Systemic determinants of brain health in ageing

Abstract

Preservation of brain health is a worldwide priority. The traditional view is that the major threats to the ageing brain lie within the brain itself. Consequently, therapeutic approaches have focused on protecting the brain from these presumably intrinsic pathogenic processes. However, an increasing body of evidence has unveiled a previously under-recognized contribution of peripheral organs to brain dysfunction and damage. Thus, in addition to the well-known impact of diseases of the heart and endocrine glands on the brain, accumulating data suggest that dysfunction of other organs, such as gut, liver, kidney and lung, substantially affects the development and clinical manifestation of age-related brain pathologies. In this Review, a framework is provided to indicate how organ dysfunction can alter brain homeostasis and promote neurodegeneration, with a focus on dementia. We delineate the associations of subclinical dysfunction in specific organs with dementia risk and provide suggestions for public health promotion and clinical management.

Introduction

Brain health has become a worldwide health priority. The WHO defines brain health as “the state of brain functioning across cognitive, sensory, social–emotional, behavioural and motor domains, allowing a person to realize their full potential over the life course”¹. Many factors determine brain health²; here, we chose to focus on highly prevalent conditions that affect cognitive function, specifically Alzheimer disease, vascular cognitive impairment, Parkinson disease and dementia. The prevalence of dementia is rising globally in association with ageing populations. Furthermore, most new cases of dementia are anticipated to occur in low-income and middle-income countries³. The rapidly increasing prevalence of dementia, a costly condition, is straining health-care systems and budgets. In response, several professional bodies, including the WHO, have called for an invigorated focus on dementia risk reduction and brain health promotion⁴.

Remarkably, most age-related dementias, except for vascular dementia, are associated with intracellular or extracellular protein aggregates in the brain. These aggregates can be found in the neuropil (for example, amyloid-β (Aβ) plaques in Alzheimer disease), neurons (for example, tau in Alzheimer disease and α-synuclein in Lewy body dementia (LBD)) or astrocytes (for example, tau in progressive supranuclear palsy)⁵. Thus, vast research has been aimed at understanding the molecular pathways that lead to abnormal protein aggregation as well as exploring the hypothesis that removing these aggregates could arrest neurodegeneration. This approach has borne some fruit. Recent phase III trials of anti-Aβ monoclonal antibodies for the treatment of Alzheimer disease, shown to be highly effective at removing plaque amyloid, reduced the rate of cognitive decline but only by 22–27%^6,7. The continuing cognitive decline, albeit at a modestly slower rate, suggests that other pathways — possibly not related to the classic protein aggregates — are active and must be addressed to stave off neurodegeneration.

Peripheral organs play an important but underappreciated role in brain health by maintaining a proper homeostatic environment, delivering energy substrates for metabolism, regulating immune system function and clearing wastes. The best-studied examples are the cardiovascular organs and the endocrine system governing glucose metabolism. Indeed, most of the risk factors for dementia identified by the Lancet Commission are also risk factors for cardiometabolic disease and stroke⁸.

Severe organ failure often has overt acute effects on the brain, causing encephalopathy⁹. What has emerged more recently in epidemiological and translational research is that a wider array of milder, and even subclinical, organ dysfunction can increase the risk of cognitive decline and dementia^10-16 (Fig. 1). While this relationship was already well documented for the cardiovascular and endocrine systems, emerging evidence increasingly implicates other organs in the maintenance of cognitive health. In this Review, we synthesize epidemiological findings on peripheral organ dysfunction and brain health, with a focus on dementia; highlight potential mechanisms by which organ dysfunction can interfere with the neurovascular function of neuronal networks, including complex bidirectional interactions with age-related neuropathologies such as Alzheimer disease and LBD; and provide suggestions for brain health promotion and clinical care. To reflect recently evolving concepts, we selected specific organ systems — cardiac, endocrine, renal, respiratory, gut and liver — based on the strength of emerging evidence that subclinical disease of these organ systems has deleterious effects on brain health.

Fig. 1 ∣. Spectrum of organ dysfunction and its association with neurodegeneration and cognitive decline.

Frank organ failure (far right) has long been recognized as a cause of acute encephalopathy mediated by various derangements, including hypoxia, ischaemia and metabolic abnormalities. More recent evidence shows that a spectrum of organ dysfunction, from asymptomatic alterations in functional tests to clinically recognized mild diseases, can increase the risk of later-life cognitive decline. Each of the abnormalities shown has been linked with future risk of cognitive decline in population-based cohort studies^{10,11,13,14,16,25,34,39,43,48,59,63}. In turn, brain dysfunction can induce organ dysfunction through neurohumoral mechanisms and poor self-care. ALT, alanine aminotransferase; AST, aspartate aminotransferase; BP, blood pressure; COPD, chronic obstructive pulmonary disease; FEV₁, forced expiratory volume in the first second; FVC, forced vital capacity; GFR, glomerular filtration rate; HbA_1c, haemoglobin A_1c; NAFLD, nonalcoholic fatty liver disease.

Dysfunctional organ systems and their effect on brain health

This section discusses epidemiological evidence on the link between subclinical organ disease and brain health. We focus on organs and systems with more recently established evidence as well as those with the greatest certainty of evidence (Table 1). A summary of associations between organ dysfunction and biomarkers of neurodegeneration is found in Table 2.

Table 1 ∣.

Organs and organ diseases associated with cognitive decline and dementia

Organ/system	Condition	Outcome	Ref.
Cardiovascular	Myocardial infarction	Cognitive decline	19
		Dementia	10
	Coronary artery bypass	Cognitive decline	198
	Heart failure	Cognitive decline	22
		Dementia	10
	Atrial fibrillation	Cognitive decline	199
		Dementia	200
Renal	Low eGFR	Cognitive decline	34
		Parkinson disease	35
Pulmonary	Restrictive lung disease	Cognitive decline	12
		Dementia	12
	COPD	Cognitive decline	40
		Dementia	39
	Asthma	Dementia	41
	Obstructive sleep apnoea	Dementia Parkinson disease	13
Gastrointestinal	Inflammatory bowel disease	Dementia	43
	Irritable bowel syndrome	Dementia	14
	GERD	Dementia	14
	Constipation	Parkinson disease	54
Hepatic	Liver fibrosis	Cognitive decline	15
		Dementia	63
Endocrine	Type 2 diabetes	Cognitive decline	16
		Dementia	16

Associations of selected organ disorders with prospective risk of cognitive decline and dementia, according to epidemiological studies. Note that we have omitted thyroid hormone disorders and hyperparathyroidism causing hypercalcaemia, as these are already well-recognized causes of reversible dementia for which diagnostic testing is already recommended by dementia guidelines²⁰¹. COPD, chronic obstructive pulmonary disease; eGFR, estimated glomerular filtration rate; GERD, gastro-oesophageal reflux disease.

