|Home | About | Journals | Submit | Contact Us | Français|
The integrity of the vascular system is essential for the efficient functioning of the brain. Ageing related structural and functional disturbances in the macro- or microcirculation of the brain make it vulnerable to cognitive dysfunction leading to brain degeneration and dementing illness. Several faltering controls including impairment in autoregulation, neurovascular coupling, blood-brain barrier leakage, decreased cerebrospinal fluid and reduced vascular tone appear responsible for variable degrees of neurodegeneration in old age. There is ample evidence that vascular risk factors are also linked to neurodegenerative processes preceding cognitive decline and dementia. Age is the strongest risk factor for brain degeneration whether it results from vascular or neurodegenerative mechanisms or both. However, several modifiable risks such as cardiovascular disease, hypertension, dyslipidaemia, diabetes and obesity enhance the rate of cognitive decline and increase the risk of Alzheimer’s disease in particular. The ultimate accumulation of brain pathological lesions may be modified by genetic influences such as apoliopoprotein E ε4 allele and the environment. Lifestyle measures that maintain or improve cardiovascular health including consumption of healthy diets, moderate use of alcohol and implementing regular physical exercise are important factors for brain protection.
The brain comprises only 2% of the body weight yet consumes a critical 20% of the body’s oxygen and other nutrients supplied via the vascular system. The integrity of the vasculature is therefore essential for the optimal functioning of the brain. In addition to the cardiovascular system, brain vascular control mechanisms are vital for the maintenance of the neurovascular milieu, created by nerve terminals, astrocytic endfeet and the microvasculature.1, 2 Numerous molecular signals including vasoactive peptides, gaseous transmitters and growth factors mediate the tight regulation of cerebral blood flow (CBF) in tandem with neural activity. Structural, chemical or functional disturbances in the macro- or microcirculation of the brain would ultimately affect cognitive function and behaviour. Increasing age is the strongest risk factor for brain degeneration and age-related cognitive disorders (Table 1). However, none of the protective mechanisms is immune to the selective or general cumulative effects of ageing and malfunction in one could impact on the other with respect to both time and space (Figure 1).
Various factors may influence the nature and severity of brain degeneration. The degree of cerebral grey matter damage, neuronal death and survival will be dictated by the multiplicity, size and laterality of the tissue injury or the extent of vascular disease. Anatomical features of the circulation including the size of vessels and vascular wall cellular elements e.g. arterioles versus capillaries are important factors in defining the pathology.1 The distribution territories of the anterior, posterior and middle cerebral and the lenticulostriate arteries affect different structures including the angular gyrus, caudate and medial thalamus in the dominant hemisphere, the amygdala and the hippocampus; all structures implicated in forms of cognitive impairment.3 The origin and degree of vascular occlusion or injury and whether this results in ischaemic or haemorrhagic lesions are further factors which define the extent and severity of damage. Alterations in specific genes associated with systemic disease or brain specific proteins and environmental or lifestyle factors may further modify the course of degeneration. This review will consider ageing-related changes at the fluid-tissue surfaces of the brain and various factors which put vascular health at risk to cause brain degeneration and dementia in old age.
Both cerebrovascular and neurodegenerative diseases increase significantly after 60 years of age in almost all populations world wide.4 Alzheimer’s disease (AD) is widely recognized as the most common cause of dementia predominantly resulting from neurodegenerative changes.5 Cerebrovascular disease causes the second most common form of brain degenerative disorder leading to dementia. Just over 100 years Alois Alzheimer and Emil Kraeplin were among the first neuropsychiatrists who had reasoned that gradual strangulation of the blood supply to the brain was the main cause of organic brain degeneration.6 They surmised that ageing related progressive hardening of the arteries leads to arteriosclerotic dementia. Even until the late sixties5 arteriosclerotic dementia, attributed to cerebral softening with loss of relatively large volume (>50 ml) of tissue was described as a major cause of dementia which was reported to be over-diagnosed clinically. In recent years vascular dementia (VaD) or more precisely vascular cognitive impairment (VCI) has been described that takes into account the consequence of a variety of cerebrovascular lesions or impaired brain perfusion.7 Subcortical ischaemic VaD appears the most significant subtype of VaD 8 which involves small vessel disease and longer survival. The original recognition of subclasses of VaD should probably be credited to Otto Binswanger who described subcortical arteriosclerotic encephalopathy upon pathological verification of cerebral white matter (WM) lesions in patients with hypertensive disease.6 Clinical, neuroimaging and pathological features have facilitated the definition of VaD or VCI subtypes and allowed the appreciation of specific risk factors and the evaluation of treatment or prevention strategies.
