Resistance Exercise-Induced Responses in Physiological Factors Linked with Cognitive Health

Atherosclerosis and Alzheimer’s disease

Atherosclerosis, a blood vessel disease of increased vessel stiffness, internal thickness, and inhibited vascular tone [23], is a key risk factor for the development of AD [5]. Individuals are three times more likely to develop AD in their lifetime if affected by atherosclerosis, the extent of which is strongly associated with AD symptom severity [24]. Whether AD accelerates atherosclerosis, or vice-versa remains equivocal; however, as atherosclerosis can develop as early as childhood [25], it is likely that atherosclerosis plays an important role in the onset and progression of AD pathophysiology and cognitive impairment [5].

Over time, atherosclerotic carotid and cerebral arteries attenuate cerebral blood flow (CBF) [26, 27] and cerebral perfusion, [5] triggering a cycle of hypoperfusion, hypometabolism, and critical neuronal degradation (Fig. 1) [28]. Atherosclerosis of the cerebral arteries may accelerate amyloid-β (Aβ) deposition inside and around the cerebral arteries, a condition termed cerebral amyloid angiopathy (CAA). CAA severity is associated with a faster decline in perceptual speed, episodic memory, and semantic memory in older individuals, [29] and is correlated with dementia severity (r _s [Spearman] = 0.57, p < 0.05) and identified biomarkers (neurofibrillary tangle [r _s = 0.62, p < 0.01] and Aβ plaque burden [r _s = 0.74, p < 0.01]) in postmortem tissue [30]. Aβ is a hallmark characteristic observed in the brains of individuals with AD and is hypothesized to be a primary cause of neuronal degeneration [31]. The role of Aβ in AD pathogenesis is strongly supported by cases of rare genetic mutations in the transmembrane proteins amyloid-β protein precursor, presenilin 1 and 2, which directly result in Aβ overproduction [32, 33]. Nevertheless, in AD cases not caused by one of these genetic mutations (i.e., most late-onset AD cases), it remains to be elucidated as to whether high levels of brain Aβ can be attributed to overproduction of Aβ, or decreased degradation and/or clearance. Regardless, accumulation of toxic Aβ peptide resulting from imbalances between production and elimination [34] can promote cerebral atherosclerosis, perpetuating the cycle of Aβ accumulation and vascular dysfunction [35].

OPEN IN VIEWER

Fig.1

Effect of aging and neurodegenerative disease on atherosclerosis, and the proposed direct () and potential downstream () influence (↑= increase; ↓= decrease; ? = unknown effect) of resistance training. Hypothetical pathways are marked (). Aβ, amyloid-β; CAA, cerebral amyloid angiopathy.

Eliminating Aβ from the brain is dependent on specialized pathways requiring extracellular drainage [36, 37] of interstitial fluid (ISF) and cerebrospinal fluid (CSF) via the blood-brain barrier and glymphatic (glial and lymphatic) system [36]. Glymphatic drainage occurs along artery and capillary basement membranes (peri-vascular pathway), [38] and immediately surrounding smooth vascular muscle carrying aquaporin channels (para-vascular pathway) [36]. The movement of ISF and CSF along the peri- and para-vascular pathways is hypothesized to be reliant on arterial pulsations [36, 39]. Within this model, the peri-vascular space is compressed during each pulse cycle with the resultant pressure wave clearing solutes like Aβ [40]. Importantly, elasticity-dependent arterial pulsations are compromised in atherosclerotic and aging vessels [39]. Consequently, older adults and especially those with atherosclerotic cerebral arteries are likely to demonstrate reduced elimination and greater accumulation of toxic Aβ, [36 , 41] accelerating the development of AD symptoms over time.

Atherosclerosis, resistance exercise, and resistance training

Data for the influence of resistance exercise and training on atherosclerosis are limited [42, 43]. As techniques to evaluate this relationship, flow-mediated dilation and reactive hyperemia (RH) [44] have provided interesting results. In healthy young adults, following an acute bout of resistance exercise (eight muscle groups; 3×10 repetitions [reps] at 10-repetition maximum [RM]) greater change in RH, indicative of greater vascular compliance, is observed compared with aerobic exercise, despite elevations in arterial stiffness [42]. A greater RH response observed after resistance training in this study was likely a compensatory mechanism to maintain homeostasis in response to elevated arterial stiffness [42]. Supporting this hypothesis, Lefferts et al. [45] observed increased (37%) carotid artery stiffness but no negative effect on downstream blood flow or pulsatility in healthy young males after an acute bout of resistance exercise (bench press and bicep curl resistance exercise at 5 to 10 RM, respectively).

