Activity and the Brain

Rapidly emerging evidence indicates that physical activity directly influences the functional and structural properties of the hippocampus, a brain region critical to learning and memory throughout the life course (Shen, Specht, De Saint Ghislain, & Li, Reference Shen, Specht, De Saint Ghislain and Li1994) and implicated in risk for dementia (Killiany et al., Reference Killiany, Hyman, Gomez-Isla, Moss, Kikinis, Jolesz and Albert2002). Change in hippocampal volume has been correlated with decline in memory performance and with increased risk for dementia (Desikan et al., Reference Desikan, Cabral, Fischl, Guttmann, Blacker, Hyman and Killiany2009; Kerchner et al., Reference Kerchner, Hess, Hammond-Rosenbluth, Xu, Rabinovici, Kelley and Miller2010; Killiany et al., Reference Killiany, Hyman, Gomez-Isla, Moss, Kikinis, Jolesz and Albert2002; Kramer et al., Reference Kramer, Mungas, Reed, Wetzel, Burnett, Miller and Chui2007; Petersen, Reference Petersen2000). Cross-sectional and longitudinal studies have reported an inverse association between self-reported physical activity and age-related decreases in mediotemporal volume (Bugg & Head, Reference Bugg and Head2011) and with mitigated rates of decline in hippocampal volume nine years later (Erickson, Raji, et al., Reference Erickson, Raji, Lopez, Becker, Rosano, Newman and Kuller2010). Greater cardiorespiratory fitness has been associated cross-sectionally with greater hippocampal volume in both children (Chaddock et al., Reference Chaddock, Erickson, Prakash, Kim, Voss, Vanpatter and Kramer2009) and healthy older adults (Erickson et al., Reference Erickson, Prakash, Voss, Chaddock, Hu, Morris and Kramer2009). In a randomized, controlled trial of exercise in healthy older adults, a walking intervention over one year increased hippocampal volume by 2% relative to 1% decreased volume in a stretching and toning control group (Erickson, Voss, et al., Reference Erickson, Voss, Prakash, Basak, Szabo, Chaddock and Kramer2010).

The hippocampus is influenced by the PFC, the evolutionarily newest region of the brain central to integration of the past and present in anticipation of future rewards and threats, and further represents an area important for many prosocial behaviors (Diamond, Reference Diamond2002). PFC maturation is not complete until one's early 20s (Gogtay et al., Reference Gogtay, Sporn, Clasen, Nugent, Greenstein, Nicolson and Rapoport2004), presumably because the ability to integrate multiple streams of information requires the maturation of sensory-specific networks (Casey, Getz, & Galvan, Reference Casey, Getz and Galvan2008; Garon, Bryson, & Smith, Reference Garon, Bryson and Smith2008; Rubia, Smith, Taylor, & Brammer, Reference Rubia, Smith, Taylor and Brammer2007). This slow maturation is reflected in adolescents’ tendency toward risk-taking behaviors and less consideration of the consequences of their actions (Casey et al., Reference Casey, Getz and Galvan2008; Steinberg, Reference Steinberg2007). With age and experience, these priorities typically reverse and thoughts of consequences more often prevail over the moment. In later life, however, the late-developing prefrontal circuits appear to be more vulnerable than other brain regions, including the hippocampus. Studies of the aging human brain show that loss of brain volume is greater in the PFC than in posterior areas of the cortex (Buckner, Reference Buckner2004; Head, Raz, Gunning-Dixon, Williamson, & Acker, Reference Head, Raz, Gunning-Dixon, Williamson and Acker2002; Madden, Reference Madden2000; Raz, Reference Raz2000; Resnick, Pham, Kraut, Zonderman, & Davatzikos, Reference Resnick, Pham, Kraut, Zonderman and Davatzikos2003). Consistent with these findings, in longitudinal study, we have found that declines in executive function preceded those in verbal memory in older community-dwelling adults (Carlson, Xue, Zhou, & Fried, Reference Carlson, Xue, Zhou and Fried2009). In this issue, Chaddock, Pontifex, Hillman, and Kramer summarize cross-sectional evidence in children demonstrating a relationship between physical fitness, physical activity, and academic achievement, and conclude that these associations may be due to enhanced cognitive control mediated by PFC networks during this critical developmental window. Effects of physical activity, however, are not limited to the prefrontal cortex. Chaddock and colleagues also discuss associations between aerobic fitness, episodic memory, and volume of the hippocampus. Moreover, they provide evidence that physical fitness predicts performance on a memory task one year later, suggesting that the benefits of fitness extend over time. These data call for more definitive randomized controlled trials of physical activity in childhood to determine its impact on different aspects of neurocognitive development.

