How Can Cognitive Reserve Promote Cognitive and Neurobehavioral Health?

Abstract

Objective

This review is aimed at understanding how cognitive reserve and related concepts contribute to promoting neurobehavioral and cognitive health, consistent with goal of the 2020 national academy of neuropsychology (NAN) Annual Meeting.

Research indicates that lifestyle factors such as achieving educational and work milestones, participating in leisure and social activities and IQ are all associated with reduced risk of cognitive decline in normal aging and of developing dementia. Many of these lifestyle factors have also been associated with better cognition in other psychiatric and neurological conditions. The cognitive reserve hypothesis posits that these lifestyle factors result in individual differences in the flexibility and adaptability of brain networks that may allow some people to cope better than others with age- or dementia-related brain changes. Recent evidence also supports the idea that specific genetic and lifestyle factors may help preserve a healthy brain or enhance brain reserve, a process that has been called brain maintenance. The complementary concept of brain reserve posits that structural brain features can guard against dementia and related conditions. This review defines these theoretical concepts, their research basis, how they are studied and their clinical applications.

Conclusion

Evidence supports the concept of reserve, which can be influenced by experiences in every stage of life. Focused research in this area can maximize the chance for successful intervention.

Introduction

The concept of reserve (CR) was developed to account for the disjunction between the degree of brain damage and the clinical manifestation of that damage. It is a common clinical observation that two people can sustain what appears to be the same amount of damage but have widely different clinical presentations. In this paper, I will briefly review several concepts associated with reserve and resilience, discuss some of the epidemiologic evidence for cognitive reserve and consider imaging studies aimed at understanding the neural instantiation of reserve. Because I focus mainly on aging and dementia, I will talk about the applicability of the concept of reserve to diseases that affects cognition, and its applicability to clinical practice.

Defining Brain Reserve, Brain Maintenance and Cognitive Reserve

There are three terms that are used in the study of reserve-related concepts: brain reserve (BR), brain maintenance (BM) and cognitive reserve. The definitions used here rely on a Whitepaper that summarizes consensus definitions across a group of 31 investigators in the field (Stern et al., 2020) and are presented here in the order that they were developed.

Brain reserve was the first concept to be proposed. The term was used by Katzman and colleagues in a paper describing 10 women who were followed in a longitudinal aging study (Katzman et al., 1989). These women were cognitively normal at death but were found to have a large degree of amyloid plaque at postmortem. The authors noted that they had larger brains than usual and speculated that these individuals “started with larger brains and more large neurons and thus might be said to have had a greater reserve.” A similar observation was made even earlier in the seminal work by Blessed, Tomlinson, and Roth (1968). Brain reserve is conceived as neurobiological capital (e.g., numbers of neurons and synapses). Brain reserve implies that individual variation in the structural characteristics of the brain allows some people to better cope with brain aging and pathology than others before clinical or cognitive changes emerge. It has been characterized as a “passive” form of reserve in that it does not actively combat the effects of pathology. However, if you have more of it, you can lose more before reaching a threshold where clinical changes occur.

I suggested the term CR to contrast it to BR (Stern, 2002). Concept of reserve is defined as the adaptability of cognitive processes that helps explain the differential susceptibility of cognitive abilities or day-to-day function to brain aging pathology or insult. Thus, the concept of CR speaks to an active process that allows some people to cope with more brain pathology than others before cognitive abilities are affected.

Brain maintenance refers to reduced development over time of age-related brain changes and pathology based on genetics or lifestyle (Nyberg, Lovden, Riklund, Lindenberger, & Backman, 2012). The key observation is that some people “maintain” their brain better than others with aging, and this is in turn associated with less cognitive decline. Thus, by definition, BM represents the process of maintaining or perhaps enhancing the brain, whereas BR represents the status of the brain at any point in time.

With many people doing research in this area, there has been a proliferation of terminology. Investigators use different terms to encompass the same concept. To help address that issue, the NIA solicited proposals for a “Collaboratory” that would bring together researchers from basic animal through human research in order to reach a consensus on operational definitions and scientific frameworks for these concepts (https://reserveandresilience.com/). We are in the third year of this process and hope to have a consensus document by our final meeting at the end of September 2021.

Epidemiologic Evidence

Epidemiologic evidence for BR relies on differential susceptibility to cognitive change or disease as a function of the extent of neurobiological capital. For example, several studies have found reduced prevalence or incidence of Alzheimer’s disease as a function of brain size or head circumference (e.g., Schofield, Logroscino, Andrews, Albert, & Stern, 1997).

