The Green Bucket

Anxiety, Social Deficits, and Loneliness in Youth with Autism Spectrum Disorders

(Journal of Autism & Developmental Disorders)

Susan W. White1, 2 Contact Information and Roxann Roberson-Nay1
(1) Department of Psychiatry, Virginia Commonwealth University, Richmond, VA, USA
(2) Present address: Department of Psychology, Virginia Polytechnic Institute and State University, 109 Williams Hall (0436), Blacksburg, VA 24061, USA

Received: 15 September 2008 Accepted: 13 February 2009 Published online: 4 March 2009

Abstract

The purpose of this study was to explore relationships among anxiety, loneliness, and degree of social skill deficit in a sample of youth with autism spectrum disorders (ASD). Participants (N = 20) were between 7 and 14 years of age, verbal, and had low average or higher assessed intelligence (average IQ = 92 ± 14.41). Youth who self-reported elevated levels of anxiety reported greater feelings of social loneliness. Those participants earning above average total anxiety scores reported significantly more loneliness than those with less anxiety (F = 6.60, p < .05). A significant relationship between parent-reported child withdrawal and depression and social disability also was found. Recommendations for assessment of co-occurring psychiatric problems in youth with ASD are offered.

There is growing recognition of the need to evaluate and treat co-occurring psychiatric conditions in youth with autism spectrum disorders (ASD). This group of disorders, which includes autistic disorder, Asperger's disorder, and pervasive developmental disorder-not otherwise specified (PDD-NOS), affects approximately one in every 152 children (Centers for Disease Control 2007). In clinical settings, anxiety-related concerns are among the most common presenting problems for school-age children and adolescents with ASD (Ghaziuddin 2002). Although there have been no large-scale epidemiological studies on the prevalence of anxiety disorders co-occurring in youth with ASD, a recent comprehensive review of studies on anxiety in youth with ASD found that between 11 and 84% of children with ASD experience impairing anxiety (White et al. 2009). Knowing how to accurately evaluate and effectively treat anxiety in this population, therefore, is imperative.

It is unclear at this time why anxiety is such a common occurrence in youth who have autism and related conditions. Possible explanations include structural (Amaral Bauman and Schumann 2003) and/or neurochemical (Apter et al. 1991) disturbances in the brain that predispose people with ASD to experience anxiety. It is also possible that there is a genetic link between ASD and anxiety (Piven and Palmer 1999; Smalley et al. 1995), the mechanisms of which at this time are unknown. Finally, anxiety may be directly or indirectly related to the social skill deficits of ASD. Bellini (2006), for example, found that physiological hyperarousal combined with social skill deficit to predict social anxiety in ASD.

The relationship between social disability and anxiety in youth with ASD, however, is not well understood. It is logical to conjecture that the social disability associated with ASD could lead to anxiety. For instance, higher functioning youth who have an awareness of their social disability may experience anxiety related to misinterpreting social cues and expectations for social failure. Alternatively, anxiety may exacerbate the child's social deficit (Myles et al. 2001a). Co-occurring problems with anxiety, for example, might contribute to social avoidance and limit the child's opportunities to practice appropriate social skills.

It is commonly assumed that individuals with ASD prefer isolation or at least minimal social contact. Although some people may be ‘protected' in a sense by lack of insight into their own social difficulties, many young people with ASD are "acutely aware of their difficulties with social integration" (Attwood 2000, p. 97) and report experiencing more loneliness than their typically developing peers (Bauminger and Kasari 2000). Consequently, children with ASD often self-report a desire for increased peer interactions. Unfortunately, the social difficulties experienced by many youth with ASD-especially involving interactions with same-age peers, may become more difficult as children develop. As both social complexity and social demands rise in middle- and high-school, a child may become acutely aware of his social disability (Schopler and Mesibov 1983; Tantam 2003), which in turn may contribute to the development of secondary mood and anxiety problems (Myles 2003; Myles et al. 2001b; Tantam 2003), especially for higher functioning individuals with ASD. In summary, it is quite possible that anxiety and social disability exert bi-directional mediating effects, which in turn impact loneliness and social isolation.

