“If He Has it, We Know What to Do”: Parent Perspectives on Familial Risk for Autism Spectrum Disorder

Abstract

Objective

Predictive testing for familial disorders can guide healthcare and reproductive decisions. Familial disorders with onset in childhood (e.g., autism spectrum disorder [ASD]) are promising targets for presymptomatic prediction; however, little is known about parent perceptions of risk to their children in the presymptomatic period. The current study examined risk perceptions in parents of infants at high familial risk for ASD enrolled in a longitudinal study of brain and behavior development.

Methods

Semistructured interviews were conducted with 37 parents of high-risk infants during the presymptomatic window (3–15 months) that precedes an ASD diagnosis. Infants were identified as high familial risk due to having an older sibling with ASD. Parent interview responses were coded and interpreted to distill emerging themes.

Results

The majority of parents were aware of the increased risk of ASD for their infants, and risk perceptions were influenced by comparisons to their older child with ASD. Parents reported a variety of negative emotions in response to perceived risk, including worry, fear, and sadness, and described impacts of perceived risk on their behavior: increased vigilance to emerging symptoms, altered reproductive and healthcare decisions, and seeking ongoing assessment through research.

Conclusions

Parents of children at high familial risk for childhood-onset disorders like ASD face a period of challenging uncertainty during early development. In anticipation of a future in which presymptomatic testing for ASD is made available, it is important to understand how parents react to and cope with the elevated—but still highly uncertain—risk conveyed by family history.

Presymptomatic prediction of heritable health conditions has received considerable research investment given the potential to improve outcomes through earlier identification and treatment. In children, predictive testing is generally recommended only in situations where effective early intervention is readily available (Borry et al., 2006; AAP/ACMG, 2013). For potentially life-threatening disorders with onset in childhood (e.g., hereditary cancers; Kodish, 1999), the motivation for presymptomatic identification and intervention is clear. Psychiatric and neurodevelopmental disorders of childhood, in contrast, present a complex set of clinical and ethical questions about the implications and utility of predictive testing. A validated predictive test could provide a more precise risk estimate than empiric risk based on family history alone, but could also create a situation in which knowledge of risk is not accompanied by options for presymptomatic intervention (Erickson et al., 2014).

Autism spectrum disorder (ASD) is an example of a neurodevelopmental disorder for which considerable research effort has been made towards earlier prediction and identification. ASD affects approximately 1.7% of U.S. children (Centers for Disease Control and Prevention, 2018) and is highly heritable (Tick, Bolton, Happé, Rutter, & Rijsdijk, 2016). Parents of children with ASD are aware of its heritability, and tend to overestimate the recurrence risk to subsequently-born children (Chen et al., 2015; Mercer, Creighton, Holden, & Lewis, 2006; Selkirk, McCarthy Veach, Lian, Schimmenti, & LeRoy, 2009; Whitelaw, Flett, & Amor, 2007). Knowledge of familial risk alone may impact healthcare and reproductive decisions in parents of children with ASD. For example, parents with one child with ASD are less likely to fully vaccinate subsequently born children (Zerbo et al., 2018), and may decide to forgo future childbearing (Selkirk et al., 2009; Wood et al., 2015).

Currently for the large majority of families there is no predictive test for ASD; however, in the research context, patterns of alteration in brain structure and function have been shown to be highly accurate in predicting later ASD diagnosis in infants at high familial risk (with positive predictive values of 81% and 100% and sensitivities of 88% and 82% for structural and functional findings, respectively; Emerson et al., 2017; Hazlett et al., 2017). These reports demonstrate that brain changes detectable on MRI occur during infancy, prior to the onset of symptoms; however, these findings require replication before MRI prediction can be used clinically. Infants identified presymptomatically could be closely monitored and enrolled in interventions as soon as symptoms emerge (typically between 18 and 24 months). Improved risk estimates could also enable research trials of preventive interventions in infancy.