This symposium also provides evidence for the benefits of physical fitness and activity in response to neuronal injury in mid-life. Prakash, Patterson and Janssen have studied the relationship between physical fitness and cognitive and brain health in multiple sclerosis (MS) (Prakash et al., Reference Prakash, Snook, Erickson, Colcombe, Voss, Motl and Kramer2007; Prakash, Snook, Motl, & Kramer, Reference Prakash, Snook, Motl and Kramer2010). In this issue, they report that higher levels of physical activity in patients with MS are associated with increased resting-state connectivity between the hippocampus and posteromedial cortex, two regions that are critical to memory. Higher connectivity, in turn, was associated with better memory performance, although physical activity itself was not associated with better memory in this small sample. These findings are intriguing, but their cross-sectional nature prevents causal inferences, in part, because higher physical activity may reflect less severe MS, which may likely be associated with increased resting state connectivity. Therefore, there is a clear need for intervention studies in MS that promote physical activity in daily life and evaluate the mechanisms of benefit.

Measuring Activity In Daily Life

There is now ample evidence that enriching environments physically, cognitively, and socially promotes the emergence of new neurons and synapses in old, as well as young, animals and reduces neuronal death, especially in the hippocampus (Holmes, Galea, Mistlberger, & Kempermann, 2004; Kempermann, Chesler, Lu, Williams, & Gage, Reference Kempermann, Chesler, Lu, Williams and Gage2006). However, measures of cognitive and physical activity have traditionally focused on the frequency of an activity, under the logical premise that more of a good thing is better than less of a good thing. In our recent work, we have directly compared the utility of frequency of cognitive activities relative to the number of different stimulating activities endorsed, under the premise that number, or variety, of distinct activities may serve as an index of an active and engaged lifestyle (Carlson et al., Reference Carlson, Parisi, Xia, Rebok, Brandeen-Roche and Fried2011). Greater variety of participation in activities, regardless of cognitive challenge, was associated with an 8 to 11% reduction in the risk of impairment in verbal memory and global cognitive outcomes over nearly 10 years among initially high-functioning older women. Participation in a variety of lifestyle activities was a better predictor of cognitive outcomes than frequency or level of cognitive challenge. We interpreted these findings as reflecting the influence of a greater variety of environments and greater requirements to flexibly coordinating time and resources across multiple activities, in keeping with the environmental complexity hypothesis (Hultsch, Hertzog, Small, & Dixon, Reference Hultsch, Hertzog, Small and Dixon1999).

A review of the epidemiologic literature on activity and risk for cognitive impairment suggests that activity in various forms promotes cognitive reserve. For example, in a 3-year follow-up study, those who participated in three or more social or leisure activities had an 80% lower dementia risk compared to those who did not participate in such activities (Fabrigoule et al., Reference Fabrigoule, Letenneur, Dartigues, Zarrouk, Commenges and Barberger-Gateau1995; Helmer et al., Reference Helmer, Damon, Letenneur, Fabrigoule, Barberger-Gateau, Lafont and Dartigues1999). The study also reported a stepwise reduction in dementia risk with increasing number of activities endorsed. In the Cardiovascular Health Study, energy expenditure in kilocalories did not predict reduced dementia risk after adjustment for demographic factors, whereas variety or number of functional and exercise activities, adjusted for covariates and energy expenditure, was a significant predictor among those without genetic risk for AD (i.e., Apoe-e4 allele carriers; Podewils et al., Reference Podewils, Guallar, Kuller, Fried, Lopez, Carlson and Lyketsos2005). These data suggest that variety may serve as an index of an active and engaged lifestyle, and may provide neurocognitive benefits by exercising PFC-mediated executive functions and hippocampally mediated memory functions, as postulated above.

