Advancing Alzheimer’s Diagnosis: Updated Criteria and the Role of Biomarkers

1. Introduction

Since 1984, the clinical diagnosis of Alzheimer’s disease (AD) has been largely guided by the NINCDS-ADRDA criteria. For over 27 years, these guidelines have served as a cornerstone for clinicians and researchers alike, proving reliable in diagnosing probable AD with a sensitivity of 81% and specificity of 70% across numerous clinical-pathological studies. Their widespread adoption in clinical trials and research underscores their significant impact on the field.

However, the landscape of Alzheimer’s disease research and clinical understanding has dramatically evolved in the ensuing decades. Our knowledge of AD’s clinical presentations, underlying biology, and the availability of advanced diagnostic tools has expanded considerably. Recognizing the necessity to incorporate these advancements, the National Institute on Aging and the Alzheimer’s Association convened a workgroup to undertake a crucial task: revising the 1984 criteria for AD dementia. This revision aims to create a more contemporary, flexible, and accurate diagnostic framework.

The primary objective of this initiative is to refine the diagnostic criteria for AD dementia – specifically, dementia that arises as a consequence of AD pathophysiology. The workgroup embarked on a thorough review of the original NINCDS-ADRDA criteria, with the intention of updating them to reflect the latest innovations in clinical assessment, neuroimaging, and biomarker analysis. The overarching goal was to develop revised criteria that are versatile enough for use across diverse healthcare settings. This includes general practitioners without access to specialized tools like neuropsychological testing, advanced imaging, and cerebrospinal fluid (CSF) analysis, as well as specialized investigators engaged in cutting-edge research and clinical trials where these sophisticated measures are readily available.

This article presents the updated diagnostic criteria for both all-cause dementia and, more specifically, for AD dementia. While maintaining the established framework for probable AD dementia from the 1984 criteria, the revisions incorporate significant changes to the clinical diagnostic process based on over a quarter-century of accumulated experience. The concept of possible AD dementia has been retained but redefined to provide a more precise and focused application. Crucially, biomarker evidence is now integrated into the diagnostic criteria for both probable and possible AD dementia, particularly for use in research settings. It is anticipated that while core clinical criteria will remain fundamental in everyday clinical practice, the incorporation of biomarker evidence will significantly enhance the pathophysiological accuracy and specificity of Alzheimer’s diagnosis. The field now faces the important task of validating the biomarker-based diagnosis of AD dementia through ongoing research and clinical application.

2. Core Clinical Criteria for All-Cause Dementia

Before addressing the specific criteria for AD dementia, it is essential to establish the fundamental criteria for diagnosing dementia of any cause. These core clinical criteria are designed for broad applicability across all clinical environments. Given the multitude of conditions that can lead to dementia, these overarching criteria serve as the initial step in the diagnostic process.

The Diagnosis of dementia is intended to encompass the full spectrum of severity, from the earliest, mildest manifestations to the most advanced stages. While the workgroup’s charge did not extend to the methodologies for staging dementia severity, the criteria themselves are applicable across all stages. Dementia is diagnosed when an individual exhibits cognitive or behavioral (neuropsychiatric) symptoms that collectively meet the following conditions:

1. Functional Impairment: The symptoms must demonstrably interfere with the individual’s ability to function in their occupation or manage their usual daily activities. This signifies a significant decline from their previous functional capacity.

2. Decline from Prior Functioning: The observed symptoms must represent a clear and documented decline from a previous level of cognitive and behavioral performance. This establishes that the current state is not a lifelong condition but rather a deterioration.

3. Exclusion of Delirium or Psychiatric Disorders: The cognitive and behavioral disturbances cannot be solely attributable to delirium, which is an acute state of confusion, or to a major psychiatric disorder, although these conditions can co-exist and complicate the diagnosis.

4. Objective Cognitive Impairment: The presence of cognitive impairment must be both detected and diagnosed through a comprehensive approach. This involves:

   • History-taking: Gathering detailed information from both the patient, when possible, and a knowledgeable informant (e.g., family member, caregiver) about the patient’s cognitive and functional changes over time.
   • Objective Cognitive Assessment: Conducting an objective cognitive assessment, which can range from a brief “bedside” mental status examination to more comprehensive neuropsychological testing. Neuropsychological testing is particularly recommended when routine history and bedside examination are insufficient to establish a confident diagnosis or to characterize the specific cognitive profile.

