Research on Child and Adolescent Psychopathology https://doi.org/10.1007/s10802-022-00910-8 Reciprocal relations between dimensions of Oppositional defiant problems and callous-unemotional traits Lourdes Ezpeleta1,2 · Eva Penelo1,3 · J. Blas Navarro1,3 · Núria de la Osa1,2 · Esther Trepat1,2 · Lars Wichstrøm4 Accepted: 21 February 2022 © The Author(s) 2022 Abstract Although irritability, headstrong/defiant behavior, and callous-unemotional traits (CU traits) often co-occur, the prospec- tive associations between them are not well known. A general population sample of 622 children was followed up yearly from ages 3 to 12 years and assessed using dimensional measures of irritability, headstrong/defiant, and CU traits with teacher provided information. A random intercept cross-lagged panel model, accounting for all unmeasured time-invariant confounding using the children as their own controls, revealed cross-lagged reciprocal associations between increased headstrong/defiant and increased CU traits at all ages and a unidirectional association from headstrong/defiant to irritabil- ity. The findings are consistent with headstrong/defiant behavior and CU traits mutually influencing each other over time and headstrong/defiant behavior enhancing irritability. School-based intervention and prevention programs should take these findings into consideration. They also suggest that irritability acts as a distinct developmental dimension of head- strong/defiant and callous-unemotional behaviors and needs to be addressed independently. Keywords Callous-unemotional traits · Cross-lagged panel model · defiant/headstrong · Irritability · Limited prosocial emotions · Oppositional defiant Oppositional defiant disorder (ODD) is a prevalent and ODD is currently conceptualized as a dimensional disorder stable heterogeneous disorder characterized by defiant/dis- with several factors comprising irritability (loses temper, obedient behavior, anger/irritability, and hostility towards angry, and touchy), headstrong/defiance (argues, defies, authority figures (American Psychiatric Association, 2013). annoys, and blames), and hurtfulness (spitefulness and vin- dictiveness) (Stringaris & Goodman, 2009). However, in a large sample, Burke et al. (2014) proposed that the best structure of ODD was a bifactor model that included irri- Lourdes Ezpeleta tability and headstrong/defiant plus a general ODD factor. lourdes.ezpeleta@uab.cat Breaking down ODD into its dimensions has proven to be 1 Unitat d’Epidemiologia i de Diagnòstic en Psicopatologia del clinically useful to explain the wide comorbidity of ODD: Desenvolupament, Barcelona, Spain the irritability dimension has been associated with depres- 2 Departament de Psicologia Clínica i de la Salut. Edifici sion and anxiety, the headstrong/defiant dimension with B, Universitat Autònoma de Barcelona, 08193 Barcelona, attention deficit/hyperactivity disorder (ADHD) and con- Bellaterra, Spain duct disorder (CD), and the hurtful dimension, the least 3 Departament de Psicobiologia i de Metodologia de les validated, with callousness. The dimensional approach is Ciències de la Salut, Universitat Autònoma de Barcelona, useful for a developmental model specifying the mecha- Barcelona, Spain nisms of ODD because it allows their heterogeneity to be 4 Department of Psychology, Norwegian University of Science and Technology, Trondheim, Norway 13 2 Research on Child and Adolescent Psychopathology disentangled (Wakschlag et al., 2018). We therefore apply been considered a manifestation of motivational deficits such an approach in the present study. such as delay aversion, which impedes ability to wait to gain ODD typically has an early onset at home and tends or avoid later reward or punishment (Griffith et al., 2019; to escalate to other contexts such as school, where it is Stringaris & Goodman, 2009). observed that affected children present significant academic While reciprocal relations are understudied, there is problems, fewer friendships, and marked functional impair- some relevant research. Whelan et al. (2013) analyzed the ment (Wesselhoeft et al., 2019). An objective pursued by cross-lagged developmental continuity of the three ODD clinicians is to try to identify groups of clients with very dimensions (irritability, headstrong/defiant, hurtful) from 8 similar symptoms who may need similar interventions. To to 13 years old, combining parent and teacher information this end, subtypes and specifiers are defined in classifica- and finding few continuities: Headstrong at age 10 years tion systems. ODD is subtyped according to the presence was positively associated with irritable at age 13; hurtful of (i) chronic irritability/anger and/or (ii) the presence of at age 10 was likewise positively associated with lower limited prosocial emotions (LPE) in ICD-11 (World Health levels of headstrong at age 13; and irritable did not relate Organization, 2019), whereas the DSM-5 (American Psy- to either headstrong or hurtful. The authors interpret these chiatric Association, 2013) does not include these subtypes, few continuities as an indication of the distinctiveness of the but rather organizes the symptoms according to the dimen- dimensions. Given that dimensions can be used to identify sions proposed by Stringaris and Goodman (2009) (angry/ etiological targets and differential treatment needs (Burke et irritable mood, argumentative/defiant behavior, and vindic- al., 2014), knowing how irritability and headstrong/defiant tiveness). These dimensions evince considerable within- dimensions are intertwined through development is essen- dimension homotypical developmental continuity (Whelan tial. There are no studies that have shown how CU traits et al., 2013), explain unique variance (Burke et al., 2014), relate to ODD dimensions (irritability and headstrong/defi- and are highly correlated (Evans et al., 2016; Krieger et al., ant) throughout development. 2013). However, whether and how these ODD dimensions may influence each other over time is poorly understood, and unraveling their heterotypical continuity holds the pros- The Relation Between Irritability and pect of increasing our understanding of the etiology and Headstrong/Defiant Dimensions, and CU course of ODD and its subtypes. The present work examines Traits the prospective relations between the three constructs irri- tability, headstrong/defiant, and LPE (as evaluated through CU traits are characterized by lack of empathy or remorse, callous unemotional traits, hereafter referred to as CU traits) reduced affect or shallow emotional responding, and not in the school context. caring about the feelings of others. CU traits distinguish a group of children with more severe ODD symptoms who present deficits in executive functioning, social cognition, Relations between Irritability and and attention, evince more instrumental aggressive behav- Headstrong/Defiant Dimensions ior, are less fearful, recover more easily after being upset, and show less negative reactivity (Blair, 2018; Hawes et al., Irritability can be defined as an elevated proneness to anger 2013; Willoughby et al., 2011). relative to that of peers at the same development level Irritability and CU traits are independent constructs that (Stringaris et al., 2018). During development, irritability is may or may not co-occur. Irritability and CU traits have dif- a normative expression when faced with frustration, which ferent developmental