Epidemiology, pathology, and pathogenesis of Alzheimer disease

INTRODUCTION — Alzheimer disease (AD) is the most common cause of dementia and one of the leading sources of morbidity and mortality in the aging population.

The hallmark neuropathologic changes of AD are diffuse and neuritic plaques, marked by extracellular amyloid beta deposition, and neurofibrillary tangles, comprised of the intracellular accumulation of hyperphosphorylated tau (p-tau) protein. The epidemiologic study of AD is being transformed by the availability of new biomarker technologies to measure such neuropathologic changes in vivo. Large randomized clinical trials are evaluating anti-amyloid and other disease-modifying therapies for the treatment and prevention of AD utilizing these imaging or cerebrospinal fluid (CSF) biomarkers [1].

The study of AD has focused on three interrelated hypotheses [2]:

●Extracellular amyloid plaques (primarily composed of amyloid beta peptides) are a unique genetic and lifestyle disease due to increased production of amyloid beta in younger, genetically high-risk individuals and reduced metabolism and removal among older individuals.

●Vascular disease is an independent determinant of vascular dementia but also of increased amyloid deposition and neurodegeneration.

●Dementia in older individuals is a syndrome that derives most commonly from an individually varying mixture of diseases that afflict the brain. The most prevalent of these is AD, but it commonly occurs along with other diseases of the brain, vascular brain injury, Lewy body disease, and hippocampal sclerosis (HS).

This topic will focus on the incidence, prevalence, and risk factors for AD specifically. Risk factors for cognitive decline and all-cause dementia more generally, derived primarily from studies that define dementia according to a Mini-Mental State Examination score cutoff or similarly nonspecific case definition, are discussed separately. Protective factors and prevention of dementia are also discussed separately.

●(See "Risk factors for cognitive decline and dementia".)

●(See "Prevention of dementia".)

PATHOLOGY — The neuropathologic assessment of AD includes both evaluation of neuropathologic changes and correlation with clinical, neuropsychologic, neuroimaging, and other laboratory data [3,4]. Essential neuropathologic changes of AD include the following:

●Extracellular deposits of amyloid beta peptides (image 1)

●Neuritic plaques, associated with neuronal injury and characterized by amyloid formed from amyloid beta with dystrophic neurites that frequently have phospho-tau immunoreactivity [5-8]

●Neurofibrillary degeneration, best exemplified by neurofibrillary tangles (image 1)

AD neuropathologic change is ranked along three parameters: amyloid beta plaque distribution score [9], neurofibrillary tangle distribution stage [10,11], and neuritic plaque density score [12]. An overall assessment of low, intermediate, or high levels of AD neuropathologic change is obtained based on these three parameters.

In addition to these essential features, several other pathologic changes are commonly observed in association with AD.

●Cerebral amyloid angiopathy is often found in cases with parenchymal amyloid beta deposits [13,14].

●Inclusions of abnormal alpha-synuclein accumulation, known as Lewy bodies, are common in the setting of intermediate-to-high levels of AD neuropathologic change [15,16], including some early-onset familial AD cases [17,18]. Lewy bodies in AD often include extensive deposition in the amygdala. (See "Epidemiology, pathology, and pathogenesis of dementia with Lewy bodies", section on 'Pathology'.)

●Vascular brain injury is encountered commonly in brains with AD. Pathologic changes of vascular brain injury are caused by oligemia, hypoxemia, or ischemia involving different caliber vessels in different regions of brain. (See "Etiology, clinical manifestations, and diagnosis of vascular dementia".)

●Hippocampal sclerosis (HS) in AD is defined by pyramidal cell loss and gliosis in the hippocampal formation that is out of proportion to AD neuropathologic change [19]. HS can be observed alone or in the context of AD, frontotemporal lobar degeneration, or vascular brain injury. A similar lesion also can be observed in chronic epilepsy and other processes that injure the hippocampus. (See "Focal epilepsy: Causes and clinical features", section on 'Hippocampal sclerosis'.)

●Immunoreactive inclusions of TAR DNA binding protein 43 (TDP-43), as seen in most cases of amyotrophic lateral sclerosis and in a subset of frontotemporal lobar degeneration cases, are also commonly observed in cases with AD neuropathologic change [20]. With the exception of individuals with specific mutations that cause frontotemporal dementia, it is not clear whether TDP-43-immunoreactive inclusions in these other neurodegenerative diseases are a primary, secondary, or coincidental event [21]. Some evidence suggests that hippocampal TDP-43 deposition is associated with faster clinical progression and hippocampal atrophy among patients with AD [20,22,23]. (See "Epidemiology and pathogenesis of amyotrophic lateral sclerosis", section on 'Intracellular inclusions' and "Frontotemporal dementia: Epidemiology, pathology, and pathogenesis", section on 'FTLD-TDP'.)