5. Impairment in Multiple Cognitive Domains: The cognitive or behavioral impairment must affect a minimum of two of the following five cognitive domains:

   a. Memory Impairment: Difficulty in acquiring and remembering new information. This manifests as symptoms such as:

   • Repetitive questioning or engaging in the same conversations repeatedly.
   • Frequently misplacing personal belongings.
   • Forgetting recent events or scheduled appointments.
   • Becoming lost or disoriented even on familiar routes.

   b. Reasoning and Executive Function Impairment: Impaired ability to reason, handle complex tasks, and exercise sound judgment. Symptoms include:

   • Poor understanding of safety risks and potential hazards.
   • Inability to manage personal finances effectively.
   • Demonstrating poor decision-making in everyday situations.
   • Difficulty planning and executing complex or sequential activities.

   c. Visuospatial Impairment: Difficulties with visuospatial abilities, affecting perception and spatial orientation. Symptoms may include:

   • Inability to recognize familiar faces or common objects, even when vision is intact.
   • Difficulty locating objects in plain sight despite good visual acuity.
   • Problems operating simple tools or implements.
   • Challenges in orienting clothing to the body when dressing.

   d. Language Impairment: Difficulties with language functions, including speaking, reading, and writing. Symptoms include:

   • Trouble finding common words during conversations (word-finding difficulties).
   • Hesitations and pauses in speech.
   • Errors in speech, spelling, and writing.

   e. Personality, Behavior, or Comportment Changes: Alterations in personality, behavior, or social conduct. Symptoms can encompass:

   • Uncharacteristic mood fluctuations, such as increased agitation or irritability.
   • Impaired motivation, initiative, and apathy.
   • Social withdrawal and decreased engagement in social interactions.
   • Loss of interest in previously enjoyed activities.
   • Reduced empathy and understanding of others’ emotions.
   • Development of compulsive or obsessive behaviors.
   • Engaging in socially inappropriate or unacceptable behaviors.

Distinguishing dementia from Mild Cognitive Impairment (MCI) is crucial. MCI, as detailed in a companion article, represents a stage of cognitive decline that is greater than expected for an individual’s age but does not yet meet the criteria for dementia. The key differentiating factor lies in the degree of functional interference. Dementia is diagnosed when cognitive impairment significantly interferes with the ability to function at work or in usual daily activities, whereas MCI involves cognitive changes without this level of functional impact. This distinction is fundamentally a clinical judgment, requiring the expertise of a skilled clinician to evaluate the patient’s individual circumstances and the detailed description of their daily life, obtained both from the patient and a knowledgeable informant.

3. Classification Criteria for Alzheimer’s Dementia

When dementia is diagnosed, the next step is to determine the underlying cause, and Alzheimer’s disease is the most common etiology. The proposed classification criteria for dementia caused by AD categorize individuals into three primary groups:

1. Probable AD Dementia: This category represents the typical clinical presentation of AD dementia and is intended for use in all clinical settings.
2. Possible AD Dementia: This category is used when the clinical picture is less clear-cut, either due to atypical features or the presence of other conditions that could contribute to cognitive impairment. It is also intended for general clinical use.
3. Probable or Possible AD Dementia with Evidence of the AD Pathophysiological Process: This category integrates biomarker evidence to increase diagnostic certainty, particularly in research settings. It is currently primarily intended for research purposes but may have increasing clinical utility as biomarker availability expands.

These categories provide a framework for refining the diagnosis of AD dementia, moving from a purely clinical assessment to one that can incorporate biological markers of the disease process.

4. Probable Alzheimer’s Dementia: Core Clinical Criteria

The diagnosis of probable AD dementia relies on specific clinical criteria being met in an individual who has already been diagnosed with dementia according to the all-cause dementia criteria outlined earlier. Probable AD dementia is diagnosed when the following characteristics are present:

4.1. Diagnostic Criteria for Probable AD Dementia

1. Meets Criteria for Dementia: The patient must first fulfill the diagnostic criteria for all-cause dementia, indicating the presence of significant cognitive and functional decline.

2. Insidious Onset: The cognitive symptoms must have an insidious onset, meaning they develop gradually over months to years, rather than abruptly over hours or days. This gradual progression is a hallmark of AD.

