Novel Strategies for Healthy Brain Aging

Devin Wahl; Alyssa N. Cavalier; Thomas J. LaRocca


Exerc Sport Sci Rev. 2021;49(2):115-125. 

In This Article

Abstract and Introduction


One of the best strategies for healthy brain aging is regular aerobic exercise. Commonly studied "anti-aging" compounds may mimic some effects of exercise on the brain, but novel approaches that target energy-sensing pathways similar to exercise probably will be more effective in this context. We review evidence in support of this hypothesis by focusing on biological hallmarks of brain aging.


As the world's population ages, the incidence of dementia and neurodegenerative diseases is expected to markedly increase. One key precursor to dementia and neurodegeneration is brain aging itself. Brain aging is characterized by declines in cognitive function (primarily memory and learning, attention/processing speed, and executive function).[1] In some people, these declines may develop into mild cognitive impairment (MCI), which increases the risk for dementia and neurodegenerative diseases like Alzheimer's disease (AD).[2]

Age-related cognitive declines are caused in part by adverse biological events in the brain known as the "hallmarks of brain aging" (Figure 1). This collection of processes mirrors the hallmarks of aging[3] in peripheral tissues in many ways, but it also includes several brain-specific phenomena. Importantly, the hallmarks of brain aging also are involved, and often more pronounced, in most neurodegenerative diseases. Thus, brain aging and neurodegeneration may be part of a continuum, and a combination of hallmarks/risk factors may determine whether dementia or disease develops.[2] As a result, there is growing interest in strategies for inhibiting the hallmarks of brain aging, which could potentially delay or prevent the onset of neurodegenerative diseases.[4]

Figure 1.

Aging is the primary risk factor for the development of cognitive dysfunction, mild cognitive impairment (MCI), and dementia. Numerous, established biological "hallmarks of brain aging" (shown at top) contribute to age-related reductions in cognitive function and increased neurodegeneration/dementia risk with aging.

Regular aerobic exercise is one of the few evidence-based ways to reduce the risk for age-related cognitive decline and neurodegenerative diseases.[5] As such, it has been suggested that exercise should be a first-line strategy for healthy brain aging. Still, there is significant interest in exercise alternatives, including pharmacological "anti-aging" compounds that may have similar or complementary effects (especially in Western societies where sedentary lifestyles are common and many find it difficult to maintain an exercise routine). Many compounds have been tested for their effects on health and longevity in model organisms, and several in particular have been shown to suppress hallmarks of aging in peripheral tissues and increase healthy lifespan in the National Institute on Aging's Intervention Testing Program (ITP).[6] These compounds are some of the most studied in the broad field of aging research and may hold some promise for attenuating hallmarks of brain aging specifically.

The literature on how exercise affects the hallmarks of brain aging is relatively new, and the extent to which select/ITP anti-aging compounds influence the hallmarks of brain aging per se (and whether these agents can mimic the effects of exercise on the brain) also is an emerging topic. However, we and others have shown that exercise remains among the best approaches to reduce hallmarks of aging and increase health/function in peripheral tissues.[4,7] Here, we hypothesize that 1) although most pharmacological agents known to increase lifespan in the ITP may reduce some hallmarks of brain aging, exercise will likely remain a more effective intervention for healthy brain aging because it stimulates key energy-sensing pathways that modulate multiple hallmarks, and 2) novel (non-ITP) compounds or interventions that target energy-sensing pathways similar to exercise will prove to be more effective for promoting healthy brain aging in this context. In the following sections, we review the hallmarks of brain aging and the effects of exercise on them, and we compare these to the effects of key ITP compounds on the same hallmarks as a framework for understanding the importance of energy-sensing pathways in healthy brain aging. Afterward, we discuss newer treatments and ideas for targeting bioenergetic signaling pathways similar to exercise that may hold the most promise as exercise alternatives.

