Advances in gene–environment interaction research have revealed genes that are associated with aggression. However, little is known about parent perceptions of genetic screening for behavioral symptoms like aggression as opposed to diagnosing disabilities. These perceptions may influence future research endeavors involving genetic linkage studies to behavior, including proactive approaches for parents to avoid events leading to aggression. The purpose of this study was to solicit the perspectives of parents who have children with autism about screening for genes associated with aggression, compared to responses from those who have children without disabilities and those planning to have children. Parents of children with autism were more likely to support screening and the use of the results to seek treatment if necessary. Results are discussed in the context of surveillance screening and systematic early intervention for behavioral symptoms related to autism. The results may provide insight for clincians, researchers, policymakers, and advocacy groups related to diagnosing and treating aggression in people with autism.
The core diagnostic features of autism consist of delayed communication, limited social behavior, and stereotypic behaviors (American Psychiatric Association, 2000; APA). Although the direct causal mechanisms have not yet been identified, researchers suggest that chemical or anatomical perturbations in the brain occur at a very early age (Hill & Furniss, 2006; Holden & Gitlesen, 2006). Furthermore, severity of the core symptoms of autism is considered a risk factor for developing psychopathologies and problem behaviors such as physical aggression (e.g., Matson & Rivet, 2008; McClintock, Hall, & Oliver, 2003). Aggression toward oneself and/or others occurs in up to 68% of children with autism aged 6 to 14 and is associated with a variety of environmental influences, including neglect, poor nutrition, caregivers who act aggressively or display low emotional attachment, communication problems, and/or operant behavioral processes (Brown & Radford, 2007; Farmer & Aman, 2010; Tremblay, 2000). In many cases, problem behavior results in more stress for parents than the core features themselves (Hastings, 2003; Lecavalier, 2006).
Recent research has focused on gene–environment interactions to explain the contribution of each of these variables to the emergence of aggression in a variety of populations. Generally speaking, 42% of the variance can be explained by environmental influences, whereas 56% can be explained through genetic predisposition (Moffitt, 2005; Rhee & Waldman, 2002). Two genes in particular have been investigated for their association with aggression. The gene encoding monoamine oxidase A (MAO-A) has been correlated with aggression in people with autism and intellectual disability (e.g., Cohen, 2003; May et al., 2009; Taylor & Kim-Cohen, 2007). For example, a recent study found that 43% of adult males with intellectual disability and aggression had the short-allele MAO-A polymorphism, compared to 20% of matched controls without aggression (May et al., 2009). Positive associations have also been found between the short MAO-A alleles and changes in cortex volume in children with autism, increased autism severity, and decreased communication skills (Cohen et al., 2003; Davis et al., 2008). In contrast, a study assessing aggression in males with Fragile X syndrome did not find a positive MAO-A association (Hessl et al., 2008). Interestingly, Hessl et al. found an association between the severity of aggression and the long variant in the serotonin transporter gene (5-HTTLPR) in males with Fragile X syndrome. These results were later confirmed in an adult sample of males with intellectual disability (May, Lightfoot, Srour, Kowalchuk, & Kennedy, 2010). Hessl et al. suggested that the likely contribution of the long 5-HTTLPR gene is to enhance reuptake of serotonin, which increases sensitivity to environmental adversity.
It is important to note, however, that there are most likely other genes, interactions between genes, and/or environmental relations occasioning aggression in people with intellectual disability. Although genes account for approximately 50% of the influence on developing certain traits, it does not imply the influence of genes is the same for everyone. For example, research has found that the short MAO-A allele alone accounts for aggression in up to 15% of children (Foley et al., 2004). However, aggression increased by 45% in children who possess the short MAO-A allele and have been exposed to environmental adversities (e.g., abuse). Contrastingly, aggression increased by only 2% among children who possessed the long MAO-A allele (Jaffee et al., 2005). Collectively, gene–environment interactions studies suggest that children expressing the short MAO-A allele are 2 to 4 times more likely to develop aggression than children expressing the long MAO-A allele. Similar results have been obtained for the serotonin transporter gene (e.g., Beitchmann et al., 2006; Brune et al., 2006). Therefore, genetic associations may support an increased propensity for an individual to engage in aggression under specific environmental conditions, but not necessarily be diagnostic of a particular disorder.
