Intraindividual variability in inhibitory function in adults with ADHD--an ex-Gaussian approach

Abstract

Objective: Attention deficit disorder (ADHD) is commonly associated with inhibitory dysfunction contributing to typical behavioral symptoms like impulsivity or hyperactivity. However, some studies analyzing intraindividual variability (IIV) of reaction times in children with ADHD (cADHD) question a predominance of inhibitory deficits. IIV is a measure of the stability of information processing and provides evidence that longer reaction times (RT) in inhibitory tasks in cADHD are due to only a few prolonged responses which may indicate deficits in sustained attention rather than inhibitory dysfunction. We wanted to find out, whether a slowing in inhibitory functioning in adults with ADHD (aADHD) is due to isolated slow responses.

Methods: Computing classical RT measures (mean RT, SD), ex-Gaussian parameters of IIV (which allow a better separation of reaction time (mu), variability (sigma) and abnormally slow responses (tau) than classical measures) as well as errors of omission and commission, we examined response inhibition in a well-established GoNogo task in a sample of aADHD subjects without medication and healthy controls matched for age, gender and education.

Results: We did not find higher numbers of commission errors in aADHD, while the number of omissions was significantly increased compared with controls. In contrast to increased mean RT, the distributional parameter mu did not document a significant slowing in aADHD. However, subjects with aADHD were characterized by increased IIV throughout the entire RT distribution as indicated by the parameters sigma and tau as well as the SD of reaction time. Moreover, we found a significant correlation between tau and the number of omission errors.

Conclusions: Our findings question a primacy of inhibitory deficits in aADHD and provide evidence for attentional dysfunction. The present findings may have theoretical implications for etiological models of ADHD as well as more practical implications for neuropsychological testing in aADHD.

Figure 1. Exemplary distribution of RT and relation of Mean, SD and the number of slow responses.

NOTE: Dotted grey arrows illustrate a rightward shift of mean RT and an increase of the SD of RT as a consequence of an increasing number of slow responses forming the right tail of the distribution.

Figure 2. Distribution of age for both adult controls and adult ADHD.

Figure 3. Frequency of intraindividual RTs with fitted ex-Gaussian probability functions exemplary for a control (left) and an ADHD (right) subject.

NOTE: Please keep in mind that RTs in ADHD encompass a broader range (200