Shaping the adult brain with exercise during development: Emerging evidence and knowledge gaps

Abstract

Exercise is known to produce a myriad of positive effects on the brain, including increased glia, neurons, blood vessels, white matter and dendritic complexity. Such effects are associated with enhanced cognition and stress resilience in humans and animal models. As such, exercise represents a positive experience with tremendous potential to influence brain development and shape an adult brain capable of responding to life’s challenges. Although substantial evidence attests to the benefits of exercise for cognition in children and adolescents, the vast majority of existing studies examine acute effects. Nonetheless, there is emerging evidence indicating that exercise during development has positive cognitive and neural effects that last to adulthood. There is, therefore, a compelling need for studies designed to determine the extent to which plasticity driven by developmental exercise translates into enhanced brain health and function in adulthood and the underlying mechanisms. Such studies are particularly important given that modern Western society is increasingly characterized by sedentary behavior, and we know little about how this impacts the brain’s developmental trajectory. This review synthesizes current literature and outlines significant knowledge gaps that must be filled in order to elucidate what exercise (or lack of exercise) during development contributes to the health and function of the adult brain.

1. Introduction

Many decades of research attest to the idea that early experiences (both positive and negative) exert profound effects on brain development, ultimately shaping the structure and function of the adult brain (Bick & Nelson, 2017; Hubel & Wiesel, 1965; Rosenzweig, 2003). Positive experience has long been modeled in laboratory rodents via an “enriched” condition, with multiple same-sex animals housed together with toys and other stimuli, thus providing an engaging environment. The effects of enrichment on brain plasticity are substantial and enduring compared to animals maintained in standard laboratory conditions, and results of enrichment studies have informed the development of research-based interventions aimed at maximizing brain development and childhood learning by modifying the environment (for a comprehensive review see Rosenzweig & Bennett, 1996).

Until relatively recently, the focus was on the benefits of mental activity, with comparably less attention devoted to those provided by physical activity. Yet the effects of mental and physical activity on brain plasticity are separable from each other (van Praag, Kempermann, & Gage, 1999). Fortunately, recent decades have seen a surge in basic and clinical research on the effects of physical activity on brain function, which include enhanced cognition (Hillman, Erickson, & Kramer, 2008) and stress resilience (Stranahan, Lee, & Mattson, 2008). With respect to children and adolescents, evidence suggests that physical activity enhances performance on a wide range of cognitive domains (Khan & Hillman, 2014) and there is growing interest in physical activity interventions to optimize brain function in these age groups as well as to prevent peripheral diseases that erode it, including type 2 diabetes and other components of the metabolic syndrome. The latter are becoming more prevalent among Western children and teens, coincident with an increasingly sedentary lifestyle. Early-life experiences configure the developing brain to match its individual environmental and behavioral challenges (Andersen, 2003), providing the basis for optimal function in adulthood. We must therefore address the possibility that physical inactivity during development will render the adult brain sub-optimally prepared to cope with life’s challenges.

We first began considering the implications of physical (in)activity for brain development after an unexpected finding several years ago, when two of us (SPR and JLL) were part of a multi-lab collaborative project aimed at determining the capacity of exercise to rebuild the brain after developmental exposure to radiation therapy (RT; as for childhood brain cancer). The brains shown in Figure 1 are from that project. The S-SHAM brain is from a rat that was sedentary its entire life. The E-SHAM brain is from a rat that exercised from adolescence to early adulthood, but that had not exercised in 9 months. We were encouraged to see that developmental exercise reversed the decrease in brain size caused by RT (E-RT vs S-RT). But we were astonished to see, with the naked eye, what appeared to be lifelong effects of developmental exercise on brain volume in the E-SHAM animals. Neuroimaging of brains from all the groups shown above substantiated the apparent increased size of the developmentally exercised brain. Specifically, total brain volume, brain length, brain width, and whole brain fiber number were significantly increased in developmentally exercised versus sedentary animals (Sahnoune et al., 2018). Moreover, among SHAM (non-irradiated) rats, those that exercised during development had significantly better cognitive control in adulthood compared to their sedentary cohort, as determined by the 5-Choice Serial Reaction Time Task (5-CSRTT). Specifically, exercise significantly improved task acquisition and reduced impulsivity. These results indicate enduring effects of developmental exercise on brain structure and function in adulthood, consistent with the idea that early-life experiences drive brain adaptation to environmental and behavioral demands.