Table 2 ∣.

Organ dysfunction and biomarker or neuropathological evidence of neurodegenerative or cerebrovascular disorders in human studies

Organ/system	Amyloid-β		Tau		Lewy bodies		Cerebral small vessel disease
Cardiovascular	+	202	+	203	?	NA	+	26,200
Renal	+	37	+a	38	?	NA	+	26,200
Pulmonary	?	NA	?		?	NA	+	42
Gastrointestinal	+b	48	+b	48	+	51-53	?	NA
Hepatic	+	64	+	64	?	NA	+	67
Type 2 diabetes	−	204	+	204	−	205	+	204

Representative references are shown. +, evidence that an association probably exists; −, evidence against an association based on results from moderate or high-quality studies; ?, insufficient evidence for or against an association; NA, not available. ^aStudies indicate that chronic renal dysfunction is associated with higher plasma tau but whether an association with cerebral tau deposition exists is unknown. ^bCaused by alterations in the gut microbiome.

Cardiovascular dysfunction

Frank cardiovascular disease has clearly established effects on brain health, most commonly via clinical stroke resulting from atrial fibrillation, left ventricular thrombus or patent foramen ovale¹⁷. Furthermore, clinically apparent cardiovascular disease, including myocardial infarction, is associated with greater risk of cognitive decline even in people without a history of stroke^18,19. In a consortium of 6 cohort studies, totalling 30,465 individuals, cognitive performance soon after myocardial infarction was comparable to that in people without myocardial infarction but decline over the years after myocardial infarction was steeper for global cognition, executive function and memory when compared with individuals in the control group¹⁸. Similar results were found for incident coronary heart disease more broadly, including myocardial infarction or angina, in the English Longitudinal Study of Ageing¹⁹. A systematic review of 16 studies including more than 1 million participants found that a history of coronary heart disease was associated with increased risk of all-cause dementia (risk ratio 1.27, 95% CI 1.08–1.50) but not the subset of dementia clinically attributed to Alzheimer disease (risk ratio 1.07, 95% CI 0.90–1.26)¹⁰. The apparent lack of association with clinically diagnosed Alzheimer disease should be interpreted with caution as Alzheimer disease diagnoses were not confirmed with biomarkers and a potential bias could be present in over-diagnoses of vascular dementia in people with vascular risk factors and coronary heart disease.

Cardiac procedures, such as coronary bypass graft surgery and coronary percutaneous intervention, have been associated with cognitive decline²⁰. However, another study suggests that the risk is primarily driven by the underlying cardiac condition itself rather than the procedure²¹. In this observational research study, the rate of cognitive decline was similar in people with coronary artery disease regardless of whether they underwent coronary artery bypass grafting with cardiopulmonary bypass or not²¹.

A heightened risk of cognitive decline and dementia is also present in individuals with heart failure²². A systematic review found that heart failure was associated with poorer performance on global cognition, memory, executive function and processing speed, with moderately large effect sizes (difference of >0.5 standard deviations)²². Another systematic review of high-quality studies that were mostly population-based and comprised 1.9 million participants found that heart failure was associated with a higher risk of all-cause dementia (risk ratio 1.59, 95% CI 1.19–2.13) and dementia related to clinical Alzheimer disease diagnoses (risk ratio 1.44, 95% CI 0.95–2.16)¹⁰.

Beyond clinically diagnosed cardiac disease, growing evidence suggests that subclinical cardiac disease contributes meaningfully to brain health. For example, structural and functional alteration of the atrium preceding atrial fibrillation (atrial cardiopathy) has been associated with incident stroke²³, silent infarcts and white matter hyperintensities on MRI²⁴ as well as with dementia²⁵. In addition, left ventricular structure changes, namely increases in left ventricular mass index and wall thickness, are associated with MRI-detected infarction and white matter hyperintensities²⁶. Moreover, left ventricular global longitudinal strain has been associated with greater stroke risk, particularly cardioembolic or cryptogenic stroke²⁷. Collectively, these studies suggest that structural and functional changes in the left atrium and left ventricle cause subclinical cerebrovascular damage and clinical stroke, which in turn are known to increase the risk of cognitive decline and dementia²⁸.

Shared risk factors, such as hypertension and diabetes, might explain some but not all of the association of subclinical and clinical cardiac disease with brain health. Vascular risk factors are present almost uniformly in individuals who develop abnormalities in cardiac structure and function, and these factors also increase the risk of stroke, subclinical cerebrovascular disease, and cognitive decline and dementia. For example, the Cerebral-Coronary Connection study demonstrated that coronary microvascular dysfunction was associated with MRI markers of cerebrovascular disease and cognitive function, consistent with both vascular beds being affected by the same risk factors²⁹. However, abundant research shows that the risk of cognitive decline and dementia in people with subclinical or clinical heart disease persists despite controlling for vascular risk factors^10,22, implying the presence of direct mechanisms that link heart disease to brain health. These mechanisms, which could include hypoperfusion, cardioembolism and the cerebral effects of systemic inflammation, warrant further study.

Renal dysfunction

Impaired kidney function is a common form of organ dysfunction encompassing subclinical impairment in glomerular excretory function, subclinical evidence of kidney damage reflected by urinary albumin loss, and clinically apparent disease requiring haemodialysis or renal transplantation³⁰. With a global prevalence of 9%, the health impact of these conditions is expected to increase in parallel to the increasing burden of its two leading causes: diabetes and hypertension³¹.

Although not all individual studies have been consistent, a meta-analysis of 28 longitudinal studies demonstrated an independent association of impaired kidney function with an increased risk of cognitive impairment and dementia¹¹. Heterogeneity across the studies might have been introduced by the use of blood creatinine to determine glomerular filtration rate (eGFR) as these estimations can be influenced by non-renal factors³². Indeed, in an analysis from the Atherosclerosis Risk in Communities prospective cohort study, lower eGFR based on cystatin C levels, which provides a more precise estimate of GFR than creatinine alone³³, was associated with an increased risk of dementia independent of diabetes and hypertension³⁴. Urine albumin to creatinine ratio, a marker of kidney damage, was also associated with higher dementia risk³⁴. Moreover, reduced excretory function and kidney damage were strikingly synergistic: individuals who had evidence of both conditions experienced more than a two-fold increased risk of dementia compared with people with neither condition³⁴. In addition, chronic kidney disease and proteinuria have been associated with increased risk of Parkinson disease³⁵.