While it is unlikely that resistance training protects against vascular dysfunction through alteration in arterial compliance, evidence does support the positive impact of resistance training on adipose-derived signaling proteins (i.e., adipokines). Varady et al. [46] observed increases in adiponectin (30–37% increase), a cytokine which enhances regional vasodilation, [47, 48] and a decrease in resistin (35% decrease), a protein which promotes oxidative stress and inhibits regional vasodilation, [49] following four sets of near-maximal leg press in experienced lifters when compared with their sedentary counterparts (Table 1). Between group differences were potentially due to greater fat-free mass (10.2%), and lower adipose mass (–18.2%) in the trained compared with sedentary group. Indeed, murine studies demonstrate that adiponectin is expressed in skeletal muscle fibers [50, 51]. Although this finding is yet to be confirmed in humans, greater skeletal muscle mass could contribute to increased secretion of adiponectin following muscle contraction. Alternatively, trained lifters demonstrate greater muscle contractile capacity and fiber recruitment when compared to untrained lifters [52]. It is possible that a greater contractile force results in enhanced contraction-induced secretion of adiponectin as observed in myokines (i.e., muscle-derived signaling proteins) [53]. The mechanism behind changes in resistin following acute resistance exercise also remain unknown; yet, may be related to elevations in anti-inflammatory, or decreases in pro-inflammatory mediators [54].

Cerebral blood flow and Alzheimer’s disease

Reduced CBF [55] is linked to cerebral hypoperfusion, decreased nutrient transport, [56] hypometabolism, and cognitive decline [57]. Declining 1% per year in healthy adults, [58] CBF velocity is significantly correlated (r = unspecified, p = unspecified) with disease severity in individuals with AD [59]. When compared with age-matched healthy controls, regional CBF in the parietal and temporal regions is ∼30% lower in individuals with AD, and up to 15% lower in the thalamus, basal ganglia, and white matter [57]. Furthermore, relationships between resting CBF in several brain areas (hippocampus: r _p [Pearson] = –0.51, p < 0.05; frontal lobe: r _p = 0.35, p < 0.05; temporal lobe: r _p = 0.46, p < 0.01; parietal lobe: r _p = 0.40, p < 0.05; occipital lobe: r _p = 0.47, p < 0.01) and cognitive performance have been reported [55, 60]. Marshall et al. [61] demonstrated this relationship in middle-aged individuals observing a reduction in CBF and a 10-fold decrease in attention during cerebral artery balloon occlusion, with a return to normal for both measures upon deflation.

Decline in CBF associated with AD is dependent on disease severity, duration of illness, and age at disease onset [57]. Although the mechanisms responsible for reduced CBF remain equivocal, greater blood viscosity, [62] blood pressure changes [63], or cerebral atrophy [64] likely contribute individually and collectively. Poorer cognitive performance has been observed in older males with higher blood viscosity when compared to individuals with normal values [65]. Importantly, blood viscosity increases with age [66] and is influenced by red-cell aggregation, white-cell mass, plasma fibrinogens, immunoglobulin, and red cell rigidity [67]. In the presence of AD, these age-related rheological changes can become more pronounced [56]. For example, fibrinogen increases in concentration with age [66] and is associated with cognitive decline (r _p = 0.17, p < 0.05) in individuals with mild cognitive impairment, and presents as a risk factor for the development of AD and dementia [68, 69]. Paradoxically, both hypo- and hypertension can negatively influence CBF [63] and have been identified as risk factors for AD [70, 71]. The mechanisms between hypotension on AD risk are poorly understood, while hypertension reduces CBF indirectly by disrupting cerebral autoregulation and promoting detrimental vascular remodeling, oxidative stress, and inflammatory responses, [63] paving the way for white matter degeneration [72].

An alternative hypothesis by Lucas et al. [64] suggests that age-related CBF decline likely reflects cerebral atrophy because less active neural mass demands less oxygen and nutrients (i.e., a hypometabolic state). However, local blood flow to neuronal tissue can become augmented intrinsically in response to metabolic demand, [73] suggesting that although hypometabolism may contribute to further decline in CBF, it may not be a root cause for CBF reductions [74]. Indeed, low CBF is present in individuals with mild dementia where cerebral atrophy and neuronal loss is minimal [75]. Furthermore, cerebral atrophy is accelerated by chronically reduced levels of CBF [76]. It is therefore more likely that chronically lowered CBF is a major contributing factor to the development of AD and dementia rather than a result of disease progression [77].

