Scarica Ricerca ADHD ATTENTION DEFECIT HYPERACTIVITY DISORDER e più Guide, Progetti e Ricerche in PDF di Pedagogia solo su Docsity!
Journal of Affective Disorders 307 (2022) 133–
Available online 1 April 2022
0165-0327/© 2022 Elsevier B.V. All rights reserved.
Research paper
Affective-cognitive-behavioral heterogeneity of Attention-Deficit/
Hyperactivity Disorder (ADHD): Emotional dysregulation as a sentinel
symptom differentiating “ADHD-simplex” and
“ADHD-complex” syndromes?
Xinxin Yue
a,b, 1
, Lu Liu
a,b, 1
, Wai Chen
c,d,e,f,g,h,*
, David A. Preece
h,i,j
, Qianrong Liu
a,b
Haimei Li
a,b
, Yufeng Wang
a,b
, Qiujin Qian
a,b,**
a (^) Peking University Sixth Hospital/Institute of Mental Health, Beijing 100191, China b (^) National Clinical Research Center for Mental Disorders (Peking University Sixth Hospital) & NHC Key Laboratory of Mental Health (Peking University), Beijing
100191, China c (^) Mental Health Service, Fiona Stanley Hospital, Perth, Australia d (^) Graduate School of Education, University of Western Australia, Australia e (^) School of Medicine, University of Notre Dame Australia, Fremantle, Australia f (^) School of Psychology, Murdoch University, Perth, Australia g (^) Curtin Medical School, Curtin University, Perth, Australia h (^) The enAble Institute, Curtin University, Perth, Australia i (^) School of Psychology, Curtin University, Perth, Australia j (^) School of Psychological Science, University of Western Australia, Perth, Australia
A R T I C L E I N F O
Keywords: ADHD Nosology Emotional symptoms Latent class analysis Subphenotypes
A B S T R A C T
Background: Current DSM and ICD classifications of Attention-Deficit/Hyperactivity Disorder (ADHD) exclude
emotional dysregulation (ED) in their diagnostic criteria, despite ED symptoms frequently co-occurring in ADHD
and likely sharing common neurobiological substrates. In this study, we examined whether consideration of ED
symptoms could delineate more informative “ADHD+ED” subphenotypes.
Method: 4106 children with ADHD were recruited. ED and inattentive (IA) and hyperactive/impulsive (HI)
symptoms were profiled using latent class analyses (LCA). The derived latent class (LC) subphenotypes were
evaluated and validated in relation to comorbidity patterns, executive functions, and functional impairments.
Results: Five LC subphenotypes with ED symptoms were identified: IA/HI + ED profile (LC1); HI + ED profile
(LC2); IA + ED profile (LC3); IA/HI profile (LC4); and IA profile (LC5). Cross-validation of the LCA model using
support vector machine analysis confirmed 83% accuracy. ED positive (ED+ve) subphenotypes were associated
with higher rates of oppositional defiant disorder, mood disorders, anxiety disorders, as well as more severe
autistic traits and sluggish cognitive tempo symptoms. Higher rates of ecological executive functioning im-
pairments (BRIEF ratings) were found among ED+ve subphenotypes (though no differences were detected by
laboratory-based measures). Functional impairments were also more severe among participants with ED+ve
subphenotypes.
Limitations: The data for our LCA were cross-sectional and based primarily on parent ratings.
Conclusion: Our classification model has parcellated IA, HI, and ED symptoms into novel informative sub-
phenotypes. These classifications provide preliminary evidence that ED symptoms could serve as sentinel fea-
tures to identify a potential “ADHD-complex” syndrome, which demarcates a more pervasive condition of greater
severity, complexity, and impairment.
* Correspondence to: W. Chen, Mental Health Service, Fiona Stanley Hospital, Perth, Australia.
** Correspondence to: Q. Qian, Peking University Sixth Hospital/Institute of Mental Health, Beijing 100191, China.
1
These authors contributed to this work equally.
Contents lists available at ScienceDirect
Journal of Affective Disorders
journal homepage: www.elsevier.com/locate/jad
https://doi.org/10.1016/j.jad.2022.03.
Received 29 January 2022; Received in revised form 23 March 2022; Accepted 28 March 2022
1. Introduction
In DSM and ICD taxonomies, Attention-Deficit/Hyperactivity Disor-
der (ADHD) is a neurodevelopmental disorder defined by the clinical
triad of inattention , hyperactivity , and impulsivity , associated with
educational, functional, and social impairments. In the DSM-5 (Amer-
ican Psychiatric Association, 2013), the core symptoms of ADHD
comprise of inattention (IA), hyperactivity (HY) and impulsivity (IM),
with the HY and IM symptoms conceptualized as a single dimension
(hyperactivity/impulsivity; HY/IM). The ICD-10 system posits a similar
definition, though HY and IM symptoms were conceived as separate
dimensions and the disorder was labelled as “hyperkinetic disorder”
rather than ADHD. In the ICD-11 revision, the term “ADHD” has
replaced the “hyperkinetic disorders” category from the ICD-10; ADHD
is now classified under “neurodevelopmental disorders”, and its het-
erogeneous clinical expressions are now represented by taxonomic
qualifiers, which include “predominantly inattentive”, “predominantly
hyperactive-impulsive”, or “combined” types across the lifespan (Reed
et al., 2019). Nonetheless, these narrow definitions of ADHD by
authoritative taxonomic systems have recently been questioned over
their accuracy (Stanton et al., 2018); in particular, the exclusion of
emotional dysregulation (ED) from the core symptom architecture has
drawn criticism (Shaw et al., 2014).