Fig 1: Lasting effects of developmental exercise.

Exercise during development increased brain size in both irradiated (RT) and SHAM rats. Brains above were harvested when the animals were 1 year old, 11 months after RT and 9 months after cessation of exercise.

While a growing body of evidence substantiates the benefits of exercise for cognition in children and teens, there is a need for research on its long-term effects on adulthood neurobehavioral outcomes. This review addresses the emerging evidence for such long-term effects and pinpoints existing knowledge gaps that need to be addressed in order to achieve a more comprehensive understanding of what developmental exercise (or lack thereof) contributes to the health and function of the adult brain. Throughout, care has been taken that the terms “physical activity” and “exercise” are not used interchangeably. Physical activity is used to refer to motor activity performed in the context of everyday life, whereas exercise refers to more vigorous, aerobic activity done in order to achieve or maintain physical fitness. In order to develop a picture of how specifically exercise during development may sculpt the adult brain, we begin with existing evidence concerning the benefits of exercise (regardless of age) for brain structure and then cover effects on brain function. To highlight the importance of understanding the contribution of developmental exercise on the brain in adulthood, we briefly review data concerning the increasingly sedentary Western lifestyle, which extends to children and teens. We close by reviewing the small but growing number of studies in which exercise occurred during development, and then brain structure and/or function was evaluated in adulthood. We conclude by identifying existing knowledge gaps and suggesting a general experimental design for studies meant to address them.

2. Exercise-driven enhancement of brain structure

Exercise plays an unequivocal role in promoting and maintaining a healthy lifestyle and several of its neuroprotective effects that contribute to maintaining brain health and decelerating aging-related cognitive decline have been identified. These include upregulating neurotrophic factors and neurotransmitter levels, maintaining blood-brain-barrier integrity, and promoting glymphatic clearance (Vecchio, et al., 2018). In contrast, less is known about how exercise may optimize the young brain for adulthood. The results obtained by Sahnoune et al. support the idea that exercise during development increases the volume of the adult brain. If this is the case, then what exactly is being increased by exercise? In the case of the brains in Figure 1, the corpus callosum was found to be larger in the E-SHAM group, and there was an overall effect of exercise on fiber number (Sahnoune et al., 2018). There are numerous other possibilities, however, given the known effects of exercise on brain structural elements. One of the most celebrated effects of exercise is its ability to enhance the generation of new neurons in the hippocampus. It has now been 2 decades since the seminal paper showing exercise-driven hippocampal neurogenesis was published (van Praag et al., 1999). This result has been replicated by hundreds of studies and holds tremendous implications for the use of exercise to enhance hippocampal-dependent functions. Exercise does not just augment generation of hippocampal neurons, however. Below, we briefly outline some of the other brain structural elements which, if increased by exercise during development, might provide lasting infrastructure for the adult brain.

2.1. Dendrites

Multiple studies indicate that exercise has the capacity to remodel dendritic arbors. Indeed, it has been shown to increase dendritic length and mean number of branches on neurons in the dentate gyrus (DG) (Redila & Christie, 2006) and has also been shown to increase spine density and length, as well as synaptic markers in the medial prefrontal cortex (mPFC) (Brockett, LaMarca, & Gould, 2015). Moreover, early-life exercise may be important for the maintenance of dendritic arborization in adulthood, in other words, preventing deleterious effects caused by inactivity. In one study, mice at postnatal day (PND) 18 were separated into standard or an enriched environment (EE) that included running wheels. EE mice were also trained to perform in various physically-demanding tasks such as swimming. After 35 days, inactive mice had lower levels of total length and number of spines on dendrites of Purkinje cells, indicating that the combination of EE and additional physical activity altered Purkinje cell dendrites during development (Pysh & Weiss, 1979). Others (Hosseiny et al., 2015) have demonstrated similar findings in juvenile female mice exposed to EE (including running wheels) or a standard environment (SE) beginning at either PND 28 (adolescence) or PND 56 (adult). EE in both age groups had similar effects on the morphology of basilar dendrites of hippocampal CA1 neurons, with an overall increase in dendritic branching due to EE (Hosseiny et al., 2015). A limitation of this research, however, is that the effects of EE are intertwined with physical activity.

