Alzheimer Disease (AD): Impact on Global Health

People all around the world are living longer. Most can expect to live beyond their 60th birthday. According to the World Health Organization,^[1] 1 in 6 people world-wide will be 60 years of age or older by 2030. Unfortunately, longer life span doesn’t always translate into longer health. Recent statistics show that the duration of life in good health has remained constant, which implies that the additional years are mired in poor health or reduced capacity for many.^[1] Common health conditions associated with advanced age include hearing loss, cataracts, osteoarthritis, diabetes, depression and dementia (Figure 1).^[1] Dementia, a loss of cognitive function, poses a significant economic burden to healthcare systems and society as a whole because of reduced productivity for both the patients and their caregivers. Because the disease progression can take many years and no cure is available, dementia is becoming a global health crises with 50 million people currently affected.^[2] A common cause of dementia, Alzheimer disease (AD) is believed to account for 60% to 80% of cases.^[3] The yearly cost of AD and other dementias in the United States alone is predicted to increase to more than $1 trillion by 2050.^[3] The emergence of COVID-19 resulted in more than 1.3 million hospitalizations among US adults age 65 and older between January 2020 and July 2021.^[3] Because critical illness and hospitalization is believed to increase the risk of long-term cognitive impairment in older people, the pandemic may increase the number of AD cases and their resulting costs beyond earlier estimates.^[3]

Figure 1. Older Age Conditions

The prevalence of these conditions will increase as the aging
population increases.^[1]
Abbreviation: COPD, chronic obstructive pulmonary disease.

Neurological Basis for AD

A neurodegenerative disease, AD is associated with cognitive, functional, and behavioral impairments.^[4] It is characterized by the progressive accumulation of extracellular amyloid beta (Aβ) plaques and intracellular neurofibrillary tangles (NFTs).^[3] Aβ plaques are generated via abnormal cleavage of the amyloid precursor protein (APP) into a small neurotoxic 42 amino acid peptide (Aβ42).^[5] As a result of either reduced Aβ42 clearance or excess Aβ42 production, plaque deposition in the brain occurs.^[6] NFTs are created when the hyperphosphorylation of the tau proteins leads to the disassembly of microtubules. These disassembled microtubules form NFTs.^[6] These abnormal aggregates are accompanied by neuronal and synaptic degeneration.^[6] A silent progression, both of these pathologies may occur 10 to 20 years prior to the onset of cognitive impairment symptoms.^[7,8]

AD Genetic Factors

Genetic factors may predispose an individual to either early-onset or late-onset AD. Early-onset disease, characterized by symptom onset before the age of 65, is seen in less than 10% of patients diagnosed with AD.^[5,9] These individuals often carry mutations on 1 of 3 genes: AAP, Presenilin1 (PSEN1), and Presenilin 2 (PSEN2).^[5]AAP mutations lead to increased Aβ production and aggregation, while PSEN1 and PSEN2 also lead to the aggregation of Aβ via interference with gamma-secretase processing. While these three gene mutations occur in only 5% to 10% of all AD diagnoses, they almost exclusively account for early-onset AD.^[5]

Another gene allele is found in the majority of individuals diagnosed with late-onset AD. APOE is the template for production of a protein that combines with fats in the human body to form lipoproteins.^[10] The ApoE protein is produced by astrocytes and microglia. It is the major cholesterol carrier of the brain, supporting lipid and Aβ transportation and injury repair.^[11,12] The APOE gene has 3 polymorphic alleles: ε2, ε3, and ε4. Individuals with the ε4 allele have increased AD risk; approximately 65% of AD patients are APOE4 carriers.^[13] Patients who are homozygous for APOE4 have the greatest risk and the lowest average age of AD onset compared to those who are heterozygous or carry ε2 or ε3 alleles.^[14] They also carry a several-fold higher burden of Aβ oligomers compared to noncarriers due to the role of ApoE in Aβ aggregation and clearance.^[12,13] Although not routinely used by clinicians, genotyping can be used as a tool for AD screening and as part of a clinical diagnostic algorithm, particularly in early stages of suspected AD.

