Relationship Among Anxiety,Depression,and Family Impact in Adolescents With Autism Spectrum Disorder and Average-Range IQ

Abstract

Little is known about how internalizing symptoms in autism spectrum disorder (ASD) affect family outcomes, despite the high rates of internalizing disorders in ASD and the dynamic relationships among child, parent, and family variables. We evaluated the predictive value of internalizing symptoms in 69 adolescents with ASD and average-range cognitive abilities on family impact. In addition, we examined the correspondence between parent and child symptom ratings. Results indicated that internalizing symptoms predict family impact above IQ and externalizing symptoms. Depression (but not anxiety) predicted more positive family functioning, and parents rated more internalizing symptoms than adolescents. The results are contrary to our initial predictions, and we offer several explanations for this difference. Differences between parent and child ratings of internalizing symptoms are also explored.

Keywords

autism spectrum disorders anxiety depression family functioning adolescents

Individuals with autism spectrum disorder (ASD) are at risk for co-occurring psychiatric conditions, with 40% or more meeting diagnostic criteria for an anxiety disorder (Strang et al., 2011; S. W. White, Oswald, Ollendick, & Scahill, 2009), as many as 34% experiencing clinical depression (Mayes, Calhoun, Murray, Ahuja, & Smith, 2011; Strang et al., 2011), and about 30% displaying externalizing disorders (Mayes, Calhoun, Mayes, & Militoris, 2012; Simonoff et al., 2008). Internalizing symptoms can occur in individuals with ASD at various ages and levels of cognitive and social functioning, including those with and without intellectual disability (S. W. White et al., 2009).

A family systems perspective suggests bidirectional effects between children’s co-occurring conditions and parent emotional experiences (Eisenhower, Baker, & Blacher, 2005; Hastings, Daley, Burns, & Beck, 2006), reflecting ongoing, dynamic interactions among the child, parent, and other family members. In the general adolescent population, anxiety and depression have been linked to negative family functioning, impaired family cohesion, and familial dissatisfaction (E. K. Hughes & Gullone, 2008; Tamplin & Gooyer, 2001); moreover, conflictual family relationships predict greater adolescent depressive symptoms such as sadness, irritability, and feelings of guilt/worthlessness (Sheeber, Hops, Alpert, Davis, & Andrews, 1997). Each of these studies examines a unidirectional relationship between family functioning and child symptomatology, but together they suggest the importance of understanding the bidirectional relationship between the two. Also, family members are more likely to report dysfunction than neutral observers during observations of parent/child interactions (Sheeber & Sorensen, 1998), suggesting that perceptions of family difficulties may differ depending on the informant (A. A. Hughes, Hedtke, & Kendall, 2008). Furthermore, preliminary evidence similarly suggests that depression does not always predict poor family relationships and may be influenced by other factors such as familial support (Sheeber et al., 1997). In the general adolescent literature, researchers have begun to use more comprehensive approaches to capture the bidirectional relationship through longitudinal designs (Cuffe, McKeown, Addy, & Garrison, 2005), but there is still a need for further research in this area (Bögels & Brechman-Toussaint, 2006).

Elucidating the nature of these relationships in ASD has become a priority because of the renewed emphasis in research and clinical practice on family-centered and integrated care (Interagency Autism Coordinating Committee, 2011). Associations between family functioning and child psychopathology have been reported in the ASD literature, although certain child characteristics (e.g., ASD symptoms, externalizing behavior) have been studied more extensively than others (e.g., internalizing symptoms; S. W. White et al., 2009). Given the nature of ASD symptomatology and high rate of associated psychiatric conditions, it is not surprising that families of children with ASD report heightened stress and reduced family functioning compared with typical families (Higgins, 2005; Sikora et al., 2013). Indeed, the negative relationship between ASD symptoms (e.g., social interaction problems, repetitive behavior) and family functioning has been extensively documented (Higgins, 2005; Zablotsky, Anderson, & Law, 2012). Other associated features of ASD also contribute to parents’ stress, anxiety, depression, and reduced feelings of parent competence, particularly externalizing problem behavior (e.g., disruption, aggression, hyperactivity, impulsivity; Firth & Dryer, 2013; Hastings et al., 2005; Herring et al., 2006).

Although children with ASD demonstrate high rates of internalizing symptoms (e.g., anxiety, depression), the extant literature on the relationships among internalizing symptoms and family functioning is somewhat limited. Findings suggesting a relationship among child internalizing symptoms, increased parenting stress, and lower family functioning (Rao & Beidel, 2009) did not include an evaluation of how these child and family factors influenced each other. In their study on depression and social relationships, Pouw, Rieffe, Stockmann, and Gadow (2013) found that core characteristics of ASD, particularly social avoidance, were associated with fewer depressive symptoms in boys with ASD, whereas the opposite was true of their typically developing counterparts; however, the authors did not evaluate these effects in family relationships. Direct associations among internalizing symptoms and family outcomes were well documented in a study of preadolescents and young teenagers (Kim, Szatmari, Bryson, Streiner, & Wilson, 2000). In this sample, anxiety and depression were associated with negative family outcomes such as parent–child relationships and limited ability to participate in social activities, with depressed mood having more of a negative impact than anxiety. While the use of correlational analyses was a strength of this article, the authors did not control for externalizing behavior, which was also quite high in this group.

To summarize, we have preliminary evidence of an association between internalizing symptoms in children with ASD and family functioning, but it is unclear whether this association would be upheld if externalizing symptoms were taken into account. In addition, bidirectional effects between child and family factors have not been well researched. Other factors further complicate our understanding of the association between internalizing symptoms and family outcomes. First, emerging research suggests that children with ASD may cope with depressive symptoms in ways that differentially affect social relationships, as compared with children without ASD (Pouw et al., 2013). Second, most existing research on the ASD population has simply measured child symptoms without examining social functioning, although some notable exceptions exist (e.g., Kim et al., 2000; Pouw et al., 2013).

