Validity of Comorbid Psychiatric Disorders in Youngsters with Autism Spectrum Disorders

Original Article

Abstract

There is mounting evidence that the Diagnostic and Statistical Manual of Mental Disorders (DSM) provides an appropriate conceptualization for psychiatric disorders in individuals with autism spectrum disorders (ASDs). The current study examined symptoms of disruptive behavior, anxiety, and mood disorders and how IQ and language skills impacted their manifestation in a sample of 61 children with ASD and significant emotional/behavioral problems. Subsyndromal presentation and the role of behavior equivalents were also explored. Analyses indicated that children with IQ < 70 had fewer reported symptoms than those with IQ ≥ 70. Symptoms endorsement also varied based on language levels. Individuals with IQ < 70 were more likely to be subsyndromal for Generalized Anxiety Disorder and nonverbal individuals were more likely to be subsyndromal for Oppositional Defiant Disorder. Only one of the seven behavior equivalents was associated with internalized disorders. Special considerations may be necessary when applying some DSM criteria to assess comorbid disorders in individuals with ASDs.

Keywords

Autism spectrum disorder DSM Psychiatric symptom Behavior problem Comorbidity Validity 

Recent studies have reported that youngsters with autism spectrum disorders (ASDs) present with high rates of behavior and psychiatric problems (e.g., de Bruin et al. 2007; Gadow et al. 2005; Lecavalier 2006; Leyfer et al. 2006; Simonoff et al. 2008). For instance, Simonoff et al. (2008) reported that 70% of their population-based sample of 112 ten to fourteen-year-old children with ASDs had at least one comorbid psychiatric disorder.

Despite these high rates of psychopathology, there is a dearth of research on the validity of the Diagnostic and Statistical Manual-Fourth Edition (DSM-IV; American Psychiatric Association 2000) in this population. For example, the impact of age, IQ, and language skills on the presentation of DSM-defined symptoms and disorders has received little attention. From a conceptual standpoint, it remains unclear if behaviors and symptoms of those with ASDs are features of traditionally defined psychiatric disorders or separate clusters of behavior which appears similar but are really part of the ASD diathesis.

Several recent studies lend support to the DSM as a valid conceptualization of psychopathology in children with ASDs (Gadow et al. 2008b; Gadow et al. 2006; Gadow, et al. 2008c; Gadow et al. 2008a). Lecavalier et al. (2008) submitted parent and teacher ratings of DSM based symptoms to confirmatory factor analysis. They found support for Attention Deficit Hyperactivity Disorder (ADHD), Oppositional Defiant Disorder (ODD), Conduct Disorder (CD), Generalized Anxiety Disorder (GAD), and Mood/Dysthymic disorder (MDD) as diagnostic categories in ASD. In fact, they reported similar indices of fit for children with ASDs and a comparison group of typically-developing children. Interestingly, fit indices improved when analyses were only conducted on children with an IQ > 70. Along the same lines, Gadow and colleagues provided additional support for the validity of ADHD and ODD as diagnostic categories in ASDs by examining patterns of comorbidity and genetic and psychosocial risk factors (Gadow et al. 2008b; Gadow et al. 2006; Gadow et al., 2008a). The differential patterns of comorbidity and risk factors observed in ASDs were similar to those observed in typically-developing children.

The language impairments and intellectual disability (ID) encountered in ASDs complicate psychopathology research. The lack of speech in some makes it difficult to determine the presence of many DSM-IV symptoms. Even mild impairments in language can make discussion of abstract concepts and subtle differences in emotion difficult to detect and assess. The high prevalence of ID in individuals with autism further complicates assessment. The intellectual limitations may lead these individuals to have difficulty understanding and expressing the more complex cognitive phenomena that occur in some psychiatric conditions such as anxiety, mood, or psychotic disorders.

Related to language and intellectual impairments are two other clinical matters: subthreshold symptomatology and behavior equivalents. It is possible that some individuals with ASDs have impairments related to comorbid psychiatric disorders, but do not meet full criteria due to an inability of clinicians to assess all symptoms. One study is noteworthy in this regard. Leyfer et al. (2006) identified children who were impaired and could benefit from treatment, but fell just short of full diagnostic criteria. They found the most common subsyndromal disorders to be ADHD (25%), MDD (14%), Separation Anxiety Disorder (7%), and Obsessive Compulsive Disorder (OCD) (6%). The specific criteria used to derive subsyndromal categories (i.e., how many symptoms were required) were not described. It was also unclear how IQ and language affected the likelihood that individuals would be identified as subsyndromal rather than meeting full criteria.

In order to more accurately identify psychiatric disorders in populations with developmental delays, some researchers have proposed to rely on behavioral equivalents. The idea is to equate DSM-IV criteria with observable behaviors that are compatible with the individual’s limited communication and cognitive delays. Guidelines for incorporating behavioral equivalents into psychiatric diagnosis in ID have emerged (Fletcher et al. 2007; Royal College of Psychiatrists 2001). For instance, proposed behavioral equivalents for depression include property destruction, aggression, or self-injury. Examples of behavioral equivalents for mania include loud inappropriate laughing, excessive giddiness, more talkative or more noisemaking/screaming.

