The Behavioral Phenotype of 7q11.23 Duplication Syndrome Includes Risk for Oppositional Behavior and Aggression

Abstract

Objective:

7q11.23 duplication syndrome (Dup7) is a genetic disorder with a variable phenotype associated with cognitive and behavioral characteristics including a high incidence of expressive language difficulties, social anxiety, and oppositional or disruptive behavior. Correlates of aggression and oppositionality were examined.

Method:

Participants were 63 children with genetically confirmed Dup7 between the ages of 4 and 17 years. A multi-method, multi-informant approach was used to assess aggression and oppositional behavior, and the contributions of cognitive functioning, expressive language, and autism spectrum, social anxiety, and hyperactivity/impulsivity symptomatology were considered.

Results:

Elevated levels of aggression and oppositional behavior were found. Cognitive functioning, expressive language, and ASD symptomatology were not significantly related to parent ratings of aggression, although young children who had language and nonverbal cognitive delays were most likely to demonstrate examiner-observed aggression. Social anxiety and hyperactivity/impulsivity symptomatology were related to defiant/aggressive and oppositional behavior.

Conclusion:

Genes in the 7q11.23 region duplicated in Dup7, in transaction with the environment, may contribute to aggressive and oppositional behavior.

Studies of individuals with disorders of known genetic etiology associated with characteristic phenotypes have the potential to contribute to the elucidation of relations between genetics and behavior and to inform investigations on chromosomal regions that may confer risk for particular phenotypes. 7q11.23 duplication syndrome (Dup7) is a recently identified genetic disorder caused by an extra copy of the same set of genes that is deleted in Williams syndrome (WS) with a prevalence of 1 in 7,500 live births.¹ Dup7 is associated with distinctive facial features, macrocephaly, hypotonia, cardiovascular abnormalities (especially aortic dilation), neurologic abnormalities including seizures, speech and language delay, mild to severe cognitive delay, and elevated risk of attention deficit/hyperactivity disorder, autism spectrum disorders (ASD) and anxiety disorders (social phobia, selective mutism, and specific phobia).^2,3 Speech and expressive language delays are described in nearly all case reports^1,3, and aggression and oppositional behavior are often reported.³ As described below, language problems, social anxiety, ADHD, and cognitive difficulties are associated with aggression and oppositional behavior both in children with other disabilities and in children in the general population. Detailed characterization of the prevalence and severity of aggression and oppositionality in children with Dup7 and their relationship to other characteristic aspects of the behavioral phenotype is needed.

Many of the characteristics included in the Dup7 behavioral phenotype have been found to be associated with increased rates of aggression and oppositionality in the broader population. In a meta-analysis, children who had language difficulties were found to have higher rates of behavior problems, including aggressive and oppositional behaviors, relative to typically developing children⁴. Based on a very large national sample, children and adolescents with ID were more likely to meet criteria for oppositional deviant disorder (ODD) than were children in the general population⁵. In an epidemiological sample, children and adults with ID who had mild ID were significantly less likely to evidence challenging behavior than individuals with moderate – profound ID⁶. In the same sample, individuals who either had no expressive language or who produced only single words were more likely to have challenging behavior than individuals who produced phrases or sentences⁶. In contrast, in a sample of 6 – 14-year-olds, the presence of aggression did not differ significantly as a function of language level⁷. Aggression was particularly common among children and adolescents with ASD, with parents of 68% reporting aggression toward caregivers and 49% reporting aggression toward non-caregivers⁸. For children with ASD, aggression was significantly negatively related to chronological age; the relation with language ability was not statistically significant. Aggression also was common among children and adolescents with ADHD, with most studies indicating that ~50% have a co-morbid diagnosis of ODD⁹. Children with hyperactivity-impulsivity were particularly likely to demonstrate physical aggression¹⁰. Finally, although individuals with social anxiety disorder are generally behaviorally inhibited and risk-averse,¹¹ characteristics usually associated with lower rates of aggression, at a population level, children with high levels of anxiety-depressive trait are very likely also to have high levels of trait aggression¹².