Whether parents would seek a predictive MRI test to increase the precision of the estimate of their infant’s risk for ASD is unknown, however. Some parents may, in fact, prefer the uncertainty of a less precise risk estimate (Meiser & Dunn, 2000; Peay, Hooker, Kassem, & Biesecker, 2009; Whitmarsh, Davis, Skinner, & Bailey, 2007). In anticipation of future in which presymptomatic testing for ASD is made available, it is important to understand how parents react to and cope with the elevated—but still highly uncertain—risk conveyed by family history. Prior work suggests that parents who report more uncertainty about a child’s undiagnosed illness also feel that they have less control over the child’s health (Madeo, O'Brien, Bernhardt, & Biesecker, 2012). If this generalizes to the context of familial risk for ASD, parents experiencing the greatest uncertainty may have more difficulty coping in the first years of their child’s life. Parental sense of self-efficacy and adaptive coping (which both contribute to improved family functioning in ASD; Estes, Swain, & MacDuffie, 2019) could potentially be increased by obtaining more precise risk estimates via predictive testing.

The goal of the current study was to gain insight into how parents perceive and respond to their child’s familial risk for ASD. Specifically, we sought to explore parent perspectives of risk in order to better understand what might motivate parents to seek (or not seek) predictive ASD testing if it becomes available. We leveraged existing qualitative data collected from parents of infants at high familial risk for ASD (by virtue of having an older sibling with ASD) who were enrolled in a longitudinal study of brain and behavior development from 3 through 24 months of age. Parent interviews were conducted as part of a supplemental study designed to investigate parent beliefs about early autism risk and associations with parenting stress, coping, and family functioning.

These interviews provided an opportunity to investigate parent perspectives on ASD risk that was distinct from prior work in two ways. First, the interviews were conducted during the presymptomatic window (3–15 months) that precedes an ASD diagnosis. During this period, perceptions of an infant’s ASD risk can influence parental decision-making (e.g., Zerbo et al., 2018), and potentially impact developmental outcomes. Prior studies have described the impact of perceived recurrence risk on reproductive choices (Chen et al., 2015; Mercer et al., 2006; Selkirk et al., 2009; Whitelaw et al., 2007), but have not explored how parental risk perceptions may impact the experience of parenting an infant at elevated risk. A second unique feature is that participants were parents enrolled in a longitudinal neuroimaging study. This group may differ from parents of infants at high familial risk who are not enrolled in research (see Limitations section for additional discussion), but are likely similar to those parents who would be the earliest adopters of a future MRI-based predictive test for ASD. For these reasons, we felt that these previously collected data were particularly well-suited for investigating our questions of interest.

Through content analysis of parent interview data, we sought to (a) assess perceptions of ASD recurrence risk in a sample of parents who have at least one child with ASD and an infant enrolled in a prospective neuroimaging study, (b) explore common factors that influence risk perception, and (c) gain insight into how these parents respond to and cope with perceived risk to their infant. By gaining a better understanding of the experience of these parents in relation to their child’s ASD risk, we can improve our predictions about what might motivate families to seek more precise risk estimates when/if a validated predictive test becomes available. In addition, knowledge generated from these interviews could inform the development of responsive supports and clinical services for families of infants at high risk for ASD.

Methods

Design of Overall Study and Parent Supplement

The interview data analyzed for the current manuscript were collected as part of a supplemental study with parents of infants enrolled in the NIH-funded, Autism Center of Excellence Infant Brain Imaging Study (IBIS)—a longitudinal, multisite study of early brain and behavior development in infants at high and low familial risk for ASD. IBIS participants were recruited through research participant lists/registries, flyers, brochures, electronic medical record searches, and community clinics at each site. Only parents of high-risk infants were interviewed for the supplemental study. High-risk infants in IBIS were defined as having an older sibling with a clinical diagnosis of ASD and no genetic, medical/neurological, or pre/perinatal exposures that could presumably impact ASD risk (see Estes et al., 2015 for full IBIS inclusion/exclusion criteria). All infants enrolled in IBIS participated in 1–2 days of behavioral and brain imaging (MRI) assessments over multiple visits (beginning at 3 or 6 months of age, ending at 24 months). A clinical best-estimate diagnosis of ASD was obtained as part of a 24-month assessment. The institutional review board at each participating site approved the IBIS research protocol, and parents provided informed consent.

The supplemental study began toward the end of the main IBIS project and was approved as a modification by institutional review boards at the three IBIS clinical sites that participated. The aim was to collect additional data from ∼50 parent participants to explore how parent risk perceptions impacted their well-being and family functioning. Following IRB approval, supplemental data were collected via qualitative interviews and parent questionnaires. The supplemental data collection was incorporated into the next scheduled study visit for all incoming and actively enrolled participants. Supplemental data were collected during the routine clinical interview portion of study visits. All parent participants completed the initial supplemental interview. A subset were interviewed a second time at a later visit, but questions in the second interview focused on parental responses to clinical feedback received through IBIS and thus have not been included in the current report.