Despite the frequency with which anxiety problems present in youth who have autism and related conditions, there is limited data on the relationship between symptoms of anxiety and associated social impairment or between anxiety and self-reported feelings of social isolation or loneliness. Likewise, there is a lack of guidance on how to assess for anxiety clinically. Measures created for use with neurotypical children might not adequately capture symptoms of anxiety in youth on the spectrum, especially if anxiety is expressed in unusual ways (e.g., exacerbation of repetitive behavior). Furthermore, youth with ASD may have difficulty with self-report instruments, particularly if they have significant cognitive limitations. Although some studies have found that self-report measures of anxiety in higher functioning youth with ASD correlate with parent-report data (Farrugia and Hudson 2006), others have reported that parent-reports may be more accurate (Russell and Sofronoff 2005). Using multiple informants (e.g., parent and child) and reliable and valid rating scales to assess psychiatric symptoms is therefore recommended (White et al. 2009).

The purpose of this study was to explore relationships among anxiety, loneliness, and social disability in a clinical sample of high-functioning youth with ASD using both parents and youth (i.e., children and adolescents) as informants. The primary aim was to explore the self-reported experience of anxiety in youth with ASD. It was hypothesized that (1) youth with more severe social disability would report more symptoms of anxiety, and (2) youth who self-reported higher anxiety would report experiencing more loneliness than less anxious peers.

Methods
Participants

The sample was comprised of 20 children and adolescents receiving services through an outpatient clinic for youth with ASD. Parents and youth participants gave consent and assent, respectively, to participate in data collection for the study, which was approved by the university institutional review board. All participants had clinical diagnoses of ASD including autistic disorder (n = 2), PDD-NOS (n = 3), or Asperger's disorder (n = 15). All ASD diagnoses were confirmed by the ADOS (Lord et al. 2002). Participants ranged in age from seven through 14 years. Eighteen (90%) of the 20 participants were male. Data on cognitive functioning (IQ level) were taken from previous psychoeducational and psychological reports provided by the parents or from assessments conducted as part of the child's clinic evaluation, when such testing was completed.
Participants in this clinical sample were not seeking treatment specifically for anxiety. The Clinic offered diagnostic evaluation and treatment services to youth with ASD. Services were independent of the research, and participation in the study was optional for clinic patients. All clients of the Clinic were offered participation in the study. Demographic information on the sample is provided in Table 1. Of the 20 participants, all attended public regular education schools with the exception of one child. Most received some type of special education services through their school (n = 14), with speech/language therapy being the most commonly reported service (n = 8, 40%), per parent report. Five (25%) participants had or were receiving social skills training interventions at school. Thirteen participants (65% of sample) were taking between one and five medications. The most commonly prescribed type of medication was antidepressants (n = 9), followed by stimulants (8), and only one child was taking an anti-anxiety medication.

Procedures

Following completion of parental consent and child assent, a packet of measures was given to the parent to complete. In addition, the child was administered several diagnostic measures and questionnaires, as described below. The investigator (SW) was available to answer any questions the parent or youth had about the measures.

Measures
Autism Diagnostic Observation Schedule (ADOS; Lord et al. 2002)

The ADOS is a structured, interview-based observational assessment conducted with the child. The child is presented with activities and questions which pull for communicative and social behaviors typically difficult for individuals with ASD. Algorithm scores for communication and socialization are calculated to support the likelihood, or lack thereof, of ASD diagnosis. The ADOS typically takes 30-45 min to complete and has excellent test-retest reliability (.82) and inter-rater reliability (.92) (Lord et al. 2002). All participants enrolled in the study obtained combined scores (communication and social interaction) above the algorithm diagnostic threshold for ASD.
Social Communication Questionnaire (SCQ; Berument et al. 1999; Rutter et al. 2003)

The SCQ is a 40-item questionnaire based on autism diagnostic interview-revised (ADI-R; Lord et al. 1994), a gold standard diagnostic interview for ASD. Completed by the child's parent or caregiver in approximately 10 min, the SCQ is a reliable instrument (internal consistency α ranging from .84 to .93; Rutter et al. 2003) for assessing the likelihood of an ASD diagnosis, with higher scores indicative of greater severity. Although the usual cutoff score for screening purposes is 15 (Rutter et al. 2003), in higher functioning individuals a lower threshold is acceptable if the diagnosis is supported by clinical assessment and other measures (Corsello et al. 2003).
Social Responsiveness Scale (SRS; Constantino and Gruber 2005)

