Neuronal Intra-Individual Variability Masks Response Selection Differences between ADHD Subtypes-A Need to Change Perspectives

Abstract

Due to the high intra-individual variability in attention deficit/hyperactivity disorder (ADHD), there may be considerable bias in knowledge about altered neurophysiological processes underlying executive dysfunctions in patients with different ADHD subtypes. When aiming to establish dimensional cognitive-neurophysiological constructs representing symptoms of ADHD as suggested by the initiative for Research Domain Criteria, it is crucial to consider such processes independent of variability. We examined patients with the predominantly inattentive subtype (attention deficit disorder, ADD) and the combined subtype of ADHD (ADHD-C) in a flanker task measuring conflict control. Groups were matched for task performance. Besides using classic event-related potential (ERP) techniques and source localization, neurophysiological data was also analyzed using residue iteration decomposition (RIDE) to statistically account for intra-individual variability and S-LORETA to estimate the sources of the activations. The analysis of classic ERPs related to conflict monitoring revealed no differences between patients with ADD and ADHD-C. When individual variability was accounted for, clear differences became apparent in the RIDE C-cluster (analog to the P3 ERP-component). While patients with ADD distinguished between compatible and incompatible flanker trials early on, patients with ADHD-C seemed to employ more cognitive resources overall. These differences are reflected in inferior parietal areas. The study demonstrates differences in neuronal mechanisms related to response selection processes between ADD and ADHD-C which, according to source localization, arise from the inferior parietal cortex. Importantly, these differences could only be detected when accounting for intra-individual variability. The results imply that it is very likely that differences in neurophysiological processes between ADHD subtypes are underestimated and have not been recognized because intra-individual variability in neurophysiological data has not sufficiently been taken into account.

Keywords: ADD; ADHD-C; conflict processing; event-related potentials; residue iteration decomposition.

Figure 1

Stimuli and timings of a flanker trial. Compatible or incompatible flanker arrows are presented 200 ms before the target stimulus. Participants are required to indicate the direction of the target stimulus by pressing a button accordingly. The response-stimulus interval (RSI) varies between 1400 ms and 1800 ms.

Figure 2

(A) Stimulus-locked waveforms (current source density, CSD) for P1_Flanker, N1_Flanker, P1_Target, N1_Target components and (B) the equivalent residue iteration decomposition (RIDE)-components (S-Cluster), depicted for both attention deficit/hyperactivity disorder (ADHD) subtypes and for compatible and incompatible flanker trials at electrodes P9/P10. Point 0 denotes the onset of the target stimulus.

Figure 3

(A) Stimulus-locked waveforms (CSD) for the N2 component and (B) the equivalent RIDE-components in the S-Cluster) and (C) the C-Cluster, depicted for both ADHD subtypes and for compatible and incompatible flanker trials at electrodes Cz/FCz. Point 0 denotes the onset of the target stimulus.

Figure 4

(A) Stimulus-locked waveforms (CSD) for the P3 component and (B) the equivalent RIDE-component, depicted for both ADHD subtypes and for compatible and incompatible flanker trials at electrode Pz. Point 0 denotes the onset of the target stimulus. Topographic maps are also shown—blue denotes negative deflections whereas red reflects positive ones.

Figure 5

R-Cluster RIDE-component depicted for both ADHD subtypes and for compatible and incompatible flanker trials at electrodes C3/C4. Point 0 denotes the onset of the target stimulus.