Cerebral blood flow, resistance exercise, and resistance training

Within the available resistance training data [45 , 78–80], blood velocity through the middle cerebral artery (MCA) is a frequent outcome measure. Indeed, the MCA supplies the greatest quantity of total blood volume to the brain among all cerebral arteries [81]. Using barbell back squats (30–90% of 1RM), Perry et al. [78] observed a consistent 31% elevation in mean MCA blood flow velocity across conditions, despite significant load-dependent increases in mean arterial pressure. No change in MCA blood flow velocity has been observed during acute sets of leg curl [79], bench press and bicep curl [45], and 10RM leg press [80], despite significant increases in heart rate, common carotid artery stiffness [45], and blood pressure. Stability of CBF throughout sub-maximal resistance exercise is likely a reflection of cerebral autoregulation [82, 83]. Indeed, fluctuations in mean arterial pressure (within pressures of 50–170 mmHg [82, 84]) during acute resistance exercise mirror changes in MCA blood flow velocity, [80, 85] indicating preserved cerebral autoregulation. It is possible that during upright high intensity or maximal resistance exercise, in which mean arterial pressure can exceed 170 mmHg, cerebral autoregulation is compromised leading to abnormal blood flow velocities. For instance, in competitive weightlifters, reductions in MCA blood flow velocity were observed during maximal leg press [86]. This style of resistance exercise is consistent with extremely high mean arterial pressures, especially when individuals utilize the Valsalva maneuver, [87] possibly leading to these findings [86, 88].

To date, one resistance training intervention has examined the influence of this exercise modality on CBF velocity [89]. However, the inclusion of aerobic exercise with the resistance training intervention limits its conclusions. In postmenopausal women, 16 weeks of bodyweight resistance training twice a week resulted in no improvement in blood flow velocity when compared to a control group [89]. Although not fully understood, [90] the loss of estrogen in postmenopausal women may contribute to the lack of vascular adaptation [91]. As a vasoprotective hormone, estrogen rapidly stimulates acute vasodilation and cyclic guanosine monophosphate activation [92]. In the longer term, estrogen regulates vascular tone, stimulates local vascular growth factor expression, and protects against atherosclerosis [92]. It is possible that an age-related decrease in estrogen inhibits the exercise-induced adaptations in CBF. To the best of our knowledge, no studies have examined changes in CBF following resistance training in male subjects.

Brain-derived neurotrophic factor and Alzheimer’s disease

Neuroplasticity is the overarching process of structural and functional brain remodeling in response to neuronal activity, injury, death, and growth [93]. However, the regions of the brain that exhibit the greatest structural plasticity such as the limbic cortex and non-primary association cortex are the most vulnerable in AD, likely due to failure in neuronal regulators of synaptic plasticity [94]. As a process of neuroplasticity, neurogenesis is the formation of neurons from neural stem and neural progenitor cells, occurring in three areas of the adult mammalian brain: the subventricular zone of the lateral ventricles for interneuron formation, the basolateral amygdala, and the dentate gyrus of the hippocampus [95, 96]. Hippocampal neurogenesis is the pathway to memory formation, and is one of the earliest brain regions affected by severe atrophy in AD [97]. Found in particularly high concentrations in the hippocampus, [98] brain-derived neurotrophic factor (BDNF) is the most widespread neurotrophin integral to neurogenesis and synaptic plasticity [99]. Through its high-affinity tropomyosin-related kinase B (TRK-B) receptor, [100] BDNF plays a major role in the growth, proliferation, survivability, and maintenance of various neurons, [101] including those in the hippocampus, cortex, and in root ganglion cells [102, 103]. Furthermore, BDNF is involved in the regulation of axonal and dendritic guidance [104] and may have bi-directional transport across the blood-brain barrier [105, 106]. In vivo and in vitro evidence also indicates BDNF-mediated neuroprotection against Aβ toxicity in rats [107]. The expression of BDNF may be mediated in part by the osteoblast-secreted hormone osteocalcin. Indeed, osteocalcin stimulates BDNF expression within hippocampal neurons in cell cultures, promotes BDNF synapse transport in rat neurons, and elevates hippocampal BDNF expression following peripheral administration in mice [108].

Normal physiological levels of BDNF outside of the brain have not been verified in humans, with large basal peripheral variations between 1.5 ng·mL^–1 and 30.9 ng·mL^–1 reported in healthy individuals [109, 110]. In individuals with AD, serum BDNF levels are elevated during the earliest stages of the disease (some 21%) in a possible compensatory response to mild neurodegeneration [111]. In late-stage AD, BDNF transcription deficiencies are evident in several cortical areas, [99 , 112–114] and BDNF protein in serum can be ∼31% lower compared with individuals in the early-stages of AD [111]. Postmortem tissue from AD brains show reduced BDNF mRNA in areas vital for memory formation and learning; the hippocampus and cortex [115 –118] Indeed, BDNF immunoreactivity in individuals with AD is approximately 70% lower in the dentate gyrus, 90% lower in the CA1 subregion, and 90% lower in the temporal cortex when compared to counterparts of a similar age [114].