Overall, little research on individuals with ASD has accounted for the dynamic processes inherent in family systems, and the available findings are limited by (a) statistical analyses that do not directly evaluate the relationships among internalizing symptoms and family functioning, and (b) inconsistent findings. The well-documented findings regarding the negative impact of internalizing symptoms on family functioning in clinical populations without ASD suggest that similar patterns may be present, but not yet specifically assessed in individuals with ASD. The current study evaluates the degree to which child internalizing behavior, as reported by multiple informants, predicts parent ratings of that child’s impact on the family. Specifically, we conducted a secondary analysis of an existing data set that included multiple measures of child and family functioning to explore the predictive value of child report of internalizing symptoms on family impact. Given the inherent limitations of a secondary analysis, we regard our study as a preliminary extension of research on internalizing symptoms in children with ASD and average-range cognitive functioning to family impact, as a step toward more comprehensively understanding this complicated relationship. This study enriches our understanding of this relationship by examining bidirectional effects, including reports from different informants, and controlling for other co-occurring characteristics (e.g., IQ, externalizing symptoms) that are known to affect family functioning.

Method

Participants

Participants (N = 69) were youth with ASD who met the following criteria: (a) age of 11 to 19 years, (b) diagnosis of ASD confirmed by the Social Communication Questionnaire (Rutter, Bailey, & Lord, 2003) and the Autism Diagnostic Observation Schedule (Lord, Rutter, DiLavore, & Risi, 2002), with (c) IQ ≥ 70 as assessed by the Stanford–Binet Intelligence Scales, Fifth Edition (SB-5; Roid, 2003), (d) placement in general education ≥80% of the school day, and (e) attendance in Grades 6 through 12. As shown in Table 1, the sample was mostly male, consistent with the male:female ratio in the ASD population. Most participants had average IQ scores (M = 93.36, SD = 15), were Caucasian, and lived in two-parent homes.

Table 1.

Participant and Family Characteristics.

Characteristic	Participants^a (n)	% of respondents	% of total sample
Sex
Male	58	85.3	84.1
Female	10	14.7	14.5
Not reported	1	NA	1.4
Race/ethnicity
African American/Black	8	11.9	11.6
Asian	13	19.4	18.8
Caucasian	30	44.8	43.5
Hispanic/Latino	7	10.4	10.1
Other	9	13.4	13.0
Not reported	2	NA	2.9
Marital status of primary caregiver
Married	30	76.9	43.5
Divorced or separated	7	17.9	10.1
Widowed	2	5.1	2.9
Not reported	30	NA	43.5
Caregivers in the home
Biological mother and father	24	61.5	34.8
Biological mother and stepfather	4	10.3	5.8
Biological mother only	6	15.4	8.7
Biological father and stepmother	1	2.6	1.4
Biological father only	2	5.1	2.9
Adoptive parents	2	5.1	2.9
Not reported	30	NA	43.5
Total number of children in household
0	1	2.6	1.4
1	8	20.5	11.6
2	20	51.3	29.0
3	6	15.4	8.7
4 or more	4	10.3	5.8
Not reported	30	NA	43.5
Family yearly income level
<US$20,000	1	2.6	1.4
US$20,000–US$40,000	3	7.8	4.3
US$40,001–US$60,000	4	10.5	5.8
US$60,001–US$80,000	1	2.6	1.4
US$80,001–US$100,000	10	26.3	14.5
US$100,001–US$150,000	8	21.1	11.6
US$150,001–US$200,000	6	15.8	8.7
>US$200,000	5	13.2	7.4
Not reported	31	NA	44.9

n = 69.

Procedures

The current project is a secondary analysis of data collected for a large multisite randomized controlled trial (Autism Intervention Research on Behavioral Health [AIR-B] Network) of school-based social skills interventions at two sites: Universtiy of California, Los Angeles and the University of Washington. Participants were recruited from local school districts. The research team worked with the school district administration to recruit adolescents with ASD and average-range IQs. Study coordinators met with school administration and ASD program faculty, and provided information about the study. Schools interested in participating in the study were given study fliers and consent forms. Once signed consent forms were returned, graduate student research assistants determined whether or not the student met criteria to participate in the study. All participants signed written assent and their parents signed written consent forms. Once consent was obtained, participants completed the second editions of the Behavior Assessment System for Children (BASC-2), the Children’s Depression Inventory (CDI-2), and the Multidimensional Anxiety Scale for Children (MASC-2) as part of their baseline assessment. Graduate student researchers facilitated the assessment, checking for comprehension and reading the questions aloud if necessary. Parents completed the Demographic questionnaire, Family Impact Questionnaire (FIQ), and BASC as part of this initial assessment.

Measures

Demographic questionnaire

The demographic questionnaire included child variables (e.g., age, gender, ethnicity), family variables (e.g., composition, income level, education, primary language), and geographic variables (e.g., site, school).

SB-5

The SB-5 is a standardized, direct observation of a person’s verbal and nonverbal cognitive skills across five factors: Fluid Reasoning, Knowledge, Quantitative Reasoning, Visual-Spatial Processing, and Working Memory. It has very strong internal consistency and reliability for the subtests (.84–.89); Factor Indexes (0.90–0.92); and the Nonverbal (0.95), Verbal (0.96), and Full Scale (0.98) IQs. Because it yields a nonverbal score and covers a wide age range, the SB-5 is commonly used in longitudinal studies of children with ASD. The SB-5 was included to characterize the sample and as a covariate in the regression analyses. It was administered at baseline in the participants’ schools, by doctoral-level graduate students.

FIQ

The FIQ is a 50-item, Likert-type scale questionnaire (Donenberg & Baker, 1993). Parents endorse statements related to the impact of their child’s behaviors on their family as compared with other children of the same age. Subscales include the following: Feelings and Attitudes About the Child, Impact on Social Life, Financial Impact, Impact on Marriage, and Impact on Siblings. It has a test–retest reliability of .43 to .71. The FIQ involves less of a family systems-based approach than some other measures of family functioning. However, because it takes into account these multiple aspects of family relationships and because it has been validated for parents of children with ASD (Donenberg & Baker, 1993), the FIQ was chosen as a good representation of family functioning for the purposes of this study. Scores on the FIQ correlate highly with other well-established measures of parenting stress such as the Parenting Stress Index (Abidin, 1995). Higher summary scores on the FIQ are indicative of negative impact on the family.