Completed by the parent, the SRS is a 65-item rating scale that measures autistic symptoms as they occur in typical social situations. Developed and normed using ASD samples, it is completed in 15-20 min and yields an overall score reflecting severity of social deficits in ASD as well as five symptom domain scores. Internal consistency of the SRS is excellent (α = .97; Constantino and Gruber 2005). The SRS provides a dimensional measure of ASD, with higher scores on the SRS indicating a greater degree of social disability.
Social Competence Inventory (SCI; Rydell et al. 1997)

Completed by the parent, the SCI is a 25-item questionnaire scored on a five-point Likert scale. It contains two subscales derived through factor analysis: the Prosocial orientation [PO] index (17 items), which assesses the child's ability to behave appropriately in normal and problematic (e.g., conflict) social situations with peers; and the social initiative [SI] index (8 items), which captures the child's capacity for making social initiations with peers. Higher scores indicate better social functioning. Both scales have been validated against behavioral observations and have excellent reliability, α = .88 and .75.
Multidimensional Anxiety Scale for Children (MASC; March 1997)

Completed by the child in less than 15 min, the MASC is a brief self-report measure that assesses four major dimensions of anxiety (physical symptoms, harm avoidance, social, separation/panic) and produces an index of total anxiety. This measure consists of 39 items and has good reliability (.88) and stability (.93) over 3 months. The MASC total score (r = .63) and all factor scores (r = .43-.71), with the exception of harm avoidance (r = -.13), correlate significantly with the revised children's manifest anxiety scale (RCMAS; Reynolds and Richmond 1978), providing support for its convergent validity. Confirmatory factor analysis also indicates that the MASC factor structure is invariant across males and females as well as younger and older youth.
Child Behavior Checklist (CBCL; Achenbach and Rescorla 2001)

The CBCL was completed by the parent. It is a syndrome-based scale that includes two parts, one assessing the child's social competence and the other assessing emotional and behavioral problems in children aged 6-18 years. The child's behavior during the past 6 months is rated on 118 items using a three-point scale: 0 = "not true", 1 = "somewhat or sometimes true" and 2 = "very true or often true." Two broad-band dimensions (internalizing and externalizing) can be derived. The CBCL covers eight syndrome dimensions including aggressive behavior, anxious/depressed, attention problems, rule-breaking behavior, social problems, somatic complaints, thought problems, and withdrawn/depressed. The scales have demonstrated reliability and validity: test-retest stability (.89), internal consistency (.90).
Loneliness Questionnaire (Asher et al. 1984)

This is a standardized self-report that assesses the child's feelings of loneliness. Bauminger et al. (2003) modification, to assess loneliness in children with ASD, was used in the current study. This 24-item scale has high internal consistency (.93). In addition to an index of ‘global' loneliness, the scale distinguishes between emotional loneliness, or feelings of isolation and lack of affective bonding (9 questions) and social loneliness, or the child's perceived lack of social involvement with peers (13 questions).
Data Analysis

The range and reliability of the youth self-reports of anxiety on the MASC were first explored, given that the MASC has not been typically used in young people with ASD. Relationships between youth self-reported symptoms of anxiety and the three parent-report measures of social disability, and the relationship between self-reported anxiety and self-reported loneliness, were examined using one-tailed bivariate correlations. Exploratory analyses examined how parent-reported anxiety symptoms in the youth related to other measures of pathology and loneliness.

Results
Self-Reported Anxiety

MASC total T scores ranged from 28 to 83 (M = 52.90). Five children (25% of sample) earned MASC total scores in the clinical range (T score > 65). Reliability was evaluated by computing internal consistency (Cronbach's α) and inter-item correlations. Cronbach's alpha provides a correlational index that reflects the homogeneity of the individual item scores with the total score (Cronbach 1951). The mean inter-item correlation provides an index of the degree to which the items comprising a scale measure the same construct. The Cronbach's α coefficient for all items (MASC total anxiety score) was .924, indicating a high level of internal consistency. For the anxiety disorder index (ADI), α = .822. For all 39 questions on the MASC, the mean inter-item correlation was .238. The mean inter-item correlation for the ADI (10 questions) was .327. These reliability coefficients are somewhat higher than those reported for the MASC's normative sample (March 1997).