Nonetheless, the effect of exercise on dendritic plasticity may be stronger during development. Recently, exercise-induced differences were found in dendritic structure among adolescent and adult animals. After 2 weeks of exercise, adolescent runners had shorter branch length and increased density of apical spines, whereas no changes were observed in the adults (Eddy & Green, 2017). To our knowledge, no studies have yet addressed whether developmental exercise enduringly changes dendritic structure or complexity of the adult brain.

2.2. Myelination/oligodendrocytes

Although still largely under-scrutinized, human and rodent studies provide evidence that early-life exercise impacts white matter and cognitive function. (Herting, Colby, Sowell, & Nagel, 2014) reported that in adolescent males (15-18 years of age), aerobic fitness relates to white matter connectivity and microstructure. Studies examining the therapeutic efficacy of aerobic exercise in clinical populations have shown that it promotes white matter recovery and improves reaction time in children (6-17 years of age) treated with cranial radiation (Riggs et al., 2017) and improves executive function and white matter integrity in deaf children (9-13 years of age) (Xiong et al., 2018). Similarly, aerobic exercise in 4-wk-old rats has been shown to mitigate cognitive impairment and white matter injury from blood-brain barrier disruption induced by chronic cerebral hypoperfusion (Lee et al., 2017). Though these studies did not include healthy controls, their findings nevertheless indicate that white matter plasticity is responsive to physical activity during development. Sahnoune and colleagues’ findings that developmental exercise induced a long-lasting increase in fiber density and width of the corpus callosum are consistent with this hypothesis. At the cellular level, it has been reported that 4 weeks of voluntary running modulated oligodendrocyte lineage development differentially in juvenile (5 weeks) and young-adult mice, in a region-specific manner (prefrontal cortex versus corpus callosum). Myelin levels in both regions, however, were unaffected by juvenile or young adult exercise (Tomlinson, Huang, & Colognato, 2018).

2.3. Astrocytes

As regulators of neuronal activity and health (Barzilai, 2011; Henneberger, Papouin, Oliet, & Rusakov, 2010) astrocytes are uniquely poised to influence brain function. It is not surprising then that the potential for exercise to alter astrocytes in the adult brain has been studied. Three or 6 weeks of treadmill exercise was found to increase the density of astrocytic (GFAP+) processes in the cortex and striatum (Li et al., 2005). Interestingly, this study was one of very few in which the longevity of exercise effects were examined after a period of inactivity, in this case 3 weeks. It was found that GFAP expression was still increased in both the cortex and striatum, relative to sedentary controls. This result is extremely interesting, as it suggests that the effect of exercise on astrocytes is long-lasting. However, it should be noted that another study also used the same 3 week on/3 week off design (albeit with wheel running) and found that although 3 weeks of exercise increased GFAP+ processes in the globus pallidus, this increase was not found after 3 subsequent weeks of inactivity (Tatsumi et al., 2016). Other studies have shown that, in the hippocampus, exercise increases astrocyte proliferation and/or GFAP expression (Saur et al., 2014; Uda, Ishido, Kami, & Masuhara, 2006) and that it alters astrocyte morphology and the orientation between astrocytes and DG granule neurons (Fahimi et al., 2017). Finally, it has also been shown that exercise increases the size of astrocyte cell bodies in the hippocampus and cortex (Brockett et al., 2015). It should be noted that one study showed a decrease in hippocampal astrocytes after 4 weeks of treadmill training (Bernardi et al., 2013). Thus, there is considerable evidence that exercise has a marked impact on astrocytes, with most studies indicating an enhancing effect.

2.4. Vasculature

Developmental exercise may also enhance vascularization of the brain, which could contribute to an increase in volume. Studies in humans have shown that chronic physical activity augments health and maintenance of the brain vasculature. Long-term, high levels of physical activity reduce the risk of vascular disease, such as atherosclerosis as measured by carotid artery (blood supply to the brain) integrity (Kwasniewska et al., 2014; Ried-Larsen, Grontved, Kristensen, Froberg, & Andersen, 2015).