Parent participants had some prior exposure to the concept of familial recurrence. Recruitment and consent materials for the main IBIS study specified that infants in the high-risk group were recruited due to having an older sibling with ASD; however, education and counseling about recurrence risk were not systematically provided to parents as part of their enrollment in IBIS. It is possible that parents may have had conversations or discussions about recurrence risk with their medical providers or members of the IBIS research team prior to participating in the supplemental interviews.

Interviews

The supplemental parent interviews were semistructured, based upon an interview guide (presented in its entirety in Table I), and intended to explore parental understanding of and response to familial risk in infants with an older sibling with ASD. The interview guide was developed collaboratively by highly experienced clinician-investigators across the IBIS network, and was designed to be brief and straightforward to administer across sites. Interviews were conducted by licensed clinical psychologists and/or clinical psychology trainees at each site who were trained in interview administration by the second author (L.Turner-Brown).

Table I.

Interview Guide Questions

1. What do you know about the risk associated with your youngest child having autism?

1.a. If you had to estimate a percent chance, what would that number be?

2. How has the information you know about risk affected you and your family?

3. In general, are you worried about your youngest child?

4. How do your concerns affect you and your family?

The complexity of integrating the supplemental interview into an ongoing research visit protocol resulted in some variation across sites (see Limitations section for additional discussion). In all cases the interviewer explained that the questions were intended to explore parental concerns about their child enrolled in IBIS due to having an older sibling with ASD. The interviews lasted 7 minutes on average (range: 2.5–17.5 minutes). Interviewers were instructed to follow up with additional questions/probes if certain responses were unclear or in need of elaboration; however, systematic probes were not provided. An additional source of variation was the timing of the interview. In some cases, the interview was conducted before parents received feedback about their infants’ developmental scores (from the Mullen Scales of Early Learning, a standardized developmental assessment of an infant’s gross and fine motor, expressive and receptive language, and visual reception skills; Mullen, 1995), and in some cases the interview was conducted after this feedback. Data on whether the interview preceded or followed developmental feedback was not systematically recorded. However, none of the parents had yet received diagnostic feedback (i.e., whether their child met diagnostic criteria for ASD) at the time of the interview, as a full diagnostic assessment was not done until the 24-month visit.

Content Analysis

Interview recordings (video or audio) and/or transcripts were obtained from the three participating IBIS sites. Videos were coded when available (31 interviews), otherwise audio recordings (2 interviews), or transcripts (4 interviews) were used. Interview data were analyzed by the first author (K. E. MacDuffie; a postdoctoral research associate and licensed clinical psychologist) and last author (H. L. Peay; a senior researcher and certified genetic counselor), using MAXQDA software (VERBI Software, 2016).

A conventional content analysis approach was used with the interview data, meaning that codes were derived directly from the data, and only content relevant to our research questions was coded (Hsieh & Shannon, 2005). Our approach was inductive and exploratory, rather than hypothesis-driven. Analysis of the data began via immersion: K. E. MacDuffie and H. L. Peay viewed, listened to, or read through all the interviews and then drafted a preliminary codebook. The preliminary codebook was applied to transcripts of five interviews by both coders, and was then revised by consensus to create the final codebook. K. E. MacDuffie coded all the interviews by systematically applying the codes to the video/audio recordings or transcripts using the MAXQDA software, and H. L. Peay reviewed the interviews and codes to ensure consensus across major coding areas. Both coders reviewed and interpreted the resulting code reports to develop a set of major themes. Interpretation of themes and subthemes were informed by a relevant conceptual model (The Common Sense Model of Self-Regulation [CSM]; Leventhal, Phillips, & Burns, 2016), which is described in more detail in the Discussion.

Results

Sample Characteristics

Data from 37 parent interviews were included in the current analysis; 6 additional parents were interviewed, but the sessions were not audio or video recorded and so could not be analyzed. Two parents were interviewed twice (at two different study visits) because they had two children enrolled in IBIS. In these cases, only the interview conducted first (for the older enrolled child) was included in the analysis. Interviews were conducted at three of the four IBIS clinical sites between 2014 and 2017: 19 interviews were conducted at the Children’s Hospital of Philadelphia, 11 at Washington University at St. Louis, and 7 at the University of North Carolina at Chapel Hill. Interviews were conducted when infants were 3–15 months of age (see Table II for detail); 32 interviews were conducted with the mother only; 4 interviews were conducted with both parents, and 1 with the father only. Parent demographics are depicted in Table III. All families interviewed had at least one older child with ASD; 7 families had two or more older children with ASD.

