether beliefs about the etiology of autism affect their decision to do so. The authors surveyed 197 (43%) of 460 eligible parents of children under 18 years of age with autism spectrum disorders who were enrolled in a state-funded agency that provides services to those with developmental disabilities in western Los Angeles County. Half of the parents discontinued or changed vaccination practices, and this was associated with a belief that vaccines contributed to autism spectrum disorders, indicating a potential subset of undervaccinated children. Educational tools should be designed to assist physicians when talking to parents of children with autism spectrum disorders about vaccination.
The prevalence of autism has risen dramatically over the past 10 to 15 years in the United States, and estimates indicate that about 1 in 110 children are diagnosed with an autism spectrum disorder (ASD; Centers for Disease Control and Prevention, 2009; Fombonne, 2001; Fombonne, 2003). Although advances in our understanding of the disorder have occurred, there are still gaps in our knowledge regarding the cause of ASD. Although the cause of autism continues to be studied, the largest body of evidence indicates that genetic factors are likely to account for most cases (Bailey et al., 1995; Folstein & Rosen-Sheidley, 2001; Muhle, Trentacoste & Rapin, 2004; Ritvo, Freeman, Mason-Brothers, Mo, & Ritvo, 1985). The specific genetic determinants, as well as their interaction with a variety of environmental factors, which are responsible for ASD are being actively researched (Buxbaum et al., 2001; Cook et al., 1997; Sebat et al., 2007). However, in a small, controversial, nonexperimental study published in The Lancet, Wakefield and colleagues (1998) proposed that autism was related to the measles-mumps-rubella (MMR) vaccine. This resulted in a media frenzy, widespread public concern, decreased MMR vaccination rates, and ultimately several measles outbreaks in the United Kingdom and elsewhere. The assertion of a link between the MMR vaccine and ASD remains unsubstantiated and there is a large and increasing body of evidence against such an association (DeStefano, Bhasin, Thompson, Yeargin-Allsopp, & Boyle, 2004; Fombonne & Chakrabarti, 2001; Madsen et al., 2002). In addition, the Wakefield study has since been discredited and officially retracted by the publishing journal (Dyer, 2010a; Fombonne & Chakrabarti, 2001; Taylor et al., 1999), with Wakefield being found guilty of professional misconduct and losing his license to practice medicine in the United Kingdom last year (Burns, 2010). However, the theory continues to receive much public and media attention. More recently, speculation that thimerosal or other preservatives are responsible for the rise in autism became popular, notwithstanding evidence that the prevalence of ASD has continued to rise despite the removal of thimerosal from almost all vaccines (Ball, Ball, & Pratt, 2001; Fombonne, 2008; Madsen et al., 2003). Despite recent court decisions that neither the MMR vaccine nor thimerosal in vaccines cause autism (Dyer, 2010b), the decision to vaccinate continues to be an emotionally charged debate for parents concerned with health risks.
The effect of these events on public perception of vaccine safety has been profound. Studies have shown that in general, parents are questioning vaccine safety and the potential link to ASD (Bardenheier et al., 2004; Freed, Clark, Butchart, Singer, & Davis, 2010; Gust et al., 2004). However, there is a dearth of literature on the beliefs and practices of parents of children with ASD. Parental beliefs have the potential to affect vaccination practices and vaccination rates adversely (Friederichs, Cameron, & Robertson, 2006; Gangarosa et al., 1998; Owens, 2002). Although U.S. immunization rates remain high, at approximately 70% to 78% (Wooten, Kolasa, Singleton, & Shefe, 2009), the public health implications of changing vaccine practices could be severe.
In particular, parents of children with ASD might be more likely to change vaccination practices than other families. In addition, parents of children with ASD may serve as sources of information for other families in their communities who may fear ASD. Given the high prevalence of ASD, understanding the concerns and vaccination practices of parents of children with ASD is necessary in order to then address those concerns and maintain immunization coverage. Thus, the objectives of this study were (a) to assess autism-related vaccine beliefs among parents of children with autism and (b) to examine the relationship between those vaccine beliefs and reported practices.