Research on behavioral equivalents has yielded equivocal results (e.g., Martson et al. 1997; Matson et al. 1999). There are disagreements on how behavioral equivalents should be applied to DSM-IV disorders in individuals with ID. Some have proposed using substitute criteria for fear of missing cases due to the individuals’ inability to report emotions (McBrien 2003). Others have suggested adding these as criteria in addition to the established DSM-IV criteria. Alternatively, some have suggested that behavior problems are non-specific indicators of distress and dysfunction and should be used as indicators of potential underlying psychopathology rather than as a characteristic or criterion for specific diagnoses (Tsiouris et al. 2003).

A reliable and valid taxonomy of psychopathology is necessary for the field of ASDs. Once clinical and associated features have been reliably described, it allows research on etiology, course, and response to treatment. Robins and Guze (1970) proposed a process to facilitate valid classification which included the following phases: clinical description, laboratory, follow-up, and family studies. The first step consists of describing the hallmark and associated features of the disorder. Once the clinical picture is elucidated, differences on external factors, such as adaptive behavior, response to treatment, or family functioning provide further support for the validity of the classification system. Follow-up studies examining marked differences in outcome and genetic studies provide further evidence for the validity of respective disorders (Cantwell 1996). According to this process, the first step toward a valid classification is to elucidate the clinical description of disorders.

The purpose of the current study was to elucidate the clinical picture of psychiatric disorders in children with ASDs. There has been little research on the specific symptoms exhibited by those with ASDs and how language and IQ impact their manifestation. Using a structured psychiatric interview, we examined the endorsement of existing psychiatric symptoms. We paid particular attention to the impact of IQ and language on endorsement rates. We also examined the presence of subsyndromal diagnoses and explored how these varied based on IQ and language levels. Subsyndromal diagnoses were defined as those individuals who had symptoms of the disorder and related impairments, but fell short of full criteria by one or two symptoms. Finally, we explored the validity of some behavioral equivalents. To do this, we measured the association between internalizing disorder diagnoses and specific behavior problems as assessed by the Nisonger Child Behavior Rating Form.

Method

Instruments

The Children’s Interview for Psychiatric Symptoms-Parent Version (P-ChIPS; Weller et al. 1999) is a structured interview designed to assess psychopathology according to DSM-IV criteria in children and adolescents aged 6 to 18 years. It assesses 20 DSM-IV disorders and psychosocial stressors. Symptoms are assessed in a yes/no format. Within each diagnostic section cardinal questions are asked. If a respondent answers “no” to a certain number of questions, the rest of the questions in that section may be skipped. Onset and duration information are obtained for each disorder. Impairment is assessed with questions asking if symptoms cause problems at home, school, or with other children. The P-ChIPS has good psychometric properties and offers the advantage of a short administration time (i.e., 1 h; Weller et al. 2000). Psychometric properties from the current sample were reported in another publication (Witwer 2009).

Rather than just relying on the cardinal questions, all questions in each mood, anxiety, and disruptive behavior disorder sections were asked. Current (i.e., last month) symptoms and the presence/duration of impairment were assessed. Items requiring the child to have functional communication were not administered to nonverbal children. For Depression, this included the following sections: Guilt, Hopelessness, and Morbid Thoughts (except items assessing suicide attempts). For Separation and Generalized Anxiety, this included items asking the parent to report on child worry. For OCD, these included items requiring parents to report on child’s thoughts. Finally, for Mania/Hypomania this includes items asking if the child felt he or she had special abilities and racing thoughts.

The Nisonger Child Behavior Rating Form (NCBRF; Aman et al. 1996) is a rating scale developed specifically for children with ID. The parent version contains two prosocial subscales (Compliant/Calm and Adaptive/Social) and six problem behavior subscales: Conduct Problems, Insecure/Anxious, Hyperactive, Self-injury/Stereotypic, Self-isolated/Ritualistic, and Overly Sensitive. Good psychometric properties have been reported in children with ASDs (Lecavalier et al. 2004) and norms are also available for this population (Lecavalier 2006). Seven NCBRF items were used to assess behavioral equivalents: Physically harms or hurts self on purpose, Rocks body or head back and forth repetitively, Repeatedly flaps or waves hands, fingers, or objects, Engages in meaningless repetitive body movements, Knowingly destroys property, Physically attacks people, and Temper tantrum.

The Autism Diagnostic Interview-Revised (ADI-R; Rutter et al. 2003) is a semi-structured interview for caregivers of children and adults for whom autism or ASDs are possible diagnoses. It contains 93 questions, 34 of which are used in the diagnostic algorithm that is based on DSM-IV autism criteria. The algorithm items were administered to confirm ASD diagnosis and to assess the presence of functional language.

