Atypically diffuse functional connectivity between caudate nuclei and cerebral cortex in autism

Abstract

Background: Autism is a neurodevelopmental disorder affecting sociocommunicative behavior, but also sensorimotor skill learning, oculomotor control, and executive functioning. Some of these impairments may be related to abnormalities of the caudate nuclei, which have been reported for autism.

Methods: Our sample was comprised of 8 high-functioning males with autism and 8 handedness, sex, and age-matched controls. Subjects underwent functional MRI scanning during performance on simple visuomotor coordination tasks. Functional connectivity MRI (fcMRI) effects were identified as interregional blood oxygenation level dependent (BOLD) signal cross-correlation, using the caudate nuclei as seed volumes.

Results: In the control group, fcMRI effects were found in circuits with known participation of the caudate nuclei (associative, orbitofrontal, oculomotor, motor circuits). Although in the autism group fcMRI effects within these circuits were less pronounced or absent, autistic subjects showed diffusely increased connectivity mostly in pericentral regions, but also in brain areas outside expected anatomical circuits (such as visual cortex).

Conclusion: These atypical connectivity patterns may be linked to developmental brain growth disturbances recently reported in autism and suggest inefficiently organized functional connectivity between caudate nuclei and cerebral cortex, potentially accounting for stereotypic behaviors and executive impairments.

Figure 1

Functional connectivity MRI effects. Functional connectivity clusters for caudate seed volumes (p < .05, corr.), overlaid onto standard MRI renderings [108] and arranged by known anatomical networks with caudate participation. Associative circuit (A). In the control group, functional connectivity for left caudate seed volumes is seen in middle and inferior prefrontal cortex bilaterally, with a few small clusters of fcMRI effects for the autism group visible in neighboring frontal regions. Direct group comparison (right) shows greater fcMRI effects in controls in the frontopolar portions of area 10 in the right hemisphere, but inverse effects (autism > control) in more posterior portions on the border of areas 9 and 10. Lateral Orbitofrontal circuit (B). FcMRI effects associated with the lateral orbitofrontal circuit (LOF) are only seen in the within-group analyses. For the bilateral caudate seed, the control group shows fcMRI effects in the superior temporal gyrus. A corresponding effect in the autism group is only seen for the left caudate seed, with an fcMRI cluster extending from area 22 into the posterior insula. Oculomotor circuit (C). The control group shows fcMRI effects for the bilateral caudate seed in close vicinity of the frontal eye fields (as expected based on published stereotactic coordinates; see main text). FCMRI effects for the autism group are only seen in distal portions of area 6 and in area 7 (precuneus). Motor circuit(D). Within-group analysis for the bilateral caudate seed shows a single cluster for the control group in left premotor area 6. However, on direct group comparison, numerous clusters showing greater fcMRI effects in the autism group are seen in pericentral and premotor cortices. Comparison with right-handed subsamples (E). Comparison between analyses for full samples (n = 8) and right-handed subsamples (n = 5) in control (red) and autism groups (blue) and overlap (yellow), showing largely consistent effects. A threshold of p = .00025 (uncorr.) was chosen for best combined visibility of effects on both analyses.

Figure 2

Overview of direct group comparisons. Each bar represents the total volume of effects (in μl) on group comparisons per subdivision of the brain (corresponding to the subdivisions used in Tables 1–3 ' [see additional file 1]'), shown separately for different seed volumes. Hemispheres of each brain subdivision are indicated by letters (L = left; R = right; B = bilateral). Upward bars show total fcMRI effects that were significantly stronger in the autism compared to the control group; downward bars show inverse effects (greater in control group). Note that for the right frontal lobe both upward and downward bars are shown because clusters of effects in both directions (autism > control; control > autism) were found in this part of the brain. No bars are shown for subdivisions without significant group differences.