Distinguishing between neurodegenerative and vascular processes which result in dementia is pertinent to identifying the causes and treatment options.. However, unselected community-based studies show that brains of demented subjects often exhibit more than one type of pathology. Vascular lesions are frequently found to co-exist with AD-type of lesions in older subjects.9, 10 Even in prospectively assessed AD subjects entirely pure neurodegenerative pathology appears to be an exception, particularly among the oldest old (>85 years).9 Autopsied brains of AD subjects frequently bear cerebrovascular pathology consisting of degenerative microangiopathy, cerebral amyloid angiopathy (CAA), cerebral infarcts and to small extent intracerebral hemorrhages.11 The co-existence of two or more pathologies is not novel to dementia. This was apparent from the description of Alzheimer’s first case of Auguste D. In his original report6 he had written that besides “one or several fibrils in otherwise normal cells, and “numerous small miliary foci …and…storage of peculiar material in the cortex, one sees endothelial proliferation and also occasionally neovascularisation.” While it is unlikely to be known whether the vascular changes represented angiogenesis as a result of ischaemic injury or microvascular degeneration per se the important question is whether the described microvascular pathology was coincident with or causal in the neurodegenerative changes leading to AD. However, it is now apparent that vascular and neurodegenerative pathologies are additive in influencing the clinical presentation of dementia.12 Cerebrovascular disease or strokes appear to not only hasten the pathological processes leading to dementia but also increases the possibility that individuals with Alzheimer lesions in their brains will exhibit dementia.13 Although cerebrovascular disease may be the main cause of degeneration and dementia, it is likely that such an event overcomes the brain's reserves14 in individuals being compromised by neurodegenerative changes related to Alzheimer pathology.
Several clinical studies suggest that blood-brain barrier (BBB) dysfunction is involved in the progression and pathogenesis of AD.15, 16 17 The cerebrospinal fluid (CSF)/ serum albumin ratio (commonly used indicators of BBB integrity) is increased in AD patients with particularly co-morbid vascular disease.16, 18 In 85-year-old AD patients,19 higher CSF/serum albumin ratios were reported but the indications of disturbed BBB function were evident even before the onset of disease. The positive correlations between CSF-albumin index and measures of disease progression over 1 year in a subgroup of AD patients also suggest that BBB impairment could modify disease progression.20 The relative dysfunction of the BBB may increase the possibility of substances penetrating the BBB more rapidly to interact with neurons, and initiating a cascade of events involving amyloid accumulation and Alzheimer encephalopathy (Figure 2).
Prior biochemical and morphological studies (Table 2) have reported a variety of age-related alterations in the intracranial vessels including the perforating arteries, arterioles and capillaries in animals and man.1, 21 Cerebral arteriosclerosis, focal constrictions and stiffening of the vessel wall resulting from increased collagen fibres and extracellular matrix components result in loss of distensibility or elasticity that would affect brain perfusion. The terminal type arterioles of the deep white matter (WM) are particularly vulnerable targets in the elderly. Thus age-related increases in both the prevalence and severity of tortuosity of vessels in the WM may contribute to a chronic hypoxic state22, 23 which is attributed to the increasing long-term disability in the elderly.24 In addition, cerebral amyloid angiopathy (CAA), microbleeds and atherosclerosis,25, 26 affecting the neocortical perfusing surfaces would not only make the circumscribed neuronal populations behind them vulnerable but also the distal fields of the end-arteries.1 When the vascular changes are compounded by neurodegenerative processes the outcomes are predicted to be worse.