Progressive AD Stages

Subtle differences exist in the nomenclature at each stage of AD. These are important differences depending on their clinical and research classifications.^[4] The National Institute on Aging -- Alzheimer's Association (NIA-AA) lists the following five stages within the AD continuum of disease progression.^[4]

Stage 1:    Preclinical

Stage 2:    AD with mild cognitive impairment (MCI)

Stage 3:    AD with mild dementia

Stage 4:    AD with moderate dementia

Stage 5:    AD with severe dementia

The earliest stage is termed preclinical AD and generally has a long asymptomatic phase lasting 6-10 years depending on the age of onset. There is evidence of AD pathology in the form of Aβ or NFTs but no evidence of cognitive decline in up to 30 years ahead of symptom onset in some patients.^[15,16] As such, a preclinical AD individual’s daily life and functional skills may not be affected while their neuropathology has already begun to change. Counterintuitively, not all persons who have these silent AD pathologies will go on to the next 4 AD stages. A 2019 meta-analysis looking at 6 longitudinal cohorts followed up for an average of 3.8 years reported that 20% of preclinical AD patients progressed to MCI due to AD.^[17]

Early clinical symptoms of patients who do progress from the preclinical stage include short-term memory loss and decline in cognitive functions such as word retrieval, forgetting recent conversations, repeating conversations or questions, difficulty in completing familiar tasks, and getting lost in familiar surroundings.^[18,19] However, with help from family or friends, these patients still remain relatively independent. For patients with MCI due to AD (Stage 2), several studies found that 33% to 70% progressed to AD with dementia (Stages 3-5) in 4 years or slightly less time.^[20-22]

Benefits of Early Diagnosis

Delivering solemn news is always challenging for healthcare providers who may hesitate to provide this information due to concern for the patient’s mental state. In a recent consortium study, learning about a positive amyloid positron emission tomography (PET) result was associated with psychological changes among affected patients.^[23] These individuals showed small increases in distress, anxiety, or depression scores compared with their amyloid PET negative peers. However, these negative reactions did not reach a threshold for clinical concern according to the authors of the study. They concluded that this disclosure to patients with subjective cognitive decline was not associated with clinically meaningful psychological risk.^[23]

Early detection and disclosure are vital because patients who progress to AD dementia phases (Stages 3-5) will require assistance with daily living activities due to ongoing cognitive defects. This is a severe burden to patients and their caregivers, requiring 24/7 care and a total loss of independence for all concerned. To prepare for a likely progression, early AD diagnosis in the preclinical stage can give patients an opportunity to plan ahead. They can create advanced care plans with their families, physicians, and other support team members. Patients can request early interventions, initiate lifestyle changes, and consider risk-reduction strategies moving forward. These actions can potentially improve the likelihood of increased medical, community, and emotional support for the patient and their caregivers.

Historical AD Diagnostic Methods

Postmortem autopsies have been used to confirm AD in deceased dementia patients, providing no preventive information during the patient’s life. It is presently the gold standard for diagnosis. But because AD is inheritable, autopsy can provide information for the genetic counseling of closely related family members.

AD diagnosis in the past has been one of exclusion and only possible in the later stages of the disease. Clinical criteria lead to an accurate diagnosis in only 70% to 80% of patients and may have a specificity for AD as low as 44%.^[24,25] Amyloid PET scans and magnetic resonance imaging (MRI), quantifying Aβ and estimating brain atrophy, respectively, have been employed for AD diagnosis in vivo. Amyloid PET scans provide direct visualization of Aβ in the brain. Abnormalities can be detected early in disease progression.^[16] Tau PET scans with second generation tracers may also have added value in early-stage tau visualization. But, due to the high cost of structural imaging, increased risk of radiation exposure, and lack of coverage by many insurance plans including Centers for Medicare & Medicaid Services (CMS), many individuals and their families reluctantly decide to "wait and see."^[26] Also, a positive amyloid PET scan does not stand alone as a definitive diagnosis of clinical AD and must be combined with other assessments.^[27] MRI excludes rather than supports a diagnosis of AD, because brain atrophy may also indicate other neurodegenerative diseases.^[28,29]

Genetic tests provide probabilities for AD disease development and progression. A patient’s APOE allele identification and number are semi-quantitative factors that may enhance the AD diagnostic tool kit. Data from these tests can be combined with structural imaging to improve the total AD risk assessment.