Several studies have found that impaired kidney function is associated with imaging markers of cerebral microvascular disease^11,36. Intriguingly, emerging data link impaired kidney function to imaging markers of neurodegeneration, including both relatively non-specific markers, such as cortical atrophy³⁶, and more specific Alzheimer disease markers such as greater cortical Aβ deposition and higher plasma Aβ levels³⁷. A significant but weaker association has also been reported between impaired kidney function and higher plasma phosphorylated and total tau levels³⁸. These data imply that kidney function should be measured when employing plasma markers to diagnose Alzheimer disease, and caution is needed when interpreting Alzheimer disease markers in individuals with kidney disease³⁸. Additionally, these data enrich our understanding of the pathophysiology of Alzheimer disease, whereby impaired kidney function might impede systemic clearance of Aβ, resulting in greater cortical accumulation (see the section ‘Systemic production and clearance of toxic metabolites’).

Respiratory dysfunction

Chronic lung disease is thought to be associated with dementia owing to overlapping risk factors such as smoking and hypertension as well as a direct effect of chronic hypoxaemia from impaired lung function. In a study of 10,975 people, restrictive but not obstructive lung dysfunction was associated with reduced cognitive score and higher dementia risk, though no association was found between lung function and cognitive decline over time¹². A follow-up study of the same cohort showed that participants with restrictive or obstructive lung disease had an increased risk of mild cognitive impairment and dementia³⁹. Consistent with this finding, another study of 4,765 people found that both chronic obstructive pulmonary disease (COPD) and restrictive lung function were each associated with worse global cognitive performance⁴⁰. Furthermore, adult asthma, another prevalent obstructive lung disease, was associated with an increased risk of dementia⁴¹. Last, a meta-analysis of 11 studies comprising more than 1 million participants showed that people with sleep apnoea have an increased risk of developing any type of neurocognitive disorder such as Alzheimer disease or Parkinson disease¹³. This evidence across multiple lung disorders demonstrates a consistent association between lung dysfunction and adverse brain health.

To date, there are few data on the cerebral pathologies that underlie dementia associated with lung dysfunction. Associations with Alzheimer disease biomarkers have not yet been explored and, to our knowledge, no autopsy studies have been conducted. However, a combined analysis of six cohort studies suggested that lower lung function affects the cerebral vasculature⁴². In this analysis, lung function, measured as the ratio of forced expiratory volume in one second to the forced vital capacity, was associated with a higher burden of white matter hyperintensities and lower brain volumes⁴².

Gut dysfunction and microbiome

Gastrointestinal dysfunction, and particularly alterations in the microbiota composition, has been proposed to be a modulator of brain health and a potential risk factor in neurodegenerative disorders. In support of a role for gut dysbiosis in dementia pathophysiology, several gastrointestinal diseases have been associated with risk of dementia. In a 16-year follow-up study of 1,742 people with inflammatory bowel disease and 17,420 healthy controls, the condition was associated with an increased risk of dementia, including clinically diagnosed Alzheimer disease⁴³. Similarly, in a cohort study of 458,181 participants over a 12-year follow-up period, an increased incidence of early-onset dementia was found in people with irritable bowel syndrome, gastritis, duodenitis, gastroesophageal reflux disease or peptic ulcer disease¹⁴. In addition, other studies of a national insurance data base found that gastroesophageal reflux disease and irritable bowel syndrome were associated with future risk of dementia^44,45.

The gut–brain axis is posited to contribute to the pathophysiology of dementia through interactions between the central and enteric nervous system, microbiota-derived metabolites, and immunological mechanisms⁴⁶. Genome-wide association studies have identified genetic overlap between the risk of gastrointestinal disorders and Alzheimer disease, specifically in pathways involved in lipid metabolism, autoimmunity and programmed cell death 1 signalling⁴⁷. However, to date, little evidence exists for a direct link between gastrointestinal disorders and neuropathological or biomarker evidence of Aβ or tau. Similarly, little evidence supports an association between gastrointestinal disorders and cerebral small vessel disease. By contrast, changes in the gut microbiome have been detected in people with biomarker-confirmed Alzheimer disease⁴⁸. These changes include overall decreased microbial diversity, along with increases in the abundance of the Bacteroidetes phylum and decreases in the abundance of the Firmicutes and Actinobacteria phyla. These shifts in microbiota phyla could affect amyloidogenesis and disease progression in Alzheimer disease not only by modulating neuroinflammation but also by shaping systemic immune responses^49,50 (see the section ‘The microbiome as a modulator of systemic immunity’).

A link between gastrointestinal dysfunction and neurodegenerative synucleinopathies, such as Parkinson disease and LBD, has also been identified. Gastrointestinal dysfunction, including dysphagia, gastroparesis and constipation, are prevalent in Parkinson disease and LBD, and are characteristic of the disease prodrome, often appearing years or decades before the onset of motor symptoms^51-53. Furthermore, early constipation predicts faster dementia onset in Parkinson disease⁵⁴. Additionally, the hallmark Lewy pathology composed of aggregated α-synuclein is found throughout the enteric nervous system, and early-stage pathology is found in the brain in the dorsal motor nucleus of the vagus, which connects the enteric nervous system to the CNS⁵⁵. These studies suggest that the spread of Lewy bodies between the brain and the gut, mediated by the vagus nerve, occurs at an early stage of LBD and causes clinically relevant gastrointestinal dysfunction.

Liver dysfunction

Chronic liver conditions exist on a spectrum; chronic liver injury from nonalcoholic fatty liver disease (NAFLD) and alcohol-associated injury, as examples, can result in the development of liver fibrosis and eventually cirrhotic liver disease⁵⁶. The prevalence of these conditions varies with the population studied and the method of ascertainment. Strikingly, the prevalence of liver fibrosis — a typically silent condition that nonetheless has critical clinical significance as a precursor to symptomatic cirrhosis⁵⁷ — ranges from 2% to 19%⁵⁸.

A broad liver–brain axis is receiving growing recognition in the literature. A 2023 hypothesis-free, machine learning analysis of participants in the UK Biobank study identified a liver-related measure as the second strongest risk factor for dementia after apolipoprotein E allelic status⁵⁹. Although the hepatic measure, that is, the ratio of aminotransferase (AST) to alanine aminotransferase (ALT), identified by this analysis is non-specific, other studies have investigated more specific chronic liver condition phenotypes, NAFLD and liver fibrosis. The association of NAFLD with cognitive impairment and dementia is confounded by shared risk factors, and studies have found mixed findings indicating a detrimental impact of NAFLD on brain health^60-62. In contrast, more consistent data demonstrate that liver fibrosis, including that identified in individuals without known chronic liver conditions, is associated with cognitive impairment and an elevated risk of dementia^15,62,63.