interpretations. Irritability is common children typically learn to control as they age (Wakschlag throughout childhood and adolescence (Morken et al., 2021) et al., 2012). The growth of prefrontal cortical structures and only if very frequent and dysregulated is considered during the preschool years facilitates the development of abnormal, whereas CU traits are not developmentally nor- executive functions, which help to self-regulate anger and mative and are considered risk factors for conduct problems respond adaptively (Wakschlag et al., 2018). High fre- and their persistence and severity (Wakschlag et al., 2018). quency, dysregulation, persistence and intensity, and a low Irritability and CU traits are related to two social emo- threshold of elicitation are indicators of abnormal irritability tions, anger and empathy, which regulate individuals’ (Wakschlag et al., 2018). aggressive response: increased anger is associated with reac- For social learning, children need to be sensitive to both tive aggression whereas reduced empathy increases the risk punishment and reward to learn to refrain from inappropri- of instrumental aggression (Blair, 2018). The two types of ate behaviors and adopt appropriate ones (Matthys et al., aggressive behavior tend to co-occur (Blair, 2018; Crapanz- 2012). Indicators of the headstrong/defiant dimension have ano et al., 2010). However, few studies have simultaneously 13 Research on Child and Adolescent Psychopathology 3 addressed irritability and CU traits in children. Stringaris common genetics, socioeconomics, gender, parenting prac- and Goodman (2009) found that the irritability dimension tices) may influence all ODD dimensions, thus producing of ODD, which is predominantly associated with emotional spurious relations between them, also prospectively. Within- problems, was also correlated with callousness. Likewise, person analytical approaches have the added advantage of the presence of CU traits with irritability was associated adjusting for all time-invariant confounding, regardless of cross-sectionally with impairment in other areas (peers, whether it is known or not (Allison et al., 2009) (although academic) (Waschbusch et al., 2020). None of these studies time-varying factors may still influence the results). To sum were longitudinal and we do not know if CU traits forecast up, this work examines how teacher ratings of within-per- increased irritability, or if increased irritability/anger/frus- son changes in CU traits, irritability, and headstrong/defiant tration fuels the difficulties in interpersonal sensitivity and are associated longitudinally in a general population of chil- behavioral inhibition observed in CU traits. Further work is dren followed yearly from ages 3 to 12. Due to the lack of needed to understand the role they play in ODD. previous research, we remain open to the direction of effects Etiological models of CU traits propose that CU behav- involving CU traits. Given that results of the developmental iors emerge from an inherited temperament risk of low associations between irritability and headstrong/defiant dif- interpersonal emotional sensitivity, which hampers the fer, we test two competing hypotheses: (i) There is a recip- development of affective empathy and facilitates deficient rocal prospective relationship between increased irritability moral emotions (lack of shame and guilt), interfering with and increased headstrong/defiant; and given that although rule internalization and fearlessness, which in turn hampers they are correlated the two dimensions bear distinguishable the development of behavioral inhibition to threat, leading developmental risks (Waldman et al., 2021), we propose to high approach, reward dominance, and difficulty learn- that (ii) increases in irritability and headstrong/defiant are ing from punishment (Wakschlag et al., 2018; Waller et al., not prospectively related. 2017; Waller & Hyde, 2018). Developmentally, it would not only be expected that chil- dren with a low response to others’ emotional states are at Method a higher risk of more severe disruptive behavior (Frick et al., 2014; Longman et al., 2016), but also that disruptive Participants behaviors and accompanying difficulties (such as difficul- ties in executive functioning, preference for immediate The sample is part of a longitudinal study of behavioral reward, poor self-control, deficiencies in social cognition, problems starting at age 3 described in Ezpeleta et al. and reduced parental warmth) (Booker et al., 2020; Deters (2014). The children were randomly selected from early et al., 2020; Matthys et al., 2013) boost CU traits. In support childhood schools in Barcelona (Spain) (N = 2,283). A two- of this view, Whelan et al. (2013) found that headstrong/ phase design was employed. In the first phase of sampling, defiant at age 13 predicted callousness three years later. 1,341 families (58.7%) agreed to participate (33.6% high School is a main context for children’s development. socioeconomic status (SES), 43.1% middle, and 23.3% low; Teachers are valid reporters of students’ social relationships 50.9% boys). To ensure the participation of children with and their respect for norms of conduct, and specific interven- possible behavioral problems, the parent-rated Strengths tions may be carried out in schools. Consequently, school is and Difficulties Questionnaire (SDQ) conduct problems a setting where ODD can be studied in a valid way (Evans et scale (Goodman, 1997) plus four ODD DSM-IV-TR symp- al., 2016). Our interest in the prospective relations between toms not included in the SDQ were used for screening. Two ODD dimensions is based on the prospect that these asso- groups were considered for the second phase of the sam- ciations may inform etiological interpretations. So far, work pling design: the screen-positive group, which included all on ODD dimensions has asked whether children higher on the children with SDQ scores ≥ 4 (i.e., percentile 90) or with an ODD dimension (e.g., irritability) than other children a positive response to any of the eight DSM-IV ODD symp- will have higher levels of another ODD dimension (e.g., toms (n = 417; 49.0% boys); and a random draw of children headstrong) than other children in the future. However, screened negative (n = 205; 51.2% boys). The minimum other (unknown) children’s ODD cannot influence the ODD sample size of the positive screening cohort was calculated of a specific child and between-person differences are not using the software nQuery Advisor 7.0 (Statistical Solu- informative. Changes within a person, however, may prove tions, 2007). A prevalence of 15% and a multiple correla- informative, validating the question “Does increased irrita- tion between covariates of 0.40 were assumed. The sample bility in this child forecast increased headstrongness in the size was determined for detecting OR = 1.8 between psy- same child?”