Deconstructing oppositional defiant disorder: Clinic-based evidence for an anger/irritability phenotype

Abstract

Objective

To examine risk factors and co-occurring symptoms associated with mother- versus teacher-reported anger/irritability symptoms (AIS) of oppositional defiant disorder (ODD) in a clinic-based sample of 1160 youth (ages 6–18).

Method

Participants completed a background history questionnaire (mothers), school functioning questionnaire (mothers, teachers), and DSM-IV-referenced symptom checklists (mothers, teachers). Youth meeting AIS criteria for ODD were compared to youth with ODD who met criteria for noncompliant symptoms (NS) but not AIS and to clinic controls.

Results

Compared with NS youth, youth with AIS were rated as exhibiting higher levels of anxiety and mood symptoms for both mother- and teacher-defined groups, and higher levels of conduct disorder symptoms for mother-defined younger and older youth. Remaining group differences for developmental, psychosocial, and psychiatric correlates varied as a function of informant and youth’s age.

Conclusions

Evidence suggests that AIS may constitute a more severe and qualitatively different ODD clinical phenotype, but informant and age of youth appear to be important considerations.

Introduction

Oppositional defiant disorder (ODD) is characterized by a pattern of negativistic, hostile, and defiant behavior toward adults and co-occurs with numerous psychiatric disorders, including attention-deficit/hyperactivity disorder (ADHD), conduct disorder (CD), and anxiety and mood disorders.^1–6 Although for many individuals ODD is relatively stable over time and predictive of later-onset CD, the anger/irritability symptoms (AIS) that in part define ODD may be more differentially associated with anxiety and mood disorders.^7–11 Given this possibility, the DSM-5 ADHD and Disruptive Behavior Disorders Workgroups recommended distinguishing between the “emotional” (loses temper, touchy or easily annoyed, and angry or resentful) and “behavioral” (e.g., argues, defies, annoys, blames others) features of ODD (www.dsm5.org).

The findings of several studies served as the basis for this study. Stringaris and Goodman^10,11 created three a priori subgroups of ODD symptoms, which they labeled irritable, headstrong, and hurtful, and examined their prognostic significance in a large mental health survey of youth aged 5 to 16 years. The 3-item irritable dimension (loses temper, angry and resentful, touchy or easily annoyed) predicted parent- and teacher-reported depressive and anxiety disorders at both baseline¹¹ and 3-year follow-up.¹⁰ The headstrong dimension (argues, defies, deliberately annoys, blames others) was associated with parent- and teacher-reported ADHD at baseline¹¹ and 3-year follow-up.¹⁰ All three dimensions were associated with parent- and teacher-reported CD at baseline; however, at 3-year follow-up, only the headstrong and hurtful (i.e., vindictive) dimensions predicted CD.¹⁰ Using epidemiological data from the Great Smoky Mountains study, Rowe et al.⁶ used factor analysis to identify an “irritable” factor identical to Stringaris and Goodman^10,11 and a “headstrong” factor comprised of the remaining five ODD symptoms. They reported both dimensions predicted CD and depressive disorders, but only the irritable dimension predicted anxiety disorders. Conversely, Mick et al.¹² found that although a 3-item irritability cluster (same items as Stringaris and Goodman^10,11 and Rowe et al.⁶) was common among youth with co-morbid ADHD, it was not associated with increased risk for mood disorders.

Additional research has considered affective (angry/irritable) and behavioral ODD symptoms, but has used items for constructing ODD symptom groups that differ from those described by the DSM-5 Workgroups and the studies referenced above.^6,10–12 For example, Burke and colleagues used factor analysis to identify a “negative affect” (touchy, angry, spiteful) and “behavioral” (argues, defies, loses temper) dimension of ODD among clinic-referred boys⁸ and among a separate sample of nonreferred girls.⁷ The behavioral dimension predicted CD (both samples), whereas negative affect predicted major depressive disorder (MDD, both samples) and CD (Caucasian girls). Last, Leibenluft et al.⁹ reported that “chronic irritability” defined a priori by 4 items (“arguing a lot” at home and school and “temper tantrums” at home and school) measured in early adolescence prospectively predicted ADHD in late adolescence and MDD in early adulthood. Note that although all studies include “loses temper” in constructing their respective AIS group, only Leibenluft et al.⁹ included “argues” in the AIS construct. Collectively, the aforementioned studies provide preliminary evidence for a distinction between AIS versus primarily noncompliant symptoms (NS) of ODD.