Age related changes in cellular elements of the neurovascular unit (Table 2) consisting of focal necrosis of the cerebral endothelium, accumulation of extracellular matrix components within the vascular basement membrane, decreased endothelial mitochondrial density, increased pinocytotic vesicles, loosening of tight junctions, loss of the perivascular nerve plexus, changes in the astrocytic end-feet and selective degeneration of the microvascular tree have also been described.1 These changes are seen to be more intense in brains from AD subjects which also exhibit convolutional abnormalities and "collapsed" or attenuated capillaries in cortical lobes.27 This vascular phenomenon, associated with amyloid β deposits in all cortical lobes, is also evident in Down's syndrome.28 The microvascular changes imply abnormalities in the patency of the brain microvasculature and breach of the BBB. These descriptions support previous conclusions on disturbances in local perfusion and decreased oxygen tension within the neuropil as a consequence such that neurons furthest away from the capillary surface are divested.1 Changes in the dynamics of supporting cells within the neurovascular unit such as astrocytes, microglia and pericytes may bear additional consequences in reactive or compensatory mechanisms. Conversely, evidence from elegant studies by Zlokovic and colleagues29 has indicated that microvascular abnormalities lead to faulty BBB clearance of amyloid β through the deregulated low density lipoprotein-related protein 1 (LRP1) and receptor for advanced glycation endproducts (RAGE)-mediated transport. These impaired clearances of amyloid β and glycation endproducts then lead to aberrant angiogenesis, remodeling of the cerebral microvasculature and eventual arterial dysfunction which in turn initiates neurovascularuncoupling.30
Selective changes in key BBB proteins including the glucose and nutrient transporters, neurotransmitter receptors and degrading enzymes (Table 2) have been recorded in the cerebral microvasculature of demented subjects.1, 2, 29 The glucose transporter (type I), an integral protein to the cerebral endothelium, is reduced in the microvasculature of AD subjects compared to ageing controls. Whether amyloid β deposition in brain microvessels in AD directly impairs synthesis or enhances degradation of the protein remains unknown, but profound effects on the permeability and function of the brain endothelium would be expected.15 These in vitro observations are corroborated by positron emission tomography studies showing that the transport of glucose (kinetic parameter k1) into brains of AD subjects is diminished 31. Moreover, uptake of 2-deoxy-D-glucose32 is also decreased further implying impairment of hexokinase activity in the cerebral microvasculature. The decreased activity directly relates to the reduced glucose transporter protein findings and suggests that microvessels deprived of glucose may be using other sources of metabolic fuel. Impairment in glucose transport or metabolism may also induce overactivation of glycogen synthase kinase-3beta and this together with down-regulation of the O-GlcNAcylation process promotes abnormal hyperphosphorylation of tau leading to neurofibrillary pathology.33
Biochemical approaches demonstrate selective changes in markers of the BBB or cerebral microvessels in AD.1, 2, 34 Specific proteins including the cholinesterases and APP cleaving proteases may promote or deregulate amyloid β deposition in the vasculature35 to affect the cerebral microcirculation which is considered a source of neurotoxic and inflammatory mediators.34 It is not unlikely that breakdown of the BBB occurs focally and transiently over a protracted period in association with reactive mechanisms that direct repair and growth.34, 36 For example, vascular endothelial growth factor reactivity is substantially increased in astrocytes and perivascular deposits in brains of subjects with AD implicating angiogenic signals are induced to counter insufficient vascularity or reduced perfusion (oligaemia).37 However, the ensuing pathology probably overwhelms the compensatory mechanisms such that the angiogenic or growth response is stunted.36, 38
The choroid plexus-cerebrospinal fluid (CSF) system ensures not only CSF secretion and clearance but the constant scrutiny of nutrients and harmful substances to maintain homeostasis.39 Both the quantity (by 50%) and quality of the CSF declines with age and in neurodegenerative disease.40, 41 The choroidal epithelium undergoes atrophy, thickening of the basement membrane, stromal fibrosis and accumulates Biondi rings tangles and lipofuscin with age and more so in AD.42 These changes together with activation of inflammatory mechanisms 43 are responsible for an estimated 50% reduction in turnover and alteration of CSF contents.39 The impaired ability of the choroid plexus to clear molecules from the CSF in older age has potentially serious implications for brain clearance of key proteins including amyloid β in AD subjects.44 This is analogous to the early vascular changes in amyloid precursor protein (APP) in over-expressing transgenic (Tg) mice2 which show that regional specific microvascular anomalies may precede the more profound pathologies which characterise neurodegeneration and AD.