3. Clear History of Worsening Cognition: There must be a clear and documented history of progressive cognitive decline, as reported by the patient, informants, or through clinical observation over time. This establishes the progressive nature of the condition.

4. Characteristic Cognitive Deficit Profile: The initial and most prominent cognitive deficits, evident through history and clinical examination, must align with one of the recognized presentations of AD dementia. These presentations include:

   a. Amnestic Presentation: This is the most common presentation of AD dementia. The defining feature is a prominent impairment in learning and recall of recently learned information (memory impairment). Additionally, there should be evidence of cognitive dysfunction in at least one other cognitive domain as defined in the all-cause dementia criteria.

   b. Nonamnestic Presentations: In some individuals, AD dementia may present with prominent deficits in cognitive domains other than memory, at least initially. These nonamnestic presentations include:

   *   **Language Presentation:** The most prominent early deficits are in language, particularly word-finding difficulties. However, deficits in other cognitive domains should also be present to support the diagnosis of AD dementia.
   *   **Visuospatial Presentation:** The most prominent deficits are in visuospatial abilities, such as object agnosia (difficulty recognizing objects), impaired facial recognition, simultanagnosia (difficulty perceiving multiple elements in a visual scene simultaneously), and alexia (acquired reading impairment).  Again, deficits in other cognitive domains should be evident.
   *   **Executive Dysfunction Presentation:**  The most prominent deficits are in executive functions, including impaired reasoning, judgment, and problem-solving. Deficits in other cognitive domains should also be present.

5. Exclusion of Other Likely Causes: The diagnosis of probable AD dementia should not be applied when there is evidence of other conditions that could independently account for or significantly contribute to the dementia. These exclusionary conditions include:

   a. Substantial Concomitant Cerebrovascular Disease: Evidence of significant cerebrovascular disease, such as a history of stroke temporally related to the onset or worsening of cognitive impairment, or the presence of multiple or extensive infarcts (areas of brain tissue death due to lack of blood supply) or severe white matter hyperintensity burden (abnormalities in brain white matter seen on imaging).

   b. Core Features of Dementia with Lewy Bodies (DLB): Presence of core clinical features suggestive of DLB, beyond just the dementia itself. These features include visual hallucinations, parkinsonism, and fluctuating cognition.

   c. Prominent Features of Behavioral Variant Frontotemporal Dementia (bvFTD): Clinical features strongly indicative of bvFTD, such as marked changes in personality and social behavior, disinhibition, and executive dysfunction with relative sparing of memory and visuospatial functions early on.

   d. Prominent Features of Primary Progressive Aphasia (PPA) Variants: Clinical features suggestive of semantic variant PPA (fluent speech with semantic comprehension deficits) or nonfluent/agrammatic variant PPA (nonfluent speech with grammatical errors).

   e. Evidence for Other Concurrent Neurological or Medical Conditions: Presence of another active neurological disease, a non-neurological medical comorbidity, or the use of medications that could have a substantial effect on cognition and explain the dementia syndrome.

It is important to note that all patients who met the criteria for “probable AD” according to the 1984 NINCDS-ADRDA criteria would also meet the current criteria for probable AD dementia. The updated criteria refine and expand upon the original framework, but the core concept of probable AD remains consistent.

4.2. Increasing Certainty in Probable AD Dementia Diagnosis

While the core clinical criteria for probable AD dementia provide a strong basis for diagnosis, certain factors can further increase the certainty that the diagnosis is accurate and reflects the underlying AD pathophysiological process.

4.2.1. Probable AD Dementia with Documented Decline

In individuals who already meet the core clinical criteria for probable AD dementia, documented evidence of progressive cognitive decline over time significantly increases the certainty of an active, evolving pathological process. However, it is important to note that documented decline alone does not specifically confirm that the underlying pathology is AD. Other neurodegenerative conditions can also cause progressive cognitive decline.

Probable AD dementia with documented decline is defined as: Evidence of progressive cognitive decline on subsequent evaluations, based on information from informants and objective cognitive testing, obtained through formal neuropsychological evaluation or standardized mental status examinations conducted over time.

4.2.2. Probable AD Dementia in a Carrier of a Causative AD Genetic Mutation

In individuals who meet the core clinical criteria for probable AD dementia, the identification of a causative genetic mutation for AD significantly increases the certainty that the dementia is indeed caused by AD pathology. These causative mutations are found in genes such as APP, PSEN1, or PSEN2. These mutations are associated with early-onset familial AD.