PATHOGENESIS — While the pathogenesis of AD remains unclear, all forms of AD appear to share overproduction and/or decreased clearance of amyloid beta peptides. Amyloid beta peptides are produced by the endoproteolytic cleavage of mature protein translated from the amyloid beta precursor protein (APP) gene and cleaved by beta-secretase and gamma-secretase. Presenilin forms part of the gamma-secretase complex, and mutations in presenilin 1 (PSEN1) or presenilin 2 (PSEN2) appear to favor production of amyloid beta overall, or more neurotoxic forms of amyloid beta. The ultimate neurotoxin in AD is debated, but experimental evidence highlights small aggregates of amyloid beta peptides called oligomers, as opposed to larger aggregates called fibrils [24]. (See "Genetics of Alzheimer disease", section on 'Presenilin 1' and "Genetics of Alzheimer disease", section on 'Presenilin 2'.)

The pathogenesis of AD also involves a second protein, tau. Tau is a microtubule-associated protein that aids in microtubule assembly and stabilization. In AD, tau becomes hyperphosphorylated and aggregates to form paired helical filament (PHF) tau, a major component of neurofibrillary tangles within the neuronal cytoplasm. The accumulation of this altered protein is toxic to neurons in experimental models. In addition, transmission of pathologic forms of tau between neurons has been proposed to account for the spread of AD in the brain, which follows a distinct progression across brain regions as AD advances [25-27].

There are several other important and potentially overlapping pathways likely involved in AD pathogenesis, many of which have been implicated by the various risk genes that have been identified. As an example, the human apolipoprotein E (APOE) gene is a pleiotropic lipoprotein involved in multiple cellular processes including cholesterol transport, development, synaptic plasticity, and immune regulation, among others. There are three alleles of APOE, called epsilon 2 (e2), e3, and e4, and their encoded isoforms also vary in several activities. At least one mechanism by which inheritance of APOE e4 may increase AD risk is impairment of amyloid beta clearance from cerebrum [28]. (See "Genetics of Alzheimer disease", section on 'Apolipoprotein E'.)

CASE DEFINITION — A definitive diagnosis of AD requires histopathologic examination, but most studies of AD rely on clinical criteria to define cases. The clinical criteria for a diagnosis of probable AD have changed over time, and much of the older literature has either not distinguished between AD and other forms of dementia or utilized relatively small case-control-type studies in selected populations. Current clinical criteria for the diagnosis of AD are discussed separately. (See "Clinical features and diagnosis of Alzheimer disease", section on 'Diagnosis'.)

Studies will continue to evolve with the adoption of more sensitive and specific clinical criteria for AD as well as the incorporation of techniques for in vivo measurement of neuropathology such as amyloid positron emission tomography (PET), which allows for antemortem evaluation of cerebral amyloid accumulation. In addition, new PET tracers are under development to assess hyperphosphorylated tau (p-tau) to further characterize pathologic peptides in the living brain. The cost of these new technologies poses challenges for the development of large-scale epidemiologic studies, however, and may limit sample sizes [2,29,30]. (See "Clinical features and diagnosis of Alzheimer disease", section on 'Neuroimaging'.)

An additional consideration is that the pathologic changes of AD are often not found in isolation; many patients who die with dementia have evidence of mixed pathology. Based on information from several large, population-based clinicopathologic studies in older adults, the following observations can be made about the relationship between brain pathology and antemortem risk factors for AD [31-40]:

●There is a very high correlation between a clinical antemortem diagnosis of AD and postmortem measures of AD.

●Approximately one-third of individuals over the age of 85 years who die without antemortem diagnoses of dementia or AD have the pathologic changes of AD.

●Cerebrovascular disease is very common among persons who die with dementia, and the presence of AD pathologic changes and cerebrovascular disease together is common among individuals who die with dementia.

●Pathologic changes of multiple diseases, known as mixed dementia, are the most common finding in older individuals with clinical diagnosis of probable AD. (See 'Cerebrovascular disease' below.)

INCIDENCE AND PREVALENCE — The overall burden of AD is substantial worldwide [41-46]. Globally, an estimated 47 million people are affected by dementia [47].

