Processing resources and executive functions

One of the questions examined in the present studies is how typical ageing affects

fundamental cognitive processes in the verbal and visuospatial domain, including processing speed and short-term and working memory capacity for verbal and visuospatial information, as well as inhibitory control and mental flexibility.

Our results revealed divergent patterns of ageing effects on processing resources, depending on the domain examined. Overall, we found that increasing age was associated with slower speed of information processing, in accordance with previous reports of age- related impairments in processing speed (Deary, Johnson, & Starr, 2010; Salthouse, 2000). However, age explained a larger proportion of the variance in processing speed for visual information (20%) than for verbal information (9%). Processing speed of visual information decreased as early as from the mid-50s whereas the speed of processing verbal information decreased approximately 20 years later, from the mid-70s. This prominent differential onset and rate of decreased processing speed emerging from our data is line with previous reports supporting that age-related slowing in processing information is more pronounced on

visuospatial tasks than verbal tasks (Jenkins et al., 2000; Lawrence, Myerson, & Hale, 1998). Even greater domain-specific discrepancies were observed in short-term and working memory capacity. On the one hand, short-term memory capacity for verbal information (i.e., strings of digits) peaked at mid-40s and then mildly dropped in the mid-50s to levels similar to those observed in individuals aged between 18-28 and remained unchanged thereafter throughout the lifespan, while the lifespan change in verbal working memory capacity was particularly subtle. On the other hand, short-term memory capacity for visual information (i.e., visual patterns) declined sharply from the mid-40s and continued to gradually decline until late adulthood, while short-term and working memory capacity for visuospatial information (i.e., spatial locations) declined linearly across the adult lifespan. These results

137 provide striking evidence of modular and asynchronous cognitive changes across the lifespan, with cognitive resources for processing verbal information remaining relatively resilient to ageing effects whilst cognitive resources for processing visual and visuospatial information being particularly susceptible to ageing effects.

Past research has shown that age-related decrease in short-term and working memory capacity is greater for visuospatial compared to verbal material (Bopp & Verhaeghen, 2007; Dolman, Roy, Dimeck, & Hall, 2000; Hale et al., 2011; Jenkins et al., 2000; Leonards, Ibanez, & Giannakopoulos, 2002; Myerson, Emery, White, & Hale, 2003; Myerson, Hale, Rhee, & Jenkins, 1999), although there is also evidence against differential ageing

trajectories for verbal and visuospatial working memory (Borella et al., 2008; Park et al., 2002). Procedural and sampling differences may be partially responsible for the inconsistent results of previous reports. Nevertheless, our results revealed clear differential ageing effects on memory for verbal (i.e., strings of digits), visual (i.e., visual patterns), and spatial (i.e., series of locations) material, highlighting the domain-specificity of the processing resources across the adult lifespan.

However, it should be noted that the tasks used in the present study (i.e., short-storage and reordering spans) emphasize rote maintenance without requiring high-demand processing operations of cognitive control. Past research on ageing effects on verbal working memory indicates that there are small age differences in tasks requiring passive storage and

maintenance of information for a short period of time, whereas age differences are more pronounced in more demanding and complex tasks that require additional processing and manipulation of information (Bopp & Verhaeghen, 2007; Vecchi, Richardson, & Cavallini, 2005; for a meta-analysis see Bopp & Verhaeghen, 2005), like tasks that include distracters (Hartman, Dumas, & Nielson, 2001). Older adults seem to engage more neural resources and additional executive processes to accomplish simple computational goals that are typically completed with fewer resources by younger adults (Reuter-Lorenz & Cappell, 2008). In cases of simple verbal working memory tasks, older adults exhibit increased recruitment of

prefrontal areas, which is postulated to reflect a functional compensation mechanism that allows older individuals to achieve equivalent memory performance with younger adults (Cappell, Gmeindl, & Reuter-Lorenz, 2010). Moreover, according to previous factor analytical and structural equation modelling results, verbal and visuospatial short-term memory are dissociable constructs, reflecting that the ability to maintain information for a limited period of time is domain-specific, whereas the ability to maintain either verbal or

138 visuospatial information whilst performing demanding processing operations reflects a more domain-general factor for working memory capacity (Kane et al., 2004).

Another dimension of cognition examined in the present study is processing that requires increased putative executive operations, including mental flexibility (using the Trail Making test) and inhibitory control (using the Stroop task). Our results showed a steep impairment in inhibitory control from the mid-40s followed by gradual declines thereafter until late adulthood. These findings corroborate previous reports revealing the detrimental ageing effects on performance in various versions of the Stroop task (Bugg, DeLosh, Davalos, & Davis, 2009; Van Boxtel et al., 2001; Zalonis et al., 2009). A later onset of age- related decline was observed for set-shifting skills, with a mild impairment emerging from the mid-50s and a sharper decline from the mid-60s until late adulthood. Several studies have also reported that mental flexibility, as assessed by the TMT, declines with increasing age (e.g., Amodio et al., 2002; Hamdan & Hamdan, 2009; Oosterman et al., 2010; Wecker, Kramer, Hallam, & Delis, 2005; Zalonis et al., 2008), with a more accentuated decline from the from the mid-60s onwards (e.g., Perianez et al., 2007; Tombaugh, 2004).

Taken together, the results of the present study indicate that different components of processing resources (processing speed, rote maintenance, and executive control functions) start to change at different timepoints across the lifespan and at different rates, highlighting the importance of recognizing the diversity of processing resources in cognitive ageing. Moreover, our data seem to be incongruent with a common factor theory of ageing and with the existence of an indistinct verbal and visuospatial pool of resources in late adulthood (e.g., Salthouse, 1995). Instead, our results provide evidence for divergent domain-specific ageing trajectories for verbal and visuospatial processing capacities and echo that basic cognitive mechanisms in the nonverbal domain are disproportionately impaired by increasing age.