Have you been waiting for a quick yet reliable test for early dementia? The AD8 might be your answer. The interview-based screening tool detects early cognitive impairment, has been validated against the clinical dementia rating scale and against neuropsychological tests, and is available in several languages. Indications are that, combined with a simple performance measure, such as recall of a word list, the AD8 is sufficiently sensitive and specific to serve as an initial screening tool in primary care. The Affordable Care Act, which comes into law beginning January 2011, stipulates that clinicians assess patients for cognitive impairment as part of their annual wellness visit. What tools will primary care physicians use to comply? Could the AD8 fit the bill?

On Tuesday, 30 November, Alzforum hosted a Webinar on the pros, cons, and potential uses of the AD8. Its developer, James Galvin, now at New York University, New York City, reviewed the latest data, including how AD8 scores correlated with AD-like biomarker signatures. Tracey Holsinger, the Durham Veteran Affairs Medical Center, North Carolina, discussed her recent finding that dementia screening is by and large well accepted among the elderly. Joining them for a panel discussion was Eric Tangalos, Mayo Clinic, Rochester, Minnesota. A geriatrician, Tangalos is the past president of the American Medical Directors Association, which promotes quality in long-term care. The Webinar was recorded for posterity and can now be viewed by clicking on the links below.




Background Text
By Tom Fagan

The AD8 is the brainchild of Jim Galvin at New York University, New York City, and John Morris and Cathy Roe at Washington University, St. Louis, Missouri. Galvin demonstrated this brief, informant-based test for early cognitive impairment to the Alzforum audience in a previous Live Discussion in 2007, when he was at WashU, where the test was developed. The test began as a 55-item questionnaire to probe cognitive function, and was whittled down to a final eight items (Galvin et al., 2005). Initially developed in a research setting, the AD8 was subsequently tested in a community-based clinical setting. The researchers reported that the AD8 gives reliable results from test to test and when different raters administer it to the same person. They also found that it is valid and reliable in comparison to the Clinical Dementia Rating Scale and neuropsychological tests of memory and executive function (Galvin et al., 2006). When combined with a brief performance measure, for example, the CERAD 10-item Word List Recall, the AD8 scored at 94 percent sensitivity and 82 percent specificity in detecting early signs of dementia (Galvin et al., 2007).

Though the scientists initially designed the AD8 to be taken by a close relative or caregiver, it later turned out that when patients rated themselves, their scores matched well with what the informants had marked on their behalf (Galvin et al., 2007). This means that in the absence of a qualified informant, the test could still be used. In primary care practice, a qualified informant is often not at hand.

New data now strengthen the case that the test actually diagnoses AD, not another cognitive problem. In the September issue of Brain, Galvin and colleagues at WashU report that AD8 scores correlate with the same AD-like biomarker signature that ADNI and numerous previous studies at WashU and elsewhere have shown to precede and predict frank AD. In a sample of 257 volunteers, a score of 2 or higher on the AD8 (high scores are more predictive of dementia) correlated with positive PIB-PET imaging of amyloid plaques in the brain, with lower cerebrospinal fluid levels of Aβ42, and with higher CSF tau and phospho-tau levels than normal (Galvin et al., 2010).

Galvin and colleagues suggest that the AD8 could be a useful early screening tool in primary care settings. Both neuropsychological testing and CDR are widely used in research settings to identify people with early cognitive impairment, but neither may be particularly suitable for community testing. The MMSE, which is commonly used to screen for dementia, by broad scientific consensus lacks sensitivity to detect early changes. The MMSE and many other brief screening measures are performance based, comparing a person’s test score to accepted norms. These tests provide a cross-sectional “snapshot” of a person’s ability, or attention, at the time of examination but do not measure change within that person, or information on whether the scores relate to any impairment in the person’s activities of daily living. The CDR does measure change in cognitive function as well as interference with ADLs but is too time-consuming for primary care. Some performance tests, such as the MMSE, may also be biased based on age, education, language, race, and culture.

The AD8 is culturally neutral. It has been translated and validated into several languages, including Spanish, Chinese, Portuguese (Correia et al., under review), and Korean (Ryu et al., 2009). Versions in French, Czech, and the Philippine language Tagalog are currently undergoing validation. According to Galvin, the AD8 is being used at a growing number of centers abroad and, in the process, proving to work well in a variety of cultural settings.

Many practitioners are reluctant to screen for early dementia, citing lack of medicines that slow progression of the disease, or inaccuracies in screening (see, e.g., Brayne et al., 2007 and Bond et al., 2010). But attitudes toward screening may soon change. For one, recent work suggests that people and caregivers overwhelmingly favor screening for dementia (see Holsinger et al., 2010 and Bond et al., 2010). In addition, the Affordable Care Act, which comes into law beginning January 2011, stipulates that clinicians assess patients for cognitive impairment as part of their annual wellness visit (see Fact Sheet for Seniors). Medicare billing codes have been set aside for tracking reimbursement for this cognitive checkup, but the Act provides no recommendations for how cognitive impairment in elders should be assessed.


  1. Thank you. As a senior services specialist with a limited and outdated biochemical background, I often have trouble following the articles. I realize the website is only marginally directed to those of my ilk. I am grateful for the early knowledge of what may prove to be a valuable tool for gerontologists.

  2. I'm just a private person who looks after those living with AD, and I think this is a wonderful tool. It will make people’s lives much easier if they know at an early stage what is wrong with them. It will give them time to deal with their diagnosis and give them time to understand what is happening to them, and hopefully give them time to accept what is happening so they will have a more peaceful life with AD instead of one of constant denial and struggle....