CDI-2: Self-Report–Short Version (CDI-2: SR(S))

The CDI-2: SR(S) is a 12-item screening measure that was used as the primary measure of depressive symptoms (Kovacs, 2011). Psychometric properties indicate it is reliable, with Cronbach’s alpha values of .67 to .91 and excellent test–retest reliability; a total score yielded by the short version is comparable with that produced by the full-length version. Although not validated in individuals with ASD, the CDI-2 has been successfully used in individuals with ASD without intellectual disability (Stewart, Barnard, Pearson, Hasan, & O’Brien, 2006).

MASC-2

The MASC-2 (March, 2013) is a 50-item self-report measure that assesses symptoms of anxiety in children aged 8 to 19 years. A total summary score is provided, in addition to the following subscale scores: Separation Anxiety/Phobias, Generalized Anxiety Disorder, Social Anxiety, Obsessions and Compulsions, Physical Symptoms, and Harm Avoidance. It demonstrates good internal consistency (.89–.92) and test–retest reliability (March, 2013). A systematic review indicated that the MASC-2 is appropriate for use in children with ASD who have communicative language (Lecavalier et al., 2014).

BASC-2

The BASC-2 (Reynolds & Kamphaus, 2004) is a norm-referenced, multidimensional rating scale. Symptoms of externalizing behavior (aggression, hyperactivity) and internalizing behavior (anxiety, depression, somatization) are calculated as both individual subscales and composites. Subscales and composite scores have parent-report reliability coefficients of .80 to .97 and self-report reliability coefficients of .90 to .97 (Reynolds, 2004). The parent-report form (BASC-2-P) was used to obtain externalizing and internalizing behavior scores only (there is no self-report externalizing behavior composite). Parent-reported externalizing behavior composites were used as the primary measure of externalizing behavior. The adolescents’ self-reported internalizing behavior scores (BASC-2-S) were used for parent/child correspondence checks. Preliminary research suggests that the BASC-2 effectively measures adaptive behavior and associated symptoms in individuals with ASD, particularly those with average IQs (Hass, Brown, Brady, & Johnson, 2012).

Data Analysis

Multiple imputation (MI) was used to replace missing values in the outcome and other measures of interest (Little & Rubin, 1987; I. R. White, Royston, & Wood, 2011). MI is a simulation-based approach to deal with incomplete data using plausible values to replace missing data. It differs from single imputation because MI replaces each missing value multiple times, preserving the error structure so that valid inferences can be made (Harel, 2009). MI uses a sequential regression multivariate imputation algorithm to generate multiple copies of the data set where each data set contains different estimates of the missing values. This algorithm was implemented using the MICE (Multiple Imputation by Chain Equations) package (van Buuren & Groothuis-Oudshoorn, 2011) for R CRAN Version 3.0.2 (R Core Team, 2014) to generate 30 imputed data sets. Results did not differ if fewer or more imputations were used (e.g., M = 20 or M = 250), indicating that the number of imputations performed was acceptable for generating plausible values to replace the amount of missing data. Variables used for the imputations included the BASC-2-P externalizing and internalizing behavior composites (19 of 69 missing), CDI-2: SR(S) total score (8 of 69 missing), MASC-2 summary score (9 of 69 missing), FIQ summary score (11 of 69 missing), and the SB-5 Full Scale IQ score (2 of 69 missing). Although the amount of missing data is not ideal, it was judged to be missing at random because it was mainly due to a decision to simplify the assessment protocol midway through the study, and because there was no evidence of differences between participants with and without missing data. Moreover, the use of MI provided robust results, as indicated by the comparable results obtained with varying numbers of imputations.

The imputed data sets were standardized within each imputation, and a linear regression was fitted for each standardized imputed data set to obtain the standardized estimated coefficients. The standardized estimated coefficients and corresponding standard errors reported in the “Results” section were calculated based on Rubin’s rule (Little & Rubin, 1987) for combining results of identical analyses performed on each of the 30 imputed data sets. The pooled estimated R² was based on the method proposed by Harel (2009), and the test for nested models was based on Meng and Rubin (1992). A hierarchical regression model was used to test the hypothesis that the child’s internalizing symptoms predicted family functioning over and above the child’s cognitive ability and externalizing behavior (Aim 1). In the first step of the analysis, the FIQ summary score was regressed onto cognitive functioning, as measured by the Full Scale IQ of the SB-5. In the second step, externalizing behavior as measured by the parent-rated BASC-2-P externalizing behavior composite was entered. In the third and final step, child-rated depression and anxiety—CDI-2:SR(S) total score and MASC-2 summary score—were entered into the model. Based on the feedback from reviewers, ASD symptom severity scores were also included in the model, but they were nonsignificant and as such are not reported.

We were also interested in the correspondence between adolescents’ self-ratings of internalizing behavior and parent’s ratings of their child’s internalizing behaviors. This was evaluated through correlational analyses using the imputed data sets as well (Aim 2). Correlations were run between the BASC-2-S internalizing summary score and the BASC-2-P internalizing summary score, and consistency of child report was examined through correlations between (a) BASC-2-S and the CDI-2:SR(S) total score, and (b) BASC-2-S and MASC-2 summary score. The correlations were combined from the sample correlation coefficients computed from the 30 imputed data sets by using the Fisher’s z transformation. A final pooled estimate of the correlation coefficient was then generated from the inverse Fisher’s z transformation.