Of note, the average inconsistency index in this ASD sample was higher than the means reported for the scale's normative, non-ASD sample (March 1997). Four participants in the current study earned inconsistency index scores above the acceptable threshold (i.e., ≥12 for 8-11-year olds; ≥10 for 12-19-year olds), which may indicate random responding or poor understanding of how to answer the questions.
As shown in Table 2, the highest elevation for the clinical scales was separation anxiety (M = 61.58), the lowest was harm avoidance (M = 48.55). Exploratory analyses were conducted to determine if self-reported anxiety varied in the sample by age or cognitive functioning. Dividing the sample into two groups based on the median age of 12 years, there was no significant difference in total anxiety between younger and older children. A significant difference in total MASC scores was present, however, between children below the median IQ of 92 and those above (t = 2.28, p = .038). The children in the lower IQ group (n = 9) had a mean total T score of 64.44 (±16.69), compared to 49.88 ± 7.20 for the higher IQ group (n = 8). Three participants did not have IQ data and were excluded from this analysis.

Anxiety, Loneliness, and Social Functioning
Hypothesis 1 Youth who self-report greater anxiety will have greater social disability. One-tailed correlations were conducted between self-reported anxiety (MASC) and the three parent-reported measures of social disability (SCQ, SRS, SCI). This hypothesis was not supported. MASC total scores were not significantly correlated with scores from the SCQ (r = .297), the SRS (r = -.339), or the SCI (SI, r = .34; PO, r = -.18).
Hypothesis 2 Youth who self-report more anxiety will report experiencing more loneliness. The bivariate correlation between total anxiety and loneliness was not significant (r = .325, ns). However, when the sample was divided into those who obtained ‘above average' anxiety scores (i.e., MASC total score ≥ 61, n = 5) and those who did not (score < 61), there were differences in reported loneliness. The high-anxiety group self-reported more ‘social' loneliness on the loneliness questionnaire, compared to their less anxious peers (t = 2.57, p < .05). Those children who experienced the highest levels of anxiety earned an average social loneliness score of 3.38 (±.58), compared to 2.44 (±.72) for those whose self-reported anxiety was not elevated. To determine if social anxiety specifically was related to loneliness, correlations were conducted. Social anxiety (raw score) was significantly correlated with social (r = .59, p = .01) and global (r = .50, p = .04) loneliness scores. Seven individuals obtained ‘above average' or higher social anxiety T scores and the median T score was 54.

Exploratory Analyses
Exploratory analyses undertaken to determine if parent-reported anxiety symptoms were related to degree of social disability yielded mixed results. Neither the anxious/depressed syndrome scale nor the anxiety problems scale of the CBCL were significantly correlated with social deficit as measured by the SCQ (see Table 3). However, the correlation between CBCL ‘anxious/depressed' T scores and the social initiation scale of the SCI was significant (r = -.59, p < .05), indicating that children whose parents rated them as more anxious or depressed also had lower social drive, or were less likely to initiate with peers.

Bivariate correlations between self-reported anxiety and parent-reported anxiety were not significant, indicating a lack of agreement between parents and their children with respect to symptoms of anxiety. In fact, MASC total scores were negatively correlated with both the anxious/depressed syndrome scale (-.165, ns) and the anxiety problems scale of the CBCL (-.045, ns). No significant differences were found between children with elevated MASC scores (i.e., those in the highest quartile) and those with the lowest MASC total scores with respect to IQ, age, parent-reported psychopathology, or loneliness.

Discussion

The purpose of this study was to explore relationships among anxiety, loneliness, and social disability in a clinical sample of high-functioning youth with ASD using both parents and youth as informants. In this sample of youth with diagnosed autism spectrum disorders, anxiety was indeed prevalent. Based on self-report, one quarter of the children (n = 5) earned MASC total scores falling in the clinical range. Seven (35% of sample) obtained elevated social anxiety scores. One of the goals of this study was to explore the utility of a commonly used self-report measure of anxiety with youth who have autism and related conditions. The MASC appears to be a reliable tool, although it is likely that some of the children had difficulty understanding the questions or they responded randomly, based on their Inconsistency Index scores.

