Abstract
ADHD is diagnosed on the basis of impairing symptoms across two core dimensions: hyperactivity-impulsivity and inattention (American Psychiatric Association [APA], 2000). Emotional lability (EL) has, from the publication of Diagnostic and Statistical Manual of Mental Disorders (3rd ed.; DSM-III; APA, 1980), been defined as an associated feature of ADHD. This view has been challenged (Barkley & Fischer, 2010), with recent support from a twin study strengthening the hypothesis of EL as etiologically relevant to the core ADHD phenotype (Merwood et al., 2014).
Mood- and emotion-related terms have been labeled differently across research groups: emotion dysregulation (Shaw, Stringaris, Nigg, & Leibenluft, 2014), deficient emotional self-regulation (Biederman et al., 2012), severe mood dysregulation (Leibenluft, 2011), EL (Sobanski et al., 2010), mood instability (Skirrow, McLoughlin, Kuntsi, & Asherson, 2009), and the new Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; APA, 2013) Disruptive Mood Dysregulation Diagnosis. All terms have in common descriptions of inappropriate levels of low frustration tolerance, negative mood, chronic irritability, and temper outbursts (APA, 2013; Leibenluft, 2011; Merwood et al., 2014; Shaw et al., 2014; Skirrow et al., 2009). In the present study, we use the term EL descriptively, similarly to how the term has been used in studies on school-age children with ADHD (Anastopoulos et al., 2011; Sobanski et al., 2010).
A bidirectional relationship between EL and ADHD has often been found in the literature. In a study of 358 schoolchildren with ADHD, nearly half were rated by parents as showing an impairing level of EL (Anastopoulos et al., 2011). Conversely, in a sample referred solely for temper tantrums, 75% were diagnosed with ADHD (Roy et al., 2013). Furthermore, in children with ADHD, EL has been associated with comorbidity (Chronis-Tuscano et al., 2010; Seymour et al., 2012) and poor prognosis (Anastopoulos et al., 2011; Barkley & Fischer, 2010; Maedgen & Carlson, 2000).
However, it has long been recognized that episodes of EL may occur during normal development (Gerson et al., 1996). Furthermore, EL is considered to be a dimensional trait (Shaw et al., 2014) and is not uniquely related to ADHD. In a large epidemiological sample (n = 1,420), severe mood dysregulation occurred in 3.3% of the children (Brotman et al., 2006). Still, among these 3%, two thirds fulfilled diagnostic criteria for an Axis I diagnosis, most commonly ADHD (27%), conduct disorder (25.9%), and/or oppositional defiant disorder (ODD; 24.5%), underlining the co-occurrence with several psychiatric disorders. In a large clinical sample of children with ADHD (n = 1,186, age = 6-18 years), hyperactivity/impulsivity, oppositional-, and anxious-shy emotional symptoms accounted for 30% of the variance in EL (Sobanski et al., 2010). In this study, oppositional symptoms were an especially important predictor (Sobanski et al., 2010).
Undoubtedly, there is a conceptual overlap between EL and the irritable dimension of ODD (Stringaris & Goodman, 2009). Thus, the conceptual overlap between EL and the aforementioned psychopathology is methodologically challenging (Caron & Rutter, 1991). Notwithstanding, several studies have shown that EL could not simply be explained by other psychiatric disorders (Barkley & Fischer, 2010; Martel, Nigg, & von Eye, 2009; Sobanski et al., 2010; Surman et al., 2011). Thus, EL has been suggested as a factor that possibly explains the high rate of comorbidity in schoolchildren with ADHD (Sobanski et al., 2010). Whether this comorbidity is present also in a young preschool sample needs to be investigated. To date, EL and ADHD have mostly been studied in school-age children (Skirrow et al., 2009; Sobanski et al., 2010). Nonetheless, research on preschool psychopathology has shown that the diagnostic entities appear to be differentiated much as it is among older children (Sterba, Egger, & Angold, 2007; Strickland et al., 2011), and it seems reasonable to expect a combination of EL and ADHD in a portion of preschoolers. To date, we lack empirical data on young children, which include measures of both psychiatric symptomatology and EL. In addition, most previous studies have relied on parent report only (Anastopoulos et al., 2011; Brotman et al., 2006; Carlson & Youngstrom, 2003; Frick & Nigg, 2012). To what degree EL is reported across settings in preschoolers with ADHD symptoms is not known.