It is important to note that carrying the ε4 allele of the apolipoprotein E (APOE) gene, a well-known genetic risk factor for late-onset AD, is not considered sufficiently specific to be included in this category. While APOE ε4 increases the risk of developing AD, it is not a causative mutation and is present in a significant proportion of individuals without AD.

5. Possible Alzheimer’s Dementia: Core Clinical Criteria

The diagnosis of possible AD dementia is utilized in situations where the clinical picture is not entirely typical or clear-cut, raising some uncertainty about the diagnosis of probable AD. A diagnosis of possible AD dementia should be considered in either of the following circumstances:

5.1. Atypical Course

An atypical course refers to cases that meet the core clinical criteria for AD dementia in terms of the nature of the cognitive deficits, but deviate from the typical progressive, insidious onset. This includes situations where the cognitive impairment has:

• Sudden Onset: The cognitive decline begins abruptly, over hours or days, rather than the characteristic gradual onset.
• Insufficient Documentation of Progressive Decline: There is a lack of sufficient historical detail or objective cognitive documentation to confirm a progressive decline over time. This may occur when the clinical history is limited or when cognitive testing has not been performed longitudinally.

5.2. Etiologically Mixed Presentation

An etiologically mixed presentation refers to cases that meet all the core clinical criteria for AD dementia but also have evidence of other conditions that could potentially contribute to cognitive impairment. These conditions include:

• Concomitant Cerebrovascular Disease: As defined in the exclusion criteria for probable AD, significant cerebrovascular disease is present, but it is not considered the primary cause of the dementia syndrome.
• Features of Dementia with Lewy Bodies (DLB): Some clinical features suggestive of DLB are present, but they do not fully meet criteria for probable DLB and AD dementia remains a strong consideration.
• Evidence for Another Neurological or Medical Condition: Another neurological disease, a non-neurological medical comorbidity, or medication use is present that could have a substantial effect on cognition, but the clinical picture is still most consistent with AD dementia.

It is important to note that a diagnosis of “possible AD” according to the 1984 NINCDS-ADRDA criteria does not automatically equate to meeting the current criteria for possible AD dementia. Patients previously diagnosed with “possible AD” based on the older criteria would need to be re-evaluated using the revised criteria to determine if they meet the current definition of possible AD dementia.

6. Probable AD Dementia with Evidence of the AD Pathophysiological Process

The integration of biomarkers into the diagnostic criteria for AD dementia represents a significant advancement. Biomarkers provide objective measures of the underlying AD pathophysiological process, enhancing diagnostic accuracy, particularly in research settings. The rationale for incorporating biomarkers is detailed in the introductory article of this series.

The major AD biomarkers that have been extensively studied fall into two main categories, reflecting different aspects of AD biology:

1. Biomarkers of Brain Amyloid-beta (Aβ) Deposition: These biomarkers directly assess the presence of amyloid plaques, a hallmark of AD pathology, in the brain. They include:

   • Low CSF Aβ42: Reduced levels of Aβ42 in cerebrospinal fluid, indicating amyloid accumulation in the brain.
   • Positive PET Amyloid Imaging: Positron emission tomography (PET) scans using amyloid-binding radioligands that show increased amyloid deposition in the brain.

2. Biomarkers of Downstream Neuronal Degeneration or Injury: These biomarkers reflect the neuronal damage and dysfunction that occur downstream of amyloid accumulation in the AD pathophysiological cascade. They include:

   • Elevated CSF Tau (Total Tau and Phosphorylated Tau [p-tau]): Increased levels of tau protein, particularly phosphorylated tau, in CSF, indicating neuronal injury and neurofibrillary tangle pathology. P-tau may be more specific for AD than total tau.
   • Decreased 18-fluorodeoxyglucose (FDG) Uptake on PET in Temporo-parietal Cortex: Reduced glucose metabolism, as measured by FDG-PET, in brain regions typically affected in AD, such as the temporal and parietal lobes, reflecting neuronal dysfunction.
   • Disproportionate Atrophy on Structural Magnetic Resonance Imaging (MRI) in AD- характерных regions: MRI showing significant shrinkage (atrophy) in brain regions vulnerable to AD, including the medial, basal, and lateral temporal lobe, and medial parietal cortex.

