top of page

MY RESEARCH

I am a Cognitive Psychologist researching how the normal aging process affects out cognitive processes. I have a particular interest in understanding how the aging brain affects language production and comprehension.  See the Publication page for links to my published work. 

 

Language and the ageing brain

Most of my research over the last 20 years has focused on understanding how normal aging affects our language abilities. Language is a critical everyday cognitive skill, and one that is influenced by the biological aging process as well as a lifetime of experience.  I use behavioral experiments and neuroimaging techniques to understand why some language abilities decline while others stay the same or even improve with age. See Shafto and Tyler (2014) and Burke and Shafto (2008) for reviews

The input/output asymmetry in language and aging. One of most consistent findings in research on language and aging is that while older adults tend to show declines in language production, they do not show declines in comprehension. This “input/output asymmetry” (MacKay et al., 1999) is a challenge to models of neurocognitive aging, as we must explain both impaired and preserved language ability in the language system where production and comprehension share many underlying processes and representations, and are carried out in the brain in largely overlapping systems. I have demonstrated this asymmetry in a spelling paradigm where older adults are impaired at error correction (output), but not at error detection (input) (Shafto, 2010). This asymmetry highlights that age-related declines have more to do with changes to how we access knowledge than changes to the integrity of underlying knowledge.

 

Language production: Word finding failures. One of the biggest worries of older adults are word finding problems such as “tip of the tongue states” (TOTs).  People fear that these “memory” failures signal general cognitive decline and encroaching dementia, but behavioral evidence suggests that these problems are normal and reflect temporary failures in language production. During production, word meaning (semantics) must be mapped onto word form (phonology) and word finding problems occur when the connections between semantics and phonology become too weak. Our research using structural and functional MRI supports behavioural evidence that age-related declines in neural integrity impairs phonological access (Shafto et al., 2007; Shafto et al., 2010; Stamatakis et al., 2011).  However, we also find evidence that during TOTs, a frontal “salience” network is activated that is associated with general cognitive control and error detection (Shafto et al 2010). Our ongoing research is using behavioral experiments, structural and functional MRI, and MEG to distinguish the temporary language failures characteristic of normal ageing from more general cognitive failures that may be related to pathological aging.

 

Language comprehension: Preservation of automatic processing. Older adults remain preserved at many of the core processes underpinning comprehension, including accessing the meaning of words and processing syntax during online sentence comprehension. Using fMRI, we have explored the possibilities that older adults maintain comprehension by reorganization of the component processes underpinning comprehension (Shafto et al., 2012), or by recruiting additional neural resources, especially from the frontal lobes (Tyler et al., 2010). Recent evidence from research from the Centre for Speech, Language and the Brain (Davis et al., 2014) suggests that despite age-related loss of grey and white matter, core automatic language processing is preserved. This is in keeping with previous behavioral evidence (see Burke and Shafto, 2008 for a review), and in keeping with recent evidence I have found from a proofreading study that older adults are preserved at error detection when processing is automatic, but impaired when it is attentionally-demanding (Shafto et al., 2015).

shafto_2010_fig.jpg
lang_brain.jpg
shafto_2010-Figure4-1.png

Cambridge Centre for Ageing and Neuroscience

In 2010 the Cambridge Centre for Ageing and Neuroscience (Cam-CAN; cam-can.com) began a large-scale, multidisciplinary investigation into the underpinnings of successful cognitive aging. We spent the next five years gathering and processing data from over 3000 participants on hundreds of measures using cognitive tests, background questionnaires, health screening, and a range of neuroimaging measures.  For our protocol paper, see Shafto et al. (2014) and for details of our neuroimaging protocol see Taylor et al (2015).

 

The structure of cognition across the lifespan: variability and diversity. A major aim of the Cam-CAN project is to understand the nature of successful agiing, and as such we have included a range of targeted cognitive experiments, aimed at identifying individual differences amongst healthy participants, and at identifying sources of good performance in old age, not just declines (for an overview of the tasks we use, see Shafto et al 2014). This differs from some other large-scale studies of ageing which are often focused on age-related declines and aim to identify general cognitive factors that decline with age and may be markers for dementia. In a project using our full range of cognitive measures I have demonstrated that lifespan cognitive performance is underpinned by a number of factors, only some of which decline with age. Additionally, the diversity of cognitive responses is maintained across the lifespan (Shafto et al., 2014). 

 

In a related study we find evidence that older adults’ performance is more variable than younger participants (Green et al., 2015). This pattern is found by some researchers but not others, and has been used to suggest that cognitive ageing by its nature involves increased neurocognitive noise. In a follow-up investigation we are asking whether increased variability reflects more noise at the individual level or reflects the divergence between better and worse agers.

“Risk factors” and “protectors” in a cognitively normal sample. A range of researchers from the Cam-CAN project are currently examining the underlying factors that protect good brain and cognitive health across the lifespan, and those that may be risk factors for decline. These projects are still underway (see cam-can.com/publications for recent papers and presentations), but will provide a unique insight into predictors in a healthy, population-derived, adult lifespan cohort. I am involved in a number of projects to examine how factors such as cardiovascular health, hobbies, employment, and sleep patterns can predict multidimensional neurocognitive health.   

 

Edited Image 2016-01-18 19-26-35
green_2015_fig1.png
gadie_etal_2016_fig.png
bottom of page