Study Design and Participants
We conducted a cross-sectional telephone survey of parents or legal guardians of children under the age of 18 who had an ASD and were enrolled in Westside Regional Center (WRC). WRC is a community-based, nonprofit agency providing services to individuals with developmental disabilities in part of Los Angeles County, regardless of income or insurance status, as part of a state-mandated entitlement program. The agency determines the diagnosis and assesses the eligibility for services for each child; these services include case management, consultations, and ongoing therapies. Using the agency's electronic database, we identified 460 parents or guardians of children with ASD. We manually reviewed each case to ensure that each parent had a child diagnosed with an ASD, including autism, pervasive developmental disorder–not otherwise specified (PDD-NOS), or Asperger syndrome. Of 460 parents, five did not speak English or Spanish as their primary language and were excluded from the study.
This telephone survey was part of a larger project designed to elicit parents' understanding of ASD diagnosis and treatments. The survey questions were intended to be brief and included demographics, socioeconomic information, parental beliefs regarding vaccinations and their children's ASD, and vaccination practices after diagnosis. We conducted pilot interviews with parents of children with developmental disabilities and other agency professionals to improve the clarity of the survey instrument and to ensure that the questions appropriately addressed key concepts. The study was also translated into Spanish by a professional who is experienced in matters of developmental disabilities.
We mailed a letter to eligible parents or guardians notifying them of the survey and of our intention to contact them. From March to September 2007, we conducted the telephone survey in English and Spanish. Up to six attempts were made to contact each potential participant by telephone during various hours of the day and on various days of the week until contact was established. Trained interviewers informed the respondents that the survey sought to obtain a better understanding of parental experiences with ASD. No incentive was offered for participation, and each respondent was assured that the survey was completely confidential, voluntary, and that their answers would not affect their agency services in any way. The average time of survey administration was 10 minutes. The Committee for the Protection of Human Subjects of the California Department of Health Services approved this study.
Respondents answered multiple-choice questions to indicate diagnostic, socioeconomic, and demographic information. Questions included the child's ASD type, perceived severity of child's ASD (on a scale of 1–10), child's insurance type, child's gender, child's age at diagnosis (in months), parental race or ethnicity, primary language spoken in the home, and parent education levels. In our analyses, child's age at diagnosis was dichotomized to less than 36 months or 36 months or older and severity of ASD was transformed to low (0–3), moderate (4–6), or high (7–10) severity.
We assessed whether parents believed that immunizations contributed to their child developing ASD with four possible responses: yes, possibly, no, and don't know. Responses from the survey question were collapsed into dichotomous variables (yes/possibly vs. no/I don't know) for our analyses. We then asked parents to specify which vaccines they believed played a role in their child's diagnosis. Parents could respond to any of a list of vaccines and preservatives and specify others in an “other” category.
Parental beliefs were explored further by a yes or no question, “Do you feel that there was any delay in the diagnosis of your child for his or her ASD?” As previous studies have shown that parents who perceived a greater delay in diagnosis tended to have less confidence in their physician (Harrington, Patrick, Edwards, & Brand, 2006), we hypothesized that parents who believed there was a delay in diagnosis may have less confidence in medical authorities and therefore may be more apt to believe that vaccines are responsible for autism.
We asked the parents questions related to parental practices and vaccines following their child's diagnosis of ASD. These questions included: (a) if parents had continued with their child's vaccine schedule after he or she was diagnosed with autism, (b) whether or not parents asked their child's doctor to change how vaccines were given in any way (such as splitting up vaccines), and (c) whether parents had subsequently vaccinated their other children. We constructed a composite variable to represent changing or discontinuing vaccine schedules, which was used as the dependent variable in our regression analyses. A positive response to this variable consisted of either a “yes” to changing the vaccine schedule or a “no” to continuing vaccines after the ASD diagnosis.
To further elaborate parental practices following their child's diagnosis with ASD, we asked whether parents changed their child's doctor because of ASD. We hypothesized that parents who changed or discontinued vaccinations after ASD diagnosis would also be more apt to change physicians due to disagreements regarding the belief that vaccines contribute to ASD.