Results

Child Symptoms and Family Functioning

The overall model of cognitive functioning, externalizing behaviors, and internalizing behaviors predicted 89% of the variance in negative family impact, R² = .885, p < .01 (see Table 2). Adding internalizing symptoms (anxiety and depression) at the third step of the model added significant predictive value, ΔR² = .013, p = .0261, to the model over and above cognitive functioning (SB-5 Full Scale IQ score entered at the first step) and externalizing behaviors (BASC-2-P externalizing behavior composite entered at the second step). Internalizing symptoms account for an additional 1.3% of the variance in negative family impact. In the overall model, anxiety, as assessed by the MASC-2 summary score, did not contribute uniquely to the model, β = .048, p = .336, ns; however, increased depressive symptoms as assessed by the CDI-2:SR(S) total score significantly predicted better family functioning, β = −.374, p = .003 (see Table 2). The negative correlation between less depression and family functioning is upheld, even without controlling for cognitive ability, externalizing behavior, and anxiety (r = −.136).

Table 2.

Contributions of Child Variables on Family Functioning at Each Level of the Hierarchical Model (N = 69).

Predictor variable and measure	ΔR²	B ^a
Step 1
IQ	.85***
SB-5		−.025
Step 2
Externalizing behavior	.03***
BASC-2-P		.308*
Step 3: Internalizing symptoms	.01*
MASC-2		.048
CDI-2:SR(S)		−.374**
Total R²	.89***

Note. SB-5 = Stanford–Binet Intelligence Scales, Fifth Edition, BASC-2-P = Behavior Assessment System for Children, Second Edition–Parent-Report Form, MASC-2 = Multidimensional Anxiety Scale for Children, Second Edition, CDI-2:SR(S) = Children’s Depression Inventory, Second Edition: Self-Report–Short Version.

β from final model.

p < .05. **p < .01. ***p < .001.

Correspondence of Child and Parent Report

Adolescent and parent ratings of overall internalizing behavior on the BASC-2 did not correlate (see Table 3). Parents rated significantly more internalizing symptoms (M = 56.33; see Table 4) than did adolescents (M = 48.0). We followed up these findings with additional analyses to determine whether adolescents rated their internalizing symptoms similarly across measures. No significant relationship was demonstrated between adolescent self-report on the BASC-2 and on the MASC-2. However, the BASC-2 and the CDI-2:SR(S) were significantly correlated, whether or not the MASC-2 was included in the model.

Table 3.

Correlations Among Parent- and Self-Reports of Adolescents’ Internalizing Symptoms.

Correlation with BASC-2-Self Internalizing Composite (IC)	r	p
BASC-2-Parent IC	−.048	.693
MASC-2 (self)	.162	.184
CDI-2 (self)	.331	.005

Note. BASC-2 = Behavior Assessment System for Children, Second Edition, MASC-2 = Multidimensional Anxiety Scale for Children, Second Edition, CDI-2 = Children’s Depression Inventory, Second Edition.

Table 4.

Parent- and Self-Report Measures.

Measure	M	SD
Parent-report measure
BASC-2-Parent IC	56.3	9.8
Self-report measures
BASC-2-Self IC	48.0	8.3
MASC-2	56.0	11.3
CDI-2	48.5	9.3

Discussion

Individuals with ASD often report co-occurring internalizing disorders such as anxiety and depression. In accordance with previous research, externalizing behaviors and IQ predicted a large amount of the variance in parent-perceived negative family impact. Internalizing behaviors also predicted family impact for depression but not anxiety. The directionality of this relationship can be inferred due to the nature of the FIQ as a measure of how specific child factors influence different aspects of family functioning (i.e., asking parents to think causally about the impact of their child’s symptoms on their family). Although it is somewhat counterintuitive that adolescent-endorsed depressive symptoms (e.g., sadness, anhedonia, guilt/worthlessness) predicted less parent-rated negative impact, it is consistent with previous research demonstrating a lack of negative impact on family relationships (Sheeber et al., 1997) and an inverse relationship between social avoidance and depressive symptoms in adolescents with ASD (Pouw et al., 2013).

In light of the conflicting evidence regarding the accuracy of self-report measures in adolescents with high-functioning ASD (Blakeley-Smith, Reaven, Ridge, & Hepburn, 2012; Mazefsky, Kao, & Oswald, 2011), we examined the relationship between child and parent report of internalizing symptoms to help inform our interpretation of the results. Adolescents are likely to perceive family functioning differently from their parents and may be naive to parent perceptions of how their behavior affects the family. Had we based the analyses upon adolescent report of family outcomes, the directionality of the effect might have also changed. For example, adolescent internalizing symptoms have been found to be more strongly related to poor family functioning per self-report than to semiobjective observer ratings (Millikan, Wamboldt, & Bihun, 2002). The effects of internalizing symptoms may thus not be well captured on measures completed by observers, rather than the child himself or herself. This is particularly salient for internalizing symptoms, which by definition are not always associated with overt, observable behavior.

Indeed, we found that parents and adolescents rated internalizing symptoms differently, with parents endorsing significantly more depressive symptoms than their children. Several potential explanations could explain this finding. First, parents may have better insight into their children’s experiences, especially considering the difficulties with emotional awareness that accompany ASD (Baron-Cohen, 2002). However, similar disagreements between parent and adolescent ratings of internalizing symptoms are noted in the general population (Youngstrom, Loeber, & Stouthamer-Loeber, 2000), suggesting that this may not be an ASD-specific issue. Furthermore, in this study, adolescents’ report of depression correlated across measures, suggesting they gave consistent ratings of their internalizing behavior, regardless of whether these ratings accurately reflected their emotional state. Second, there is some overlap between depressive symptoms and ASD symptoms, such as social withdrawal, low motivation, and atypical speech (Stewart et al., 2006), which could confound what symptoms are attributed to depression versus ASD core symptomatology, particularly for children with limited communication (Mayes et al., 2011; Stewart et al., 2006).