In individuals who meet the core clinical criteria for probable AD dementia, biomarker evidence can significantly increase the certainty that the clinical dementia syndrome is indeed due to the AD pathophysiological process. However, it is crucial to emphasize that the use of AD biomarker tests for routine diagnostic purposes is not currently advocated. Several reasons underpin this cautious approach:

1. Accuracy of Clinical Criteria: The core clinical criteria for probable AD dementia already provide very good diagnostic accuracy and clinical utility in the majority of patients in typical clinical settings.
2. Need for Further Research: More research is needed to rigorously validate diagnostic criteria that incorporate biomarkers, ensuring they are optimally designed and interpreted.
3. Limited Standardization: There is currently limited standardization of biomarker assays and imaging protocols across different laboratories and clinical centers, which can affect the reliability and comparability of results.
4. Limited Access: Access to advanced biomarker testing, particularly PET imaging and CSF analysis, is variably limited in community settings, restricting their widespread clinical applicability.

Currently, the use of biomarkers to enhance the certainty of AD pathophysiology is considered most valuable in specific contexts:

• Investigational Studies: In research studies aimed at understanding AD pathogenesis, progression, and treatment.
• Clinical Trials: In clinical trials of potential AD therapies, to ensure that participants have biomarker-confirmed AD pathology.
• Optional Clinical Tools (When Available and Clinically Indicated): In clinical practice, when biomarkers are available and clinicians deem them appropriate to refine the diagnosis in specific cases, particularly when the clinical picture is ambiguous or when there is a need to increase diagnostic certainty for management decisions.

Biomarker test results can be categorized into three interpretations: clearly positive, clearly negative, and indeterminate. The application of biomarkers in the diagnostic process is outlined in Table 1.

Table 1. AD dementia criteria incorporating biomarkers

Diagnostic category	Biomarker probability of AD etiology	Aβ (PET or CSF)	Neuronal injury (CSF tau, FDG-PET, structural MRI)
Probable AD dementia
Based on clinical criteria	Uninformative	Unavailable, conflicting, or indeterminate	Unavailable, conflicting, or indeterminate
With three levels of evidence of AD pathophysiological process	Intermediate Intermediate High	Unavailable or indeterminate Positive Positive	Positive Unavailable or indeterminate Positive
Possible AD dementia (atypical clinical presentation)
Based on clinical criteria	Uninformative	Unavailable, conflicting, or indeterminate	Unavailable, conflicting, or indeterminate
With evidence of AD pathophysiological process	High but does not rule out second etiology	Positive	Positive
Dementia-unlikely due to AD	Lowest	Negative	Negative

Abbreviations: AD, Alzheimer’s disease; Aβ, amyloid-beta; PET, positron emission tomography; CSF, cerebrospinal fluid; FDG, 18fluorodeoxyglucose; MRI, magnetic resonance imaging.

7. Possible AD Dementia with Evidence of the AD Pathophysiological Process

This diagnostic category addresses individuals who present clinically with features suggestive of a non-AD dementia syndrome (e.g., DLB, FTD) but who also have biomarker evidence indicating the presence of AD pathophysiology. This could also include individuals who, despite clinical features not typical for AD, meet neuropathological criteria for AD at autopsy.

For example, an individual might clinically meet criteria for DLB but have a positive AD biomarker study (e.g., positive amyloid PET scan and elevated CSF tau). In such cases, the diagnosis of possible AD dementia with evidence of AD pathophysiological process can be considered.

As indicated in Table 1, a conservative approach is currently recommended: for an individual with a non-AD clinical phenotype to meet criteria for possible AD dementia with biomarker evidence, both categories of biomarkers (amyloid and neuronal injury) should be positive. This cautious approach may be refined as more data emerge regarding the long-term clinical outcomes associated with different combinations of biomarker findings.

It is crucial to recognize that a diagnosis of possible AD dementia with evidence of AD pathophysiological process does not exclude the possibility that a second, co-existing pathophysiological condition is also present and contributing to the clinical picture. Mixed pathologies are increasingly recognized in neurodegenerative diseases.