Data from survey questions were analyzed using descriptive statistics and logistic regression modelling. We performed univariate analyses and calculated unadjusted odds ratios for our independent and control variables. We then estimated a multiple logistic regression model with the outcome variable of “changed or discontinued vaccines,” calculating adjusted odds ratios for all of our independent variables, including our main independent variable (vaccine beliefs) and other potential confounding variables (perceived severity of ASD, age of child at diagnosis, child's gender, child's insurance type, parent's race or ethnicity, parent's education level, and primary language). We included variables in the final multivariate model if p values were less than .10 in bivariate analyses. We did not include in our regression other parental beliefs (delayed diagnosis) or practices (vaccinating other children) due to concerns of co-linearity and reverse causality. The Hosmer-Lemeshow test was used to measure goodness-of-fit of the regression model. Level of significance was predetermined to be at p < .05. SPSS version 16.0 (Chicago, IL) was used for analyses.
One hundred ninety-eight parents of children with ASD participated for a response rate of 43%. Of those, 23 (of 48 eligible) participated in Spanish. One parent was removed from analyses because the single child was diagnosed with Rett Syndrome, which has a known genetic cause, giving a final sample of 197 participants. Most children were diagnosed with autistic disorder (84%); fewer had PDD-NOS (10%) or other ASDs; and most parents believed that their child's disorder was of low (30%) or moderate (40%) severity (Table 1). Two-thirds of the children had private insurance, and one-third had public insurance. Median age at diagnosis was 30 months. English was the primary language at home for 81% of respondents, Spanish was the primary language for 14% of parents, and other languages were spoken by 5% who were bilingual (English/other language). Forty-six percent of parents were Caucasian; 26% were Latino; and 13% were African American. Two-thirds of the parents were college graduates, and 21% finished high school or less. Eighty-one out of ninety-one (89%) Caucasian parents had completed college or had a graduate degree, compared to 13 out of 52 (25%) Hispanic/Latino parents. African American parents (65%) and those who reported another race or ethnicity (82%) had education levels similar to the Caucasian parents in our sample.
There was an approximately even split between those who believed that vaccines contributed to autism and those who did not or who did not know. Thirty-two percent (32%) of parents believed that vaccines did contribute to their children's ASD and 17% believed that vaccines possibly contributed, whereas 17% didn't know and 34% did not believe that vaccines led to the ASD. Over half of parents made different vaccine decisions following their child's diagnosis, with one-third (35%) of parents asking their doctor to change how vaccines were administered and one-fifth (21%) discontinuing vaccines altogether for their child (Table 2).
Latino (54%) and African American (58%) parents more often believed that vaccines led to their child's ASD compared to 37% of the Caucasian parents (Table 2). However, more Caucasian parents (43%) changed their child's vaccination schedule as compared to Latino (25%) or African American (15%) parents. Similarly, one-quarter (24%) of Caucasian parents, but fewer Latino (19%) and African American (12%) parents, discontinued their child's vaccines.
Parents who did not vaccinate their other children reported high rates of believing that vaccines caused ASD (67%) and discontinuing vaccines for their children with ASD (62%). In comparison, parents who vaccinated their other children were less likely to believe that vaccines caused ASD (49%), more likely to change their vaccine practices (37% vs. 29%), and less likely to discontinue vaccines (18%).
About two-thirds of parents who believed that their child's ASD diagnosis was delayed also believed that vaccines were responsible for the ASD. Over two-thirds (68%) of parents who changed their child's doctor believed that vaccines cause ASD. Almost half (47%) of parents who changed their child's doctor changed their child's vaccination practices and one-third (34%) discontinued their child's vaccination schedule altogether.
Table 3 summarizes parental beliefs about which specific vaccines contribute to ASD. Parents most often cited the MMR vaccine (49%), multiple vaccines (26%), additives or preservatives (13%), and many believed that vaccines led to ASD but they did not know which vaccine (23%). Twenty-three parents (24%) expressed other concerns about vaccines, such as causing “immune system overload” or “vaccines predisposed to autism.”
In multiple logistic regression analysis, parents who believed that vaccines contribute to ASD had twice the odds of changing or discontinuing vaccinations as compared to parents who did not believe vaccines contributed to ASD, aOR (adjusted OR): 2.01, 95% CI [1.04, 3.88] (Table 4). Having a child diagnosed with ASD at less than 36 months and having a female child were also associated with changing or discontinuing vaccine schedules, insurance type and parental education level were not associated with changing or discontinuing vaccination. African American parents were less likely to change or discontinue their children's vaccination schedule, aOR: 0.24, 95% CI [0.08, 0.74], than Caucasian parents. There was also a trend for Latino parents to change or discontinue vaccines less often. Adequate fit of this model was assessed using the Hosmer-Lemeshow test, χ2 (8) = 4.27, p = .83.