Autoregulation of CBF is the first order mechanism that ensures that the flow and supply of oxygen, glucose and nutrients through the vascular beds remain within the upper and lower limits of the autoregulatory range (50–160 mm Hg) during fluctuations in systemic arterial pressure.3 Cerebral resistance arteries dilate or constrict during changes in arterial pressure. Ageing related changes in the systemic circulation and degenerative changes in the extracerebral resistance arteries1 may shift the lower and upper limits of the autoregulatory plateau to cause hypertensive encephalopathy or cerebral hypoperfusion, which may not necessarily cause ischaemic injury as obvious as in stroke but oligaemia or hypovolemia that leads to disruption of the microcirculation, damage to the cerebral endothelium, breach of the BBB and edema.17 Neurogenic responses via the brainstem autonomic nuclei, which degenerate in age-related brain disorders,45, 46 may also disrupt CBF autoregulation that may increasingly cause orthostatic hypotension and syncope. In addition, pathological changes in vascular wall smooth muscle (myogenic effect) and altered release of metabolic factors could profoundly influence autoregulatory responses, which would have a high impact in the oldest old (>85 years of age). Animal studies reveal that autoregulation fails in Tg mice over-expressing the APP and accumulating Aβ deposits. In the APP Tg mice vascular alterations are observed at 3 months of age much in advance of Aβ deposition and cognitive decline.2
Functional hyperemia or neural activity coupling with CBF is another vital mechanism to protect the tight dynamic relationship between the cellular elements of the neurovascular unit and homeostasis of the cerebral microenvironment.2 Sensitive neuroimaging methods confirm ageing-related regression in global and regional measures of CBF (~4mL/min/year), cerebral metabolic rate for oxygen, glucose oxidation, cerebral blood volume and not surprisingly CSF secretion and flow.47 In AD, more pronounced regional effects are apparent.2 Under certain pathological and pharmacological conditions, CBF and metabolism are uncoupled and likely result from weaker muscle tone, failure of angiogenesis or breakdown in the molecular communication between neurons and microvessels.38 The latter undergo several changes including loss of pericytes1 that indicate the BBB in the elderly is less likely to compensate for leaks.17 The characteristic deposition of amyloid in cerebral vessels in AD exacerbates vascular dysfunction and promotes hypoperfusion. APP expressing Tg mice also show that the reactivity of the cerebral circulation to endothelium-dependent vasodilators, a major vasoregulatory mechanism, is impaired. This is consistent with the diminished vasodilator responses attributed to altered modulatory function of the endothelium rather than impairments in vascular smooth muscle per se during ageing.48
Recent advances indicate that dementia risk is modified by perinatal events, education status, nutritional intake, degree of physical activity and cognitive and social engagement.49 Several of these factors impact on adult-onset vascular disorders including strokes, hypertension, atherosclerotic disease, atrial fibrillation, diabetes mellitus, dyslipidemia, hyperinsulinaemia, hyperglycaemia, hyperhomocysteinemia, and obesity (Table 3).50 It is increasingly recognized that factors which increase cardiovascular disease or brain vascular pathology exacerbate the onset or progression of late-onset dementias (Figure 2). While data vary in individual studies that may relate to the clinical diagnosis of dementia, carriers of the apolioprotein E ε4 allele in general appear at greater at risk in the presence of vascular disease. However, the evidence for single clinically defined cardiovascular risk factors being associated with incident AD is inconsistent.51 Some genetic influences or environmental factors may modify the progressive changes which define the final phenotype and burden of brain pathology.11 While randomized or controlled trials of risk factor modification (with multiple simultaneous interventions) are lacking it is encouraging that interventions of cognitive and physical activity were shown to improve cognitive performance and slow cognitive decline.52
In the elderly, stroke episode or history of transient ischaemic attack increases the risk of AD up to three-fold.53–55 Cerebral infarcts appear to accelerate AD progression and account for a large proportion of mixed pathologies cases in older age groups.54, 56–58 Silent brain infarction (without overt neurological sequelae) occurs with high prevalence in the healthy elderly, estimated at ~5% in those 60 yrs of age in both genders.59, 60 Almost 90% of these occur subcortically (and appear as lacunae in basal ganglia) with the rest in the neocortex. Silent infarcts double the risk of dementia, cause steeper cognitive decline (memory performance) and impact on executive dysfunction.61, 62
Patients with cerebrovascular disease who survive long may progress to develop dementia characterized by Alzheimer type of changes.63, 64 Conversely, Henon et al 65 have shown that severe cognitive impairment or low cognitive ability, which could represent AD, is also a risk for cerebrovascular disease or strokes.66 Stroke incidence was highest in those with severe impairment and the association between cognitive impairment and incident stroke appears not to be mediated by systemic vascular disease. The elevated risk of subsequent strokes in older persons with cognitive impairment suggests that cerebrovascular disease may play a direct role in cognitive impairment. Therapeutic monitoring or treatment of the vascular component should ameliorate or prevent worsening of dementia symptoms in individuals with a diagnosis of AD or other insidious neurodegenerative disease.7