There are several possible reasons why a positive relationship between depression and family functioning could exist. These are hypotheses that we present for future study, rather than conclusions drawn from the current findings. In the general population, depression is often associated with psychomotor slowing, flat affect, and a tendency not to engage in previously enjoyed activities (American Psychiatric Association [APA], 2013). In ASD, depression could temper externalizing and stereotyped behaviors, which may be seen as a positive for families. Depressed adolescents may be less likely to engage in motor mannerisms, sensory-seeking behaviors, or circumscribed interests, which on the surface could effectively mimic decreases in core ASD symptoms. Although there is no literature to our knowledge examining this particular question, it could be addressed in future research by adding standardized rating scales of stereotyped and repetitive behaviors. Parents may also interpret the moodiness and irritability that accompany adolescent depression as related to puberty, which could be a normalizing experience for parents, who may be relieved to feel as though their child is going through the “normal” trials of the teenage years. Finally, informant choice on the FIQ could be relevant. Parents who endorsed lower child impact may see their child as more socially involved with the family as a whole. Although this may be positive from the parents’ perspective, social involvement may be unwanted by the child and could prevent the adolescent from coping with depressive symptoms in an effective way.

Despite the significant findings with respect to depression, similar trends did not occur for symptoms of anxiety, which raises the question of why these two aspects of internalizing symptoms differ. Part of this phenomenon likely relates to the presentation of anxiety in individuals with ASD, who are less likely to demonstrate “classic” anxiety symptoms and more likely to exhibit different patterns of anxiety (Kerns & Kendall, 2012; Kerns et al., 2014; Weisbrot, Gadow, DeVincent, & Pomeroy, 2005; S. W. White et al., 2009). Anxiety in ASD may differ both in terms of topography and content (Kerns, & Kendall, 2012; Wood & Gadow, 2010). For example, many individuals with ASD become anxious around change and novelty, and rather than exhibiting a fear response, they often show behavioral rigidity or an increase in stereotyped behaviors. As such, the overlap between what are perceived as ASD symptoms and anxiety symptoms is considerable (Hallett, Ronald, & Happé, 2009; Hallett, Ronald, Rijsdijk, & Happé, 2012). Moreover, children and adolescents with ASD and anxiety are likely to exhibit emotional outbursts and aggressive behaviors in response to anxiety (S. W. White et al., 2009), behaviors that are typically considered to be reflective of externalizing problems rather than internalizing problems. Given the “ASD-specific” nature of anxiety in this population, existing anxiety measures may not map well onto anxiety symptoms in individuals with ASD.

Finally, families of children with ASD often adapt their daily routines to accommodate for anxiety and to avoid situations that are likely to provoke distress. By the teenage years, families may have become overadapted, such that they are either extremely effective at preventing the expression of anxiety symptoms or have normalized the anxiety that adolescents do exhibit. Furthermore, adolescent self-report ratings were consistent between measures for depression but not anxiety, and the measures used could affect this correspondence. The MASC-2 is more nuanced and concrete in the sense that it provides more specific examples across anxiety domains than the BASC-2 (e.g., “I get scared riding in the car or on the bus” vs. “is fearful”). As the BASC-2 parent version was included in the family functioning analyses, we may have been tapping into anxiety symptoms that do not accurately capture anxiety in this population and missing out on others. A similar divergence between parent and child ratings has been reported elsewhere (S. W. White & Roberson-Nay, 2009).

Limitations and Future Directions

The findings of this study are qualified by certain limitations. Participants represented a small research sample of individuals with ASD who had average-range IQ and placement in general education, so our findings may not be generalizable to the larger ASD population. Given the diverging presentation of internalizing symptoms in individuals with ASD across levels of cognitive functioning (Ghaziuddin, Ghaziuddin, & Greden, 2002; Mayes et al., 2011), it would be important to evaluate internalizing behavior and family functioning in younger children and individuals with comorbid intellectual disability. Regarding our statistical methodology, the small sample size limited the power to detect smaller effect sizes once our analyses controlled for externalizing behaviors and IQ, which have a significant and large effect. Missing data (particularly for the BASC-2-P) was another limitation, although our use of MI to replace missing data mitigated this issue. Graham and Schafer (1999) showed that with sample sizes as small as n = 50, MI performs very well even with large multiple regression models (as large as 18 predictors). Although statistically significant, the effect size for internalizing and, to some extent, externalizing behavior was small, which may not necessarily translate into clinically meaningful information. However, we believe these preliminary findings suggest that this is a worthwhile topic for further exploration.

As previously mentioned, the potentially limited insight of adolescents with ASD is of concern, both regarding confidence in self-report of internalizing symptoms and when comparing this self-report with parent report of family functioning. Furthermore, the findings are potentially confounded by symptom overlap, specifically the topographical similarity between ASD and symptoms of anxiety and depression. Some research suggests that the MASC-2 and CDI-2 are not well suited for use in individuals with ASD (Mazefsky et al., 2011). Others suggest that while the MASC-2 can be inaccurate, it is among the best measures of anxiety available, especially in high-functioning individuals (Lecavalier et al., 2014). Even in self-report measures that may be promising (e.g., Revised Children’s Manifest Anxiety Scale; Mazefsky et al., 2011), clinicians are still encouraged to interpret the results with caution. A possible alternative is to use semistructured interviews such as the Anxiety Disorders Interview Schedule (ADIS) and Kiddie- Schedule for Affective Disorders and Schizophrenia (SADS). Although time-consuming, these interviews have the advantage of allowing for follow-up questions and clinical judgment and can be modified to include questions related to the specific presentation of anxiety in ASD (as described in Kerns et al., 2014). Finally, studies that compare individuals with ASD who do and do not present with associated anxiety and depression would further inform our understanding of internalizing symptoms in ASD.

Adolescent report of family functioning should be collected in future studies, as the literature suggests that ratings of family functioning vary widely by reporter. Multiple perspectives would support bidirectional analyses of the relationship between adolescent-rated depression and better family functioning. Our previous hypothesis regarding how moodiness might normalize the teenage experience for parents is one avenue to explore. The potential role of depression on restricted and repetitive behavior is also of interest, given that, to our knowledge, it has yet to be examined.