. Hence, we apply within-person analysis to our chopathology and risk factors, using a test of hypothesis for data (Usami et al., 2019). Moreover, a range of factors (e.g., risk alpha = 0.05 and power of 0.80. The negative screening 13 4 Research on Child and Adolescent Psychopathology group had to be included to obtain unbiased estimates of Measures prevalence and incidences. A sample of 30% of the negative screening group was considered sufficient. As the planned Dimensions of ODD. The symptoms of the Strengths and follow-up was over 12 years, the sample size was increased Difficulties Questionnaire (SDQ) (Goodman, 1997) conduct by 50% in anticipation of attrition. The research team also problem scale (loses temper, defies rules, argues, spiteful) established that it could manage the size of the final sample. plus four symptoms of DSM-IV ODD not covered by the The final sample for the follow-up was comprised of 622 SDQ (deliberately annoys, blames others, touchy, angry- children (mean age = 3.77 years; SD = 0.33; 96.9% born in resentful) (0 = not true; 2 = certainly true) were used to Spain) followed yearly from ages 3 to 12 years. Table 1 obtain the dimension scores of ODD following Rowe’s shows the descriptive at age 3 for the sample of 621 children 2-factor model (Rowe et al., 2010) (Ezpeleta, Granero, de based on the available data (one family did not complete la Osa, Penelo, & Domènech, 2012). Teachers answered the SDQ from the first follow-up). The retention rates for the questionnaire every year from when the child was aged the follow-ups from age 4 to 12 years were 97.8%, 90.7%, 3 to 12 years. The irritability dimension included three 74.3%, 76.2%, 70.5%, 72.1%, 69.0%, 73.1% and 60.0%, items, ‘touchy-easily annoyed’, ‘angry and resentful’, and respectively. The mean number of assessments was 7.8. ‘loses temper’; the median (Mdn) of the ordinal alpha in Regarding differences due to attrition, the available sample the sample through follow-ups was 0.91. The headstrong/ at age 12 showed lower headstrong/defiant (p = .005) and defiant dimension included five items (‘argues with adults’, lower CU traits (p < .001), no differences in irritability, no ‘defies rules’, ‘deliberately annoys’, ‘blames others’, ‘spite- differences in sex, and higher SES (p = .001) at age 3 than ful’) (Mdn of ordinal alpha = 0.89). The dimensions were those who abandoned. obtained as the sum of the scores of the corresponding The level of nesting between children-teachers-schools items. Higher scores indicated greater irritability and head- was low. A total of 92 schools took part when the children strong/defiant problems. were 12 years old, with 1 or 2 teachers contributing to the Callous-unemotional Traits (CU traits). The Inventory study from 71% of the schools, and up to 5 teachers from of Callous-Unemotional Traits (ICU) (Frick, 2004), which the other schools (the mean was 1.97 and the median was includes 24 items coded on a 4-point Likert-type scale (0: not 2 teachers per school). A total of 122 teachers participated at all true; 3: definitely true), evaluates callous-unemotional at age 12 years, 62.3% of whom evaluated only 1 or 2 chil- traits and was answered by the teachers at each follow-up. dren, with a maximum of 19 children per teacher (the mean The total score is the sum of the raw scores, providing what was 2.81 and the median was 2 children per teacher). The is assumed to be a reliable and valid continuous measure of nesting of teachers-schools at the previous ages was similar, CU traits (Ray et al., 2016). Higher scores indicate higher while the nesting of children-teachers was slightly higher, CU-traits. The Mdn of Cronbach’s alpha for the total scores with a maximum mean of 6.81 children per teacher at age 3. through follow-ups was 0.90. Throughout the 10 follow-ups, all 621 children were taught by more than one teacher, while 513 children changed school at least once and only 108 attended the same school Procedure from ages 3 to 12 years. Regarding the most usual teach- ers and the most attended school per child throughout the The Ethics Committee on Animal and Human Experimenta- 10 follow-ups (mode by rows in wide format), 187 teach- tion of the Universitat Autònoma de Barcelona approved the ers and 77 schools were observed, respectively. These 187 project. The families were recruited at the schools and gave teachers reported on between 1 (37.4%) and 12 (0.5%) written consent for the assessment, and the children and children each (M = 3.3, SD = 2.7; Mdn = 2; mode = 1 [n = 70; adolescents gave their assent to participate. All the families i.e., 70 of the 187 teachers rated 1 child]). Regarding the 77 of the 3-year-old children from participating schools were most attended schools, there were between 1 (27.3%) and invited to answer the screening questionnaire. The families 42 (1.3%) children per school (M = 8.1, SD = 7.4; Mdn = 7; who agreed and met the screening criteria were requested mode = 1 [n = 21; i.e., 21 of the 77 schools had only 1 child). permission to ask the teachers to answer the questionnaires. Given that the level of nesting regarding teachers could Over the 10 follow-ups teachers knew the children a mean of therefore be considered negligible, we nested according to 8.9 months (SD: 3.7) before answering the questionnaires, the most attended school over follow-ups. with a minimum of 7.5 and a maximum of 11.2 months at ages 6 and 7 years, respectively. 13 Research on Child and Adolescent Psychopathology 5 Statistical Analysis loading on all the observed measures of the construct in question (ages 3 to 12 years) and setting factor loadings to The statistical analyses were carried out using Mplus8.7. 1. In the cross-lagged within-person part, the variables at Given the multistage sampling process, the analyses were each time point were represented as latent factors loading on weighted by the inverse probability of selection in the sec- the corresponding observed variable at that point, setting the ond phase of sampling. No correction of the global type I loading at 1 and fixing the variance in the observed variable error rate was carried out. We fitted a Cross-Lagged Panel at 0, thereby transferring the variance to the latent variable. Model with Random Intercepts (RI-CLPM; Hamaker et al., The latent variables at time point t were regressed on all the 2015) to separate between-person effects from within-person variables at t−1, enabling concurrent residuals to correlate. effects and thereby adjust for all unmeasured time-invariant Hence, the latent measures of irritability, headstrong/defi- confounding. The model consisted of a between-person part ant, and CU traits at each time point all captured within- and a within-person part. The between-person time-invari- person variance (i.e., deviation from the child’s own mean) ant part had three latent variables, one for each construct, and were consequently free of unmeasured time-invariant Fig. 1 Schematic example for a RI-CLPM of relationship b/w indicators. Note. For the between-person level: X# and Y# (# corresponds to each wave) are observed indicators over time (whose variances are constrained to 0); RI-x and RI-y represent underlying latent stability/trait over time, and their correlation is denoted by i; factor loadings of observed scores X and Y on this underlying latent stability/trait of over-time are denoted by x and y paths, respectively (which are constrained to 1); paths g# and h# indicate time-specific latent variables for observed X# and Y#, respectively.For the within-person level: cX# and cY# denote time-specific latent variables for X and Y, respectively. Paths a and b indicate cross- lagged effects from X to Y and vice versa; paths d and e indicate autoregressive paths from one occasion to the next within X and Y, respectively; and f# (double-headed arrows) indicate contemporaneous correlations between X and Y at each time point (f1 as correlation between the two constructs at the first time point, and f2 and f3 as correlated residuals between the two constructs at subsequent time points). 