An important conceptual and methodological issue to consider is the use of data from multiple informants when defining caseness.^13,14 For example, symptom severity, impairment, levels of comorbidity, and psychosocial correlates differ depending on the informant whose evaluations serve as the basis for establishing ODD status.^3,4,15,16 Some evidence suggests that compared with parents’ reports, teachers’ ratings (a) have better predictive power for the diagnosis of ODD,¹⁷ (b) are more strongly related to peer-reported impairment criteria,¹⁶ and (c) have greater specificity for predicting co-occurring symptoms.^4,15 Three studies that compared different strategies for considering parent and teacher data found source-specific classifications demonstrated better internal and differential validity than grouping strategies based on combining data from multiple informants.^3,15,18

To further examine the validity of AIS and NS as ODD clinical phenotypes, we compared associated clinical, developmental, and psychosocial features of referred youth with ODD with AIS, ODD with NS, and non-ODD clinic controls. The present study expands on prior research by examining a wider range of psychosocial validators of group differences, as well as informant-specific classification strategies. The primary focus was co-occurring symptoms previously found to be differentially associated with AIS.^6,10–12 We hypothesized that youth with AIS would have more severe symptoms of anxiety, depression, and mania than youth with NS, but both groups would not differ in terms of ADHD or CD. We also examined additional variables associated with ODD,^1,3,4,19,20 but which heretofore have not been studied in terms of their potential differential associations with AIS (e.g., “difficult” temperament, academic and language difficulties, familial stressors, parental discipline, treatment history). Findings have implications for nosology and, more specifically, DSM-5 Workgroup recommendations.

Method

Participants

Participants were parents (primarily mothers) and teachers of 1160 youth who were consecutive referrals to a university hospital child psychiatry outpatient service that serves an ethnically and economically diverse clientele. Given well-established developmental differences in the emergence of psychiatric symptomatology, we divided the sample into a younger (6–11 year olds; n=546; 72.7% males) and older (12–18 year olds; n=614; 67.1% males) cohort (full sample M=12.1, SD=3.4 years; 69.6% males). Caregiver-identified ethnicity was as follows: European-American (n=977; 84%), African-American (n=81; 7%), Hispanic-American (n=127; 11%), Native-American (n=12; 1%), Asian-American (n=30; 3%), and Other (n=11; 1%). Most youth live with their biological mothers (84%) and fathers (62%), and in 65% of families, parents were married. Maternal and paternal education, respectively, was as follows: without high school diploma (6.9%, 9.5%); high school graduate without college education (23.5%, 26.4%); some college education (26.0%, 27.3%); and degree from 4-year college or more (43.6%, 36.8%). With regard to family income, 12% of families reported income < $20,001/year, 15% reported income from $20,001–$40,000/year, 23% reported income from $40,001–$70,000/year, and 50% reported income >$70,001. The two most common clinician-assigned diagnoses were ADHD and ODD, and almost all youth with ODD were also co-morbid for ADHD. This retrospective chart review study was approved by a university Institutional Review Board, and appropriate measures were taken to protect patient (and rater) confidentiality.

Measures

DSM-IV symptoms

Mothers and teachers rated youth’s symptoms using the Child and Adolescent Symptom Inventory-4R (CASI-4R).²¹ The CASI-4R is a parent- (158-item) and teacher- (153-item) completed behavior rating scale for evaluating youth 5 to 18 years old and combines the symptom modules from the Child Symptom Inventory-4^22,23 and the Adolescent Symptom Inventory-4.^24,25 Individual items bear one-to-one correspondence with DSM-IV symptoms and are rated on a scale from 0 (never) to 3 (very often). Item responses were summed to create Symptom Severity scores for co-occurring symptoms. ODD items rated as occurring often or very often were considered clinically significant and used for ODD subgroups (see Subgrouping section below). The present study used the following CASI-4R subscales (mother/teacher alphas in the present sample): ADHD, Inattentive type (ADHD:I, α=0.92/0.93; range=0–27); ADHD, Hyperactive-Impulsive type (ADHD:HI, α=.090/0.95; range=0–27); ODD (α=0.93/0.95; range=0–24); CD (α=0.86/0.83; range=0–44); Generalized Anxiety Disorder (GAD; α=0.80/0.75; range=0–21); Social Anxiety Disorder (α=0.87/0.85; range=0–12); Separation Anxiety Disorder (parent only, α=0.84; range=0–24); Major Depressive Disorder (MDD; α=0.88/0.79; range=0–30); and Manic symptoms (α=0.89/0.86; range=0–27). Scores of studies indicate that the CASI-4R demonstrates adequate internal consistency, test-retest reliability, clinical utility, and convergent and discriminant validity.^3,15,24–27

Developmental difficulties

Mothers indicated (yes/no) whether they had ever experienced any of 10 family stressors (α=0.61; M=0.99, SD=1.39; range=0–9); whether youth experienced any of 7 behaviors during the first year of life (e.g., sleep problems, feeding problems, excessive crying, difficult to comfort; α=0.75; M=0.85, SD=1.44; range=0–7); or any of 6 problems during the preschool period such as aggressive toward peers (α=0.60; M=1.11, SD=1.33; range=0–6). Speech or language difficulties (15 items) were indicated for past (α=0.79; M=1.37, SD=2.12; range=0–14) and present (α=0.77; M=1.19, SD=1.92; range=0–14).