While aggression has been noted in some case studies of Dup7,^13–15 detailed characterization of aggression and/or oppositionality is not available. Further, language delays or difficulties, lower cognitive functioning, ASD, social anxiety, and ADHD are vulnerabilities for children with Dup7 and represent potential risk factors for the presence of oppositionality and/or aggression. However, there is little research examining interrelationships between level of cognitive functioning or severity of ASD symptomatology and oppositionality and/or aggression within the populations studied. Especially when variability is present, better understanding of these relationships may be useful to practitioners guiding families. The current study used a multimethod, multi-informant approach, with both parent report and examiner-based parent interview, as well as examiner observation, to simultaneously take a dimensional and a categorical approach to describing aggression and/or oppositionality in children with Dup7 as well as constructs that may confer vulnerability to these difficulties. We hypothesized that oppositional behavior and aggression would be commonly reported by parents and observed by examiners and we explored whether oppositionality and/or aggression was significantly related to cognitive functioning, expressive language, ADHD symptomatology, ASD symptomatology, and social anxiety symptoms.

Method

Participants and Procedure

Participants were 63 children with Dup7 who participated in a larger study designed to comprehensively describe the cognitive, behavioral, and medical phenotype of children with Dup7. All children had genetically-confirmed classic Dup7 (duplication of the 25–27 protein-coding genes in the classic WS region). Inclusion criteria were that the child had a 7q11.23 duplication that included all and only the 25 – 27 genes in the classic WS region, was between ages 4.00 and 17.99 years at the time of the assessment, and had English as a native language or was fluent in English. Children were excluded if they had an additional copy number variant or syndrome associated with intellectual disability and/or ASD. Participants were recruited by referrals from geneticists or other physicians, the family support organization for Dup7 (Duplication Cares), or the child’s parent or legal guardian.

This research was reviewed and approved by the Institutional Review Board at the University of Louisville. Parents or legal guardians of all participants provided written informed consent and participants provided assent as appropriate. Assessments took place over three days and included evaluations by psychologists, speech-language pathologists, and a clinical geneticist. All measures were administered according to their standardized procedures. This is the same group of children as was reported in Klein-Tasman and Mervis (2018), where the ASD diagnostic procedures are described in detail.¹⁶ As reported there, 19% met clinical diagnostic criteria for an ASD based on gold-standard diagnostic assessment practices that included administration of the Autism Diagnostic Interview- Revised¹⁷, the Autism Diagnostic Observation Scale – second edition¹⁸, assessment of cognitive functioning, and clinical judgment of a Licensed Psychologist with expertise in ASD assessment. Unless otherwise noted, all measures listed below were administered to all participants. Parents or legal guardians were asked to list their child’s medications and reasons for taking them in open-ended format. All diagnoses resulted from the study procedures (rather than reflecting diagnoses by outside practitioners).

Measures

Differential Ability Scales-Second Edition (DAS-II).¹⁹ The DAS-II is a comprehensive measure of cognitive abilities (mean = 100, SD = 15) for children ages 2½ - 17 years. It yields an overall composite score, or General Conceptual Ability (GCA), which is equivalent to a full-scale IQ and based on the DAS-II’s six core subtests. The Special Nonverbal Composite (SNC) was derived for all participants from the four Nonverbal Reasoning and Spatial core subtests and provides an assessment of cognitive functioning separate from language and/or speech problems.

Expressive Language Index (mean = 100, SD = 15) from the Clinical Evaluation of Language Fundamentals- Preschool, Second Edition (CELF-Preschool-2; ages 4–5.99 years) ²⁰ or Clinical Evaluation of Language Fundamentals Fourth Edition (CELF-4; ages 6.0–17.99 years)²¹ that both measure expressive vocabulary and grammar.

Autism Diagnostic Interview-Revised (ADI-R).¹⁷ The ADI-R is a semi-structured standardized clinical interview containing items focusing on behaviors in three content areas: quality of social interaction; communication and language; and repetitive, restricted, and stereotyped interests and behavior. Responses are scored by the clinician according to operational definitions based on the caregiver’s description. Items relating to aggression (Item 81: Aggression toward Caregivers or Family Members; Item 82: Aggression toward NonCaregivers or NonFamily Members) were examined for presence and severity of aggression. Item 81 is coded for episodes of aggression toward family and caregivers of sufficient severity and/or frequency to constitute a significant cause for concern. Similarly, Item 82 is coded for episodes of aggression directed toward individuals who are not caregivers or family members, including both children and adults. Codes range from 0 – 3 [0 = no aggression or only rare episodes; 1 = “mild aggressiveness only (e.g., threatening without physical contact, or momentary, provoked lashing out);” 2 = “definite physical aggression involving hitting or biting but no use of implements);” 3 = “violence that involves the use of implements”)]. Separate codes are assigned based on the caregiver’s description of current behavior (over the past 3 months) and history of ever showing the behavior. The ADI-R was administered by one of two clinicians, both of whom had achieved research reliability.