The liver could influence brain health through several potential mechanisms. Substantiating the association of liver conditions with cognitive impairment and dementia, studies in the past 5 years have demonstrated links between liver conditions and imaging biomarkers of Alzheimer disease and cerebrovascular disease. For example, NAFLD and liver fibrosis have been linked to lower brain volume, which is a relatively non-specific Alzheimer disease imaging biomarker^15,62. Further, liver fibrosis (but not NAFLD) has also been associated with greater cortical Aβ deposition and tau accumulation on PET brain imaging⁶⁴. So far, few studies have explored the association of NAFLD with cerebrovascular disease, with conflicting results^65,66. However, biopsy-proven NAFLD has been associated with imaging markers of cerebral small vessel disease⁶⁷. Last, a gut–liver–brain axis has been proposed in which gut dysbiosis leads to gut bacteria and their metabolites entering the liver and affecting its function with secondary effects on the brain⁶⁸ (see the section ‘The microbiome as modulator of systemic immunity’).

Endocrine dysfunction, using diabetes as an example

The brain and the endocrine system are closely intertwined, both in health and disease. The brain directs the activity of most, if not all, endocrine glands and, in turn, peripheral hormones derived from the glands provide feedback to the brain⁶⁹. Furthermore, hormones exert complex actions on the brain, including trophic support to neurons and short-term and long-term modulation of essential aspects of behaviour, for example, feeding and reproduction. Thus, subtle and prolonged alterations in endocrinological systems can have a profound impact on brain health in the long term.

Here, we focus on type 2 diabetes as an example of endocrine dysfunction. A large body of literature shows that type 2 diabetes is associated with a range of brain changes that extend well beyond the acute changes in brain function that relate to acute, severe hypoglycaemia or hyperglycaemia⁷⁰. These brain changes manifest clinically through an increased risk of stroke and dementia but also through an increased occurrence of depressive disorders⁷⁰. For example, in dementia and ageing-related cognitive decline, a meta-analysis of 122 studies reported that diabetes conferred a 1.25–1.91-fold excess risk of cognitive disorders, including mild cognitive impairment, any type of dementia and a clinical diagnosis of Alzheimer disease¹⁶. This increased risk is already present in prediabetic stages of glucose dysmetabolism⁷¹, which aligns with the notion that the associations between type 2 diabetes and worse brain health are not entirely owing to abnormal glucose levels. Indeed, little modulation of dementia risk is seen across different levels of glycaemic control in people with diabetes, except for individuals with poor control reflected in HbA_1c concentrations of 9% or more⁷².

The pathways from diabetes to dementia are not fully understood. Biomarker studies have failed to find associations between diabetes and either PET or cerebrospinal fluid (CSF) evidence of Aβ^73,74, suggesting that the link between diabetes and dementia is not mediated by an increased risk of Alzheimer disease pathology. Neuropathology studies have also failed to show correlations between diabetes and Aβ pathology^75,76; instead, they suggest that diabetes causes cognitive decline and dementia through its effects on the cerebral vasculature, increasing the risk of symptomatic stroke and silent brain infarction. Some studies suggest that brain tau is elevated in people with diabetes, even in those who are negative for Aβ, for unclear reasons^74,77.

Diabetes-associated brain changes are likely to be highly multifactorial; type 2 diabetes develops in the context of environmental and societal factors, affecting, among other factors, physical activity and diet on a background of genetic predisposition. Furthermore, prediabetic conditions, such as metabolic syndrome, not only predict development of diabetes but also changes to other organs that are addressed in this Review, including the cardiovascular system and kidneys, and are associated with alterations in the immune system, gut microbiome and sleep⁷⁸. Once diabetes has developed, further complications in these other organ systems can arise, which might in turn further affect the brain. Therefore, rather than focusing on single pathways, a systems biology approach is likely needed to understand the impact of type 2 diabetes on the brain, identifying key driving pathways to brain injury, which will probably show substantial interindividual variation.

Systemic mechanisms mediating brain dysfunction

This section examines the basic mechanisms by which organ dysfunction can lead to cognitive impairment. These systemic pathogenic factors are likely to act in concert with intrinsic pathogenic processes in the brain — vascular, neuroimmune and neurodegenerative — to contribute to age-related cognitive decline. We examine pathogenic factors common to multiple organs that could affect the brain as well as important organ-specific mechanisms (Fig. 2).

Fig. 2 ∣. Mechanisms linking peripheral organ dysfunction to brain health.

Endothelial dysfunction⁸⁹, reduced peripheral clearance of toxins¹²⁶, altered microbiome composition¹⁸² and systemic inflammation¹⁵⁶ have been linked to neurovascular unit dysfunction in experimental studies. CSF, cerebrospinal fluid.

Hypoxaemia and hypercapnia

Hypoxaemia and hypercapnia are most directly related to lung dysfunction resulting in subclinical respiratory insufficiency in the setting of COPD and obstructive sleep apnoea. Hypoxaemia has profound effects on brain structure and function owing to neuronal loss in vulnerable regions such as the hippocampus, resulting in sustained cognitive impairment⁷⁹. Several potential mechanisms could underly neuronal loss associated with hypoxaemia. A study in rats suggested a link between cerebral hypoxia and tau hyperphosphorylation, leading to cognitive impairment⁸⁰. In addition, an increase in blood–brain barrier (BBB) permeability has been reported in ultra-elite breath-hold divers following prolonged breath holding that reduced arterial pO₂ from a mean of 97 to 36 mmHg (ref. 81). Because increased BBB permeability has in turn been linked to the development of MRI white matter hyperintensity⁸², this experiment suggests a potential mechanism by which intermittent severe hypoxia could injure cerebral white matter⁴². Moreover, hypoxaemia is associated with hypercapnia, which could also have deleterious effects on the brain especially in combination with hypoxaemia^83,84. However, a clear link between chronic hypoxaemia and cognitive impairment in COPD has not yet been established⁸⁵, and COPD is also associated with alterations in brain metabolism⁸⁶ and brain connectivity between neuronal hubs⁸⁷ as well as with markers of inflammation⁸⁸ (see the section ‘Systemic inflammation’), which could additionally play a part.