There was little agreement between parent and self-report on the youths' experience of anxiety. Parent-reported child anxiety was associated with the youth being less likely to initiate with peers, based on parent report. Child-reported anxiety, however, was not related to parent-reported social impairment. These findings are consistent with previous research finding that different informants usually do not agree in their ratings of child behavior problems. As De Los Reyes and Kazdin (2005) conceptualized, raters' attributions of the causes of the child's behavior, their own biases, and the contexts in which they form their ratings all impact their reports and likely contribute to such discrepancies in behavior ratings. The Attribution Bias Context Model (De Los Reyes and Kazdin 2005) may help partially explain the lack of agreement between parent- and child-ratings in the present study. Characteristics typically associated with ASD (e.g., poor insight) and the frequently reported unusual expression of symptoms of anxiety in children with ASD (White et al. 2009) add further complexity to parent-child discrepancies in ratings, however. It is unclear which report should be relied upon-parent or youth. If one assumes the parent is the more valid reporter, is the social interest present but the child lacks confidence in initiating? If so, intervention targeting anxiety would theoretically impact (mediate) social functioning. If, on the other hand, the youth self-report is relied upon, then anxiety and social disability should be viewed separately from each other. The data from this preliminary study do not indicate which report is more or less valid.

The youth with ASD who reported experiencing elevated anxiety also reported experiencing more social loneliness, but not emotional loneliness. Perhaps these youngsters desire more peer social partners but get their ‘need' for social affective bonding met through relationships other people in their lives, such as parents. Alternatively, they may not have a great need for emotional closeness, and thus do not experience emotional loneliness. The social and emotional loneliness scores reported by the present study's sample are consistent with those reported by the 18 children with high-functioning autism in the Bauminger et al. (2003) sample; social (M: 2.72, 2.73), emotional (M: 2.74, 2.44), respectively. Bauminger and colleagues found that children and teens with high-functioning autism self-reported more feelings of both social and emotional loneliness than did typically developing peers. Therefore, an alternative and perhaps more plausible explanation is that, regardless of whether or not youth with ASD report experiencing anxiety, they may tend to feel emotionally disconnected or alone. Anxiety, when present, may mediate the child's degree of involvement in activities with peers. Of particular interest is the significant relationship between social anxiety and social loneliness.

Follow-up studies should examine potential age effects on youth's experience of anxiety, loneliness, and social relationships. In a hypothetical scenario, a young boy with Asperger's may be highly socially motivated and relatively unencumbered by anxiety, which in turn leads to frequent yet awkward social interactions with peers. As he approaches adolescence, he develops greater insight about his ‘uniqueness' and he begins to experience heightened social anxiety. This developing insight and anxiety lead him to become more aloof or passive, due not to social disinterest but rather to previous negative learning experiences and his growing social anxiety. This is but one possible scenario representing a complex developmental pathway, however, and longitudinal research is needed. Understanding anxiety in ASD and its possible influence on social functioning will require examination of the possible influence of poor social skills and experiences with social failure on the development of anxiety in youth with ASD (e.g., Attwood 1998; Bellini 2006).

The most notable limitation of this study is its small sample size. Clearly, firm conclusions based on such a preliminary study are difficult to draw. Ideally, future studies should examine larger, well-characterized samples with sufficient power to compare the ASD diagnostic subgroups: autistic disorder, aspergers disorder, and pervasive developmental disorder-not otherwise specified. It is, for instance, quite possible that youth with Aspergers subjectively experience more loneliness and social anxiety than peers with other forms of ASD. Such subgroup differences could not be explored this given the configuration and size of the current sample. A control group of typically developing youth without ASD would also help to place the findings of the present study in perspective.

A second limitation is the lack of secondary anxiety measures. The MASC is not the only youth self-report measure of anxiety, although it is widely used and appropriate for the studied age range. Data on the relative utility, and face validity, of self-report instruments for this population is imperative, especially given the apparent difficulty with random responding in some of the participants. It is indeed concerning that four participants (20% of sample) obtained Inconsistency Index scores above the typically accepted threshold. Related to this, it would be ideal to have blinded or masked assessments of social functioning, rather than relying on parent-report for cross-comparison. Observational data or teacher ratings may be useful in this regard in future studies.

Despite these limitations, this study confirms that anxiety is often a clinical concern in youth with ASD. Furthermore, the experience of anxiety may be related to the experience of social loneliness. The MASC may be a useful tool in clinical settings with higher functioning youth with ASD. But, as with most child-based evaluation scenarios, the self-report should be complimented by parent report and when possible, teacher report as well.