In a previous article, we found that a nonstandardized measure of emotional control (EC) at age 18 months was an unspecific predictor for symptoms of both anxiety and ADHD, as well as co-occurring anxiety and ADHD at age 3 (Overgaard et al., 2014). The aim of the present study was a further investigation of EL in preschool children with symptoms of ADHD, using a standardized measure, through the following research questions:
Method
Participants
The Norwegian Mother and Child Cohort Study (MoBa) is an ongoing birth cohort study (n = 107,000 pregnancies) designed to identify environmental and genetic factors for diseases in pregnancy and childhood (Magnus et al., 2006). In brief, about 89% of the participating mothers were Norwegian and predominantly White Caucasians (Mellingen, Torsheim, & Thuen, 2013). The mean household income equaled US$52,000, the same as the population mean, and the participation rate was about 44% (Nilsen et al., 2009). Questionnaires were completed during pregnancy and when the child was 6, 18, and 36 months. The questionnaire at 36 months included 6 questions from the Child Behavior Checklist (CBCL; Achenbach & Rescorla, 2000) and 5 items from the Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev., DSM-IV-TR; APA, 2000) criteria for ADHD, in total 11 items about hyperactivity, impulsivity, and attention problems. To identify children who at age 3 might be at risk of developing ADHD, children of the MoBa population scoring ≥90th percentile on the 11 questions on hyperactivity, impulsivity, and attention problems were invited to a clinical assessment. Children with severe medical conditions and children with higher scores on autistic symptoms were excluded. In total, 2,798 children were invited from August 2007 to January 2011, and 1,048 (37.5%) participated in the clinical assessments (the ADHD study). There were no statistically significant differences between participants and invited nonparticipants regarding background factors and pre- and perinatal risk factors except for a higher level of maternal education in participants. The clinical assessments were done when the children were 36 to 44 months old and included questionnaire data from parents and preschool teacher on development, temperament, language, and behavior; diagnostic interview with the mother; and neuropsychological testing. The children in the present study (n = 368) were selected when meeting all or nearly all Diagnostic and Statistical Manual of Mental Disorders (4th ed., DSM-IV; APA, 1994) criteria for ADHD according to parent report in the diagnostic interview (see below) and had complete data on EL.
The Regional Committee of Ethics in Medical Research, the Data Inspectorate, the Norwegian Institute of Public Health, and Oslo University Hospital approved the study. Informed written consent was obtained from the parents of the children in the study. The parents also consented to preschool teachers being contacted and answering questionnaires.
Measures
Psychiatric symptoms: Diagnostic assessment of the child was based on the Preschool Age Psychiatric Assessment (PAPA) interview with caregiver (Egger & Angold, 2004). The interview was developed for preschool children from 2 to 5 years of age and included questions on psychiatric symptoms and possible impairment of these symptoms. Symptoms of ADHD (all subtypes), ODD, and anxiety that had persisted for 3 months or longer were included. Impairment was considered present if the parent reported that the child showed at least some impairment in one or more areas. Interviews were conducted by graduate students in psychology trained in administration and scoring of the interview. Specialists in clinical psychology or child psychiatry supervised the scoring of the interview. An interrater reliability check was carried out by a second rater, blind to any knowledge about the child and family, and rescored from audiotapes of 79 randomly selected interviews. The average intraclass correlations (ICCs) were .98 for total number of ADHD symptoms, .99 for inattentive symptoms, and .97 for hyperactivity-impulsivity symptoms. ICC for anxiety and ODD symptoms were .86, and .98, respectively.
