Regular articleWhite matter hyperintensities and cognitive reserve during a working memory task: a functional magnetic resonance imaging study in cognitively normal older adults
Introduction
The concepts of brain and cognitive reserve (CR) emerged to account for the observation that individuals with the same extent of brain damage may differ in the expression of their clinical or cognitive status (Stern, 2009). Brain reserve models posit that individuals exhibiting greater measures of brain size (Guo et al., 2013) or neural counts or synaptic density (Perez-Nievas et al., 2013) can sustain higher amounts of brain damage until functional decline becomes evident. In these passive models, once injury reduces the brain reserve below a critical threshold, clinical deficits become apparent. In contrast, CR models are considered as active models; they propose that the threshold for functional decline is not fixed and can be modified based on experience. Several clinical and epidemiological studies have confirmed that education level and cognitively stimulating or leisure activities protect against the manifestation of clinical conditions, including the expression of dementia (Valenzuela and Sachdev, 2006), cognitive and functional decline in multiple sclerosis (Sumowski et al., 2009, Sumowski et al., 2013a), traumatic brain injury (Sumowski et al., 2013b), and stroke (Nunnari et al., 2014). Moreover, interventional studies have shown that training in cognitive stimulating activities may have an impact on other cognitive abilities (Chan et al., 2014, Park et al., 2014 Anguera et al., 2013, Carlson et al., 2008, Dahlin et al., 2008a, Dahlin et al., 2008b), brain morphology (Lövdén et al., 2012), and function (Anguera et al., 2013, McDonough et al., 2015). In this regard, McDonough et al. (2015) reported decreases in brain activity during the easy condition of a semantic task after engaging in high-challenge learning activities.
The functional brain mechanisms through which CR operates are still under debate. Two concepts have been proposed: neural reserve and neural compensation (Stern, 2009). Neural reserve is related to networks that have developed over the lifespan; it addresses differences in the network efficiency, capacity, or flexibility that healthy older individuals deploy to perform tasks and to cope with increasing task difficulty. In this regard, neural reserve may be evident when aged individuals invoke the same brain networks as young participants and may help explain the susceptibility to brain injury and age-related brain changes. On the other hand, neural compensation refers to the recruitment of alternative brain networks to accomplish cognitive tasks when the brain has been injured or affected by age; thus, individuals with higher CR would be able to better recruit additional resources that help them maintain their performance (Barulli and Stern, 2013). Both neural reserve and neural compensation have been studied using functional magnetic resonance imaging (fMRI). Neural efficiency has been previously observed in healthy older adults. As an example, in a previous study by our group, we observed that healthy older adults with higher estimated levels of CR showed reduced activation in frontal regions during a working memory task when matched for cognitive performance with those with lower CR (Bartrés-Faz et al., 2009). In terms of neural compensation, Steffener et al. (2011) showed that the increased expression of compensatory networks has a less negative impact on older adults with higher CR than in those with lower CR. Thus, higher CR decreased the impact of the expression of the compensatory network on task performance.
A growing number of studies use both structural and functional neuroimaging techniques to investigate CR in aging and dementia (Bartrés-Faz and Arenaza-Urquijo, 2011). In particular, the study of the core mechanisms of CR, including neural efficiency or neural compensation, is particularly well suited for investigations in which participants are faced with demanding cognitive tasks because compensation networks may not emerge until task demands exceed a threshold (Steffener and Stern, 2012). Yet, to our knowledge, there has been little research specifically addressing how cognitive reserve estimates modulate the expression of brain activity patterns during cognitively challenging tasks in individuals suffering from different degrees of common age-related brain structural white matter damage. White matter hyperintensities (WMHs), hyperintense areas that appear in T2-weighted images and are presumed to have a vascular origin (Wardlaw et al., 2013) tend to accumulate with age (de Leeuw, 2001) and are particularly detrimental for executive functioning (Birdsill et al., 2014, Tullberg et al., 2004). Besides, WMH may also underlie cerebral dysfunction and cognitive decline (Prins and Scheltens, 2015, Tuladhar et al., 2015) and predict an increased risk of developing dementia (Debette and Markus, 2010). Furthermore, WMHs appear to have an impact on functional brain activation patterns and have been associated with both higher (Lockhart et al., 2015) and lower functional brain activation (Nordahl et al., 2006, Venkatraman et al., 2010). As CR is supposed to mediate the relation between brain pathology and cognition, WMHs can serve as a model of neural injury to study the reorganization of brain networks in the presence of brain damage among cognitively normal older adults with high and low levels of CR. In support of this view, both Nebes et al. (2006) and Dufouil et al. (2003) found that the relationship between WMH and cognition was attenuated in individuals with a high level of education. Other studies also suggest that CR moderates the impact of WMH on cognition (Brickman et al., 2011, Vemuri et al., 2015). However, there is scarce evidence of the functional mechanisms of CR in the context of WMHs. In terms of neural compensation, Venkatraman et al. (2010) found that higher WMHs were associated with lower brain activity during an executive control task. However, the recruitment of task-related and non-task-related areas in the presence of WMHs was associated with higher accuracy, indicating possible compensatory mechanisms. Recently, Griebe et al. (2014) found higher activation in all levels of a working memory task in the group with greater WMHs, suggesting the presence of compensatory mechanisms already at low task demands. However, performance was not addressed in that study. While these reports investigated the impact of WMHs on brain activation, to our knowledge, none of the ones published to date have explored the interaction between measures of WMH and CR estimates during a working memory task.