Studies using animal models have further demonstrated that exercise promotes angiogenesis in the brain. Classic studies on the effects of mental activity versus exercise showed that exercise promotes angiogenesis, whereas learning promotes synaptogenesis (Black, Isaacs, Anderson, Alcantara, & Greenough, 1990; Isaacs, Anderson, Alcantara, Black, & Greenough, 1992). Three weeks of running wheel access increased the proliferation of vascular endothelial cells in the hippocampal molecular layer (Ekstrand, Hellsten, & Tingstrom, 2008). Exercise also increases angiogenesis in the motor cortex and promotes vascular flow when this region is activated (Swain et al., 2003). A similar angiogenic response has been found in female middle-aged rats (Huang et al., 2013) and monkeys (Rhyu et al., 2010). However, it should be noted that not all research replicates the exercise-induced increases in brain vascularization (Cudmore, Dougherty, & Linden, 2017).

Interestingly, exercise-induced vascularization is associated with enhanced cognition. For example, chemically-induced inhibition of angiogenesis during exercise negatively affected visual spatial learning and memory assessed in a Morris water maze (Kerr, Steuer, Pochtarev, & Swain, 2010). Specifically, runners with inhibited angiogenesis showed long-term spatial memory impairment as demonstrated by poorer performance on probe trials 30 days after the cessation of exercise and task acquisition. This suggests that exercise in adulthood has long-term beneficial effects on cognitive behaviors that is partly a result of running-induced vascularization. However, others have reported that in middle-aged animals the enhanced cortical vasculature found immediately after exercise is no longer apparent several months after the cessation of exercise (Rhyu et al., 2010). Perhaps, then, development is a time to maximize the effects of a positive experience, such as exercise, on brain structure.

3. Exercise-driven enhancement of brain function

As described in numerous reviews, these exercise-induced brain structural alterations are in turn associated with enhanced cognition in adults, both human and animal (Cotman & Berchtold, 2002; Cotman, Berchtold, & Christie, 2007; Hillman et al., 2008). Both brief and mild exercise have been shown to enhance memory processing and increased functional connectivity among hippocampal and cortical regions in humans (Suwabe et al., 2018), an effect that may be an outcome of increased concentrations of plasma brain derived neurotrophic factor (BDNF; (Griffin et al., 2011)). In aged populations, exercise is frequently cited to slow cognitive decline. There is a positive correlation between aerobic exercise and hippocampal volume in elderly humans (Erickson et al., 2009), suggesting better memory function. Further investigation (Erickson et al., 2011) demonstrated that aerobic exercises were successful in reversing hippocampal volume loss whereas controls displayed a decline. Animal models further confirm the finding that exercise enhances behavioral outcomes, particularly within learning and memory (for review see van Praag, 2009). For example, both short-term (10 days) and long-term (3 months) running enhanced memory consolidation on a passive avoidance task when exercise occurred prior to training (Saadati, Babri, Ahmadiasl, & Mashhadi, 2010). Voluntary running has also improved spatial learning and memory in aged mice (van Praag, Shubert, Zhao, & Gage, 2005). Aerobic exercise also increases neurotrophic factors, neurotransmitters such as glutamate and dopamine, and blood brain barrier integrity, all of which reduce the decline of normal aging (Vecchio et al., 2018).

At the other end of the lifespan, physical activity in youth possesses great potential for improving cognitive function. For example, physical education programs aimed at increasing daily physical activity demonstrate some of the greatest benefits, including working memory and inhibition (for review see Álvarez-Bueno et al., 2017). A rapidly growing literature shows that exercise-induced brain structural alterations are also associated with enhanced cognition in youth. For example, higher-fit children have demonstrated greater executive control and differences in frontal cortex activation when compared to lower-fit children (Khan & Hillman, 2014). Yet little is known regarding whether early-life physical activity produces lasting adulthood outcomes. However, emerging evidence suggests that physical activity during childhood/adolescence is associated with better mental health outcomes 20 years later (McKercher et al., 2014). Additionally, investigators have recently established a protocol for measuring the effect of physical activity on cognitive development and are currently conducting a longitudinal study to measure adult outcomes (Zhao et al., 2017).