Treatment

Mothers indicated (yes/no) whether youth had ever received 7 different types of behavioral and medical interventions (α=0.67; M=1.64, SD=1.62; range=0–7).

Parental discipline

Mothers endorsed (yes/no) whether they used any of 7 strategies to discipline their child (e.g., spanking, withdrawing love; α=0.51; M=2.13, SD=1.29; range=0–7).

Academic difficulties

Mothers reported (yes/no) special education services and current difficulties in any of 8 subjects (α=0.77; M=4.69, SD=2.98; range=0–12). Teachers rated academic performance for 5 subjects from 0 (2 or more years below) to 4 (2 or more years above grade level) (α=0.94; M=7.20, SD=3.80; range=0–20) and classroom performance (i.e., tests, homework, participation, behavior) on a scale from 0 (poor) to 4 (superior) (α=0.78; M=5.12, SD=3.15; range=0–16).

Interpersonal difficulties

Mothers rated (yes/no) five types of peer (e.g., α=0.69; M=1.02, SD=1.32; range=0–5) and four types of sibling (α=0.75; M=1.06, SD=1.30; range=0–4) difficulties (e.g., frequent arguments, fights, teasing).

Procedure

Prior to scheduling their initial clinic evaluation, parents were mailed a packet of materials including behavior ratings scales for both parent and teacher, background information questionnaire, and permission for release of school, psychoeducational, and special education evaluation records. Teacher ratings were given to the school by parents, completed by teachers (96% of participants), and mailed to the clinic prior to the evaluation. Procedures for evaluating patients are described in detail elsewhere.^23,25,27 Youth with both parent and teacher reports (n=1111) and youth with reports from only one informant (n=49) did not differ on any categorical or dimensional ODD scores, ethnicity, family income, or parental education. However, compared to youth with one informant’s report, youth with both informants’ reports, respectively, were younger (Ms=13.3 vs. 12.0 years), more likely to be male (55% vs. 70%), and more likely to have married parents (47% vs. 66%).

Subgrouping

The present study builds on previous research by using a priori operationalized criteria for AIS (i.e., proposed DSM-5 anger/irritability mood symptoms).^6,10–12 Parent and teacher reports of ODD symptoms were considered separately.^3,15,18 Youth with severity ratings of often or very often for “loses temper,” “is angry and resentful,” and “is touchy or easily annoyed by others” were placed in the AIS subgroup. The remaining youth were classified as follows: primarily noncompliant symptoms (NS) if they met severity criteria for four ODD symptoms, but two or fewer AIS, and clinical comparison (C) if they had three or fewer ODD symptoms.

Statistical Analyses

For count (background) variables, we conducted nonparametric Kruskal-Wallis tests with follow-up Bonferroni-corrected Mann-Whitney U-tests to localize differences among groups. For co-occurring symptom variables, we conducted one-way ANOVAs with follow-up Scheffe tests and, if variances across groups were not homogeneous, Games-Howell tests. We used a Bonferroni correction within measure. Effect sizes (eta²) are reported for the main effects of the ANOVAs and can be interpreted as follows: 0.01 (small), 0.06 (medium), and 0.14 (large).²⁸ We also conducted secondary analyses of covariance for co-occurring symptom variables, controlling for (a) youth’s gender and (b) ODD symptom severity for the informant that was used to identify subgroups. This stringent test allowed examination of whether group differences were associated with only greater symptom severity, or both symptom severity and symptoms used to define groups (AIS vs. NS). We were unable to conduct these secondary analyses with the count (background) variables because of the nonparametric nature of these data.

Results

Distribution of Subgroups

When mothers’ ratings were the basis of group classification, 212 younger (69.8% male) and 284 older (69.4% male) youth met symptom criteria for ODD of whom 53% (63.4% male) and 64% (70.2% male), respectively, were classified AIS. Using teachers’ ratings to construct groups, 204 younger (78.9% male) and 195 older (76.9% male) youth were ODD of whom 61% in each age group (younger: 82.4% male, older: 77.3% male) were classified AIS.