The Stanford-Binet-V (Roid 2003) is an individually-administered standardized intelligence test for individuals aged 2 years through adulthood. It can provide full scale, verbal, nonverbal, and brief IQs. A Brief IQ was obtained on children who did not have another IQ test administered within the last year. A Nonverbal IQ was obtained for children without conversational language.

A standard Demographic Form was used to collect information on the child and informant. This form contained questions concerning the child’s date of birth, gender, ASD diagnosis, psychiatric diagnoses, and treatments used. Informants were asked their age, relationship to the child, and highest level of education.

Participants

Table 1 lists child and informant characteristics in detail. A total of 61 children and adolescents (mean = 11.2 ± 3.8 years; range 6–17) with an ASD participated in the study. Eighty-two percent (n = 50) of the sample was male and 77% (n = 47) was Caucasian. IQs ranged from 42 to 150 with a mean of 68.4 (SD = 23.3). The sample contained 16 children with parent-reported diagnoses of Asperger’s Disorder, 17 with autism, and 26 with Pervasive Developmental Disorder-Not Otherwise Specified (PDDNOS). Fourteen (23%) of the 61 children did not have conversational language based on the ADI-R. Mean ADI-R diagnostic algorithm scores are presented in Table 1. Thirty-six children were receiving treatment from a psychiatrist and 12 from a developmental pediatrician. Psychotropic medications are listed in Table 1.
Table 1

Child and informant characteristics

Participant Characteristics Mean (SD)/n, (%)

Children

 Age (Mean and SD; in years)

11.2 (3.8)

 Gender (n, % male)

50 (82%)

 Race (n, % Caucasian)

47 (77%)

 IQ (Mean and SD; n = 58)

68.4 (23.3)

   IQ ≥ 85

13 (22.4%)

   IQ 70–84 (Borderline)

9 (15.5%)

   IQ 69–55 (Mild)

18 (31%)

   IQ 54–40 (Moderate)

18 (31%)

ADI-R Diagnostic Algorithm Scores

 ADI-R Social

20.4 (5.3)

 ADI-R Communication (nonverbal ; n = 14)

12.5 (3.6)

 ADI-R Communication (verbal ; n = 47)

15.0 (4.8)

 ADI-R Repetitive

6.7 (2.6)

Medication

 Antipsychotic

37 (60.7%)

 ADHD medication

25 (41.0%)

 Beta blocker

28 (32.8)

 Mood stabilizer

13 (21.3%)

 SSRI

12 (19.7%)

 Anxiolytic

3 (4.9%)

 Atypical antidepressant

3 (4.9%)

 Antiepileptic

3 (4.9%)

Informants

 Age (Mean and SD; in years)

40.4 (3.4)

 Relationship (n, % mother)

59 (97%)

Education (n, %)

 Attended High School

1 (1.8%)

 Graduated High School

6 (10.9%)

 Attended College

19 (34.5%)

 Graduated College

27 (49.1%)

 Graduate/Professional School

2 (3.6%)

Ninety-seven percent (n = 59) of informants were mothers. Informant ages ranged from 25 to 56 years (Mean = 40.4 years; SD = 3.4 years). A large proportion of the 55 informants who provided information on their education level had graduated college (n = 29, 52.7%).

Procedure

Inclusion criteria for the current study were as follows: a.) aged between 6 and 17 years, b.) previous diagnosis of an ASD confirmed by the ADI-R, c.) presence of significant emotional/behavior problems requiring behavior/ psychiatric treatment or special school placement, d.) IQ > 40; and e.) parent/guardian fluent in English. Since the ADI-R does not have specific criteria for Asperger’s disorder or PDD-NOS, confirming these diagnoses was done by requiring those with Asperger’s disorder to be above cutoffs on the Social and Restricted Repetitive Behavior domains and those with PDD-NOS to be above cutoff on the Social domain and one of the other two domains. This is consistent with DSM-IV criteria and has been done in previous studies (Posey et al. 2007).

Children were recruited from University-based clinics, local psychiatrists’ offices, local support groups, and listservs for parents of children with ASDs. Agencies/clinics posted advertisements in their offices and distributed fliers to their patients. Previous OSU psychiatric research participants who gave permission to be contacted for future studies were also recruited with the help of research staff.

The appointment(s) took place either in the family home or at the University, depending on parent availability/convenience. First, the ADI-R was completed to confirm ASD diagnoses. Then, the NCBRF and P-ChIPs were completed. The NCBRF was completed before the structured interview because this is typically the order in which instruments are completed in applied settings. The IQ tests were administered at various times depending on the parent’s schedule and child’s behavior. Interviews were conducted by an author (ANW) who is trained and research certified on the ADI-R and P-ChIPS. Blindness was protected by not inquiring about diagnoses and treatment until after the interview and assessments were completed. The assessments took approximately 4–5 hours and were split into more than one visit when necessary. Families received a summary of interview results and were compensated for their time. Three children did not have IQ scores because parent/guardians did not return calls to schedule the testing. Therefore, 58 children were used for analyses involving IQ scores.