Older adults — AD is increasingly prevalent with advancing age. In the United States in 2011, there were an estimated 4.5 million individuals over the age of 65 years living with clinical AD; this included 0.7 million people ages 65 to 74 years, 2.3 million ages 75 to 84, and 1.8 million 85 years and older [46]. This figure is projected to rise to 13.8 million in the United States and >130 million worldwide by 2050. The age-standardized prevalence of dementia ranges from 5 to 7 percent in most developed countries [42,45,47-50].

Estimates of the incidence of dementia vary across studies and depend heavily on age. In general, the incidence of dementia doubles every 10 years after age 60 years [44]. In the longitudinal Cardiovascular Health Study-Cognition Study (CHS-CS), only 19 of 160 participants alive at age 93 years remained free of mild cognitive impairment or dementia [51]. There is very little sex difference in incidence and prevalence of dementia or AD, although by absolute numbers there are more females than males with the disease, particularly over the age of 85 years, due to differences in life expectancy.

In 24 longitudinal studies, the age-specific incidence of dementia ranged from as low as 5 of 1000 for individuals ages 65 to 70 years and as high as 60 to 80 of 1000 for those age 85 years and older [45]. Examples of individual studies include the following:

●The CHS-CS followed over 3000 adults who were free of dementia at baseline over a five-year period and captured cases of AD or a mixture of AD and vascular dementia. Incidence estimates varied from 32 of 1000 person-years ages 75 to 79 at entry to 96 of 1000 person-years age ≥85 at entry, with little sex difference [52].

●In a population-based study in Rotterdam, the short-term incidence of dementia in the year 2000 was 1 of 1000 for ages 60 to 69 years, 6.4 of 1000 for ages 70 to 79 years, and 26 of 1000 for ages 80 to 89 years [53].

Dementia is a common cause of death in older adults. In a Medicare survey of 22,896 adults age 65 years and older, 15 diseases accounted for 70 percent of all deaths [54]. Dementia was second to heart failure as a leading cause of mortality, accounting for 19 percent of the deaths. Traditional mortality statistics are not a good measure of deaths attributed to dementia, however; these figures underestimate the contribution of dementia to mortality [54,55]. While individuals do not die of AD per se, advanced disease increases vulnerability for other disorders, commonly infections, which ultimately lead to death.

Younger adults — AD is also prevalent in younger adults with dementia, although there are fewer population-based studies in individuals under the age of 65 years [56,57].

A study from the United Kingdom estimated that the incidence of dementia in individuals ages 30 to 65 years was approximately 54 per 100,000 person-years [58]. The most common causes of dementia in these patients were AD (34 percent), vascular dementia (18 percent), frontotemporal dementia (12 percent), dementia with Lewy bodies (7 percent), alcohol-related dementia (10 percent), and others (19 percent). A separate population-based study in England reported an incidence of AD of 4.2 cases per 100,000 person-years among individuals ages 45 to 64 years [59].

Temporal trends — Multiple studies have reported that the prevalence of all-cause dementia may be declining over time in high-income countries [41,60-66]. As an example, a population-based study in Rotterdam reported a decreased incidence of dementia in all age groups between 1990 and 2000 as well as an increase in brain volume as a percent of intracranial volume [53]. In another study using data from the Framingham Heart Study in the United States, the cumulative five-year hazard rate of dementia fell from 3.6 per 100 persons during the late 1970s to 2.0 per 100 persons in the late 2000s [67]. The magnitude of the decline was greater for vascular dementia than for AD dementia, and rates of hypertension, atrial fibrillation, and stroke also fell over the same time period. Possible reasons for such a trend include improved educational levels of the more recent birth cohorts who are reaching the age of dementia risk or improved prevention and treatment of risk factors for vascular disease. By contrast, age-adjusted mortality rates from AD specifically have risen in the United States over the last decade [68].

It remains possible that the apparent decline in overall dementia is an artifact introduced by the definition of dementia as a specific level of cognitive impairment, especially memory, and related disability, rather than a relative change. Better-educated, cognitively higher-functioning individuals can tolerate a greater extent of neuropathologic change prior to reaching the threshold of dementia (this concept is termed "cognitive reserve") [69]. If dementia is defined as a change in cognition to below an absolute level and education improves over time, the prevalence of dementia may appear to decline when the only thing that has changed is the initial (higher) level of education. In support of this, a population-based study in the northern United States found no change in the incidence of dementia over time in a relatively homogenous higher-educated population [41].

