**3. Compensatory mechanisms of CR and EF**

Traditionally, late-mature regions, such as the frontal lobes, are considered especially vulnerable to normal age changes, inspiring theories of cognitive aging, such as the "last in, first-out" or "retrogenesis" hypothesis. This hypothesis considers an anteroposterior gradient of age vulnerability, which explains the decline in EF often observed in healthy older adults [70].

Executive functions, such as processing speed, working memory, inhibitory control, top–down suppression, or shifting ability, are shaped by education and by other CR proxies. A decline in executive performance has been shown to be associated with low performance in activities of daily living and to predict conversion from MCI to dementia [71, 72]. Moreover, EF are known to be sensitive to damage in other parts of the brain, such as subcortical white matter changes [34], thalamic nuclei, the limbic system, and basal ganglia [73] apart from prefrontal lobe damage.

The perspective that age-related cognitive decline emerges when a person is no longer able to compensate for the reduced functioning of the primary brain structures and circuits, is largely supported in the literature. Relevant conceptual models have emerged over the last 20 years, aiming to describe and understand brain reorganization in response to age-related changes and brain injury. Older adults may use alternative networks to aim for the same level of functioning as younger individuals, which can represent a mechanism of neural compensation [74, 75]. The "Scaffolding Theory of Aging and Cognition" (STAC) model proposed by Park and Reuter-Lorenz [76] claims the recruitment of additional circuits as a way to strengthen the declining structures whose functioning has become inefficient. These strategies lose efficiency in the aged brain and are eventually no longer accessible when there is cerebral pathology, as in the case of AD. The "normalcy-pathology homology" phenomenon suggests that there are regions more vulnerable to age-related changes and that this age vulnerability renders them more susceptible to additional pathological AD-related changes. This is particularly clear in frontotemporal regions where the elderly, even with a low risk of AD, present prominent cortical reductions [70].

The Cognitive Reserve framework suggests that individual differences in cognitive performance are based on more efficient recruitment of brain networks (neural reserve) or the enhanced ability to recruit alternate (compensation) brain networks [15, 77]. Regarding neural reserve, it is postulated that inter-individual variability related to the efficiency, capacity, or flexibility of the brain networks will influence how the healthy brain can deal with the demands imposed by the emergence of brain injuries or pathologies. The neural reserve allows healthy young individuals with greater CR to solve tasks more efficiently and more capably and, in turn, may better confront the disruptions imposed by brain pathology due to the increased flexibility of brain networks. Neural compensation concerns task-related activation, a mechanism that only appears when new resources are needed to maintain or improve performance due to changes in the brain structure. Hence, neural compensation is a mechanism usually referring to people who have brain pathology [15, 77]. The degree of compensation can also vary in individuals in terms of expression and effectiveness. In fact, neural compensation refers to inter-individual variability to compensate for the disruption of standard processing networks. In this situation, brain structures or networks that are not normally used by individuals with intact brains become activated. Both neural reserve and neural compensation support CR, with compensation being the most common mechanism in more advanced stages of the aging spectrum [78].

As previously stated, individuals with higher CR can maintain a more efficient and capable network or compensate advantageously in the face of a comparable amount of brain pathology [79]. Accordingly, Scarmeas et al. [80], using a set of memory tasks, identified brain regions where systematic relationships between CR and brain activation differed as a function of aging. Thus, when facing certain tasks, young and older people activate similar brain regions but as the difficulty of the memory task increased the magnitude of activation was often higher in older individuals, suggesting more efforts to achieve a comparable level of performance, which can be related to network efficiency. In addition, the older adults recruited additional regions of the brain not used by young people while performing certain memory tasks, which can represent a form of active neural compensation [80]. A similar pattern of compensation was also found when comparing old adults schooled later in life with old adults schooled at the proper age, in a memory recognition task using Magnetoencephalography (MEG), and the first ones displayed additional activations to keep the level of performance [81].

#### *The Role of Cognitive Reserve in Executive Functioning and Its Relationship to Cognitive Decline… DOI: http://dx.doi.org/10.5772/intechopen.104646*

In the last few decades, scientific studies have tried to capture the "neural implementation" of CR through functional neuroimaging [78]. This approach seeks to identify resting state or task-related functional activation brain networks that may underlie CR. Potentially, the expression of these networks may be associated with the influence of CR proxies, moderating the effect of brain changes on cognition. If these networks were identified through functional neuroimaging research methods (not properly used in clinical practice), their degree of manifestation would be a more direct measure of CR than other types of proxies [17]. Tucker and Stern [37] suggested that there may be at least one "generic CR network" that can be activated during the performance of many tasks, explaining how CR protects against brain pathology, which seems to be a promising line of future investigation [17].

A recent systematic review indicates that a resting-state network, implicating medial temporal regions and cingulate cortex (anterior or posterior), is associated with neural reserve, whereas frontal regions and the dorsal attentional network (DAN), activated during the cognitive engagement, are related to neural compensation [78]. Task-related studies have found a positive correlation between CR proxies (mostly premorbid IQ and education) and higher frontal activity in healthy older adults compared to young adults [82–85]. Moreover, a positive association between CR (i.e., education-occupation attainment, premorbid IQ, and leisure activities) and frontal activity in MCI and AD patients compared to healthy older adults has been found [86, 87].