Despite substantial, credible evidence that vaccines are not responsible for ASD, almost half of parents of children with ASD in our study believed that vaccines contributed to their child's diagnosis. Furthermore, more than half of parents went on to change or discontinue vaccines. These results indicate that there remains a large gap between (a) parental belief and practices and (b) medical practice and recommendations for childhood immunizations.
Parents of children with autism have rarely been the focus of previous studies on vaccines, though significant differences between this population and parents in general may exist. Overall, U.S. parents surveyed in 1999 believed that vaccines were generally safe and important (Gellin, Maibach, & Marcuse, 2000). However, in the wake of the prolonged, negative publicity regarding vaccine safety, it is important to continue to assess parental concerns and beliefs. In fact, a 2009 survey found that although 90% of U.S. parents shared the belief in the importance of immunization, these same parents expressed concerns regarding the safety and adverse effects of vaccines. For example, 25% of parents' surveyed believed that ASD could be caused by vaccines and 11.5% had refused a vaccine (Freed et al., 2010). Similarly, we found that many parents of children with ASD were concerned about vaccine safety. However, families we surveyed much more frequently believed that vaccines were responsible for ASD and over twice as often refused vaccinations as compared to the general population of parents.
The large proportion of parents who believe vaccines caused their child's ASD indicates a potential communication barrier between physicians and parents, as well as the powerful effect of media on vaccines practices. Physicians may believe that parents of children with ASD may not be open to discussion about vaccines or that education will not change beliefs. Parents may be fearful and perceive that physicians do not value their beliefs (Young, Davis, Shoen, & Parker, 1998). One recent study found that parents who did not vaccinate felt estranged from and unable to trust their pediatricians (Benin, Wisler-Scher, Colson, Shapiro, & Holmboe, 2006) and another study indicated that those who refused the MMR vaccine felt that the good intentions for their child were not recognized (Casiday, Cresswell, Wilson, & Panter-Brick, 2006). Parents of children with ASD frequently feel tremendous guilt that by vaccinating they might have caused their child's disorders (Hilton, Hunt, & Petticrew, 2007). They may be more likely to withhold vaccinations because of the concern that they could worsen their child's ASD or cause ASD in their other children. Parents perceive that physicians underestimate their safety concerns and the impact of ASD on their family (Hilton et al., 2007). Nonetheless, studies have found that in general, parents most commonly look to their pediatric providers for information about vaccines (Gellin et al., 2000; Gust et al., 2004). Though they frequently turn to outside resources such as media sources, most parents of children with ASD will likely continue to seek information from their physicians. Thus, if physicians promote open discussion and make it clear that there is a shared primary concern for the welfare of the child, then an alliance and opportunity for respectful education may be available.
One-fifth of parents in our study were so concerned about the association between vaccines and autism that they discontinued vaccinations altogether. Furthermore, believing that vaccines caused ASD resulted in a twofold increase in changing or discontinuing vaccinations. We also found that many parents who did not vaccinate their child diagnosed with an ASD also did not vaccinate their other children. In the only other similarly published study, two-thirds of parents who reported an ASD to the vaccine adverse event reporting system (VAERS) subsequently withheld vaccines from their children (Woo et al., 2004). These results indicate that families of children with ASD, which constitute a large and growing group of children in pediatric practices, may be at risk for vaccine-preventable diseases due to lack of vaccination. Given that children with ASD may congregate in special education classes or other groupings, the risk of vaccine-preventable diseases in those settings in particular could rise. Further research should investigate the rates of vaccinations and childhood vaccine-preventable illness in families of children with ASD and in these settings in particular.
An interesting finding of this study was the relationship between race and ethnicity and vaccine beliefs and practices. Latino and African American parents more often believed that vaccines contributed to ASD but were less likely to act on that belief by changing or discontinuing vaccinations. For African Americans, this result persisted regardless of education level or other demographic characteristics, which may reflect a longstanding distrust of medical and research communities within minority groups (Corbie-Smith, Thomas, & St. George, 2002). Previous studies have reported that vaccine refusal rates are lower among African American and Latino parents in both the general population (Friederichs et al., 2006) and among parents of children with special health care needs (Wood et al., 1995). This division between beliefs and practices may be due to daycare or school requirements (for example, these families may not be aware of the availability of vaccine exemption forms for school immunization requirements) or differing cultural norms related to questioning a physician's recommendations (Shui, Weintraub, & Gust, 2006; Schouten & Meeuwesen, 2006).