Mother-defined ODD Subgroups

Background variables

The two ODD groups differed from Controls for most variables, with the exceptions of school-related functioning (Table 1). Compared with the NS group, the youth with AIS were rated as being more difficult as toddlers (both age groups). Among younger youth, AIS and NS groups differed on ever in treatment (AIS>NS) and teacher-rated academic functioning (NS>AIS).

Table 1.

Means (SDs) and comparison statistics for background variables among mother-rated oppositional defiant disorder subgroups and clinic controls

Variable	AIS	NS	Controls	K-W χ2	Post hoc
Youth ages 6–11 years	n=112	n=100	n=328
Difficult infant	1.3 (1.7)	1.0 (1.4)	0.6 (1.1)	17.9***	AIS,NS>C
Difficult toddler	2.2 (1.5)	1.8 (1.6)	0.9 (2.1)	86.6***	AIS>NS>C
Past language problems	2.1 (2.5)	1.8 (2.4)	1.4 (2.1)	13.2**	AIS,NS>C
Present language problems	2.0 (2.5)	1.7 (2.2)	1.0 (1.7)	30.0***	AIS,NS>C
Parent discipline	2.6 (1.5)	2.6 (1.1)	2.0 (1.1)	31.9***	AIS,NS>C
Peer difficulties	2.2 (1.6)	1.8 (1.5)	0.7 (1.0)	113.8***	AIS,NS>C
Sibling difficulties	1.9 (1.4)	1.7 (1.5)	0.8 (1.1)	70.3***	AIS,NS>C
Familial stress	1.3 (1.7)	1.1 (1.4)	0.7 (1.2)	16.4***	AIS,NS>C
Ever in treatment	2.3 (1.6)	1.8 (1.6)	0.9 (1.1)	83.3***	AIS>NS>C
School functioning (parent)	4.7 (3.3)	4.6 (3.3)	4.7 (2.7)	0.3	NA
Academic performance (teacher)	6.7 (3.5)	8.5 (4.2)	6.9 (4.0)	7.6*	NS>AIS,C
Classroom functioning	4.8 (3.1)	5.9 (3.1)	5.0 (3.3)	10.9**	NS>AIS,C
Youth ages 12–18 years	n=181	n=103	n=319
Difficult infant	1.2 (1.6)	1.0 (1.7)	0.6 (1.3)	20.8***	AIS>C
Difficult toddler	1.3 (1.3)	1.0 (1.3)	0.6 (0.9)	53.8***	AIS>NS>C
Past language problems	1.2 (1.8)	1.3 (1.9)	0.9 (1.7)	8.9*	AIS,NS>C
Present language problems	1.2 (1.8)	1.2 (1.9)	0.6 (1.3)	33.8***	AIS,NS>C
Parent discipline	2.5 (1.4)	2.2 (1.3)	1.7 (1.1)	41.1***	AIS,NS>C
Peer difficulties	1.5 (1.5)	1.2 (1.3)	0.5 (0.9)	78.3***	AIS,NS>C
Sibling difficulties	1.6 (1.4)	1.3 (1.4)	0.6 (1.0)	67.3***	AIS,NS>C
Familial stress	1.4 (1.6)	1.5 (1.7)	0.8 (1.2)	26.3***	AIS,NS>C
Ever in treatment	2.6 (1.7)	2.8 (1.9)	1.5 (1.6)	68.0***	AIS,NS>C
School functioning (parent)	5.1 (3.1)	5.0 (3.0)	4.7 (3.0)	1.6	NA
Academic performance (teacher)	7.2 (3.6)	7.2 (3.8)	7.1 (3.8)	0.2	NA
Classroom functioning	5.0 (2.9)	5.1 (3.5)	5.4 (3.0)	2.2	NA

Note: AIS=anger/irritability symptoms, C=clinic-referred controls, NS=noncompliant symptoms, K-W=Kruskal-Wallis test, NA=not applicable, post hoc=Mann-Whitney U-test.

^*p<0.05;

^**p<0.01;

^***p<0.001.

Co-occurring symptoms

The ODD groups obtained more severe symptom ratings than Controls for most symptoms (Table 2). Compared with the NS group, youth with AIS were rated as having more severe symptoms of ODD, CD, GAD, MDD, and manic symptoms (both age groups). Younger youth with AIS had more severe social anxiety disorder symptoms than the NS group. Older youth with AIS obtained higher ADHD:I and ADHD:HI scores than the NS group. Effect sizes ranged from medium to large among the younger group, and from small to large among the older group.