Anxiety Disorders Interview Schedule for DSM-IV: Parent Interview Schedule (ADIS-P).²² Primary caregivers completed the ADIS-P, a semi-structured interview assessing current anxiety and related disorders, including externalizing disorders, in children and adolescents. The Social Phobia (Social Anxiety Disorder), Disruptive Behavior Disorders (DBD; including ODD), and ADHD sections that include hyperactivity/impulsivity (H/I; ADHD combined and ADHD H/I) were examined. The ADIS-P was administered by one of two trained personnel and interviews were reviewed by a Licensed Psychologist with anxiety diagnostic expertise.

Conners Early Childhood (Conners EC)²³ or Conners Comprehensive Behavior Rating Scales (Conners CBRS).²⁴ The Conners EC (used here for ages 4 – 5 years; 1 participant with missing data) and the Conners CBRS (used here for ages 6 – 17 years; 2 participants with missing data) were completed by primary caregivers and assess behaviors, emotions, and social problems using a questionnaire format. Ratings on the Defiant/Aggressive Behaviors scale were examined for children administered the Conners EC. Ratings on the Defiant/Aggressive Behaviors (DAB), Oppositional Defiant Disorder (ODD), ADHD Predominantly Hyperactive-Impulsive Type (ADHD-H/I), and Social Phobia scales were examined for those administered the Conners CBRS. All available data were used for each analysis. The Conners measures yield T-scores (M = 50, SD = 10) for each scale. Based on the test author’s classifications, T-scores < 40 are considered Low, from 40 – 59 Average, from 60 – 64 High Average, from 65 – 69 Elevated, and ≥ 70 Highly Elevated.

Autism Diagnostic Observation Schedule – 2^nd Edition (ADOS-2).¹⁸ Children were administered the appropriate module from the ADOS-2 (Module 1, n = 9; Module 2, n = 21; Module 3, n = 33). The ADOS-2 is a semi-structured, standardized assessment of socio-communication, social interaction, play/imaginative play, and restricted and repetitive behaviors or interests, considered the “gold standard” observational assessment for diagnosing ASD. ADOS Classifications of nonspectrum, autism spectrum, and autism are determined based on empirically-derived cutoffs. ADOS-2 Comparison Scores, indicating the severity of ASD symptomatology as observed during the ADOS-2 assessment, were computed; these are calibrated relative to children who have ASD, with consideration of chronological age and language level. Item E2 was examined, reflecting aggression toward the examiner during ADOS-2 administration. This item is coded as follows: 0 = no disruptive, destructive, negative, or aggressive behavior; 1 = mild disruption, anger, aggression, or negative behavior (verbal threats, swearing, deliberately loud voice); 2 = more than one intentionally disruptive or negative incident (loud talking or repeated swearing); 3 = marked or repeated temper tantrums or significant aggression (throwing things, hitting, biting others, screaming or yelling).

Analytic Plan

The data were analyzed using IBM SPSS version 27. Nonparametric statistics were used when variables were not continuous or group size was uneven. To identify the contributors to parent-reported defiance/aggression and oppositionality, two multiple linear regressions were conducted, with the independent variables entered at the same time. All multiple linear regression assumptions were met, and no outliers were identified.

Results

Sample Characteristics

See Table 1 for demographic information. Thirty-two participants were currently taking no prescription medications. Twelve participants were taking medications for ADHD, 12 for anxiety, and 12 for mood/aggression/behavior. Seven participants were currently prescribed medications for seizures, and one additional participant had been prescribed seizure medication in the past. Descriptive statistics for the standardized assessments are presented in Table 2.

Table 1.