The ADHD symptom group included 368 preschoolers who were reported to meet all (n = 161) or nearly all (n = 207) DSM-IV diagnostic criteria for an ADHD subtype. Diagnostic criteria for ADHD included the presence of at least six out of nine symptoms of either inattention or hyperactivity-impulsivity (comprising either the inattentive or the hyperactive/impulsive subtypes of ADHD, respectively) or both (comprising the combined subtype of ADHD) when impairment due to symptoms is also present. Children meeting nearly all criteria for an ADHD diagnosis included those showing six or more symptoms of either subtype without impairment, and children showing subthreshold ADHD (three-five symptoms of either subtype) and were reported to show impairment.
The ODD symptom group included all participants in whom the DSM-IV criteria of four or more of the eight ODD symptoms were present, with and without impairment.
Anxiety symptoms consisted of the anxiety DSM-IV diagnoses known to be most frequent in preschoolers, that is, specific phobia, social anxiety, separation anxiety disorder, or generalized anxiety disorder (Costello, Egger, & Angold, 2005; Muris & Broeren, 2009), which were included if the anxiety symptoms were inappropriate and excessive. An anxiety sum score was derived by adding symptoms from the four specific anxiety symptom subtypes.
In sum, psychiatric symptom sum scores were calculated for ADHD total and subcomponents inattention and hyperactive-impulsive, ODD symptoms, and anxiety symptoms, respectively.
EL was measured with the EC subscale from the preschool version of the Behavior Rating Inventory of Executive Function–Preschool Version (BRIEF-P; Gioia, Espy, & Isquith, 2003) rated by both parents and preschool teachers. BRIEF-P has good test–retest stability and a high internal consistency, and for the relevant subscale, the manual reports Cronbach’s alphas to be .86 (parents) and .91 (teachers), an interinformant correlation of .25, as well as convergent validity by a significant correlation (.65, p < .001) with the Emotional Reactive subscale from the Achenbach System of Empirically Based Assessment (ASEBA) CBCL/1.5 to 5 (Achenbach & Rescorla, 2000; Gioia et al., 2003). In our sample, the parent and teacher EC subscales had Cronbach’s alphas of .84 and .92, very similar to findings from three large preschool studies (Bonillo, Araujo Jimenez, Jane Ballabriga, Capdevila, & Riera, 2012; Duku & Vaillancourt, 2014; Ezpeleta, Granero, Penelo, de la Osa, & Domenech, 2013). The Norwegian BRIEF-P version was recently found to have good psychometric properties, including support for the Emotional Control subscale as a clearly differentiated factor in a study of 1,134 three-year-olds (Skogan et al., 2015). This subscale measures a child’s ability to modulate emotional responses, and these children may show overblown emotional reactions to seemingly minor events. It consists of 10 items: (a) Overreacts to small problems; (b) has explosive, angry outbursts; (c) becomes upset too easily; (d) has outbursts for little reason; (e) mood changes frequently; (f) small events trigger big reactions; (g) angry or tearful outbursts are intense but end suddenly; (h) reacts more strongly to situations than other children; (i) is easily overwhelmed or overstimulated by typical daily activities; and (j) after having a problem, will stay disappointed for a long time. Following scoring instructions, items were scored as 1, 2, or 3; referring to never-, sometimes-, or often a problem, respectively; EC subscales with 1 to 2 missing items were replaced with score 1 (12 parents and 15 teachers). In the present study, we used the term emotional lability (EL) as defined in the BRIEF-P manual where EL is a description of deficient EC and present with high scores on the EC subscale (T-scores ≥ 65; Gioia et al., 2003). We used both continuous raw scores and a dichotomization of EL as either present or not present.
A measure of general intellectual ability was derived from a short version of the Stanford–Binet Intelligence scales, the “Vocabulary” and “Object matrices” subtests, combined into an abbreviated IQ (ABIQ; Roid, 2003).