General aging and genetics likely contribute to the decline in cognitive function through altered BDNF expression (Fig. 2). Lommatzsch et al. [119] observed a decrease in plasma BDNF that negatively correlated with advancing age (r _s = –0.20, p < 0.05). This finding was reproduced in a larger scale investigation (n = 259) of healthy older adults, during which serum BDNF demonstrated a negative correlation (r _s = –0.15, p = 0.02) with age [120]. Importantly, yet unverified in humans, a strong association exists (r _p = 0.81, p < 0.01) between basal peripheral levels and cortical levels of BDNF in rats [121]. In primates, the normal aging process is associated with reduced BDNF mRNA in the cortex and hippocampus [122]. Moreover, reduced TRK-B expression is observed in the pituitary gland of aging rats, which is potentially linked to aging-related endocrine dysfunction [123]. Independent of the aging process, certain genetic polymorphisms can negatively impact BDNF. A common single nucleotide polymorphism (SNP) for BDNF exists, whereby valine (Val), an amino acid responsible for the proper encoding of BDNF, becomes substituted with methionine (Met), a sulphur-containing amino acid, at codon 66 (i.e., Val66Met SNP). Met substitution impairs activity-dependent secretion and intracellular distribution of BDNF in hippocampal neurons [124]. Furthermore, Met carriers have poorer episodic memory, hippocampal function, and reduced prefrontal and temporal lobe grey matter volume when compared with Val carriers [124, 125].

OPEN IN VIEWER

Fig.2

Effect of aging and neurodegenerative disease on brain-derived neurotrophic factor expression, and the proposed direct () and potential downstream () influence (↑= increase; ↓= decrease) of resistance training. Hypothetical pathways are marked (). Aβ, amyloid-β; BDNF, brain-derived neurotrophic factor; SNP, single-nucleotide polymorphism; TRK-B, tropomyosin-related kinase B.

Brain-derived neurotrophic factor, resistance exercise, and resistance training

The influence of acute resistance exercise on BDNF expression in various mediums remain equivocal with reports of no effect [126 –128] or highly variable (13–100%) increases in BDNF following a single bout of high intensity resistance exercise in young adults (Table 1) [129 –132]. It is possible that the influence of resistance exercise on BDNF is intensity dependent, [133] an idea supported by previous aerobic-based studies [134]. As an example, Rojas Vegas et al. [132] investigated the acute BDNF response following 50 reps, without rest, of single-legged leg extension at low and high intensities (40–110% of maximal effort) in eight healthy young males. Mean BDNF concentrations were approximately 40% greater immediately post-high intensity (effect size ∼0.85), and 4% lower immediately post-low intensity leg extension when compared to baseline, indicating a potential difference between conditions based on intensity. However, variability in the measure combined with a small sample size likely contributed to the reported non-significant difference from baseline within the high-intensity condition.

Following longer-term resistance training (Table 2), many studies show no changes in resting plasma or serum BDNF [127 , 136]. Unlike acute resistance training studies, it does not appear that exercise intensity is the reason for the lack of change as the highlighted interventions appear to be adequate in intensity, frequency, and volume. However, in all but one [135] of the above-mentioned studies, the participants were young healthy adults. In a study of 56 older individuals, high (2×10–15 reps at 80% of 1RM), low (1×80–100 reps at 20% of 1RM), and low+ (1×60 reps at 20% of 1RM, plus 1×10–20 reps at 40% of 1RM without rest) intensity resistance training was performed three times per week for 12 weeks. Males in the low+ group alone experienced a 20% increase in resting BDNF from baseline; a change greater (p < 0.05) than observed in the high and low intensity groups where non-significant BDNF reductions were seen [137]. Although the programs appear similar for total volume performed, the low+ program incorporated a “superset” element maximizing muscle fatigue, [138] likely leading to a greater BDNF response. In frail older females, Coelho et al. [139] measured a 65.2% increase in plasma BDNF following 10 weeks of lower extremity resistance training three times per week. Consistent with this finding, Pereira et al. [140] observed an approximate 20% increase in plasma BDNF following 10 weeks of full-body resistance training three times per week at 50% to 75% of 1RM.