Dysregulated Energy Metabolism: A Central Mechanism

Dysregulated energy metabolism is an underlying contributor to all hallmarks of brain aging. During aging, fasting glucose levels increase as cells become less effective at importing glucose in response to insulin (insulin resistance). Peripheral insulin resistance and elevated fasting glucose are linked with accelerated brain aging, poorer cognitive function, and dementia. Brain function is particularly sensitive to the adverse effects of insulin resistance as glucose is a key energy source for neurons, and this can be compounded by metabolic and cardiovascular changes with aging (e.g., dyslipidemia, increased triglycerides and low-density lipoprotein cholesterol). Consequently, sedentary lifestyles and unhealthy diets that impair peripheral function and metabolic health also are linked with accelerated brain aging.[4,8]

Importantly, exercise is an energetic stress (reduced cellular energy levels) that peripheral tissues and the brain react to with many metabolic, mitochondrial, and cellular responses (known as hormesis). These adaptive responses involve the activation of energy-sensitive pathways that restore bioenergetic homeostasis in both peripheral tissues and in the brain. Key examples include upregulation of glucose transporters (GLUT) and increased signaling through the low-energy sensors AMP protein kinase (AMPK), sirtuin 1 (SIRT1; an NAD+ sensor), and the amino acid–sensing mammalian target of rapamycin (mTOR) pathway (Figure 2).[8,9] The activation of these systems results in enhanced glucose utilization (GLUT), supports neuronal energy levels by regulating glycolysis and respiration (AMPK), controls the activity of transcription factors and energy-metabolizing enzymes (SIRT1), and modulates the turnover of proteins/macromolecules that can be used for energy (mTOR). These signaling proteins also may directly impact neural gene expression, synaptic plasticity, and memory formation. Moreover, prolonged exercise leads to glycogen depletion, fatty acid oxidation, and the production of ketone bodies, which are another important energy source for the brain and can independently activate these same sensors.[8] In addition to all of this, exercise increases the expression of brain-derived neurotrophic factor (BDNF), which powerfully regulates neuron function and development as well as energy intake/behavior and can interact directly with or be activated by energy sensors like AMPK, mTOR, and SIRT1.[10,11] Importantly, all of these signaling pathways 1) contribute to stress resistance by enhancing cellular function and quality control, 2) have been reported to decline with aging, and 3) modulate other hallmarks of aging/brain aging, as emphasized in Figure 2 and described in the following sections.[3,8,9] Thus, the activation of these energy-regulating pathways is central and perhaps the most important mechanism underlying the beneficial influence of exercise on the aging brain.

Figure 2.

Exercise is perhaps the best strategy for inhibiting all major hallmarks of brain aging, largely because it activates key cellular energy-sensing pathways. This enhanced energy metabolism/regulation with exercise is important because the proteins and cellular pathways involved (in bold) can also directly influence other hallmarks. GSH, glutathione; VEGF, vascular endothelial growth factor.

Mitochondrial Dysfunction

Aging is associated with a decline in efficiency of the electron transport chain, resulting in reduced ATP generation and electron leakage that propagates reactive oxygen species (ROS). These ROS can damage cellular structures, including mitochondria and even mitochondrial DNA (mtDNA). This mitochondrial dysfunction is a universal hallmark of aging that impairs overall tissue function, and it is exacerbated by age-related declines in mitochondrial biogenesis and disposal of damaged mitochondria (mitophagy).[3]

Mitochondria are centrally important in the brain, where they play key roles in calcium homeostasis and neuronal metabolism. Numerous studies have shown that mitochondrial dysfunction, especially in neurons, increases with brain aging.[8] Dysfunctional mitochondria also are centrally implicated in neurodegenerative diseases, including Parkinson disease (PD) and AD. However, evidence shows that exercise reduces mitochondrial dysfunction in peripheral tissues and the brain. For example, in old mice, chronic high-intensity running increases mtDNA copy number and enhances mitochondrial biogenesis by activating proliferator-activated receptor gamma coactivator-1α (PGC-1α),[12] which is downstream of the key energy sensors AMPK and SIRT1. Long-term wheel running also activates autophagy (the cellular process of recycling damaged internal components, including mitochondria),[13] and similar experiments in rats show that running enhances mitochondrial network formation and mitochondrial respiration.[14]

As with many of the hallmarks of brain aging, there is limited direct evidence for the effects of exercise on brain mitochondrial function with aging in humans (because access to brain tissue/cells is limited). However, at least one study has shown that hippocampal gene expression patterns in older adults who are more physically active reflect enhanced mitochondrial energy production.[15] As methods for monitoring mitochondrial function in vivo become more advanced, there may be opportunities to demonstrate such cellular effects of exercise more directly.