Emotions manifest along experiential (e.g., feeling of anger), physi-
ological (e.g., increased heart rate), and behavioral (e.g., urge to fight)
channels of the emotion system, occurring in response to stimuli
appraised as meaningful for one's goals (Gross, 2015). ED refers to dif-
ficulties in inhibiting or regulating the occurrence, intensity, and dura-
tion of such emotional states, the result of which is extreme emotional
lability (Barkley and Fischer, 2010). Emotion regulation skills, more
broadly, have been demonstrated to be an important determinant of
health and wellbeing outcomes (e.g., Gross and John, 2003), and thus
levels of ED are of substantial clinical interest. In recent years, ED has
been proposed to represent a core defining feature of ADHD (Barkley
and Fischer, 2010). Interestingly, such a proposal is not entirely new.
Prior to DSM-III, there was a long historical tradition (between 1902 and
- of depicting ADHD as closely linked with ED. For instance, in
1902, Still described the precursor of ADHD with excessive emotionality
as a prominent feature (Still, 1902). In 1937, Bradley described “mood
lability” as one of the six cardinal symptoms (Lange et al., 2010). In
1948, Rosenfeld and Bradley firstly defined “unpredictable variability in
mood” as diagnostic symptom. In 1981, the “Utah Criteria” were pub-
lished to define the “Adult ADHD” phenotype (Wender et al., 1981). In
this definition, ED symptoms occupy 3 out of the 7 cardinal features
(Reimherr et al., 2005).
Overall, ED symptoms clearly assume a central position in historical
definitions of the ADHD syndrome, and this historical account is rele-
vant to our present consideration of ED. It is possible that ED forms an
integral part of ADHD psychopathology, though not all ADHD cases may
express ED. While ED symptoms are neither necessary nor sufficient to
define ADHD diagnostically, their centrality in ADHD psychopathology
nevertheless poses an unresolved conceptual conundrum. Here, we
propose a hypothesis which could potentially resolve this conundrum.
We hypothesize that two broad distinctive categories of ADHD may
exist: what we will call (i) ADHD-complex (i.e., ADHD+ED) and (ii)
ADHD-simplex (i.e., ADHD without ED). From this perspective, we
expect that ED symptoms may demarcate a more pervasive, severe, and
complex variant of ADHD, thus acting as sentinel features rather than defining features. This representation may explain why ED can be criti- cally important in syndromal consideration of ADHD, despite being neither necessary nor sufficient in defining ADHD. As such, the explo- ration, validation, and exposition of this hypothesis form the central theme of this study.^2 Several avenues of empirical work suggest links between ADHD and ED. Several studies have demonstrated that ADHD and ED are strongly associated (Shaw et al., 2014), as measured by constructs such as emotional lability (EL), emotional impulsiveness, affective lability (AL), or deficient emotional self-regulation (Sobanski et al., 2010). A review by Shaw et al. (2014), for instance, suggested 34%–70% of adults with ADHD experience ED problems, a finding also supported by two recent meta-analytical studies (Beheshti et al., 2020; Graziano and Garcia, 2016). The study by Beheshti et al. (2020) concluded that adults with ADHD have significantly higher levels of ED compared to controls. Graziano and Garcia (2016), similarly, conducted a meta-analysis on 77 studies on children and adolescents ( n = 32,044) and evaluated several aspects of emotionality (emotional negativity, ED, callous-unemotional traits, and emotional understanding) finding strong associations be- tween ADHD and emotional reactivity/negativity/lability. In contrast, there were weaker associations with callous-unemotional traits and with emotional understanding, thus demonstrating a degree of specificity in the correlation patterns. For the factorial architecture, there is also recent factor analytic support for poor emotion regulation skills cohering statistically with other components of ADHD, and that ADHD factor models including ED can demonstrate strong fit index values (Hirsch et al., 2018). Interestingly, a family risk analysis has found that siblings of pro- bands with ADHD+ED have elevated risk of expressing ED, but not ADHD per se. ADHD and ED cosegregated in siblings, showing a pattern of inheritance which suggests that an “ADHD+ED” phenotype may present a distinctive familial subtype of ADHD (Surman et al., 2013). Partitioning subtypes can be based on familial risk and also by syndro- mal clustering. Recently, the concept of a “subphenotype” has been used in clinical medicine in parcellating phenotypic heterogeniety. For example, in acute respiratory distress syndrome (ARDS), two distinct biological subphenotypes have been identified: a (i) hypoinflammatory and (ii) hyperinflammatory subphenotype. The hyperinflammatory subphenotyope was found to be a more serious condition, associated with a shock state, metabolic acidosis, and worse clinical prognosis (Wildi et al., 2021). In the same vein, we postulate that “ADHD+ED” may represent a more severe subphenotype, associated with greater biological risk loading, with ED symptoms signifying more complex expression of ADHD, i.e. “ADHD-complex”, analogous to the “hyper- inflammatory” variant of ARDS. Yet to date, it has been under-explored whether “ADHD+ED” indeed represents a more severe and complex ADHD subphenotype, signifying greater loading of aberrant neurobiological substrates, which embody more comorbidities and impairments (such as Sluggish Cognitive Tempo (SCT) symptoms, Autism Spectrum Disorder (ASD) symptoms, neuro- cognitive deficits and other psychopathologies). This study aimed to fill this key gap in the literature, by using latent class analysis to empirically test whether an “ADHD+ED” subphenotype (i.e., latent subclass) emerges statistically in children/adolescent data, and whether such a subphenotype is indeed associated with greater comorbidities and im- pairments. More specifically, we hypothesized that (i) not all ADHD
2 In proposing the term ADHD-complex to denote ADHD subphenotypes with
ED throughout this manuscript, it is not our intention to imply that other
variables (other than ED) cannot also present alongside ADHD and make the
presentation “complex”. However, given the historical status of ED as a po-
tential core part of the ADHD presentation (Barkley and Fischer, 2010), we
think there is strong justification for proposing the presence of ED as a key
marker for a more complex ADHD presentation.