Table II.

Interviews by Age of Infant and Study Site

Age (months)	Philadelphia	St. Louis	Chapel Hill	Total
3	7	7	2	16
6	7	2	3	12
9	0	1	0	1
12	2	1	0	3
15	3	0	2	5
Total	19	11	7	37

Note. 3 of 4 IBIS clinical sites conducted interviews for the supplemental study: The Children’s Hospital of Philadelphia (Philadelphia, PA), Washington University in St. Louis (St. Louis, MO), and The University of North Carolina at Chapel Hill (Chapel Hill, NC). The fourth IBIS clinical site (The University of Washington, Seattle, WA) did not participate in the supplemental study.

Table III.

Parent Demographics

Gender
Female	36
Male	5
Race/Ethnicity
White/Caucasian	85%
Non-Hispanic	93%
Highest level of education
High school diploma	5%
Some college	24%
College degree	39%
Graduate degree	32%

Note. 32 interviews were conducted with the mother only; 4 interviews were conducted with both parents, and 1 with the father only.

Parent Knowledge of ASD Recurrence Risk

Parents’ prior knowledge of ASD recurrence risk for younger siblings of children with ASD was directly queried (see Table I). Two parents said they were unaware or unsure about the recurrence risk; all those remaining reported prior knowledge of elevated recurrence risk. Seventeen parents gave quantitative estimates of recurrence risk, which ranged widely: 3 parents estimated the recurrence risk accurately at 18–20% (Ozonoff et al., 2011), 10 parents overestimated the recurrence risk as numerically higher than the empiric risk (ranging from 25 to 50% chance of ASD with one affected sibling, and 60–80% chance with two affected siblings), and 4 parents underestimated the risk (2–15% chance of ASD). Eighteen parents were aware of the higher recurrence risk, but did not provide a quantitative estimate. For example, one parent stated: “I just know he is at higher risk of having autism. It doesn’t mean he is going to have it, but he is at higher risk than typical children would be.” (3 month visit). Some parents spontaneously mentioned an awareness of the sex difference in ASD risk (higher for boys than girls; Halladay et al., 2015).

Comparison of Infant to Older Child with ASD

Unprompted, most of the parents made comparisons between the behavior of their infant and affected child. Parents related comparisons to perceptions of their infant’s risk. Some parents described their infant as similar to their affected sibling. For example, “He was showing a lot of signs that my oldest showed… He reminds me a lot of my oldest and so of course that was a little bit of a panic.” (3 month visit). Others described differences in the development of their infant and affected sibling. Parents who noticed similarities tended to report increased vigilance about potential ASD symptoms and greater worry, whereas parents who noticed differences tended to report being less worried. For example, one parent stated: “Honestly, I’m not worried at all because I’ve already noticed him doing things that [affected child] never did.” (12 month visit). Finally, a number of parents reported that their infant was too young for comparisons to be relevant. In some cases, parents related this perspective to the prior experience of observing a regression or loss of skills in the older child. As one parent put it, “we haven’t yet hit the danger zone” (12 month visit).

Affective Response to Perceived Risk

Parents reported a range of emotional responses to the perceived risk for ASD. The most common emotion reported was worry; however, it is important to note that the interview guide asked directly about worry, which may have inflated this response. As one parent noted: “We’re worried, of course we are. You don’t want your child to suffer or to have any difficulties in life.” (3 month visit). Though the phrasing of this question asked specifically about worry, in their answers parents reported other affective responses besides worry. Some parents described living with uncertainty as “scary,” and a few described a fear that failing to constantly interact with their infant might affect their developmental outcome. For example, one parent noted: “It’s like, ‘Oh my God, he is going to become autistic because he is lying by himself!’” (15 month visit). Finally, although the interview guide was not designed for in-depth exploration of emotional reactions, two parents became tearful while describing their experience, and one conveyed a sense of loss: “It's extremely hard because many times I feel like I'm losing his babyhood. I don't get to be like the mom who gets to ooh and ahh over sweet little things he learns or not have to worry about it” (6 month visit).