We did not fully explore other family variables that might relate to family functioning and internalizing symptoms. For instance, we did not measure parent psychopathology, despite evidence that parent mental health predicts a significant family dysfunction, even when controlling for child characteristics (Frye & Garber, 2005; A. A. Hughes et al., 2008; Johnson, Frenn, Feetham, & Simpson, 2011). This would be especially important for adolescents with internalizing disorders, as mothers of children with ASD and co-occurring psychiatric disorders have a higher risk of depression and lower quality of life than mothers of children with just ASD (Zablotsky et al., 2012). Other variables of interest include the impact on siblings, who might be at risk of depressive and anxiety symptoms (Orsmond & Seltzer, 2009). Finally, although we explored dynamics within the parent/child relationship, many other important pathways are inherent within the family system (e.g., mother/child, father/child, sibling, spousal) that affect the family as a whole. The notable variability in family structure (e.g., marital status, number of children) in our sample also raises possibilities for future research, as family characteristics and resilience have been linked to parent stress and coping (Gray & Holden, 1992; Weiss et al., 2013). Including these factors in future analyses would help create a cohesive picture of family systems that include individuals with ASD.

Clinical Recommendations

In light of the complications to assessment and treatment of internalizing symptoms raised in this article, we offer clinical recommendations for those working with individuals with ASD. These suggestions are not derived solely from the findings presented in this article; rather, they integrate our findings with findings reported by others (e.g., Kerns & Kendall, 2012; Stewart et al., 2006; Vickerstaff et al., 2007; White et al., 2010).

Assessment considerations

In initial evaluations, consider the use of a gradated assessment process.

(a) Gather information directly from the individual with ASD, particularly for persons with at least average-range IQ and language skills. The overlap in internalizing symptoms and ASD symptoms may result in observer over- or underreport of anxiety and depression.

(b) When indicated (i.e., via clinical judgment or meeting screener cutoffs), use semistructured, standardized interviews (e.g., ADIS, Kiddie-SADS). Although time-consuming, interviews allow for follow-up questions to disentangle anxiety and depressive symptoms from ASD symptomatology. They also allow for probing of “atypical” anxiety or depression topographies that are not well captured on standardized parent- or self-report rating scales.

When talking with families, assess typical precursors and reactions to the individual’s behavior. Identifying the functions of certain behaviors (e.g., withdrawal, repetitive behavior) may help indicate internalizing symptoms or ASD symptoms and can help inform interventions.

Use visual supports (e.g., feelings thermometers, pictures of emotion faces) for assessing anxiety and depression to help address difficulties with emotion recognition and abstract concepts (particularly for those who are young and those with delayed cognitive or language skills).

Treatment considerations

Encourage productive communication between parents and children. Counsel parents to explicitly ask about potential internalizing symptoms on an ongoing basis and in different ways (e.g., open-ended and yes/no questions, with visual supports).

Traditional cognitive-behavioral therapy (CBT) is still indicated for individuals with ASD who are at least school age and have at least average-range IQ (Reaven, Blakeley-Smith, Culhane-Shelburne, & Hepburn, 2012; Szatmari & McConnell, 2011). Clinicians should help families recognize any nontraditional presentations of internalizing symptoms in their children.

Target behaviors with child or family interference (e.g., those that cause distress, family avoidance of activities, significant accommodation, academic or social problems). Evaluate interference and preference for individuals with ASD, given the potentially counterintuitive effects of internalizing symptoms on child and family functioning. The presence of some symptoms may exert a positive effect on the child’s mood or the family’s ability to get along. Conversely, what families may think is enjoyable for their children (e.g., participating in activities outside the house) may actually exacerbate anxiety and/or feelings of incompetence.

Footnotes

Acknowledgements

We express our gratitude toward the students with autism spectrum disorder (ASD), their families, and their schools for their participation in this study.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Grant R01HD65291 from the National Institutes of Health; analyses supported by Grant UA3 MC 11055 AIR-B from the Maternal and Child Health Research Program, Maternal and Child Health Bureau (Combating Autism Act Initiative), Health Resources and Services Administration, Department of Health and Human Services.

References

Abidin

(1995). Manual for the Parenting Stress Index. Odessa, FL: Psychological Assessment Resources.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.

Baron-Cohen

(2002). Is Asperger syndrome necessarily viewed as a disability? Focus on Autism and Other Developmental Disabilities, 17, 186–191.

Blakeley-Smith

Reaven

Ridge

Hepburn

(2012). Parent-child agreement of anxiety symptoms in youth with autism spectrum disorders. Research in Autism Spectrum Disorders, 6, 707–716.

Bögels

S. M.

Brechman-Toussaint

M. L.

(2006). Family issues in child anxiety: Attachment, family functioning, parental rearing and beliefs. Clinical Psychology Review, 26, 834–856.

Cuffe

S. P.

McKeown

R. E.

Addy

C. L.

Garrison

C. Z.

(2005). Family and psychosocial risk factors in a longitudinal epidemiological study of adolescents. Journal of the American Academy of Child & Adolescent Psychiatry, 44, 121–129.

Donenberg

Baker

B. L.

(1993). The impact of young children with externalizing behaviors on their families. Journal of Abnormal Child Psychology, 21, 179–198.

Eisenhower

Baker

B. L.

Blacher

(2005). Preschool children with intellectual disability: Syndrome specificity, behaviour problems, and maternal well-being. Journal of Intellectual Disability Research, 49, 657–671.

Firth

Dryer

(2013). The predictors of distress in parents of children with autism spectrum disorder. Journal of Intellectual & Developmental Disability, 38, 163–171. doi:10.3109/13668250.2013.773964

10.

Frye

A. A.

Garber

(2005). The relations among maternal depression, maternal criticism, and adolescents’ externalizing and internalizing symptoms. Journal of Abnormal Child Psychology, 33, 1–11.

11.

Ghaziuddin

Greden

(2002). Depression in persons with autism: Implications for research and clinical care. Journal of Autism and Developmental Disorders, 32, 299–306.

12.

Graham

J. W.

Schafer

J. L.

(1999). On the performance of multiple imputation for multivariate data with small sample size. Statistical Strategies for Small Sample Research, 50, 1–27.

13.

Gray

D. E.

Holden

W. J.

(1992). Psycho-social well-being among the parents of children with autism. Australia and New Zealand Journal of Developmental Disabilities, 18(2), 83–93.

