Item Response Theory Analysis of ADHD Symptoms in Children With and Without ADHD

Impulsivity as key bridge symptoms in cross-sectional and longitudinal networks of ADHD and ODD

BACKGROUND

Impulsivity is viewed as key to attention-deficit/hyperactivity disorder (ADHD) and disruptive behavior disorders (DBD). Yet, to date, no work has provided an item-level analysis in longitudinal samples across the critical developmental period from childhood into adolescence, despite prior work suggesting items exhibit differential relevance with respect to various types of impairment. The current study conducted a novel longitudinal network analysis of ADHD and oppositional defiant disorder (ODD) symptoms between childhood and adolescence, with the important applied prediction of social skills in adolescence.

METHODS

Participants were 310 children over-recruited for clinical ADHD issues followed longitudinally for six years in total with gold standard diagnostic procedures and parent and teacher ratings of symptoms and social outcomes.

RESULTS

Findings from baseline, Year 3, and Year 6 suggested Difficulty waiting turn, Blurts, and Interrupts/intrudes were key bridge items across cross-sectional and longitudinal parent-reported DBD networks. Furthermore, shortened symptom lists incorporating these symptoms were stronger predictors of teacher-rated social skills 5 years later compared to total DBD scores.

CONCLUSIONS

Such findings are consistent with the trait impulsivity theory of DBD and ADHD and may inform useful screening tools and personalized intervention targets for children at risk for DBD during adolescence.

Data-Driven Methods for Predicting ADHD Diagnosis and Related Impairment: The Potential of a Machine Learning Approach

Current diagnostic criteria for ADHD include several symptoms that highly overlap in conceptual meaning and interpretation. Additionally, inadequate sensitivity and specificity of current screening tools have hampered clinicians' ability to identify those at risk for related outcomes. Using machine learning techniques, the current study aimed to propose a novel algorithm incorporating key ADHD symptoms to predict concurrent and future (i.e., five years later) ADHD diagnosis and related impairment levels. Participants were 399 children with and without ADHD; multiple informant measures of ADHD symptoms, global impairment, academic performance, and social skills were included as part of an accelerated longitudinal design. Results suggested eight symptoms as most important in predicting impairment outcomes five years later: (1) Has difficulty sustaining attention in tasks or play activities, (2) Does not follow through on instructions and fails to finish work, (3) Has difficulty organizing tasks and activities, (4) Avoids tasks (e.g., schoolwork, homework) that require sustained mental effort, (5) Is often easily distracted, (6) Is often forgetful in daily activities, (7) Fidgets with hands or feet or squirms in seat, and (8) Interrupts/intrudes on others. The algorithm comprising this abbreviated list of symptoms performed just as well as or significantly better than one comprising all 18 symptoms in predicting future global impairment and academic performance, but not social skills. It also predicted concurrent and future ADHD diagnosis with 81-93% accuracy. Continued development of screening tools will be key to ensuring access to clinical services for youth at risk for ADHD.

Explaining the Prospective Association of Positive and Negative Parenting Behaviors and Child ADHD Symptoms: Pathways Through Child Executive Function and Reward Responsivity

Objective: Parenting behavior is a well-established correlate of offspring ADHD. However, little is known about how parenting exerts its effects on offspring ADHD symptomatology. We examined whether prospective associations between positive and negative parenting behaviors and child ADHD symptoms are mediated by deficits in child executive function (EF) and reward responsivity (RR). Method: One hundred and thirty-five children with and without ADHD were assessed across two Waves, when children were mean ages 6 and 8 respectively. Children completed tasks on EF, and parents completed questionnaires about their parenting behaviors and their children's RR and ADHD symptoms. Results: Negative parenting behavior at Wave 1 was indirectly associated with offspring ADHD symptoms at Wave 2 via offspring EF. Conclusion: Individual differences in EF, but not RR, during early childhood may constitute a potential pathway by which negative parenting behaviors exerts its effects on subsequent offspring ADHD symptomatology. Treatment implications are discussed.