Medical narratives and family histories are also used in conjunction with structural tests to evaluate a patient’s risk for developing AD with dementia. Caregiver input and an individual’s lifestyle data can be combined with cognitive, behavioral, and functional tests to calculate risk. While these historical methods can provide risk estimates, they are not a definitive diagnosis. Fortunately, emerging clinical tools are now available to provide a more accessible, accurate, and quantitative diagnosis for AD.

Biomarker Guidelines for AD Diagnostics

Because past structural and behavioral assessments have proven costly or too vague, alternative in vitro biomarker tests that are rapid, less expensive, and more quantitative have been sought. A biomarker is any measure of biological activity (e.g., weight, complete blood count, blood pressure, electrocardiogram). Certain fluid biomarkers, proteins found in cerebral spinal fluid (CSF) or blood plasma (BP), are associated with AD pathology and are useful for early AD detection. The presence of Aβ and NFTs in human CSF or BP can identify preclinical individuals at risk for developing MCI due to AD, the second stage of progression.

The US Food and Drug Administration (FDA) has endorsed their use while the NIA-AA has created a research framework recognizing the use of biomarkers for not only diagnosing AD in vivo but monitoring disease progression as well.^[8,30] The International Working Group (IWG) recommends that AD diagnosis be restricted to individuals with positive biomarkers as well as clinical manifestations.^[31] Those with positive biomarkers but no clinical signs, the aforementioned Preclinical Stage 1 or cognitively unimpaired, should only be considered at risk for development of AD.^[31]

CSF Biomarkers

The IWG recommends CSF biomarkers along with clinical phenotype for the diagnosis of AD.^[31] Alternative or additional tests to detect amyloid PET are the analyses of CSF phosphorylated tau (p-tau) and isoforms of Aβ. They provide additional clinical information useful in the diagnosis of AD without requiring the cost and time associated with structural scans (Figure 2). Because CSF biomarkers exhibit strong correlation with amyloid PET, they are widely accepted in the AD community as supporting a diagnosis of early stage AD.^[32,33]

Figure 2. CSF Biomarkers and AD Diagnosis Functionality ^[34-46]

Abbreviations: Aβ, amyloid beta; AD, Alzheimer disease; CSF, cerebrospinal fluid; MCI, mild cognitive impairment; NFT, neurofibrillary tangle; PET, positron emission tomography.

The tau protein is associated with the development of NFTs in AD. In the CSF of individuals diagnosed with AD, total tau (t-tau) is often elevated and predicts neurodegeneration.^[34] However, t-tau is associated with other neurodegenerative and acute brain disorders such as stroke.^[35] While CSF t-tau may be a prognostic indicator of MCI and later to AD dementia, its usefulness in AD diagnosis is hampered by its lack of specificity for the disease.^[36]

As an alternate measure of the tau protein, CSF p-tau is another direct measurement strongly correlated to the presence of NFTs in the brain.^[34] p-tau181 is the species normally elevated specifically in early AD,^[36] although some studies have shown specificity of p-tau217 and other isoforms. p-tau is more specific for AD than t-tau, with a sensitivity for AD of over 90%.^[35] p-tau has also not been found to be elevated in other neurodegenerative diseases, indicating its potential as a diagnostic marker.

Aβ is the protein of AD-associated amyloid plaques. As such, several isoforms of Aβ have been evaluated for diagnostic sensitivity and specificity for AD. Two Aβ isoforms are commonly evaluated for AD in CSF. Aβ40 is the most abundant Aβ in CSF and is used to represent total Aβ levels in cortical tissue.^[37] However, Aβ40 concentration does not differ between AD, non-AD dementia, and healthy controls, thus eliminating its ability to be used as a stand-alone biomarker for AD. Another Aβ isoform, Aβ42, has been associated with AD in multiple studies.^[35,36,38] CSF levels of Aβ42 are decreased due to increased Aβ aggregation, leading to deposits in the brain.^[34] Low Aβ42 alone shows very high concordance with amyloid PET.^[35]