Insulin-like growth factor 1 and Alzheimer’s disease

The primary function of insulin-like growth factor 1 (IGF-1) is the growth and maintenance of skeletal muscle, [141] particularly in developmental stages of life. Surprisingly, IGF-1 also plays a direct role in the survival of newly generated neurons [142] and the growth and branching of dendrites [143]. Indirectly, sufficient levels of IGF-1 in circulation are necessary for the maturation of pro-BDNF, an immature precursor, resulting in a greater abundance of mature BDNF [144]. Readily transported across the blood-brain barrier, [145] IGF-1 interacts with target tissues through its high-affinity binding protein, insulin-like growth factor binding protein 3 (IGFBP-3), which is responsible for approximately 80% of circulating insulin-like growth factors [146]. Serum IGF-1 concentrations in healthy individuals are age-dependent, peaking at 500 ng·mL^–1 at approximately 15 years of age before steadily declining by more than 80% by the age of 85 [147]. When comparing individuals with AD to healthy controls of the same age, increased [148, 149] and decreased [150, 151] IGF-1 levels in serum and plasma have been reported. Importantly, only one of the aforementioned studies also examined IGFBP-3, which reflects total IGF-1 bioavailability and impact [152, 153]. Although this investigation reported no difference in plasma IGFBP-3 between AD and healthy controls, the ratio between plasma IGF-1 and IGFBP-3 was ∼16% greater in individuals with AD. Greater IGF-1 bioavailability in individuals with AD may indicate a compensatory response to cells becoming resistant to IGF-1 signaling [154] or reduced expression of IGF-1 receptors [148, 155]. It is possible that different mechanisms are involved in instances of abnormally low resting IGF-1, which may reflect dysfunction in the transcription of IGF-1 mRNA. Indeed, individuals with AD can have 80% lower levels of IGF-1 mRNA when compared to a healthy counterpart, [155] diminishing total IGF-1 in circulation.

Regardless of the mechanism, IGF-1 likely plays a significant role in the pathogenesis of AD and long-term neuroprotection (Fig. 3). For instance, increased cerebral Aβ toxicity, tau phosphorylation, and Aβ accumulation are observed following IGF-1 inhibition within the choroid plexus [154]. Since IGF-1 is known to modulate the amyloid carrying lipoproteins albumin and transthyretin, [154] inhibition of IGF-1 disrupts this amyloid regulation pathway. Further, low IGF-1 in animal models is highly correlated (r _unspecified = 0.77, p < 0.05) with cerebral microvasculature degeneration, [156] and associated with impaired cognition [156, 157] and reduced hippocampal neurogenesis [158]. Interestingly, microvascular density and cognition is improved in IGF-1 deficient mice following central nervous system IGF-1 administration [159]. An early investigation in middle-aged males found that higher unbound plasma IGF-1 concentrations trends toward greater cognitive performance later in life [152]. More recently, middle-aged and older individuals with serum IGF-1 in the lowest quartile were shown to be 51% more likely to develop AD [160]. In contrast, higher IGF-1 levels are associated with greater total brain volume in cognitively healthy individuals, [160] and greater cognitive performance in individuals with dementia [161].

OPEN IN VIEWER

Fig.3

Effect of aging and neurodegenerative disease on insulin-like growth factor 1 expression, and the proposed direct () and potential downstream () influence (↑= increase; ↓= decrease) of resistance training. Hypothetical pathways are marked (). Aβ, amyloid-β; BDNF, brain-derived neurotrophic factor; IGF-1, insulin-like growth factor 1; IGFBP-3, insulin-like growth factor binding protein 3.

Insulin-like growth factor 1, resistance exercise, and resistance training

With several key actions in the central nervous system, IGF-1 is a likely mediator for improvements in cognitive function following resistance exercise and training [13, 162]. Following a single bout of resistance exercise (Table 1), a ∼15% increase in serum IGF-1 has been reported in young healthy females (p < 0.05) performing 3-12RM loads [163]. In young healthy males, IGF-1 increases were observed following protocols with varied set, repetition, and load schemes [164]. In older individuals with mild cognitive impairment, serum IGF-1 is elevated by ∼15% following intense resistance exercise in comparison to a control group [128]. Inconsistent with these findings, Manini et al. [165] and Rojas Vega et al. [132] observed no change in IGF-1 following knee extension exercise in young and older males. It is possible that the total training volume or the use of a single-joint exercise alone (i.e., isolated muscle mass), was not a great enough stimulus for acute IGF-1 changes.

The majority of evidence supports a beneficial effect of resistance training on IGF-1 (Table 2) [141 , 166]. An early investigation by Fiatarone-Singh and colleagues [166] observed some 500% increase in intra-muscular IGF-1 in the vastus lateralis of older adults undertaking a 6-month progressive resistance training program two to three days per week. Within this study, major muscle groups were exercised for 3 sets of 8 reps at an undeclared load intensity, and changes in IGF-1 were associated with muscle damage (r _unspecified = 0.85, p = 0.03) and developmental myosin (a marker of muscle regeneration) [166]. These findings indicate that a degree of exercise-induced muscle damage may be necessary for increases in intramuscular IGF-1; however, it remains undetermined whether the IGF-1 produced within muscle fibers is capable of reaching tissues such as the brain. In young middle-aged adults [141] and older males [13], IGF-1 significantly increased when exercising at 50–80% of 1RM during ∼25-week interventions. Interestingly, Borst et al. [141] observed a peak in IGF-1 levels at 13 weeks, indicative of a possible ceiling effect, or a decline in IGF-1 as a consequence of excessive training load [167]. In comparison, no changes in serum IGF-1 in healthy young adults, [127] and decreased plasma IGF-1 in young [136] and older adults [168] are reported following 10 to 12 weeks of resistance training. Together, the above findings indicate that IGF-1 is responsive to resistance training and that intensity of exercise is a key determinant for IGF-1 upregulation.

