|Home | About | Journals | Submit | Contact Us | Français|
The National Institute on Aging and the Alzheimer’s Association charged a workgroup with the task of revising the 1984 criteria for Alzheimer’s disease (AD) dementia. The workgroup sought to ensure that the revised criteria would be flexible enough to be used by both general healthcare providers without access to neuropsychological testing, advanced imaging, and cerebrospinal fluid measures, and specialized investigators involved in research or in clinical trial studies who would have these tools available. We present criteria for all-cause dementia and for AD dementia. We retained the general framework of probable AD dementia from the 1984 criteria. On the basis of the past 27 years of experience, we made several changes in the clinical criteria for the diagnosis. We also retained the term possible AD dementia, but redefined it in a manner more focused than before. Bio-marker evidence was also integrated into the diagnostic formulations for probable and possible AD dementia for use in research settings. The core clinical criteria for AD dementia will continue to be the cornerstone of the diagnosis in clinical practice, but biomarker evidence is expected to enhance the pathophysiological specificity of the diagnosis of AD dementia. Much work lies ahead for validating the biomarker diagnosis of AD dementia.
In the fall of 1983, a group was convened by the National Institute of Neurological and Communicative Disorders and Stroke (NINCDS) and the Alzheimer’s Disease and Related Disorders Association (ADRDA) to establish criteria and to describe the clinical diagnosis of Alzheimer’s disease (AD). The group addressed issues of medical history, clinical examination, neuropsychological testing, and laboratory assessments and then produced a report, which was published in July 1984 . The criteria in this report, commonly referred to as the NINCDS-ADRDA criteria, have been quite successful, surviving for over 27 years. These criteria have been reliable for the diagnosis of probable AD, and across more than a dozen clinical pathological studies have had a sensitivity of 81% and specificity of 70% , They have been widely used in clinical trials and clinical research.
However, now 27 years later, these criteria require revision. Therefore, the National Institute on Aging and the Alzheimer’s Association charged a workgroup with the task of revising the 1984 criteria for AD dementia. Details of the charge to the workgroup are described in the Introduction that accompanies this article . The characterization of the preclinical  and mild cognitive impairment (MCI)  phases of the AD pathophysiological processes is described in the companion articles.
Our knowledge of the clinical manifestations and biology of AD has increased vastly. The features of the original criteria that required revision include the following:
The objective of our committee was to focus on the criteria for AD dementia, that is, dementia secondary to the pathophysiology of AD. It was our intention to first review the NINDS–ADRDA criteria and then to update them, incorporating more modern innovations in clinical, imaging, and laboratory assessment. We will first propose (1) Criteria for all-cause dementia and then, (2) Criteria for dementia caused by AD. We set ourselves the goal of ensuring that the revised criteria would be flexible enough to be used by both general healthcare providers without access to neuropsychological testing, advanced imaging, and CSF measures, as well as specialized investigators involved in research or in clinical trial studies who would have these measures available.
In this section, we outline core clinical criteria to be used in all clinical settings. Because there are many causes of dementia, we will first outline the criteria for all-cause dementia.
The diagnosis of dementia is intended to encompass the spectrum of severity, ranging from the mildest to the most severe stages of dementia. The methodology for staging of dementia severity was beyond the charge of the workgroup. Dementia is diagnosed when there are cognitive or behavioral (neuropsychiatric) symptoms that:
The differentiation of dementia from MCI (see companion article  on the diagnosis of MCI) rests on the determination of whether or not there is significant interference in the ability to function at work or in usual daily activities. This is inherently a clinical judgment made by a skilled clinician on the basis of the individual circumstances of the patient and the description of daily affairs of the patient obtained from the patient and from a knowledgeable informant.
We propose the following terminology for classifying individuals with dementia caused by AD: (1) Probable AD dementia, (2) Possible AD dementia, and (3) Probable or possible AD dementia with evidence of the AD pathophysiological process. The first two are intended for use in all clinical settings. The third is currently intended for research purposes.
Note: All patients who met criteria for “probable AD” by the 1984 NINCDS–ADRDA criteria  would meet the current criteria for probable AD dementia mentioned in the present article.
In persons who meet the core clinical criteria for probable AD dementia, documented cognitive decline increases the certainty that the condition represents an active, evolving pathologic process, but it does not specifically increase the certainty that the process is that of AD pathophysiology.
Probable AD dementia with documented decline is defined as follows: evidence of progressive cognitive decline on subsequent evaluations based on information from informants and cognitive testing in the context of either formal neuropsychological evaluation or standardized mental status examinations.