The purpose of genetic susceptibility screening for symptoms related to disabilities is to identify children at risk as soon as possible. Early detection of symptoms promotes educational planning and treatment, family supports and education, reduction of family stress, and delivery of appropriate medical care to the child (Filipek et al., 1999). However, since there are currently no genetic tests available for accurately diagnosing autism, professionals (e.g., pediatricians, psychologists, audiologists, language pathologists) rely on their clinical judgment and a battery of developmental and psychological assessments to distinguish behavioral symptoms in children with autism from that of other populations. Thus, despite the undesirable effects of aggression, problem behavior typically emerges before it is assessed and treated (Richman & Lindauer, 2005).
Although it is unlikely genetic susceptibility screening for genes associated with aggression will lead to preventive medical practices (e.g., genetic engineering), its use as part of a comprehensive surveillance and evaluation practice might direct behavioral intervention efforts prior to the onset of these behaviors. However, genetic screening of children for susceptibility purposes is highly controversial (Nuffield Council on Bioethics, 1998; Ross, 2002). As a result of low predictive utility, genetic screenings may be misinterpreted among families and/or clinicians. For example, parents may attribute gene association with aggression as a predictor of future prognosis while disregarding the role of the child's environment in developing aggressive tendencies. Furthermore, multiple genes most likely contribute to the overall behavioral phenotype expressed by an individual; so, overly emphasizing the role of a single gene can cause undue stress for a family or lead to false-positive diagnoses. Other issues of contention regarding susceptibility screening include the potential for stigmatization and discrimination, confidentiality, lack of adequate follow-up to testing, using untested treatments and/or preventive measures in children not necessarily displaying symptoms, and selective abortion (Clayton, 2003; DeLisi & Bertisch, 2006; de Vries et al., 1997; Hoop, Roberts, Green-Hammond, & Cox, 2008; McConkie-Rosell et al., 1999; Milner, Collins, Connors, & Petty, 1998; Milner, Han, & Petty, 1999; Ross, 2002; Smith et al., 1996).
Nonetheless, many studies also suggested families find genetic susceptibility screening helpful to some extent. For example, Campbell and Ross (2004) interviewed health care professionals and families about the hypothetical testing of children for genes conferring a propensity toward violence. Most of the health care professionals did not support susceptibility screening unless medical treatment was available, and the behavior had actually emerged. Parents, on the other hand, perceived the benefits of genetic screening as providing information for seeking appropriate educational, therapeutic, and social services irrespective of “cures” for health-related disorders. Similarly, Milner et al. (1998) surveyed undergraduate students about their perceptions of genetic screening for a variety of psychiatric conditions and behavioral traits. Participants generally supported early genetic screening for violent behaviors; however, these participants had no vested interest in the results of this study (e.g., participants did not have children, children with aggression, or children with disabilities).
Little is known about the perceptions of parents of children with autism toward genetic susceptibilty screening for behavioral symptoms. Therefore, the purpose of the current survey was to solicit the perspectives of parents who have children with autism about their views of genetic susceptibilty screening for genes associated with aggression. Responses to genetic screening statements were compared to responses obtained from parents who have children without disabilities, and young adults planning to have children within the next 5 years. We hypothesized parents of children with autism would be more inclined to support genetic screening as opposed to other survey participant groups.
Method and Materials
The survey was pilot tested with 45 demographically diverse students who were majoring in special education at Southern Illinois University. Participants were asked to complete the survey and provide written feedback about the clarity of the stated research purpose and the content of the questions being asked. The survey consisted of 10 questions about demographics, including gender, age, ethnicity, education level, relationship status, family income, and child status (including having a child with autism). The second section of the survey asked respondents to rate the degree to which they agreed (or disagreed) with 30 statements pertaining to genetic screening for aggression on a 4-point Likert-type scale (4 = strongly agree, 3 = agree, 2 = disagree, 1 = strongly disagree). The statements chosen for the survey were derived from previous research on genetic screening for health disorders and/or specific behavioral traits (Campbell & Ross, 2004; Clayton, 2003; DeLisi & Bertisch, 2006; de Vries et al., 1997; Hoop et al., 2008; McConkie-Rosell et al., 1999; Milner et al., 1998, 1999; Ross, 2002; Smith et al., 1996). Respondents were given an explanation of the purpose of the study and the following instructions:
Thank you for taking the time to answer these questions. We are going to ask you about your opinions for preventative genetic screening for aggression. Aggression is defined as a tendency to physically harm others. Genetic screening consists of using a precise DNA test to identify the presence or absence of specific mutations in genes associated with aggression in humans. It is unlikely that any single gene directly causes the development of aggression. However, genes increase the chances that particular behaviors may develop depending upon how they interact with the environment. There is no right or wrong answer, and your responses will remain confidential and anonymous.