The aim of this study was to investigate the concepts of neural reserve and neural compensation in a sample of healthy older adults during a working memory task. Participants were grouped according to their WMH burden and years of education (used as a proxy of CR) to explore the effect of the interactions between WMH burden and CR on cognitive performance and on brain activity associated with increasing loads of an n-back working memory task (3 > 2 > 1 > 0, see the following). Subsequently, we aimed to compare the pattern of differences in brain activation with those of a group of young participants to characterize areas of task-related activation. We hypothesized that the effect of WMHs on brain activation would differ as a function of education, in so far as individuals with high WMHs and high education would show putative compensatory mechanisms to maintain performance. We further expected participants with low education and high WMH rates to present the poorest performance, due to compromised neural reserve or compensation mechanisms. Finally, we predicted that older adults with high education and low WMH burden would exhibit high WM performance and would tend to make a greater use of the networks recruited by young individuals.
Section snippets
Participants
One hundred sixteen cognitively normal individuals (79 women, mean age = 68 years; range: 63–76 years) participated in this study. Candidates were recruited from retirement homes and centers registered with the Institut Català de l'Envelliment, a nonprofit organization caring for older adults, in the area of Barcelona. All participants underwent a neuropsychological assessment covering major cognitive domains (memory, language, attention, processing speed) and had a normal cognitive profile
Behavioral results
Univariate ANOVA revealed a main effect of YoE on the hits [F(1,84) = 6.290; p = 0.014] and d′ [F(1,84) = 8.357; p = 0.005; ηp2 = 0.070], such that the highest educated participants had a greater proportion of hits and d′ values at 3 back than the least educated (Table 1). The main effect of WMH burden [hits: F(1,84) = 1.31; p = 0.256; ηp2 = 0.015; d′: F(1,84)=1.229; p = 0.271; ηp2 = 0.014] and the interaction between WMHs and YoE were not significant [hits: F(1,84) = 0.119; p = 0.731; ηp2 =
Discussion
The main goal of this study was to assess the functional brain mechanisms engaged by healthy older adults during a cognitively challenging task as a function of educational status, a proxy measure of CR, and WMH burden. We observed that high level of education was related to better cognitive performance, regardless of WMH status. Functionally, higher education was globally associated with broader patterns of brain activity, but the specific brain mechanisms subtending the cognitive advantage
Disclosure statement
The authors have no conflicts of interest to disclose.
Acknowledgements
Sara Fernández-Cabello was supported by the Doctoral College “Imaging the Mind” (FWF-W1233) of the Austrian Science Foundation (FWF-W1233). Partially funded by a Spanish Ministry of Economy and Competitiveness (MINECO) grant to David Bartrés-Faz (PSI2015-64227-R) and the Walnuts and Healthy Aging (WAHA) study (http://www.clinicaltrials.gov NCT01634841) funded by the California Walnut Commission, Sacramento, California, USA.
References (65)
- et al.
Interactions of cognitive reserve with regional brain anatomy and brain function during a working memory task in healthy elders
Biol. Psychol.
(2009) - et al.
Efficiency, capacity, compensation, maintenance, plasticity: emerging concepts in cognitive reserve
Trends Cogn. Sci.
(2013) - et al.
Regional white matter hyperintensities: aging, Alzheimer's disease risk, and cognitive function
Neurobiol. Aging
(2014) - et al.
White matter hyperintensities and cognition: testing the reserve hypothesis
Neurobiol. Aging
(2011) - et al.
Age differences in prefontal recruitment during verbal working memory maintenance depend on memory load
Cortex
(2010) - et al.
Brain size and the compensation of Alzheimer's disease symptoms: a longitudinal cohort study
Alzheimers Dement.
(2013) - et al.
Age effects on load-dependent brain activations in working memory for novel material
Brain Res.
(2009) - et al.
Working memory circuitry in schizophrenia shows widespread cortical inefficiency and compensation
Schizophr. Res.
(2010) - et al.
Spatial navigation training protects the hippocampus against age-related changes during early and late adulthood
Neurobiol. Aging
(2012) - et al.
Bilateral basal ganglia activity in verbal working memory
Brain Lang.
(2013)