In persons who meet the core clinical criteria for probable AD dementia, evidence of a causative genetic mutation (in APP, PSEN1, or PSEN2), increases the certainty that the condition is caused by AD pathology. The workgroup noted that carriage of the ε4 allele of the apolipoprotein E gene was not sufficiently specific  to be considered in this category.
A diagnosis of possible AD dementia should be made in either of the circumstances mentioned in the following paragraphs.
Atypical course meets the core clinical criteria in terms of the nature of the cognitive deficits for AD dementia, but either has a sudden onset of cognitive impairment or demonstrates insufficient historical detail or objective cognitive documentation of progressive decline,
Etiologically mixed presentation meets all core clinical criteria for AD dementia but has evidence of (a) concomitant cerebrovascular disease, defined by a history of stroke temporally related to the onset or worsening of cognitive impairment; or the presence of multiple or extensive infarcts or severe white matter hyperintensity burden; or (b) features of Dementia with Lewy bodies other than the dementia itself; or (c) evidence for another neurological disease or a non-neurological medical comorbidity or medication use that could have a substantial effect on cognition
Note: A diagnosis of “possible AD” by the 1984 NINCDS-ADRDA criteria  would not necessarily meet the current criteria for possible AD dementia. Such a patient would need to be re-evaluated.
The rationale for including biomarkers for the pathophysiological process of AD in the diagnostic criteria is summarized in the Introduction to this series of articles , The major AD biomarkers that have been widely investigated at this time (see  for review) may be broken into two classes based on the biology which they measure. Biomarkers of brain amyloid-beta (Aβ) protein deposition are low CSF Aβ42 and positive PET amyloid imaging [22,23]. The second category is that of biomarkers of downstream neuronal degeneration or injury. The three major bio-markers in this category are elevated CSF tau, both total tau and phosphorylated tau (p-tau); decreased 18fluorodeoxyglucose (FDG) uptake on PET in temporo–parietal cortex; and disproportionate atrophy on structural magnetic resonance imaging in me-dial, basal, and lateral temporal lobe, and medial parietal cortex. Total tau and p-tau are treated equivalently in this study, although p-tau may have more specificity for AD than other dementing diseases.
In persons who meet the core clinical Criteria for probable AD dementia biomarker evidence may increase the certainty that the basis of the clinical dementia syndrome is the AD pathophysiological process. However, we do not advocate the use of AD biomarker tests for routine diagnostic purposes at the present time. There are several reasons for this limitation: (1) the core clinical criteria provide very good diagnostic accuracy and utility in most patients; (2) more research needs to be done to ensure that criteria that include the use of biomarkers have been appropriately designed, (3) there is limited standardization of biomarkers from one locale to another, and (4) access to biomarkers is limited to varying degrees in community settings. Presently, the use of biomarkers to enhance certainty of AD pathophysiological process may be useful in three circumstances: investigational studies, clinical trials, and as optional clinical tools for use where available and when deemed appropriate by the clinician.
Biomarker test results can fall into three categories–clearly positive, clearly negative, and indeterminate. We envision that application of biomarkers for the AD pathophysiological process would operate as outlined in the Table 1.
This category is for persons who meet clinical criteria for a non-AD dementia but who have either biomarker evidence of AD pathophysiological process, or meet the neuropathological criteria for AD. Examples would include persons who meet clinical criteria for dementia with Lewy bodies or for a subtype of frontotemporal lobar degeneration, but who have a positive AD biomarker study or at autopsy are found to meet pathological criteria for AD. In the biomarker table, we indicate that both categories of biomarkers must be positive for an individual who presents clinically with a non-AD phenotype to meet criteria for possible AD. This is a conservative approach that may change as more information is gained concerning the long-term outcomes of different combinations of biomarker findings. A diagnosis of possible AD dementia with evidence of AD pathophysiological process does not preclude the possibility that a second pathophysiological condition is also present.
As described in the two companion articles on the preclinical  and MCI  phases of the AD pathophysiological process, AD dementia is part of a continuum of clinical and biological phenomena. AD dementia is fundamentally a clinical diagnosis. To make a diagnosis of AD dementia with biomarker support, the core clinical diagnosis of AD dementia must first be satisfied.