effects, and also by Latent Profile Analysis (LPA) for categorical
threshold effects. The six domains of the WFIRS-P were input as vari-
ables for a Latent Profile Analysis (LPA), using Mplus Version 7 (with
500 set as the starting value to verify each class solution) (Muth´en and
Muth´en, 1998). The process of determining the optimal number of
classes was similar to the classification of the LCAs. Finally, three latent
profile classes were identified based on WFIRS-P scores , which were used
for subsequent multinomial logistic regression (MLR) analyses.
2.3.4. Analysis of covariance (ANCOVA) and multinomial logistic
regression (MLR)
Comparisons of demographic and clinical characteristics were con-
ducted using either Pearson's chi-square test or variance analysis as
appropriate. To compare the behavioral difficulties (ATs and SCT), EFs,
and functional impairment levels among the different ADHD classes
extracted by the LCA, we first performed hierarchical multiple regres-
sion to explore whether ED symptoms could explain extra variance in-
dependent from ADHD core symptoms and comorbidity status. When
independent effects were indicated, further analyses of covariance
(ANCOVA) with age, IQ, gender, comorbidity status, and ADHD severity
(total symptoms) as covariates were performed.
To further explore the threshold effects of categorical classification
of functional impairments (i.e., as extreme, moderate, and impaired
groups), multinomial logistic regression (MLR) was applied to evaluate
the effects of ED subphenotypes upon functions grouped as 3 latent
profile classes. All analyses were done with SPSS, except for MLR with
STATA software.
For the MLR, LC5 was assigned as the base group, and therefore used
as the ‘reference group’ in predicting the membership of the LPA 3-class
latent profiles of functional impairment (‘WFIRS Mild’ group was used
as the reference group in relation to ‘WFIRS Moderate’ and ‘WFIRS Se-
vere’ groups). In other words, LC5 is represented as zero and LC1 as 1 in
the LC1 vs LC5 comparison. Then we repeated the same process with the
LC2 vs LC5, LC3 vs LC5, and LC4 vs LC5 comparisons, in predicting
membership of WFIRS Moderate (represented as 1) and Severe group
(represented as 2, while compared with WFIRS Mild group represented
as zero). As a correction for multiple comparisons, Bonferroni correc- tions were applied. Sixteen variables were analyzed (ATs, SCT, 5 laboratory-based EFs variables, 3 BRIEF subfactors, and 6 WFRIS sub- factors), the alpha value for significance was adjusted to p < 0. (0.05/16). For the group comparisons surviving Bonferroni corrections, post hoc analyses of variance (ANOVA) were further performed to compare the estimated marginal means of each pair of the classes.
3. Results
3.1. Determining the optimum number of latent classes (LC)
Based on the parameters from fit statistics, the 5-LCs solution model (Model 5) was identified as the best fitting model by our LCA (Appendix Table S2). The latent class probability for all models (2-LCs, 3-LCs, 4- LCs, 5-LCs, and 6-LCs) can be found in Appendix Table S3. For the 5-LCs model, the average posterior class probability and the odds of correct classification further indicated high classification accuracy of the 5-LCs model (Appendix Tables S4 & S5). The model class assignment pro- portion using the estimated posterior class probabilities was reasonably distributed (Appendix Table S6). SVM validation of Model 5 confirmed the best classification accuracy of 83% (Appendix Table S7). Repetition of 5-fold cross validation 10 times showed robust results, indicating good stability of the model (Appendix Table S8). The distribution of symptoms across these 5 classes is shown visually in Fig. 1. 49% of participants expressed ED symptoms, with these participants principally classified within 3 of the classes: LC1, LC2, or LC3. The 5 LC profiles included:
- LC1 ( n = 824, 20.52%): with inattention, hyperactivity/impulsivity, and emotional dysregulation, designated “IA/HI + ED”;
- LC2 ( n = 414, 10.31%): with hyperactivity/impulsivity and moder- ate emotional dysregulation, designated “HI + ED”;
- LC3 ( n = 731, 18.20%): with inattention and emotional dysregula- tion, designated “IA + ED”;
Fig. 1. Conditional response probabilities of ADHD and ED symptoms in all five latent classes. LC1: with inattention, hyperactivity/impulsivity, and emotional
dysregulation (IA/HI + ED); LC2: with hyperactivity/impulsivity and moderate emotional dysregulation (HI + ED); LC3: with inattention and emotional dysre-
gulation, IA + ED; LC4: with inattention and hyperactivity/impulsivity, IA/HI; LC5: with inattention, IA. Items in black indicated inattentive symptoms, items in
black green indicated hyperactive/impulsive symptoms, and items in brown indicated emotional dysregulation symptoms. (For interpretation of the references to
color in this figure legend, the reader is referred to the web version of this article.)