14.

Hallett

Ronald

Happé

(2009). Investigating the association between autistic-like and internalizing traits in a community-based twin sample. Journal of the American Academy of Child & Adolescent Psychiatry, 48, 618–627.

15.

Hallett

Ronald

Rijsdijk

Happé

(2012). Disentangling the associations between autistic-like and internalizing traits: A community based twin study. Journal of Abnormal Child Psychology, 40, 815–827.

16.

Harel

(2009). The estimation of R² and adjusted R² in incomplete data sets using multiple imputation. Journal of Applied Statistics, 36, 1109–1118.

17.

Hass

M. R.

Brown

R. S.

Brady

Johnson

D. B.

(2012). Validating the BASC-TRS for use with children and adolescents with an educational diagnosis of autism. Remedial and Special Education, 33, 173–183.

18.

Hastings

R. P.

Daley

Burns

Beck

(2006). Maternal distress and expressed emotion: Cross-sectional and longitudinal relationships with behavior problems of children with intellectual disabilities. American Journal on Mental Retardation, 111, 48–61.

19.

Hastings

R. P.

Kovshoff

Ward

N. J.

Degli Espinosa

Brown

Remington

(2005). Systems analysis of stress and positive perceptions in mothers and fathers of pre-school children with autism. Journal of Autism and Developmental Disorders, 35, 635–644.

20.

Herring

Gray

Taffe

Tonge

Sweeney

Einfeld

(2006). Behaviour and emotional problems in toddlers with pervasive developmental disorders and developmental delay: Associations with parental mental health and family functioning. Journal of Intellectual Disability Research, 50, 874–882.

21.

Higgins

D. J.

(2005). Factors associated with functioning style and coping strategies of families with a child with an autism spectrum disorder. Autism, 9, 125–137. doi:10.1177/1362361305051403

22.

Hughes

A. A.

Hedtke

K. A.

Kendall

P. C.

(2008). Family functioning in families of children with anxiety disorders. Journal of Family Psychology, 22, 325–328. doi:10.1037/0893-3200.22.2.325

23.

Hughes

E. K.

Gullone

(2008). Internalizing symptoms and disorders in families of adolescents: A review of family systems literature. Clinical Psychology Review, 28, 92–117.

24.

Interagency Autism Coordinating Committee. (2011). 2011 IACC strategic plan for autism spectrum disorder research. Retrieved from http://iacc.hhs.gov/strategic-plan/2011/index.shtml

25.

Johnson

Frenn

Feetham

Simpson

(2011). Autism spectrum disorder: Parenting stress, family functioning and health-related quality of life. Families, Systems, & Health, 29, 232–252.

26.

Kerns

C. M.

Kendall

P. C.

(2012). The presentation and classification of anxiety in autism spectrum disorder. Clinical Psychology: Science and Practice, 19, 323–347.

27.

Kerns

C. M.

Kendall

P. C.

Berry

Souders

M. C.

Franklin

M. E.

Schultz

R. T.

. . . Herrington

(2014). Traditional and atypical presentations of anxiety in youth with autism spectrum disorder. Journal of Autism and Developmental Disorders, 44, 2851–2861.

28.

Kim

J. A.

Szatmari

Bryson

S. E.

Streiner

D. L.

Wilson

F. J.

(2000). The prevalence of anxiety and mood problems among children with autism and Asperger syndrome. Autism, 4, 117–132.

29.

Kovacs

(2011). Children’s Depression Inventory 2: Technical manual (2nd ed.). Toronto, Ontario, Canada: Multi-Health Systems.

30.

Lecavalier

Wood

J. J.

Halladay

A. K.

Jones

N. E.

Aman

M. G.

Cook

E. H.

. . . Scahill

(2014). Measuring anxiety as a treatment endpoint in youth with autism spectrum disorder. Journal of Autism and Developmental Disorders, 44, 1128–1143.

31.

Little

R. J. A.

Rubin

D. B.

(1987). Statistical analysis with missing data. New York, NY: John Wiley.

32.

Lord

Rutter

DiLavore

Risi

(2002). Autism Diagnostic Observation Schedule: ADOS. Los Angeles, CA: Western Psychological Services.

33.

March

J. S.

(2013). Multidimensional anxiety scale for children (2nd ed.). Toronto, Ontario, Canada. Multi-Health Systems.

34.

Mayes

S. D.

Calhoun

S. L.

Mayes

R. D.

Molitoris

(2012). Autism and ADHD: Overlapping and discriminating symptoms. Research in Autism Spectrum Disorders, 6, 277–285.

35.

Mayes

S. D.

Calhoun

S. L.

Murray

M. J.

Ahuja

Smith

L. A.

(2011). Anxiety, depression, and irritability in children with autism relative to other neuropsychiatric disorders and typical development. Research in Autism Spectrum Disorders, 5, 474–485.

36.

Mazefsky

Kao

Oswald

(2011). Preliminary evidence suggesting caution in the use of psychiatric self-report measures with adolescents with high-functioning autism spectrum disorders. Research in Autism Spectrum Disorders, 5, 164–174.

37.

Meng

X. L.

Rubin

D. B.

(1992). Performing likelihood ratio tests with multiply-imputed data sets. Biometrika, 79, 103–111.

38.

Millikan

Wamboldt

M. Z.

Bihun

J. T.

(2002). Perceptions of the family, personality characteristics, and adolescent internalizing symptoms. Journal of the American Academy of Child & Adolescent Psychiatry, 41, 1486-1494. doi: 10.1097/00004583-200212000-00021

39.

Orsmond

G. I.

Seltzer

M. M.

(2009). Adolescent siblings of individuals with an autism spectrum disorder: Testing a diathesis-stress model of sibling well-being. Journal of Autism and Developmental Disorders, 39, 1053–1065. doi:10.1007/s10803-009-0722-7

40.

Pouw

L. B. C.

Rieffe

Stockmann

Gadow

K. D.

(2013). The link between emotion regulation, social functioning, and depression in boys with ASD. Research in Autism Spectrum Disorders, 7, 549–556. doi:10.1016/j.rasd.2013.01.002

41.