Statistics
We used the SPSS (Version 21) for statistical analyses. We collapsed into one group the children who showed EL at home, in preschool, or in both settings; for group comparisons, we used Pearson’s chi-square statistics and independent-sample t tests. We investigated the relationship between parent- and teacher-reported EL raw scores and the psychiatric symptoms (ADHD, anxiety, and ODD) through correlation analyses. Finally, we applied linear regression analyses to address how much of the variance in EL reported at home or in preschool could be explained by psychiatric symptoms. The parent- and teacher-reported EL raw scores were entered as dependent variables in two different models, and the parent-reported psychiatric symptoms were entered as independent variables in both models. Statistical significance was defined at the p < .05 level, all statistical tests were two-tailed, and preliminary analyses were done to ensure no violation of the assumptions of normality and linearity.
Results
A total of 25% of the children with ADHD (n = 90/368) were reported to show EL at home (n = 46), in preschool (n = 34), or in both settings (n = 10). This was significantly more frequent than in control children, where 7% (n = 9/127, 6 in preschool and 3 in both settings) showed EL, χ2(1, N = 495) = 17.8, p < .0001. In the ADHD group, no significant differences were found between the children with or without EL regarding age, gender, or IQ (Table 1).
Background Characteristics in the ADHD Symptom Group (n = 368) With and Without Deficient EL.
Note. EL = emotional lability.
Modest agreement was found between parent and teacher reports of EL (r = .21, p < .0001). Significant correlations were found between parent-reported EL scores and symptoms of ADHD, anxiety, and ODD, whereas the teacher-reported EL scores only correlated significantly with the ADHD subcomponents hyperactivity and impulsivity (Table 2).
Correlations Between Parent and Preschool Teacher Ratings of EL and Parent-Reported Psychiatric Symptoms in the ADHD Symptom Group (n = 368).
p < .05. **p < .001.
The children with EL were reported to have significantly higher mean symptom scores of inattention, hyperactivity-impulsivity, and ODD compared with those without EL (Table 3). Furthermore, among children with EL, 29% (n = 26/90) fulfilled criteria for the DSM-IV diagnosis of ODD, significantly more frequent than the 11% found in children without EL, n = 31/278, χ2(1, N = 368) = 16.3, p < .0001.
Symptom Comparisons in the ADHD Symptom Group With and Without EL.
Note. EL was considered present if reported by parent-, teacher-, or both raters. EL = emotional lability.
In the model with parent-reported EL as dependent variable, linear regression analyses showed that all entered variables (symptoms of inattention, hyperactivity-impulsivity, anxiety, and ODD) explained 15% of the variance, R2 = .15, F(368) = 15.6, p < .0001, and only symptoms of ODD (R2 = .10, p < .0001) made a significant contribution. In the model with the teacher-reported EL, all entered variables explained 3% of the variance, R2 = .032, F(344) = 2.8, p = .03, and only the ADHD subcomponents hyperactive-impulsive symptoms contributed significantly to the model (R2 = .03, p = .002).
Discussion
In the present study of EL and ADHD in preschoolers, EL was reported in 25% of the children in the ADHD group, which was significantly more frequent than in the control group. When reported by parents, EL was associated with psychiatric symptoms of ADHD, anxiety, and ODD, in contrast to teacher report where the only significant association was found with hyperactivity and impulsivity (ADHD subcomponents). Although EL was only partly explained by psychiatric symptoms, there was an association, and it appeared more closely related to ODD symptoms than to ADHD and anxiety.
Concerning the first aim of our study, we found that approximately one in four of the preschoolers with ADHD symptoms was reported to have EL (at home, in preschool, or in both settings). This is a somewhat lower percentage than those described with EL in schoolchildren with ADHD using a mixed community- and clinic-based sample (Anastopoulos et al., 2011) or a clinical sample (Sobanski et al., 2010). Differences in sample selections, variation in participants’ age, as well as different measures of EL across studies are all factors that possibly contribute to the percentage differences. However, our findings extend previous findings in schoolchildren (Anastopoulos et al., 2011; Shaw et al., 2014; Sobanski et al., 2010), by showing that EL is present also in a subgroup of 3-year-old children with symptoms of ADHD.