Systemic vascular factors

The cerebral vasculature is a prime target of early-stage organ dysfunction⁸⁹. A complicating factor in assessing the cerebrovascular impact of subclinical organ disease is the coexistence with vascular risk factors such as hypertension, diabetes and hyperlipidaemia, which can also affect brain vessels⁹⁰. Cerebrovascular alterations have profound effects on brain health by reducing cerebral perfusion⁹¹ or by impairing the clearance of potentially toxic by-products of brain activity⁹².

Cerebrovascular insufficiency.

The brain requires a well-regulated and constant delivery of O₂ and glucose through the cerebral blood vessels⁹³. Severe reductions in cerebral blood flow lead to ischaemic brain injury, and more subtle alterations in cerebrovascular function are often associated with cognitive impairment⁹¹. Thus, cerebral endothelial cells are critically important for the maintenance of brain health⁹⁴ but are particularly susceptible to organ dysfunction⁸⁹. Endothelial function is altered in the setting of cardiac, kidney and liver diseases as well as diabetes^95-97, and is worsened by coexisting pathologies such as systemic inflammation, oxidative stress and gut dysbiosis (see the section ‘Systemic inflammation’)⁹⁶.

Cerebral endothelial dysfunction can damage the brain through a wide variety of mechanisms. Dysfunction of microvascular endothelial cells impairs microvascular flow regulation and results in increased BBB permeability⁹¹. Cerebral perfusion might also be compromised by endothelial dysfunction through the promotion of leukocyte adhesion, thrombosis and embolism⁹⁸. In both systemic and brain vessels, endothelial dysfunction could alter cerebral perfusion by promoting atherosclerosis and vascular stiffening, resulting in large artery obstruction or mechanical damage from the pressure wave reaching the neurovasculature⁹⁹, respectively. The mechanisms by which endothelial dysfunction alters vascular structure and promotes atherosclerosis remain to be fully elucidated but might involve vascular oxidative stress and inflammation¹⁰⁰, reduction in endothelial nitric oxide¹⁰¹, as well as genetic and epigenetic factors affecting the extracellular matrix, ultimately leading to increased collagen and reduced elastin¹⁰².

Beyond direct effects of impaired perfusion, loss of endothelial growth factors undermines trophic support to neuronal and oligodendrocyte precursors, reducing neuroplasticity and promoting diffuse white matter injury^103-105, which is common in heart, liver and kidney diseases^66,106,107. Furthermore, endothelial dysfunction is associated with a reduced circulating endothelial progenitor pool, which impairs the renewal of the cerebral endothelium, causing endothelial senescence¹⁰⁸ — a major causative factor in cerebrovascular pathologies¹⁰⁹. Moreover, endothelial dysfunction also results in loss of endothelial nitric oxide, which can promote neurodegenerative pathology¹¹⁰.

Beyond shared risk factors, subclinical cardiac disease itself can confer a specific risk of adverse brain health¹¹¹. A key mechanism involved is reduced cerebral perfusion, which is observed in people with subclinical and clinical disease¹¹². The reduction in cerebral blood flow can result from systolic dysfunction as well as age-related alterations in the ability of cerebral blood vessels to maintain constant blood flow during changes in blood pressure (cerebrovascular autoregulation)¹¹²; these cardiac-induced reductions in blood flow have been linked to increased permeability of the BBB¹¹². In addition, cardiac dysfunction has been linked to increased circulating markers of inflammation (see the section ‘Systemic inflammation’), which could lead to endothelial dysfunction and contribute to the alterations in cerebral perfusion and BBB permeability.

Reduced clearance of brain toxins.

The brain produces toxic metabolites and proteins during neural activity, including Aβ and tau, which need to be quickly removed to maintain the integrity of the brain’s internal milieu¹¹³. Research in animals and humans has begun to elucidate how the brain rids itself of these potentially toxic agents. Proposed clearance mechanisms include a perivascular pathway that carries solutes retrogradely through the perivascular space up to the subarachnoid space; a ‘glymphatic’ pathway that moves solutes anterogradely into the perivascular space, which then reach the subarachnoid space through astrocytic end-feet and the venous perivascular space; and a transvascular pathway that transports solutes directly into the bloodstream¹¹⁴. From the subarachnoid space, solutes are then removed through CSF clearance systems, including the arachnoid granulations and the lymphatic system¹¹⁵. The efficiency of these vascular and lymphatic clearance systems depends on the integrity of cerebral vasomotor function, and experimental studies suggest that ageing and vascular risk factors reduce brain clearance^92,116. Therefore, vascular dysfunction occurring in the setting of subclinical organ failure is likely to compromise the removal of potentially neurotoxic agents like Aβ and tau, promoting their accumulation in the brain. This hypothesis is supported by studies showing that the deposition of Aβ and tau is increased in people with cardiovascular and cerebrovascular diseases^117-120.

Systemic production and clearance of toxic metabolites

Overproduction or clearance of metabolites can be pathologically affected by peripheral organ dysfunction and might alter brain health. In kidney disease, increased levels of cystatin C, homocysteine, fibroblast growth factor 23 and guanidine compounds have been suggested to contribute to brain pathology¹¹. In liver disease, AST and ALT metabolic products as well as modified lipids have been linked to cognitive impairment^121,122, and in atrial cardiopathy, elevated levels of n-terminal pro-B-type natriuretic peptide — a marker of heart failure — has been associated with cognitive decline, smaller brain volume and higher MRI white matter hyperintensity burden¹²³. However, the contribution of these factors, relative to other consequences of organ dysfunction, such as inflammation and oxidative stress, remains unclear.

Failure of the systemic organs to clear potentially neurotoxic substances might also contribute to cognitive decline and risk of dementia. For example, Aβ is transported from the brain to the circulation¹²⁴ and is rapidly removed by peripheral clearance mechanisms, involving gut, liver, kidney and skin^125,126. The kidney filters and excretes Aβ in the urine and, consequently, kidney dysfunction is associated with increased plasma Aβ¹²⁷. In the liver, Aβ is metabolized in hepatocytes and excreted in the bile, and plasma Aβ levels are inversely related to liver function¹²⁸. One hypothesis supported by animal studies is that liver disease further reduces the age-related decrease in the ability of the liver to clear Aβ¹²⁹; however, this hypothesis needs to be tested in human studies using Alzheimer disease biomarkers.