Physical activity, particularly aerobic exercise, during postnatal development has positive outcomes on brain plasticity because the brain is particularly malleable during this period (for review see Gomes da Silva & Arida, 2015). Although the preclinical literature is sparse, the evidence suggests that developmental exercise can modulate learning and memory even after its cessation (i.e. chronic effects). For example, four months after the cessation of developmental exercise, rats tested in a contextual fear conditioning task showed activation of a greater number of adult-born neurons in the DG in comparison to their sedentary controls (Shevtsova, Tan, Merkley, Winocur, & Wojtowicz, 2017). However, the overall quantity of activated neurons, both new/adult-born and non-adult-born neurons, did not differ between groups, suggesting that developmental exercise enhances the usage of new neurons in adulthood memory function. Others (Gomes da Silva et al., 2012) have also demonstrated long-term behavioral effects of developmental exercise in adulthood. Specifically, runners demonstrated better spatial memory retention than sedentary animals 5 weeks after the cessation of adolescent exercise and training on a spatial learning and memory task in the Morris water maze. Similarly, exercise during adolescence, but not adulthood, improved object recognition memory 4 weeks after the cessation of voluntary running (Hopkins, Nitecki, & Bucci, 2011). Interestingly, both groups of animals showed enhanced recognition memory in comparison to their age-matched sedentary counterparts immediately after voluntary running, but this effect remained after a resting period only in the juvenile runners. However, enhanced object recognition memory was not replicated by Soch and colleagues (Soch et al., 2016), though this study included a short-term memory (4 hour inter-trial interval) task, while Hopkins and colleagues showed adolescent exercise to be beneficial in a long-term memory (24 hours inter-trial interval) version of object recognition.

Developmental exercise, however, may not always produce the expected positive effects on behavior. For example, though O’Leary and colleagues (O’Leary, Hoban, Cryan, O’Leary, & Nolan, 2019) found increases in mRNAs associated with neuroplasticity in juvenile runners, their behavioral findings were not as clear. Adolescent runners displayed a lower level of freezing in a fear conditioning task than the adult runners, which displayed greater freezing levels than their sedentary counterparts, suggesting that adolescent exercise results in impaired memory and adulthood exercise improves memory performance. However, greater activity levels were also shown in adolescent runners, potentially confounding the results of the fear conditioning task. Juvenile exercise, upon cessation, is also reported to enhance anxiety-like behaviors and partially impair neurogenesis in mice (Nishijima et al., 2013), relative to sedentary controls, indicating that the effects of a reduction in physical activity may extend beyond a mere return to a sedentary phenotype. Furthermore, it is important to test the long-term effects of developmental exercise given some challenge to the system in adulthood, to assess the potential for early-life exercise to optimize brain health and plasticity and adaptively respond to challenges in adulthood, as discussed below.

Indeed, one key way in which exercise enhances brain function is by conditioning stress circuitry, thus preparing it to cope with stressful life events (Shors, 2004; Silverman & Deuster, 2015; Stranahan et al., 2008). Stress is a risk factor for mental illness (Duman, 2009; Duman, Aghajanian, Sanacora, & Krystal, 2016), accelerated brain aging (Wolkowitz, Epel, Reus, & Mellon, 2010) and addiction relapse (Koob & Le Moal, 2001, 2005, 2008), and indeed, one of the most difficult and yet important aspects of life that any organism faces is responding appropriately to stressors. It is possible that inactivity during development could result in a brain that is sub-optimally prepared to respond to adulthood stressors.