The mechanisms on which the function and resilience of large-scale brain networks are based are still poorly understood. Early lifespan environmental influences can contribute to understanding phenomena such as reserve, as, at least partially, to determine the variance of the underlying structural network. This may have implications for global and regional network controllability. A dynamic network theory can be crucial for advancing the understanding of the resilience of the human brain, reinforcing the need for a spatiotemporal analysis in complex systems [88]. In fact, the human capacity to perform a variety of tasks seems to be associated with cognitive control networks, specifically the frontoparietal control network (FPN) in the left posterior parietal cortex. The adaptability of this network, whose global connectivity pattern seems to change more than other networks, and the connectivity patterns that can be used to identify task performance, point to the importance of this network in cognitive control and task performance. It seems to be possible through "flexible hubs," that is, regions that quickly update their connectivity pattern according to task demands [89].

This greater variability in FPN connectivity, both between networks and between tasks, supports the notion that this network implements core flexible hubs, allowing cognitive control across various and distinct tasks [89]. This is especially relevant for this chapter's purpose as the existence of this control network appears to be crucial for reserve. Specifically, one of its hubs, the left frontal cortex (LFC, covering BA 6/44) [90, 91], is a likely candidate for the neural implementation of CR. The resting-state connectivity of that LFC hub region had previously been associated with protective factors such as high IQ and high cognitive performance. Concretely, it had already been demonstrated that the lateral prefrontal cortex (LPFC) is a hub region with an especially high global connectivity but, more than that, it has been shown that this global connectivity could predict the fluid intelligence of individuals, appearing to be a global hub connector [92]. This level of the organization thus appears to be especially relevant for understanding the brain and CR that involve distributed circuits and complex psychological constructs.

Global connectivity of the LFC hub (close to the Broca area), in resting-state fMRI, is associated with more years of education (CR proxy) and with milder effects of FDG-PET hypometabolism on memory performance in prodromal AD [91]. This can be important for instance in the selection of participants for intervention trials since MCI patients with higher CR seem to have a higher likelihood to benefit from a cognitive intervention [93].

Increased frontoparietal activation may reflect a compensatory mechanism, helping to protect memory task performance in early-stage AD. Additionally, increased global connectivity of LFC can support frontoparietal increased activation and that is associated with CR, moderating the association between AD neuropathology and cognitive decline, and helping to maintain better memory performance [90]. In a task-related fMRI study, the authors tried to understand if LFC hub connectivity during an episodic memory task was associated with a reserve in aging and MCI. More years of education were associated with increased LFC connectivity during memory processing, and increased LFC connectivity was associated with a higher reserve in the memory domain. This result pattern was found in controls and MCI groups, which was interpreted as suggestive that connectivity of a key hub of the frontoparietal control network contributes to reserve in both normal and pathological aging. This conclusion reinforces that LFC is a good candidate for the neural basis of reserve and that a higher LFC connectivity may be a long-lasting trait that is influenced by environmental stimuli, namely education [91]. In fact, CR, being the result of multiple and distinct stimulations, involves connectivity between different tasks and domains. Consistently with this, the LFC (BA 6/44) ranks among the top 5% of brain regions in terms of number of connections in the brain, being high and globally connected to the rest of the brain and is a key connector hub between different functional networks [91]. Taken together the results seem to point out that the cognitive control network, particularly LFC, works as a hub of the frontoparietal control network, which is associated with greater reserve. Later work showed that education is associated with better performance on memory tasks thanks to greater efficiency of functional networks, clearly demonstrating the effects of education on DMN/DAN small-worldness, mediated via LFC connectivity, and reinforcing its role as a neural basis of the reserve [94].

Moreover, evidence also shows that education facilitates the brain's ability to form segregated functional groups of networks, with stronger signals in parietal and occipital regions [95]. This fact reinforces the perspective that more years of schooling trigger a more specialized use of neural processing. However, CR (residual variance in memory and general executive functioning) was also associated with higher global network efficiency (i.e., functional integration). In this sense, this study corroborates that CR is associated both with increased functional connectivity and better organization of the network topology.

The protective role of higher global functional connectivity in the FPN and higher local connectivity between the salience network (anterior cingulate cortex) and medial frontal cortex can significantly mitigate the impact of white matter lesions on EF [96], emphasizing the role of the cognitive control network as a neural substrate for CR. As pointed out by the authors, both the salience and the FPN are important cognitive control networks, that are crucial for appropriate brain functioning, with the FPN flexibly regulating the activity of other networks and the salience network integrating inputs from different sources. Their results reinforce the notion that cognitive control networks may play a role in brain resilience mechanisms with increased connectivity being linked to better cognition.

#### *The Role of Cognitive Reserve in Executive Functioning and Its Relationship to Cognitive Decline… DOI: http://dx.doi.org/10.5772/intechopen.104646*

Overall, these findings suggest that greater activity of frontal regions, namely via LFC connectivity, is a potential component of functional networks underlying neural compensation. Conversely, MTL regions, which are known to be critical for the conversion from MCI to AD, may reflect the capacity of the neural reserve [97, 98].