Among respondents, the most commonly implicated vaccine was MMR. However, many respondents expressed concern with the use of multiple vaccines, especially when given together. This result is similar to parents reporting ASD to VAERS as an adverse vaccine reaction (Woo et al., 2004). Though some parents were concerned about preservatives such as thimerosal, the majority were not, perhaps indicating that concerns about preservatives are decreasing. Many parents were concerned about vaccines, but unable to state which one in particular had contributed to ASD, suggesting considerable levels of incomplete information and confusion.
The parents in our study were generally well educated, with the majority having a college or graduate degree. We also found an association between higher education levels and belief that vaccines led to ASD. This result indicates that some educational materials and discussions should be tailored for a highly educated group of parents. We suggest that the discourse should acknowledge parents' educational level, accurately appraise the successes and shortcomings of research on causes of autism, and explain why vaccines are not related to ASD. In addition, because many parents of children with ASD believed that their child's diagnosis was delayed, we recommend that, in general, the conversation about vaccinations should occur in conjunction with discussion about ASD evaluation or management plans. This discussion requires significant time, education, and practice for physicians. Parents of children with ASD who did not believe that their child's ASD was caused by vaccines, who comprised half of the parents in our study, may inform physicians of the timing, content, and style of discussions that would best reach other parents of children with ASD. Further efforts should be made by organizations, such as the new AAP Immunization Alliance, to collaborate with parents to develop and disseminate a program to better support these discussions. Translating science into physician practice and public knowledge can take more than a decade (Contopoulos-Ioannidis, Ntzani, & Ioannidis, 2003). In conjunction with physician education programs, media attention given to ASD and vaccines may actually encourage physicians to increase their knowledge and speed the process of scientific translation to public knowledge.
This study is subject to several limitations. First, we had a response rate of 43%, which is not unusual for survey data, but lower than we had anticipated. Second, the sample was taken from one geographic location and the sample was highly educated, which may not reflect parents of children with ASD in other areas of the United States and elsewhere. Still, this study represents a relatively large sample of children with ASD and other surveys of parents of children with ASD had similar or smaller sample sizes (Rhoades, Scarpa, & Salley, 2007). Third, parents' self-reported practices of changing or discontinuing vaccinations may not reflect actual vaccination records due to questionable recall. This study did not seek to document actual immunizations rates, but rather beliefs and practices from a parental perspective. Further research should document actual vaccination rates for families of children with ASD as compared to the general population. Fourth, because this project involved a brief telephone interview, questions that would have further elucidated parents' attitudes about vaccinations were not asked (for example, why did a parent change from a family physician to a medical specialist?). More in-depth research may be able to ascertain additional information regarding parents' motivations. Finally, the survey was administered before the Wakefield article was officially retracted in January 2010. Even though the findings had already been widely discredited by the medical community, recent data could indicate changes in parental beliefs about vaccines due to the official retraction. Nonetheless, this study enhances our understanding of parents' views and practices and provides targets for intervention.
A large portion of parents of children with ASD believe that vaccines contribute to ASD and half either change or discontinue vaccinations, indicating the potential for a subset of under-vaccinated and at-risk children. Given that any link between vaccines and autism has been refuted, this indicates a lengthy delay in the translation of science to practice and public knowledge. Therefore, discussion about vaccinations should be emphasized and specifically tailored to meet the needs of parents of children with ASD. Educational tools should also be designed to assist pediatricians when talking to parents of children with ASD. Further research should focus on assessing national vaccination coverage rates in children with ASD and how best to communicate with parents of children with ASD about vaccinations.
Authors: Alicia Bazzano (e-mail: email@example.com), UCLA Fielding School of Public Health, Department of Health Services, 650 Charles Young Dr., S., 31-269 CHS Box 951772, Los Angeles, CA, 90095-1772, USA; Ari Zeldin, Navy Medical Center, San Diego; Erica Schuster, Westside Regional Center, Culver City, CA; Christopher Barrett, Westside Regional Center, Culver City, CA; Danise Lehrer, Westside Regional Center, Culver City, CA.
Editor-in-Charge: Gael Orsmond