- LC4 ( n = 1007, 25.07%): with inattention and hyperactivity/ impulsivity, designated “IA/HI”;
- LC5 ( n = 1040, 25.90%): mainly with inattention, designated “IA”.
3.2. Clinical characteristics of the 5 latent classes (LCs)
Table 1 shows the demographic and clinical characteristics of the 5
LCs. LC2 showed male preponderance. However, gender distributions
between LC1 and LC4 (males, 85.6% versus 84.9%, p = 0.696), and
between LC3 and LC5 (males, 81.9% versus 80.4%, p = 0.410) were
comparable. The age distribution showed a significant group difference
( p < 0.001): LCs with prominent inattentive symptoms (LC3 and LC5)
were older, while the LCs with prominent hyperactivity/impulsivity
symptoms (LC2) were younger. No group difference in IQ distribution
was found ( p = 0.570). Table 1 also cross-tabulates LCs and the DSM-
based subtypes.
3.3. Comorbidity patterns associated with the 5 LCs
The comorbidity patterns of the 5 LCs are shown in Table 2. Those
LCs with ED showed higher rates of comorbidity than those without ED,
including ODD, Anxiety Disorders, and Mood Disorders.
Hierarchical multiple regression indicated that the presence of ED
symptoms could explain extra variance (beyond ADHD core symptoms
and comorbidity status) in ATs and SCT (Supplementary Tables S9–10).
Further group comparisons indicated significant differences among the
five LCs (Table 2). More specifically, between LC1 and LC4, and between
LC3 and LC5, thus indicating higher levels of ATs in the ED sub-
phenotypes. For SCT, a significant difference was only indicated be-
tween LC1 and LC4. In addition, LCs with more prominent inattentive
symptoms (LC3 and LC5) had the higher SCT scores.
3.4. Ecological and laboratory-based EF correlates with ED
subphenotypes
Hierarchical multiple regression indicated that ED symptoms could
explain extra variance in EF deficits, as rated by the BRIEF, for Inhibi-
tion, Shifting, and Working Memory (Supplementary Tables S11–13).
The ED positive (ED+ve) subphenotypes were associated with more
Inhibition and Shifting deficits (Table 2) when comparing LC1 and LC4,
and comparing LC3 and LC5. For Working Memory, there was no sig-
nificant difference between the LCs with and without ED (LC1 = LC4,
LC3 = LC5). However, such differences were not found for the laboratory-based EFs tasks (Appendix Table S14).
3.5. Functional impairment correlations with the 5 LCs
As aforementioned, the Chinese version of the WFIRS was translated in 2009 (Qian et al., 2011), and was therefore only available for par- ticipants recruited between 2009 and 2017, yielding approximately 43.9% of the sample (i.e., 1766/4016 participants). There was however no significant difference in the demographic or clinical characteristics between the pre-2009 and post-2009 subsamples, except for age which was included as a covariate in all subsequent analyses (details provided in Supplementary materials Table S15). No differences in the 5 LC rates were detected between these two subsamples ( p = 0.100). Hierarchical multiple regression indicated that ED symptoms could explain independent variance in functional impairments, especially for the “children's self-concept” subdimension (Supplementary Tables S16–21). The differences between the 5 LCs on these 6 functional impairments are shown in Table 3. The results from the ANCOVA (with age, IQ, gender, comorbidity status, and ADHD severity as covariates) indicated that the ED+ve subphenotypes were associated with more functional impairments when comparing LC1 with LC4, and comparing LC3 with LC5, except for in the “School and learning” and “Life skills” domains. All results survived Bonferroni corrections. We further parcellated WFIRS scores into profile groups using LPA. Functional impairment based WFIRS scores was classified into three categories: mild, moderate, and severe LCs (Appendix Tables S22– 23 and Fig. S2). We also examined the average posterior class probability, the odds of correct classification, and the model class assignment pro- portion using the estimated posterior class probabilities. The results indicated that the three categories had high accuracy (Appendix Tables S24–26). Table 4 shows the effect sizes as Relative Risk Ratios, in terms of the ED+ve LCs as predicted by the WFIRS-P LCs. There is 4. fold greater risk of having severe functional impairment if one is assigned to the IA/HI + ED LC membership category (compared with LC5/IA as the reference group). In contrast, the risk is nonsignificant for IA/HI without ED, in terms of having severe functional impairment.