Parental Coping with Perceived Risk

Parents described a variety of strategies or ways that they coped with perceived risk to their infant. We conceptualized these responses as falling into two general categories: behavioral (i.e., related to taking or opting out of an action) or cognitive (i.e., changing perspective or cognitive reframing techniques). The most commonly reported behavioral strategy was increased vigilance: parents mentioned being watchful, aware, or cautious about early signs of autism-related symptoms. For example, one parent reported that she is “Just more aware, keeping an eye on the little ‘-isms’. Like how frequently they are spinning that tire… [laughs]” (3 month visit). Some parents also discussed the impact of perceived risk on their reproductive behavior. A number of parents mentioned that their knowledge of familial risk had made them question or decide not to have additional children. However, a similar number of parents reported no impact on childbearing decisions. As one parent put it, “We knew we wanted kids, however they came” (6 month visit).

As part of their enrollment in IBIS, these parents were receiving periodic developmental assessments and feedback from clinicians trained in the early detection of ASD; a number of parents explicitly mentioned their enrollment in IBIS as a step that they had taken to ensure that their child’s development was being monitored. Parents also mentioned an impact on other types of healthcare decisions: for example, choosing not to vaccinate their infant, or enrolling their infant in pre-emptive early intervention services. One parent reported a change in interactions with her pediatrician: “I feel like when I go to the pediatrician I’m a little less inclined to take everything they say, just because that is what happened with [affected child], when I brought up concerns and they were brushed off” (6 month visit).

Many parents also reported engaging in cognitive coping strategies, which served to put into perspective or reframe the perceived risk. Some parents described their infant’s risk for ASD as outside of their control. For example: “It is what it is, and he’s born the way he’s born.” (6 month visit). Others reported an acceptance of a potential ASD outcome, often based on the progress shown by their affected child. For example, one parent described how her conceptualization of ASD shifted over time: “I think that our opinion has changed a little bit over the last, like, 6 month or so, from the original diagnosis [of affected child] … I think a lot of our scare like ‘he’s never going to do anything in life' has, you know, changed. We have different outlooks through therapy and stuff” (15 month visit).

The most common cognitive strategy reported by parents was feeling ready or prepared for another diagnosis, based on their experience with their affected child. For example, “We know autism now, so if he has it, we know what to do.” (15 month visit). A few parents reported taking the additional step of assuming their infant would develop ASD: “I feel like sometimes I'm already labeling him, but it's like it's stupid not to. I mean, it's only going to help him. And if the end result is that he doesn't have it, then maybe that’s part of that.” (15 month visit).

Discussion

Leveraging existing data collected as part of a longitudinal neuroimaging study, we explored parent perspectives on risk from 37 parents of high-risk infants interviewed during the presymptomatic window that exists prior to the emergence of ASD symptoms. Our results can be summarized in five key points: (a) The majority of parents were aware of the increased risk of ASD for their infants, and most parents who provided quantitative estimates overestimated the empiric recurrence risk. (b) Parent perceptions of risk were influenced by comparing the behavior of their infant to their affected child; noted similarities were associated with more worry, noted differences were associated with less worry. (c) Parents reported a variety of negative emotions in response to perceived risk, including worry, fear, and sadness. (d) Parents altered their behavior in response to perceived risk—they reported being more watchful or vigilant for early signs of ASD, considering familial risk in reproductive and healthcare decisions, and enrolling in IBIS as a way of monitoring their infant’s early development. (e) Parents described cognitive strategies for coping with uncertainty, such as perceiving ASD risk as out of their control, or reflecting upon how prior experience had prepared them to accept and/or manage another ASD diagnosis.

Our results are consistent with prior studies of recurrence risk perception in parents of children with ASD. Prior work has shown that parents tend to overestimate the empiric recurrence risk (Chen et al., 2015; Mercer et al., 2006; Selkirk et al., 2009; Whitelaw et al., 2007), and similarly report impacts on reproductive and vaccination decisions (Chen et al., 2015; Selkirk et al., 2009; Zerbo et al., 2018). Our data take a step beyond this prior work by exploring parent perspectives on familial risk during the first 2 years of life, when parents are faced with uncertainty about whether, when, and with what severity ASD symptoms will emerge for their infant. Parental perceptions of ASD risk—in this presymptomatic period prior to diagnosis—have potential to shape the healthcare decisions that parents make for their families (e.g., vaccination, seeking developmental or early intervention services, and family planning; Selkirk et al., 2009; Webb, Jones, Kelly, & Dawson, 2014; Wood et al., 2015; Zerbo et al., 2018).