The pilot group did not suggest any changes to the wording of survey statements or instructions for completing the survey. Cronbach's alpha test for internal consistency was calculated for each category of statements. Statements in a category with an alpha level below 0.70 were excluded from the survey. The survey was then updated to include the original demographic items and 20 statements equally divided into four categories (see Table 1). Specifically, the statements asked for opinions on the use of genetic screening for determining available interventions (Category 1), exploratory screening for purposes of understanding the etiology of aggression and development of interventions (Category 2), who should bear the financial responsibility for screening (Category 3), and the possibility of discrimination by others based on screening outcomes (Category 4).
To participate in the current genetic screening survey, respondents had to be (a) a parent of a child with autism (with or without aggression), (b) a parent of a child without a diagnosed disability (with or without aggression), or (c) an adult planning to have children within the next 5 years. A Universal Resource Locator (URL) address containing the online survey was disseminated to potential respondents through the Interactive Autism Network (IAN). IAN is an online longitudinal database and research registry at the Kennedy Krieger Institute in Baltimore, Maryland. A total of 1,982 announcements of the survey link were randomly distributed to members on the IAN listserv between January and April 2010 and also posted on the IAN Community Research Opportunities Bulletin Board for the public at large. During the same period, parents of children without disabilities and nonparents were recruited using a chain-referral sampling technique. Pilot study participants were asked to locate one parent of a child without disabilities and one nonparent to complete the online survey. Emphasis was placed on identifying participants from different demographic backgrounds (e.g., age, ethnicity, gender, etc.). All pilot study participants who recruited respondents for the survey were awarded course credit for their efforts. Pilot study participants were assigned a unique code to serve as confirmation that they recruited participants for the study. Parents of children without disabilities and nonparents who responded to the survey request entered this code on the survey to track recruitment efforts for course credit. Internet protocol addresses (IP addresses) of returned surveys were evaluated to ensure only one member of a household participated. None of the returned surveys came from the same IP address, and a total of 224 surveys were received for the study. The Southern Illinois University at Carbondale Human Subjects Committee and IAN research committee approved the survey study. Informed consent was implied by submitting the online survey.
Demographic information pertaining to the three participant groups is presented in Table 2. Most of the respondents were Caucasian (79%), female (79%), and over 33 years of age (67%). Sixty-two percent of respondents earned over $55,000 per year, and most respondents (59%) reported having a college or graduate degree. In terms of family dynamics, over half of the respondents were married (64%), had children (85%), had a child with autism (55%), and had a child with aggression regardless of having an autism diagnosis (51%). Of those respondents reporting they had a child with autism, 64% claimed their child engaged in aggression. Overall, 124 (55%) respondents had children with autism; 67 (30%) had children without disabilities; and 33 (15%) were nonparents. Chi-square analyses were conducted to explore associations between respondent demographics per group and the likelihood of completing the survey. Two of these analyses produced significant associations. Groups differed in terms of age, χ2(6) = 65.47, p = .000, with over 70% of respondents in the autism group and no disabilities groups who were 36 years of age and older, and 84% of respondents in the nonparent group who were 18 to 33 years of age. The autism and no disabilities groups differed in terms of having a child behaving aggressively, χ2(1) = 25.57, p = .000, with twice as many respondents in the autism group reporting their child is aggressive compared to the no disabilities group.
The decision was made a priori to exclude surveys with missing data. None of the returned surveys was incomplete. Responses to categorical statements were aggregated for each participant group. A one-way analysis of variance (ANOVA) was conducted to identify group differences associated with perceptions of genetic screening for aggression for each category. ANCOVA was also calculated for interaction effects on responding between participant group membership and each demographic variable. Post hoc pairwise comparisons of significant differences were conducted using the Bonferroni correction statistic.
Table 3 displays the average response to each category of statements and ANOVA results per participant group. Overall, participant groups differed significantly in perceptions of using genetic screening for treatment decisions, F(2,221) = 4.71, p = .01. Post hoc analyses revealed parents of children with autism differed from parents of children without disabilities in relation to using genetic screening for treatment decisions (p = .01). Although over 50% of respondents in each group claimed genetic screening should be used for treatment considerations, 70% of parents of children with autism were more likely to agree or strongly agree that genetic screening would help identify treatments, their doctor would be able to better identify available treatments if genetic predisposition was known, screening should be made available to parents, and that both medical and behavioral treatments may be supported by results of genetic screening.