PRESYMPTOMATIC PERIOD — There is a long presymptomatic period between the onset of pathophysiologic processes in the brain and the development of clinical symptoms of AD, suggesting that long-term epidemiologic studies beginning at an early age are needed to properly study the gene-lifestyle environmental determinants of amyloid vascular disease and neurodegeneration. In the Rotterdam Study, for example, the average age at diagnosis of dementia was 82 years, and memory complaints in these patients began 16 years before diagnosis [70].

The long presymptomatic period has also been demonstrated in families with dominantly inherited AD due to genetic mutations in amyloid beta precursor protein (APP), presenilin 1 (PSEN1), or presenilin 2 (PSEN2) [71,72]. Such mutations carry a nearly 100 percent risk of early-onset AD, which is typically diagnosed between 35 and 50 years of age, depending on the family. In one study, 88 carriers were evaluated at a mean age of 39 years [71]. Parental age at onset of AD in the index cases was used to estimate the relationship between disease biomarkers and the onset of clinical AD. A decline in amyloid beta 1 through 42 in cerebrospinal fluid (CSF) appeared to precede disease onset by 25 years; amyloid in the brain measured by Pittsburgh compound-B (PiB) retention on positron emission tomography (PET) scan, phospho-tau in CSF, and magnetic resonance imaging (MRI) measures of brain atrophy appear to be present approximately 15 years prior to onset of disease. Measurable cognitive impairment appears to occur approximately five years prior to clinical diagnosis.

These biomarkers appear to identify a presymptomatic period for AD in the general population as well. Neuroimaging findings including amyloid positivity on PET, as well as white matter lesions, neurodegeneration (cortical thickness), and small hippocampal volumes measured on MRI, have been found to be associated with cognitive decline and incident dementia, particularly when these findings occur in combination [73-76].

GENETIC RISK FACTORS — Aside from age, the most clearly established risk factors for AD are a family history of dementia, rare dominantly inherited mutations in genes that impact amyloid in the brain, and the apolipoprotein E (APOE) epsilon 4 (e4) allele.

Early-onset Alzheimer disease — The genetic basis for AD is understood most clearly in the early-onset form, which accounts for less than 1 percent of cases. Early-onset AD follows an autosomal dominant inheritance pattern related to mutations in genes that alter amyloid beta protein production, aggregation, or clearance, including amyloid beta precursor protein (APP), presenilin 1 (PSEN1), and presenilin 2 (PSEN2). Such mutations are highly penetrant, meaning that carriers have a nearly 100 percent chance of developing the disease in their lifetime.

Other early-onset cases without mutations can present clinically prior to 65 years of age. While these individuals have higher than expected inheritance of APOE e4, not all do, and this early-onset group is not fully understood [77]. (See "Genetics of Alzheimer disease", section on 'Early-onset Alzheimer disease'.)

Late-onset Alzheimer disease — The genetic basis of late-onset AD is more complex, with susceptibility likely conferred by a variety of more common but less penetrant genetic factors interacting with environmental and epigenetic influences. The most firmly established genetic risk factor for late-onset AD is APOE; carriers of one e4 allele are at two- to threefold increased odds of developing AD compared with noncarriers, and those with two e4 alleles are at approximately 8- to 12-fold increased odds. The strength of the association may be modified by several factors, including sex, race, and vascular risk factors. (See "Genetics of Alzheimer disease", section on 'Apolipoprotein E'.)

Many more candidate susceptibility genes have been identified through genome-wide association studies and other techniques. The average allelic summary odds ratios for these candidates are modest, ranging from approximately 1.1 to 1.5 in most cases. (See "Genetics of Alzheimer disease", section on 'Other candidate genes'.)

Family history — Family history of dementia is a risk factor for the development of AD; patients with a first-degree relative with dementia have a 10 to 30 percent increased risk of developing the disorder [78]. Individuals in families with two or more affected siblings with late-onset AD have a threefold increased risk of AD compared with the general population [79].

One study found that the increased risk in first-degree relatives was lower if the patient developed AD later in life (age 85 or older); their risk was similar to that of a control group [80]. Additionally, the excess risk of AD in relatives of patients with early-onset, nonmutation-associated disease was highest when the relatives were younger and diminished as the relatives aged: The hazard rate ratio of developing AD in relatives of a patient with early-onset disease was 19.7 when the relative was ages 50 to 54 but decreased to 1.2 by ages 90 to 94.