Experimentally-induced increases in stress hormones, such as corticosterone, as well as environmental stressors are known to effectively reduce cell proliferation in the DG (Gould & Tanapat, 1999). Paradoxically, exercise increases stress hormones but is instead associated with beneficial behavioral outcomes (Chen et al., 2017) that is, it activates the hypothalamic-pituitary-axis without psychological indicators of fear (Stranahan et al., 2008). Because developmental exercise has the potential to enhance the adult brain as described above, it may optimize the response to a challenge, such as stress, in adulthood. Exercise influences epigenetic changes (Collins et al., 2009), dendritic complexity (Hoffman et al., 2016), and growth factors (Munive, Santi, & Torres-Aleman, 2016) after a stressor, typically in a positive direction. Exercise has also been shown to promote stress coping behaviors (Lalanza et al., 2015; Munive et al., 2016) and reduce behavioral impairment after early-life maternal separation (Baek et al., 2012). Moreover, juvenile exercise has prevented stress-induced activation of the DG and reduced anxiety-like behavior after a stressor in mice (Schoenfeld, Rada, Pieruzzini, Hsueh, & Gould, 2013). However, it is not known if these effects persist once exercise ceases. Collectively, the findings by Sahnoune et al. and the emerging evidence on the long-term effects of developmental exercise suggest that early-life exercise may help optimize brain development and promote increased stress resilience.

4. Physical inactivity in American youth: Consequences for the adult brain?

There is a growing amount of literature attesting to the sedentary lifestyle of American children and teens. Contrary to adults, adolescents do not typically face the risk of acquiring chronic diseases. However, they are prone to other risk factors such as obesity, elevated insulin and blood pressure when their lifestyle lacks sufficient physical activity (Piercy et al., 2018). More recently, however, there has been a higher occurrence of children developing type 2 diabetes, along with the increase in obesity (Castelli, Hillman, Buck, & Erwin, 2007). If a trend of inactivity continues into adulthood, the risk for other health problems such as cardiovascular disease, type 2 diabetes, asthma and an overall lower quality of life will increase in all likelihood (Colberg et al., 2016; Piercy et al., 2018). In 2015-2016, there was an 18.5% prevalence of obesity among adolescents in the United States alone (Hales, Carroll, Fryar, & Ogden, 2017). It is believed that increased rates of overweight and obese adolescents are connected to the age-related decline of activity (Dishman, Dowda, McIver, Saunders, & Pate, 2017). Furthermore, inactivity places important physical needs such as bone strength, muscle development, muscle strength, and lower cardiorespiratory fitness (CRF) at risk, where CRF alone is a potential primary indicator of mortality versus any other risk factor (Booth, Roberts, Thyfault, Ruegsegger, & Toedebusch, 2017).

According to Piercy et al., 80% of adults and adolescents in the US are “insufficiently active.” A longitudinal study by Gordon-Larsen et al. found that there has been an increase in inactivity (Gordon-Larsen, Nelson, & Popkin, 2004). The investigators found that over a period of time, the majority of adolescents did not achieve a moderate amount of physical activity throughout the week. Throughout adulthood, that trend remained unchanged. Additionally, there is a known decrease in physical activity from age 9, whereby a child is active about 3 hours daily, to the age of 15, whereby an adolescent is active approximately 40 minutes daily (Nader, Bradley, Houts, McRitchie, & O’Brien, 2008). Yet, the recommended amount of physical activity for this age group is a minimum of 60 minutes of moderate-to-vigorous physical activity (MVPA) per day, including aerobic, muscle-strengthening and bone-strengthening activities at least 3 days a week (Nader et al., 2008; Piercy et al., 2018).

In keeping with the theme of this special issue on biological consequences of early-life experience, we advocate the idea that inactivity during development does more than engender obesity, type 2 diabetes and other morbidities. Given the beneficial cognitive effects associated with exercise in children, combined with the evidence from animal models that developmental exercise increases brain volume, and the evidence that exercise in adulthood increases brain infrastructure, a picture emerges in which developmental exercise is necessary in order to produce an adult brain that can withstand the physical, cognitive and psychological challenges of adulthood. At present, however, we know little about the potential negative effects of inactivity on development and neurobehavioral outcomes in adulthood.