Additional psychometric research on anxiety measures in youth with ASD is paramount. An area that deserves immediate attention is the determination of whether anxiety measures are performing similarly across ASD youth and non-ASD youth. Differential item functioning (DIF) analyses are able to determine the presence of qualitatively distinct response processes that bias assessment of severity within the construct of interest (cf. Douglas et al. 1998; Roussos and Stout 2004; Zumbo and Witarsa 2004). Thus, DIF would be present when ASD and non-ASD youth with the same level of estimated anxiety severity do not have the same probability of endorsing an item option(s) (Holland and Wainer 1993). This type of data would inform researchers as to whether or not youth with and without ASD presenting with anxiety difficulties can be directly compared on anxiety measures. Moreover, it could provide information on the sensitivity of a given measure for capturing psychopathology in ASD.

Youth with ASD often struggle with anxiety. However, anxiety may not symptomatically present like anxiety in typically developing, non-ASD youth and the young person with ASD may have difficulty recognizing their feelings and behaviors as symptomatic of anxiety. It is often very difficult to distinguish other Axis I conditions, such as anxiety disorders, from the ASD itself. As such, further investigation on reliable and valid assessment tools would benefit both future clinical research and practitioners who must make difficult diagnostic decisions.
Acknowledgments This project was supported by a grant from the National Institute of Mental Health [1K01MH079945-01; PI: S. W. White]. The authors acknowledge the consultation of Jessica Schneider, MA.

References
Achenbach, T. M., & Rescorla, L. A. (2001). Manual for the ASEBA school-age forms & profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, & Families.

Amaral, D. G., Bauman, M. D., & Schumann, C. M. (2003). The amygdale and autism: Implications from non-human primate studies. Genes Brain & Behavior, 2, 295-302. doi:10.1034/j.1601-183X.2003.00043.x.
CrossRef

Apter, A., Brown, S.-L., Korn, M. L., & van Praag, H. M. (1991). Psychiatric disorders of childhood: The role of serotonin. In S.-L. Brown & H. M. van Praag (Eds.), Clinical and experimental psychiatry monograph: No. 4. The role of serotonin in psychiatric disorders (pp. 215-238). Philadelphia, PA: Brunner/Mazel.

Asher, S. R., Hymel, S., & Renshaw, P. D. (1984). Loneliness in children. Child Development, 55, 1456-1464. doi:10.2307/1130015.
CrossRef

Attwood, T. (1998). Asperger's syndrome: A guide for parents and professionals. Philadelphia: Kingsley.

Attwood, T. (2000). Strategies for improving the social integration of children with Asperger syndrome. Autism, 4, 85-100. doi:10.1177/1362361300004001006.
CrossRef

Bauminger, N., & Kasari, C. (2000). Loneliness and friendship in high-functioning children with autism. Child Development, 71, 447-456. doi:10.1111/1467-8624.00156.
PubMed CrossRef

Bauminger, N., Shulman, C., & Agam, G. (2003). Peer interaction and loneliness in high-functioning children with autism. Journal of Autism and Developmental Disorders, 33, 489-507. doi:10.1023/A:1025827427901.
PubMed SpringerLink

Bellini, S. (2006). The development of social anxiety in adolescents with autism spectrum disorders. Focus on Autism and Other Developmental Disabilities, 21, 138-145. doi:10.1177/10883576060210030201.
CrossRef

Berument, S., Rutter, M., Lord, C., Pickles, A., & Bailey, A. (1999). Autism screening questionnaire: Diagnostic validity. The British Journal of Psychiatry, 161, 839-843.

Constantino, J. N., & Gruber, C. P. (2005). Social responsiveness scale (SRS). Los Angeles: Western Psychological Services.

Corsello, C., Cook, E., & Leventhal, M. (2003). A screening instrument for autistic spectrum disorders. Paper presented at the Society for Research in Child Development Biennial Conference, Tampa, Florida.

Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16, 297-334. doi:10.1007/BF02310555.
SpringerLink

De Los Reyes, A., & Kazdin, A. E. (2005). Informant discrepancies in the assessment of childhood psychopathology: A critical review, theoretical framework, and recommendations for further study. Psychological Bulletin, 131, 483-509. doi:10.1037/0033-2909.131.4.483.
PubMed CrossRef

Douglas, J. A., Stout, W., & DiBello, L. V. (1998). A kernel-smoothed version of SIBTEST with applications to local DIF inference and function estimation. Journal of Educational and Behavioral Statistics, 21, 333-363.