Table 2.

Means (SDs) and comparison statistics for oppositional defiant disorder and co-occurring symptoms among parent-rated oppositional defiant disorder subgroups and clinic controls

Variable	AIS	NS	Controls	F	eta2	Post hoc
Youth ages 6–11 years	n=112	n=100	n=328
Oppositional defiant disorder	19.0 (3.6)	14.8 (2.6)	5.5 (3.3)	869.5***	.76	AIS>NS>C
ADHD: Inattentive	18.1 (6.5)	18.0 (6.5)	14.0 (6.3)	25.1***	.09	AIS,NS>C
ADHD: Hyperactive/Impulsive	17.6 (8.1)	16.6 (7.7)	9.0 (6.7)	78.1***	.23	AIS,NS>C
Conduct disorder	10.9 (6.6)	8.3 (5.1)	2.5 (2.6)	183.8***	.42	AIS>NS>C
Generalized anxiety disorder	11.2 (4.3)	8.4 (3.8)	5.6 (3.6)	94.2***	.27	AIS>NS>C
Social anxiety disorder	4.2 (3.6)	2.7 (2.7)	2.2 (2.9)	16.9***	.06	AIS>NS,C
Separation anxiety disorder	4.9 (5.2)	3.9 (4.7)	2.2 (3.0)	23.5***	.08	AIS,NS>C
Major depressive disorder	12.6 (8.0)	8.0 (6.7)	4.2 (4.9)	79.0***	.24	AIS>NS>C
Manic symptoms	7.1 (6.7)	4.3 (5.1)	1.9 (3.1)	58.2***	.18	AIS>NS>C
Youth ages 12–18 years	n=181	n=103	n=319
Oppositional defiant disorder	19.1 (3.9)	14.4 (2.8)	6.0 (3.1)	959.1***	.76	AIS>NS>C
ADHD: Inattentive	19.4 (5.8)	17.2 (6.1)	13.4 (6.7)	52.6***	.15	AIS>NS>C
ADHD: Hyperactive/Impulsive	12.7 (7.3)	10.2 (7.1)	5.8 (5.9)	65.3***	.18	AIS>NS>C
Conduct disorder	14.5 (7.6)	12.0 (6.0)	4.8 (4.6)	167.0***	.37	AIS>NS>C
Generalized anxiety disorder	10.8 (4.3)	8.7 (4.5)	6.7 (4.2)	52.1***	.15	AIS>NS>C
Social anxiety disorder	3.5 (3.5)	2.7 (3.3)	2.2 (3.0)	9.3***	.03	AIS>C
Separation anxiety disorder	2.6 (3.8)	2.2 (4.1)	1.4 (2.8)	7.5***	.02	AIS>C
Major depressive disorder	17.6 (9.4)	14.1 (8.8)	8.3 (7.7)	68.4***	.20	AIS>NS>C
Manic symptoms	7.4 (5.8)	5.6 (4.6)	2.3 (3.2)	81.2***	.22	AIS>NS>C

Note: ADHD=attention-deficit/hyperactivity disorder, AIS=anger/irritability symptoms, C=clinic-referred controls, eta² = effect size, NA=not applicable, NS=noncompliant symptoms.

^***p<0.001.

Teacher-defined ODD Subgroups

Background variables

Compared with mother-defined groups, there were many fewer group differences in background and developmental variables for the teacher-defined groups (Table 3). For both age groups, ODD groups differed from Controls on difficult toddler and peer difficulties. Age-related differences between ODD groups and Controls were as follows: younger youth (language problems, familial stress, ever in treatment) and older youth (parent discipline, academic difficulties). There were no ODD group differences in developmental characteristics.

Table 3.

Means (SDs) and comparison statistics for background variables among teacher-rated oppositional defiant disorder subgroups and clinic controls