Concerning our second aim, we found a modest degree of agreement between parent and preschool teacher reports of EL (r = .21) in accordance with what is reported in the BRIEF-P manual (r = .22; Gioia et al., 2003). Identical correlations (r = .23) were also found in a study on school-age children (Sobanski et al., 2010) between parent and teacher ratings of EL using the Conners EL-index (Conners, 2003) with very similar items; (a) temper outbursts, (b) explosive unpredictable behavior, (c) crying often and easily, and (d) demands must be met immediately—easily frustrated; the fourth item to teachers only. Children may vary in different behavioral and emotional expressions across settings, as pointed out in a recent review (Dirks, De Los Reyes, Briggs-Gowan, Cella, & Wakschlag, 2012). Thus, there is growing recognition that children’s behavior varies meaningfully across situations, and evidence indicates that these differences, in combination with informants’ unique perspectives, are at least partly responsible for interrater discrepancies in reports of symptomatology.
With regard to our third aim, parents reported that preschool children with ADHD and EL were more severely affected with overall psychiatric symptoms compared with those without EL. This finding is in line with studies on both older children (Anastopoulos et al., 2011) and adults (Barkley & Fischer, 2010). Furthermore, for parents, ODD symptoms had the strongest statistical association with EL. This was not unexpected, as anger- and temper outbursts are included in the definition of both ODD and EL. However, that 29% of the preschoolers with EL also fulfilled DSM-IV criteria for ODD was less frequent than the percentages reported in clinical studies (Dickstein et al., 2005; Sobanski et al., 2010). This may be due to the younger age and the recruitment from the general population in our study. A total of 71% of children with EL in our study did not fulfill criteria for an ODD diagnosis, underlining that EL is not the same as ODD, and is important to recognize in preschool children.
In contrast to the parent report of EL, we found that the teacher report only correlated with hyperactivity and impulsivity (ADHD subcomponents). This could, on one hand, suggest that the link between EL and psychiatric symptoms might be due to bias when only one informant is included (Abikoff, Courtney, Pelham, & Koplewicz, 1993; Caron & Rutter, 1991). On the other hand, it could suggest a link between EL and the hyperactivity-impulsivity subcomponent of ADHD across settings.
It is noteworthy that 11% of those without EL also fulfilled the ODD criteria, and ODD only explained 10% of the EL variance in the linear regression analysis with parent-reported EL as the dependent variable. ADHD symptoms did not contribute significantly, in contrast to the teacher-reported EL, where symptoms of hyperactivity and impulsivity explained a small, but significant fraction of the EL variance. That most of the variance of EL remained unexplained is somewhat surprising in light of the overlap between phenomena although in line with family studies showing the presence of EL without concurring ADHD and ODD symptoms in both probands and siblings (Sobanski et al., 2010; Surman et al., 2011). The unexplained variance, both in our and other studies could be due to a methodological challenge, as EL is reported in questionnaires, which might more easily tap into emotional mood swings and temper outbursts than the more rigid criteria used in a diagnostic interview. In conclusion, the present study suggests that EL is a feature important to recognize as it cannot simply be explained by other symptoms.
Strengths and Limitations
Strengths of the present study were the use of a diagnostic interview with parents and a standardized measure of EL from BRIEF-P in a large group of preschoolers from a population-based cohort. Information about EL from two different settings is unique as earlier studies have mostly relied on parent report alone, which might bias findings (Caron & Rutter, 1991). The study design, with recruitment of children with reported ADHD symptoms, allowed us to study EL and co-occurring psychiatric symptoms, without the selection biases of clinical studies that would probably be particularly large in a study of preschool children as they are seldom referred to specialist clinics (Furniss, Beyer, & Guggenmos, 2006).