Farrugia, S., & Hudson, J. (2006). Anxiety in adolescents with Asperger syndrome: Negative thoughts, behavioral problems, and life interference. Focus on Autism and Other Developmental Disabilities, 21, 25-35. doi:10.1177/10883576060210010401.
CrossRef

Ghaziuddin, M. (2002). Asperger syndrome: Associated psychiatric and medical conditions. Focus on Autism and Other Developmental Disabilities, 17, 138-144. doi:10.1177/10883576020170030301.
CrossRef

Holland, P. W., & Wainer, H. (1993). Differential item functioning. Hillsdale, NJ: Erlbaum.

Lord, C., Rutter, M., DiLavore, P. C., & Risi, S. (2002). Autism diagnostic observation schedule. Los Angeles: Western Psychological Services.

Lord, C., Rutter, M., & Le Couteur, A. (1994). Autism diagnostic interview-revised: A revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. Journal of Autism and Developmental Disorders, 24, 659-685. doi:10.1007/BF02172145.
PubMed SpringerLink

March, J. S. (1997). Multidimensional anxiety scale for children manual. North Tonawanda, NY: Multi-Health Systems.

Myles, B. (2003). Behavioral forms of stress management for individuals with Asperger syndrome. Child and Adolescent Psychiatric Clinics of North America, 12, 123-141. doi:10.1016/S1056-4993(02)00048-2.
PubMed CrossRef

Myles, B., Barnhill, G., Hagiwara, T., Griswold, D., & Simpson, R. (2001a). A synthesis of studies on the intellectual, academic, social/emotional and sensory characteristics of children with Asperger syndrome. Education and Training in Mental Retardation and Developmental Disabilities, 36, 304-311.

Myles, B., Bock, S., & Simpson, R. (2001b). Asperger syndrome diagnostic scale. Austin, Texas: Pro-Ed.

Piven, J., & Palmer, P. (1999). Psychiatric disorder and the broad autism phenotype: Evidence from a family study of multiple-incidence autism families. The American Journal of Psychiatry, 156, 557-563.
PubMed

Reynolds, C. R., & Richmond, B. O. (1978). Factor structure and construct validity of "What I Think and Feel". The revised children's manifest anxiety scale. Journal of Personality Research, 43, 281-283.

Roussos, L. A., & Stout, W. F. (2004). Differential item functioning analysis: Detecting DIF items and testing DIF hypotheses. In D. Kaplan (Ed.), The SAGE handbook of quantitative methodology for the social sciences (pp. 107-115). Thousand Oaks, CA: Sage.

Russell, E., & Sofronoff, K. (2005). Anxiety and social worries in children with Asperger syndrome. The Australian and New Zealand Journal of Psychiatry, 39, 633-638. doi:10.1111/j.1440-1614.2005.01637.x.
PubMed CrossRef

Rutter, M., Bailey, A., & Lord, C. (2003). The social communication questionnaire. Los Angeles: Western Psychological Services.

Rydell, A., Hagekull, B., & Bohlin, G. (1997). Measurement of two social competence aspects in middle childhood. Developmental Psychology, 33, 824-833. doi:10.1037/0012-1649.33.5.824.
PubMed CrossRef

Schopler, E., & Mesibov, G. (1983). Autism in adolescents and adults. New York: Plenum Press.

Smalley, S., McCracken, J., & Tanguay, P. (1995). Autism, affective disorders, and social phobia. American Journal of Medical Genetics, 60(1), 19-26. doi:10.1002/ajmg.1320600105.
PubMed CrossRef

Tantam, D. (2003). The challenge of adolescents and adults with Asperger syndrome. Child and Adolescent Psychiatric Clinics of North America, 12, 143-163. doi:10.1016/S1056-4993(02)00053-6.
PubMed CrossRef

United States Centers for Disease Control and Prevention. (2007). Prevalence of autism spectrum disorders-autism and developmental disabilities monitoring network, 14 sites, United States, 2002. Surveillance Summaries, MMWR 2007; 56(No. SS-1).

White, S.W., Oswald, D., Ollendick, T., & Scahill L. (2009). Anxiety in children and adolescents with autism spectrum disorders. Clinical Psychology Review (in press).

Zumbo, B. D., & Witarsa, P. M. (2004). Nonparametric IRT methodology for detecting DIF in moderate-to-small scale measurement: Operating characteristics and a comparison with the Mantel Haenszel. Paper presented at the American Educational Research Association Meeting, San Diego, CA.