Parents of children with autism also differed significantly from parents of children without disabilities and nonparents in terms of exploratory research into etiology or treatment development, Category 2: F(2, 221) = 14.83, p = .00, and financial obligation for genetic screening, Category 3: F(2, 221) = 6.47, p = .00. Although responses from all groups were generally positive, 55% of parents of children with autism were strongly in agreement that genetic screening would support discovery of treatments and/or etiologies of aggression, as opposed to 18% of parents of children without autism and 36% of nonparents. Thus, parents of children with autism are more likely to enroll their children in genetic research studies and support newborn screening if it helps to discover treatments, reduce or prevent the development of aggression, and helps lead to an understanding about why aggression occurs. For Category 3, 87% of parents of children with autism agreed that insurance should pay for genetic screening but that they would also accept the financial burden of paying for treatments if their child was identified through genetic screening. The other two groups were less in agreement with statements in this category, especially if payment for assessment and treatment was to come out-of-pocket.
No differences were observed between groups in terms of agreement that discrimination may occur as a result of genetic screening results. Therefore, all groups agreed that if genetic predisposition toward aggression was known, the child and/or family may experience discrimination from employers, insurance companies, peers, and law enforcement, particularly if the affected child was a member of a minority group. Finally, separate one-way ANCOVA calculations did not reveal any significant differences in responses based on gender, age, ethnicity, income, relationship status, education, or whether or not the respondent's child engaged in aggression. Based on the values obtained from the statistical analyses, responses to survey items appear to be related to whether or not the respondent had a child with autism.
The aim of the current genetic screening survey was to explore the perceptions of parents about using screening as a susceptibility measure of aggression. This study extends previous research on parent perceptions by comparing survey responses of parents with children diagnosed with autism to parents of children without disabilities and nonparents (e.g., Selkirk, Veach, Lian, Schimmenti, & LeRoy, 2009). It appears the results of the current study were indeed associated with whether or not a respondent's child had a diagnosis of autism. The reader should be cautioned, however, that hypothetical interest in genetic screening for susceptibility has been shown to be a poor predictor of actual test uptake (Lerman, Croyle, Tercyak, & Hamann, 2002). It is possible that the high numbers of participants in our study expressing intention to test may not mirror the actual future demand for genetic screening for aggression.
A small number of studies have focused on parents' opinions about genetic susceptibility screening. As mentioned previously, parents generally support genetic susceptibility research when there are effective interventions, and the information obtained from screening provides insight into the etiology or prevention of specific traits (e.g., Campbell & Ross, 2004; Clayton, 2003; DeLisi & Bertisch, 2005; Hoop et al., 2008). Parents of children with autism in the current study were more likely to respond positively to using genetic screening for treatment decisions, exploratory research into etiology or treatment development, and financing genetic screening than the comparison groups. The results are not surprising, since behavioral, social, and cognitive characteristics of autism are associated with stress in parents (e.g., Hastings, 2003). In other words, it is possible that parents of children with autism are more sensitive to activities that would lead to the resolution of problems they experience with their children, including aggression.
Some studies have shown parents perceive genetic predisposition or heritability as a primary cause of autism (e.g., Elder, 2001; Harrington, Rosen, Garnecho, & Patrick, 2006; Mercer, Creighton, Holden, & Lewis, 2006; Selkirk et al., 2009). Thus, parents of children with autism may be more likely to support genetic susceptibility screening because they have a more direct need for genetic assessments. For example, Bailey, Armstrong, Kemper, Skinner, and Warren (2009) pointed out that parents informed about their child's genetic linkage to a disorder want to learn more about its associated symptoms, prognosis of health and development, reproductive risks, potential treatments and interventions, and how to communicate with other families who have children with the same diagnosis. Studies have also shown parental decision to have additional children was associated with a current child being diagnosed with an autism spectrum disorder (Jones & Szatmari, 1988; Selkirk et al., 2009).
Although the current study provides insight into the perceptions of susceptibility screening for aggression, there are some limits to the generalizability of our findings. First, survey results may be skewed because individuals with neutral or negative attitudes toward genetics may have been less likely to participate in the survey. Previous research has shown parents of children with autism not only perceive genetics perturbations to be a primary etiology, but also have positive perceptions of genetic screening if symptoms of autism are severe (Selkirk et al., 2009). A related limitation is that the current study did not differentiate among autism spectrum disorders. Therefore, responses to the survey could not be examined based on characteristics other than the presence of aggression. Parents of children with less severe symptoms may not have participated in the study. Third, it was not possible to determine the actual response rate for the online survey. Because efforts were made to reduce correlations between multiple respondents from a single household, it is possible additional surveys were excluded from the sample size.
