Categorical and dimensional approaches to the developmental relationship between ADHD (Attention-Deficit/Hyperactivity Disorder) and irritability

Rania Johns-Mead, Doctor of Clinical Psychology Candidate/Provisional Psychologist, at Deakin University, delivers a video abstract on her JCPP paper ‘Categorical and dimensional approaches to the developmental relationship between ADHD and irritability‘.

Authors; Rania Johns-Mead, Nandita Vijayakumar, Melissa Mulraney, Glenn Melvin, George Youssef, Emma Sciberras, Vicki A. Anderson, Jan M. Nicholson, Daryl Efron, Philip Hazel, Tim J. Silk

First published: 11 May 2023

Open Access paper doi.org/10.1111/jcpp.13818

Rania Johns-Mead is a candidate in the Doctor of Psychology (Clinical) program at Deakin University, Melbourne Australia. She is working with the Cognitive Neuroscience Unit to complete her thesis, investigating the course, socioemotional correlates, and neurodevelopmental underpinnings of irritability in the context of ADHD. Rania has a professional background in youthwork, and is particularly interested in trajectories of mental health issues during childhood and adolescence. Over the course of her studies, she has developed a further interest in how neurodevelopmental and neurocognitive research inform interventions and diagnostic considerations in clinical practice.

Transcript

[00:00:18.060] Rania Johns-Mead: Hi, my name is Rania Johns-Mead, and I’m completing a Doctorate in Clinical Psychology at Deakin University.  I’ve been working with a Research Team to investigate irritability in ADHD.  Recently, part of our research was published in the Journal of Child Psychology and Psychiatry, in a paper called “Categorical and Dimensional Approaches to the Developmental Relationship Between ADHD and Irritability.”  In this video abstract, I’ll provide a brief overview of our methods and findings and discuss possible implications.

But first some background.  Irritability is known to be common in ADHD and is of interest due to its association with poorer outcomes in various domains.  Interestingly, irritability is thought to have similar neurological underpinnings to impulsivity and follows a similar trajectory to hyperactive-impulsive symptoms, declining over childhood and adolescence.  In contrast, inattentiveness seems to remain stable over development.  Some Researchers have suggested because of this, that irritability is a form of emotional impulsivity, an extension of behavioural impulsivity and difficulty with self-regulation.  Others argue that irritability in ADHD arises from comorbidities, rather than being specifically associated with ADHD symptoms.

Despite the overlaps between irritability and ADHD symptoms, there is limited longitudinal research on whether irritability trajectories differ between children with and without ADHD, or whether this is influenced by the type, severity and course of ADHD symptoms.  Accordingly, our study sought to answer three questions.  One, “Does change in irritability from late childhood to early adolescence differ based on childhood history of ADHD diagnosis?”  Two, “Is it predicted by baseline symptom counts for each domain, i.e., inattentive and hyperactive-impulsive?”  And three, “Is it associated with change in severity of ADHD symptoms over time?”

Our study used a community sample of 337 participants recruited as a part of the Children’s Attention Project, of which around 45% met diagnostic criteria for ADHD.  At baseline, participants had a mean age of ten and a half and follow-up was conducted 18 months later.  Irritability was measured at baseline and at follow-up, using the Affective Reactivity Index or ARI.  The DISC-IV was used at baseline to assess ADHD status and number of symptoms in each symptom domain, and the Conners-3 ADHD Index was administered at both baseline and follow-up as an indicator of symptom severity.

Our analyses used Latent Change Score models, as they allow the estimation of variants within the change factor and capture the extent to which change scores depend on baseline levels.  For models with two or more longitudinal variables, they also assessed covariance at baseline, cross-domain coupling, which is the extent to which change in one domain is predicted by baseline levels of the other, and correlation of concurrent change, so the extent to which change in one variable is correlated with change in another.

We conducted three Latent Change Score models, so co-varying for age, gender, other diagnoses and medication status.  The first was a univariate model of ARI, split by ADHD group status.  We used the Wald test to compare group differences on parameters of interest.  The second was a univariate model of ARI, with baseline counts of inattention and hyperactivity-impulsivity as predictors.  The third was a bivariate model of ARI and Conners-3 ADHD Index, which assessed baseline covariance, cross-domain coupling and simultaneous latent change between two longitudinal variables.

Our results showed that irritability was significantly higher among the ADHD group than controls, but followed a similar trajectory, suggesting stable differences, which are likely to persist into adolescence.  We also found that latent change in irritability was predicted by higher counts of baseline hyperactive-impulsive symptoms, but not by inattentiveness.  Interestingly, and counter to predictions, greater levels of baseline hyperactivity-impulsivity predicted greater declines in irritability.  Based on existing knowledge of typical ADHD symptom trajectories, one feasible explanation is that hyperactive-impulsive symptoms tend to decline with age, while inattentiveness remains stable.  The decrease in irritability predicted by higher baseline hyperactivity may have been accompanied by a concurrent decrease in hyperactivity, but unfortunately, we did not capture in our data.

This explanation is consistent with the results of our bivariate analysis, as well, which showed that change in ADHD symptoms and change in irritability were correlated, with greater declines in ADHD symptoms predicting correspondingly greater declines in irritability over time.  Even though this analysis didn’t differentiate between symptom domains, existing knowledge of typical symptom trajectories suggests decreasing ADHD severity is likely to reflect changes in hyperactive-impulsive symptoms, rather than inattention.

The bivariate analysis also found that change in irritability was predicted by baseline ADHD symptoms, while the opposite cross-coupling relationship between baseline irritability and change in ADHD symptoms was not significant, indicating ADHD symptoms drive change in irritability, rather than the reverse.

Considered together, we can make more nuanced inferences from these results.  In each analysis, results supported the hypothesis that irritability is associated with ADHD symptoms, not solely attributable to comorbidities.  This makes sense if, as our findings suggest, irritability in ADHD is at least partially driven by ADHD symptoms.  Findings regarding the relationship between irritability and hyperactive impulsivity align with views of irritability as a form of emotional impulsivity, which again, supports the specific association between irritability and ADHD.

Finally, the fact that categorical ADHD status didn’t predict change in irritability independently of baseline levels, but dimensional measures of both hyperactive-impulsive symptoms and of ADHD symptom severity did, show that this relationship is not limited to cases above the diagnostic thresholds, a fact with important clinical indications.  And crucially, we would not have been able to make these inferences using traditional regressions or categorical diagnostic variables, demonstrating the importance of using dimensional and longitudinal approaches to capture heterogeneity and developmental change within the ADHD population.

Thanks for watching and we hope you’ll have a look at our full article.