13 6 Research on Child and Adolescent Psychopathology Table 1 Description of the Sample Analyzed 0.18–0.82; headstrong/defiant-CU traits: 0.10–0.72; irrita- At age 3 (N = 621) bility-CU traits: −0.01–0.60). Moreover, correlation values Age (years); M (SD) 3.8 (0.33) between each pair of different processes in the same wave Sex; % Female 50.4 ranged from 0.32 to 0.82 in absolute value (irritability- SES; % High 35.4 headstrong/defiant: 0.62–0.82; headstrong/defiant-CU Medium-High/Medium 46.3 Medium-low/Low 18.3 traits: 0.55–0.72; irritability-CU traits: 0.32–0.60). For the Born in Spain; % Yes 97.2 RI-CLPM, covariance coverage of data ranged from 50.2 Ethnicity; % Caucasian 91.1 to 99.8%. Latino 4.7 Other 4.2 Model Fits confounding, whereas the correlations between the random intercept factors captured the between-person associations. Considering all parameters freely estimated, the goodness- In other words, we tested whether changes in irritability, of-fit indices for the baseline model (configural model) headstrong/defiant, and CU traits predicted later changes in established were acceptable [χ2 (318) = 453.7, CFI = 0.978, these constructs, using participants as their own controls. TLI = 0.970, RMSEA = 0.026 (CI90%: 0.021–0.032)]. For the sake of clarity, the current 3-indicator and 10-wave Complete equivalence was not achieved [Δχ2 (72) = 108.5, model is schematically summarized in Fig. 1, showing the p = .004] when fixing each type of lagged parameter (cross- simplest possible 2-indicator and 3-wave model. lagged between pairs of processes and autoregressive RI-CLPM was conducted using the Robust Maximum within each process) to be equivalent over time. Almost full Likelihood (MLR) method of estimation, which is robust equivalence was attained after freeing one parameter (see to non-normality for continuous dependent variables. It is below) based on modification indices [Δχ2 (71) = 100.8, also a full information method (Enders & Bandalos, 2001; p = .011], and this final model showed a satisfactory fit [χ2 Graham, 2009). To account for the hierarchical data struc- (389) = 554.4, CFI = 0.973, TLI = 0.970, RMSEA = 0.026 ture due to cluster sampling, nesting of children according (CI90%: 0.021–0.031]. All cross-lagged parameters to the most attended school throughout the 10 follow-ups between pairs of processes were equivalent over time. All was performed with the Type = COMPLEX and CLUSTER but one of the autoregressive parameters were also equiva- commands in MPlus. Goodness-of-fit was assessed using lent over time. The unique non-equivalent autoregressive the common fit indices: scaled χ2, the Comparative Fit Index path was irritability at age 6 on age 5 (p = .805), which was (CFI), the Tucker-Lewis Index (TLI), and the Root Mean not statistically significant. The rest of the autoregressive Square Error of Approximation (RMSEA). Reasonable parameters over time showed a small or close to medium and adequate model fits above 0.90 and 0.95, respectively, effect size, all of which were statistically significant (r 0.16– were considered for CFI and TLI, and below 0.08 and 0.06, 0.21, p < .001). respectively, for RMSEA (Hu & Bentler, 1999). A base- line model with all parameters freely estimated was estab- lished to test equivalence between types of associations, Relation Between Headstrong/Defiant, and invariance for paths over time was subsequently tested Irritability, and CU Traits using the scaled chi-square difference (Bryant & Satorra, 2012) for nested models (α level set at 0.01). Magnitudes of The standardized parameters for the final model are shown effect sizes for path coefficients were considered following in Table 1. The cross-lagged effects from headstrong/defi- Kline (1998): < 0.10 was regarded as small, around 0.30 ant to irritability were small but statistically significant and medium, and > 0.50 large. positive (0.061 to 0.079, p ≤ .029), meaning that the children that displaying increased headstrong/defiant behavior in a given year relative to their average level over time evinced Results increased irritability the following year. Moreover, when children showed increased headstrong/defiant behavior they Absolute values of Pearson’s correlation coefficients for were likely to exhibit increased CU traits at the next assess- observed scores of the same process between the 10 waves ment point (small effect size paths with standardized values (Supplementary Table S1) ranged from 0.18 to 0.63 (irri- ranging from 0.057 to 0.079, but all of them statistically tability: 0.18–0.57; headstrong/defiant: 0.30–0.63; CU significant, p ≤ .042). The reverse was small but also signifi- traits: 0.18–0.61), and between each pair of different pro- cant, given that when children showed increased CU traits cesses from − 0.01 to 0.82 (irritability-headstrong/defiant: they were likely to exhibit increased headstrong/defiant 13 Research on Child and Adolescent Psychopathology 7 behavior at the next assessment point (standardized path p < .001), between headstrong/defiant and CU traits (medium ranging from 0.051 to 0.065, p ≤ .043). Notably, the cross- or large effect sizes, r 0.36–0.65, p < .001), and between lagged effects from irritability to CU traits, from CU traits to irritability and CU traits (four with small effect sizes and irritability, and from irritability to headstrong/defiant were the rest medium or large, r 0.18–0.54, p ≤ .002) (Table 2, not statistically significant (p ≥ .232). bottom). At the between-person level, the three constructs Significant contemporaneous associations (within-time were moderately to highly correlated (r headstrong/defiant- correlations reflecting within-person change associations) irritability 0.89, r headstrong/defiant-CU traits 0.73, r irri- were found at all the ages between headstrong/defiant tability-CU traits 0.59; all p < .001), indicating that children and irritability (medium or large effect sizes, r 0.47–0.75, with higher levels of headstrong/defiant, irritability, or CU Table 2 Standardized Parameters (p-value) for the RI-CLPM Paths Indicator/s Wave Age 3 Age 4 Age 5 Age 6 Age 7 Age 8 Age 9 -> Age 10 Age 11 -> 4 -> 5 -> 6 -> 7 -> 8 -> 9 10 -> 11 -> 12 Cross- Irritability -> - 0.014 0.014 0.015 0.014 0.013 0.015 0.013 0.013 0.012 lagged 1 Defiant (0.601) (0.606) (0.606) (0.605) (0.603) (0.604) (0.603) (0.600) (0.601) Defiant -> - 0.072 0.079 0.070 0.065 0.061 0.071 0.066 0.072 0.068 Irritability (0.025) (0.023) (0.025) (0.019) (0.026) (0.023) (0.029) (0.025) (0.021) Defiant -> CU - 0.069 0.079 0.064 0.066 0.057 0.071 0.061 0.064 0.062 (0.028) (0.030) (0.033) (0.036) (0.042) (0.034) (0.037) (0.024) (0.031) CU -> Defiant - 0.053 0.058 0.060 0.064 0.051 0.065 0.054 0.057 0.056 (0.027) (0.031) (0.031) (0.030) (0.038) (0.029) (0.034) (0.043) (0.030) Irritability -> - −0.035 −0.033 −0.030 −0.033 −0.028 −0.033 −0.029 −0.029 −0.027 CU (0.242) (0.233) (0.241) (0.232) (0.250) (0.238) (0.241) (0.238) (0.252) CU -> - 0.026 0.026 0.025 0.027 0.022 0.027 0.024 0.027 0.025 Irritability (0.307) (0.317) (0.302) (0.317) (0.311) (0.338) (0.319) (0.327) (0.311) Autore- Irritability -> - 0.194 0.176 0.014 0.174 0.160 0.179 0.172 0.175 0.161 gressive Irritability (< 0.001) (< 0.001) (0.805) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (lagged) 2 Defiant -> - 0.174 0.213 0.204 0.184 0.165 0.202 0.173 0.183 0.181 Defiant (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) CU -> CU - 0.197 0.201 0.178 0.212 0.162 0.213 0.177 0.187 0.179 (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) Correla- RIIrritability <-> 0.889 - - - - - - - - - tions 3 RIDefiant (< 0.001) RIDefiant <-> 0.732 - - - - - - - - - RICU (< 0.001) RIIrritability <-> 0.590 - - - - - - - - - RICU (< 0.001) Age 3 Age 4 Age 5 Age 6 Age 7 Age 8 Age 9 Age 10 Age 11 Age 12 Contempo- Irritability 0.551 0.471 0.512 0.633 0.516 0.750 0.534 0.574 0.635 0.633 raneous 4 <-> Defiant (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) Defiant <-> 0.508 0.490 0.513 0.586 0.359 0.646 0.503 0.557 0.623 0.512 CU (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) Irritability 0.296 0.256 .252 0.341 0.183 0.541 0.233 0.344 0.491 0.315 <-> CU (< 0.001) (< 0.001) (< 0.001) (< 0.001) (0.002) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) Factor load- RIIrritability -> 0.527 0.572 0.581 0.588 0.592 0.565 0.581 0.581 0.586 0.560 ings 5 Irritability (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) RIDefiant -> 0.621 0.594 0.646 0.680 0.675 0.630 0.661 0.634 0.628 0.617 Defiant (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) RICU-> CU 0.584 0.597 0.619 0.594 0.636 0.573 0.617 0.588 0.582 0.558 (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) (< 0.001) Note. Defiant Problems based on Rowe’s Headstrong/Defiant dimension; Irritability based on Rowe’s Irritability dimension; CU: Callousness based on ICU-teachers; RI: random intercept In bold: p < .05; in italics: non-equivalent lagged path (cross-lagged or autoregressive) parameter over contiguous follow-ups 1 Paths a and b in Fig. 1; 2 Paths d and e in Fig. 1; 3 and correlation between RIs in Fig. 1; 4 f# correlations in Fig. 1; 5 Factor loadings x# and y# in Fig. 1 (paths g# and h# in Fig. 1 are not detailed; all these values were statistically significant at the 0.001 level ranging between 0.806–0.850 for Irritability, 0.733–0.804 for Defiant, and 0.772–0.830 for CU) 13 8 Research on Child and Adolescent Psychopathology traits than other children also tended to present higher levels punishment (Frick & Viding, 2009; Matthys et al., 2012). of the other problems than other children. Therefore, it is possible that time-varying changes in these intrapersonal factors influence both headstrong/defiant and CU behavior, also longitudinally, creating a spurious relation Discussion between the two. In any event, this pattern of association, which reflects severe disruptive behavior at school, suggests This is the first study to analyze how irritability, head- that teachers may need special support to manage defiant strong/defiant dimensions, and CU traits are longitudinally behavior and lack of empathy, lack of guilt, and insensi- interrelated. We did so in a general population of children tivity to others’ distress. At the same time, these findings assessed yearly over a 10-year period from ages 3 to 12 highlight the important role teachers can play in preventing years using RI-CLPM, thereby adjusting for all unmeasured escalation between different forms of aggression (reactive time-invariant confounding using the children as their own and instrumental) typically associated with headstrong/defi- controls. This was done at the school, an important develop- ant and callous traits. According to Stringaris and Good- mental context, where ODD may be both shaped and influ- man (2009), decomposing ODD in different dimensions can ence development. From preschool to early adolescence, help to identify their etiological variability and may help to a reciprocal cross-lagged association was found between predict different developmental trajectories. In this line, by headstrong/defiant and CU traits, and a unidirectional rela- breaking ODD down into dimensions, reciprocal associa- tion was found from headstrong/defiant to irritability. Chil- tions could be found that may not have emerged in previous dren who increased in headstrong/defiant behavior at one cross-lagged analyses, such as Servera et al. (2020), who point evinced increased CU traits at the subsequent point in found that higher levels of ODD at age 6 predicted more time and vice versa. Moreover, increased headstrong/defi- CU traits at age 9, while the reverse association was not sig- ant behavior at one year forecasted increased irritability the nificant and higher levels of CU traits did not predict ODD. subsequent year. Irritability did not show any cross-lagged Additionally, previous research has not adjusted for the fact association with the other variables. Of the two competing that CU traits and ODD may have a common etiology, pro- hypotheses about the relation between irritability and head- ducing spurious relations between them. Our results suggest strong/defiant, our findings are consistent with the view that that the association between ODD and CU traits, displayed they are not mutually involved in their etiology but have in the subtype of ODD with LPE, may be due to its relation different developmental courses. Conversely, the findings with headstrong/defiant. agree with the hypothesis that headstrong/defiant and CU Although there is extensive literature available on ODD traits influence each other. Possibly, from preschool age dimensions, few studies have tested their mutual develop- to early adolescence, headstrong/defiant behavior and CU mental association. Therefore, we contrasted two opposing traits at school may act longitudinally as mutual associated hypotheses regarding whether they were reciprocally asso- factors and headstrong/defiant behavior may drive increases ciated or not. In line with the results of Whelan et al. (2013), in irritability, whereas irritability is likely not involved in the headstrong/defiant predicted irritability. Increased head- development of headstrong/defiant behaviors or CU traits. strong/defiant one year consistently predicted increased Using a within-subject approach, headstrong/defiant irritability the subsequent year from ages 3 to 12 years. and CU traits were consistently and reciprocally associ- Unrewarded defying, arguing, bothering others, etc., at ated throughout development. This means that adjusting for school may cause frustration and subsequent increased the child’s underlying level of these problems, those who irritability. In this line, abnormal irritability is explained as increased in insensitivity to others and in lack of empathy an aberrant response to frustrative non-reward and threat and guilt (i.e., CU) also evinced a developmental propen- (Brotman et al., 2017). The relation was in one direction, sity towards subsequent increased defiance and vice versa. implying that what the child does determines how they feel The present study was not positioned to explained why this and so focusing intervention on improving misbehavior at should be so. Even though there is a direct link via interper- school may, as a result, improve affect. Hence, headstrong/ sonal or other mechanisms (e.g., defiant behavior contribut- defiant is a target to diminish irritability at school. ing to coercive cycles with parents and teachers, thus fueling Irritability is considered a core component of ODD (Evans a callous view of others), a range of time-varying factors et al., 2017). Most children with ODD present marked irrita- may also contribute to prospective associations between the bility (Ezpeleta et al., 2016; Rowe et al., 2010). However, in two. In this regard, we should consider that