Our study has several limitations. The MoBa had a participation rate of 44% and an underrepresentation of risk groups like young mothers, mothers living alone, and children with previous birth complications (Nilsen et al., 2009). Furthermore, the ADHD study had a participation rate of 37.5% (Overgaard et al., 2014). One would expect that a low representation of families at risk might give an underestimation of EL and ADHD symptoms in our sample. Selection bias affects population representativeness of symptom distribution but should not affect associations between symptom clusters and associated features (Nilsen et al., 2009). Both interview data and questionnaire data were based largely on parent report, mostly from mothers, with the possibility of report bias as in other studies on co-occurrence (Abikoff, Courtney, Pelham, & Koplewicz, 1993; Caron & Rutter, 1991).
Our defined ADHD symptoms group consisted of children who fulfilled, or nearly fulfilled, the DSM-IV criteria for ADHD. We chose to report psychiatric symptoms that had lasted for 3 months or longer without excluding any particular items such as specific phobic symptoms. This poses a risk for including normative behavior and thus elevated symptom rates. There is reason to assume that this would affect the groups equally and thus not influence group comparisons. Strictly using DSM-IV criteria may, however, result in an underestimation of clinically significant symptoms that may represent later onset of disorders (Sonuga-Barke, Auerbach, Campbell, Daley, & Thompson, 2005). Furthermore, subthreshold ADHD was recently reported to have clinical relevance as it was associated with functional impairments across domains (Hong et al., 2014). Nevertheless, caution should be exercised regarding the generalizability of our findings.
The BRIEF-P EC scale has, to our knowledge, not been used as a measure for EL, which is a limitation when comparing our study with findings from studies that have used various measures (Barkley & Fischer, 2010; Leibenluft, 2011; Sobanski et al., 2010; Stringaris & Goodman, 2009). Although the BRIEF-P manual provides a definition for EL (T-scores ≥ 65 on the EC scale), the EC scale was initially constructed to measure the manifestation of executive functions within the emotional realm and measures a child’s ability to modulate emotional responses (Gioia et al., 2003). However, the scale does include the EL-features of angry outbursts, mood changes, and emotional reactivity (for items, see “Measures” section). At present, validation studies that compare EL as measured by BRIEF-P, with EL defined by other more widely used measures, are lacking. In the future, the new Disruptive Mood Dysregulation Disorder category of the DSM-V might prove useful as a more clearly defined category (APA, 2013).
We based our analyses on reports of psychiatric symptoms from parents only, which is in accordance with other studies on preschool psychopathology utilizing diagnostic entities (Sterba et al., 2007; Wichstrom et al., 2012). It is however likely the diagnostic symptom groups would be different if direct information from preschool teachers were part of the assessments.
Conclusion
This study confirms that EL is identifiable in 3.5-year-old children with ADHD symptoms and was found to occur in 25% of these preschoolers. Clinicians should look for these symptoms in preschoolers with ADHD as these children may be at particular risk of poor prognosis. Referral and systematic treatment should be considered in children who are more severely affected. It is important to recognize this phenomenon even in very young children with ADHD symptoms as EL could not simply be explained by the presence of psychiatric symptoms. Improving detection of EL and ADHD in young children should provide for the development of intervention programs, which might potentially alter the severe developmental trajectory associated with these features. Longitudinal follow-up studies from preschool age onwards are warranted.
Footnotes
Acknowledgements
We are grateful to all the participating children and their families in Norway who took part in these studies.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The present study was supported by a grant from the South Eastern Health Region, Grant No. 2009039. The study, “Preschool ADHD: Early characteristics, developmental trajectories, risk and protective factors in a prospective birth cohort (The ADHD study),” from which the present data were drawn was supported by funds and grants from the Norwegian Ministry of Health, The Norwegian Health Directorate, The South Eastern Health Region, G & PJ Sorensen Fund for Scientific Research, and from The Norwegian Resource Centre for ADHD, Tourettes syndrome, and Narcolepsy. The Norwegian Mother and Child Cohort Study is supported by the Norwegian Ministry of Health and the Ministry of Education and Research, NIH/NIEHS (Contract No. NO-ES-75558), NIH/NINDS (Grant No. 1 UO1 NS 047537-01), and the Norwegian Research Council/FUGE (Grant No. 151918/S10).