Statistical Analyses

Frequency counts of those who 1.) were above proposed cutoffs, 2.) were above symptom count criteria, 3.) fell short of proposed cutoffs by one or two symptoms and impairment was reported (subsyndromal), and 4.) had at least one impairing symptom were calculated. Chi squares were used to examine the relationship between various diagnostic groupings (above vs. below criteria; subsyndromal vs. above criteria) and specific symptoms and IQ (IQ ≥ 70/IQ < 70) and language (presence/absence of conversational language). Fisher’s exact test was used for analyses when expected cell counts were less than 5. Chi square analyses were also used to examine the distribution of behavioral equivalents among those above and below the clinical P-ChIPS cutoffs for internalizing disorders. NCBRF scores were dichotomized for analyses of symptom presence such that a score of 0 or 1 was coded as a 0 and scores of 2 or 3 were coded as a 1. Associations were examined between diagnoses and the seven NCBRF items listed above. Due to the number of comparisons, p values less than .01 were considered statistically significant.

Results

Impact of IQ and Language on Diagnostic Classification

Table 2 lists the frequency of those meeting diagnostic criteria for disruptive, mood, and anxiety disorders. Those with ID were less likely to meet criteria for GAD (χ2 = 12.95, p < .001). Verbal children were significantly more likely to be diagnosed with ODD (Fischer exact, p = .01) and GAD (Fischer exact, p = .01).
Table 2

Rates of disorders: impact of IQ and language

Disorder/ Symptom

Entire Sample (n = 61)

IQ ≥ 70 (n = 22)

IQ < 70 (n = 36)

No Lang (n = 14)

Lang (n = 47)

ADHD-C

41 (67.2)

16 (72.7)

24 (66.7)

9 (64.3)

32 (68.1)

ADHD-I

10 (16.4)

5 (22.7)

4(11.1)

0

10 (21.3

ADHD-H

5 (8.2)

0

3 (8.3)

1 (7.1)

3 (6.4)

ODD

46 (75.4%)

16 (72.7)

28 (77.8)

7 (50.0)

39 (83.0)*

CD

30 (49.2%)

11 (50.0)

19 (52.8)

5 (35.7)

25 (53.2)

Phobia

41 (67.2)

15 (68.2)

24 (67.2)

9 (64.3)

32 (68.1)

Soc Pho

10 (16.4)

5 (22.7)

5 (13.9)

1 (7.1)

9 (19.1)

Sep Anxiety

9 (14.8)

3 (13.6)

5 (13.9)

0

9 (19.1)

GAD

15 (24.6)

11 (50.0)

3 (8.3)**

0

15 (31.9)*

OCD

3 (4.9)

3 (13.6)

0

0

3 (6.4)

OCD-C

6 (9.8)

3 (13.6)

2 (5.6)

0

6 (12.8)

OCD-O

6 (9.8)

4 (18.2)

2 (5.6)

0

6 (12.8)

MDD

9 (14.8)

5 (22.7)

4 (11.1)

0

9 (19.1)

Dysthymia

8 (13.1)

3 (13.6)

5 (13.9)

0

5 (13.9)

Mania

9 (14.8)

4 (18.2)

4 (13.9)

2 (14.3)

7 (4.9)

*p ≤ 01, **p≤.001; ADHD-C = Attention Deficit Hyperactivity Disorder-Combined Type; ADHDH = Attention Deficit Hyperactivity Disorder-Hyperactive Type; ADHDI = Attention Deficit Hyperactivity Disorder-Inattentive Type; ODD = Oppositional Defiant Disorder; CD = Conduct Disorder; Phobia = Specific Phobia; Soc Pho = Social Phobia; Sep Anx = Separation Anxiety Disorder; GAD = Generalized Anxiety Disorder; OCD = Obsessive Compulsive Disorder; OCD-C = Obsessive Compulsive Disorder Compulsions; OCD-O = Obsessive Compulsive Disorder Obsessions; MDD = Major Depression Disorder

Table 3 lists frequency counts of those who: 1) were above proposed cutoffs (symptoms, duration, and impairment criteria met), 2) were above symptom count criteria, 3) fell short of proposed cutoffs by only one or two symptoms but reported impairments (subsyndromal), and 4) had at least one impairing symptom. Those with ID were more likely to be subsyndromal for GAD (χ2 = 7.04, p = .008; n = 10) versus meeting full criteria (n = 3). Those without ID were more likely to meet full GAD criteria (n = 11) than to be subsyndromal (n = 4). Chi Square/Fischer exact results examining the impact of IQ were not significant for the other disorders. No participants were subsyndromal for specific phobia, so it was not included in the analyses.
Table 3

Rates of individuals with symptoms endorsed, impairment noted, and meeting full diagnostic criteria (N = 61)

Disorder

Symptoms reported

Impairment reported

1–2 Symptoms short of criteria

Symptom criteria met

Criteria met for disorder

ADHD-C

61 (100)

55 (90.2)

5 (8.2)

41 (67.2)

41 (67.2)

ADHD-I

61 (100)

55 (90.2)

6 (9.8)

10 (16.4)