First-degree relatives of African Americans with AD appear to have a higher cumulative risk of dementia than relatives of White Americans. This was illustrated in an epidemiologic study of 17,639 first-degree biologic relatives of 2339 White probands and 2281 first-degree biologic relatives of 255 African American probands with AD [81]. The cumulative risk of the disorder in the African American and White relatives was 43.7 and 26.9 percent, respectively (relative risk [RR] 1.6, 95% CI 1.4-1.9).

ACQUIRED RISK FACTORS — A variety of polygenic or acquired factors influence risk for dementia and perhaps specifically the risk for AD, including hypertension, dyslipidemia, cerebrovascular disease, altered glucose metabolism, and brain trauma. Many of these risk factors appear to be most relevant when they are present in midlife [82,83]. In one study, individuals with two or more vascular risk factors in midlife had nearly threefold higher odds of brain amyloid deposition in late life as measured by amyloid positron emission tomography (PET) [82].

As such, aggressive management of vascular risk factors during midlife represents a key strategy for reducing risk, progression, and severity of AD and other forms of dementia [84,85]. It has been estimated that up to one-third of AD cases worldwide might be attributable to modifiable risk factors such as diabetes, midlife hypertension, and physical inactivity [86].

The discussion of risk factors for dementia (not specific to AD) is provided separately. (See "Risk factors for cognitive decline and dementia".)

Hypertension — Midlife hypertension has been consistently associated with risk of dementia and AD in cross-sectional and longitudinal cohort studies [83,87-91]. The risk is probably mediated through cerebrovascular disease and the long-term effects of elevated blood pressure, smoking, and diabetes [83,92-94]. The association may be stronger in females than males, although this finding has not been consistent [91].

Arteriolosclerosis and blood pressure variability may play an important role in the relationship between blood pressure and the risk of AD [95-97]. Cerebral arteriolosclerosis increases with age, blood pressure levels, and possibly diabetes and smoking [98]. In a cohort of dementia-free older adults ages 83 to 96 years, for example, arterial stiffness was directly related to amyloid beta plaque deposition, independent of blood pressure and apolipoprotein E (APOE) genotype [99,100]. Other studies have shown similar associations [101].

While observational studies suggest that treatment of hypertension reduces the risk of dementia, clinical trials of antihypertensive drug therapies have not consistently demonstrated a reduction in the risk of dementia. However, most trials have measured short-term outcomes, and it remains possible that long-term treatment of hypertension, particularly beginning in midlife, alters AD risk later in life. Studies evaluating antihypertensive treatment in the prevention of dementia are discussed separately. (See "Prevention of dementia", section on 'Antihypertensive therapy'.)

Dyslipidemia — While some epidemiologic studies have reported an association between total or low-density lipoprotein cholesterol (LDL-C) levels, especially in midlife, and risk of AD, the interpretation of these studies is complex. Peripheral blood LDL-C does not cross the blood-brain barrier unless the barrier is damaged by factors such as vascular disease. Most of the cholesterol in the brain is synthesized by astrocytes and neurons and provided to brain cells by high-density lipoprotein (HDL) complexes, and there is little or no LDL-C in the brain [102-106].

Nonetheless, observational data generally support an association between elevated LDL-C or total cholesterol at midlife and risk of AD. In one longitudinal cohort study, high total cholesterol at midlife (mean age 50 years) conferred threefold higher odds of AD, independent of APOE genotype, education level, smoking status, and alcohol consumption [107]. Other studies have reported similar results [87,108]. When late-life cholesterol levels are examined, results have been mixed, with some studies finding a positive association between cholesterol and AD or dementia risk [109-111], and others finding either no correlation [112] or an inverse relationship [113,114]. Elevated total cholesterol and LDL-C have also been associated with higher rates of cognitive decline in patients with AD [109].

It is speculated that brain cholesterol may increase the risk of AD by enhancing the formation and/or deposition of amyloid beta, or that it may affect nonamyloid factors such as cerebrovascular risk, local inflammation, or tau metabolism [115].

Based on these observed associations, there has been considerable interest in cholesterol-lowering therapy as a potential preventive therapy for AD. Observational data have been mixed, however, and randomized trials in patients with cardiovascular disease or at higher risk of cardiovascular disease have not shown a protective effect of statins for cognition (although these studies were not designed to identify incident dementia as a primary endpoint). Oral statins also do not affect the rate of decline in mild to moderate AD.