Development of the short Creative Expression Interest Scale based on item response theory

This study develops a short Creative Expression Interest Scale (CEIS) among Chinese freshmen based on the perspective of item response theory (IRT). Nine hundred fifty-nine valid Chinese freshmen participated in the Creative Expression Interest survey. Researchers applied the initial data for unidimensionality, item fit, discrimination parameter, and differential item functioning to obtain a short CEIS. The results show that the Short CEIS meets the psychometric requirements of the IRT. Pearson correlation coefficient of theta between the short and long CEIS is 0.922. The marginal reliability of the short CEIS is 0.799. These indicate that the short CEIS developed in this study among Chinese freshmen, meets the psychometric requirements. Although the Short CEIS can eliminate redundant, uninformative items, save time, and improve the quality of data collection. However, the validity of this short scale needs further validation.

Network analysis of 18 attention-deficit/hyperactivity disorder symptoms suggests the importance of "Distracted" and "Fidget" as central symptoms: Invariance across age, gender, and subtype presentations

The network theory of mental disorders conceptualizes psychiatric symptoms as networks of symptoms that causally interact with each other. Our present study aimed to explore the symptomatic structure in children with attention-deficit/hyperactivity disorder (ADHD) using network analyses. Symptom network based on 18 items of ADHD Rating Scale-IV was evaluated in 4,033 children and adolescents with ADHD. The importance of nodes was evaluated quantitatively by examining centrality indices, including Strength, Betweenness and Closeness, as well as Predictability and Expected Influence (EI). In addition, we compared the network structure across different subgroups, as characterized by ADHD subtypes, gender and age groups to evaluate its invariance. A three-factor-community structure was identified including inattentive, hyperactive and impulsive clusters. For the centrality indices, the nodes of "Distracted" and "Fidget" showed high closeness and betweenness, and represented a bridge linking the inattentive and hyperactive/impulsive domains. "Details" and "Fidget" were the most common endorsed symptoms in inattentive and hyperactive/impulsive domains respectively. On the contrary, the "Listen" item formed a peripheral node showing weak links with all other items within the inattentive cluster, and the "Loss" item as the least central node by all measures of centrality and with low predictability value. The network structure was relatively invariant across gender, age and ADHD subtypes/presentations. The 18 items of ADHD core symptoms appear not equivalent and interchangeable. "Distracted" and "Fidget" should be considered as central, or core, symptoms for further evaluation and intervention. The network-informed differentiation of these symptoms has the potentials to refine the phenotype and reduce heterogeneity.

Differential utility of teacher and parent-teacher combined information in the assessment of Attention Deficit/Hyperactivity Disorder symptoms

BACKGROUND

Consistent research findings indicate that parents and teachers observe genuinely different Attention Deficit/Hyperactivity Disorder (ADHD) behaviours in their respective settings.

OBJECTIVE

To evaluate the utility of information provided by teacher informant assessments (INFAs) of ADHD symptoms, and the implications of aggregation algorithms in combing parents' information, i.e. using 'or-rule' (endorsement by either one informant) versus 'and-rule' (endorsement by both informants).

METHOD

Teacher ratings on Conners scales and clinical data from parental accounts on 1383 probands and their siblings from the IMAGE study were analysed. The psychometric properties of teacher and combined ratings using the item response theory model (IRT) are presented. Kappa coefficients, intraclass correlations and linear regression were employed.

RESULTS

First, teacher endorsement of symptoms is located in a narrow part of the trait continuum close to the average levels. Symptoms exhibit comparable perception in the measurement of the trait(s) with similar discrimination ability and information (reliability). Second, the IRT properties of the 'or-rule' ratings are predominantly influenced by parent-INFAs; and the 'and-rule' ratings predominantly by teacher-INFAs ratings. Third, parent-teacher INFAs agreement was low, both for individual items (κ = 0.01-0.15) and for dimensional scores (r = 0.12-0.16). The 'or-rule' captured milder expressions of ADHD symptoms, whereas the 'and-rule' indexed greater severity of ADHD.

CONCLUSIONS

Parent and teacher-INFAs provide different kinds of information, while both are useful. Teacher-INFA and the 'and-rule' provide a more accurate index of severity than an additive symptom count. Parent-INFA and the 'or-rule' are more sensitive for detecting cases with milder ADHD.