Neuropatholgical studies provide further evidence for considerable overlap between AD and cerebrovascular disease.53 High proportions of individuals fulfilling the neuropathological diagnosis of AD have significant cerebrovascular lesions11–13, 67–69 while VaD patients frequently show extensive AD–type changes.70, 71 Thus in one prospective study70 in clinically diagnosed VaD patients at autopsy nearly 60% were found to have AD pathology alone and 40% had AD pathology in combination with cerebrovascular disease. Older patients diagnosed with VaD may accumulate some Alzheimer pathological features e.g. cerebral amyloid β42 levels comparable to those seen in AD.72 The concomitant vascular pathology in terms of lacunae, microinfarcts and arteriosclerosis appear to coexist by more than chance alone.73 For example, studies have emphasised that a significant association exists between cortical microinfarcts and AD (32% cases v 3% controls).57, 74 The microinfarcts were restricted to watershed cortical zones indicating that disturbed haemodynamic factors were involved in the genesis of these vascular lesions which may precede Alzheimer’s symptoms.75 When both neurodegenerative e.g. amyloid β deposits and neurofibrillary tangles and vascular pathologies are present the burden of dementia tends to be associated with small or micro ischemic lesions76, 77 rather than large infarctions. These findings corroborate the importance of microvascular disease rather than macroscopic infarction as the critical substrate in dementia.78
Further clinical importance of the combined neurodegenerative and vascular pathologies is emphasised by the finding that dementia was worse and compounded by the co-existence of both neuropathological features of AD and VaD in elderly nuns.13 In this prospectively assessed cohort with low vascular disease risk and where there was no association between cerebral infarcts and AD type of pathology, it was estimated that an eight-fold greater burden of neocortical NFT would be necessary to develop dementia in the absence of strategic cerebral infarction.13, 79 Similarly, the Religious Orders Study found that although there was no interaction between infarctions and AD pathology13 the presence of one or more infarctions independently increased the odds of dementia by nearly three-fold.80 The Honolulu Aging in Asians study (HAAS) provides additional support for the independence of AD and vascular neuropathology.81 These reports do not show whether ischaemic cerebrovascular disease contributes directly to the development of Alzheimer’s pathology but they indicate that vascular changes have an additive effect and reduces the threshold for dementia diagnosis in patients with an otherwise asymptomatic low-grade Alzheimer’s pathology.82
Progressive age-related changes in intracerebral vessels or presence of infarcts may result in WM abnormalities in as many as 96% of those over 65 years of age.83 WM abnormalities are seen as WM74, 84 hypersignals in T2-weighted or fluid attenuated inversion recovery (FLAIR) sequences on magnetic resonance imaging (MRI). Significant risk factors for more severe WMH include older age, history of hypertension, prior stroke, presence of diabetes, smoking habit and clinically silent strokes.85 The pathological correlates of WMH comprise regions of deep WM pallor or myelin loss, axonal disruption, arteriosclerosis within small to medium sized arteries with perivascular spaces. These changes are worsened by oedema and diffuse damage of the BBB with chronic leakage of fluid and macromolecules. Multi-disciplinary correlative studies show that both deep WM and periventricular hyperintensities represent primarily oligaemic tissue damage86 which may be accompanied by a Wallerian type of axonal degeneration secondary to neurodegenerative processes and lead to dementia (Figure 2).
Past studies have attempted to assess correlations between the degree of WMH from MRI and cognitive function or disability in the elderly. Although distinctions between periventicular and deep WM may be arbitary24, 87 it appears that the total WMH volume is a critical measure. Most studies with cohorts of variable size reveal declines in frontal lobe functions which include executive tasks, attention, motor and processing speed rather than memory per se and appear to be associated with increased WMH volume.88, 89 In the LADIS cohort executive dysfunction was attributed to WM lesion severity in the non-disabled elderly.24 On the other hand, volume of periventricular WMH rather than deep WMH at baseline and progression was longitudinally associated with decline in processing speed.90 Severe WMH has likewise been reported to increase the risk of disability24 and death91 in community-dwelling elderly without stroke or other neurological disease. Increased WMH and the presence of lacunar infarcts, indicative of subcortical vascular disease, are associated with greater global atrophy.92, 93 An association between WMH severity and medial temporal atrophy (MTA) was reported in one study of siblings affected and at risk for AD94 consistent with demyelination secondary to AD neuronal damage.