Participant Demographics

Measure	Categories	Descriptive Statistics
Chronological Age		M: 8.64 years, SD: 3.77, Range: 4.01 – 17.70
Sex	Female	26 (41.3%)
	Male	37 (58.7%)
Race/Ethnicity	White non-Hispanic	46 (73.0%)
	White Hispanic	4 (6.3%)
	Asian non-Hispanic	3 (4.8%)
	Black – African American non-Hispanic	2 (3.2%)
	Multiracial non-Hispanic	8 (12.7%)
Maternal Education	High school diploma or GED	8 (12.7%)
	Associate degree or some college	20 (31.7%)
	Bachelor degree	25 (39.3%)
	Advanced degree	10 (15.9%)

Table 2.

Descriptive Statistics for Standardized Assessments

Scale	n	Mean	SD	Range	Percent with Difficultya
DAS-II General Conceptual Ability (GCA) SS	63	79.38	19.45	33 – 132	46.0
DAS-II Special Nonverbal Composite (SNC) SS	63	80.22	20.34	31 – 139	36.5
CELF Expressive Language Index SS	63	72.59	20.59	45 – 116	61.9
Conners Defiant/Aggressive Behaviors T	60	57.12	15.88	36 – 90	26.7
Conners Oppositional Defiant Disorder (ODD) T	44	61.36	15.61	40 – 90	43.2
Conners Social Phobia T	44	73.02	15.20	39 – 90	75.0
Conners ADHD-Hyperactive/Impulsive T	44	66.52	15.63	43 – 90	50.0

Note: DAS-II = Differential Ability Scales-II; CELF = Clinical Evaluation of Language Fundamentals; SS = standard score (M = 100, SD = 15) with lower scores indicating more difficulty; T = T-score (M = 50, SD = 10) with higher scores indicating more difficulty. Further information about score ranges can be found in the respective measure technical manuals.

^aDAS-II standard score of 79 or below (test author’s Low or Very Low classifications, CELF standard score of 77 or below (test authors’ Low or Very Low classifications), or Conners T-score of 65 or above (test author’s Elevated or Very Elevated classifications)

Presence and Severity of Defiance/Aggression and Oppositionality in Children with Dup7

Parent-reported questionnaire measures.

One-sample Wilcoxon signed rank tests indicated significantly higher T-scores than the normative mean (T = 50) for both DAB (Z = 2.35, p = .019) and ODD (Z = 3.70, p < .001). ODD T-score was significantly higher than DAB T-score (t (42) = 3.39, p = .002). Figure 1 details the distribution of parent-reported aggression levels on the Conners scales. Very elevated scores (expected for 3% of the general population) were observed for 20% of children with Dup7 on the DAB scale and 38% on the ODD scale.

Figure 1.

Percentage of sample (with frequencies indicated) at different levels of parent-reported aggression on the Defiant/Aggressive Behaviors scale on the Conners Early Childhood (for children aged 4 – 5 years)/Conners Comprehension Behavior Rating Scales (for children aged 6 – 17 years) and on the Oppositional Defiant Disorder (ODD) scale on the Conners Comprehensive Behavior Rating Scales (for children aged 6 – 17 years). General Population distribution, which is the same for both Conners scales, is also indicated. The T-score ranges corresponding to each of the five levels are < 40 for Low, 41–59 for Average, 60 – 64 for High Average, 65 – 69 for Elevated, and ≥ 70 for Very Elevated. The frequency of participants falling within each range is indicated within the figure.

Parent interview.

ADI-R parent interview aggression item ratings are detailed in Table 3. Thirty-six children (57%) were reported as currently showing some degree of aggression toward family members and 15 (24%) toward non-family members. Four additional children did not show aggression currently but had evidenced some degree of aggression toward family members in the past. Children showing higher levels of aggression toward family members were likely to show higher levels of aggression toward non-family members both current (rho = .40, n = 63, p = .001) and ever (rho = .53, n= 63, p = .001) aggression ratings. Eighty percent (12/15) of the children who currently demonstrated aggression toward non-family members also demonstrated aggression toward family members. More than 1/3 of the participants (39.7%) met criteria for a DBD diagnosis based on the ADIS.

Table 3.

