An NIA research team led by Dr. Susan Resnick images the living brain to detect the earliest changes that signal the onset of Alzheimer’s disease.
Many researchers believe that treatments for Alzheimer’s are more likely to be effective if initiated early in the disease process. It is now thought that Alzheimer’s-related changes in the brain can begin years, even decades, before cognitive impairment becomes evident. Researchers are developing methods to detect these changes at their earliest stages. These efforts are designed to determine who is at the highest risk for Alzheimer’s so that possible treatments can be tested more rapidly and effectively. Identifying those at high risk will also improve diagnosis in clinical practice, better serving patients and their families.
Scientists continue to explore three main approaches to early diagnosis: measurements of biomarkers in blood and cerebrospinal fluid (CSF), brain imaging, and standardized clinical tests of memory and thinking abilities to determine cognitive health. Through the National Institute on Aging-led Alzheimer’s Disease Neuroimaging Initiative (ADNI) and other studies, these efforts are already showing some success. (See “Supporting Infrastructure and Initiatives ” for more on ADNI.) Scientists now are beginning to explore the combined use of biomarkers and brain imaging to predict disease risk. A growing body of research is devoted to looking for blood proteins whose levels change during the early stages of Alzheimer’s, which could lead to the development of routine blood tests for risk assessment.
Cognitive decline may begin earlier in life than researchers had previously believed. A study led by researchers at University College, London, and the Institut National de la Santé et de la Recherche Médicales (INSERM) in Villejuif, France, analyzed 10 years of cognitive test data for more than 10,000 British civil servants participating in the long-running Whitehall II study  (Singh-Manoux et al, 2011). Age 45 to 70 years old when the study began, the participants were tested on verbal and mathematical reasoning, verbal memory, verbal fluency, and vocabulary.
Performance in all areas except vocabulary declined over time in all age groups. For example, reasoning test scores declined by 4 percent in the 45- to 49-year-old group and by 10 percent in the 65- to 70-year-old group. Relatively recent improvements in women’s educational attainment resulted in higher scores, and although the scores still declined with age, the extent of the decline was less than might have been expected from previous cross-sectional data. This study suggests that efforts to detect and prevent declines in cognitive health may need to start in adults as young as 45 years.
Scientists have developed a number of biomarkers that help detect Alzheimer’s disease, track its progression, and evaluate promising therapies. For example, lumbar punctures are used to identify Alzheimer’s-related proteins in CSF, and brain scans detect changes in brain structure and function. Because the use of these biomarkers is limited primarily to research settings, scientists are seeking more readily accessible tests.
Toward that goal, researchers at the University of Pennsylvania, Philadelphia, screened levels of 190 different proteins in blood samples from 600 older people being treated for memory complaints at the University’s clinical center and at the Washington University School of Medicine, St. Louis (Hu et al., 2012). Seventeen of the proteins tested were associated with a diagnosis of mild cognitive impairment (MCI) or Alzheimer’s in both groups. Of those proteins, four were associated with MCI or Alzheimer’s in a third group of more than 550 volunteers from ADNI . The four proteins—apolipoprotein E, B-type natriuretic protein, C-reactive protein, and pancreatic polypeptide—were also strongly associated with CSF Alzheimer’s biomarker levels. These findings add to a growing body of studies aiming at a possible blood test to help detect Alzheimer’s.
To make progress toward developing such a blood test, the association of individual markers with disease status will have to be further tested. For example, the University of Pennsylvania researchers associated reduced levels of blood C-reactive protein, a marker of inflammation, with MCI and Alzheimer’s.
In contrast, a similar study at Tufts University, Medford, MA, reached a different conclusion. Among some 800 participants (average age, 76) in the Framingham Heart Study , Tufts researchers found no association between C-reactive protein levels at the start of the study and risk of developing dementia over the next 13 years (van Himbergen et al., 2012). Among the women in the Framingham group, however, increased levels of plasma adiponectin, a protein involved in glucose metabolism, were associated with increased risk of developing Alzheimer’s disease and related dementias. Conflicting findings in blood biomarker research underscore the importance of testing and standardizing these measurements before introducing them into clinical practice.