4. Discussion
This study set out to address the conceptual conundrum of whether ED is a core symptom of ADHD. We hypothesized that ED symptoms
Table 1
Demographic and clinical characteristics of the participants.
LC1 LC2 LC3 LC4 LC5 p Post hoc analyses
IA/HI + ED HI + ED IA + ED IA/HI IA
N (%) 824 (20.52) 414 (10.31) 731 (18.20) 1007 (25.07) 1040 (25.90) Male (%) 85.6 88.4 81.9 84.9 80.4 0.001 LC1 = LC2 = LC4 > LC5, LC2 > LC3 = LC FSIQ (mean ± SD) 102.47 ±
Age in months (mean ± SD) 114.19 ±
< 0.001 LC3 = LC5 > LC1 > LC4 = LC
Symptoms (mean ± SD)a Inattention 8.25 ± 0.84 5.31 ± 1.32 7.42 ± 1.11 7.65 ± 1.09 7.22 ± 1.14 < 0.001 LC1 > LC4 > LC3 > LC5 > LC Hyperactivity/ impulsivity
7.45 ± 1.17 6.97 ± 1.17 3.35 ± 1.28 6.37 ± 1.35 1.93 ± 1.17 < 0.001 LC1 > LC2 > LC4 > LC3 > LC
Emotional dysregulation 7.50 ± 1.80 5.44 ± 1.83 6.88 ± 1.70 2.98 ± 1.48 2.90 ± 1.59 < 0.001 LC1 > LC3 > LC2 > LC4 = LC DSM-IV subtypes (%) Inattentive 1.1 1.5 34.4 12.9 50. Hyperactive/impulsive 0.5 85.8 1.1 10.4 2. Combined 45.0 12.5 1.3 40.7 0.
Note. FSIQ, full-scaled intelligence quotient; SD, standard deviation; DSM-IV, Diagnostic and Statistical Manual of Mental Disorders 4th edition; LC, latent class; IA,
inattention; HI, hyperactivity/impulsivity; ED, emotional dysregulation.
a Inattention and hyperactivity/impulsivity symptoms and ADHD diagnostic subtypes were evaluated using the Clinical Diagnostic Interview Scale (CDIS);
emotional dysregulation symptoms were measured using the ‘emotional lability’ subscale items of the Conners' Parent Rating Scale.
of increased ED severity (Leaberry et al., 2020); patterns which are
consistent with our findings. ED has been found to represent a separate
construct from ODD and its “mood/affect” subdimensions (Liu et al.,
2019), despite potential symptom overlaps based on face validity.
Future studies with longitudinal designs may be able to unravel the
causal mechanisms.
For internalizing disorders, ED+ve subphenotypes in our data
showed higher rates of anxiety disorders and mood disorders. This is
consistent with the findings from a prospective cohort study suggesting
that ED profiles in children predicted later depression or anxiety in
adolescence (Wang et al., 2018). For adults, ED also significantly
mediated the comorbidity of ADHD and internalizing problems (Murray
et al., 2021). The high co-occurrence may arise from the shared
neurobiological mechanisms between ADHD and ED, in particular
aberrant networks identified within the prefrontal cortex (PFC) circuitry
and amygdala (Hulvershorn et al., 2014). A longitudinal follow-up study
incorporating serial neuroimaging, while tracking symptom develop-
ment (based on our novel nosology model), could represent an impor-
tant avenue to better understand symptom evolution over time; this
could help to elucidate potential changes in neurobiological mecha-
nisms and the developmental determinants linking ADHD and ED.
We also explored the potential association between the ED+ve sub-
phenotypes and ATs and SCT. We found higher levels of AT symptoms in
participants with ED+ve subphenotypes. Recently, impaired emotional
functioning, as assessed using the CBCL, was found to be associated with
ATs over the long term, and such findings (Joshi et al., 2020) are in line
with our results. For SCT, the inattention LCs showed the higher SCT
scores, consistent with previous reports that SCT is associated with
inattention (Becker et al., 2016a). Nevertheless, our study also showed
that the ED+ve LCs were associated with higher SCT scores, and par-
ticipants of LC1 (IA/HI + ED) had higher scores than LC4 (IA/HI). The
SCT phenotype was indeed associated with ED (Flannery et al., 2016)
and high rates of internalizing disorders as well as suicide attempts
(Becker et al., 2016b).
Past evidence that a poorer working memory capacity could predict
higher levels of ED (Jensen et al., 2018) supports our present findings of
greater EF impairments. This close correlation between ED and 'inhibi-
tion' is not surprising, as ED is putatively secondary to an inhibition
deficit in Barkley's (1997) model. It is worth noting, though, that in our
data only ecological assessment showed those significant correlations
(not laboratory-based tests), which is consistent with previous reports
(Gisbert et al., 2019). One potential reason for this inconsistency across
measurement approaches might be that the lab-based evaluation is
conducted in a highly structured setting, with explicit guidance on
performance; such properties provide scaffolding for performance and
may conceal the true EF deficits of some children with ADHD.
Finally, we examined the effects on functional impairment. The re-
sults showed that ED increased impairments across multiple domains.