Rather than use the concentration of Aβ42 singly, research suggests a ratio between Aβ42 and Aβ40 to adjust for individual differences in a patient’s Aβ production.^[38] Because Aβ40 is abundant, it can be utilized to anchor the ratio so that results are indicative only of changes in Aβ42 in reference to total Aβ load.^[37] Decreased Aβ42 in absence of reduced Aβ40, or a positive (low) Aβ42/Aβ40 ratio, is indicative of AD, while reduction of both biomarkers may indicate other pathologies involving subcortical damage, such as Parkinson’s dementia, vascular dementia, and Lewy body dementia.^[39] More patients are classified correctly when using the Aβ42/Aβ40 ratio versus Aβ42 alone (94% vs 86.7% when comparing AD to controls, 90% vs 85% when comparing AD to non-AD dementias, and 90.8% vs 87% when comparing AD to non-AD dementias plus controls).^[38,40] Use of the Aβ42/Aβ40 ratio mitigates sample collection issues associated with individual biomarker tests, is less affected by racial differences, and is more consistent with amyloid PET results with a 97% positive and 84% negative concordance.^[41,42] To date, only 1 CSF Aβ42/Aβ40 ratio assay has been approved by the FDA for the diagnosis of AD.^[41]

An alternative ratio often used with Aβ42 is the p-tau/Aβ42 ratio. p-tau/Aβ42 predicts worsening cognitive impairment in individuals with AD^[43] and has a higher percent agreement with amyloid PET results than either marker alone.^[44] p-tau/Aβ42 may also predict a greater 2-year decline in patients with MCI.^[32] Providing a 90% concordance with amyloid PET, this ratio eliminates some of the patient variability seen with individual markers. There is 1 p-tau/Aβ42 assay that has obtained FDA approval.

Neurofilament light (NFL) is another CSF biomarker that is currently being evaluated for use in AD diagnosis. A biomarker of subcortical large-axonal degradation, CSF levels of NFL have been shown to be elevated in multiple neurodegenerative disorders including AD.^[45] NFL alone has sensitivity for AD of around 81%^[46] and is primarily increased in individuals with a rapid disease progression.^[26] Like many biomarkers evaluated for AD, NFL in combination with other markers may provide increased sensitivity and specificity. Utilizing NFL as part of a ratio with Aβ42 increases the area under the receiver operating characteristic curve from 0.84 (NFL) to 0.90 (NFL/Aβ42) when comparing AD to healthy individuals.^[46]

Patients who have cognitive or behavioral symptoms suggestive of AD can obtain a lumbar puncture. The lumbar puncture can be performed safely as an outpatient procedure. CSF biomarker testing is not recommended for patients who already have received an AD diagnosis or in lieu of genotyping for APOE alleles.^[47]

Blood Plasma Biomarkers

Because of ease of collection, BP biomarkers are being developed to detect early AD, identify those at risk for future AD pathology, and monitor disease progression.^[48-50] Assays measuring p-tau181 and p-tau217 use these isoforms to differentiate AD from other neurodegenerative diseases.^[48] Other assays evaluate isoforms of Aβ42 or Aβ42/Aβ40 ratio for plaque status in presumptive AD and have shown high correlation with amyloid PET.^[51] However, due to the limitations of plasma assays, all current BP assays are for clinical or research use only and are not covered by insurance. For plasma biomarkers to be useful to AD diagnosis, several limitations such as low biomarker concentrations and issues with plasma matrix would have to be overcome.

In their 2021 update, IWG does not recommend plasma biomarkers for the diagnosis of AD due to the need for standardization and validation.^[31] To date, BP assays for assistance in AD diagnosis are not FDA approved and are for clinical use only in the United States. Two assays are close to approval by the FDA after being granted Breakthrough Device Designation for the assessment of AD. One is a qualitative assay that utilizes p-tau181 and APOE4. A positive result requires confirmation with CSF markers or an amyloid PET scan.^[52] The second assay uses a plasma Aβ42/Aβ40 ratio and APOE proteotype to provide an AD likelihood score.^[53,54] This BP assay has both an 86% positive and negative predictive value, as well as a positive amyloid PET concordance of 92% and a negative concordance of 77%.^[53,54] While BP assays have a lower PET concordance than the FDA-approved assay for CSF Aβ42/Aβ40, they may have a role in future AD diagnostics.