Cerebrovascular disease — Cerebrovascular disease and AD frequently coexist, and the co-occurrence of AD pathologic changes and vascular brain injury is a common form of mixed dementia. Cerebral small vessel disease is common in aging, does not occur in isolation, and is often comorbid with AD. Risk factors include hypertension, diabetes, smoking, obesity, low physical activity, and hyperlipidemia [116].

Decreased blood flow before beta amyloid deposition has been observed in both mouse models of AD and human studies and has been proposed to contribute directly to amyloid accumulation, probably by impairment of clearance of amyloid. Vascular factors may also contribute to the breakdown of the blood-brain barrier [117]. In a population-based cohort study, decreased cerebral blood flow (CBF) was associated with an increased risk of AD, particularly in combination with white matter abnormalities [118].

In autopsy series, vascular brain injury is found in 34 to 50 percent of those with AD pathologic changes [119-127]. Conversely, approximately one-third of patients diagnosed with vascular dementia will have AD pathologic changes at autopsy [122]. Using the consensus definitions of AD pathologic changes and vascular brain injury, patients with a diagnosis of dementia are more likely to have combined pathologic features rather than either AD or vascular brain injury in isolation [128]. One large autopsy study found that coexistence of lacunar and large infarcts was associated with higher likelihood of a clinical diagnosis of AD dementia only when the AD pathology burden was low [129].

The APOE epsilon 4 (e4) allele is associated with an increased risk of cardiovascular disease and with AD [120,130]. In an autopsy study of 99 individuals, the severity of coronary artery disease was significantly associated with the density of AD neuropathologic features, primarily in APOE e4 carriers [131]. In another autopsy series, APOE e4 was associated with subcortical microvascular changes but not gross cerebrovascular pathology in patients with AD [132].

Cerebrovascular disease has been associated with worse cognitive performance in patients with AD, and clinicopathologic studies suggest that cerebrovascular disease lowers the threshold for clinical dementia in patients with a neuropathologic diagnosis of AD [74,119,133-141]. Numerous studies have found an increased risk of AD in association with various neuroimaging or pathologic markers of cerebrovascular disease, including atherosclerosis in the circle of Willis [142-145], periventricular white matter lesions [138,146-148], cerebral microbleeds [149], and cortical infarcts [138,139,150]. From these and other studies, some conclude that a vascular mechanism may be a primary etiologic factor in AD [151].

Peripheral and cardiovascular atherosclerosis — Markers of peripheral atherosclerotic disease burden, including measures of carotid intima media thickness and coronary artery calcification, have also been associated with an increased risk of dementia and AD. It is possible that such associations are modulated by arteriosclerosis and vascular brain disease, especially white matter lesions and microinfarcts in the brain [152,153].

Type 2 diabetes and obesity — Obesity and type 2 diabetes have been associated with an approximately 1.5-fold increased risk of AD [82,154,155]. As with other vascular risk factors, the data are strongest for midlife obesity and more mixed for late-life body mass index. (See "Risk factors for cognitive decline and dementia", section on 'Obesity and body mass index'.)

The direct effects of hyperinsulinemia and insulin resistance in the brain, as well as a possible relationship between insulin and amyloid beta metabolism, are areas of active investigation [155-164]. Some studies have suggested a role of insulin-degrading enzyme (IDE), which metabolizes both amyloid beta and insulin, in the accumulation of oligomeric beta amyloid [165]. Other studies are investigating a pathogenic role for the accumulation of advanced glycation end products within brain tissue [166].

Small pilot studies of intranasal insulin and dietary manipulation in patients with mild cognitive impairment or AD showing improved cognitive function or biochemical profiles provide further support for the potential link between brain insulin and AD pathogenesis [167-169]. Larger follow-up studies are underway. (See "Mild cognitive impairment: Prognosis and treatment", section on 'Others'.)

Additional data supporting these associations in dementia more broadly are discussed separately. (See "Risk factors for cognitive decline and dementia", section on 'Cardiometabolic risk factors' and "Estrogen and cognitive function".)

Lifestyle and activity — Physically active individuals have lower incidence and prevalence of cognitive decline and dementia, including AD [170-173]. A meta-analysis of 16 prospective studies suggested a 28 percent reduction in overall dementia and a 45 percent reduction in AD among those who were physically active compared with those who were less active, even after adjusting for confounding variables [174].