ED symptoms have previously been reported to increase functional im-
pairments, exerting unique effects, independent of those from ADHD
core symptoms (Barkley and Fischer, 2010). Moreover, ED persists in the
life course of ADHD, and can become the main driver of functional
impairment (Shaw et al., 2014). Meanwhile, greater variability in affect
among those with ADHD has been associated with worse social func-
tioning (Breaux et al., 2020). According to the results of our LPA based
on the WEIRS, we found that the ED+ve LCs had a higher relative risk
ratio compared with the IA subphenotype. In short, there was roughly a
3 – 4 fold increased risk of severe functional impairments (adjusted for
IQ, age, gender, comorbidity status, and ADHD severity) for those
expressing ED in pairwise comparisons of analogous ADHD LCs (i.e. LC
(Relative Risk Ratio [RRR] = 4.56) vs LC4 (RRR = 0.94, nonsignificant);
and LC3 (RRR = 3.01) vs LC5 (RRR = 1 as reference group) as shown in
Table 4).
Our results highlight that ED should be routinely assessed in ADHD
clinical assessments, and that ED could be a sentinel sign to demarcate a
more severe clinical phenotype associated with greater levels of
comorbidities and impairments with clinical utility. Given the cross- sectional nature of our sample, it is not possible to determine whether ED is a ‘gateway symptom cluster’ driving other comorbidities and/or impairments, or ED is an epiphenomenon of more severe latent pheno- types. In the former case, direct treatment of ED may reduce morbidities and impairments, if future studies can demonstrate that ED plays a causal role in downstream complications. In the latter case, more research is needed to understand the underlying driver of psychopa- thologies beneath the pleiotropic expressions before extrapolation on treatment implications can be made. If replicated, our findings support a more prominent inclusion of ED as a pattern qualifier (rather than a core symptom) in ADHD within future revisions of taxonomic systems (i.e., DSM and ICD) (Bach and First, 2018).
4.1. Limitations and future directions
Several limitations of our study need to be considered. First, the validation analyses in our present study were not conducted in an in- dependent sample. Second, our sample was cross-sectional, thus the temporal stability of our classification model could not be examined. Future studies with longitudinal designs can evaluate the temporal stability of detected LCs and comorbidity patterns. Third, the ED mea- sure was based only on rating scale evaluation, which lacks objective or physiological measurement. Problematic influences from recall bias and information bias from the informants can therefore not be ruled out. Similar to our approach, many published LCA and LPA studies generate and test latent classes based on a set of variables all collected using the same assessment modality (e.g., Girard, 2021; Preece et al., 2021; Sul- livan et al., 1998), thus we consider that our derived solutions still make a strong contribution. Overall, these issues can be addressed by future research which could incorporate a multi-method approach with parent/teacher report, as well as laboratory measurements, such as real- time ascertainment of emotions, peripheral physiological characteristics (cardiac measures of respiratory sinus arrhythmia), and central nervous system functioning via portable devices (e.g., Skirrow and Asherson, 2013). Fourth, future studies can explore whether the complex syn- drome is driven by a higher level of disinhibition, which is expressed as cognitive, behavioral, and emotional impulsivity as well as a higher level of comorbidities. Finally, the treatment response and clinical outcome of different subphenotypes could be explored in order to establish the clinical utility and predictive validity of the proposed classification.
5. Conclusions
About 49% of our ADHD sample expressed ED. For those with ED, there were elevated risks of having more severe symptoms of ADHD, as well as expressing symptoms of ODD, anxiety disorders, mood disorders, ATs, and SCT. These participants had more functional impairment and more EFs deficits (as reported by carers, though not detected when measured by laboratory testing). Our findings therefore provide pre- liminary evidence to support the existence of an “ADHD-complex” syndrome, characterised by ED sentinel symptoms and clustering of associated comorbidities, EF deficits, and functional impairments. If our findings are extensively replicated in the future, a more prominent in- clusion of ED as a pattern qualifier in ADHD diagnostic systems may be warranted.
Notes
We have listed all abbreviations and the corresponding full names in the Supplementary materials (Table S27).
Funding
This work was supported by the National Science Foundation of
China (81873802; 81641163; 81571340), Beijing Natural Science
Foundation (7172245), the National Key Basic Research Program of
China (973 program 2014CB846104).
CRediT authorship contribution statement
Xinxin Yue, Lu Liu and Qiujin Qian contributed conception and
design of the study; Haimei Li and Xinxin Yue organized the database;
Xinxin Yue, Lu Liu, Wai Chen designed the statistical analysis; Xinxin
Yue and Qianrong Liu performed the statistical analysis. Xinxin Yue, Lu
Liu, Wai Chen, David Preece, Qianrong Liu, Yufeng Wang and Qiujin
Qian interpreted the results and wrote sections of the manuscript. All
authors contributed to manuscript revision, and read and approved the
submitted version.
Conflict of interest
The authors declare that the research was conducted in the absence
of any commercial or financial relationships that could be construed as a
potential conflict of interest.
Acknowledgments
We would like to thank all subjects who participated in this study.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.jad.2022.03.065.