ADI-R Ratings of Parent-Reported Aggressive Behavior

Code	Description	Family		Non-family		Family or non-familya
		Current	Ever	Current	Ever	Current	Ever
0	No Aggression	43%	37%	76%	67%	38%	32%
1	Mild Aggression	12%	14%	6%	6%	16%	17%
2	Definite Physical Aggression	40%	38%	16%	24%	41%	40%
3	Violence with Implements	5%	11%	2%	3%	5%	11%

^aThis column reflects the highest level of aggression (whether to family member or non-family member)

Examiner observation.

Eleven children (17%) demonstrated aggression during the ADOS-2 assessment. On Item E2, seven (11%) children (Module 1 n = 2/9; Module 2 n = 4/21; Module 3 n = 1/33) were given a rating of 1 for displaying mild disruption, anger, aggression or negative behavior; three children (5%) (Module 1 n = 3/9) were given a rating of 2 for being intentionally disruptive, and one child (2%) (Module 2 n = 1/21) was given a rating of 3 for displaying marked or repeated temper tantrums or significant aggression.

Contributors to the Presence and Severity of Aggression in Children with Dup7

Age.

There were no significant correlations between dimensional parent-reported norm-referenced ratings for aggression (Conners DAB, ODD T-scores) and age (all rs < .08, ps > .51) and no significant difference in DAB T-scores between parent ratings on the Conners EC (M = 58.27, SD = 16.82) and parent ratings on the Conners CBRS (M = 56.18, SD = 15.56), t(59) = 0.82, p = .401. On the categorical measures of aggression from the ADI-R, a significant positive correlation was found between age and Caregiver “ever” aggression (rho = .30, n = 63, p = .016). No significant correlations with age were found for the other ADI-R aggression variables (ps > .30). Children who showed aggression during the ADOS had a significantly younger chronological age distribution than those who did not (z = −2.79, p = .005). No effect of age was seen for categorical DBD diagnosis (z = 1.10, p = .273).

Sex.

No significant sex differences were found for T-scores on the DAB scale, t(59) = 1.05, p = .296 or the ODD scale, t(43) = −0.19, p = .85. Similarly, results of χ² tests indicated no significant sex differences on the ADI-R categorical measures of aggression (all ps > .60). No significant relation between sex and examiner-observed aggression (ADOS-2 E2 item, p = .744) or categorical DBD diagnosis (p = .439) was observed.

Cognitive and expressive language functioning.

No significant correlations were observed between DAB T-score and cognitive (DAS-II) and expressive language (CELF) SSs (all rs < .15, ps > .50). Significant negative correlations with DAS-II SNC were found for both ADI-R Caregiver “current” (rho = - .25, n = 63, p = .047) and “ever” (rho = - .25, n = 63, p = .041) aggression, indicating that children with weaker nonverbal abilities relative to their age-peers evidenced more aggression toward family members than did children with stronger nonverbal abilities. All other correlations for ADI-R Caregiver items were not statistically significant (rhos < .19, ps > .15) as were correlations with Non-caregiver items (rhos < .05, ps > .51). Mann-Whitney U tests indicated that children who showed aggression during the ADOS had significantly lower GCA (z = −2.47, p = .013), and SNC (z = −3.16, p = .002) distributions than those who did not. The two groups did not differ significantly on CELF Expressive Language SS distributions (z = −1.14, p = .253). However, a Fisher exact test indicated that the proportion of children who showed aggression during the ADOS-2 was significantly larger for children who completed Module 1 than for children who completed Module 2 and 3 (p = .006); age and module were inherently confounded as participants administered Module 1 were also significantly younger than those administered Module 2 or 3, z = 3.30, p < .001.

Social anxiety.

Significant positive correlations were found between Social Phobia T-scores and DAB (r = .38, n = 44, p = .013) and ODD T-scores (r = .41, n = 44, p = .006), indicating that children with more aggression-related symptoms also evidenced more social anxiety symptoms. Thirty-four of 63 participants (54%) were diagnosed with Social Anxiety Disorder. No significant relation was observed between ADIS-P oppositionality-related diagnosis (either DBD or ODD; n = 25) and diagnosis of Social Anxiety Disorder [x² (1, n = 63) = 0.27, p = .603]. No significant relationship between aggression on the ADOS and social anxiety was observed (p = .526).

ADHD H/I symptomatology.