The Alzheimer's Association and the World Dementia Council reviewed the evidence for modifiable risk factors for cognitive decline and dementia and concluded that sufficient evidence supports the link between regular physical activity and management of cardiovascular risk factors (diabetes, obesity, smoking, and hypertension) and reduced risk of cognitive decline and possibly dementia [175]. They also found strong evidence to conclude that a healthy diet and lifelong learning (cognitive training) may also reduce the risk of cognitive decline.

Small clinical trials have also shown that exercise improves cognitive function and may also increase hippocampal and total brain volume [171]. Most of these trials were of short duration, however, and no trials have evaluated the effects of sustained exercise on the long-term risk of AD [176].

Studies supporting an association between other lifestyle factors and dementia, potential mechanisms by which lifestyle factors might influence risk, and prevention strategies based on these data are discussed separately. (See "Risk factors for cognitive decline and dementia", section on 'Lifestyle and activity' and "Prevention of dementia", section on 'Lifestyle and activity'.)

Brain trauma — History of severe brain trauma with loss of consciousness of 30 minutes or more has been inconsistently associated with an increased risk of AD [177-180]. A causal relationship has been suggested by the finding of increased amyloid in the brain shortly after severe brain trauma [179]. However, at least two well-designed cohort studies have not confirmed an association between traumatic brain injury (TBI) and risk of AD or AD biomarkers [181]. In one autopsy study examining 1589 brains, no association was found between self-reported TBI and neuropathologic changes consistent with AD nor with incident AD [180]. Similarly, A Veterans Affairs study did not find a relationship between TBI and PET amyloid deposition [182].

Repeated milder forms of head trauma have been associated with neuropathologic changes of a tauopathy that is distinct from AD. (See "Sequelae of mild traumatic brain injury", section on 'Chronic traumatic encephalopathy'.)

The association of TBI with cognitive decline and dementia more generally is discussed separately. (See "Risk factors for cognitive decline and dementia", section on 'Head trauma'.)

Medications — Multiple studies have found an association between use of certain medication classes (eg, benzodiazepines, anticholinergics, antihistamines, opioids) and cognitive impairment in older adults, but the effects have been presumed to be transient and reversible [183]. Long-term exposure has been linked to an elevated risk of AD or all-cause dementia in several large studies, however, raising the possibility that cognitive effects may not be reversible in some patients or that these exposures somehow enhance the penetrance of presymptomatic AD.

●Benzodiazepines – Available data on the association between benzodiazepine use and dementia risk are conflicting. In a nested case-control study of nearly 2000 older adults enrolled in a public drug plan in Canada, use of benzodiazepines for >180 days was associated with a 1.5-fold increase in risk of AD after adjusting for multiple potential confounders, including anxiety, depression, and insomnia [184]. A dose-response effect was observed, with longer exposure and longer half-life drugs associated with increased risk. It remains possible that benzodiazepines were used to treat prodromal symptoms of dementia, however, even though the study limited exposure to prescriptions begun at least five years before the diagnosis of AD [185,186].

Other large observational studies have failed to find a significant association between long-term benzodiazepine use and incident dementia; these are discussed separately. (See "Risk factors for cognitive decline and dementia", section on 'Medications'.)

●Anticholinergics – The case for anticholinergics increasing the risk of irreversible effects is probably stronger [187,188] and is plausible given the prominence of cholinergic deficits in AD. Studies associating these medications with incident dementia are discussed separately. (See "Risk factors for cognitive decline and dementia", section on 'Medications'.)

●Proton pump inhibitors – There are conflicting data on the association between proton pump inhibitors (PPIs) and risk of AD. In one longitudinal study, PPI use had similar associations with all-cause dementia (hazard ratio [HR] 1.3) and with AD dementia (HR 1.4) [189].

By contrast, two other observational studies, including a large case-control study with over 70,000 cases of AD and over 280,000 age-, sex-, and region-matched controls, found no association between PPI use and AD, nor evidence of a dose-response relationship in either dose or duration of use [190,191].

Limited preclinical data suggest a possible interaction between PPIs and both amyloid and tau [192-194]. Alternatively, malabsorption of vitamin B12 or other nutrients due to long-term PPI use could play a role [195]. On the other hand, the association may reflect residual confounding by factors related to both use of PPIs and the development of dementia, and more studies are needed [196].