References
Achenbach, T.M., 1991. Manual for the Child Behavior Checklist And Youth Self-report. University of Vermont Department of Psychiatry, Burlington, VT. American Psychiatric Association, 1994. Diagnostic And Statistical Manual of Mental Disorders, 4th edition. Author, Washington, DC. American Psychiatric Association, 2013. Diagnostic And Statistical Manual of Mental Disorders, 5th ed. Author, Arlington, VA. Bach, B., First, M.B., 2018. Application of the ICD-11 classification of personality disorders. BMC Psychiatry 18 (1), 351 https://bmcpsychiatry.biomedcentral.com/ articles/10.1186/s12888-018-1908-3. https://bmcpsychiatry.biomedcentral.co m/articles/10.1186/s12888-018-1908-3. Barkley, R.A., 1997. Behavioral inhibition, sustained attention, and executive functions: constructing a unifying theory of ADHD. Psychol. Bull. 121 (1), 65–94 https:// psycnet.apa.org/record/1997-02112-004. https://psycnet.apa.org/record/1997- 02112-004. Barkley, R.A., Fischer, M., 2010. The unique contribution of emotional impulsiveness to impairment in major life activities in hyperactive children as adults. J. Am. Acad. Child Adolesc. Psychiatry 49 (5), 503–513 https://www.sciencedirect.com/science/ article/pii/S0890856710002030. https://www.sciencedirect.com/science/article/ pii/S0890856710002030. Becker, S.P., Leopold, D.R., Burns, G.L., Jarrett, M.A., Langberg, J.M., Marshall, S.A., Willcutt, E.G., 2016. The internal, external, and diagnostic validity of sluggish cognitive tempo: a meta-analysis and critical review. J. Am. Acad. Child Adolesc. Psychiatry 55 (3), 163–178 https://www.sciencedirect.com/science/article/pii/ S0890856715008229?via%3Dihub. https://www.sciencedirect.com/science/artic le/pii/S0890856715008229?via%3Dihub. Becker, S.P., Withrow, A.R., Stoppelbein, L., Luebbe, A.M., Fite, P.J., Greening, L., 2016. Sluggish cognitive tempo is associated with suicide risk in psychiatrically hospitalized children. J.Child Psychol.Psychiatry 57 (12), 1390–1399 https:// acamh.onlinelibrary.wiley.com/doi/10.1111/jcpp.12580. https://acamh.onlineli brary.wiley.com/doi/10.1111/jcpp.12580. Beheshti, A., Chavanon, M.-L., Christiansen, H., 2020. Emotion dysregulation in adults with attention deficit hyperactivity disorder: a meta-analysis. BMC Psychiatry 20 (1), 120 https://bmcpsychiatry.biomedcentral.com/articles/10.1186/s12888-020-2442-
- https://bmcpsychiatry.biomedcentral.com/articles/10.1186/s12888-020-2442-7. Breaux, R., Langberg, J.M., Swanson, C.S., Eadeh, H.-M., Becker, S.P., 2020. Variability in positive and negative affect among adolescents with and without ADHD: differential associations with functional outcomes. J. Affect. Disord. 274, 500– 507 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8010584/. https://www.ncbi. nlm.nih.gov/pmc/articles/PMC8010584/. Conners, C.K., 1997. Conners' Rating Scales-revised (Technical Manual). Multi-Health Systems, Toronto, Ontario, Canada. Faraone, S.V., Rostain, A.L., Blader, J., Busch, B., Childress, A.C., Connor, D.F., Newcorn, J.H., 2019. Practitioner review: emotional dysregulation in attention- deficit/hyperactivity disorder - implications for clinical recognition and
intervention. J.Child Psychol.Psychiatry 60 (2), 133–150 https://acamh. onlinelibrary.wiley.com/doi/10.1111/jcpp.12899. https://acamh.onlinelibrary.wil ey.com/doi/10.1111/jcpp.12899. Flannery, A.J., Becker, S.P., Luebbe, A.M., 2016. Does emotion dysregulation mediate the association between sluggish cognitive tempo and college students' social Impairment? J. Atten. Disord. 20 (9), 802–812 https://journals.sagepub.com/doi/ 10.1177/1087054714527794?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref. org&rfr_dat=cr_pub%20%200pubmed. https://journals.sagepub.com/doi/10. 77/1087054714527794?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr _dat=cr_pub%20%200pubmed. Gioia, G.A., Isquith, P.K., Guy, S.C., Kenworthy, L., 2000. Behavior rating inventory of executive function. Child Neuropsychol. 6, 235–238 http://europepmc.org/article/ MED/11419452. http://europepmc.org/article/MED/11419452. Girard, L.C., 2021. Concomitant trajectories of internalising, externalising, and peer problems across childhood: a person-centered approach. Res.ChildAdolesc. Psychopathol. 49 (12), 1551–1565 https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC8557151/. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8557151/. Gisbert, L., Richarte, V., Corrales, M., Iba´˜nez, P., Bosch, R., Bellina, M., Ramos- Quiroga, J.A., 2019. The relationship between neuropsychological deficits and emotional lability in adults with ADHD. J. Atten. Disord. 