Alterations in arterial pressure during ageing have been strongly implicated in brain degeneration and dementia.95 Many studies have confirmed that a history of hypertension during midlife may increase risk of dementia but particularly AD. In the first longitudinal study, Skoog et al96 showed that increased systolic (>160 mmHg) and diastolic (>95 mmHg) blood pressure 10–15 years before were factors in the onset of AD. The risk of developing dementia between the ages of 80 and 85 increased with increasing blood pressure at the age of 70 years even within the lower ranges of blood pressure values obtained. Long-standing increase in blood pressure may increase risk for dementia and AD by inducing small-vessel disease, WM changes and cerebral hypoperfusion by disrupting vasoregulatory functions. However, atherosclerotic disease leading to cerebral hypoperfusion has also been implicated.97 In later life and after onset of dementia blood pressure actually decreases suggesting that cerebral atrophy or intrinsic neuropathological changes are responsible for lowering pressure.
Pathological studies have revealed that hypertensives with elevated systolic and diastolic pressures in midlife exhibit an increased burden of AD type of neurodegenerative lesions including neuritic plaques and neurofibrillary tangles98 and which are also linked to low brain weight.99 A few trials also show that the use of antihypertensives offers some protection from the risk of dementia.100 However, there is no convincing evidence from the currently published trials that blood pressure lowering in late-life prevents the development of dementia or cognitive impairment in hypertensive patients with no apparent prior cerebrovascular disease.101
In comparison to the role of hypertension,81 the risk of developing dementia after hypercholesterolemia is less robust. This may relate to how studies are conducted and the age at which cholesterol concentrations are determined.81, 102 However, using doppler scanning, the Rotterdam investigators103 concluded that atherosclerosis exhibited as carotid artery thickening or accumulation of wall plaques in the elderly is a strong risk factor for late onset AD and VaD. On the other hand cholesterol levels measured in midlife were associated with higher risk of AD but not VaD.104 Total cholesterol and low density liopoprotein (LDL) concentrations and a history of diabetes have also been shown to be associated with more rapid cognitive decline in individuals with AD.105 Consistent with the increased risk of AD, subjects with coronary heart disease also exhibit an increased amount of brain amyloidosis.106, 107 It is clear that high cholesterol and other lipids play a role in increasing risk but the course of the converging pathways and how this occurs are debated.108
Several epidemiological show that a history of adult onset diabetes mellitus increases the risk of cognitive impairment and dementia in the elderly.109, 110 Risk for AD and particularly VaD was reported to be 2–2.5 fold greater among type II diabetics, irrespective of age at which diabetes occurs. Consistent with this a meta-analysis111 showed that the risk for AD was significantly higher (RR 1.4) but almost two fold greater for VaD (RR 2.4). Neuroimaging evidence suggests an association between diabetes and cerebral atrophy and lacunar infarcts but the association with WM lesions is less robust.112 These studies emphasise the link between glucose dysregulation and brain degeneration even in the very old but do not explain the link between diabetes and the AD type of neurodegenerative changes. Several scenarios including impaired insulin signalling induced neurodegenerative changes, advanced glycation of neuronal components, oxidative stress and inflammatory mechanisms have been proposed.