Clinical Subtypes in Children with Attention-Deficit Hyperactivity Disorder According to Their Child Behavior Checklist Profile

This study sought to identify subgroups of attention-deficit hyperactivity disorder (ADHD) defined by specific patterns of emotional and behavioral symptoms according to the parent-rated Child Behavior Checklist (CBCL). Our clinical sample comprised 314 children (aged 4 to 15 years) diagnosed with ADHD according to the DSM-5. In addition, comorbid psychiatric disorders, general functioning, and medication status were assessed. Cluster analysis was performed on the CBCL syndrome subscales and yielded a solution with four distinct subgroups. The "High internalizing/externalizing" group displayed an overlap between internalizing and externalizing problems in the CBCL profile. In addition, the "High internalizing/externalizing" group revealed a high rate of comorbid autism spectrum disorder and elevated autistic traits. The "Inattention and internalizing" group revealed a high rate of the predominantly inattentive presentation according to ADHD specifier from the DSM-5. The "Aggression and externalizing" group revealed a high rate of comorbid oppositional defiant disorder and conduct disorder. The "Less psychopathology" group scored low on all syndrome scales. Children with ADHD were subdivided into four distinct subgroups characterized by psychopathological patterns, with and without internalizing and externalizing problems. The overlap between internalizing and externalizing problems may be mediated with emotional dysregulation and associated neurobiological bases.

Item Response Theory Analyses of Barkley's Adult Attention-Deficit/Hyperactivity Disorder Rating Scales

There are many challenges associated with assessment and diagnosis of ADHD in adulthood. Utilizing the graded response model (GRM) from item response theory (IRT), a comprehensive item-level analysis of adult ADHD rating scales in a clinical population was conducted with Barkley's Adult ADHD Rating Scale-IV, Self-Report of Current Symptoms (CSS), a self-report diagnostic checklist and a similar self-report measure quantifying retrospective report of childhood symptoms, Barkley's Adult ADHD Rating Scale-IV, Self-Report of Childhood Symptoms (BAARS-C). Differences in item functioning were also considered after identifying and excluding individuals with suspect effort. Items associated with symptoms of inattention (IA) and hyperactivity/impulsivity (H/I) are endorsed differently across the lifespan, and these data suggest that they vary in their relationship to the theoretical constructs of IA and H/I. Screening for sufficient effort did not meaningfully change item level functioning. The application IRT to direct item-to-symptom measures allows for a unique psychometric assessment of how the current DSM-5 symptoms represent latent traits of IA and H/I. Meeting a symptom threshold of five or more symptoms may be misleading. Closer attention given to specific symptoms in the context of the clinical interview and reported difficulties across domains may lead to more informed diagnosis.

Toward a Revised Nosology for Attention-Deficit/Hyperactivity Disorder Heterogeneity

Attention-deficit/hyperactivity disorder (ADHD) is among the many syndromes in the psychiatric nosology for which etiological signal and clinical prediction are weak. Reducing phenotypic and mechanistic heterogeneity should be useful to arrive at stronger etiological and clinical prediction signals. We discuss key conceptual and methodological issues, highlighting the role of dimensional features aligned with Research Domain Criteria and cognitive, personality, and temperament theory as well as neurobiology. We describe several avenues of work in this area, utilizing different statistical, computational, and machine learning approaches to resolve heterogeneity in ADHD. We offer methodological and conceptual recommendations. Methodologically, we propose that an integrated approach utilizing theory and advanced computational logic to address targeted questions, with consideration of developmental context, can render the heterogeneity problem tractable for ADHD. Conceptually, we conclude that the field is on the cusp of justifying an emotionally dysregulated subprofile in ADHD that may be useful for clinical prediction and treatment testing. Cognitive profiles, while more nascent, may be useful for clinical prediction and treatment assignment in different ways depending on developmental stage. Targeting these psychological profiles for neurobiological and etiological study to capture different pathophysiological routes remains a near-term opportunity. Subtypes are likely to be multifactorial, cut across multiple dimensions, and depend on the research or clinical outcomes of interest for their ultimate selection. In this context parallel profiles based on cognition, emotion, and specific neural signatures appear to be on the horizon, each with somewhat different utilities. Efforts to integrate such cross-cutting profiles within a conceptual dysregulation framework are well underway.