10 (16.4)

ADHD-H

61(100)

55 (90.2)

9 (14.8)

5 (8.2)

4 (6.6)

ODD

61 (100)

54 (88.5)

10 (16.4)

48 (78.7)

46 (75.4)

CD

50 (82.0)

33 (54.1)

3 (4.9)

35 (57.4)

30 (50.0)

Phobia

53 (86.9)

41 (67.2)

0

43 (70.5)

41 (67.2)

Social phobia

25 (41.0)

14 (23.0)

3 (4.9)

11 (18.0)

10 (16.4)

Separation anxiety

37 (60.7)

16 (26.2)

7 (11.5)

13 (21.3)

9 (14.8)

GAD

42 (68.9)

30 (49.2)

14 (23.0)

16 (26.2)

15 (24.6)

OCD

29 (47.5)

19 (31.1)

3 (4.9)

3 (4.9)

3 (4.9)

OCD-O

21 (34.4)

13 (21.3)

7 (11.5)

7 (11.5)

6 (9.8)

OCD-C

18 (29.5)

14 (23.0)

8 (13.1)

6 (9.8)

6 (9.8)

MDD

60 (98.4)

37 (60.7)

5 (8.2)

11 (18.0)

9 (14.8)

Dysthymia

60 (98.4)

37 (60.7)

21(34.4)

8 (13.1)

8 (13.1)

Mania

58 (95.1)

35 (57.4)

2 (3.3)

25 (41.0)

9 (14.8)

Hypomania

58 (95.1)

35 (57.4)

15 (24.6)

13 (21.3)

9 (14.8)

The only significant effect noted for language was with ODD (Fischer exact, p < .001). Children who were verbal were more likely to meet full ODD criteria than to be subsyndromal (38 of 42 verbal children compared to 4 of 11 nonverbal children). Those who were nonverbal were more likely to be subsyndromal (7 nonverbal children vs. 3 verbal children). No nonverbal individuals met full or subsyndromal criteria for obsessions. Therefore, significance testing could not be conducted. No participants were subsyndromal for specific phobia, so it was not included in the analyses.

Impact of IQ and Language on Symptom Frequency

Symptom Variation by IQ. Table 4 lists symptom frequencies for the entire sample and those broken up by IQ (≥70 or <70). Due to space constraints, only symptoms whose frequency rates were statistically significant are presented. Tables containing all symptoms are available by contacting the authors. The frequency of the ADHD symptom, Pushes into groups was inversely related to IQ. The ADHD symptom Blurts out answers was positively related to IQ. The specific phobia symptom Fear keeps from school was positively associated with IQ. Symptoms positively related to IQ for GAD included: Worry more than others, Hard to relax when worries, and Trouble letting go of worry. Only one depression symptom was positively associated with IQ, Depressed every day.
Table 4

Symptom variation by IQ: rate of endorsement and percentage (N = 61)

Disorder/Symptom

Entire Sample a (n = 61)

IQ ≥ 70 (n = 22)

IQ < 70 (n = 36)

χ2

ADHD

 Pushes into groups

30 (49.1)

8 (36.4)

21 (58.3)

7.86*

 Blurts out answers

34 (55.7)

17 (77.3)

16 (44.4)

6.0*

Phobia

 Fear interfere: school

26 (42.6)

15 (68.2)

9 (25.0)

10.5**

GAD

 Worry more

18 (29.5)

12 (54.5)

5(13.89)

10.89**

 Hard to relax

31 (50.8)

16 (72.7)

13 (36.11)

7.32*

 Trouble letting go

27 (44.3)

15 (68.2)

9 (25.00)

10.5**

Depression

 Depressed every day

14 (29.8)

9 (40.9)

4 (11.11)

b

* p ≤ 01, ** p ≤ .001; a Summary of item phrasing only; b Fischer exact p < .01

Symptom Variation by Language. Table 5 lists symptom frequencies for the entire sample and those broken up by language (presence/absence of conversational language). Due to space constraints, only symptoms whose frequency rates were statistically significant are presented. Tables containing all symptoms are available by contacting the authors. The following ADHD symptom frequencies were inversely related to the presence of language: Teachers say forgetful, Talking too much, and Blurts out answers. ODD symptoms included: Argues with parents, Argues with teachers, Bugs others, Blames others for mistakes, Angry a lot of time, and Gets even. The specific phobia symptom Fear keeps from school was also directly related to the presence of language. The following GAD symptoms were directly related to the presence of language: Hard to relax when worries and Trouble letting go of worry. OCD symptoms associated with language included: Symptoms are a problem at home, school, and Time spent interferes with routine >1 hour. The depression symptom, Gets into arguments and the mania symptom, Talks way too much/loudly, was also directly related to the presence of language.
Table 5

Symptom variation by language: rate of endorsement and percentage (N = 61)

Disorder/ Symptom a

Entire sample (n = 61)

No lang (n = 14)

Lang (n = 47)

χ2

ADHD

 Teachers say forgetful

35 (57.4)

3 (21.4)