8. Considerations for Incorporating Biomarkers into AD Dementia Criteria

As highlighted in companion articles discussing the preclinical and MCI stages of AD, AD dementia is understood as part of a continuum of clinical and biological changes. It is fundamentally a clinical diagnosis, meaning that the diagnosis is primarily based on clinical assessment. To diagnose AD dementia with biomarker support, the core clinical criteria for AD dementia must first be met. Biomarkers are used to enhance the certainty of the clinical diagnosis, not to replace it.

CSF biomarkers typically require quantitative interpretation in comparison to established normative ranges. Imaging biomarkers can be interpreted both qualitatively (visually assessed as positive or negative) and quantitatively (measured numerically). In many cases, biomarker results will be clear-cut, either definitively positive or negative. Qualitative interpretation can often readily identify “positive” findings suggestive of underlying AD pathophysiology or “negative” findings suggesting its absence.

However, it is inevitable that some cases will yield ambiguous or indeterminate biomarker results. This is inherent because all biomarkers are continuous measures, and the application of diagnostic labels (“positive” or “negative”) necessitates the use of cutoff values, which can lead to uncertainty for values near the cutoff. While sophisticated quantitative image analysis methods exist, standardized protocols for quantitative analysis of AD imaging tests are still evolving. Current clinical practice in diagnostic imaging often relies on qualitative assessments. Therefore, quantitative interpretation of imaging biomarkers typically relies on laboratory-specific standards. Similar considerations apply to CSF biomarkers, although standardization efforts are more advanced in this domain. Quantitative analytical techniques for biomarkers are continuously evolving. Thus, the practical use of biomarkers must adhere to best-practice guidelines within laboratory-specific contexts until full standardization is achieved across centers.

The temporal sequence of AD pathophysiological events suggests that Aβ pathology becomes abnormal earlier in the disease process, followed by downstream neuronal injury biomarkers becoming abnormal later. This might imply a hierarchical ranking of Aβ biomarkers over neuronal injury biomarkers for diagnostic purposes. However, the reliability and clinical utility of such a hierarchical scheme are not yet sufficiently established for routine use in AD dementia diagnosis. Given the expanding array of AD biomarkers, diverse combinations of test results are expected. For example, cases may present with positive Aβ biomarkers but negative neuronal injury biomarkers, or positive FDG-PET but negative tau measures, and so on. Currently, the data are insufficient to provide definitive recommendations for arbitrating among all possible biomarker combinations. Further research is essential to prioritize biomarkers, determine their relative value and validity in both clinical practice and research settings, and develop algorithms for interpreting complex biomarker profiles.

9. Pathophysiologically Proved AD Dementia

The definitive diagnosis of AD dementia, termed pathophysiologically proved AD dementia, is made when an individual meets both the clinical and cognitive criteria for AD dementia and neuropathological examination of brain tissue confirms the presence of AD pathology. This neuropathological confirmation is based on widely accepted criteria for the postmortem diagnosis of AD, which typically involve assessing the burden of amyloid plaques and neurofibrillary tangles in specific brain regions. This category is primarily relevant in research contexts and in cases where autopsy is performed.

10. Dementia Unlikely to Be Due to AD

The category of dementia unlikely to be due to AD is applied in two main scenarios:

1. Failure to Meet Clinical Criteria for AD Dementia: The individual does not meet the core clinical criteria for probable or possible AD dementia. Their clinical presentation is more consistent with another type of dementia or cognitive disorder.

2. Evidence for an Alternative Diagnosis or Negative Biomarkers:

   a. Sufficient Evidence for an Alternative Diagnosis: Regardless of whether the individual meets clinical criteria for probable or possible AD dementia, there is compelling evidence for an alternative diagnosis that is unlikely to overlap with AD. Examples include HIV-associated dementia, Huntington’s disease dementia, or other rare dementias that have distinct clinical features and underlying pathologies.

   b. Negative Amyloid and Neuronal Injury Biomarkers in Possible AD: In individuals who meet clinical criteria for possible AD dementia, but both amyloid (Aβ) and neuronal injury biomarkers are negative, the diagnosis of AD dementia becomes less likely. Negative biomarkers in this context suggest that the underlying cause of the dementia syndrome is likely not AD pathology.

These criteria for “dementia unlikely to be due to AD” help to refine the differential diagnosis and identify individuals whose cognitive impairment is likely attributable to a different etiology than Alzheimer’s disease.