Finally, the sampling technique used to recruit parents of children without a disability and nonparents could introduce bias to the survey responses. Chain-referral methods start with a set of initial respondents (e.g., pilot study participants) who refer their acquaintances for study participation. Biases may include demographic sameness, volunteerism, and differential recruitment (i.e., one peer group overrepresented or underrepresentation of those who are less socially connected). However, demographic sameness was viewed as a strength of the chain-referral method for the current study. The pilot group represented a wide range of demographic variables, including gender, ethnicity, income, relationship status, and age. Thus, recruiting participants for the study increased the likelihood of a demographically diverse participant pool. Nonetheless, voluntary completion of the survey did not provide the opportunity to investigate differences between study participants and nonparticipants. Furthermore, not all of the initial pilot group members chose to recruit participants for the study. Factors such as the degree of relationship (e.g., family members versus friends) and desirability of earning course credit (e.g., course credit may not have been perceived as important) could influence who recruited and who responded to the survey. Although these factors could also influence the over- or underrepresentativeness of a demographic variable in this study, chi-square analyses revealed no significant differences among the participant groups.
Parents often need assistance understanding and evaluating the credibility of medical and behavioral services related to aggression (Hardey, 1999; James & Levy, 2006). Thus, preservice training of health care professionals on the genetic susceptibility of autism and traits associated with the disorder may help alleviate confusion or misconceptions held by parents. For example, parents are most likely to approach health care professionals about the strengths and weaknesses of assessments and interventions, etiology, and prognosis of disabilities and/or specific behavioral traits and ramifications for participating in activities aimed at minimizing or preventing problem behavior. Although there may be a genetic predisposition to aggression, parents should be informed that the risk of developing aggression based on genetic results is relatively low, and overreliance on one assessment could have untoward consequences (e.g., increased stress, seeking out unreliable interventions). In addition, genetic predisposition is not prescriptive for the types of interventions chosen or the relative success of those interventions. Likewise, insurance providers may not cover specific interventions (e.g., applied behavior analysis), so finding the resources to pay for effective treatments for aggression may be a challenge. Despite evidence that insurance companies do, in some cases, reimburse genetic screening, genetic counseling services historically have been poorly reimbursed (Ostrer et al., 1993). With reduced financial support for clinical genetic services, parents may be required to pay out-of-pocket for the extra expenses of treatment once genetic susceptibility has been identified.
None of the groups differed in its perception that genetic information could be used against the individual expressing genes associated with aggression. As discussed previously, insurance companies and employers could use screening results to determine insurance rates and employment status. Although the enactment of the Health Insurance Portability and Accountability Act of 1996 (HIPAA) explicitly prohibits insurance companies and employers from using presymptomatic genetic diagnoses to profile high-risk groups, it does not prohibit mandatory testing or disclosure of genetic information to third parties. There is also the possibility of genetic information being used by law enforcement to profile individuals, which may lead to harassment, unnecessary surveillance, or misrepresentation of an individual's character. Therefore, parents should be aware of their legal rights regarding privacy and confidentiality, which protect citizens from undue judgment.
Genetic susceptibility screening can be part of a comprehensive and proactive approach to identifying and treating symptoms of autism. For example, information on genetic susceptibility to challenging behavior could be provided as an option for parents. In the event parents choose genetic screening and find their child is susceptible to aggressive tendencies, additional resources could be made available to them. Behavior management classes could be made available that allow parents to learn behavioral techniques for identifying and preventing environmental contingencies setting the occasion for aggression to occur. If those children identified as having a predisposition toward aggression deviate from typically developing children in the development and/or severity of aggression, these children could be referred for specialized training such as in-home behavioral services. If behavioral interventions are not as effective as intended, parents could then have the option to seek medical advice for pharmacological options to reduce aggression. However, parents should be aware that very little data exists currently supporting the efficacious benefits of medication on aggression (e.g., Tsiouris, 2010).
Editor-in-Charge: Glenn T. Fujiura
Michael E. May (e-mail: email@example.com), Educational Psychology and Special Education, Southern Illinois University at Carbondale, Mailcode: 4618, Carbondale, IL 62901, USA; Rachel C. Brandt and Joseph K. Bohannan, Southern Illinois University.