32 (68.1)

9.6**

 Talking too much

22 (36.7)

1 (7.1)

21 (44.7)

6.0**

 Blurts out answers

34 (55.7)

1 (7.1)

33 (70.2)

17.39***

ODD

 Argues with parents

46 (75.4)

5 (35.7)

41 (87.2)

15.44***

 Argues with teachers

23 (37.7)

1 (7.1)

22 (46.8)

7.23**

 Bugs others

35 (57.4)

4 (28.6)

31 (66.0)

6.17**

 Blames others for mistakes

29 (47.5)

1 (7.1)

28 (59.6)

11.89***

 Angry a lot of time

23 (37.7)

0

23 (48.9)

11.0***

 Gets even

24 (39.3)

1 (7.1)

23 (48.9)

7.9**

Phobia

    

 Fear interferes: school

26 (42.6)

1 (7.1)

25 (53.2)

9.35**

GAD

 Hard to relax when worries

31 (50.8)

2 (14.3)

29 (61.7)

9.7**

 Trouble letting go of worry

27 (44.3)

1 (7.1)

26 (55.3)

10.15***

OCD

 Symptoms problem

18 (30.0)

0

18 (38.3)

b

 Interferes; >1 h

15 (25.0)

0

15 (31.9)

b

MDD

 Gets into arguments

35 (58.3)

2 (14.3)

33 (70.2)

14.58***

Mania

 Talks way too much/loudly

31 (50.8)

3 (21.4)

28 (59.6)

6.28**

* p ≤ 01, ** p ≤ .001; a Summary of item phrasing only; b Fischer exact p < .01

Behavioral Equivalents

Analyses only indicated one association among behavioral equivalents and P-ChIPS internalizing diagnoses. The item “Engages in meaningless body movements” was associated with MDD (Fischer exact, p = .01). The other items were not associated with any internalizing diagnoses.

Discussion

The purpose of this study was to contribute to the growing body of knowledge on psychiatric disorders in children with ASDs. We set out to explore systematically DSM-IV criteria using a structured interview and assessed the impact of IQ and language skills on endorsement rates. Looking at differential rates of diagnoses and symptoms among different subgroups of children can provide valuable insight on the appropriateness of the DSM as a classification system for this population. It could also potentially suggest certain modifications to existing criteria. The other objective of this study was to assess the validity of behavioral equivalents in defining anxiety and mood disorders.

Results indicated that endorsement rates were impacted by IQ and language skills. This suggests that certain modifications may be needed when applying DSM criteria to assess for comorbid disorders in individuals with ASDs. Across disorders, children with an IQ < 70 had fewer reported symptoms than those with IQ ≥ 70. This was especially true for symptoms of ADHD and GAD. As expected, those with an IQ < 70 and no conversational language were less likely to have a diagnosis of GAD. The associated physical symptoms (e.g., feel edgy, have trouble sleeping) were endorsed equally across groups. The separation occurred with questions requiring the child to voice worry. Fifty-five percent of children with an IQ ≥ 70 worried more than other children their age, compared with only 14% of those with an IQ < 70. Interestingly, 36% of parents of children with an IQ < 70 reported that their child had difficulty calming or relaxing when worrying. Because both of these symptoms are required by the P-ChIPS, only 8% of children with an IQ < 70 were diagnosed with GAD, compared to 50% of those with an IQ > 70. Ten parents (28%) of children with an IQ < 70 reported impairment from GAD symptoms but only 8% were identified with the P-ChIPS. Also at issue here is how the diagnostic criteria of GAD are interpreted. The P-ChIPS asks if the child “worries more than other kids his/her age (for example about things coming up in the future such as school, small mistakes he/she made in the past, taking a test, or seeing the dentist).” The DSM-IV-TR criteria are more vague about the nature of worries (events or activities). This may be more appropriate for children on the autism spectrum rather than the examples given in the P-ChIPS item, which appears to require an individual to have some sense of social awareness to recognize symptoms and report them.

Those with lower IQs were more likely to exhibit the ADHD symptom, Push their way into groups (58%). It is unclear if this reflects an interaction of low IQ paired with impaired social understanding often present in ASDs rather than symptoms of ADHD. This may be adversely impacting the specificity of DSM criteria. Of similar concern is the item Interrupts others, which was endorsed by 97% of the IQ < 70 group.

Surprisingly, OCD symptom endorsement was not impacted by IQ. One would have expected those with higher IQs to be more likely to express why they engaged in compulsions. Apparently, expressing these ideas is difficult for most children with ASDs, regardless of IQ. A total of 34% of the sample exhibited the symptom of having to say or do something over and over. However, only 10% of the sample met criteria for compulsions and 10% for obsessions. What remains unclear is if the individuals identified by this item on the P-ChIPS were exhibiting OCD or the repetitive behaviors associated with an ASD diagnosis. The DSM-IV-TR does not address this differential diagnosis and very little empirical work has been done to delineate this distinction. A better understanding of the distinction between repetitive behaviors and OCD will likely be necessary to better understand the pathogenesis of disorders.