23 (12), 1514– 1525 https://journals.sagepub.com/doi/10.1177/1087054718780323?url_ver=Z39.88- 2003 &rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed. https://jour nals.sagepub.com/doi/10.1177/1087054718780323?url_ver=Z39.88-2003&rfr _id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed. Graziano, P.A., Garcia, A., 2016. Attention-deficit hyperactivity disorder and children's emotion dysregulation: a meta-analysis. Clin. Psychol. Rev. 46, 106–123 https:// www.sciencedirect.com/science/article/pii/S0272735816301350?via%3Dihub. https://www.sciencedirect.com/science/article/pii/S0272735816301350?via% 3Dihub. Gross, J.J., 2015. Emotion regulation: current status and future prospects. Psychol. Inq. 26, 1–26 https://www.tandfonline.com/doi/full/10.1080/1047840X.2014.940781. https://www.tandfonline.com/doi/full/10.1080/1047840X.2014.940781. Gross, J.J., John, O.P., 2003. Individual differences in two emotion regulation processes: implications for affect, relationships, and well-being. J. Pers. Soc. Psychol. 85 (2), 348 – 362 https://psycnet.apa.org/record/2003-05897-016. https://psycnet.apa. org/record/2003-05897-016. Hirsch, O., Chavanon, M., Riechmann, E., Christiansen, H., 2018. Emotional dysregulation is a primary symptom in adult attention-deficit/hyperactivity disorder (ADHD). J. Affect. Disord. 232, 41–47 https://www.sciencedirect.com/science/ article/pii/S0165032717323662?via%3Dihub. https://www.sciencedirect.com/scie nce/article/pii/S0165032717323662?via%3Dihub. Hulvershorn, L.A., Mennes, M., Castellanos, F.X., Di Martino, A., Milham, M.P., Hummer, T.A., Roy, A.K., 2014. Abnormal amygdala functional connectivity associated with emotional lability in children with attention-deficit/hyperactivity disorder. J. Am. Acad. Child Adolesc. Psychiatry 53 (3), 351–361.e1 https://www. ncbi.nlm.nih.gov/pmc/articles/PMC3961844/. https://www.ncbi.nlm.nih.gov/pmc /articles/PMC3961844/. Jensen, D.A., Høvik, M.F., Monsen, N.J.N., Eggen, T.H., Eichele, H., Adolfsdottir, S., Sørensen, L., 2018. Keeping emotions in mind: the influence of working memory capacity on parent-reported symptoms of emotional lability in a sample of children with and without ADHD. Front. Psychol. 9, 1846 https://www.ncbi.nlm.nih.gov/ pmc/articles/PMC6176092/. https://www.ncbi.nlm.nih.gov/pmc/articles/PM C6176092/. Joshi, G., DiSalvo, M., Faraone, S.V., Wozniak, J., Fried, R., Galdo, M., Biederman, J.,
- Predictive utility of autistic traits in youth with ADHD: a controlled 10-year longitudinal follow-up study. Eur.ChildAdolesc.Psychiatry 29 (6), 791–801 https:// www.ncbi.nlm.nih.gov/pmc/articles/PMC6176092/. https://www.ncbi.nlm.nih. gov/pmc/articles/PMC6176092/. Lange, K.W., Reichl, S., Lange, K.M., Tucha, L., Tucha, O., 2010. The history of attention deficit hyperactivity disorder. Atten.DeficitHyperact.Disord. 2 (4), 241–255 https:// link.springer.com/content/pdf/10.1007%2Fs12402-010-0045-8.pdf. https://link.sp ringer.com/content/pdf/10.1007%2Fs12402-010-0045-8.pdf. Leaberry, K.D., Rosen, P.J., Fogleman, N.D., Walerius, D.M., Slaughter, K.E., 2020. Comorbid internalizing and externalizing disorders predict lability of negative emotions among children with ADHD. J. Atten. Disord. 24 (14), 1989–2001 https:// journals.sagepub.com/doi/10.1177/1087054717734647?url_ver=Z39.88- 2003 &rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed. https://jour nals.sagepub.com/doi/10.1177/1087054717734647?url_ver=Z39.88-2003&rfr id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed. Liu, L., Chen, W., Vitoratou, S., Sun, L., Yu, X., Hagger-Johnson, G., Wang, Y., 2019. Is emotional lability distinct from "Angry/Irritable mood," "Negative affect," or other subdimensions of oppositional defiant disorder in children with ADHD? J. Atten. Disord. 23 (8), 859–868 https://journals.sagepub.com/doi/10.1177/ 1087054715624228?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr pub%20%200pubmed. https://journals.sagepub.com/doi/10.1177/ 624228?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%20% 200pubmed. Lubke, G., Muth´en, B.O., 2007. Performance of factor mixture models as a function of model size, covariate effects, and class-specific parameters. Struct.Equ.Model. 14 (1), 26 – 47 https://psycnet.apa.org/record/2007-00656-002. https://psycnet.apa.org/re cord/2007-00656-002. Masyn, K.E., 2013. Latent class analysis and finite mixture modeling. In: Little, T.D. (Ed.), The Oxford Handbook of Quantitative Methods. Oxford University Press, New York, NY, pp. 551–611. Muth´en, L., Muth´en, B., 1998. Mplus User's Guide, 6th ed. Muth´en & Muth´en, Los Angeles, CA.