Variable	AIS	NS	Controls	K-W χ2	Post hoc
Youth ages 6–11 years	n=125	n=79	n=333
Difficult infant	0.7 (1.3)	0.9 (1.3)	0.9 (1.4)	0.8	NA
Difficult toddler	1.5 (1.4)	1.4 (1.3)	1.2 (1.4)	8.7*	AIS>C
Past language problems	1.8 (2.4)	2.1 (2.4)	1.4 (2.1)	7.2*	NS>C
Present language problems	1.4 (1.8)	2.2 (2.8)	1.2 (1.8)	10.6**	NS>C
Parent discipline	2.3 (1.3)	2.4 (1.5)	2.2 (1.2)	2.1	NA
Peer difficulties	1.8 (1.5)	1.8 (1.5)	0.8 (1.2)	68.1***	AIS,NS>C
Sibling difficulties	1.1 (1.3)	1.5 (1.5)	1.1 (1.3)	4.0	NA
Familial stress	1.1 (1.4)	1.3 (1.8)	0.8 (1.2)	8.4*	NS>C
Ever in treatment	1.7 (1.5)	1.9 (1.6)	1.1 (1.4)	29.2***	AIS,NS>C
School functioning (parent)	4.7 (3.0)	5.1 (3.1)	4.6 (2.9)	0.7	NA
Academic performance (teacher)	7.0 (4.2)	7.7 (3.4)	7.2 (4.0)	1.8	NA
Classroom functioning	4.7 (3.2)	4.8 (3.0)	5.2 (3.3)	2.2	NA
Youth ages 12–18 years	n=119	n=76	n=396
Difficult infant	0.9 (1.5)	0.9 (1.5)	0.8 (1.6)	2.3	NA
Difficult toddler	1.1 (1.2)	1.0 (1.4)	0.8 (1.1)	7.4*	AIS>C
Past language problems	1.2 (1.8)	1.4 (2.3)	1.0 (1.7)	3.5	NA
Present language problems	1.1 (1.8)	1.1 (1.9)	0.8 (1.4)	3.5	NA
Parent discipline	2.3 (1.4)	2.3 (1.4)	1.9 (1.6)	11.3**	AIS,NS>C
Peer difficulties	1.6 (1.6)	1.2 (1.4)	0.7 (1.0)	38.7***	AIS,NS>C
Sibling difficulties	1.1 (1.4)	1.1 (1.2)	1.0 (1.2)	0.7	NA
Familial stress	1.2 (1.5)	1.1 (1.5)	1.1 (1.4)	0.9	NA
Ever in treatment	2.3 (1.8)	2.0 (1.8)	2.0 (1.7)	3.0	NA
School functioning (parent)	5.3 (2.9)	5.8 (3.3)	4.6 (3.0)	10.7**	AIS,NS>C
Academic performance (teacher)	7.0 (4.0)	7.5 (4.4)	7.2 (3.4)	0.2	NA
Classroom functioning	4.6 (2.8)	5.5 (3.2)	5.3 (3.1)	4.9	NA

Note: ADHD = attention-deficit/hyperactivity disorder; AIS=anger/irritability symptoms, C=clinic-referred controls, K-W=Kruskal-Wallis test, NA=not applicable, NS=noncompliant symptoms, post hoc=Mann-Whitney U-test.

^*p<0.05;

^**p<0.01;

^***p<0.001.

Co-occurring symptoms

The AIS and NS groups generally obtained more severe ratings than Controls (Table 4), particularly among younger youth. As was the case for mother-defined ODD groups, the teacher-defined AIS group had more severe ODD and GAD symptoms (both age groups) than youth with NS, and effect sizes were large. Younger youth with AIS had more severe MDD symptoms than the NS group (large effect size). Unlike the mother-defined groups, teacher-defined AIS and NS groups exhibited comparable levels of ADHD, CD, social anxiety disorder, and manic symptoms.

Table 4.

Means (SDs) and comparison statistics for oppositional defiant disorder and co-occurring symptoms among teacher-rated oppositional defiant disorder subgroups and clinic controls

Variable	AIS	NS	Controls	F	eta2	Post hoc
Youth ages 6–11 years	n=125	n=79	n=333
Oppositional defiant disorder	19.3 (3.5)	14.7 (2.4)	3.4 (3.6)	1108.8***	.81	AIS>NS>C
ADHD: Inattentive	18.7 (5.8)	18.8 (6.9)	15.1 (7.8)	16.0***	.06	AIS,NS>C
ADHD: Hyperactive/Impulsive	17.9 (8.3)	17.2 (8.5)	8.5 (8.2)	76.1***	.23	AIS,NS>C
Conduct disorder	7.2 (5.9)	5.5 (4.5)	0.9 (1.8)	154.2***	.37	AIS,NS>C
Generalized anxiety disorder	10.8 (3.6)	8.1 (4.0)	5.8 (3.5)	87.5***	.25	AIS>NS>C
Social anxiety disorder	2.6 (3.0)	2.2 (3.1)	2.3 (3.1)	0.5	.00	NA
Major depressive disorder	12.4 (7.3)	7.7 (6.4)	6.7 (6.0)	35.0***	.12	AIS>NS,C
Manic symptoms	11.8 (5.0)	10.1 (5.5)	5.0 (4.4)	105.8***	.29	AIS,NS>C
Youth ages 12–18 years	n=119	n=76	n=396
Oppositional defiant disorder	19.1 (3.5)	15.0 (2.7)	3.6 (3.7)	1050.4***	.78	AIS>NS>C
ADHD: Inattentive	18.9 (6.2)	18.5 (6.3)	13.2 (7.4)	40.4***	.12	AIS,NS>C
ADHD: Hyperactive/Impulsive	14.3 (8.6)	14.2 (9.1)	4.2 (5.1)	147.4***	.34	AIS,NS>C
Conduct disorder	9.0 (6.2)	7.1 (4.9)	2.2 (2.9)	142.9***	.34	AIS,NS>C
Generalized anxiety disorder	10.2 (3.7)	8.3 (3.6)	5.5 (3.7)	77.9***	.21	AIS>NS>C
Social anxiety disorder	1.8 (2.4)	1.7 (2.3)	2.1 (3.0)	1.1	.00	NA
Major depressive disorder	11.2 (6.8)	9.0 (6.5)	7.7 (6.0)	14.0***	.05	AIS>C
Manic symptoms	10.5 (5.6)	9.8 (5.3)	3.5 (3.7)	153.6***	.34	AIS,NS>C