Features of the insulin resistance syndrome have also been associated with low cognitive function113, 114 and with AD.115 The risk is even higher in individuals expressing components of the metabolic syndrome including high blood pressure, increased triglycerides, high blood glucose, low LDL cholesterol and obesity.116, 117 Moreover, obesity and overweight in midlife, measured by body mass index and skin-fold thickness, are strongly associated with an increased risk of both AD and VaD, independent of the development of diabetes or other cardiovascular-related morbidities. Conversely, higher baseline body mass index and slower declining body mass in late life appear to reduce risk of dementia.118 This suggests that a faster decline in body mass index in late life is a preclinical indicator of an underlying dementing illness, especially for those who were initially overweight.118
Outcomes of large multicentre randomized trials are awaited but the bulk of currently reported observational studies in elderly cohorts119, 120 suggest that development or progression of cognitive impairment and dementia can be slowed by use of healthy foods and vitamin supplements121 and by the avoidance of nutritional deficiencies and balanced calorie restriction. Independent evidence suggests regular consumption of fish is related to lower risk of AD122, 123 and slower rate of cognitive decline.124, 125 Fish may offer protection by countering inflammation and enhancing vascular tone and countering atherosclerosis. Diets that emphasize polyphenols including reservatrol, omega-3 fatty acids, docosahexaenoic and eicosapentaenoic acids, flavonoids and the B vitamins, especially folate, B6 and B12 are implicated as likely to sustain or improve cognitive function in older age.126 Higher adherence to the Mediterranean diet127 consisting of whole grains, fish and olive oil and moderate consumption of alcohol is also associated with lower risk for AD128 and significant improvement in incidence neurodegenerative disorders.129
Findings from clinical and experimental studies show that chronic accumulation of reactive oxygen species in older brains may exhaust antioxidant capacity and trigger neurodegenerative processes as characterized in AD. Thus dietary supplementation with fruit or vegetable extracts high in antioxidants help to decrease the enhanced vulnerability to oxidative stress and improve neuronal communication via increases in neuronal signaling and animal behavior.130 Onset of AD was significantly delayed by use of antioxidant vitamins and polyphenols derived from fruits and vegetables.129 131 Results from the Kame Project suggest that drinking fruit juices which are high in polyphenolic compounds, was associated with lower risk of incident AD.129 Congruent with the notion that vascular health is key to maintaining cognitive function polyphenols from wine, cocoa, coffee, grape seed, blueberries, strawberries, tea, curcumin, pomegranate and green leafy vegetables also have beneficial effects on endothelial function and cardiovascular performance.129 The beneficial actions of resveratrol have been implicated in anti-oxidant defence, regulation of the cell cycle, mitochondrial energy production, vascular reactivity, oncogene suppression and activation of sirtuins (silent information regulator-related enzymes), as anti-ageing inhibitors.132 In contrast to prior assertions that it may be protective there is strong consensus that smoking is a vascular risk factor for dementia because of its association with several degenerative features. These include cerebral infarcts, white matter hyperintensities,102 subcortical atrophy and neuronal death by increasing amyloid burden.133, 134 Although medicinal nicotine may be beneficial133 these findings reinforce the notion that any measure that decreases atherosclerosis and increase brain blood flow are valuable for cognitive function in old age (Figure 2).
Accumulating evidence suggests changes in lifestyle factors such as increasing physical activity will decrease the risk of developing dementia in later life.120, 135 Most studies136, 137 studies show reduced rate of age-related cognitive decline, decreased risk of incident dementia or AD in individuals who exercise regularly. In a prospective cohort study138 higher physical activity was identified to reduce risk of dementia in community-dwelling elderly people. In the largest clinical trial of its kind139 physically active volunteers with memory problems exhibited significant improvement in cognition after a 24-week intervention.
The effects of physical exercise on vascular health may be obvious but neuroimaging studies 140 suggest that the more aerobically fit older adults lose less grey matter. Likewise, whole brain volume was found to be positively associated with aerobic fitness levels among patients with early stage AD.141 Higher aerobic fitness levels also have been shown to be associated with larger hippocampus, increased blood flow, oxygen delivery and better spatial memory150.142 In related studies on prevention, increased physical activity in mid-life136 have been found to be associated with less neocortical atrophy in the elderly. The latter association was independent of atherosclerosis and brain infarcts. The likely mechanisms through which exercise induces brain growth and improves cognition include upregulation of brain-derived neurotrophic factor and preservation of synapses.143
Several lines of evidence suggest a strong role for vascular factors in the aetiology and pathogenesis of brain degeneration. Ageing associated cerebral microangiopathy inflicted at several control levels may explain deficits in autoregulation and neurovascular coupling and can be predicted to occur in subjects developing AD, the most common cause of dementia. Prevailing evidence from population and community based studies suggests that pre-existing features of cardiovascular disease are strongly linked to late-onset dementia. These may lead to cerebral hypoperfusion involving key pathological substrates mainly white matter changes, axonal damage and microinfarction. Unlike age and genetic background, vascular risk factors are modifiable through management and lifestyle changes which then help to protect the brain.
I thank Yumi Yamamoto for helping to create Figure 1. Our work has been supported by grants from the Medical Research Council (UK) for G0500247 and G0700718 awards, US National Institutes of Health (NINDS) and the Alzheimer’s Research Trust (UK).