Cross-Validation and Development of Empirically Derived ADHD Assessment Strategies: Insights From the National Longitudinal Study of Adolescent Health (Add Health)

Objective: There is replicated evidence that individual ADHD symptoms differentially predict ADHD diagnostic status, and that non-Diagnostic and Statistical Manual of Mental Disorders (DSM)-based symptom algorithms are superior to DSM criteria for ruling in or ruling out ADHD. However, these findings have been limited to case-control samples, despite the need to replicate them in independent and more representative samples and to increase their generalizability. Method: We analyzed the base rates and psychometric properties of ADHD symptoms in the population-based National Longitudinal Study of Adolescent Health (Add Health; http://www.cpc.unc.edu/projects/addhealth ) sample, and evaluated the predictive utility of empirically derived ADHD symptoms against the DSM approach with respect to academic, health, and relational functional outcomes. The sample consisted of 11,247 (54% female) ethnically diverse (45% non-White) individuals who retrospectively self-reported the severity of their ADHD symptoms from 5 to 12 years. Results: Individual ADHD symptoms variably predicted ADHD status, and whereas most symptoms were most predictive when endorsed at the two highest severity levels in previous studies, symptoms in Add Health were often most predictive of ADHD when endorsed at the highest severity level. Conclusion: The divergence in optimal severity thresholds for ADHD symptoms potentially reflects the different symptom base rates in clinical versus population-based samples. The accuracy of the different classification approaches also suggested that different symptom algorithms may be superior to evaluate ADHD depending on the assessment setting.

Measurement Invariance of the ADHD Rating Scale-IV Home and School Versions Across Age, Gender, Clinical Status, and Informant

The present study aimed to investigate the measurement invariance across age, gender, clinical status, and informant of the Attention-Deficit/Hyperactivity Disorder Rating Scale-IV (ADHD-RS-IV) Home and School versions. The participants were 1,106 Romanian children and adolescents (mean age = 12.74 years, standard deviation = 2.84, age range 6-18 years). Both parents and teachers assessed ADHD symptoms. The factorial structure of the scale was assessed using confirmatory factor analysis, and measurement invariance was assessed using multigroup confirmatory factor analysis. The results supported the reliability of the ADHD-RS-IV, with high internal consistency coefficients for both versions. Confirmatory factor analysis validated a two-factor model. Multigroup confirmatory factor analysis confirmed the measurement invariance of ADHD-RS-IV across age, gender, clinical status, and informant. ADHD-RS-IV had good psychometric properties in a sample of Romanian children and adolescents. It is a reliable instrument given its strong invariance. Implications for evidence-based assessment of ADHD are discussed.

A data mining and item response mixture modeling method to retrospectively measure Diagnostic and Statistical Manual of Mental Disorders-5 attention deficit hyperactivity disorder in the 1970 British Cohort Study

OBJECTIVE

To facilitate future outcome studies, we aimed to develop a robust and replicable method for estimating a categorical and dimensional measure of Diagnostic and Statistical Manual of Mental Disorders-5 (DSM-5) attention deficit hyperactivity disorder (ADHD) in the 1970 British Cohort Study (BCS70).

METHOD

Following a data mining framework, we mapped DSM-5 ADHD symptoms to age 10 BCS70 data (N = 11,426) and derived a 16-item scale (α = 0.85). Mapping was validated by an expert panel. A categorical subgroup was derived (n = 594, 5.2%), and a zero-inflated item response theory (IRT) mixture model fitted to estimate a dimensional measure.

RESULTS

Subgroup composition was comparable with other ADHD samples. Relative risk ratios (ADHD/not ADHD) included boys = 1.38, unemployed fathers = 2.07, below average reading = 2.58, and depressed parent = 3.73. Our estimated measures correlated with two derived reference scales: Strengths and Difficulties Questionnaire hyperactivity (r = 0.74) and a Rutter/Conners-based scale (r = 0.81), supporting construct validity. IRT model items (symptoms) had moderate to high discrimination (0.90-2.81) and provided maximum information at average to moderate theta levels of ADHD (0.5-1.75).

CONCLUSION

We extended previous work to identify ADHD in BCS70, derived scales from existing data, modeled ADHD items with IRT, and adjusted for a zero-inflated distribution. Psychometric properties were promising, and this work will enable future studies of causal mechanisms in ADHD.

Identifying potentially marker symptoms of attention-deficit/hyperactivity disorder

Background

For the diagnosis of attention-deficit/hyperactivity disorder (ADHD), the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) proposes that adherence to six symptoms in either group (inattention and hyperactivity/impulsivity) will lead to the diagnosis of one of three presentations of the disorder. Underlying this diagnostic algorithm is the assumption that the 18 symptoms have equal relevance for the diagnosis of ADHD, all are equally severe, and all have the same power to detect the presence of the disorder in all its degrees of severity, without considering the possibility of using marker symptoms. However, several studies have suggested that ADHD symptoms differ in both their power to discriminate the presence of the disorder and the degree of severity they represent. The aim of the present study was to replicate the results of previous research by evaluating the discriminative capacity and relative severity of ADHD symptoms, as well as to extend the investigation of this topic to Spanish-speaking Latin American samples.