ENVIRONMENTAL RISK FACTORS — As large-scale genetic studies have yielded limited predictive risk alleles beyond apolipoprotein E (APOE) epsilon 4 (e4), there is a renewed interest in environmental and toxic exposures as potential risk factors for AD [197]. Areas of investigation include the following:

●Secondhand smoke – In a cross-sectional study of 2692 never-smokers aged 60 years and older in China, secondhand smoke exposure was associated with an increased risk of AD (adjusted relative risk [RR] 2.28, 95% CI 1.82-2.84) [198]. Risk was higher for in-home exposure compared with at-work exposure.

●Air pollution – Animal studies [199,200] and limited human epidemiologic data provide support for air pollution as a potential risk factor for AD. An increase in diffuse amyloid plaques and inflammation in the olfactory bulb, hippocampus, and frontal lobes have been described in brain autopsy specimens from adults living in Mexico City and Monterrey, areas known for high levels of air pollution, compared with adults living in smaller cities in Mexico with lower levels of air pollution [201]. Similar findings have also been reported in children and young adults [202].

●Pesticides – Several studies have implicated occupational or environmental exposure to pesticides as a risk factor for AD [203,204]. As an example, a case-control study examined serum dichlorodiphenyldichloroethylene (DDE) levels in 86 pathologically confirmed AD cases and 79 controls and found that levels of DDE were 3.8-fold higher in those with AD [204].

SUMMARY

●Hallmark neuropathologic features – Neuropathologic changes of Alzheimer disease (AD) include neuritic plaques, extracellular deposits of amyloid beta, and neurofibrillary degeneration. (See 'Pathology' above.)

●Pathogenesis – While the pathogenesis of AD remains unclear, all forms of AD appear to share overproduction and/or decreased clearance of a family of proteins known as amyloid beta peptides. The pathogenesis of AD also involves tau, a microtubule-associated protein that aids in microtubule assembly. (See 'Pathogenesis' above.)

●Case definition – A definitive diagnosis of AD requires histopathologic examination. Clinical criteria for the diagnosis of AD have evolved over time, and the ability to accurately diagnose AD has improved with the development of techniques for in vivo measurement of pathophysiologic features of AD. The diagnosis of AD requires a decline in both cognition, especially memory, and function, as well as specific neuropathology. (See 'Case definition' above.)

●Presymptomatic disease – There is a long presymptomatic period between the onset of biochemical changes in the brain and the development of clinical symptoms of AD, suggesting that long-term epidemiologic studies beginning at an early age are needed to properly study the gene-lifestyle environmental determinants of amyloid vascular disease and neurodegeneration. (See 'Presymptomatic period' above.)

●Epidemiology and risk factors – AD is increasingly prevalent with advancing age, and the overall burden of AD is substantial worldwide. It has been estimated that the global prevalence of dementia will rise to >100 million by 2050.

•Advanced age – The age-standardized prevalence of dementia ranges from 5 to 7 percent in most regions of the world. (See 'Incidence and prevalence' above.)

•Genetic factors – Aside from age, the most clearly established risk factors for AD are a family history of dementia, rare dominantly inherited mutations in genes that impact amyloid in the brain, and the apolipoprotein E (APOE) epsilon 4 (e4) allele. (See 'Genetic risk factors' above.)

•Vascular risk factors – Risk factors for vascular disease, including hypertension, obesity, and diabetes, increase the risk of dementia and may impact AD, particularly when they are present in midlife. The pathogenic mechanisms linking various cardiometabolic factors and AD are not well understood, and both vascular and nonvascular mechanisms may be involved. (See 'Hypertension' above and 'Type 2 diabetes and obesity' above.)

Brain cholesterol metabolism may be an important determinant of AD. The relationship between diet, genetics, blood lipoproteins levels, and AD is complex and inconsistent. (See 'Dyslipidemia' above.)

Cerebrovascular disease and AD frequently coexist. Hypertension is the primary risk factor for vascular brain injury. Cerebrovascular disease is associated with worse cognitive performance in patients with AD, and clinicopathologic studies suggest that cerebrovascular disease lowers the threshold for clinical dementia in patients with a neuropathologic diagnosis of AD. (See 'Cerebrovascular disease' above.)

•Lifestyle risk factors – Accumulating data suggest that social, cognitive, and physical activities are inversely associated with the risk of AD and other forms of dementia, and there is considerable interest in their potential as preventive strategies. (See "Risk factors for cognitive decline and dementia", section on 'Lifestyle and activity' and "Prevention of dementia", section on 'Lifestyle and activity'.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Lewis H Kuller, MD, DrPH (deceased), who contributed to earlier versions of this topic review.