Significant positive correlations were observed between ADHD H/I T-scores and both DAB T-scores (r =.50, n = 43, p = .001) and ODD T-scores (r =.47, n = 44, p = .001). A significant relation was observed between ADIS-P ODD/DBD diagnosis (n = 25 of 63) and any ADHD diagnosis involving hyperactivity/impulsivity (n = 27 of 63) [x² (1, n = 63) = 10.70, p = .001]. No significant relationship between aggression on the ADOS and ADHD diagnosis involving hyperactivity/impulsivity was found (p = .425).

ASD symptomatology.

No significant correlations were found between ADOS-2 ASD classification (all ps > .5), ASD severity score (all ps > .5), or ASD clinical diagnosis (all ps > .3) and T-scores on any of the aggression-related Conners scales. In addition, there were no significant differences in frequency of ASD diagnosis as a function of the presence of an oppositionality-related diagnosis (either DBD or ODD) (x²(1, n = 63) = 0.59, p = .617. No significant relationship between aggression on the ADOS and ASD classification (p = .394) or ASD diagnosis (p = .432) was observed.

Contributions to parent-reported defiance/aggression and oppositionality.

For children ages 6 years and older (for whom parent ratings of social anxiety, ADHD-H/I symptomatology, and oppositional/aggressive behavior were all available), linear regressions were used to assess the contributions of social anxiety and hyperactivity/impulsivity symptoms to individual differences in oppositional/defiant and aggressive behavior. Because there were no significant sex differences and no significant effects of age, expressive language, or GCA in the bivariate analyses, these variables were not included in the regressions. Conners Social Phobia T-score and Conners ADHD-H/I T-score together explained 34.4% of the variance in Conners DAB T-score. ADHD-H/I T-score (B = .461, p = .001) uniquely accounted for 20.34% and Social Phobia T-score (B = .312, p = .022), uniquely accounted for 9.24% of the total variance. Social Phobia T-score and ADHD-H/I T-score together explained 33.5% of the variance in ODD T-score, with both ADHD-H/I (B = .416, p = .002) and Social Phobia T-scores (B = .342, p = .011) making statistically significant contributions, uniquely accounting for 16.89% and 11.42% of the total variance respectively.

Discussion

In the current study, we provided a detailed quantitative characterization of oppositionality and defiant/aggressive behavior in a substantial sample of children with Dup7 and explored relationships to social anxiety symptomatology, ADHD H/I symptomatology, ASD symptomatology, cognitive functioning, and expressive language abilities. A notable proportion of children with Dup7 showed elevated rates of aggression and oppositional behavior on dimensional and categorical measures. For the categorical measures, age and nonverbal cognitive functioning were related to both parental reports of aggression and observations of aggression during a brief play interaction with the examiner. While norm-referenced scores from dimensional measures of aggression/oppositionality did not show significant correlations with age, sex, ASD symptom severity, overall cognitive abilities, or expressive language abilities, both ADHD hyperactivity/impulsivity symptomatology and social anxiety symptomatology were found to significantly contribute to parental ratings of defiance/aggression and oppositionality.

Presence and Severity of Aggression

Children with Dup7 displayed elevated rates of aggression and oppositionality on a parent reported norm-referenced dimensional measure of aggression and oppositional behavior. On the ODD scale, which reflects argumentative, defiant, and non-compliant behaviors, 43% of children were rated as demonstrating behaviors at an elevated or very elevated level. In contrast, only 27% of children showed elevated or very elevated levels on the DAB scale, which measures more markedly aggressive behavior. Hence, children with Dup7 tend to demonstrate more argumentative or noncompliant behaviors than violent, physically aggressive behaviors. More than 1/3 met diagnostic criteria for a DBD diagnosis. These rates of oppositionality and aggression for children with Dup7 are higher than previously reported for individuals with ID of mixed etiology (13%),⁵ ASD (22%),²⁵ or both ASD and ID (15–18%).²⁶

