Research in 2012 and 2013 offered new insights into the trajectory of the Alzheimer’s disease process. New evidence suggests that cellular changes associated with the disease begin years—even decades—before people first show clinical symptoms of memory loss or cognitive difficulties. Increasingly, we are using biomarkers—specific proteins in blood or cerebrospinal fluid (CSF), for example, or imaging of brain structure and function—to measure risk for Alzheimer’s, even in symptom-free people. Biomarkers are also used in research settings to track the onset and progression of the disease, as well as the effectiveness of promising interventions.
The findings described below highlight ways in which biomarkers are advancing our understanding of Alzheimer’s disease.
View a short video of NIA Division of Neuroscience Director Dr. Neil Buckholtz discussing the use of biomarkers:
In 2010, experts at the Mayo Clinic, Rochester, MN, first proposed a model suggesting that Alzheimer’s biomarkers emerge at different stages of the disease (Jack et al., 2010). The model proposed markers that tracked disease onset and progression: changes in levels of amyloid protein in the brain and CSF were the first sign of disease onset, followed by changes in the level of the protein tau in CSF, changes in brain structure, and finally memory loss and other clinical symptoms. Subsequent research has generally supported this model, with one key exception: the relationship of abnormal tau to other biomarkers and disease progression was puzzling. In people with Alzheimer’s disease, tau abnormalities in CSF become detectable only after brain amyloid accumulation has begun. However, autopsy analyses have detected abnormal tau in the brains of most middle-aged people, almost always before amyloid markers change.
In 2013, Jack and colleagues revised their model to show that tau changes begin before amyloid changes, but that amyloid changes occur faster and are the first ones detectable (Jack, Knopman et al., 2013). Like many other researchers, they suggest beta amyloid accumulates due to its overproduction and/or under-clearance. Jack and his team further suggest that beta amyloid, rather than tau, may initiate disease onset. One hypothesis now being considered is that amyloid pathology exacerbates typical, age-related accumulation of abnormal tau, leading to the severe tau pathology seen in people with Alzheimer’s.
Alzheimer’s Disease Progression
This diagram illustrates how Alzheimer’s disease-related changes may occur in the brain long before symptoms of cognitive decline first appear in people with mild cognitive impairment (MCI). The curves represent the sequence in which specific markers may play a role as people progress from normal cognition, to MCI, and finally, to dementia. This model suggests that in typical late-onset Alzheimer’s disease, tau changes may begin before amyloid changes, but that amyloid changes occur faster and are usually the first ones detectable. It also suggests that amyloid accumulation drives progression of tau and other downstream events in the disorder (Jack, Knopman et al., 2013).
Researchers at the Mayo Clinic, Rochester, MN, identified a window of several years during which treatments targeting beta-amyloid buildup might be most effective (Jack, Wiste et al., 2013). The researchers used positron emission tomography (PET) scans to study beta-amyloid accumulation in the brains of 260 people age 70 to 92. At the start of the study, 22 percent of the volunteers had mild cognitive impairment or Alzheimer’s, and the rest were cognitively normal. The PET images revealed that beta-amyloid levels built up slowly over an average period of 15 years, then reached a plateau. These results suggest that there may be a long window of opportunity for future treatments that slow beta-amyloid buildup.
In a study from the Dominantly Inherited Alzheimer Network (DIAN), at Washington University, St. Louis, researchers found that years—even decades—could pass between initial changes in Alzheimer’s-related biomarkers and the expected onset of symptoms in people with the rare, familial form of the disease (Bateman et al., 2012). In the familial form of Alzheimer’s, clinical symptoms typically begin in the mid-40s in people who inherit certain gene mutations (in either the amyloid precursor protein gene or the presenilin gene) from a parent. This study involved 128 volunteers with a parent with the familial form of Alzheimer’s; of these volunteers, 88 had inherited one of the gene mutations. The expected age of symptom onset in mutation carriers can be estimated based on the age at which the parent with the mutation first showed symptoms.
The researchers examined CSF and brain-imaging biomarkers and conducted cognitive testing on all of the study participants. Those with gene mutations showed declines in CSF beta-amyloid 42 levels a full 25 years before the age at which they would be expected to show clinical symptoms. Beta-amyloid 42 is a major component of amyloid plaques in the brain, so decreased levels in CSF may indicate Alzheimer’s-related changes are taking place in the brain.
Other Alzheimer’s-related biomarkers, such as beta-amyloid deposits in the brain, brain atrophy, and increased CSF tau concentrations, emerged on average 15 years before clinical symptoms were expected. Decreased brain metabolism and impaired memory were detected some 10 years before expected clinical symptoms. Significantly, the volunteers experienced a wide array of cognitive deficits 5 years before the expected onset of clinical symptoms, although they did not meet diagnostic criteria for dementia until 3 years after the expected age of onset.
A second study of people carrying mutations for early-onset Alzheimer’s disease complements these findings. Led by researchers at Banner Alzheimer’s Institute, Phoenix, the study involved 30 young adults, all of whom carried the same mutation in the presenilin 1 gene (Reiman et al., 2012). Magnetic resonance imaging scans revealed functional and structural changes in their brains more than 2 decades before they would, on average, be expected to show signs of cognitive impairment. A number of the participants also had increased levels of beta-amyloid 42 in their CSF and blood. This finding is consistent with the idea that this presenilin mutation leads to overproduction of beta-amyloid 42, a more toxic form of the protein.