Some aspects of the experiences of parents in our study may be specific to ASD. For example, the concept of “the danger zone” invoked by one parent (referring to the uncertain window of potential symptom onset) and the desire for early monitoring/intervention both reflect parental understanding that ASD unfolds during early development (rather than existing from birth) and is potentially treatable with early intervention. These characteristics of ASD make it distinct from other familial disorders with onset in childhood (e.g., Fragile X syndrome, discussed in the next section). However, other findings may generalize to other disorders. Our results are generally consistent with predictions from the CSM (Leventhal et al., 2016)—a widely used framework for modeling how individuals understand and cope with health threats. For example, the CSM predicts that individuals will interpret symptoms based on previous health/illness experiences when they assess health threats. This is similar to the tendency of parents in our study to compare their infant’s behavior to that of the affected older sibling. The CSM also postulates that individuals make a variety of efforts to cope with perceived risk and engage in a dynamic process of appraising their coping efficacy and adjusting their strategies as needed. Future research that directly assesses parental coping during the presymptomatic period in ASD could provide additional insights into the relationship between parental perceptions of risk, feelings of uncertainty, and adaptive or maladaptive coping (Madeo et al., 2012).

Limitations

The sample of parents interviewed (all enrolled in an infant neuroimaging study) is both a strength and limitation of this study. These existing interview data were uniquely suited to answering our questions of interest, given the timing of the interviews (after birth but prior to onset of ASD symptoms) and the similarity of the parents to future early adopters of predictive testing if/when it is offered clinically. However, it is important to recognize that all parents interviewed had, a priori, taken the active step of enrolling in a longitudinal research study that would provide them with regular feedback about their infant’s development. Similar to other predictive testing contexts, it may be that those who actively seek this type of information may be the most emotionally equipped to deal with what they learn (Meiser & Dunn, 2000). Alternatively, it could be that the most worried parents choose to participate in research. Recruitment into IBIS targeted infants with one or more older siblings with ASD, which may have alerted parents to elevated ASD recurrence risk and/or made the risk more salient. Our sample of parents (like the full IBIS sample) was also highly educated and predominantly Caucasian, and thus not representative of all parents of children with ASD. Capturing the full range of parent attitudes about familial risk and predictive testing, therefore, will require research with a broader, more representative sample of families.

Additional limitations affect the interpretability of specific results. One question in the interview guide asked directly about worry, and thus may have inflated parent reports of worry beyond what might have been reported spontaneously or with a more open-ended question about emotional responses. We also do not have precise data about the timing of reproductive decisions made by these parents; for example, we do not know whether the infant enrolled in IBIS was conceived before or after the older sibling was diagnosed.

Finally, the interviews were conducted by available clinicians at each of the three clinical sites. This meant that different clinicians conducted interviews at each site (and often, within site), and while all clinicians were trained to administer the interview, there was variability in the phrasing of interview questions, number of follow-up questions, and in the environment in which the interviews occurred. In most cases, the clinician interviewing the parents had also completed the developmental assessment with the child. The dual nature of this relationship meant that parents may have been more inclined to ask for information or advice from the clinician, or that the clinician was more likely to offer information or reassurance than if the interview had been conducted by a neutral third party. Finally, in some cases the interviews were conducted after the parents had received feedback from the developmental testing (scores on the Mullen Scales of Early Learning), whereas in some cases in interview preceded the feedback.

Implications for Presymptomatic Prediction in Childhood

These data and prior work suggest that a subset of parents of children with ASD are aware of elevated risk to subsequently born children and motivated to engage in intensive monitoring in the hopes that early detection of ASD symptoms would benefit their infant. The parents we interviewed (enrolled in a longitudinal study of brain development in high-risk infants) represent the types of early adopter parents who might seek predictive testing for ASD in a research or clinical context if it were offered. Worth considering, therefore, is how these data can inform our understanding of how parents might approach predictive testing for ASD.