Aggression was evident within the context of categorical interview and observation-based measures (ADI-R, ADOS-2). On the categorical structured interview measure, aggression was reported by parents at a rate similar to that seen for children and adolescents with ASD.²⁷ Over half of parents indicated that their child engaged in some form of aggressive behavior toward family members currently, and rates of past aggression were even higher, with nearly two-thirds of parents indicating that their children had demonstrated some form of aggressive behavior toward family members either currently or in the past. Children who demonstrated aggression toward a family member were more likely to demonstrate it toward a non-family member. The majority of parents who reported the presence of aggression described definite physical aggression (e.g., hitting or biting), rather than mild aggression (e.g., threatening without physical contact). Expressions of aggression were also evident for some children based on examiner observational coding during a structured play interaction, especially for young children who had very low nonverbal abilities and were either pre-verbal or using single words to communicate. Eighteen percent of children were rated by examiners as showing some degree of aggression during the administration of the ADOS-2, with many of those children evidencing mild disruption, anger, aggression, or negative behavior rather than physical aggression. It is not surprising that rates of aggression were lower based on direct observation than based on parental report; the reduced observation and interaction time that clinicians have compared to parents, the novel context of interaction with a new communicative partner, and the generally different relationship of children with family members and professionals make it less likely that aggression will be captured by the observational measure.

Contributions to Aggression and Oppositionality

A second aim of this study was to identify relations of cognitive functioning, expressive language abilities, age, ADHD H/I symptomology, social anxiety, and ASD symptomatology to levels of aggression in children with Dup7. On dimensional norm-referenced measures, social anxiety and hyperactivity/impulsivity showed relations to defiance/aggression and oppositionality. On categorical measures, nonverbal cognitive functioning showed relations to both observed aggression and parent report of aggressive acts. On the ADOS, younger age, weaker nonverbal cognitive functioning, and weaker spontaneous language are confounded in our sample; hence we cannot disentangle the possible contributions of younger age, poorer cognitive functioning, or weaker spontaneous language communication skills to the presence of aggression. Each of these factors is consistent with prior literature, in which increased rates of aggression among younger children²⁸ and individuals with low IQ⁵ and vulnerability to externalizing problems among children with language difficulties⁴ have been reported. No sex differences or relations to ASD severity were observed. Relations to age were complex.

As a group, younger children with Dup7 were more likely to demonstrate aggression during the ADOS-2 assessment. In fact, aggression during the ADOS was only observed for children under 9 years old. Similarly, only one child over 8 showed aggression currently based on the ADI-R. Hence, when aggression is shown it is unlikely to be seen in older children and adolescents. At the same time, no significant effect of age on T-scores from norm-referenced parental ratings of aggression was found, suggesting that standing relative to same-aged peers in aggressive symptomatology may not change substantially with age. Similarly, cognitive functioning, particularly nonverbal cognitive functioning, showed significant relations to examiner-observed aggression and aggression based on parental interview, but was not significantly related to norm-referenced parental ratings of defiance/aggression or oppositionality. Performance on standardized measures of expressive language did not show relations to measures of aggression, consistent with meta-analytic findings that failed to find such a relation for other groups.⁷ However, over half of children administered Module 1 demonstrated some aggression as observed by the examiner. Children who were pre-verbal or using single words to communicate, as a group, demonstrated examiner-observed aggression at elevated rates compared to children using phrase speech or fluent language. It should be noted that the younger children in this sample were also more likely to have more significant nonverbal cognitive and language delays, such that age, language ability, and degree of delay were confounded. The role of language functioning in aggression in Dup7 remains unclear and relations to age were inconsistent.

ADHD H/I symptomatology showed relations to parent-reported aggression based on both dimensional and categorical measures. Children who met diagnostic criteria for oppositionality-related diagnoses (either DBD or ODD) were significantly more likely also to meet criteria for ADHD H/I or Combined. This is consistent with well-established previous research demonstrating aggression as a correlated and distinct associating feature of ADHD H/I.²⁹ ADHD hyperactive/impulsive symptomatology uniquely accounted for approximately 20% of the variance in defiant/aggressive behavior and 17% of the variance in oppositional behavior. It has long been hypothesized that emotional dysregulation in children with ADHD H/I or Combined can partially explain these children’s high rates of aggressive behavior,³⁰ and this is likely also the case for children with Dup7.