Possessing one or more copies of the apolipoprotein E4 (APOE4) allele is a strong risk factor for Alzheimer disease¹³⁰. The liver is responsible for 90% of peripheral APOE production¹³¹. Expressing human APOE4 in the mouse liver resulted in neurovascular endothelial dysfunction and enhanced brain Aβ accumulation¹³², raising the possibility that liver dysfunction impacts brain health through changes in lipid metabolism. Other emerging molecular mechanisms in the liver–brain axis include bile acid dysregulation¹³³ and perturbations in levels of exercise-induced neuroprotective hepatokines¹³⁴.

The neurotoxicity of tau is well characterized, although our knowledge of the peripheral clearance of this protein is limited. However, an endothelial tau transport mechanism from the brain to blood has been identified¹³⁵. Furthermore, tau is also thought to be transported from the brain to blood by perivascular and glymphatic interstitial fluid and CSF clearance routes¹³⁶ (see the section ‘Reduced clearance of brain toxins’), and is cleared mainly by the kidney and metabolism by blood cells¹³⁷. The role of the kidney in tau clearance is supported by studies from the past few years that indicate an association of kidney dysfunction with higher levels of circulating phosphorylated tau^38,138.

Systemic inflammation

Systemic inflammation has emerged as a common feature across dysfunction of various organs, exhibiting a stereotypic immunological response. Although the effects of systemic inflammation on brain function under conditions of severe infection, such as sepsis, have been well described, the transduction of inflammatory signals from the periphery to the brain in other organ dysfunction and cognitive decline are surprisingly insufficiently understood.

In animal models and humans, sepsis leads to microglial activation, disturbance of astroglial function, neuronal cell death and cognitive decline via inflammatory mediators, including inducible nitric oxide synthase and pro-inflammatory cytokines¹³⁹. However, inflammation can also be induced by non-infectious, sterile mediators of organ dysfunction ranging from subclinical changes in organ-specific tissue dyshomeostasis to acute tissue injury^140,141. Whether mild forms of systemic inflammation induced by systemic conditions such as liver disease, atherosclerosis or diabetes can also affect neurovascular unit function and influence neurodegenerative disease processes is not yet fully clarified. Yet, for other conditions, including pulmonary dysfunction, their impact on brain health and cognition via systemic inflammation is more apparent; for example, COPD, which has been identified as a risk factor for dementia^39-41, is widely recognized by pulmonologists as a systemic inflammatory disorder¹⁴². Likewise, studies suggest that intestinal inflammation, immune cell migration and auto-reactive T cells are involved in the pathogenesis of Parkinson disease¹⁴³.

Circulating cytokines and their effect on brain health.

An increase in blood concentration of pro-inflammatory cytokines, including IL-1β, tumour necrosis factor, IL-6 and others, has been demonstrated for a range of subclinical organ dysfunctions of the heart, lung, kidney and liver^144,145. The most likely shared mechanism across peripheral organ disease that results in this stereotypical increase in pro-inflammatory cytokine secretion is the sensing of mediators (damage-associated molecular patterns) that are released from dysfunctional organs under conditions of cell stress or injury¹⁴⁶ and induce similar inflammatory cascades as pathogens¹⁴⁷. For example, COPD has been associated with increased circulating levels of cytokines and a pro-inflammatory polarization of circulating myeloid cells, which could negatively affect brain health and contribute to cognitive decline¹⁴⁸.

Physiological functions of the neurovascular unit, and particularly the increase in cerebral blood flow evoked by neural activity (neurovascular coupling), are tightly regulated multicellular processes that can be modulated and negatively affected by external factors, including circulating cytokines⁹¹. For example, experimental induction of a systemic immune response is sufficient to alter the neurovascular response evoked by neural activity in a rat¹⁴⁹. Endothelial cells are likely to be the first ‘sensors’ and cellular mediators of changes in cerebrovascular function caused by a systemic inflammatory response^150,151. Systemically derived cytokines induce profound changes in endothelial function, including upregulation of adhesion molecules, endothelial phagocytosis and antigen presentation¹⁵¹. Such disruption of physiological endothelial function in conditions of inflammatory immune-to-brain communication results in suppression of neurovascular coupling, which has been linked to cognitive decline^152-155.

In addition to affecting the cerebrovascular endothelium, cytokines can enter the brain parenchyma by passing through the BBB¹⁵⁶. Thereby, astrocytes and oligodendrocytes can be directly exposed to cytokines, resulting in phenotypic changes and alterations of their housekeeping functions. For example, IL-1β stimulation of astrocytes can alter neuronal excitability, survival and extracellular neurotransmitter transport¹⁵⁷. Likewise, IL-1β reduces oligodendrocyte maturation, resulting in myelination defects in rats¹⁵⁸. These direct adverse effects of cytokines that pass through the BBB might promote neurodegeneration or decrease brain resilience to age-related neuropathologies such as Alzheimer disease.

Cellular immunity: adaptive and innate immune memory.

Organ dysfunction results not only in an acute, systemic inflammatory response but can have long-term effects on the reactivity of the immune system to consecutive stimuli. Adaptive immune response, antigen-specific clonal expansion and T cell memory have been described in various organ disorders, including lung (asthma, fibrosis), kidney (ischaemia), muscle (ischaemia, trauma) and liver injury (toxic hepatitis)^159-161. However, over the past decade, the memory of the innate immune system has also drawn a lot of interest, challenging the traditional view that only the adaptive arm of the immune system can build immunological memory. Epigenetic changes in innate immune cells have been demonstrated mainly after infections but also in some emerging studies of sterile tissue injuries¹⁶². As such, persistent epigenetic changes in the innate immune system can increase the production of pro-inflammatory mediators and the reactivity to secondary pathologies, including brain disorders¹⁶².

Ageing, organ dysfunction (including COPD and cardiac insufficiency) and metabolic disorders (including obesity and diabetes) have been associated with the occurrence of clonal expansion of haematopoietic cells¹⁶³. Clonal haematopoiesis results in chronic systemic inflammation owing to expansion of myeloid cell clones with a pro-inflammatory phenotype; in turn, inflammation itself is a strong driver of clonal haematopoietic selection that leads to propagation of chronic inflammation¹⁶⁴. This predisposition to elevated levels of chronic inflammation could promote neurodegeneration. However, in contrast with this assumption, a population-based study including 3,180 people found that clonal haematopoiesis is associated with reduced incidence of Alzheimer disease¹⁶⁵. This potential protective effect was hypothesized to be owing to increased phagocytosis of Aβ by marrow-derived cell clones that infiltrated the brain. Therefore, the health impact of clonal haematopoiesis seems to be organ and disease specific, a hypothesis that needs to be tested in other brain and systemic diseases.