children with both our study and Whelan’s et al. (2013), irritability did not headstrong/defiant and CU traits share difficulties in social show any cross-lagged association with the other variables, learning, and both show dysfunctions in delaying gratifica- acting as a distinct developmental dimension needing to be tion, emphasizing reward, and minimizing the effects of addressed independently. The distinct developmental course 13 Research on Child and Adolescent Psychopathology 9 of irritability may be showing its ubiquitous non-specificity expected, the prevalence of psychopathology was low. The in different disorders and their transdiagnostic nature (Wak- resulting restricted variability may have implied that some schlag et al., 2018). This approach is in line with the dimen- associations did not emerge. Furthermore, the size of the sional Research Domain Criteria (Insel et al., 2010), which sample meant that we could not study the mediating indirect proposes the identification of underlying cross-cutting effects that may have explained further associations, such dimensions of individual functioning. as between irritability and headstrong/defiant and CU traits. Within-person stability for all three variables was low but Studies with larger samples are needed to further investi- significant, with only one exception (autoregressive param- gate the indirect mechanisms between these variables. Last, eters around just under 0.20). The positive within-time cor- although attrition was not high considering the length of the relations indicate that within-child increases in irritability, study, the results must be interpreted with caution given that headstrong/defiant, or CU traits above their mean level are the cases that dropped out were those with a lower SES, accompanied by child-specific increases in the same respec- possibly introducing a bias in the outcomes. Notably, our tive variable compared to their average. The previous litera- decision to not correct the global type I error rate may have ture reports a moderate stability for irritability (Beauchaine led to reporting associations as significant that would be & Tackett, 2020), headstrong/defiant (Whelan et al., 2013), deemed non-significant under a more conservative statis- and CU traits (Wakschlag et al., 2018), but these studies tical approach. Furthermore, correcting the p-values in an contain information on the between-person rank-order sta- exploratory analysis like ours could have hidden important bility of these variables and so it was to be expected that associations which may have served as hypotheses for fur- our parameters would be lower as they only included one ther investigation (Armstrong, 2014). source of variation (within-person). Stability for CU traits, Ascertaining how irritability, headstrong/defiant, and CU for example, may be due to common causes such as genetic traits covary may aid a better understanding of the develop- influences or fearlessness (Frick et al., 2014), having an ment of behavior problems at school and may be informative effect throughout development; or it may also be argued that for intervention planning. The present results have potential CU traits have effects producing more CU traits through a value for early targeting prevention of ODD through its range of feedback-loops involving, for example, parents or dimensions that goes beyond the concurrent and long-term peers. Stable causes leading to stability could at least be comorbidity associated with these conditions (depression, ruled out by removing time-invariant common causes. On anxiety, CD, substance abuse) (Nock et al., 2007). Given the other hand, the trait-like between-person differences that irritability forecasts changes in the other variables, the captured by the random intercepts were moderate to high results highlight the need to intervene in this dimension (rank 0.59 to 0.89), showing that children with higher levels directly or through improving headstrong/defiant behav- in these measures than other children tended to score higher iors. Targeting headstrong/defiant behavior may also help than other children on the other two measures. Although the to prevent the impact of diverse aggressive behavior and the reasons for these rather high between-person associations increase of CU traits through development, and vice versa. cannot be discerned by the present design, common genes, As headstrong/defiant and CU traits reciprocally relate, it temperamental traits, stable parenting styles or social adver- may be helpful to intervene in their common processes such sity may be at play. as reward and punishment processing. Schools are an ideal This is the first study to report on the developmental links setting to provide interventions for young children at risk of between irritability, headstrong/defiant dimensions, and CU mental health disorders (Guerra et al., 2019). They are not traits simultaneously and using the RI-CLPM methodology. only an opportunity for children to learn and to generalize The strengths are the length of the longitudinal study, with social and emotional skills, but they also hold potential for 10- year follow-ups from preschool to early adolescence and addressing significant gaps in children’s mental health ser- a low attrition rate in a community sample, the dimensional vice delivery, supported by estimates that only around half assessment of ODD dimensions, and the availability of a the children needing services actually receive them (Meri- measure to specifically assess CU traits. However, some lim- kangas et al., 2009). Systematic reviews show that school- itations should be considered when interpreting the results. based mental health interventions can be effective (Baskin First, as the focus of the study was school, the teachers were et al., 2010) and many evidence-based programs can be the only reporters. However, different teachers reported over implemented in schools (Kratochwill et al., 2008). Our the ten-year period, and therefore common method variance results indicate that school intervention programs should should not have inflated prospective associations. Further- have two essential components: a training block for teachers more, teachers are one of the primary sources of referral for in behavior modification techniques to manage the behavior treatment of children’s problems (De los Reyes & Kazdin, of headstrong/defiant children, and a second component for 2005). Second, we studied a community sample in which, as 13 10 Research on Child and Adolescent Psychopathology Beauchaine, T. P., & Tackett, J. L. (2020). Irritability as a transdiag- children aimed at promoting strategies for emotional self- nostic vulnerability trait: current issues and future directions. regulation and empathic capacity-prosocial behavior. Behavior Therapy, 51(2), 350–364. https://doi.org/10.1016/j. beth.2019.10.009 Supplementary Information The online version contains Blair, R. J. R. (2018). Traits of empathy and anger: Implications for supplementary material available at https://doi.org/10.1007/s10802- psychopathy and other disorders associated with aggression. 