New perspective on the impact of aging on motor adaptation
Because of increased life expectancy and falling fertility rates, populations are aging worldwide, which puts pressure on our health- and social care system. An aging population is not problematic if the added years are lived in good health. However, a large gap appears to exist between life expectancy and healthy life expectancy. Therefore, healthy aging is an important target for the World Health Organization (WHO). They define healthy aging as developing and maintaining the functional ability that enables well-being in older age. Functional ability is the ability to meet basic needs, to learn, grow and make decisions, to be mobile, to build and maintain relationships, to contribute to society. These abilities require mobility, while movement function appears to decline with aging. An improved knowledge of movement function of a healthy older population can help us to design strategies to improve functional independence of older people and provide us with a reference population for older adults encountering neurological disease affecting movement. For instance, as we age, muscle fiber properties change, bone mass and density falls, articular cartilage undergoes significant changes, joints become more rigid and fragile which results in more frequent falls, muscle strength decrease, movement and gait speed decrease. Beside movement function decline, also the ability to adjust movement to environmental and body changes declines with aging. In Chapter 1, we introduce the concept forward model, which allows predicting sensory outcome of a given movement. This prediction function allows the brain to deal with limitations of the human body, such as slow information transfer, noisy sensory information and slow neural computation. Forward models are only useful when they are accurate in all circumstances, which is enabled through motor adaptation. We discuss motor adaptation, its properties and underlying components. Previous work has shown that motor adaptation reduces with aging, however, several findings were inconclusive and many opportunities for further exploration remain. Therefore, this PhD study aimed to provide further insights in how motor adaptation declines with healthy aging.
Three major hypotheses have been put forward to explain the age-related decline in adaptation: 1) deterioration of internal model recalibration due to age-related cerebellar degeneration (Internal model hypothesis), 2) impairment of the cognitive component of motor adaptation (Strategy hypothesis, and 3) deficit in the retention of the learned movement (Retention hypothesis). In Chapter 2, we systematically investigated these 3 hypotheses in a large sample of older women and men. We demonstrated that age-related deficits in motor adaptation are not due to impaired internal model recalibration or impaired retention of motor memory. Rather, we found that the cognitive component was reduced in older people. Therefore, this first experimental study suggests the interesting possibility that cerebellar-based mechanisms do not deteriorate with age despite cerebellar degeneration. In contrast, internal model recalibration appears to compensate for deficits in the cognitive component of this type of learning.
In Chapter 3, we elaborated on the decline of the cognitive component of motor adaptation. In other motor tasks than motor adaptation, older adults appear to rely on cognition to improve their motor performance. It is unknown why older adults are not able to do so in motor adaptation. In order to solve this apparent contradiction, we proposed the cognitive resources hypothesis. This hypothesis suggested that older adults require more cognitive resources in unperturbed reaching compared to younger adults, which leaves fewer resources available for the cognitive aspect of motor adaptation. Two dual-task experiments were designed to verify this hypothesis. Dual-task costs in the baseline of visuomotor rotation experiments would provide an estimation for the amount of cognitive resources applied during unperturbed reaching. However, we failed to confirm the cognitive resources hypothesis. Instead, we observed that explicit adaptation was associated with visuospatial abilities (visual working memory capacity and visual memory). Therefore, we suggest that the general cognitive status (amount of cognitive resources available) of an individual appears to predict the development of explicit adaptation.
In Chapter 2, we discovered that internal model recalibration remained intact or even increased with aging, when assessed independently. In Chapter 4, we examined the influence of an age-related decline of sensory acuity on the increase of internal model recalibration. One possible explanation was that the reduced sensory acuity results in an increased reliance on predicted sensory feedback. This led to our preregistered hypothesis: we expected a relation between the decline of somatosensation and the increased internal model recalibration with aging. However, we failed to support this hypothesis and we concluded that the increase of internal model recalibration was not linked with an age-related decline of somatosensation. Instead, we suggest that life-long learning from sensory prediction errors increases internal model recalibration with aging.
In this PhD study, we took a hypothesis-driven approach with careful experimental testing according to the best possible research practices. This approach allowed us to conclude that the cognitive component of motor adaptation was causing the decline of adaptation with aging. On the other hand, an increased recalibration of the internal model could compensate for this reduced cognitive component. The reduced cognitive component was not linked to an increased amount of resources required for unperturbed reaching in elderly. Instead, general cognitive status appears a better predictor for the development of explicit strategy. Finally, increased internal model recalibration was unrelated to age-related declines of somatosensation. Future research should investigate whether these insights can be generalized to other aspects of motor learning and how we can implement these findings in new or improved rehabilitation strategies for elderly patients with or without neurodegenerative disease.