Motor differences in autism

Researchers reported motor coordination difficulties in autistic children, adolescents, and adults.^[1][4] Meta-analytic evidence indicates that autistic participants, on average, perform differently from non-autistic comparison groups on a wide range of motor measures with findings spanning both gross and fine motor domains.^[1] More recent reviews have reinforced the view that motor concerns are widespread and heterogeneous, affecting balance, postural control, praxis, gait, imitation, and manual dexterity.^[4][5]

In the SPARK study, many autistic children were identified as being at risk for motor impairment, suggesting that these difficulties are not limited to small clinical samples.^[5] While this may be true, the motor profile in autism is not uniform. Some autistic individuals show pronounced difficulties across multiple domains, whereas others show more selective or task-specific differences.^[4][6] This heterogeneity has led many researchers to emphasize careful measurement and individualized interpretation rather than assuming a single motor phenotype for all autistic people.^[4][2]

Motor differences can emerge early in development. Studies of infancy and toddlerhood have described differences in milestone timing, posture, gait, movement quality, and object-related motor behavior in children later identified as autistic or already diagnosed with autism.^[7][8] Reviews of early motor signs have suggested that movement-related features may be observable before, or alongside, more widely recognized social-communication differences.^[8]

These motor differences often continue beyond early childhood. Reviews and empirical studies have reported persistent differences in postural control, motor planning, imitation, coordination, and movement execution across later developmental stages.^[1][4] Older studies using historical subgroup labels, including “Asperger’s syndrome” and “high-functioning autism,” also reported motor planning and execution differences among participants without intellectual disability^[9]. Across development, motor function has also been linked to adaptive behavior, suggesting that movement differences may influence participation in everyday activities, independence, and access to learning opportunities.^[10][11]

Research on sensorimotor control has explored how autistic movement may differ in sensory weighting, motor planning, error correction, and adaptation to changing task demands.^[12][13][14] Rather than treating movement variability only as a deficit marker, this literature often examines variability as a source of information about how sensory input is integrated during action and how motor output is updated over time.^[12][2]

Some studies have suggested that autistic participants may rely differently on proprioceptive and visual information during movement control.^[13][12] There has been research that shows differences in the representation of internal models of action, interpreting their findings as evidence that autistic participants in their sample relied more heavily on proprioceptive feedback during motor learning.^[13] Other work has examined cerebellar and cortical circuitry as possible contributors to motor adaptation and predictive control in autism.^[14]

A study of feeding-related action chains reported reduced anticipatory activation of mouth-opening musculature in autistic children relative to comparison participants, which the authors interpreted as evidence of differences in action planning.^[15] Taken together, these studies have supported growing interest in motor function as an important domain for understanding autistic behavior, development, and support needs.^[12][4]

Summarize Timeline Fact Check

Kinesthetic priors are internal expectations about how movement will unfold, including expectations about timing, trajectory, force, and sensory feedback. These expectations are shaped through repeated action and updated through sensory and kinesthetic feedback. In the movement perspective, differences in how these priors are formed or calibrated may help explain why some autistic people show differences in timing, movement fluidity, and transitions between actions.^[2]

A movement-based framework in which trial-to-trial variability in micro-movements can be analyzed as a stochastic process rather than summarized only by averages across trials.^[2] This approach treats small variations in movement as meaningful information about sensorimotor control. Using a statistical platform based on the two-parameter Gamma family of probability distributions, the authors estimated individualized distributions for velocity-dependent micro-movement measures, including normalized peak velocity and average speed.^[2]

The resulting shape and scale parameters were plotted in a Gamma plane to characterize each participant’s stochastic movement signature and to examine how that signature changed across task conditions and over time.^[2] Within this framework, distributions closer to the Exponential limit of the Gamma family were interpreted as more memoryless, meaning that current movement patterns provided less information for predicting future movement from prior movement history. By contrast, distributions farther from that limit were interpreted as more stable, reliable, and predictive.^[2]

Applying this approach to hand-movement data, many autistic participants in the sample showed micro-movement patterns closer to the Exponential end of the Gamma family than age-matched non-autistic comparison participants.^[2] The authors interpreted this pattern as consistent with differences in the calibration of kinesthetic re-afference and movement-related prediction in the autistic participants studied, rather than as a direct measure of cognitive ability or intention.^[2]

These findings may have implications for volitional motor control, adaptive exploration, and behavioral flexibility.^[2] If movement-related feedback is less predictable or more variable, building stable expectations from prior movement history may be more difficult. This could contribute to greater variability in motor performance across trials and may require more frequent online corrections during movement. In this sense, the framework presents both an empirical statistical method for quantifying micro-movement variability and a broader interpretive account linking those results to kinesthetic priors and motor control in autism.^[2]

Independent motor-control research is consistent with this interpretation. A computational review of autistic motor abilities concluded that motor differences in autism may involve differences in integrating information for efficient motor planning, along with increased variability in sensory inputs and motor outputs.^[16] In a motor-learning study, autistic children showed differences in internal models of action, with stronger-than-typical reliance on the relationship between self-generated motor commands and proprioceptive feedback.^[17] A subsequent study reported increased sensitivity to proprioceptive error and decreased sensitivity to visual error during motor learning in autistic children, suggesting differences in how sensory feedback is weighted when movements are learned and adapted.^[18]

Together, these studies suggest that Torres’s Gamma-plane findings are consistent with a broader motor-learning literature. Although these authors do not necessarily use the phrase “kinesthetic priors,” their findings address closely related mechanisms, including internal models of action, proprioceptive feedback, sensory-error weighting, motor planning, and movement variability. This literature is consistent with the view that movement differences in autism may reflect differences in how the nervous system learns from prior movement experience, predicts the sensory consequences of action, and updates motor control in real time.^{[2][16][17][18]}

Motor demands can influence performance on measures of cognition, language, and adaptive behavior, especially when assessments rely on rapid speech, pointing, writing, fine motor precision, or other time-sensitive forms of response.^[10][11] For this reason some research argues that performance on conventional tests may reflect not only the intended cognitive construct but also the motor and expressive demands built into the assessment format.^[19][20]

Some researchers reported that some autistic children performed more strongly on strength-informed and motor-reduced perceptual tasks than on standard school-based assessments, raising concerns that cognitive abilities may be underestimated when assessment formats require forms of responding that are difficult for the child.^[19] Other researchers likewise argue that limited spoken-language output can obscure receptive abilities and recommended multimodal assessment approaches that allow alternative forms of response.^[20] This literature has contributed to broader calls for assessment practices that distinguish cognitive demands from motor and expressive demands whenever possible.^[19][20][11]

Emerging genetic and neurobiological evidence also supports the importance of distinguishing autistic behavioral presentation from general cognitive ability. Bradley et al. reported that Ptchd1-as knockout mice showed autism-like behavioral features while learning and memory measures and hippocampal synaptic plasticity were preserved.^[21] This dissociation suggests that some autism-relevant mechanisms may be separable from broad cognitive differences, reinforcing the need for assessment approaches that do not interpret reduced or atypical task performance as direct evidence of reduced cognitive ability.^[21]

Emerging genetic and neurobiological evidence also supports the importance of distinguishing autistic behavioral presentation from general cognitive ability. In a Ptchd1-as knockout mouse model, autism-like behavioral features were reported while measures of learning, memory, and hippocampal synaptic plasticity were preserved^[21]. This dissociation suggests that some autism-relevant mechanisms may be separable from broad cognitive differences, reinforcing the need for assessment approaches that do not treat reduced or atypical task performance as direct evidence of reduced cognitive ability^[21].