Most of the evidence for BM relies on studies of normal aging. As mentioned previously, longitudinal studies have shown that fewer age-related changes in the brain over time, as assessed by measures like brain volume and cortical thickness, are associated with more preserved cognition. Many lifestyle and genetic factors have been related to BM. More recently, data have begun to suggest that the concept of brain maintenance can also be applied to differential aggregation of pathologies such as neuritic plaques (Landau et al., 2012).

The initial epidemiologic evidence for CR came from longitudinal studies of incident Alzheimer’s disease. In an early study, my group showed that individuals with higher levels of educational attainment or occupational attainment had reduced risk of developing Alzheimer’s disease (Stern et al., 1994). Many other groups have replicated these findings and have added in other proxies for CR including IQ, engagement in late-life leisure activities and social networks. In these studies, the underlying assumption is that on average, all of the participants were accumulating Alzheimer’s pathology at the same rate. Thus, differences in time to developing dementia are based on differential ability to cope with the underlying pathology.

Another early finding regarding CR in patients with Alzheimer’s disease was that patients with higher reserve have more rapid rates of decline (Stern, Albert, Tang, & Tsai, 1999). The explanation for this is that these people can withstand the ongoing development of the underlying pathology longer, so it is more severe by the time dementia emerges. Therefore, there is less time before the pathology overwhelms the system.

Findings in studies of CR in normal aging have been more diverse. Although some studies have found that higher education is associated with a lower rate of age-related cognitive decline (Manly, Touradji, Tang, & Stern, 2003; Zahodne, Stern, & Manly, 2015b), others have not noted this (Zahodne et al., 2011). However, other factors presumably associated with CR, such as IQ and occupational attainment, have been associated with more successful aging. Some studies have directly quantified age-related brain changes and can directly determine whether various reserve-related exposures moderate the effects of these changes on cognition. These are examples of the types of epidemiologic studies that have been used to examine these concepts. Many studies now take advantage of a very large data sets and can account for multiple interacting exposures that might enhance brain maintenance and cognitive reserve.

Studying Cognitive Reserve

In my first paper about CR (although I did not call it that at that time; Stern, Alexander, Prohovnik, & Mayeux, 1992), we used an imaging measure of neurodegeneration, considered to be the final path of Alzheimer’s pathology including plaques and tangles, in patients with Alzheimer’s disease. When controlling for disease severity, the study noted that individuals with higher education had more advanced neurodegeneration. This suggested that those with higher education could cope with more pathology while maintaining the same level of clinical severity. This finding has been replicated in multiple studies using autopsy, CSF and neuroimaging measures of Alzheimer’s disease pathology. This study demonstrates the three components necessary to study CR: a brain change, a cognitive or functional change related to the brain change and a theoretical moderator associated with cognitive reserve. This finding was subsequently replicated using aspects of occupational attainment (Stern et al., 1995) as well as a measure of leisure activities (Scarmeas et al., 2003). It has also been replicated numerous times using autopsy data (Bennett et al., 2003) as well as various biomarkers (Rentz et al., 2017) to measure the severity of Alzheimer’s pathology.

In most studies, CR is estimated based on proxies presumably associated with increased reserve, such as educational or occupational attainment or engagement in leisure activities. An alternate residual approach (Reed et al., 2010) attempts to directly measure the degree of cognitive reserve that is present at any time. This approach attempts to use some select demographic variables as well as a set of relevant brain measures such as cortical thickness or white matter hyperintensities to predict a measure of cognitive function. The unpredicted portion, or residual, is considered a measure of CR. The logic behind this is that those who do better than expected based on brain and demographic measures must have higher cognitive reserve. The advantage of this approach is that it directly quantitates cognitive reserve and can change as reserve is depleted (Zahodne, Manly, et al., 2015a). Disadvantages include the fact that the residual always consists of much more than cognitive reserve, including the effect of unmeasured pathologic changes and other undetermined influences on cognition beyond CR. From a research point of view, the residual measure cannot provide insight into factors that might promote reserve.

Imaging Cognitive Reserve

Many investigators are using functional imaging to explore how CR is “implemented.” Several early papers addressed the concept of compensation, where someone with more limited brain resources might recruit an additional brain area in order to maintain function (Grady et al., 1994). However, these studies were not linked to CR, where individual differences in the ability to compensate could be explored. My group proposed two forms for the neural implementation of CR (Stern et al., 2005). The first, which I call neural reserve, refers to individual differences across healthy people in how networks underlying cognitive function are implemented. Before the system is challenged by age-related changes or disease, there are individual differences in the efficiency, capacity, or flexibility of neural networks that might be the result of IQ, natural abilities and various life exposures. If this person was challenged by brain damage or aging, these pre-existing differences might allow some people to do better than others. The second concept is neural compensation, where someone affected by brain damage or disease recruits a new network not typically used by intact individuals. This ability to recruit a new network may allow the person to either maintain performance or perform better in the face of these brain changes.