CSF Biomarker Clinical Implications

The clinical diagnosis of AD in the MCI and dementia stages are evolving from a diagnosis of exclusion to a diagnosis of inclusion. Specifically, we need to be able to confirm AD pathology either by CSF or imaging and perhaps by plasma biomarkers in the near future. Lumbar puncture for CSF is more accessible, minimally invasive, used in a variety of neurological indications, and is part of routine neurological practice. The CSF assays that measure the Aβ42/Aβ40 ratio, t-tau, and p-tau have proven to be excellent diagnostic tools with high predictive value for the patient’s progression from MCI to AD dementia.^[47] Thus, a diagnosis of AD is evolving into a biomarker-driven diagnosis of inclusion (Figure 3).

Figure 3. Algorithm for AD Diagnosis^{[8,30-31,41,47]}

Abbreviations: AD, Alzheimer disease; CSF, cerebrospinal fluid; FDA, US Food & Drug Administration.

Biomarker Limitations

Biomarker assays hold great promise for early AD diagnosis. Presently, all BP and most CSF tests are laboratory developed and unvalidated. There are currently only 2 FDA approved biomarker assays for the diagnosis of AD, each evaluating a different biomarker ratio. Both the Aβ42/Aβ40 and p-tau181/Aβ42 assays utilize ratios of biomarkers from CSF samples. The Aβ42/Aβ40 assay has a 97% positive and 84% negative concordance with amyloid PET,^[41] while the p-tau181/Aβ42 assay has an overall 90% concordance rate.^[55] Neither assay should be used alone, but as part of a diagnostic pathway with consideration for other clinical evaluations.

A substantial limitation to the use of AD biomarkers is the racial disparity between White and non-White groups (Figure 4). Black American adults have a 64% higher rate of progression to AD and other dementias than White adults.^[56] Recent studies have suggested differences in AD-related neurobiology between different racial groups.^[56] Cutoffs for AD biomarkers have been established in largely White populations, bringing into question the basis for determining a positive or negative assay result.^[57] Several studies with BP and CSF have indicated that Aβ40, p-tau isoforms, including p-tau181 and p-tau181/Aβ42 ratio, and other presumptive biomarkers such as NFL, are consistently lower in non-White individuals.^[56,58] However, CSF Aβ42 and the Aβ42/Aβ40 ratio are not significantly different between racial groups,^[58] although BP Aβ42/Aβ40 ratios were higher in Black Americans.^[59] Additionally, studies have found racial differences in the statistical interaction of p-tau and APOE4 status.

Figure 4. Racial Differences in CSF and BP Biomarkers of AD ^[56-59]

Abbreviations: CSF, cerebrospinal fluid; MCI, mild cognitive impairment.

Healthcare providers should consider these racial disparities when diagnosing non-White patients. Whether these differences are genetic, environmental, or a product of other health-related disparities, further analysis and inclusion of racial minorities in clinical trials is imperative in establishing standard ranges and cutoffs of AD-related biomarkers.

Looking Ahead

Between 2015 and 2050, the proportion of the world’s population over age 60 will nearly double -- from 12% to 22%.^[1] Healthcare and social systems will have to adapt to make the most of this demographic shift. Older adults bring deep knowledge, hard-won experience, and the ability to synthesize information -- all for the benefit of society. "A mind is a terrible thing to waste," is the iconic slogan used by the United Negro College Fund since 1972.^[60] More than ever, this expression applies to our elders facing AD and other age-related dementias. To help prevent this societal "brain drain," scientists must be able to identify MCI with more confidence. They must also be able to evaluate racial differences in AD biomarkers. Future biomarker protocols may be developed to further evaluate at-risk patients and the efficacy of treatments.

AD is a disease requiring long-term care; a multistage process involving family, friends, social services, health care professionals, and specialized memory units (Figure 5). Our current model for care must evolve to better serve the affected patient and the entire support system. This lengthy disease continuum begins with primary care and requires the integration of several clinical disciplines. Early diagnosis is the first step in mitigating the effects and consequences of AD. No longer a diagnosis of exclusion, technologies such as biomarkers and multi-pronged assessments are becoming available that greatly improve AD diagnostic accuracy while avoiding expensive structural and genetic testing.

Figure 5.

Early diagnosis of AD is the first important step.