Comparative Analysis of the WISC between Two ADHD Subgroups

Prediction of attention deficit hyperactivity disorder based on explainable artificial intelligence

Accurate assessment of Attention Deficit Hyperactivity Disorder (ADHD) is crucial for the effective treatment of affected individuals. Traditionally, psychometric tests such as the WISC-IV have been utilized to gather evidence and identify patterns or factors contributing to ADHD diagnosis. However, in recent years, the use of machine learning (ML) models in conjunction with post-hoc eXplainable Artificial Intelligence (XAI) techniques has improved our ability to make precise predictions and provide transparent explanations. The objective of this study is twofold: firstly, to predict the likelihood of an individual receiving an ADHD diagnosis using ML algorithms, and secondly, to offer interpretable insights into the decision-making process of the ML model. The dataset under scrutiny comprises 694 cases collected over the past decade in Spain, including information on age, gender, and WISC-IV test scores. The outcome variable is the professional diagnosis. Diverse ML algorithms representing various learning styles were rigorously evaluated through a stratified 10-fold cross-validation, with performance assessed using key metrics, including accuracy, area under the receiver operating characteristic curve, sensitivity, and specificity. Models were compared using both the full set of initial features and a well-suited wrapper-type feature selection algorithm (Boruta). Following the identification of the most suitable model, Shapley additive values were computed to assign weights to each predictor based on their additive contribution to the outcome and to elucidate the predictions. Strikingly, a reduced set of 8 out of the initial 20 variables produced results comparable to using the full feature set. Among the ML models tested, the Random Forest algorithm outperformed others on most metrics (ACC = 0.90, AUC = 0.94, Sensitivity = 0.91, Specificity = 0.92). Notably, the principal predictors, ranked by importance, included GAI - CPI, WMI, CPI, PSI, VCI, WMI - PSI, PRI, and LN. Individual case examples exhibit variations in predictions depending on unique characteristics, including instances of false positives and negatives. Our ML model adeptly predicted ADHD diagnoses in 90% of cases, with potential for further enhancement by expanding our database. Furthermore, the use of XAI techniques enables the elucidation of salient factors in individual cases, thereby aiding inexperienced professionals in the diagnostic process and facilitating comparison with expert assessments. It is important to note that this tool is designed to support the ADHD diagnostic process, where the medical professional always has the final say in decision-making.

Is Processing Speed (Gs) related to Hyperactivity (As a Narrow Cognitive Area of Gps): A Dimensional Approach to Heterogeneity of Clinical and WISC-IV Cognitive Profiles in ADHD From RDoC/HiTOP Perspective

OBJECTIVE

The aim of this study was to investigate the relationship between symptomatology and the Wechsler Intelligence Scale for Children, Fourth Edition (WISC-IV) cognitive profiles in Attention Deficit and Hyperactivity Disorder (ADHD), taking into account clinical heterogeneity.

METHOD

The WISC-IV was administered to 60 children aged 6 to 13 years with ADHD who had not previously taken psychotropic medication. Conners Teacher-Rating Scale (CTRS-R/L) and Conners Parent-Rating Scale (CPRS-R/L) were completed by parents and teachers.

RESULTS

We found a statistically significant positive correlation between the symptoms of hyperactivity/impulsivity and PSI (Processing Speed Index). A statistically significant negative correlation was found between the CTRS-R/L's DSM-IV Inattentive subscale and PRI (Perceptual Reasoning Index). No relationship was found between the inattention or hyperactivity/impulsivity symptomatology with Working Memory Index (WMI). WMI stood out as the most frequently reported weakness among the four index scores.

CONCLUSION

Considering the relationship between the cognitive domains of RDoC (Research Domain Criteria) and dimensions of HiTOP with the ADHD symptom clusters, the assessment of cognitive and behavioral symptoms may be useful for phenotyping ADHD. According to the CHC (Cattell-Horn-Carroll) theory; the positive correlation level between PSI, which is accepted as Gs, and hyperactivity/impulsivity symptoms in people with ADHD may be due to the fact that hyperactivity is one of the narrow cognitive domains of Gps.

Children and adolescents with neurodevelopmental disorders show cognitive heterogeneity and require a person-centered approach

We aimed to identify patterns of cognitive differences and characterize subgroups of Mexican children and adolescents with three neurodevelopmental disorders (NDD): intellectual disability (ID), autism spectrum disorders (ASD) and attention deficit/hyperactivity disorder (ADHD). The sample included 74 children and adolescents 6–15 years; 34% had ID, ASD or ADHD, 47% had ID in comorbidity with ASD, ADHD or both, 11% had ASD + ADHD, 8% were children without NDD. We applied WISC-IV, Autism Diagnostic Interview-Revised, Mini-International Neuropsychiatric Structured Interview, Child Behavior Checklist, and UNICEF Child Functioning Module. We evaluated the normality of the WISC-IV sub-scales using the Shapiro-Francia test, then conducted a latent class analysis and assessed inter-class differences in terms of household, parent and child characteristics. The following four-class solution best fit the data: “Lower Cognitive Profile” (LCP), “Lower Working Memory” (LWM), “Higher Working Memory” (HWM), “Higher Cognitive Profile” (HCP). LCP included most of the children with ID, who had a low Working Memory (WM) index score. LWM included mainly children with ASD or ID + ADHD; their Perceptual Reasoning (PR) and Processing Speed (PS) index scores were much higher than those for Verbal Comprehension (VC) and WM. HWM included children with ASD or ADHD; their scores for PR, PS and VC were high with lower WM (although higher than for LWM). HCP included children without NDD and with ASD or ADHD or both and had the highest scores on all indices. Children with NDD show cognitive heterogeneity and thus require individualized treatment plans.