Many of my imaging studies focused on the concepts of efficiency, capacity and compensation (for review see Stern, 2009). We often used a variant of the Sternberg working memory task where individuals are exposed to one, three, or six letters. After a delay period, they see a letter and must indicate whether it was present in the set they had studied. We focused on the change in task-related activation as the task got more difficult because we felt this might give us some insight into efficiency or capacity. In early studies, we found that individuals with higher IQ showed less of an increase from one to three to six letters as if they had higher efficiency. Using alternate tasks with more challenging stimuli, we saw that individuals with higher IQ had higher capacity, in that they could increase the utilization of the task related areas to a higher degree than people with lower IQ. In subsequent studies, we are trying to repeat these observations while carefully controlling for BR. That is, we are measuring brain volume, cortical thickness, white matter tract integrity, white matter hyperintensity burden, and amyloid and tau. Our working hypothesis is that given a certain level of brain reserve, people with higher CR will show greater efficiency or capacity of specific networks, in turn leading to better performance. We are also following participants longitudinally so that we can examine the relationship between change in brain, change in task-related activation and change in performance. Again, these are the three key features needed to study CR.

Instead of looking for individual differences in task-related activation as a neural implementation of CR, we investigated the possibility of a task-invariant CR network, that is, a CR-related network that is expressed across multiple tasks. In one study, we identified a pattern of functional magnetic resonance imaging activation that correlated with IQ and that was expressed during the performance of 12 different tasks (Stern, Gazes, Razlighi, Steffener, & Habeck, 2018). Degree of network expression was also associated with better cognitive performance after controlling for cortical thickness, a measure of brain status, or BR. Alternately, we (Stern, Varangis, & Habeck, 2021) and other groups (Franzmeier et al., 2017) have investigated the presence of resting blood oxygen level-dependent networks that moderate the effect of brain change on cognition.

Clinical Implications

There are direct clinical implications of CR for clinical characterization of aging and Alzheimer’s disease. For example, the optimal descriptive evaluation of someone for Alzheimer’s disease should include measures of cognition and function, disease pathology and the individual’s CR assessed either by proxy or by imaging-based techniques. This would supply the best overall description of patient status. Such a characterization would be important for early diagnosis. It would also provide information about prognosis, because rates of decline differ as a function of cognitive reserve. Similarly, clinical trials should take note of CR since they rely on differential rates of change in cognition in the drug and placebo groups. Lack of good randomization for CR could result in unwanted skewing of trial results.

Although this discussion has focused on aging and dementia, the concept of CR is applicable to any condition that affects the brain. This includes multiple sclerosis (Sumowski, Wylie, Deluca, & Chiaravalloti, 2010), HIV (Farinpour et al., 2003), traumatic brain injury (Kesler, Adams, Blasey, & Bigler, 2003), stroke and psychiatric disorders such as schizophrenia (Barnett, Salmond, Jones, & Sahakian, 2006).

Since CR and BM are recognized to be associated with more successful aging, it is important to consider how interventions might promote them. Many articles have focused on risk factor production for Alzheimer’s disease, and have included features associated with these concepts (e.g., Livingston et al., 2020). Healthy diets, exercise, reduction of risk for stroke and diabetes can all help maintain the brain. Similarly better education, more challenging occupations, richer social networks and engagement in leisure activities can promote CR. Most likely all of these experiences enrich both brain maintenance and cognitive reserve, but in different ways. It is important to recognize that each of these factors can account for unique components of cognitive reserve and brain maintenance. Recently, large-scale studies combining many of these activities have been initiated (Ngandu et al., 2015).

Conclusion

In summary, epidemiologic and imaging evidence support the concept of reserve. Reserve is malleable and can be influenced by experiences in every stage of life. Here, I have discussed CR and BR. The level of BR at any point in time is determined by BM. The concept of reserve is applicable to a wide range of conditions that impact on brain function at all ages. Imaging studies can help clarify the neural implementation of reserve. Influencing reserve and brain maintenance may delay or reverse the effects of aging or brain pathology. Careful, focused research will help us maximize the chance for successful intervention.

Funding

This work was supported by grant R01 AG26158 from the National Institute on Aging.

Conflict of Interest

Consultant for Eisai: Columbia University licenses the Dependence Scale, and in accordance with University policy, Dr. Stern is entitled to royalties through this license.