Note: ADHD = attention-deficit/hyperactivity disorder; AIS=anger/irritability symptoms, C=clinic-referred controls, effect size=eta², NA=not applicable, NS=noncompliant symptoms.

^***p<0.001.

Secondary Analyses

When controlling for youth’s gender, the significance of the main effects of ODD group for co-occurring psychiatric symptoms did not change. When covarying ODD symptom severity using the same informant as was used to define subgroups, several main effects of ODD group remained significant, consistent with the idea that AIS and NS are qualitatively different. Regarding co-occurring symptoms, variables that remained significant were (a) parent-reported GAD (both age groups), social anxiety disorder (younger group), and MDD (older group); and (d) teacher-reported ADHD-I and ADHD-HI (both age groups), GAD and MDD (younger group), and manic symptoms (older group). As noted above, these secondary analyses were not conducted with the background variables because of the nonparametric nature of these data.

Discussion

Recent evidence suggests that different types of ODD symptoms have distinct correlates and outcomes,^6–12 which in part contributed to a suggestion by the DSM-5 ADHD and Disruptive Behavior Disorders Workgroups to consider AIS and NS separately. Nevertheless, few studies have explicitly examined their divergent diagnostic validity or considered the implications of contextual variation. Our findings for a large, heterogeneous, clinic-based sample provide additional support for divergent ODD syndromes. Specifically, AIS is characterized by a range of differentially more severe emotional reactions as evidenced by consistencies across age and informant for GAD and, with one exception (older teacher-defined groups), MDD symptoms (AIS>NS). Furthermore, AIS and NS group differences were significant for manic symptoms (mother-defined groups) and social anxiety disorder symptoms (mother-defined younger group). Collectively, these results, in conjunction with the findings of other investigators,^6–12 suggest that AIS may be (a) one component of a broader emotion dysregulation syndrome or (b) a particularly virulent form of ODD. This latter possibility is consistent with two-hit genetic models for various diseases, which indicate that differentially greater severity may actually indicate unique pathogenic processes.²⁹ Findings also raise questions regarding the ODD diagnosis, given that youth meeting the 3-symptom criteria for AIS would not meet DSM criteria for ODD without a fourth symptom. However, the results of previous work^30,31 indicate that youth with subthreshold ODD according to DSM criteria exhibit significant impairment and comorbidity, comparable to youth who meet DSM criteria for ODD. Thus, findings also lend partial support to the ICD-10 model of conceptualizing ODD, which takes a developmental approach to ODD and CD and permits diagnosis of ODD with fewer than 4 ODD symptoms. Nevertheless, all youth with 3 AIS also exhibited at least one NS in the present sample. Given that the DSM model may under-identify youth with impairing ODD,^30,31 future research should evaluate the prevalence of youth with AIS who do not meet criteria for ODD, as well as multiple models for considering ODD symptoms and diagnoses (e.g., DSM, ICD).