022-00910-8. Philosophical Transactions of the Royal Society B: Biological Sciences, 373, 1744. https://doi.org/10.1098/rstb.2017.0155 Funding This work was supported by the Spanish Ministry of Science, Booker, J. A., Capriola-Hall, N. N., Greene, R. W., & Ollendick, T. Innovation and Universities [Grant PGC2018-095239-B-I00 (MICIU/ H. (2020). The parent-child relationship and posttreatment child FEDER)] outcomes across two treatments for oppositional defiant disor- der. Journal of Clinical Child and Adolescent Psychology, 49(3), 405–419. https://doi.org/10.1080/15374416.2018.1555761 Open Access Funding provided by Universitat Autonoma de Barcelona. Brotman, M. A., Kircanski, K., Stringaris, A., Pine, D. S., & Leiben- luft, E. (2017). Irritability in Youths: A Translational Model. Code Availability Not applicable. American Journal of Psychiatry, 174(6), 520–532. https://doi. org/10.1176/appi.ajp.2016.16070839 Compliance with Ethical Standards Bryant, F. B., & Satorra, A. (2012). Principles and practice of scaled difference chi-square testing. Structural Equation Modeling: A Conflicts of Interest The authors declare that they have no conflicts Multidisciplinary Journal, 19, 372–398. https://doi.org/10.1080 of interest. /10705511.2012.687671 Burke, J. D., Boylan, K., Rowe, R., Duku, E., Stepp, S. D., Hipwell, A. E., & Waldman, I. D. (2014). Identifying the irritability dimen- Ethics Approval The project was approved by the Ethics Committee sion of odd: Application of a modified bifactor model across five on Animal and Human Experimentation of the Universitat Autònoma large community samples of children. Journal of Abnormal Psy- de Barcelona. chology, 123(4), 841–851. https://doi.org/10.1037/a0037898 Crapanzano, A. M., Frick, P. J., & Terranova, A. M. (2010). Patterns Consent to Participate Families gave written consent for the assess- of physical and relational aggression in a school-based sample ment and assent were collected from participants. of boys and girls. Journal of Abnormal Child Psychology, 38(4), 433–445. https://doi.org/10.1007/s10802-009-9376-3 Consent for Publication All the authors have reviewed the manuscript Deters, R. K., Naaijen, J., Rosa, M., Aggensteiner, P. M., Ban- and consented to its publication. aschewski, T., Saam, M. C. … Dietrich, A. (2020). Executive functioning and emotion recognition in youth with oppositional Open Access This article is licensed under a Creative Commons defiant disorder and/or conduct disorder. World Journal of Bio- Attribution 4.0 International License, which permits use, sharing, logical Psychiatry, 21(7), 539–551. https://doi.org/10.1080/1562 adaptation, distribution and reproduction in any medium or format, 2975.2020.1747114 as long as you give appropriate credit to the original author(s) and the Enders, C. K., & Bandalos, D. L. (2001). The relative performance of source, provide a link to the Creative Commons licence, and indicate full information maximum likelihood estimation for missing data if changes were made. The images or other third party material in this in structural equation models. Structural Equation Modeling, 8, article are included in the article’s Creative Commons licence, unless 430–457. https://doi.org/10.1207/S15328007SEM0803_5 indicated otherwise in a credit line to the material. If material is not Evans, S. C., Burke, J. D., Roberts, M. C., Fite, P. J., Lochman, J. E., included in the article’s Creative Commons licence and your intended de la Peña, F. R., & Reed, G. M. (2017). Irritability in child and use is not permitted by statutory regulation or exceeds the permitted adolescent psychopathology: An integrative review for ICD-11. use, you will need to obtain permission directly from the copyright Clinical Psychology Review, 53, 29–45. https://doi.org/10.1016/j. holder. To view a copy of this licence, visit http://creativecommons. cpr.2017.01.004 org/licenses/by/4.0/. Evans, S. C., Pederson, C. A., Fite, P. J., Blossom, J. B., & Cooley, J. L. (2016). Teacher-Reported Irritable and Defiant Dimensions of Oppositional Defiant Disorder: Social, Behavioral, and Aca- demic Correlates. School Mental Health, 8(2), 292–304. https:// References doi.org/10.1007/s12310-015-9163-y Ezpeleta, L., de la Osa, N., & Doménech, J. M. (2014). Prevalence Allison, T. C., Miller, A. K., & Inman, D. J. (2009). A time-varying of DSM-IV disorders, comorbidity and impairment in 3-year- identification method for mixed response measurements. Jour- old Spanish preschoolers. Social Psychiatry and Psychiat- nal of Sound and Vibration, 319(3–5), 850–868. https://doi. ric Epidemiology, 49(1), 145–155. https://doi.org/10.1007/ org/10.1016/j.jsv.2008.06.031 s00127-013-0683-1 American Psychiatric Association. (2013). Diagnostic and statisti- Ezpeleta, L., Granero, R., de la Osa, N., Penelo, E., & Doménech, J. M. cal manual of mental disorders (5th ed.). American Psychiatric (2012). Dimensions of oppositional defiant disorder in 3-year-old Association preschoolers. Journal of Child Psychology and Psychiatry, 53(11), Armstrong, R. A. (2014). When to use the Bonferroni correction. 1128–1138. https://doi.org/10.1111/j.1469-7610.2012.02545.x Ophthalmic & Physiological Optics, 34, 502–508. https://doi. Ezpeleta, L., Granero, R., de la Osa, N., Trepat, E., & Domènech, org/10.1111/opo.12131 J. M. (2016). Trajectories of oppositional defiant disorder irri- Baskin, T. W., Slaten, C. D., Sorenson, C., Glover-Russell, J., & Mer- tability symptoms in preschool children. Journal of Abnor- son, D. N. (2010). Does youth psychotherapy improve academi- mal Child Psychology, 44, 115–128. https://doi.org/10.1007/ cally related outcomes? A Meta-Analysis. Journal of Counseling s10802-015-9972-3 Psychology, 57(3), 290–296. https://doi.org/10.1037/a0019652 13 Research on Child and Adolescent Psychopathology 11 Frick, P. J. (2004). The Inventory of Callous-Unemotional traits. Matthys, W., Vanderschuren, L. J. M. J., Schutter, D. J. L. G., & Loch- Unpublished rating scale. http://psyc.uno.edu/Frick%20Lab/ICU. man, J. E. (2012). Impaired neurocognitive functions affect social html learning processes in oppositional defiant disorder and conduct Frick, P. J., Ray, J. V., Thornton, L. C., & Kahn, R. E. (2014). A disorder: Implications for interventions. Clinical Child and Fam- developmental psychopathology approach to understanding cal- ily Psychology Review, 15(3), 234–246. https://doi.org/10.1007/ lous-unemotional traits in children and adolescents with serious s10567-012-0118-7 conduct problems. Journal of Child Psychology & Psychiatry, Merikangas, K. R., Nakamura, E. F., & Kessler, R. C. (2009). Epi- 55, 532–548. https://doi.org/10.1111/jcpp.12152 demiology of mental disorders in children and adolescents. Dia- Frick, P. J., & Viding, E. (2009). Antisocial behavior from a develop- logues in clinical neuroscience, 11(1), 7–20 mental psychopathology perspective. Development and Psycho- Morken, I. S., Viddal, K. R., Ranum, B., & Wichstrom, L. (2021). pathology, 21, 1111–1131 Depression from preschool to adolescence - five faces of stabil- Goodman, R. (1997). The Strengths and Difficulties Questionnaire: A ity. Journal of Child Psychology and Psychiatry, 62, 1000–1009. Research Note. Journal of Child Psychology and Psychiatry, 38, https://doi.org/10.1111/jcpp.13362 581–586. https://doi.org/10.1111/j.1469-7610.1997.tb01545.x Nock, M. K., Kazdin, A. E., Hiripi, E., & Kessler, R. C. (2007). Life- Graham, J. W. (2009). Missing data analysis: Making it work in the time prevalence, correlates, and persistence of oppositional defi- real world. Annual Review of Psychology, 60, 549–576. https:// ant disorder: Results from the National Comorbidity Survey doi.org/10.1146/annurev.psych.58.110405.085530 Replication. Journal of Child Psychology & Psychiatry, 48, 703– Griffith, S. F., Arnold, D. H., Rolon-Arroyo, B., & Harvey, E. A. (2019). 