Verbal working memory and processing speed: Correlations with the severity of attention deficit and emotional dysregulation in adult ADHD

Objectives

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM‐5), emphasizes symptoms severity with regard to the diagnosis of attention deficit hyperactivity disorder (ADHD). Many clinicians use neuropsychological test results as objective measures of cognitive functions as part of the diagnostic work‐up. The aim of this study was to investigate whether the psychometric test results regarding verbal working memory and processing speed are useful as indicators of the severity of attention deficits and emotional dysregulation in adults with ADHD.

Methods

This observational cross‐sectional clinical study included 418 adults diagnosed with ADHD according to the DSM‐5. Attention deficit severity was defined based on the inattentive subscale of the Adult ADHD Self‐Report Scale. Emotional dysregulation was assessed with the Deficient Emotional Self‐Regulation scale. Verbal working memory was measured with the Working Memory Index (WMI), and processing speed was measured with the Processing Speed Index (PSI) from the Wechsler Adult Intelligence Scale, third edition.

Results

The full‐scale intelligence quotients of the participants were in the normal range, with expected reductions in verbal working memory and processing speed. Only processing speed was associated with attention deficits (β = −.056, p = .003). The association between the psychometric test result for verbal working memory and processing speed and that between the severity of attention deficits and emotional dysregulation were weak (R² < .1) and mostly non‐significant.

Conclusion

The psychometric index scores for verbal working memory (WMI) and processing speed (PSI) seem to have limited utility as indicators of the severity of attention deficits and emotional dysregulation in adult ADHD patients.

Sleep, academic achievement, and cognitive performance in children with attention-deficit hyperactivity disorder: A polysomnographic study

The objective of this study was to examine the influence of parent-reported and polysomnography (PSG)-measured sleep patterns on the academic and cognitive performance of children with attention-deficit hyperactivity disorder (ADHD). We assessed 91 children (18 girls) diagnosed with ADHD aged 7-11 years (29 ADHD-Inattentive, 32 ADHD-Hyperactive/Impulsive, and 31 ADHD-Combined). The Paediatric Sleep Questionnaire (PSQ) and Paediatric Daytime Sleepiness Scale (PDSS) were used to assess subjective sleep quality, as perceived by parents, and objective sleep variables were assessed by PSG. Cognitive performance was evaluated using the Wechsler Intelligence Scale for Children (WISC), and the final average grade of the last school year was used as a measure of academic performance. Academic performance was predicted by the following sleep variables: Sleep time, time in bed, night awakenings, and daytime sleepiness. The best predictors of cognitive performance in children with ADHD were rapid eye movement latency, light sleep, periodic limb movements index (PLMs), awakenings, and daytime sleepiness. In conclusion, sleep parameters are closely associated with the academic and cognitive functioning of children with ADHD.

Deconstructing Cognitive Heterogeneity in Puerto Rican Spanish-Speaking Children With ADHD

OBJECTIVE

There is limited understanding of the cognitive profiles of Spanish-speaking children with Attention-Deficit/Hyperactivity Disorder (ADHD). The current study investigated the cognitive cluster profiles of Puerto Rican Spanish-speaking children with ADHD using the Wechsler Intelligence Scales for Children-Fourth Edition Spanish (WISC-IV Spanish) Index scores and examined the association between cognitive cluster profiles with other potentially relevant factors.

METHOD

Hierarchical cluster analysis was used to identify WISC-IV clusters in a sample of 165 Puerto Rican children who had a primary diagnosis of ADHD. To examine the validity of the ADHD clusters, analysis of variances and chi-square analyses were conducted to compare the clusters across sociodemographics (e.g., age and education), type of ADHD diagnosis (ADHD subtype, Learning Disorder comorbidity), and academic achievement.

RESULTS

Clusters were differentiated by level and pattern of performance. A five-cluster solution was identified as optimal that included (C1) multiple cognitive deficits, (C2) processing speed deficits, (C3) generally average performance, (C4) perceptual reasoning strengths, and (C5) working memory deficits. Among the five clusters, the profile with multiple cognitive deficits was characterized by poorer performance on the four WISC-IV Spanish Indexes and was associated with adverse sociodemographic characteristics.

CONCLUSIONS

Results illustrate that there is substantial heterogeneity in cognitive abilities of Puerto Rican Spanish-speaking children with ADHD, and this heterogeneity is associated with a number of relevant outcomes.