Researchers from the Biomarkers Consortium ‘s Alzheimer’s Disease Plasma Proteomics Project identified a panel of plasma proteins with levels strongly associated with apolipoprotein E (APOE) allele status in people with Alzheimer’s, regardless of the patients’ disease status (Soares et al., 2012). The Consortium, a public-private biomedical research partnership managed by the Foundation for the National Institutes of Health,  conducted the study with support and volunteers from ADNI . This finding highlights the strong influence a person’s genetic makeup can have on blood protein profiles and may help explain why certain proteins that appear to be linked with Alzheimer’s in some groups of people fail to show that link in other groups. Studying plasma proteins with levels associated with APOE allele status could also offer further insight into the biological functions of the different alleles.
The diagnostic category “preclinical Alzheimer’s” identifies individuals who are cognitively normal but whose brains are undergoing changes that may eventually lead to clinical symptoms of Alzheimer’s. Harvard University, Cambridge, MA, researchers studied a brain imaging biomarker for its potential value in identifying people with preclinical Alzheimer’s: a pattern of thinning in specific cortical areas that is detectable on magnetic resonance imaging scans and common in people with mild Alzheimer’s (Dickerson et al., 2012).
The scientists scanned nine areas of the cortex—all involved in the brain’s higher-level processing of complex information—to determine their average thickness. Among a group of 159 cognitively normal people (average age, 76), the researchers identified 19 whom they judged to be at high risk of Alzheimer’s based on their patterns of cortical thinning. During the subsequent 3 years, participants classified as high risk were three times more likely than low-risk participants to experience cognitive decline and more likely to show Alzheimer’s-like changes in CSF beta-amyloid levels.
Declining gait speed may predict cognitive decline in older adults. However, it can be difficult to recognize and assess gradual changes in walking speed. To obtain continuous daily assessments of gait speed, researchers at the Oregon Health & Sciences University, Portland, OR, used an in-home device that unobtrusively monitors gait speed by collecting measurements from infrared motion sensors (Dodge et al., 2012).
Over more than 3 years, the researchers collected data from 93 volunteers (average age, 85) who had either normal cognition, amnestic MCI (the form that usually precedes Alzheimer’s), or nonamnestic MCI. They found that participants with amnestic MCI were more likely to experience declines in walking speed. The study suggests that the use of in-home activity monitors offers a new method for detecting changes in motor function that may signal early stages of progression to Alzheimer’s dementia.
Examination of brain tissue after death is commonly used to diagnose Alzheimer’s disease. Scientists are trying to improve methods for processing and analyzing brain tissue samples to increase the accuracy of diagnosis and better compare results obtained in different laboratories. In 2012, an expert panel convened by the National Institute on Aging and the Alzheimer’s Association published revised guidelines for evaluating Alzheimer’s pathology  in post mortem brain tissue (Hyman et al., 2012).
In a follow-up report, the panel offered practical guidelines for pathologists to assess autopsy tissue for signs of Alzheimer’s disease and other dementias (Montine et al., 2012). The report suggests standardized methods for staining brain tissue for cellular markers associated with Alzheimer’s (plaques and tangles) and for “scoring” the level of Alzheimer’s-like pathology in the tissue. It also emphasizes the importance of looking for signs of other neurodegenerative processes and offers guidelines for assessing signs of Lewy body disease and cerebrovascular disease, two other diseases that commonly contribute to dementia in older people. These advances in post mortem brain analysis will offer new insights into the cellular and structural changes caused by dementia disorders.
Neuropathological criteria may be of less use in identifying causes of dementia in people who live to age 90 or older than in people who die younger. University of California, Irvine, researchers performed post mortem analyses of the brains of 104 people from The 90+ Study, one of the largest studies of the “oldest old” in the world. All of the volunteers had received clinical diagnoses before their deaths (Corrada et al., 2012).
The researchers found that post mortem signs of Alzheimer’s and cerebrovascular disease were common in the brains of both demented and cognitively normal participants. However, cellular changes seen in the brains after death did not correlate well with the clinical diagnoses made before death. For example, Alzheimer’s pathology was seen in 49 percent of the people who were deemed cognitively normal before they died and in 57 percent of those who had had dementia. Yet among those diagnosed with dementia, more than 20 percent did not show sufficient brain pathology of any kind to account for their cognitive decline. This study suggests that dementia among people of very advanced age may result from multiple cellular pathologies and/or combinations of pathologies, possibly including some yet to be identified.