The risks and benefits of predictive testing during infancy and childhood have been addressed for other childhood-onset disorders. Newborn screening is the most comprehensive predictive testing effort that occurs during childhood, and strict criteria must be met before a new test is added to the screening panel. A central consideration for inclusion is whether effective, immediately available treatments exist (Kemper et al., 2014). Predictive tests for ASD, if validated, would not meet this criterion, as there are currently no evidence-based treatments for presymptomatic infants (Landa, 2018; Zwaigenbaum et al., 2015). Interestingly, the promise of early intervention was mentioned by a number of parents who were interviewed, and no parents mentioned the absence of evidence-based presymptomatic interventions. However, intervention was not a focus of the interview guide questions—parents may have had concerns about efficacy or access that were not spontaneously mentioned, and these should be assessed in future research. It is possible that concerns about access to intervention would have been reported had we recruited a more representative sample of families across a larger geographic area.

The unavailability of treatment, however, does not necessarily mean predictive testing has no value. Fragile X is an inherited neurodevelopmental disorder which can be detected with a genetic test, but which is not currently part of newborn screening panels due to the absence of medical treatment and the ambiguity of the risk facing children identified as carriers, but who do not have the full mutation (Skinner et al., 2011). However, the majority of parents of children with Fragile X are in favor of newborn screening, noting that earlier diagnosis enables earlier enrollment in supportive developmental services and informed reproductive and healthcare decisions (Skinner, Sparkman, & Bailey, 2003). While there are a number of parallels between Fragile X and ASD (25–35% of males with Fragile X are diagnosed with ASD; Bailey, Raspa, Olmsted, & Holiday, 2008), the etiology and inheritance pattern of Fragile X is clearly identified, whereas the etiology and inheritance pattern in most cases of ASD remains unknown. Predictive testing of ASD via emerging MRI methods (Emerson et al., 2017; Hazlett et al., 2017) could guide clinical decision-making, but would not reveal information about etiology or inheritance patterns to families. Some parents in our study described the subjective benefits of simply knowing that their infant was at elevated risk for ASD (e.g., mental preparation, earlier monitoring), which parallel the reported benefits in the Fragile X literature (Skinner et al., 2003). This suggests that a more precise risk estimate for ASD could have personal utility for families, even if it does not provide etiological clarity (Kohler, Turbitt, & Biesecker, 2017).

Given the complex, multifactorial nature of ASD, a parallel can be drawn to psychiatric disorders that manifest in late adolescence, such as bipolar disorder and schizophrenia. These disorders are also familial but cannot be predicted with a simple genetic test. Parents of children at high familial risk for bipolar disorder described engaging in symptom monitoring (similar to our ASD parents) but some were ambivalent about obtaining a more precise quantitative estimate of their child’s risk (Peay et al., 2009). Other studies have found that parental interest in predictive testing for mood disorders varies depending on the predictive power of the test and/or the availability of effective early interventions (reviewed in Erickson et al., 2014). Therefore, it may be that the attitudes of parents towards presymptomatic prediction of ASD are intermediate between the Fragile X and mood disorder examples: a combination of ambivalence due to uncertain etiology and lack of available treatments, and a desire to know for the sake of informed decision-making and subjective wellbeing (Kohler et al., 2017). Evidence of parental ambivalence towards ASD risk information comes from two recent intervention trials targeting 12-month-olds via community-based screening: approximately half of parents told their infant was at high risk for ASD declined further assessment (59%, Baranek et al., 2015; 43%, Watson et al., 2017). Clearly, future work to reveal attitudes of parents toward predictive testing for ASD is necessary to assess the motivations of parents to seek (or not seek) testing during the presymptomatic window.

Parents of children at high familial risk for childhood-onset disorders face a period of challenging uncertainty during early development. For example, in the case of ASD, parents may be aware that their infant is at higher risk but must wait for diagnostic confirmation until the age at which ASD symptoms usually manifest (∼18–24 months). This study explored the perspectives and experiences of parents of high-risk infants during this window of presymptomatic uncertainty. Efforts to develop predictive tests for ASD are currently underway (Emerson et al., 2017); if validated, infants identified at highest risk could be targeted for ongoing symptom monitoring and/or trials of presymptomatic interventions. Participating in predictive testing, ongoing monitoring, and targeted intervention trials will require considerable motivation and investment from parents (on top of caring for an older child with ASD). Future work will reveal whether the potential benefits and risks of predictive testing for ASD align with the motivations and concerns of parents of infants at high familial risk.