Border-associated macrophages (BAMs) in the meninges and perivascular space contribute substantially to the cerebral immunological milieu and the pathophysiology of cognitive decline in vascular and amyloid-related dementia^166,167. Additionally, BAMs are the predominant macrophage population in an α-synuclein model of Parkinson disease¹⁶⁸. Following brain injury, BAMs, unlike microglia cells, can be repopulated by circulating monocytes¹⁶⁹. Thus, the response of BAMs to local vascular injury or protein aggregates might be influenced by systemic alterations of circulating monocytes owing to epigenetic-trained immunity or clonal selection. Indeed, repopulation by BAMs is increased in ageing and leads to a phenotypic shift in the cerebral macrophage population with the ability to promote neuroinflammation¹⁶⁹. Moreover, this response is associated with impaired BBB integrity and reduced cognitive function^170-172.

The microbiome as a modulator of systemic immunity.

The human gut microbiome, a complex ecosystem of trillions of microorganisms residing in the gastrointestinal tract, has a crucial role in modulating immunity and influencing the inflammatory response in both health and disease. In a state of symbiosis, the gut microbiota interacts extensively with the host immune system, contributing to its development, maturation and maintenance of homeostasis¹⁷³. Alterations in the gut microbiome, known as dysbiosis, can have profound consequences for immune regulation and contribute to the development of various inflammatory diseases¹⁷⁴. A reduced abundance of beneficial commensal bacteria and an overgrowth of potentially pathogenic species have been associated with multiple organ disorders remote of the intestinal tract, including heart failure, asthma, fatty liver disease, diabetes and others^175-178. Additionally, growing evidence suggests that the lung microbiome plays a part in modulating CNS immune responses by influencing microglial activation¹⁷⁹; however, the relevance to Alzheimer disease and brain health more generally is not yet known.

Several mechanisms underlie the influence of the gut microbiome on immunity. Microbial metabolites, such as short-chain fatty acids (SCFAs), act as ligands for host immune receptors, promoting immune polarization including the differentiation of regulatory T cells that suppress excessive inflammation⁴⁶. Metabolites derived from the gut microbiome, particularly SCFAs, have been recognized as key humoral mediators in the gut–brain crosstalk that influence gut motility, inflammatory responses and neurophysiology¹⁸⁰. Notably, SCFAs have been implicated in both Alzheimer disease and Parkinson disease, highlighting their potential as therapeutic targets. Specifically, SCFAs are decreased in stool samples from people with Parkinson disease and correlate with some indicators of disease severity, including non-motor symptoms¹⁸¹. In Alzheimer disease, SCFAs might be neuroprotective through a number of mechanisms, including rectifying hypometabolism, modulating disease-associated epigenetic mechanisms, promoting brain microglia maturation, conferring more appropriate responsiveness of the brain microglia to stimulating factors, and directly interfering with protein aggregation¹⁸².

Microbiota-derived metabolites can also affect secondary organ function with consequences on brain health. For example, gut dysbiosis and resulting changes in the systemic metabolome have been shown to affect liver function with secondary effects on the brain⁶⁸. Further, the vagus nerve serves as a critical intermediary in the gut–brain axis and orchestrates gastrointestinal functions and immune responses^46,183,184. Correspondingly, people who received truncal vagotomy for the now obsolete treatment of peptic ulcer disease have a lower risk of developing Parkinson disease¹⁸⁵, and data from animal models support a gut-to-brain spread of pathology along the vagus nerve¹⁸⁶. Nevertheless, the exact mechanism by which the vagus nerve facilitates the connection between the gut microbiome and disease progression requires further detailed investigation¹⁸³.

Implications for brain health promotion and dementia prevention

The knowledge that even subclinical dysfunction of many organs plays an important part in the risk of dementia offers new avenues for brain health promotion. However, to capitalize on these opportunities, changes will be needed in defining and managing several stages of risk. Here, we define these stages as, first, the systemic predisposition to dementia, where peripheral organs are currently functioning normally but a predisposition to systemic disturbance is present, determined by genetic and lifestyle factors; second, the systemic at-risk stage, where systemic factors are moving out of balance and beginning to contribute to peripheral organ and brain dysfunction; and last, the systemic disease stage, where frank organ disease is damaging the brain.

For the predisposition stage, individuals should be made aware that maintaining a healthy heart, liver, lungs and kidneys will sustain brain health as well as body health. At this stage, lifestyle modification is the preferred approach. Unfortunately, many people are unaware of risk factors for dementia, particularly the risk factors that are related to peripheral organs. The relationship between cardiovascular risk and dementia is an illustrative example and a cautionary tale; despite more than 15 years of knowledge that cardiovascular disease is strongly related to dementia risk, public knowledge remains low. For example, some surveys show that as few as 25% of the general population recognize that hypertension is a risk factor for dementia¹⁸⁷.

In the systemic at-risk stage, a role for health-care practitioners is required to identify and manage risk factors such as hypertension and obesity. Spreading awareness of the relationship between organ dysfunction and dementia risk could help to motivate individuals to comply with prevention and management programmes for at-risk conditions not previously known to also confer risk of dementia such as NAFLD (see the section ‘Liver dysfunction’). However, this strategy relies on raising awareness among health-care workers of the links between dementia and organ dysfunction. For example, many health-care practitioners (46% in one survey)¹⁸⁸ are unaware of the association of cardiovascular risk factors with dementia, and awareness of other organ-related risk factors is undoubtedly lower. Beyond improving awareness in general, future advances in proteomics and metabolomics could uncover signatures of metabolic dysfunction that provide opportunities for individual precision prevention or even targets for future drug development.

In the systemic disease stage, greater focus is needed on identifying and supporting people with cognitive consequences of organ dysfunction. This stage provides an area of opportunity as people with known organ dysfunction generally receive medical care, yet accompanying cognitive dysfunction often goes undetected¹⁸⁹. Efforts should be made to improve the identification of cognitive dysfunction among such individuals, for example, by implementing cognitive screening in people at high risk such as those with diabetes or cardiac disease¹⁹⁰.

An important benefit of identifying cognitive impairment is to allow personalized support for better management of systemic disease. Enhancing the capacity for early detection of cognitive decline can start with quick screening questions on cognitive symptoms and functions followed, based on the level of concern, by standardized cognitive assessment. Identification and support of people with cognitive impairment could improve quality of life, increase time spent living well in the community, and help individuals to better manage their organ diseases. A ‘heart–brain clinic’ model proposed in 2023 can serve as a guide for other diseases associated with a high risk of cognitive impairment¹⁹⁰.

