Procedural Learning Differs In Autism, Parkinson’s, And Other Disorders

We are surrounded by information that only makes sense when presented in a specific order. Many fundamental skills are reliant on learning sequences of information. When learning a motor skill such as kicking a ball, we need to carry out a series of movements in a sequential order, so that our foot contacts the ball at the right time, with the right force, and so that we remain standing at the end.

When learning our first language, we need to pick up on the way that the sounds we hear follow sequences to form words and sentences. For both the motor and non-motor skills, repeated practice or exposure to the information is often required for learning to take place. This type of learning is undertaken by the “procedural memory system.”

What happens when this memory system is compromised? One symptom of abnormalities in the procedural memory system is motor control problems. This is perhaps most clearly demonstrated by the experience of those with Parkinson’s disease. Parkinson’s disease involves degeneration of cells within the basal ganglia – a core structure of the procedural memory system –  causing the well-known difficulties in controlling body movements.

In recent years, a range of neurodevelopmental disorders have also been associated with procedural memory dysfunction. These include disorders of motor coordination (developmental coordination disorder), reading (dyslexia), language (specific language impairment), as well as schizophrenia. Procedural memory has also been studied in individuals with autism spectrum disorder, though seems to be unaffected in this group. The core symptoms of each disorder are different, and the part of the brain that is affected in each disorder is also different, yet each disorder may be associated with some dysfunction of the procedural memory system.

The majority of studies that have investigated procedural memory functioning in disorders have used the same behavioral task to assess procedural memory. Furthermore, each disorder has had the relevant procedural memory literature statistically synthesized using meta-analysis. In each meta-analysis, findings are pooled from all available studies that compare procedural memory performance in a disordered group to an age-matched control group. This provides an indication of the overall averaged result. For example, after pooling results from 27 studies, it was shown that people with Parkinson’s disease perform about half a standard deviation worse on a procedural memory task than their peers.

We aimed to determine whether the severity of procedural memory deficits is different depending on the disorder.  Due to the known motor-related function of the procedural memory system, we hypothesized that disorders with a core motor control problem – Parkinson’s disease and developmental coordination disorder – would be associated with the largest procedural memory deficits.

The presence of several meta-analyses investigating performance on the same task provided us the rare opportunity to compare procedural memory performance across a range of disorders using ‘second-order meta-analysis’. An advantage of using this technique is that, because each meta-analysis pools together multiple studies, we can compare large groups of matched participants. Our analysis summarised data from 1412 individuals with a disease/disorder and 1415 age-matched control participants.

Results revealed that there was a difference in procedural memory functioning between disorders, but this difference was only attributable to those with autism spectrum disorder. The other five disorders (Parkinson’s disease, schizophrenia, dyslexia, developmental coordination disorder, and specific language impairment) were statistically indistinguishable in terms of the size of the procedural memory problem in comparison to their own control groups. Interestingly, this shows that the clear neuropathology of Parkinson’s disease leads to the same level of procedural memory problem as disorders with much less pronounced neural abnormalities. This also highlights that behavioral measures do not provide information about which part of the brain is affecting procedural memory performance.

Results also showed that individuals with autism spectrum disorder perform at a similar level to their age-matched peers. There was even a trend for those with autism to perform slightly better than their peers on the procedural memory task.  We suggest that the better performance in autism spectrum disorder in comparison to the other disorders is due to over-activity in the neural circuits that underlie the procedural memory system. This means that while each disorder may have some level of abnormal procedural memory functioning, this can lead to either deficient or superior procedural memory. Other disorders which are also associated with over-active procedural memory circuitry, such as Tourette’s syndrome and obsessive-compulsive disorder, may also show superior performance on procedural memory tasks.

Overall, our analyses demonstrate that procedural memory is affected in a range of neurodevelopmental disorders. Importantly, it was also shown that not all disorders are associated with poor procedural memory. Avenues for future research include investigating a wider range of disorders to determine whether superior procedural memory is common to several groups. Coupling the procedural memory task with neuroimaging techniques, in order to understand how such a diverse range of disorders are associated with one of the same underlying problems, is also required.

These findings are described in the article entitled Procedural learning in Parkinson’s disease, specific language impairment, dyslexia, schizophrenia, developmental coordination disorder, and autism spectrum disorders: A second-order meta-analysis, recently published in the journal Brain and Cognition. This work was conducted by Gillian M. Clark and Jarrad A. G. Lum from Deakin University.