Accumulation of Oxidatively Damaged Molecules

Oxidatively damaged biomolecules accumulate with aging as a result of oxidative stress, an imbalance between antioxidant defenses and ROS production.[3] These ROS can damage biomolecules like proteins, lipids, and DNA, as well as organelles — all of which can impair cellular function directly (because damaged components do not function properly) or indirectly (because damaged biomolecules may interfere with function in other cellular compartments). Neurons are especially sensitive to ROS because of their high, oxidizable lipid content and rate of oxidative metabolism.[8]

Interestingly, exercise is associated with oxidative stress, as increased metabolism with exercise increases ROS generation. However, this effect is likely hormetic because exercise-associated oxidative stress leads to greater long-term antioxidant capacity.[9] In fact, habitual aerobic exercise is associated with reduced levels of biomolecule damage in the brain. In old rodents, for example, both long-term running and swimming improve cognitive function and reduce ROS, oxidized proteins (carbonyls), and peroxidized lipids in the brain, in part by increasing antioxidants like glutathione and various neurotrophic factors[16] — many of which are modulated by AMPK, mTOR, and SIRT1. Long-term wheel running even reduces peroxidized lipids, improves memory, and protects against AD pathology in transgenic mouse models.[17] Some evidence suggests that these beneficial effects require longer exercise interventions in older mice, which is consistent with the idea that exercise is a hormetic stress on the brain.[18] Although there is little direct evidence for the effects of exercise on oxidative damage in the human brain, clinical studies have shown that exercise increases antioxidant defenses and reduces pro-oxidant processes systemically,[9] and, similar effects likely occur in the brain.

Reduced Molecular Repair and Disposal

One main reason for both mitochondrial dysfunction and oxidative damage accumulation with aging is reduced activity of cellular repair and disposal systems. Many reports have documented systemic, age-related reductions in 1) autophagic-lysosomal degradation of proteins and organelles, and 2) activity of the proteasome (degrading old/damaged proteins). These systems are critical in neurons, which are postmitotic and must maintain their functional capacity throughout the lifespan. Moreover, the accumulation of undegraded, damaged, or aggregated proteins due to reduced autophagy and proteasome activity is central to the pathology of many neurodegenerative diseases (e.g., amyloid beta plaques in AD, alpha synuclein deposits in PD).[2]

Exercise increases autophagy and proteasome activity in both preclinical models and clinical settings,[9] and although there is limited direct evidence for these effects in the brain, rodent studies indicate that swimming enhances autophagy and mitochondrial function in the hippocampus of older animals[19] and may protect against pharmacologically induced "premature" brain aging by activating the same systems.[20] Running also increases proteasome activity in the brain,[21] which may explain reductions in smaller, damaged biomolecules (e.g., protein carbonyls). These effects of exercise on molecular quality control likely involve multiple signaling networks for which autophagy and the proteasome are common downstream effector mechanisms. For example, studies in mice show that exercise activates autophagy in the brain by increasing the activity of AMPK and SIRT1.[13] Exercise also increases mitophagy and enhances mitochondrial network dynamics, in part by activating PGC-1α.[14] In transgenic mice, the activation of these protective quality control systems by exercise may even reduce neurodegenerative protein aggregation.[22]

Dysregulated Calcium Homeostasis

Cellular control of important ions and minerals becomes dysregulated in most aged tissues, but impaired calcium (Ca2+) homeostasis with aging is a particular problem in neurons. Ca2+ plays a role in numerous brain functions, including neurotransmission, neuronal excitability, synaptic plasticity, and long-term memory consolidation. Ca2+ levels also modulate gene expression, as Ca2+ influx through N-methyl-D-aspartate (NMDA) receptors can activate transcription factors like cyclic AMP response element-binding protein (CREB) and PGC-1α. With aging, neuronal Ca2+ homeostasis becomes impaired due to increased Ca2+ influx and dysregulation of intracellular Ca2+ modulators like mitochondria. This leads to altered gene expression and neuronal function that are linked with cognitive decline and can also contribute directly to neuronal death via Ca2+-driven excitotoxicity.[8]