Vascular endothelial growth factor and Alzheimer’s disease

Angiogenesis, the stimulated growth of capillaries, [169] requires endothelial cell division, basement membrane degradation, and endothelial cell migration to form new vessels from pre-existing vasculature. Angiogenesis is physiologically regulated by hypoxia, [170] muscle contraction, fluid shear stress [171], and changes in metabolism [172]. Under normal circumstances, angiogenesis plays a role in wound healing and tissue regeneration, but may occur in pathologies such as tumor growth or rheumatoid arthritis [173]. Disrupted angiogenesis is a proposed mechanism explaining the progression of AD pathology [174]. Aging is accompanied by decreased cerebral angiogenesis, [175] diminished responsiveness to hypoxia, and reduced expression of angiogenic growth factors [176]. Furthermore, reduced angiogenesis in individuals with AD is associated with cerebral vasoconstriction, blood vessel degeneration, and CAA [177].

Vascular endothelial growth factor (VEGF) is a potent stimulator of endothelial cell proliferation, sprouting, and blood vessel survival [178, 179] and is capable of inducing endothelial vasodilation in a dose-dependent manner [180]. Secondarily, VEGF stimulates neural growth, survival, and function [95]. Administration of VEGF in animal models promotes hippocampal angiogenesis and cognitive improvement, while reducing toxic Aβ deposition and hyperphosphorylated tau [181, 182]. However, it is unknown whether VEGF freely crosses the blood-brain barrier [183]; thus, peripheral changes may not affect central neuronal targets directly. Regardless, in animal models, peripheral VEGF is suggested to be an essential downstream mediator of adult neurogenesis [184]. It is hypothesized that peripheral VEGF modulates hepatic expression of IGF-1, known to cross the blood-brain barrier [145] and promote neural growth [184]. In individuals with AD, VEGF can be 30% lower in serum [185] and has the propensity to bind directly to Aβ plaques, [186] an action that likely reduces VEGF bioavailability and cerebral perfusion [181]. In addition, Aβ plaque burden in temporal region capillaries of AD affected brains has been inversely correlated with expression of both VEGF (r _s = unspecified, p < 0.01) and endothelial nitric oxide synthase ([eNOS]; r _s = unspecified, p < 0.01), [187] an enzyme integral to vascular tone.

Of the VEGF family (VEGF A-F and placental growth factor), VEGF-A has the strongest link to angiogenesis, neurogenesis, and neural maintenance [95]. Although predominantly angiogenic, VEGF-A induced blood vessel growth can enhance neurogenesis in tissues local to the growing vessels [95]. Mackenzie et al. [95] presents two hypotheses for this process: that VEGF-A instigates neurogenesis through the stimulation of neural progenitor cells directly; or the blood vessels acted upon by VEGF-A release neurogenic compounds (Fig. 4). Indeed, in animal models, blood vessels stimulated by VEGF-A promote the release of BDNF [188, 189]. Interestingly, VEGF-A inhibition leads to disrupted neurogenesis, cognitive function and capillary growth [190, 191].

OPEN IN VIEWER

Fig.4

Effect of aging and neurodegenerative disease on vascular endothelial growth factor, and the proposed direct () and potential downstream () influence (↑= increase; ↓= decrease) of resistance training. Hypothetical pathways are marked (). Aβ, amyloid-β; BDNF, brain-derived neurotrophic factor; IGF-1, insulin-like growth factor 1; VEGF, vascular endothelial growth factor.

In humans, hypoxia, growth factors (e.g., tissue growth factor-β), inflammatory cytokines (e.g., tumor necrosis factor-α), active muscle contraction, and fluid shear stress [171, 192] regulate VEGF expression. Myocytes, a cell abundant in muscle tissue, contain high concentrations of VEGF, which may be released following muscle contraction-mediated agitation [193, 194]. In skeletal muscle, greater extracellular VEGF expression following muscle contraction is partially dependent on interstitial adenosine content [192]. Indeed, adenosine infusion upregulates VEGF from myocytes or endothelial cells through interaction with the angiogenic adenosine A_2B receptor [192]. When implemented concurrently, muscle contraction and blood-flow restriction, a method used to maximize local muscle hypoxia, can increase VEGF expression in serum by some 150% [195]. In addition to hypoxic stimulation of VEGF, high levels of blood flow, and the resultant shear stress on the endothelial wall, upregulate VEGF expression in an intensity-dependent manner [196]. The expression of VEGF mRNA is ∼77% greater when human endothelial cells are exposed to high fluid stress when compared to low fluid stress [196].