According to their nature, CSF biomarkers rely on a quantitative interpretation in comparison with normative standards. Imaging biomarkers can be interpreted in both a qualitative or quantitative manner. In many cases, biomarker results will be clearly normal or abnormal. In these cases, a qualitative interpretation of a biomarker test will unequivocally identify “positive” findings that imply the presence of the underlying AD pathophysiological process, or negative findings that unequivocally imply absence of an AD pathophysiological process. However, in some cases, ambiguous or indeterminate results will be obtained. This is inevitable given that all biomarkers are continuous measures, and the diagnostic labels of “positive” or “negative” require that cutoff values be applied to continuous biological phenomena. Although sophisticated quantitative and objective image analysis methods do exist, at present, accepted standards for quantitative analysis of AD imaging tests are lacking. Standard clinical practice in diagnostic imaging is qualitative in nature. Therefore, quantification of imaging biomarkers must rely on local laboratory specific standards. The same holds true for CSF biomarkers, although standardization efforts are more advanced for CSF biomarkers than for the imaging tests. Quantitative analytic techniques are, and will continue to be in evolution for some time. Therefore, practical use of biomarkers must follow best-practice guidelines within laboratory-specific contexts, until standardization has been fully accomplished.
A sequence of events has been described with Aβ pathophysiological processes becoming abnormal first and downstream neuronal injury biomarkers becoming abnormal later [6,7], This might imply a hierarchical ranking of Aβ biomarkers over downstream neuronal injury biomarkers for diagnostic purposes. However, at this time, the reliability of such a hierarchical scheme has not been sufficiently well established for use in AD dementia. Given the number of different AD biomarkers, it is inevitable that different combinations of test results can occur. For example, individual cases might be encountered with a positive Aβ and negative neuronal injury biomarker, or a positive FDG PET and negative tau measure, and so on. At present, the data are insufficient to recommend a scheme that arbitrates among all different biomarker combinations. Further studies are needed to prioritize biomarkers and to determine their value and validity in practice and research settings.
The diagnosis of pathophysiologically proved AD dementia would apply if the patient meets the clinical and cognitive criteria for AD dementia outlined earlier in the text, and the neuropathological examination, using widely accepted criteria , demonstrates the presence of the AD pathology.
The authors acknowledge the assistance of Dr. Cerise Elliott at the National Institute on Aging.
Guy McKhann serves on a Data Safety Monitoring Board for Merck. David Knopman serves on a Data Safety Monitoring Board for Lilly Pharmaceuticals and is an investigator for clinical trials sponsored by Elan Pharmaceuticals, Forest Pharmaceuticals, and Baxter Healthcare; he is deputy editor of Neurology and receives compensation for editorial activities. Howard Chertkow serves as a consultant to Pfizer Canada, Lundbeck Canada, Janssen Ortho, Novartis Canada, and Bristol Myers Squibb; he receives a research grant from Pfizer Canada. Bradley Hyman serves as a consultant to EMD Serrano, Janssen, Takeda, BMS, Neurophage, Pfizer, Quanterix, foldrx, Elan, and Link, and receives funding from the NIH, the Alzheimer’s Association, and Fidelity Biosciences. Clifford Jack serves as a consultant for Eli Lilly, Eisai, and Élan; he is an investigator in clinical trials sponsored by Baxter and Pfizer Inc., and owns stock in Johnson and Johnson. Claudia Kawas serves on a Data Safety Monitoring Board for Lilly Pharmaceuticals, Elan Pharmaceuticals, and Lundbeck; she is an investigator in a trial sponsored by Avid Radiopharmaceuticals. William Klunk serves as a consultant to GE Healthcare and receives research grants from the same; he also receives royalties from GE Healthcare for PiB PET technology and owns stock or options in Neuroptix, a company seeking to commercialize detection of amyloid in the eye. Walter Koroshetz are employees of the U.S. Government and report no conflicts. Jennifer Manly reports no conflicts of interests. Richard Mayeux serves on scientific advisory board of Psycho-Genics. Richard Mohs is a full-time employee of Eli Lilly and Company and holds stock in Lilly. Avid Radiopharmaceuticals is a wholly owned subsidiary of Eli Lilly and Co. John Morris serves as a consultant to Astra Zeneca, Bristol-Myers Squibb, Eisai, Janssen, Genetic, Eli Lilly, Merck, Novartis, Otsuka, Pfizer, and Schering Plough. University College London receives payment for Martin Rossor serving on the Safety Monitoring Committees for Janssen and Servier trials in AD. Philip Scheltens serves as a consultant to Pfizer Pharmaceuticals, Genetech, Danone Research, Lundbeck Pharmaceuticals, GE Healthcare, Roche, and Novartis; he also serves on a speakers bureau for Lundbeck Pharmaceuticals. Maria Carrillo and Bill Thies are employees of the Alzheimer’s Association and reports no conflicts. Sandra Weintraub reports no conflicts of interest and Creighton Phelps.