With Alzheimer’s trials starting to recruit cognitively normal and mildly impaired people at risk for Alzheimer’s dementia, researchers need tests that can detect subtle cognitive changes. Many groups are developing cognitive composites that might fit the bill. These composites generally combine those individual tests from larger established batteries that have proven most sensitive to early cognitive decline in longitudinal studies of the preclinical phase (see May 2012 news story; Nov 2012 news story; and April 2013 Webinar). Several composite measures are in the early stages of validation. At the Sixth Clinical Trials Conference on Alzheimer’s Disease (CTAD) held November 14-16 in San Diego, Dorene Rentz at Massachusetts General Hospital, Boston, described two such measures. One is in a traditional pencil-and-paper format, the other a computerized battery. Individual tests were previously shown to capture the first hints of cognitive decline and correlate with disease biomarkers. At CTAD, Rentz reported that the composites appear to be reliable. Not only do they yield similar results when taken multiple times by a given person, but scores on the computerized composite also correlate well with the paper version. The findings support the use of these measures in prevention trials, and suggest that computerized testing could replace paper versions, Rentz said.

The paper composite, formally called the Alzheimer’s Disease Cooperative Study-Preclinical Alzheimer’s Cognitive Composite (ADCS-PACC), will be the primary outcome measure in the upcoming Anti-Amyloid Treatment in Asymptomatic AD (A4) Trial. This study will enroll cognitively normal older adults with evidence of brain amyloid. The ADCS-PACC includes tests that measure learning of word lists, paragraph recall, timed executive function, and global cognition. These items were shown to track cognitive decline in the normal controls in several large longitudinal cohorts such as the Alzheimer’s Disease Neuroimaging Initiative and the Australian Imaging, Biomarkers, and Lifestyle Study. One component of the composite, the Free and Cued Selective Reminding Test, showed the most sensitivity for predicting progression to mild cognitive impairment (MCI) out of a wide range of tests evaluated in a four-year observational study (see Ferris et al., 2006; Mar 2007 Webinar; Dubois et al., 2010).

A4 study participants will take two alternating versions of the ADCS-PACC to avoid the “practice effect,” by which people often improve on repeated testing because they learn the test questions. Rentz and colleagues administered these versions one to two weeks apart to 52 cognitively normal volunteers with an average age of 71. Participants’ two scores showed a correlation of 0.94 (1.0 would be a perfect match), demonstrating that the versions are equivalent, Rentz reported.

In addition, the A4 study will deploy a computerized cognitive composite (C3) on an iPad as an exploratory outcome measure. This battery includes memory tasks such as the Face Name Associative Memory Exam, which asks participants to remember what name goes with a face. Cognitively normal people with brain amyloid perform worse on this task than do those without amyloid (see Rentz et al., 2011). Other memory measures in the composite also correlate with amyloid deposits and functional MRI changes. Some of these come from the Cogstate computerized test battery; they measure reaction time, attention, and processing speed. The researchers are developing numerous versions of the C3 so that participants will see different test questions each time they take it. So far, Rentz and colleagues have validated six versions, which have an overall test-retest coefficient of reliability of 0.83 (researchers consider scores above 0.7 to be good). C3 scores had a 0.75 correlation with those from ADCS-PACC.

The findings validate these tests for use in preclinical populations, Rentz claimed. Paul Maruff at the University of Melbourne, Australia, agreed. “The individual measures in the composite have all been shown to be sensitive to amyloid-related cognitive impairment in healthy older individuals and in MCI,” he wrote to Alzforum. “I think they are ready for use in clinical trials in this population.”

The potential of computerized tests to replace paper versions particularly interests Rentz. Tablet-based tests would have several advantages, she noted. Potentially, trial participants could take the tests at home at regular intervals. This would allow researchers to better track cognitive change over time while reducing clinic visits and keeping down costs. In addition, many of the computer tasks, such as remembering faces and names, are similar to real-world tasks and thus relevant to people’s lives, Rentz said.

Other current prevention trials, such as the ones being run by the Dominantly Inherited Alzheimer Network and the Alzheimer’s Prevention Initiative, use slightly different composites that nonetheless test the same cognitive domains. All the composites also have some elements in common, which will help researchers compare results across trials. Data from these first prevention trials may show which tests are the most sensitive for tracking decline or stability, Rentz said. Maruff sees some advantage to having multiple composites. “Different groups may develop different tests, but ultimately the tests all reflect the same cognitive and clinical characteristics of the disease. It will be reassuring to the public if we can detect drug-related improvement on a variety of appropriate outcome measures, including different composites,” he wrote.—Madolyn Bowman Rogers.


  1. Aside from the obvious conflict-of-interest issues involved with selecting commercial computerized tests, the use of tablet hardware is indicative of a failure to adequately consider usability. At face value, sending people home with nifty tablets may seem cost-efficient compared with clinic visits. However, the wide range of distractions and uncontrolled assistance at home is compounded by the touch and text size problems associated with tablets when used by the elderly. The poor test-retest reliability evident in the current report (and on existing tablet tests, 0.30 to 0.74 on the tablet-administered Computer Assessment of MCI) reflects inadequate attention to usability with geriatric subjects.

    (Disclosure: Dr. Hill was the founder of the CANS-MCI touch screen test battery, administered in doctors' offices only.)


    . Computer assessment of mild cognitive impairment. Postgrad Med. 2009 Mar;121(2):177-85. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. DIAN Grows, Gets Ready for Therapeutic Trials
  2. CTAD: Adaptive Antibody Trial to Try Bayesian Statistics

Webinar Citations

  1. New Frontier: Developing Outcome Measures for Pre-dementia Trials
  2. ADCS Prevention Instrument Project

Paper Citations

  1. . ADCS Prevention Instrument Project: overview and initial results. Alzheimer Dis Assoc Disord. 2006 Oct-Dec;20(4 Suppl 3):S109-23. PubMed.
  2. . Revising the definition of Alzheimer's disease: a new lexicon. Lancet Neurol. 2010 Nov;9(11):1118-27. PubMed.
  3. . Face-name associative memory performance is related to amyloid burden in normal elderly. Neuropsychologia. 2011 Jul;49(9):2776-83. PubMed.

External Citations

  1. Anti-Amyloid Treatment in Asymptomatic AD (A4) Trial

Further Reading