Not surprisingly, the majority of significant differences with respect to language skills were obtained with symptoms which require language to be endorsed (e.g., arguments, blaming others, talking too much). Additionally, nonverbal individuals were more likely to be subsyndromal for ODD. Perhaps some modification of DSM criteria when assessing nonverbal individuals should be considered; some criteria might need to be refined or omitted and the number of criteria necessary to meet cut-off may need to be amended.

To our knowledge, this was the first study to examine behavioral equivalents in an ASD population. Results do not seem to support their use in this population. Only one item, “Engages in meaningless body movements,” was associated with MDD. However, other stereotypy and self-injury items were not associated with MDD. This significant finding may also be a result of Type 1 error as a large number of analyses were conducted. Potential behavioral equivalents seem to be indicators of overall impairment, rather than specific disorders. Of course, it may be argued that this study was underpowered for such analyses.

While administering the P-ChIPS, a number of challenges related to face and content validity came to light. Some questions had content that was not appropriate or applicable to most individuals with ASDs. This ranged from questions about grades in school to understanding social norms such as personal possession. Consideration of the child’s level of functioning, social awareness, and educational setting is needed when interpreting parent responses. A more pervasive content issue is the P-ChIPS’ assessment of impairment. Parents often had difficulty assessing/reporting on the impairment caused by symptoms of a particular disorder. They seemed to interpret most of the symptoms as part of the ASD. They had made so many accommodations for their child that they often were unable to recognize the actual impairment caused by the comorbid psychiatric symptoms.

Interpretation of these results is constrained by sample size and characteristics. A larger sample would have obviously yielded more power to the IQ and language analyses. This was especially true for the language analyses which were underpowered. It is important to stress that the current sample was comprised of clinically-referred children. We felt this was important in order to learn about the presentation of psychiatric symptoms. Base rates might have been too low to be informative in an epidemiological sample. The high rate of psychiatric problems does not come without problems. For instance, it is possible that behavior equivalent analyses were muddled because of the high rate of disruptive behavior in the sample (i.e., many children met criteria for several disorders). While this study sought to recruit participants from a wide range of treatment venues, it is unclear how well the current sample represented all ASD youngsters with significant emotional and behavioral problems. Presumably those in the behavioral intervention programs were referred largely for disruptive behaviors. It is possible that those with internalizing disorders were underrepresented in this sample.

This study was but one small step in the elucidation of the clinical picture of psychiatric disorders in ASDs. Clarification of the clinical picture will only come from a collection of studies from a wider breadth of disorders and more external variables. Further, follow-up studies examining differential outcome will be essential to this process. Additional research is clearly needed to better understand presentation of psychiatric symptoms in individuals who are nonverbal and/or have lower IQs. Careful consideration should be given to how impairment is defined and assessed in these individuals.

Notes

Author Note

This study was supported, in part, by a grant from the Ohio Department of Mental Health.

