Anomalous Patterns of Inferior Parietal Lobule Functional Connectivity and Handwriting Impairment in Children with Autism

Background: Children with Autism Spectrum Disorder (ASD) exhibit handwriting impairments, particularly in letter formation. Although multiple studies have examined neural networks involved in handwriting in neurotypical children and adults, the neural features of handwriting in children with ASD are unknown. The inferior parietal lobule (IPL) is a multi-modal association cortex hub that is heavily involved in skilled motor behavior and handwriting. Differences in functional connectivity of the IPL may account for handwriting impairments in ASD.

Objectives: To examine differences in intrinsic functional connectivity of the IPL, as it relates to handwriting letter formation in typically developing (TD) children and children with ASD.

Methods: Resting-state fMRI data was acquired from seventy-six children (38 ASD) aged 8-12 years. Groups were matched on age, gender and framewise displacement. Masked independent components analysis (ICA) estimated functional connectivity of the IPL, separately for left and right hemisphere (5 independent components/hemisphere), followed by dual regression to estimate whole-brain connectivity maps for each IPL subnetwork. Outside of the scanner, participants completed three conditions of the Minnesota Handwriting Assessment (MHA), on a digitizing tablet. Participants first copied a sentence (Copy), traced the same sentence at a comfortable speed (Trace), and lastly, traced the same sentence as fast as possible (Fast Trace, FT). Letter-form was assessed using Large Deformation Diffeomorphic Metric Mapping (LDDMM) to quantify a participant’s deviation from a template, where high form scores indicated poor performance. MHA measures were included as covariates in a second-level GLM of dual-regression whole-brain maps, with significant results meeting FWE cluster-correction (p<.05).

Results: A significant effect of diagnosis was observed for FT, such that children with ASD demonstrated worse letter formation than their TD peers (F(1,73) = 7.33, p<.01). There was no effect of diagnosis for Copy or Trace conditions. IPL functional connectivity analyses revealed a significant interaction effect of FT letter formation with diagnosis (ASD vs. TD) on connectivity of the posterior supramarginal gyrus (pSMG) component bilaterally: While better letter formation was associated with greater pSMG connectivity in children with ASD, for TD children better letter formation was associated with lesser pSMG connectivity. Specifically, the diagnosis x FT interaction was observed for: a) left pSMG connectivity with left precuneus, left dorsolateral prefrontal cortex (DLPFC) and bilateral occipital fusiform gyrus, and b) right pSMG connectivity with the right superior lateral occipital gyrus.

Conclusions: The interaction of connectivity between diagnosis and FT letter formation highlights the known role of the pSMG in the cognitive-motor control of everyday goal-directed skilled actions, including handwriting. The autism-specific association of greater pSMG connectivity with better FT letter formation may reflect increased necessity to recruit higher-order cognitive-motor control during speeded tracing of overlearned letters, which taxes internal model output. In contrast, the detrimental impact of pSMG network recruitment for TD children suggests that these children may do best when relying on more efficient primary motor circuitry to guide speeded tracing of highly practiced orthography. These results indicate a promising first step in understanding the neural circuitry underpinning handwriting: a critical developmental skill that is often impaired in ASD.