Vascular endothelial growth factor, resistance exercise, and resistance training

Few studies have examined acute or chronic influence of resistance exercise on VEGF, with the current findings equivocal (Table 1). No changes in serum VEGF are observed following intense resistance exercise in healthy young adults [132] or older adults with mild cognitive impairment [128]. In contrast, elevated VEGF-A (∼38%) has been observed 10 minutes after completing a high intensity resistance exercise session [197]. While following a single session of double-legged knee extension (3×10 reps at 60–80% 1RM with 120 s recovery) greater concentrations of serum and muscle VEGF were observed at two (∼44%; p = 0.02 and 20%; p = 0.02, respectively) and 4 hours (∼78%, p = 0.02 and 16%; p = 0.02, respectively) post-session when compared with baseline [198]. In contrast, no changes in serum VEGF were reported by Rojas Vega et al. [132] following a single-legged knee extension set of 50 reps at a load intensity of 40–110% of maximal effort. This is surprising considering the large repetition range without recovery prescribed at a high intensity, yielding an intense endurance stimulus. As effect sizes were not reported and the data was not displayed graphically, further interpretation of this finding is not possible.

Currently, one study has examined chronic resistance training and the VEGF response (Table 2). Following 12 weeks of resistance training in older females, no changes in plasma concentrations of VEGF were observed [168]. However, limitations with study design possibly contributed to the null findings as the intensity of exercise was low, as was the total training dose (i.e., average of 1.2 sessions per week). Exercise protocols with higher reps or shorter inter-set recovery periods, pursuing an endurance stimulus, are more likely to result in an enhanced VEGF expression due to greater blood flow in the periphery, muscle hypoxia or contraction-induced VEGF secretion [192, 197].

Nitric oxide and Alzheimer’s disease

Optimal vascular tone is maintained through a healthy endothelium by careful regulation of vasoactive mediators of vessel constriction (i.e., endothelin, angiotensin 2) and dilation. Nitric oxide (NO) is the single most potent chemical vasodilator [199] derived from oxygen, NO synthase, and L-arginine, and controls blood flow [200] and blood viscosity [73], prevents the adhesion of platelets and leukocytes to the endothelium [201], and protects against endothelial dysfunction (i.e., atherosclerosis) [199, 202]. Importantly, NO is also associated with neuroprotection through its interaction with the guanylate cyclase receptor to reduce oxidative stress and excess Ca^2 + by inhibiting the N-methyl-D-aspartate (NMDA) receptor [203]. Within neurons, NO is a messenger molecule and influences neuronal plasticity in the hippocampus and cerebellum [200]. Diminished NO activity is associated with systemic vasoconstriction, increased blood pressure, atherosclerosis, inflammation, and platelet aggregation (Fig. 5) [199, 204]. Oxidative stress is recognized as a leading factor influencing NO bioavailability [73, 199] and Aβ aggregates are known to generate superoxide anions, [205] a form of reactive oxygen species [206].

OPEN IN VIEWER

Fig.5

Effect of aging and neurodegenerative disease on endothelial nitric oxide synthase and the proposed direct () and potential downstream () influence (↑= increase; ↓= decrease; ? = unknown effect) of resistance training. Hypothetical pathways are marked (). Aβ, amyloid-β; eNOS, endothelial nitric oxide synthase; NO, nitric oxide.

Endothelial NO is dependent on the eNOS enzyme, a NO isoform which primarily regulates the production of NO through the substrate L-arginine [204]. Activation of eNOS in animal models preserves CBF, promotes VEGF driven angiogenesis, [207] and protects against platelet aggregation, inflammation and programmed cell death [204]. Neurologically, eNOS mobilizes progenitor cells, which are crucial for the repair of ischemia-induced neural tissue damage [207, 208]. The activation of eNOS occurs through several mediators such as insulin, [209] estrogen, [210] acetylcholine [211], and fluid shear stress upon the walls of the endothelium [212]. Considered the most important physiological stimuli for upregulation of eNOS expression and mRNA activity, [213] fluid shear stresses result from tangential blood flow against the endothelium. Indeed, even brief periods of these stresses cause rapid changes in eNOS induced NO concentrations [214].

Within individuals with AD, eNOS levels are decreased and negatively associated with Aβ plaque burden severity in temporal (r _s = –0.43, p < 0.01) and occipital cortices (r _s = –0.41, p < 0.01) [215]. The multifaceted action of eNOS can be observed in eNOS knockout or eNOS deficient animal models whereby hypertension, [204] BDNF depression, [207] mitochondrial dysfunction, [216] and insulin resistance [217] are side-effects. Prolonged exposure to cerebral hypoperfusion due to hypertension, dyslipidemia, or hyperglycemia activates eNOS through homeostatic control [218]. Under these conditions, a chronic upregulation of eNOS is observed leading to a decrease in the responsiveness to eNOS signaling, endothelial cell dysfunction and, consequently, a reduction of basal NO concentration [73].

