Periodic Reporting for period 1 - NEURO-AGE (Addressing Neural Declines to Promote Healthy Ageing)
Reporting period: 2017-09-01 to 2019-08-31
The European Union estimates that by 2020, 1 in 5 Europeans will be at least 65 years of age. The increasing number of older adults presents challenges for society that have never been encountered before. Age-related declines in the ability to think, and the ability to control movements, lead to problems with everyday life, the ability of older people to live alone, and their ability to be part of our modern society. Developing approaches to understand and address age-related declines is therefore an important scientific challenge.
The ability to quickly select the correct action to perform is essential to everyday life; for example, when driving a car, if we see a red light, it is essential that we can quickly choose the correct response of applying the brakes. This depends firstly on our ability to learn that we must respond to a red light by stopping, and secondly on our ability to make what we have learned automatic by practicing it. The ability to select responses therefore depends on an important combination of both our ability to think, and to move. Both of these abilities are affected by ageing, making it is important for us to improve our understanding of how we learn to select responses, understand how this is affected by ageing, and identify ways in which we could improve our ability to select appropriate responses.
The overall objectives of this project are to measure the problems that older adults have when attempting to select appropriate responses, to develop new approaches to improve our ability to select responses, to identify the regions of the brain that are involved in action selection tasks, and to learn whether the ability to select responses can be predicted by combining data from performing action selection tasks with data from scans of the brain.
Based on the findings of the project, we are able to conclude that older adults do not slow down to favour accuracy - the slowing of their actions is part of a general age-related reduction in performance, that we can enhance the rate of learning using novel training paradigms developed during this project, and that a cerebellar-cortico network underlies behavioural response selection.
The ability to quickly select the correct action to perform is essential to everyday life; for example, when driving a car, if we see a red light, it is essential that we can quickly choose the correct response of applying the brakes. This depends firstly on our ability to learn that we must respond to a red light by stopping, and secondly on our ability to make what we have learned automatic by practicing it. The ability to select responses therefore depends on an important combination of both our ability to think, and to move. Both of these abilities are affected by ageing, making it is important for us to improve our understanding of how we learn to select responses, understand how this is affected by ageing, and identify ways in which we could improve our ability to select appropriate responses.
The overall objectives of this project are to measure the problems that older adults have when attempting to select appropriate responses, to develop new approaches to improve our ability to select responses, to identify the regions of the brain that are involved in action selection tasks, and to learn whether the ability to select responses can be predicted by combining data from performing action selection tasks with data from scans of the brain.
Based on the findings of the project, we are able to conclude that older adults do not slow down to favour accuracy - the slowing of their actions is part of a general age-related reduction in performance, that we can enhance the rate of learning using novel training paradigms developed during this project, and that a cerebellar-cortico network underlies behavioural response selection.
We first examined how ageing affected performance in a simple response selection task (reaching to different targets). It has been suggested that older adults favour speed over accuracy - that when making decisions, they will wait longer, as they don't want to make mistakes. We measured the minimal time that participants were capable of selecting correct responses by applying time pressure. We compared the minimal time required to make a response under time pressure to the time that participants selected themselves when asked to respond as quickly as possible. Our results showed that people could always make accurate responses with less reaction time than they would select themselves, and that older adults needed more time to respond; however, the delay between when older adults could act, and when they themselves chose to act, was not different to the delay for younger adults. This indicated that it is the physical capabilities of older adults - rather than a change in their attitude - that lead to age-related declines in performance.
We then examined a more complex response selection task (pressing buttons in response to different symbols), asking participants to try to stop themselves from performing 'bad habits' that they had made automatic through practice. We predicted that this would be simple to overcome a bad habit when people had plenty of time, but that applying time pressure would make it more difficult. After training participants for several days, we found that they could easily overcome these bad habits when given plenty of time - at first appearance, it would seem that they did not have a bad habit to overcome. However, when we applied time pressure, the participants found it much more difficult to hide their bad habits. We are in the final stages of examining how this effect changes with age.
We then developed a new approach to speed up the process of learning to select appropriate responses. We compared two groups of participants; one group who sometimes had to select the same response twice in a row, and another group who rarely had to select the same response twice in a row. We found that participants who rarely had to select the same response twice in a row learned the task faster; we believe this is because the constant switching of the required response made the task more challenging.
The final part of the project is examining how the structure of the brain affects our ability to select and perform appropriate responses. Using data from over 300 participants, we separate the brain into 96 different areas, and are developing computer code that can learn how the volume of these different areas relates to the ability to select responses. Using advanced (machine-learning) computing techniques, we are currently developing tools that attempt to predict how well a person is able to select appropriate movements based on the volume of different areas of their brain.
To date work from the project and related research has so far led to three publications in scientific journals, and five presentations at international research meetings.
We then examined a more complex response selection task (pressing buttons in response to different symbols), asking participants to try to stop themselves from performing 'bad habits' that they had made automatic through practice. We predicted that this would be simple to overcome a bad habit when people had plenty of time, but that applying time pressure would make it more difficult. After training participants for several days, we found that they could easily overcome these bad habits when given plenty of time - at first appearance, it would seem that they did not have a bad habit to overcome. However, when we applied time pressure, the participants found it much more difficult to hide their bad habits. We are in the final stages of examining how this effect changes with age.
We then developed a new approach to speed up the process of learning to select appropriate responses. We compared two groups of participants; one group who sometimes had to select the same response twice in a row, and another group who rarely had to select the same response twice in a row. We found that participants who rarely had to select the same response twice in a row learned the task faster; we believe this is because the constant switching of the required response made the task more challenging.
The final part of the project is examining how the structure of the brain affects our ability to select and perform appropriate responses. Using data from over 300 participants, we separate the brain into 96 different areas, and are developing computer code that can learn how the volume of these different areas relates to the ability to select responses. Using advanced (machine-learning) computing techniques, we are currently developing tools that attempt to predict how well a person is able to select appropriate movements based on the volume of different areas of their brain.
To date work from the project and related research has so far led to three publications in scientific journals, and five presentations at international research meetings.
This project has introduced several state-of-the-art developments. First, our approach of making participants respond under time pressure has allowed us to gain a much finer understanding of how people select actions. Secondly, by developing a new approach that combines several different software packages designed to separate the brain into different regions, we were able to improve our undersanding of how both areas of the cortex and cerebellum are involved in response selection. Finally, the use of advanced computing techniques to predict how participants will perform based on measures of the structure of their brain is a cutting-edge approach that goes beyond traditional research approaches. The new insights into response selection, and the approaches and tools developed during the project, will have lasting impact on our understanding of response selection and how response selection is affected by ageing, as well as how we can improve our ability to select appropriate responses.