R Core Team. (2014). R: A language and environment for statistical computing [software]. Vienna, Austria: R Foundation for Statistical Computing. Available from http://www.r-project.org

42.

Rao

P. A.

Beidel

D. C.

(2009). The impact of children with high-functioning autism on parental stress, sibling adjustment, and family functioning. Behavior Modification, 33, 437–451. doi:10.1177/0145445509336427

43.

Reaven

J. A.

Blakeley-Smith

Culhane-Shelburne

Hepburn

(2012). Group cognitive behavior therapy for children with high-functioning autism spectrum disorders and anxiety: A randomized trial. Journal of Child Psychology and Psychiatry, 53, 410–419.

44.

Reynolds

C. R.

(2004). Behavior assessment system for children. In Weiner

Craighead

(Eds.), Corsini Encyclopedia of Psychology. doi: 10.1002/9780470479216

45.

Reynolds

C. R.

Kamphaus

R. W.

(2004). Behavior assessment system for children. Los Angeles, CA: American Guidance Service.

46.

Roid

G. H.

(2003). Stanford-Binet intelligence scales: Technical manual (5th ed.). Itasca, IL: Riverside.

47.

Rutter

Bailey

Lord

(2003). The Social Communication Questionnaire: Manual. Torrance, CA: Western Psychological Services.

48.

Sheeber

Hops

Alpert

Davis

Andrews

(1997). Family support and conflict: Prospective relations to adolescent depression. Journal of Abnormal Child Psychology, 25, 333–344.

49.

Sheeber

Sorensen

(1998). Family relationships of depressed adolescents: A multimethod assessment. Journal of Clinical Child Psychology, 27, 268–277. doi:10.1207/s15374424jccp2703_4

50.

Sikora

Moran

Orlich

Hall

T. A.

Kovacs

E. A.

Delahaye

J., …

Kuhlthau

(2013). The relationshisp between family functioning and behavior problems in children with autism spectrum disorders. Research in Autism Spectrum Disorders, 7, 307–315. doi:10.1016/j.rasd.2012.09.006

51.

Simonoff

Pickles

Charman

Chandler

Loucas

Baird

(2008). Psychiatric disorders in children with autism spectrum disorders: Prevalence, comorbidity, and associated factors in a population-derived sample. Journal of the American Academy of Child & Adolescent Psychiatry, 47, 921–929. doi:10.1097/CHI.0b013e318179964f

52.

Stewart

M. E.

Barnard

Pearson

Hasan

O’Brien

(2006). Presentation of depression in autism and Asperger syndrome: A review. Autism, 10, 103–116.

53.

Strang

J. F.

Kenworthy

Daniolos

Case

Wills

M. C.

Martin

Wallace

G. L.

(2011). Research in autism spectrum disorders. Research in Autism Spectrum Disorders, 6, 406–412. doi:10.1016/j.rasd.2011.06.015

54.

Szatmari

McConnell

(2011). Anxiety and mood disorders in individuals with autism spectrum disorder. In Amaral

D. G.

Dawson

Geschwind

(Eds.), Autism spectrum disorders (pp. 330–338). New York, NY: Oxford University Press.

55.

Tamplin

Gooyer

(2001). Family functioning in adolescents at high and low risk for major depressive disorder. European Child & Adolescent Psychiatry, 10, 170–179.

56.

van Buuren

Groothuis-Oudshoorn

. (2011). mice: Multivariate imputation by chained equations in R. Journal of Statistical Software, 45(3), 1–67. Retrieved from http://www.jstatsoft.org/v45/i03

57.

Vickerstaff

Heriot

Wong

Lopes

Dossetor

(2007). Intellectual ability, self-perceived social competence, and depressive symptomatology in children with high-functioning autistic spectrum disorders. Journal of Autism and Developmental Disorders, 37, 1647–1664.

58.

Weisbrot

D. M.

Gadow

K. D.

DeVincent

C. J.

Pomeroy

(2005). The presentation of anxiety in children with pervasive developmental disorders. Journal of Child & Adolescent Psychopharmacology, 15, 477–496.

59.

Weiss

J. A.

Robinson

Fung

Tint

Chalmers

Lunsky

(2013). Family hardiness, social support, and self-efficacy in mothers of individuals with autism spectrum disorders. Research in Autism Spectrum Disorders, 7, 1310–1317.

60.

White

I. R.

Royston

Wood

A. M.

(2011). Multiple imputation using chained equations: Issues and guidance for practice. Statistics in Medicine, 30, 377–399. doi:10.1002/sim.4067

61.

White

S. W.

Oswald

Ollendick

Scahill

(2009). Anxiety in children and adolescents with autism spectrum disorders. Clinical Psychology Review, 29, 216–229. doi:10.1016/j.cpr.2009.01.003

62.

White

S. W.

Roberson-Nay

(2009). Anxiety, social deficits, and loneliness in youth with autism spectrum disorders. Journal of Autism and Developmental Disorders, 39, 1006–1013. doi:10.1007/s10803-009-0713-8

63.

White

S. W.

Albano

A. M.

Johnson

C. R.

Kasari

Ollendick

Klin

… Scahill

(2010). Development of a cognitive-behavioral intervention program to treat anxiety and social deficits in teens with high-functioning autism. Clinical Child and Family Psychology Review, 13, 77-90.

64.

Wood

J. J.

Gadow

K. D.

(2010). Exploring the nature and function of anxiety in youth with autism spectrum disorders. Clinical Psychology: Science and Practice, 17, 281–292.

65.

Youngstrom

Loeber

Stouthamer-Loeber

(2000). Patterns and correlates of agreement between parent, teacher, and male adolescent ratings of externalizing and internalizing problems. Journal of Consulting and Clinical Psychology, 68, 1038.

66.

Zablotsky

Anderson

Law

(2012). The association between child autism symptomatology, maternal quality of life, and risk for depression. Journal of Autism and Developmental Disorders, 43, 1946–1955. doi:10.1007/s10803-012-1745-z