Endothelial nitric oxide synthase and resistance training

Within the literature, a single study examined the influence of resistance training on eNOS availability (Table 2). After 10 weeks of resistance training (3×of 8–10 reps at 60–80% of 1RM), serum eNOS levels were unchanged in an exercise only and non-exercise control group. However, in a group prescribed both exercise and testosterone supplementation (100 mg/week), a 62% increase in peripheral serum eNOS was observed post-intervention when compared with an exercise + placebo group [219]. Testosterone treatment in animal models has been shown to stimulate microvascular growth through downstream mediation of VEGF production, and expression of its receptor [188]. Considering a substantial quantity of skeletal muscle-derived eNOS is sourced from the microvasculature, [220] it is possible that testosterone-mediated microvascular growth is responsible for the observed elevation in eNOS in this study. It is important to note that the participants recruited for this study were middle-aged and older men with already low testosterone levels and chronic obstructive pulmonary disease, a disease that is known to reduce resting eNOS levels [221]. From these findings, it appears resistance training alone provides inconclusive benefit to eNOS without testosterone administration.

Homocysteine and Alzheimer’s disease

Homocysteine is a homologue to cysteine, a proteinogenic amino acid with several crucial biological functions including protein structure maintenance. Cysteine catabolizes further to glutathione, a powerful antioxidant that minimizes oxidative stress-induced lipid peroxidation in cells [222]. Levels of homocysteine in circulation are typically low (4.0–10.0μmol·L^–1 in serum [223]) due to consistent folate and vitamin B12 driven re-methylation to methionine, or catabolism to cysteine and glutathione [223, 224]. However, homocysteine levels increase with age [225, 226] and are markedly elevated in individuals with AD, including preclinical cases [227 –229]. In individuals with AD, serum homocysteine concentrations may be some 80% higher than that of healthy age-matched controls [230]. Indeed, there is a five-fold greater risk of developing AD in individuals with serum homocysteine levels greater than 14.0μmol·L^–1 when compared with individuals with low serum homocysteine [228]. Furthermore, when compared with individuals presenting with plasma homocysteine concentrations less than 8.6μmol·L^–1, individuals with values greater than 11.9μmol·L^–1 are three-times more likely to experience white matter damage within the brain [231].

Hyperhomocysteinemia is associated with vascular and neuronal damage (Fig. 6) [232]. Within the vascular system, hyperhomocysteinemia elevates plasma asymmetric dimethylarginine, strongly inhibiting NO synthase leading to endothelial dysfunction [232]. Furthermore, there is a greater risk of atherosclerosis, blood clots, and stroke under such physiological conditions [233, 234]. Neurologically, homocysteine induces oxidative stress by generating reactive oxygen species and stimulating lipid peroxidation in the brain, processes that are associated with Aβ neurotoxicity [232 , 236]. Indeed, antioxidant supplementation can partially counteract the neurotoxicity of homocysteine [235]. Sustained high levels of homocysteine accelerates the deposition of Aβ in the brain, indicating a significant role in cognitive dysfunction using current theorems. High levels of homocysteine are positively associated with diminishing cognitive performance over time, [229, 237] and changes in homocysteine levels over a six-year period have been shown to predict cognitive decline in older individuals [238]. Furthermore, functional changes reflect morphological changes in cerebral tissue, whereby hyperhomocysteinemia results in hippocampal atrophy in cognitively healthy older individuals [239].

OPEN IN VIEWER

Fig.6

Effect of aging and neurodegenerative disease on homocysteine, and the proposed direct () and potential downstream() influence (↑= increase; ↓= decrease; ? = unknown effect) of resistance training. Hypothetical pathways are marked (). Aβ, amyloid-β; CBS, cystathionine beta-synthase; CBSP, cystathionine beta-synthase polymorphism; NO, nitric oxide.

Homocysteine and resistance training

A number of studies have assessed basal homocysteine levels following chronic resistance training in healthy older participants [240, 241] and individuals with AD, [242] with varying findings (Table 2). When compared with control groups, 6 [241] and 12 months [240] of intensive resistance training (75–80% of 1RM) resulted in significant declines in serum homocysteine levels, as did 12 months of low-intensity (50% 1RM) resistance training [240]. These findings, however, were observed in healthy older adults as the benefits of resistance training on homocysteine in AD is less conclusive. Indeed, no difference in serum homocysteine levels was observed after 16 weeks of resistance training within an AD cohort [242]. In this study, participants exercised three times per week using three sets of 20 reps at 85% of an approximate 20RM. It is possible that other factors such as diet and poor nutrient assimilation, medications, supplementation, or comorbidities not controlled for in this study could have influenced this result [243].