Neurologic basis of inhibitory deficits in ADHD

Principal Investigator: Stewart Mostofsky

Prevailing hypotheses implicate deficient response inhibition as one of the fundamental features driving the pathophysiology of ADHD. Results from studies of lesions in animals and humans led many investigators to conclude that response inhibition may be localized specifically to ventral prefrontal regions (with a right preponderance) and consequently that ADHD may be the result of developmental abnormalities localized to this region. More recent data from electrophysiology and imaging studies, however, have lent support to a multiple-domain model of response inhibition according to which the specific region of the frontal lobe crucial for response inhibition depends on the nature of the task (skeletomotor, oculomotor, cognitive, socioemotional) being performed. With respect to ADHD, evidence from neurobehavioral and imaging studies has suggested that abnormalities within the frontal lobe may not be restricted to ventral prefrontal regions. Accordingly, we propose to investigate the hypothesis that in ADHD there exist abnormalities of several frontal lobe regions, each of which contributes to one of several parallel deficits in response inhibition, involving skeletomotor, oculomotor, cognitive, and socioemotional domains. Alexander’s model of frontal-subcortical circuits (recently updated by Middleton and Strick to include skeletomotor, oculomotor, dorsolateral prefrontal, anterior cingulate, and medial/lateral orbitofrontal circuits) will serve as a basis for these experiments. We propose to use neurobehavioral testing, anatomic MRI (aMRI), and functional MRI (fMRI) to examine the hypothesis that ADHD-associated abnormalities involve multiple frontal circuits. For the current proposal we will focus on examining neuroanatomic deficits at the cortical level for both pragmatic and theoretical reasons: We are currently better able to measure at the cortical level circuit-specific regions; furthermore, “open-loop” connections that allow for interaction between circuits, thereby creating a mechanism for regulation of complex human behavior, exist primarily at the cortical level. Examination of subcortical contributions (at the level of the basal ganglia, cerebellum and thalamus) is recognized as equally important but deferred until future study.

Specific Aim 1 To test the hypothesis that in children with ADHD, reduced volumes within the frontal lobes, as determined using aMRI, will not be confined to one region. Rather, reduced volumes (gray, white, or total) will be cortically sublocalized within more than one frontal circuit (as above), including: skeletomotor (premotor cortices including supplementary motor area (SMA) and/or ventral premotor (PMv) area); oculomotor (frontal eye field (FEF), supplementary eye field (SEF)); dorsolateral prefrontal (DLPF) cortex (including areas 9 and 46), and medial/lateral orbitofrontal (OF) cortex (including areas 11 and 12).

Specific Aim 2 To test the hypothesis that, compared to control subjects, children with ADHD will show deficits on tasks reflecting different components governing response inhibition, i.e. representing multiple domains of response inhibition (skeletomotor, oculomotor, cognitive, and socioemotional).

Specific Aim 3 To test the hypothesis that for each domain of inhibitory function, a specific frontal-regional contribution is necessary, such that the neural substrate for ADHD-associated deficits in response inhibition is dependent upon the nature of the action being inhibited. To investigate this hypothesis, the brain-behavior relationships underlying response inhibition will be examined using two approaches:
a) correlations of regional frontal lobe volumes (as determined using aMRI) with levels of functional performance on measures of response inhibition (skeletomotor, oculomotor, cognitive, and socioemotional) within each circuit, and
b) fMRI.

Aim 3a Domain specific measures of response inhibition within and across groups of children with ADHD and controls will correlate with volumes of specific frontal lobe regions as measured using anatomic MRI (aMRI), such that:
3a.1 Volumes of skeletomotor regions (SMA and/or PMv) will predict the presence of mirror overflow movements on neurologic examination and performance on measures of motor response inhibition (commission errors on Go/No-go task and echopraxic errors on the conflicting motor response).
3a.2 Volumes of oculomotor regions (FEF/SEF) will predict measures of oculomotor response inhibition (commission errors on oculomotor Go/No-go task, and directional errors on antisaccades).

3a.3 Volumes of DLPF regions will predict a measure of response inhibition during performance of a Stroop-like Go/No-go task in which the inhibition of a habitual, prepotent response is governed by a rule that is held in working memory.
3a.4 Volumes of medial/lateral OF regions will predict measures of socioemotional inhibition (Go/No-go task with reward/response cost and parent/teacher reports of disinhibited behavior).

Aim 3b As a first step towards examining a multiple domain model of response inhibition, fMRI in adult controls will be used to examine the hypothesis that the distinct set of neural substrates contributing to response inhibition are task dependent and that different (but neighboring) regions will be active during skeletomotor and oculomotor Go/No-go tasks. Specifically, activation in the pre-SMA will be associated with occurrence of No-go trials during a simple Go/No-go task; activation in the FEF/SEF will be associated with the occurrence of No-go trials during an oculomotor Go/No-go task. Furthermore, DLPF regions will be recruited during performance of a more complex Go/No-go task in which it is necessary to retain a novel rule that governs inhibition of a habitual prepotent response. These paradigms will then be extended to children with ADHD, beginning with skeletomotor and oculomotor tasks; the hypothesis being that, compared with age-matched controls, differences in activation associated with No-go cues will be seen in the SMA during skeletomotor Go/No-go and in the FEF/SEF during oculomotor Go/No-go.