Associations Between Handwriting Fluency and Motor Control in Children with Autism

Background: Children with ASD experience difficulty performing a host of skilled motor behaviors, including handwriting. Children develop automaticity in handwriting which reduces interference of motor demands with higher-order cognitive processes related to composition. Therefore, the dynamics of writing account for a large portion of variance in composition fluency.

Objectives:  To use advanced mathematical modeling to compare handwriting performance under different demands and relate handwriting to measures of motor and cognitive function.

Methods:  Writing performance was assessed using a digitizing tablet (Wacom Intuos4) in 56 children, 20 with ASD (17M and 3F) and 36 typically developing (TD; 29M and 7F), ages 8-12 years. Groups were balanced for age, gender, IQ (PRI), and SES. Each participant completed a handwriting task in which they copied and traced 6 non-latin alphabet characters. Letter speed and form, as well as speed inflections were extracted using Matlab. Letter Form was analyzed using LDDMM to map shape morphometry. Spectral power (SP) was also assessed using Fourier transform on the velocity curve for each letter.

Results:  Repeated measures ANOVA revealed a significant effect of diagnostic group for Letter Form (F=7.88, p=.007), Speed (F=4.65, p=.036), and Speed Variability (F=13.17, p=.001), with TD performing better/slower than ASD for all measures. An effect of condition (better when tracing) was significant across diagnostic groups for Letter Form, Speed, and Speed Variability (F>29.28, p<.001). A significant groupxcondition interaction effect was only seen for Letter Speed (F=4.65, p=.036). Post-hoc tests revealed significant group (TD>ASD) differences in both conditions for Letter Form (Copy: F=3.89, p=.05; Trace: F=13.78, p<.001) and Speed Variability (Copy: F=6.07, p=.017; Trace: F =10.64, p=.002). Letter Speed showed group differences (ASD>TD) only in the Copy condition (Copy: F=6.71, p=.012; Trace: F=.351, p=.556).  SP analysis revealed significant increases in 4 and 5Hz frequency bins for children with ASD in the Copy condition only (4Hz: F=12.53, p =0.001; 5Hz: F=8.84, p=.005). Pearson’s correlations revealed significant correlations between Letter Form and Working Memory Index (WISC-IV) in the ASD group, particularly during the Copy condition (Copy: r=-0.72, p<0.001; Trace: r=-0.45, p=0.046). Four hertz SP was significantly correlated with mABC-2 Aiming and Catching standard score (r=-.58 p=0.008) during the copy condition. Significant correlations were not observed for the TD group.

Conclusions: These results provide further evidence that children with ASD have substantial difficulties with handwriting, manifested as poor letter formation and fluency. Tracing resulted in improvements in letter formation, speed, and variability in both groups, and was particularly advantageous for normalizing speed in children with ASD. Analysis of speed variability revealed that children with ASD showed increased 4-5Hz periodicity when copying, which has previously been associated with cerebellar impairments. For children with ASD, letter formation was associated with working memory particularly during copying suggesting that difficulty maintaining the visual representation of the letter during motor preparation may contribute inaccuracies in the formation of letters. Further, increased 4-5Hz periodicity was associated with poorer aiming and catching in children with ASD, suggesting that impairments in visual-motor integration, perhaps at the level of the cerebellum, contribute to poor writing fluency.