References

  1. Aman, M. G., Tassé, M. J., Rojahn, J., & Hammer, D. (1996). The Nisonger CBRF: A child behavior rating form for children with developmental disabilities. Research in Developmental Disabilities, 17, 41–57.CrossRefPubMedGoogle Scholar
  2. American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revisions. Washington: Author.CrossRefGoogle Scholar
  3. Cantwell, D. P. (1996). Classification of child and adolescent psychopathology. Journal of Child Psychology and Psychiatry, 37, 3–12.CrossRefPubMedGoogle Scholar
  4. de Bruin, E. I., Ferdinand, R. F., Meester, S., de Nijs, F. A., & Verheij, F. (2007). High rates of psychiatric co-morbidity in PDD-NOS. Journal of Autism and Developmental Disorders, 37, 877–886.CrossRefPubMedGoogle Scholar
  5. Fletcher, R., Loschen, E., Stavrakaki, C., & First, M. (2007). Diagnostic Manual Intellectual Disability: A Textbook of Diagnosis of Mental Disorders in Persons with Intellectual Disability. Kingston: NADD Press.Google Scholar
  6. Gadow, K. D., DeVincent, C. J., Pomeroy, J., & Azizian, A. (2005). Comparison of DSM-IV symptoms in elementary school-age children with PDD versus clinic and community samples. Autism, 9, 392–415.CrossRefPubMedGoogle Scholar
  7. Gadow, K. D., DeVincent, C. J., & Pomeroy, J. (2006). ADHD symptom subtypes in children with pervasive developmental disorder. Journal of Autism and Developmental Disorders, 36, 271–283.CrossRefPubMedGoogle Scholar
  8. Gadow, K. D., Roohi, J., DeVincent, C. J., et al. (2008a). Association of ADHD, tics, and anxiety with dopamine transporter (DAT1) genotype in autism spectrum disorder. Journal of Child Psychology and Psychiatry, 49, 1331–1338.CrossRefGoogle Scholar
  9. Gadow, K. D., DeVincent, C. J., & Drabick, D. A. G. (2008b). Oppositional defiant disorder as a clinical phenotype in children with autism spectrum disorder. Journal of Autism and Developmental Disorders, 38, 1302–1310.CrossRefGoogle Scholar
  10. Gadow, K. D., DeVincent, C., & Schneider, J. (2008c). Predictors of psychiatric syndromes in children with autism spectrum disorders. Journal of Autism and Developmental Disorders, 38, 1710–1720.CrossRefGoogle Scholar
  11. Lecavalier, L. (2006). Behavior and emotional problems in young people with pervasive developmental disorders: Relative prevalence, effects of subject characteristics, and empirical classification. Journal of Autism and Developmental Disorders, 36, 1101–1114.CrossRefPubMedGoogle Scholar
  12. Lecavalier, L., Aman, M. G., Hammer, D., Stoica, W., & Matthews, G. L. (2004). Factor analysis of the Nisonger Child Behavior Rating Form in children with autism spectrum disorders. Journal of Autism and Developmental Disorders, 34, 709–721.CrossRefPubMedGoogle Scholar
  13. Lecavalier, L., Gadow, K., DeVincent, C. J., & Edwards, M. C. (2008). Validation of DSM-IV model of psychiatric syndromes in children with autism spectrum disorders. Journal of Autism and Developmental Disorders, 39, 278–289.CrossRefPubMedGoogle Scholar
  14. Leyfer, O. T., Folstein, S. E., Bacalman, S., Davis, N. O., Dinh, E., Morgan, J., et al. (2006). Comorbid psychiatric disorders in children with autism: Interview development and rates of disorders. Journal of Autism and Developmental Disorders, 36, 849–861.CrossRefPubMedGoogle Scholar
  15. Martson, G. M., Perry, D. W., & Roy, A. (1997). Manifestations of depression in persons with developmental disabilities. Journal of Intellectual Disability Research, 41, 476–480.CrossRefGoogle Scholar
  16. Matson, J. L., Rush, K. S., Hamilton, M., Anderson, S. J., Bamburg, J. W., Baglio, C. S., et al. (1999). Characteristics of depression as assessed by the Diagnostic Assessment for Severely Handicapped-II (DASH-II). Research in Developmental Disabilities, 20, 305–313.CrossRefPubMedGoogle Scholar
  17. McBrien, J. A. (2003). Assessment and diagnosis of depression in people with intellectual disability. Journal of Intellectual Disability Research, 47, 1–13.CrossRefPubMedGoogle Scholar
  18. Posey, D. J., Aman, M. G., McCracken, J. T., Scahill, L., Tierney, E., Arnold, L. E., et al. (2007). Positive effects of methylphenidate on inattention and hyperactivity in pervasive developmental disorders: An analysis of secondary measures. Biological Psychiatry, 61, 538–544.CrossRefPubMedGoogle Scholar
  19. Robins, E., & Guze, S. B. (1970). Establishment of diagnostic validity in psychiatric illness: Its application to schizophrenia. American Journal of Psychiatry, 126, 983–987.PubMedGoogle Scholar
  20. Roid, G. H. (2003). Stanford-Binet Intelligence Scales, Fifth Edition. Itasca: Riverside Publishing.Google Scholar
  21. Royal College of Psychiatrists. (2001). Diagnostic Criteria for Psychiatric Disorders for use with Adults with Learning Disabilities/Mental Retardation. London: Gaskell.Google Scholar
  22. Rutter, M., LeCouteur, A., & Lord, C. (2003). Autism Diagnostic Interview-Revised. Los Angeles: Western Psychological Services.Google Scholar
  23. Simonoff, E., Pickles, A., Charman, T., Chandler, S., Loucas, T., & Baird, G. (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 and Adolescent Psychiatry, 47, 921–929.CrossRefPubMedGoogle Scholar
  24. Tsiouris, J. A., Mann, R., Patti, P. J., & Sturmey, P. (2003). Challenging behaviors should not be considered as depressive equivalents in individuals with intellectual disability. Journal of Intellectual Disability Research, 47, 14–21.CrossRefPubMedGoogle Scholar
  25. Weller, E. B., Weller, R. A., Teare, M., & Fristad, M. A. (1999). Parent Version Children’s Interview for Psychiatric Syndromes (P-ChIPS). Washington: American Psychiatric Press.Google Scholar
  26. Weller, E. B., Weller, R. A., Fristad, M. A., Rooney, M. T., & Schecter, J. (2000). Children’s Interview for Psychiatric Symptoms (ChIPS). Journal of the American Academy of Child and Adolescent Psychiatry, 39, 76–84.CrossRefPubMedGoogle Scholar
  27. Witwer, A. N. (2009). Psychopathology in youngsters with Autism Spectrum Disorders (Doctoral dissertation). Available from Ohio Link ETD. (Accession Number: osu1243561855).Google Scholar

Copyright information

© Springer Science+Business Media, LLC 2010

Authors and Affiliations

  1. 1.Nisonger Center and Department of PsychologyOhio State UniversityColumbusUSA

Personalised recommendations