Public health policy should focus on spreading awareness of systemic risks for dementia as well as providing incentives for good general preventive health care and access to health care. Improving organ health might address some of the current racial and socioeconomic disparities in dementia burden as many of the diseases discussed here have a higher prevalence in racial minority groups and people with lower socioeconomic status and are already known to contribute to differences in mortality depending on race and socioeconomic status¹⁹¹.

Conclusions

The evidence discussed here highlights the association of peripheral organ dysfunction and subclinical disease with brain health and increased dementia risk. This relationship poses an important alternative narrative to the traditional notion that the major threats to the ageing brain come from within the brain itself, creating new opportunities for intervention. The brain requires a supportive environment maintained by healthy organ systems to function optimally and to confer resilience to age-related neurodegenerative diseases such as Alzheimer disease and Parkinson disease. Important associations exist between dementia risk and diseases of the cardiovascular system (including myocardial infarction and heart failure), kidney (chronic kidney disease), lungs (including obstructive and restrictive disease), intestine (inflammatory bowel disease), liver (cirrhosis, fibrosis and possible NAFLD) and endocrine system (including diabetes).

Clinical and basic science studies, including in model systems, suggest that organ dysfunction can cause endothelial dysfunction, reduced clearance of wastes (such as Aβ), systemic inflammation, and alterations in the microbiome that affect the cerebral vasculature, brain resident immune system and glia. These alterations can disrupt the normal functions of the neurovascular unit leading to cognitive impairment and risk of dementia. Optimizing peripheral organ health to prevent cognitive decline and dementia can address three stages of vulnerability: predisposition to organ dysfunction (by promoting healthy lifestyle and behaviour), people at risk of organ dysfunction (by addressing risk factors such as hypertension) and people with frank organ disease (by disease-specific treatments).

More research is needed to confirm early studies of association, identify mechanisms that link organ dysfunction to dementia and, most importantly, to determine if improving organ function reduces dementia risk. Basic science studies can further explore potential mechanisms of inflammation, reduced waste clearance, endothelial dysfunction and altered microbiota. This endeavour should involve animal models of organ diseases, including crosses with Alzheimer disease animal models, and brain readouts.

Clinical research studies can include genomics, proteomics and metabolomics to identify the alterations in homeostasis and metabolism that precede or result from organ dysfunction and affect the function of the neurovascular unit. For cardiovascular, renal, pulmonary and gastrointestinal diseases, more data are needed to determine whether their associations with dementia are mediated by a higher risk of Alzheimer disease or Parkinson disease. Because biomarkers of Alzheimer disease can now be measured accurately in peripheral blood¹⁹², the determination of whether peripheral organ disease or dysfunction is associated with Alzheimer disease pathophysiology is increasingly feasible in cohort studies with stored blood samples. The same approach can likely be extended to Parkinson disease given the improving accuracy of blood biomarkers¹⁹³.

Verification that organ function improvement reduces the risk of dementia would ideally be done via randomized controlled trials, but the required size and duration of the studies might be prohibitive. To start, a more pragmatic approach would be to include secondary assessments of cognition and cognitive-related function in clinical trials of treatments for systemic organ diseases. Indeed, advances in the accuracy of plasma Aβ and tau measurements enable the acquisition of evidence of the Alzheimer disease pathological continuum in larger-scale studies.

Other approaches to confirm the relationship between organ function and dementia risk would be to use modern epidemiological techniques, including Mendelian randomization, trial simulation using propensity weighting and mediation analyses, to make stronger causal inferences from observational data. Epidemiological studies should consider whether improvement in organ health accounts for some of the variation in dementia incidence over time. For example, a pooled analysis of 7 cohorts with 49,202 people across Europe and North America suggests that the age-standardized incidence of dementia has declined since the 1990s¹⁹⁴; improvement of cardiovascular health over time has been cited as a potential contributor. Our Review highlights that improvements in the function of other organs, such as the lung or kidney, could also have a role in the declining incidence of age-standardized dementia. On the other hand, some adverse temporal trends, such as the rising prevalence of diabetes in young people¹⁹⁵, might be holding back gains in dementia prevention.

In this Review, we focused on six specific organ systems and their association with risk of Alzheimer disease, Parkinson disease, vascular cognitive impairment and dementia. However, these organs are only a sample of the organ systems that have been linked to brain health. Other organ systems that have been associated with a risk of later-life cognitive decline and dementia include the hypothalamic–pituitary–adrenal axis, female and male reproductive systems, muscles (that signal to the brain based on physical activity), thyroid, parathyroid and the immune system.

The focus of this Review was on the causes of dementia because dementia has a major impact on brain health in later life. However, the mechanisms linking peripheral organ dysfunction to dementia are also likely to influence other aspects of brain health. To provide one of many examples, our understanding of links between the microbiome and mental health is growing¹⁹⁶. Future research is needed on the relationship between peripheral organ health and brain health more generally, including social, behavioural and mental health, as well as physical function and mobility. The influence of ageing on the effect of organ dysfunction on the brain is also of great interest. A 2023 study on human plasma proteins revealed previously unrecognized organ-specific signatures of ageing that are accelerated by organ disease¹⁹⁷. However, more research is needed to elucidate how cellular ageing and senescence underlie age-related organ changes with secondary effects on brain health as well as the role of multi-morbidity and frailty in these processes.

In summary, growing evidence demonstrates that systematic organ dysfunction is a substantial contributor to cognitive decline and dementia and that improving overall organ health might be an underappreciated strategy for promoting brain health and preventing dementia.

Key points.

• Brain health is a worldwide priority, and dementia is the biggest threat to healthy brain ageing.

• The brain requires a supportive environment maintained by healthy organ systems to function optimally and to confer resilience to neurodegenerative diseases such as Alzheimer disease and Parkinson disease.

• Increasing evidence links subclinical dysfunction of the cardiovascular, pulmonary, renal, gastrointestinal, hepatic and endocrine systems with age-related cognitive decline and dementia.

• Cardiovascular, renal and hepatic dysfunction as well as changes in the gut microbiome have been associated with the presence of Alzheimer disease biomarkers in human research.

• Pathways that might lead from organ dysfunction to brain dysfunction include systemic endothelial dysfunction, reduced peripheral clearance of wastes (including amyloid-β and tau), systemic inflammation and alterations in the microbiome.

• Future research needs to explore mechanisms that link peripheral organ dysfunction to risk of age-related neurodegenerative and cerebrovascular diseases and to elucidate whether improving organ function reduces this risk.