713. https://doi.org/10.1111/j.1469-7610.2007.01733.x Neuropsychological Predictors of ODD Symptom Dimensions in Ray, J. V., Frick, P. J., Thornton, L. C., Steinberg, L., & Cauffman, Young Children. Journal of Clinical Child and Adolescent Psy- E. (2016). Positive and negative item wording and its influence chology, 48(1), 80–92. https://doi.org/10.1080/15374416.2016.1 on the assessment of callous-unemotional traits. Psychological 266643 Assessment, 28(4), 394–404. https://doi.org/10.1037/pas0000183 Guerra, L. A., Rajan, S., & Roberts, K. J. (2019). The Implementation Rowe, R., Costello, E. J., Angold, A., Copeland, W. E., & Maughan, B. of Mental Health Policies and Practices in Schools: An Examina- (2010). Developmental pathways in oppositional defiant disorder tion of School and State Factors. Journal of School Health, 89(4), and conduct disorder. Journal of Abnormal Psychology, 119(4), 328–338. https://doi.org/10.1111/josh.12738 726–738. https://doi.org/10.1037/a0020798 Hamaker, E. L., Kuiper, R. M., & Grasman, R. P. P. P. (2015). A cri- Servera, M., Seijas, R., Garcia-Banda, G., Barry, C. T., Beauchaine, T. tique of the crosslagged panel model. Psychological Methods, 20, P., & Burns, G. L. (2020). Longitudinal associations of callous- 102–116. https://doi.org/10.1037/a0038889 unemotional and oppositional defiant behaviors over a three-year Hawes, D. J., Dadds, M. R., Brennan, J., Rhodes, T., & Cauchi, A. interval for Spanish children. Development and Psychopathology, (2013). Revisiting the treatment of conduct problems in chil- 32(2), 481–490. https://doi.org/10.1017/s0954579419000221 dren with callous-unemotional traits. Australian and New Statistical Solutions, L. (2007). nQuery Advisor 7.0. In Statistical Zealand Journal of Psychiatry, 47(7), 646–653. https://doi. Solutions Ltd. www.statsol.ie org/10.1177/0004867413484092 Stringaris, A., & Goodman, R. (2009). Three dimensions of Hu, L. T., & Bentler, P. (1999). Cutoff criteria for fit indexes in covari- oppositionality in youth [Article]. Journal of Child Psy- ance structure analysis: Conventional criteria versus new alter- chology and Psychiatry, 50(3), 216–223. https://doi. natives. Structural Equation Modeling, 6, 1–55. https://doi. org/10.1111/j.1469-7610.2008.01989.x org/10.1080/10705519909540118 Stringaris, A., Vidal-Ribas, P., Brotman, M. A., & Leibenluft, E. Insel, T., Cuthbert, B., Garvey, M., Heinssen, R., Pine, D. S., Quinn, K. (2018). Practitioner Review: Definition, recognition, and treat- … Wang, P. (2010). Research Domain Criteria (RDoC): Toward ment challenges of irritability in young people. Journal of Child a New Classification Framework for Research on Mental Disor- Psychology and Psychiatry, 59, 721–739. https://doi.org/10.1111/ ders. American Journal of Psychiatry, 167(7), 748–751. https:// jcpp.12823 doi.org/10.1176/appi.ajp.2010.09091379 Usami, S., Todo, N., & Murayama, K. (2019). Modeling reciprocal Kline, R. B. (1998). Principles and practice of structural equation effects in medical research: Critical discussion on the current modeling. Guilford practices and potential alternative models. PloS One, 14(9), Arti- Kratochwill, T., Hoagwood, K. E., Frank, J. L., Mass-Levitt, J., cle e0209133. https://doi.org/10.1371/journal.pone.0209133 Romanelli, L. H., & Saka, N. (2008). Evidence-based practices in Wakschlag, L. S., Choi, S. W., Carter, A. S., Hullsiek, H., Burns, school psychology: Challenges and opportunities for the profes- J., McCarthy, K. … Briggs-Gowan, M. J. (2012). Defining the sion. In C. R. Reynolds, & T. B. Gutkin (Eds.), The handbook of developmental parameters of temper loss in early childhood: school psychology (4 ed., pp. 25–37). Wiley implications for developmental psychopathology. Journal of Krieger, F. V., Polanczyk, V. G., Robert, G., Rohde, L. A., Graeff- Child Psychology and Psychiatry, 53(11), 1099–1108. https://doi. Martins, A. S., Salum, G. … Stringaris, A. (2013). Dimensions org/10.1111/j.1469-7610.2012.02595.x of oppositionality in a Brazilian community sample: Testing the Wakschlag, L. S., Perlman, S. B., Blair, R. J., Leibenluft, E., Briggs- DSM-5 proposal and etiological links. Journal of the American Gowan, M. J., & Pine, D. S. (2018). The neurodevelopmental basis Academy of Child and Adolescent Psychiatry, 52(4), 389–400. of early childhood disruptive behavior: Irritable and callous phe- https://doi.org/10.1016/j.jaac.2013.01.004 notypes as exemplars. American Journal of Psychiatry, 175(2), Longman, T., Hawes, D. J., & Kohlhoff, J. (2016). Callous-unemotional 114–130. https://doi.org/10.1176/appi.ajp.2017.17010045 traits as markers for conduct problem severity in early childhood: Waldman, I. D., Rowe, R., Boylan, K., & Burke, J. D. (2021). Exter- A meta-analysis. Child Psychiatry and Human Development, nal validation of a bifactor model of oppositional defiant disor- 47(2), 326–334. https://doi.org/10.1007/s10578-015-0564-9 der. Molecular Psychiatry, 26, 682–693. https://doi.org/10.1038/ Matthys, W., Vanderschuren, L. J. M. J., & Schutter, D. J. L. G. s41380-018-0294-z (2013). The neurobiology of oppositional defiant disorder and Waller, R., & Hyde, L. W. (2018). Callous-unemotional behaviors in conduct disorder: Altered functioning in three mental domains. early childhood: the development of empathy and prosociality Development and Psychopathology, 25(1), 193–207. https://doi. gone awry. Current Opinion in Psychology, 20, 11–16. https:// org/10.1017/s0954579412000272 doi.org/10.1016/j.copsyc.2017.07.037 13 12 Research on Child and Adolescent Psychopathology Waller, R., Shaw, D. S., Neiderhiser, J. M., Ganiban, J. M., Natsuaki, Whelan, Y. M., Stringaris, A., Maughan, B., & Barker, E. D. (2013). M. N., Reiss, D. … Hyde, L. W. (2017). Toward an understand- Developmental continuity of oppositional defiant disorder sub- ing of the role of the environment in the development of early dimensions at ages 8, 10, and 13 years and their distinct psy- callous behavior. Journal of Personality, 85, 90–103. https://doi. chiatric outcomes at age 16 years. Journal of the American org/10.1111/jopy.12221 Academy of Child Adolescent Psychiatry, 52, 961–969. https:// Waschbusch, D. A., Baweja, R., Babinski, D. E., Mayes, S. D., & Wax- doi.org/10.1016/j.jaac.2013.06.013 monsky, J. G. (2020). Irritability and limited prosocial emotions/ Willoughby, M. T., Waschbusch, D. A., Moore, G. A., & Propper, C. callous-unemotional traits in elementary-school-age children. B. (2011). Using the ASEBA to screen for callous unemotional Behavior Therapy, 51(2), 223–237. https://doi.org/10.1016/j. traits in early childhood: Factor structure, temporal stability, and beth.2019.06.007 utility. Journal of Psychopathology and Behavioral Assessment, Wesselhoeft, R., Stringaris, A., Sibbersen, C., Kristensen, R. V., Boje- 33, 19–30. https://doi.org/10.1007/s10862-010-9195-4 sen, A. B., & Talati, A. (2019). Dimensions and subtypes of World Health Organization (2019). The ICD-11 classification of mental oppositionality and their relation to comorbidity and psychoso- and behavioural disorders. Author. https://icd.who.int/browse11/ cial characteristics. European Child and Adolescent Psychiatry, lm/en#/http://id.who.int/icd/entity/1545599508 28(3), 351–365. https://doi.org/10.1007/s00787-018-1199-8 Publisher’s Note Springer Nature remains neutral with regard to juris- dictional claims in published maps and institutional affiliations. 13