In keeping with the known benefits of social networks (Fratiglioni, Paillard-Borg, & Winblad, Reference Fratiglioni, Paillard-Borg and Winblad2004), a growing body of evidence also suggests that greater mid- and late-life social engagement is associated with better cognitive and physical health and greater longevity (Carlson, Helms, et al., Reference Carlson, Saczynski, Rebok, Seeman, Glass, McGill and Fried2008; Saczynski et al., Reference Saczynski, Pfeifer, Masaki, Korf, Laurin, White and Launer2006). In this issue, James, Wilson, Barnes, and Bennett report that, among a large, community-based cohort of older adults initially free of dementia, more frequent social activity was associated with reduced rates of cognitive decline across several domains over 5 years, independent of physical and cognitive activity. Because their analyses adjusted for size of social network, their findings speak to the unique value of social activity. Benefits are postulated to derive, in part, from navigating complex and interpersonal exchanges and the associated intellectual engagement (Hultsch et al., Reference Hultsch, Hertzog, Small and Dixon1999), and through the adoption of social roles that may provide a sense of purpose in later life (Berkman, Glass, Brissette, & Seeman, Reference Berkman, Glass, Brissette and Seeman2000; Hultsch et al., Reference Hultsch, Hertzog, Small and Dixon1999). Both of these areas represent emerging topics of study incorporated in the design and evaluation of a new generation of social health promotion interventions that target cognition.

Enriching Environment Through Multiple Activity Pathways

This approach is exemplified by Experience Corps, a community-based model in which retired adults are trained and placed in teams in elementary schools with the goal of improving the academic achievement and behavioral outcomes of children during a critical period of development in reading skills (Fried et al., Reference Fried, Carlson, Freedman, Frick, Glass, Hill and Zeger2004). Through this form of volunteer activity, physical, cognitive, and social activities have been preliminarily shown to increase (Fried et al., Reference Fried, Carlson, Freedman, Frick, Glass, Hill and Zeger2004), through walking to the school (Tan, Xue, Li, Carlson, & Fried, Reference Tan, Xue, Li, Carlson and Fried2006), supporting literacy and math skill, aiding behavior management and positive conflict resolution, and the need to flexibly shift between these roles (see Rebok, Carlson, Barron, Frick, & McGill, Reference Rebok, Carlson, Barron, Frick and McGill2011). In harnessing older adults’ knowledge and wisdom accumulated over a lifetime, they become simultaneously the agents of social health promotion for themselves and for the children.

A 7-month, randomized, controlled pilot trial of Experience Corps enhanced executive functions and memory among those community-dwelling older adults with poor executive function at baseline relative to matched controls (Carlson, Saczynski, et al., Reference Carlson, Helms, Steffens, Burke, Potter and Plassman2008). A pilot brain-imaging study over 6 months in Experience Corps volunteers showed improved executive function and increased task-related brain activity in the PFC relative to matched controls (Carlson, Erickson, et al., Reference Carlson, Erickson, Kramer, Voss, Bolea, Mielke and Fried2009). Both studies show that individuals at greatest risk for cognitive impairment maintain great potential for activity-induced plasticity. These preliminary program results are encouraging in both volunteers and among the children they serve (Rebok et al., Reference Rebok, Carlson, Glass, McGill, Hill, Wasik and Rasmussen2004) and await confirmation in a large-scale randomized trial of the Experience Corps Program in Baltimore.

Aging and Neuropathology: Alzheimer's Disease

Although the causes of AD are not well understood, observational studies consistently demonstrate that risk may be modified by education, physical activity, and cognitive activity (Barnes & Yaffe, Reference Barnes and Yaffe2011; see Qiu, De Ronchi, & Fratiglioni, Reference Qiu, De Ronchi and Fratiglioni2007, for a review). Verghese and colleagues have previously shown in the Bronx Aging Study that, among older adults who ultimately developed dementia, participation in cognitively stimulating leisure activities delayed the onset of accelerated memory decline. In this issue, Pillai, Hall, Dickson, Buschke, Lipton, and Verghese show that participation in crossword puzzles delayed onset of accelerated memory decline by 2.5 years compared to non-puzzlers. However, once memory decline began, the rate of decline was more rapid for crossword puzzlers than for non-puzzlers. These findings are consistent with the cognitive reserve hypothesis, which postulates that cognitively stimulating activities may help delay the onset of clinical deficits until the point that pathological burden overwhelms reserve (Stern, Reference Stern2009). Along the same line, we have observed in the population-based Cache County Dementia Progression Study that weekly participation in cognitively stimulating activities early in the course of AD was associated with slower cognitive decline (Treiber et al., Reference Treiber, Carlson, Corcoran, Norton, Breitner, Piercy and Tschanz2011).

An important cautionary note is the potential for reverse causality, namely, the possibility that higher levels of engagement in stimulating activities may be a reflection of less severe illness. Therefore, randomized and matched case-control intervention studies represent an important next step to elucidate whether the very pathways by which inactivity may adversely affect childhood and older adult development are the same malleable pathways by which to enrich and boost cognitive health to help circumvent advancing neuropathology.