5. Knowledge gaps and how to fill them

In Figure 2A we present an overview of the experimental approach that must be undertaken in order to investigate the long-term consequences of developmental exercise on adult brain and behavior. This begins with a period of exercise during development, complemented by an appropriate sedentary control group. Adolescence is most feasible, as animals are by this time strong enough to rotate a standard exercise wheel. Additionally, adolescence may be key in promoting lasting exercise-induced effects as animals at this stage have been weaned but are continuing to experience growth, approximately up to postnatal day 60 (Spear, 2000). During adolescence, sexual maturation (McCormick, Green, & Simone, 2017) and brain development, particularly maturation of limbic system structures and prefrontal cortex (Fu, Rusznak, Herculano-Houzel, Watson, & Paxinos, 2013; Juraska, Sisk, & DonCarlos, 2013; Spear, 2000), are not yet complete. This exercise period could be limited strictly to development or extend into early adulthood. After the exercise period, we suggest a period of inactivity in order to permit assessment of long-term exercise effects on brain structure and function. A period of inactivity is also relevant to the natural age-related decline in human activity levels noted above. In Figure 2B we summarize rodent studies to date that have adopted this general approach, to highlight what is known and the knowledge gaps that need to be addressed. From these studies it is immediately apparent that while the neural and behavioral effects of a transient period of exercise during a critical period of plasticity such as adolescence can persist into adulthood, there remain substantial gaps in our understanding.

Figure 2.

A. Experimental design to investigate the chronic effects of developmental exercise. B. Summary of studies on the neural and/or behavioral effects in adulthood of developmental exercise (EX) following a sedentary (SED) interval. All experiments used experimentally naïve rodents and summarized results reflect comparisons to always sedentary groups, with one exception. *Comparisons include additional continuous (no SED interval) exercise group.

First, we need to broaden the scope of investigation outside hippocampal neurogenesis and, indeed, outside the hippocampus to other brain regions also undergoing maturation and modulated by exercise. Chief among these is the PFC whose maturation underlies the expression of several higher-order cognitive abilities such as working memory, decision-making, inhibitory control, and cognitive flexibility (reviewed in Caballero, Granberg, & Tseng, 2016 and Larsen & Luna, 2018). As such, exercise has the potential to alter the course of PFC development by augmenting dopaminergic neurotransmission (Chen et al., 2016), neurotrophic support (Uysal et al., 2015), as well as dendritic, synaptic, astrocytic, and glial plasticity (Brockett et al., 2015; Tomlinson et al., 2018). Consistent with this, we have shown that developmental exercise induces long-term reductions in impulsivity and processing speed in cranially irradiated and sham-irradiated juvenile rats, relative to sedentary controls (Sahnoune et al., 2018). Second, as most studies have examined only males, we must include both sexes in our investigations, when possible, because adolescence is a phase of increased divergence between the sexes in terms of brain development (Lenroot & Giedd, 2010) and because there are sex differences in voluntary physical activity (Rosenfeld, 2017). Third, it also remains to be determined if a lack of exercise during adolescence is a missed opportunity that cannot be completely recouped at a later stage in life given that, in humans, voluntary physical activity declines with age, particularly during adolescence (Cairney, Veldhuizen, Kwan, Hay, & Faught, 2014; Dumith, Gigante, Domingues,& Kohl, 2011; Nader et al., 2008; Troiano et al., 2008) and early-life physical activity is positively associated with physical activity in adulthood (Malina, 2001). Furthermore, preclinical evidence suggests that exercise-induced modulation of neural plasticity can be both region- and age-dependent. For example, in an investigation of changes in hippocampal mRNA expression, robust increases in the expression of neural plasticity markers were found in adolescent but not adult rats after 7 weeks of voluntary running (O’Leary et al., 2019). This suggests that the conditions present in the juvenile brain are more amenable to positive, exercise-induced changes than the adult brain. Tomlinson et al (2018) found that 4 weeks of voluntary running in juvenile or young-adult mice modulated oligodendrocyte lineage development differentially in the prefrontal cortex and corpus callosum in an age-dependent manner. Exercise also influenced spatial working memory differentially in juvenile and young-adult runners. Others (Soch et al., 2016) have also demonstrated that the inflammatory response in various brain regions differs between adolescent and adult runners, though it is unclear in both these studies how, if at all, observed cellular changes in response to exercise correspond to observed behavioral changes. Finally, in order to observe effects of developmental exercise, it is likely necessary to challenge the adult brain with a cognitive task or stressor. For example, if developmental exercise indeed conditions neural stress circuitry, then it is reasonable to expect that this would only become apparent if a stressor were presented in adulthood, as effects on unchallenged circuitry may not be detectable.

