687 – Bifactor Models Support ICD-10 Construct of ADHD Against DSM-IV, Both in Children Rated by Teachers and Self-rated in Adults

Abstract

Introduction:

Attention Deficit-Hyperactivity Disorder (ADHD) is present in both DSM-IV and ICD-10, with slight differences. For instance, one item is allocated to hyperactivity in DSM-IV and to impulsivity in ICD-10. ADHD is best conceived as a set of core symptoms with some specificity in phenotypal symptomatology. Regarding the specific components, their number remains an open question, the answer oscillating between two (Attention and Hyperactivity-Impulsivity) and three (Attention, Hyperactivity, and Impulsivity).

Aims:

The aim of this study is to contrast alternative measurement models of instruments assessing ADHD symptoms (2 versus 3 specific factors, hierarchical versus bifactor models) across instruments, and age groups.

Methods:

We analysed data from the new ChiP-ARD study. Youths (n = 892) aged 5 to 18 years-old were randomly selected and rated by their teachers (ADHD-Rating Scale, and SWAN). Parents (n = 1,171) of these youths rated their own behaviour using the Adult ADHD Symptom Rating Scale (ASRS). The fit of alternative models for ordered-categorical items was tested using the robust Weighted Least Square Estimator (WLSMV).

Results:

Results support a bifactor model including one global ADHD factor and two specific Inattention and Hyperactivity- Impulsivity factors. The results also show that the Hyperactivity-Impulsivity factor is improperly defined, and unreliable, calling into question the existence of this subtype. Furthermore, the fit improved when Impulsivity was measured by 4 (ICD-10) versus 3 (DSM-IV) items.

Conclusion:

ADHD is a continuous condition from childhood to adulthood, and the “Talk too much” item is best conceived as a measure of Impulsivity than Hyperactivity.