Methods

The properties of ADHD symptoms rated by the parents of 474 Chilean children were analyzed. Symptom parameters were estimated using the graded response model.

Results

The results suggest that symptoms of ADHD differ substantially in both the accuracy with which they reflect the presence of the disorder, and their relative severity. Symptoms “easily distracted by extraneous stimuli” and “have difficulty sustaining attention in tasks” (inattention) and “is on the go, acting as if driven by motor” (hyperactivity/impulsivity) were the most informative, and those with relatively lower severity thresholds.

Discussion

The fact that symptoms differ substantially in the probability of being observed conditionally to the trait level suggests the need to refine the diagnostic process by weighting the severity of the symptom, and even to assess the possibility of defining ADHD marker symptoms, as has been done in other disorders.

Evaluating the Hierarchical Structure of ADHD Symptoms and Invariance Across Age and Gender

The bifactor model of attention-deficit/hyperactivity disorder (ADHD) has been extensively explored, yet the tendency of the bifactor model to overfit data necessitates investigation of alternative, more parsimonious models, such as a modified bifactor structure. The present study used item response theory to compare unidimensional, correlated factors, bifactor, and modified bifactor models of ADHD symptoms in a clinical sample of youth ( N = 1,612) and examined differential item functioning (DIF) by age (<11 and ≥11 years) and gender. Results suggested that two restricted bifactor models showed superior fit compared with alternative models, and support strong general and inattention dimensions, with unreliable hyperactivity and impulsivity dimensions. No DIF was found across gender or age. The present study confirms that the general dimension (i.e., inhibition) and one specific dimension (i.e., sustained attention) represent distinct variability in ADHD symptoms that may improve prediction of symptom persistence, treatment response, or functional outcomes.

Bifactor Models of Attention-Deficit/Hyperactivity Disorder (ADHD): An Evaluation of Three Necessary but Underused Psychometric Indexes

BACKGROUND

In the past decade, the bifactor model of attention-deficit/hyperactivity disorder (ADHD) has been extensively researched. This model consists of an ADHD general dimension and two specific factors: inattention and hyperactivity/impulsivity. All studies conclude that the bifactor is superior to the traditional two-correlated factors model, according to the fit obtained by factor analysis. However, the proper interpretation of a bifactor not only depends on the fit but also on the quality of the measurement model.

OBJECTIVE

To evaluate the model-based reliability, distribution of common variance and construct replicability of general and specific ADHD factors.

METHOD

We estimated expected common variance, omega hierarchical/subscale and H-index from standardized factor loadings of 31 ADHD bifactor models previously published.

RESULTS AND CONCLUSION

The ADHD general factor explained most of the common variance. Given the low reliable variance ratios, the specific factors were difficult to interpret. However, in clinical samples, inattention acquired sufficient specificity and stability for interpretation beyond the general factor. Implications for research and clinical practice are discussed.

Hierarchy and Psychometric Properties of ADHD Symptoms in Spanish Children: An Application of the Graded Response Model

The Diagnostic and Statistical Manual of Mental Disorders (DSM) diagnostic criteria assume that the 18 symptoms carry the same weight in an Attention Deficit with Hyperactivity Disorder (ADHD) diagnosis and bear the same discriminatory capacity. However, it is reasonable to think that symptoms may differ in terms of severity and even in the reliability with they represent the disorder. To test this hypothesis, the aim of this study was to calibrate in a sample of Spanish children (age 4–7; n = 784) a scale for assessing the symptoms of ADHD proposed by Diagnostic and Statistical Manual of Mental Disorders, IV-TR within the framework of Item Response Theory. Samejima’s Graded Response Model was used as a method for estimating the item difficulty and discrimination parameters. The results showed that ADHD subscales (Attention Deficit and Hyperactivity / Impulsivity) had good psychometric properties and had also a good fit to the model. However, relevant differences between symptoms were observed at the level of severity, informativeness and reliability for the assessment of ADHD. This finding suggests that it would be useful to identify the symptoms that are more important than the others with regard to diagnosing ADHD.