There is limited direct evidence for the effects of exercise on calcium homeostasis in the context of brain aging per se. However, many studies show that exercise increases BDNF, which has been reported to reduce neurotoxic extrasynaptic Ca2+ influx mediated by the NMDA receptor.[23] Some of the beneficial effects of exercise on neuronal mitochondria likely improve Ca2+ homeostasis as well, as exercise enhances mitochondrial Ca2+ handling. In fact, studies in transgenic mice show that exercise increases the expression of the protective mitochondrial sirtuin, SIRT3, and that this is required for protection against Ca2+-related excitotoxicity.[24] Studies of neuronal Ca2+ homeostasis in humans are lacking, but there is strong interest in related systemic biomarkers like S100 calcium-binding protein β (S100β), which is linked with brain aging, injury, and neurodegeneration, and is reduced by exercise.[25]

Impaired Adaptive Cellular Stress Response

All cells activate stress resistance networks (e.g., antioxidant and anti-inflammatory pathways) in response to stressors. In the brain, these adaptive responses are important in settings of electrochemical, ionic, and even psychological stress. With aging, however, most cells become markedly less effective at mounting adaptive stress responses.[3] In neurons, this is partly due to age-related decreases in the expression/activity of stress sensors like AMPK and SIRT1, and protective neurotrophic factors, such as BDNF and nerve growth factor (NGF). These proteins stimulate gene expression that promotes antioxidant defenses, healthy mitochondrial function, and Ca2+ handling.[8]

As a hormetic stress itself, exercise activates adaptive cellular stress responses, and it increases the ability of these same systems to respond to other stressors. The metabolic stress associated with exercise activates antioxidant systems in mouse neurons, such as those controlled by nuclear factor erythroid 2-related factor 2 (Nrf2), and these protect against aging-relevant neurotoxic stressors.[8,26] This exercise-induced Nrf2 activation increases endogenous antioxidants like catalase and superoxide dismutases that directly reduce ROS, and it activates other protective processes (e.g., autophagy) that further protect against oxidative stress. Studies in aged rodents also show that exercise potentiates stress response systems that protect neurons against inflammatory insults, in part by activating BDNF and anti-inflammatory cytokines like interleukin 10 (IL-10).[27]

In humans, there is some evidence that routine exercise increases tolerance for psychological stressors with aging, but direct data on cellular stress responses are limited. Still, many clinical studies have shown that aerobic exercise increases the activity of systemic antioxidant and anti-inflammatory defense systems, and some have even demonstrated that these effects are associated with increases in BDNF and reductions in S100β.[28]

Aberrant Neuronal Network Activity

The brain relies on communication among billions of neurons. Optimal function of neuronal networks requires balanced activity of glutamatergic (excitatory) neurons and GABAergic (inhibitory) interneurons. However, during aging, activity within these circuits (largely white matter communication via myelinated axons) is dysregulated, which can result in hyperexcitability and excitotoxic damage. This may lead to degeneration of fiber systems involved in decision making and learning/memory.[8,26]

Some studies have investigated the influence of exercise on neuronal network activity. One showed that pharmacologically induced brain aging is mitigated by swimming, which reduces neurotoxicity and improves synaptic transmission by influencing the expression of glutamate-receptor 1 and synaptophysin (a marker of synaptic integrity) in rats.[29] Exercise may also confer neuroprotection by modulating GABA disinhibition[26] and by enhancing dendritic outgrowth, maintaining structural integrity, and improving long-term potentiation in aged animals via BDNF and other proteins also involved in energy sensing. These changes coincide with increased hippocampal glutamate synthesis and an increase in NGFs.[30]

In humans, there is evidence that exercise may positively influence neuronal transmission and connectivity. For example, greater cardiovascular fitness in older adults is associated with improved prefrontal cortex (important for decision making) processing speed and heightened synaptic plasticity,[31] and meta-analyses find that exercise is associated with greater hippocampus volume and may influence network functional connectivity in this region.[32]


