Age related prefrontal compensatory mechanisms for inhibitory control in the antisaccade task

Highlights

• Inhibitory control age-related changes can be measured within the oculomotor saccade system.

• Aging results in increased reaction time that correlates negatively with BOLD signal across the saccade system.

• The older group shows frontal pole recruitment whose activity also correlates with reaction time.

• Dorsolateral prefrontal cortex activation in the older group negatively correlates with the number of direction errors.

Abstract

Cognitive decline during aging includes impairments in frontal executive functions like reduced inhibitory control. However, decline is not uniform across the population, suggesting individual brain response variability to the aging process. Here we tested the hypothesis, within the oculomotor system, that older adults compensate for age-related neural alterations by changing neural activation levels of the oculomotor areas, or even by recruiting additional areas to assist with cognitive performance. We established that the observed changes had to be related to better cognitive performance to be considered as compensatory. To probe this hypothesis we used the antisaccade paradigm and analyzed the effect of aging on brain activations during the inhibition of prepotent responses to visual stimuli. While undergoing a fMRI scan with concurrent eye tracking, 25 young adults (21.7 y/o ± 1.9 SDM) and 25 cognitively normal older adults (66.2 y/o ± 9.8 SDM) performed an interleaved pro/antisaccade task consisting of a preparatory stage and an execution stage. Compared to young adults, older participants showed a larger increase in antisaccade reaction times, while also generating more antisaccade direction errors. BOLD signal analyses during the preparatory stage, when response inhibition processes are established to prevent an automatic response, showed decreased activations in the anterior cingulate and the supplementary eye fields in the older group. Moreover, older adults also showed additional recruitment of the frontal pole not seen in the younger group, and larger activations in the dorsolateral prefrontal cortex during antisaccade preparation. Additional analyses to address the performance variability in the older group showed distinct behavioral-BOLD signal correlations. Larger activations in the saccade network, including the frontal pole, positively correlated with faster antisaccade reaction times, suggesting a functional recruitment of this area. However, only the activation in the dorsolateral prefrontal cortex during the antisaccade events showed a negative correlation with the number of errors across older adults. These findings support the presence of two dissociable age-related plastic mechanisms that result in different behavioral outcomes. One related to the additional recruitment of neural resources within anterior pole to facilitate modulation of cognitive responses like faster antisaccade reaction times, and another related to increased activation of the dorsolateral prefrontal cortex resulting in a better inhibitory control in aging.

Introduction

There is ample evidence that cognition declines during aging (Salthouse, 2010). This decline, however, is not uniform across the population; some healthy individuals maintain constant cognitive levels across aging, while others show cognitive decay despite being otherwise healthy (Nyberg et al., 2012, Rapp and Amaral, 1992). A range of possibilities could explain this variability, including the preservation of a healthy brain, or the implementation of compensatory mechanisms that could counteract age related brain deterioration (Morcom and Johnson, 2015). Some of the most investigated mechanisms proposed to help maintain the cognitive performance in the elderly emerged from results obtained from imaging studies. These studies have shown significant changes in activation levels in areas known to participate in specific cognitive processes (Cabeza, 2002, Eyler et al., 2011, Reuter-Lorenz and Cappell, 2008). Furthermore, they have even shown the functional recruitment of additional regions not previously associated with the performance of a given cognitive task (Eyler et al., 2011, Reuter-Lorenz and Cappell, 2008). However, more information is needed to establish the specific impact that the observed changes could have on different cognitive processes, including executive functions (Spreng et al., 2010).

A well suited task to analyze the effects of aging on executive functions is the antisaccade task (Guitton et al., 1985, Hallett, 1978). To complete an antisaccade correctly, participants must first inhibit an automatic eye movement towards the presentation of a suddenly appearing peripheral stimulus, and then initiate a voluntary saccade in the opposite direction. Areas involved in this process include the dorsolateral prefrontal cortex (DLPFC) (Guitton et al., 1985, Pierrot-Deseilligny et al., 1991), the frontal, supplementary and parietal eye fields (FEF, SEF, PEF respectively) (Brown et al., 2007, Connolly et al., 2002, Curtis and D'Esposito, 2003, DeSouza et al., 2003, Ford et al., 2005), and the basal ganglia (Watanabe and Munoz, 2011). These brain imaging studies have demonstrated that a network of cortical and subcortical structures must be activated prior to the appearance of the visual stimulus so the motor system can then generate the appropriate action, a phenomenon referred to as ‘task set’ or ‘preparatory set’ (Munoz and Everling, 2004). The addition of preparatory trials to the antisaccade task allows separate examination of activation related to the preparatory stage including the inhibitory control mechanisms for a pro or an antisaccade, compared to activation related to the execution of the response (Alahyane et al., 2014, Cameron et al., 2012, Witiuk et al., 2014).

Age related deterioration of executive functioning includes deficits in inhibitory control (Eyler et al., 2011), as suggested by previous antisaccade studies showing that inhibitory response performance of elderly participants is considerably more impaired than automatic prosaccade responses (Abel and Douglas, 2007, Abrams et al., 1998, Munoz et al., 1998, Peltsch et al., 2011, Peltsch et al., 2014, Yang and Kapoula, 2006). A number of imaging studies have further demonstrated the effect of aging on the response inhibition in the saccade system (Alichniewicz et al., 2013, Mirsky et al., 2013, Nelles et al., 2009, Pa et al., 2014, Raemaekers et al., 2006). These studies showed that age related activation increases are evident throughout the frontoparietal network during saccade performance in the elderly (Nelles et al., 2009), and particularly during antisaccade trials, suggesting compensatory processes (Raemaekers et al., 2006). However, the possible behavioral correlates of the observed activation changes in the saccade network have not been studied. That is, it is not known if activity-changes including increase, decrease, lateral or antero-posterior activation shifts in saccade related areas, or even the additional recruitment of new areas to the oculomotor circuity, correlate with a better behavioral performance, or if those changes are irrelevant for inhibitory control or saccade execution in the elderly.

Here, our first goal was to test if older adults showed local BOLD activation changes primarily during the preparatory stage of the antisaccade task, when the response inhibition process is established (Alahyane et al., 2014, Munoz and Everling, 2004). Once we obtained the activation differences between the young and the older participants, our second goal was to test the hypothesis that those changes fit mechanisms suggested as compensatory in the elderly. To evaluate this hypothesis we analyzed the areas with age related changes and tested if the changes in activation correlated with the participants’ performance, specifically with reaction times and error responses. If those changes showed no correlations, or even correlations with behavioral deficits, that would suggest a maladaptive plastic response. On the contrary, if the changes showed correlations with better behavioral outcomes, it would be suggestive of beneficial aging compensatory mechanisms.