Use of Alzheimer's Disease Biomarkers to Predict Longevity and Disability

Catherine Roe, Ph.D. 

Project Overview:
Deaths due to Alzheimer’s disease (AD) increased 66% between 2000 and 2008, although other leading causes of death have actually decreased in recent years. As the average age of Americans increases, the number of deaths due to Alzheimer’s disease will increase even more unless scientists can develop an effective treatment soon. 

The mission of the Longer Life Foundation is to study factors that help to predict death and illness. This increased knowledge should contribute to longer lifespans and to a healthier older adulthood with less disability and more independence. In 2011, we received one year of funding from the Foundation to examine and compare the accuracy of AD biomarkers in predicting who will develop symptoms of AD in the future. An “AD biomarker” is obtained by taking images of the brain, or by testing spinal cord fluid, to see whether telltale proteins are present indicating that the person has AD, even though he or she may not be showing any problems with thinking and memory yet. 

We now would like to test whether these biomarkers predict time to death and other difficulties that often accompany AD, such as problems carrying out activities of daily living, weight loss, emotional problems, depression, and the need to go to a nursing home. 

AD biomarkers are beginning to be used to help develop drugs. By using biomarkers, scientists can identify people who are likely to develop AD symptoms within a few years and can enroll them in drug trials. This speeds up the ability to know whether the drug is working. Also, it is thought that drugs developed using biomarkers may work better if they are given very early in the course of AD, before the person has developed many symptoms and many brain cells have died. 

The Charles F. and Joanne Knight Alzheimer’s Disease Research Center (ADRC) at Washington University School of Medicine in St. Louis, has substantial experience working with AD biomarkers and testing how they are associated with memory and thinking problems. We now want to study how biomarkers are associated with other important consequences of having AD. If we know how biomarkers are related to disease outcomes, this can help us figure out how to intervene and prevent future disability and death.

Final Report Abstract:
We tested whether cerebrospinal fluid (CSF) biomarkers of Alzheimer’s disease (AD) predicted future decline in “non-cognitive” outcomes among individuals who were cognitively normal at baseline. We used data from 430 participants aged 50 years and above, who donated CSF within one year of a clinical assessment indicating normal cognition. Mixed linear models were used to test whether baseline values of the biomarkers (Aβ42, tau, ptau181, tau/Aβ42, and ptau181/Aβ42) predicted future decline in function (instrumental activities of daily living), weight, behavior and mood. Clinical Dementia Rating (CDR) Sum of Boxes and Mini-Mental State Exam (MMSE) scores were also examined. Results showed that abnormal levels of each biomarker were related to greater impairment with time in behavior (p<.035) and mood (p<.012) symptoms, and more difficulties with independent activities of daily living (p<.012).


However, biomarker levels were unrelated to weight change with time (p>.115). As expected, abnormal biomarker values also predicted more rapidly changing scores on the cognitive measures compared to normal values. We found that CSF biomarkers among cognitively normal individuals are associated with future decline in some, but not all, non-cognitive AD symptoms studied. More work is needed to determine the extent to these findings generalize to other samples. Read the full Final Report.