  3. As a geriatrician practicing for over 25 years, I have yet to see or administer a "patient-friendly" screening tool that has been well accepted by patients in my practice. If this new tool, the AD8, indeed proves reliable, accurate, reproducible, and possibly reimbursable, it will be very well accepted into many physicians’ offices.

    I look forward to hearing more about it on 30 November.

  4. You cite Galvin et al., 2007, as evidence that "when a qualified informant is not at hand" (or, for that matter not conscientious, unbiased relatives most likely to accompany research subjects to their research appointments), the participant can be used. The Clinical Dementia Rating was adequately correlated with the informant's AD8 scores (rho = 0.75), but not with the participant's (rho -0.34). The significance levels for both were high but irrelevant when you have a lot of subjects. It appears that the AD8 is a snapshot that requires a high-quality informant. Comment?


    . Patient's rating of cognitive ability: using the AD8, a brief informant interview, as a self-rating tool to detect dementia. Arch Neurol. 2007 May;64(5):725-30. PubMed.

  5. Reply to comment by Emory Hill
    The AD8 was designed as an informant assessment to screen for the presence of cognitive impairment of any cause. However, during its development, we also assessed whether in the absence of an informant, the AD8 would be sufficient as a self-rated instrument for detecting impairment. In a study of 325 consecutive patient-informant dyads, we studied the utility of the patient AD8. The CDR was correlated with both informant   (rho = 0.75, p <.001) and participant (rho = 0.34, p <.001) AD8 scores (Table 1 in Galvin et al. 2007).

    However, further review of the table reveals that the correlation is not the most meaningful interpretation for the participant AD8 scores. The Informant AD8 scores continue to worsen with increased CDR staging, hence the strong correlation. In the case of the Participant AD8, individuals with very mild impairments (CDR 0.5) have a positive AD8 score. However, it should be noted that as dementia increases in severity, participants can rate the presence of impairment, but not its severity. Thus, as a screening tool, the AD8 still conforms to its original intent and design: It detects impairment compared with a Gold Standard.

    Another way to view this is to look at the effect size of the AD8 in detecting impairment. Since the greatest challenge to clinicians is the very mildest cases, we can consider only the CDR 0 and CDR 0.5 cases. The Cohen d for the informant AD8 is 1.6, supporting a very large effect for detecting impairment. The Cohen d for the participant AD8 is 0.9, also a very large effect. Thus, the AD8 can be used as a screening tool for informant (preferable) or for patients if no informant is available. The AD8 should not be used as a staging tool for patients since they appear not to be able to rate severity of impairments.


    . Patient's rating of cognitive ability: using the AD8, a brief informant interview, as a self-rating tool to detect dementia. Arch Neurol. 2007 May;64(5):725-30. PubMed.

  6. The AD8 can be a very useful tool for admission screening at Adult Day Health Care Centers. The strained families or caregivers could be answering reliably, since the problem brings them to the health centers for their elderly loved ones. Having administered over 1,500 Mini-Mental Status Questionnaires (MMSQs) at adult day health care facilities, I have seen a significant difference between the MMSQ scores using the word concentration versus that of serial seven countdown calculation. Aside from reliable gender differences in MMSQ scores using word versus calculation, stroke, dementia, and Alzheimer’s, traumatic amnesia, Parkinson’s, schizophrenia and substance use amnestic patients differ with these scores—generally depending on the side of the brain having been affected by the disorder. I have yet to do a statistical analysis on the data since the setting is not permissible. On the face of it, questions number 4 and 6 on the AD8 (“Does your family member have trouble learning how to use a tool, appliance, or gadget, e.g., VCR, computer, microwave, remote control?” and “Does your family member have trouble handling complicated financial affairs, e.g., balancing checkbook, income taxes, paying bills, etc.?”) may positively correlate with the three-stage command (pick paper with right hand, fold, and place down) and serial seven calculation concentration on the MMSQ. There may also be other inter-item correlations that can substantiate patient-family inter-rating reliability measures using both instruments.

Make a Comment

To make a comment you must login or register.


Webinar Citations

  1. The AD8: A Brief Screening Tool for Very Mild Dementia

Paper Citations

  1. . The AD8: a brief informant interview to detect dementia. Neurology. 2005 Aug 23;65(4):559-64. PubMed.
  2. . Validity and reliability of the AD8 informant interview in dementia. Neurology. 2006 Dec 12;67(11):1942-8. PubMed.
  3. . Evaluation of cognitive impairment in older adults: combining brief informant and performance measures. Arch Neurol. 2007 May;64(5):718-24. PubMed.
  4. . Patient's rating of cognitive ability: using the AD8, a brief informant interview, as a self-rating tool to detect dementia. Arch Neurol. 2007 May;64(5):725-30. PubMed.
  5. . Relationship of dementia screening tests with biomarkers of Alzheimer's disease. Brain. 2010 Nov;133(11):3290-300. PubMed.
  6. . Validity and reliability of the Korean version of the AD8 informant interview (K-AD8) in dementia. Alzheimer Dis Assoc Disord. 2009 Oct-Dec;23(4):371-6. PubMed.
  7. . Dementia screening in primary care: is it time?. JAMA. 2007 Nov 28;298(20):2409-11. PubMed.
  8. . Screening for cognitive impairment, Alzheimer's disease and other dementias: opinions of European caregivers, payors, physicians and the general public. J Nutr Health Aging. 2010 Aug;14(7):558-62. PubMed.
  9. . Acceptability of dementia screening in primary care patients. Int J Geriatr Psychiatry. 2011 Apr;26(4):373-9. PubMed.

External Citations

  1. Fact Sheet for Seniors

Further Reading


  1. . Impaired prion replication in spleens of mice lacking functional follicular dendritic cells. Science. 2000 May 19;288(5469):1257-9. PubMed.