Neuroinflammation is characterized by increased numbers of immune-activated, proinflammatory astrocytes and microglia that secrete neurotoxic cytokines.[26] This glial cell activation is linked with brain aging, MCI, and most neurodegenerative diseases. In fact, markers of neuroinflammation and related neurodegeneration [e.g., glial fibrillary acidic protein (GFAP) and neurofilament light chain (NFL)] even have been shown to increase with aging in cognitively unimpaired people.[2] Evidence shows that exercise decreases brain inflammation, marked by lower levels of the proinflammatory transcription factor nuclear factor κB and tumor necrosis factor α (TNF-α), and that these changes are associated with improved spatial memory.[26] Exercise even prevents obesity-induced cognitive decline in rodents by influencing inflammatory genes, reducing the number of reactive astroglia and microglia, and reducing major markers of hippocampal inflammation.[33] Finally, in old mice, running reduces brain proinflammatory cytokine levels, in part by activating BDNF and RE1-silencing transcription factor (REST), which is linked with numerous neurological disorders.[34]

In humans, peripheral inflammation has been linked with brain structural abnormalities, reduced cognitive function, and greater dementia risk. As an example, recent studies demonstrate that high levels of C-reactive protein (CRP; a common clinical marker of inflammation) are associated with reduced brain white matter integrity in older adults,[35] and epidemiological data show that systemic inflammation may precede neurodegenerative diseases by decades. However, meta-analyses show that habitual exercise is associated with reduced proinflammatory cytokines including CRP and TNF-α.[36]

Impaired DNA Repair

The accumulation of DNA damage is an important hallmark of aging, and it plays a causal role in several premature aging syndromes. DNA damage also is closely linked with brain aging, cognitive decline, and neurodegenerative diseases.[8] The genomic damage that accumulates with aging can contribute to cellular senescence (in which cells cease dividing and begin to produce proinflammatory molecules) and apoptosis (programmed cell death), leading to functional decline of organs/systems (including the brain) and reduced longevity.[3]

The mechanisms by which exercise reduces DNA damage during aging are not well understood, but evidence points toward several DNA repair pathways, many of which are modulated by SIRT1 and other energy sensors.[37] Rodent studies have shown that running stimulates DNA repair in brain tissue by upregulating the expression of CREB and apurinic/apyrimidinic endonuclease 1, which is a key enzyme for DNA base excision repair.[38] Treadmill exercise has also been shown to reduce hippocampal DNA fragmentation and neuronal apoptosis in rat models of traumatic brain injury.[39] Moreover, in animal models of AD, treadmill exercise is associated with reduced DNA damage and fewer double-stranded breaks in the hippocampus.[40]

Impaired Neurogenesis/Stem Cell Exhaustion

Most neurons do not proliferate, but the dentate gyrus of the hippocampus, subventricular zone of the lateral ventricle, and the olfactory bulbs all contain stem cells.[41] As is the case with most stem cells, the ability of these stem cells to self-renew and generate new progenitors (neurogenesis) declines with aging, which may account for some of the structural declines observed in the aging brain.[8] However, exercise enhances neurogenesis within these areas, and there is strong interest in determining underlying mechanisms in the hippocampus, which is crucial for learning and memory.

To date, most studies on exercise and hippocampal neurogenesis have been in rodents. Some studies have demonstrated that running induces neurogenesis in the hippocampal dentate gyrus and is associated with increased neurogenesis and improved memory.[42] Others have investigated the influence of long-term exercise on hippocampal neurogenesis in old mice and found that exercise increases neurogenesis and improves neuronal structure,[43] and these results likely linked with increased activity of BDNF. In humans, neurogenesis may correlate with cerebral blood flow, which has been shown to increase with exercise.

Other: Telomere Attrition and Cellular Senescence

Telomeres are repetitive DNA sequences that cap the ends of chromosomes and protect DNA from degradation. They also protect overall cellular function, as critically short telomeres can lead to cell death or senescence. Telomere shortening and cellular senescence are established hallmarks of aging in peripheral tissues. Their role in brain aging is less clear, but telomere maintenance, which contributes to the production of neuronal stem cells, may protect against neurodegenerative diseases — and there is evidence that habitual exercise is associated with longer telomeres and reduced cellular senescence in both mice and humans[3,8,9]