Across mother-defined age groups, the AIS group was rated as exhibiting higher levels of CD symptoms than the NS group, and both AIS and NS youth received higher CD ratings than Controls for mother- and teacher-defined groups. These results are consistent with previous research with epidemiologic samples linking AIS and NS to CD symptoms concurrently¹¹ and prospectively.⁶ Nevertheless, our findings diverge from studies that considered AIS and NS symptoms dimensionally, which found NS (but not AIS) predicted CD symptoms.^7,10 It is noteworthy that these relations were observed with and without controlling for youth’s gender, indicating AIS and NS group differences were similar among clinic-referred males and females despite different prevalence rates associated with the co-occurring symptoms considered.²

ADHD

It is well established that ADHD and ODD commonly co-occur in referred samples,^1,3,4,32 and the present study is no exception. Nevertheless, co-occurring ADHD symptom severity (both inattentive and hyperactive-impulsive behaviors) was comparable among the younger AIS and NS groups, as well as teacher-defined AIS and NS groups. This pattern of findings is consistent with research suggesting that AIS does not differentially predict ADHD severity compared with NS^10,11 and suggests that co-occurring ADHD was not likely a primary contributor to group differences in other variables. We did find, however, that the older, mother-defined AIS group had more severe ADHD symptoms than the NS group, and others have found AIS is associated with elevated ADHD symptoms.^9,12 Future research will need to determine the best approach for conceptualizing AIS and NS (categorical vs. dimensional), evaluating data from different informants, assessing ODD symptoms, and considering different developmental periods to improve our understanding of relations between ADHD and ODD.

Contextual Variables

The modest convergence in group differences for the two informant-based classification systems is consistent with our strategy for classifying youth. Only 97 youth in the entire sample met criteria for AIS according to both informants, and they comprised only a minority of mother-defined (33%) and teacher-defined (40%) AIS groups. It is likely that different contextual variables (e.g., consistency, routine, socialization) are the progenitors of AIS reactions in home versus school settings,⁸ which is well-illustrated in our results. For example, mothers describe their offspring with AIS as “difficult as a toddler,” whereas teachers (who have spent a lot less time with the youth) do not. As parents are typically responsible for seeking intervention, it is not surprising that mother-defined ODD groups were more likely to have received treatment, particularly among younger youth. Other setting-laden variables, such as academic performance and classroom functioning, differentiated informant-specific syndromes. Nevertheless, differences in environmental triggers for AIS are likely a gross oversimplification of pathogenic processes that may index distinct neurobiological processes³³ and contribute to situation specificity. Future research will need to identify distinct neurobiological substrates that contribute to individual differences in AIS.^33,34 For example, previous clinical and developmental investigations have considered constructs similar to AIS (e.g., emotional lability, poor executive functioning, exploratory/novelty-seeking behavior)^3,15,35 and provide a foundation for such research examining biologic substrates for NS and AIS and considering contextual variables.^15,20,35 Collectively, our results suggest that although emotional reactivity is influenced by environmental variables that in turn influence informant perceptions of symptom severity, (a) the characteristics of the emotional response are correlated within setting and (b) the expression of anger, anxiety, and depression likely share common neurobiologic pathways.³³

ODD Symptoms

Several main effects for mood and anxiety disorder symptoms, as well as ADHD symptoms (teacher-defined groups), remained significant when controlling for ODD symptom severity. These findings indicate that group differences cannot be attributed only to differential levels of ODD, and that AIS and NS groups are qualitatively different. Moreover, ancillary analyses comparing individual ODD symptoms indicated that the AIS subgroups had more severe ratings of “takes anger out on others” and “blaming others” than the NS subgroups and were more likely to “annoy others” (mothers’ ratings, both age groups) and “argue” (teachers’ ratings, older youth), but were not more “defiant” (both raters and age groups). These findings further support the notion that AIS and NS differ in both severity and quality of co-occurring symptoms, and informant, environmental precipitants, age, and symptom specification will likely play an important role in future modifications of phenotypic criteria.

Summary

The strengths of the study include the use of proposed DSM-5 criteria for distinguishing youth with AIS from NS;^6,10–12 both parents and teachers as informants; large, heterogeneous, clinic-based sample; broad age range (stratified as 6–11 years and 12–18 years); and a diverse array of background and clinical characteristics. Nevertheless, we did not examine source exclusivity (i.e., differences among youth who were symptomatic according to one, but not both informants⁴). Although we were able to contrast early versus later referrals, which provides useful clinical information, cross-sectional designs do not address developmental issues. Future research should also consider a wider range of risk factors as potential validators of teacher-defined diagnostic groups.

Our findings provide further empirical validation for the recommendation to consider AIS and NS separately, though this recommendation will most assuredly undergo future modifications as the significance of informant, environmental precipitants, age, and symptom specification become more salient. Moreover, they also suggest different developmental processes, course, and possibly response to intervention,²⁷ and when considered in conjunction with the results of related studies, provide further support for a broad-based, emotion dysregulation syndrome whose clinical features are influenced by individual differences in neurobiology, personal experiences, and contextual characteristics.