Social anxiety symptomatology was also significantly related to aggression and oppositional behavior and uniquely accounted for approximately 9 and 11% of the variance in defiance/aggression and oppositionality respectively. In contrast, categorical diagnoses of ODD/DBD and Social Anxiety Disorder were not significantly related. The categorical approach used an interviewer-based measure in which the clinician interviewing is able to integrate working knowledge of diagnostic criteria and consider the context of reported behaviors. Clinicians may consider that a child’s non-compliant behavior in social situations derives from social anxiety rather than oppositionality. While in the general population social anxiety has often been associated with low levels of violence and aggression,³¹ for individuals with ASD, social anxiety has been shown to increase hostility and aggression in adults and has shown a strong relationship with physical aggression in children and adolescents.³² Anxiety in early childhood has been shown to predict aggression later in childhood,³³ and hostile and aggressive behaviors also predict internalizing emotional difficulties later in childhood.³⁴ Emerging evidence from theoretical developmental psychopathology literature suggests several mechanisms associated with the fight-flight response that are common to both anxiety and aggression that could explain the observed co-occurrence of anxiety and aggression.³⁵ Berkowitz posited that while the flight response is a conscious experience of fear, the activation of the fight response gives rise to feelings of irritation and anger.³⁶ In this model, internalizing experiences, such as fear or anxiety, precede aggressive behavior.

In sum, for children with Dup7, the vulnerability to ADHD (particularly hyperactive/impulsive symptomatology) and social anxiety difficulties, possibly together with early language difficulties and/or very limited nonverbal cognitive abilities, may converge as a perfect storm for the development of aggression. The findings from the current study provide clear evidence for a role of ADHD H/I and social anxiety symptomatology in aggression in children with Dup7, evidence that both younger age and lower nonverbal cognitive functioning may be vulnerability factors, and some suggestive evidence (based on observational assessment) for a role of very limited levels of expressive language in aggression in children with Dup7. It is notable that difficulties with aggression are not described as a core vulnerability for children with Williams syndrome, a condition that involves deletion of one copy of the same set of genes as is duplicated in Dup7. Children with Williams syndrome also show elevated rates of ADHD and expressive language difficulties (albeit with a different profile than in Dup7), but do not show social anxiety.³⁷ Further elucidation of the role of social anxiety in aggression/oppositionality for children with Dup7 is warranted.

Limitations and Future Directions

There are several limitations to the current work. First, while the sample size is substantial in the context of the current literature, it is relatively small given the large age span included. In cases in which significant interrelations were not observed (e.g., age, sex, cognitive or language functioning) this may be due to lack of sufficient power. Further, future directions with a larger sample could include exploration of the role of specific medical comorbidities (e.g., seizures), medication, and other treatment experiences on the observed phenotype as well as potential mediating variables, especially emotion regulation. Second, as Dup7 is a recently identified syndrome that requires detailed genetic testing, there may be an ascertainment bias in favor of children with more significant difficulties, as their parents may be more likely to seek medical attention and are more likely to be referred to a geneticist. Finally, this work is cross-sectional. Researchers studying children in the general population have differentiated subgroups of children based on patterns of low and high aggression, as well as declining and increasing rates of aggression with age.³⁸ Longitudinal work will be important to provide information about the trajectories of aggressive behavior among children with Dup7. Furthermore, longitudinal study may elucidate the predictive utility of social anxiety- and ADHD H/I-related difficulties for the degree of aggression.

Conclusion

This work contributes to the scarce Dup7 behavioral phenotype literature and to the literature linking social anxiety and ADHD H/I with risk for aggression and oppositional behavior. Clinicians who observe young children with clear expressive language impairments (which are generally characteristic of young children with Dup7), together with social anxiety, hyperactivity/impulsivity, oppositional behavior, and aggression should consider a referral for genetic testing. While only a proportion of those children would have Dup7 (there are certainly other routes to this phenotype), it is also possible that such genetic testing might identify additional CNVs that are clinically meaningful. Further, understanding the role of social anxiety and ADHD in aggression/oppositionality may assist clinicians in developing more targeted interventions; identification and behavioral and/or psychosocial interventions to address social anxiety and/or impulsivity may be useful to reduce aggressive or oppositional behaviors by addressing underlying factors that set the stage for this behavior. Overall, this study’s findings support the possibility that increased dosage of one or more genes in the 7q11.23 region duplicated in Dup7, in transaction with the environment, may contribute to aggressive and oppositional behavior.