Other considerations regarding the chronic effects of developmental exercise include its potential to shape brain development and function by mitigating long-term negative effects of early-life stress (reviewed in Harrison & Baune, 2014 and Hueston, Cryan, & Nolan, 2017) and via its peripheral actions such as promoting adaptive gut bacteria and metabolic health (Mika & Fleshner, 2016; Mika et al., 2015) and regulating energy homeostasis (Patterson & Levin, 2008).

We have begun preliminary studies that utilize the general experimental approach outlined in Figure 2A. We reasoned that if exercise during development increases brain volume in adulthood, it must also increase the number and/or size of various structural elements. Given the results obtained by Li et al (outlined in section 2.3 above), which suggest longevity of exercise effects on GFAP expression, we examined the effect of developmental exercise on astrocytes in adulthood. Juvenile rats had access to exercise wheels beginning at 35 days of age.

They had access to exercise 5 days/week for 6 weeks and were then sacrificed immediately (“Exercise”), or following 6 additional weeks, during which they were sedentary (“Exercise-rest”). Consistent with many published reports of exercise increasing astrocytes, we found markedly enhanced GFAP expression in animals that exercised during adolescence (see Figure 3). Although an increase in GFAP-labeled astrocyte fibrils appear to remain in adulthood, stereological quantification will be needed to determine whether this is the case.

Figure 3.

Astrocytes in the DG in rats immediately after exercise (Exercise), after a rest period (Exercise-rest) and rats that remained sedentary throughout the entire experiment (Sedentary). Tissue was labeled for cell nuclei (blue), S100β (astrocytes; red) and GFAP (astrocyte fibrils; yellow).

Using the same experimental approach, we assessed the effect of developmental exercise on granule neurons in the adult DG. Exercise has been well-established to increase hippocampal neurogenesis in the adult brain, and more recently this has been shown in the adolescent brain as well (Victorino et al., 2017). Others have shown that if exercise was begun in adolescence but ceased during adulthood, the augmenting effect on neurogenesis did not last (Merkley, Jian, Mosa, Tan, & Wojtowicz, 2014). We reasoned that if exercise during adolescence increases hippocampal neurogenesis, the number of mature dentate gyrus granule neurons would be increased in adulthood. Using the optical fractionator method, we stereologically quantified the number of NeuN+ neurons in the dentate gyrus granule layer, but found no difference between sedentary animals and those sacrificed either immediately after exercise or following exercise + 6 weeks of inactivity (see Figure 4). Ongoing analyses will determine whether developmental exercise results in lasting enhancement of other structural elements, such as dendrites and vasculature.

Figure 4.

Mature neuron (NeuN+) cells in the DG of rats immediately after exercise (Exercise) or after a resting period (Exercise-rest).

Conclusions

In recent years it has become apparent that physical activity during development builds motor as well as cognitive skills, and that physically active children and teens enjoy better health and brain function compared to their inactive peers. What is currently lacking is an understanding of the lifetime neurobehavioral effects of physical activity during development. Given their short life spans and enthusiasm for exercise, laboratory rodents are an ideal means by which to assess this, and emerging evidence from animal models suggests that developmental exercise permanently alters brain size and enhances adult brain function. It is also crucial that we understand the lifetime implications of developmental inactivity, as lack of physical activity during development might translate into lessened ability to undergo activity-dependent plasticity in adulthood. Moreover, it may rob the developing brain of the benefits of exercise for cognition and stress resilience, ultimately leading to an adult brain that is sub-optimally prepared to withstand life’s challenges. Overall, there is a paucity of studies in which exercise occurred during development, ceased and then neurobehavioral outcomes were measured in adulthood. Such studies are badly needed if we are to understand and appreciate the contributions of physical exercise in early life to adult brain health and function.

Highlights:

• Exercise during development increases the volume of the adult brain

• Emerging evidence suggests that exercise during development enhances adult cognition

• Effects of developmental inactivity and adult brain function are under-studied

• Studies on the enduring effects of developmental exercise in adulthood are necessary