Improving Presymptomatic Detection of Impending Alzheimer’s Disease
Catherine Roe, Ph.D.
So far, no medications have been shown to be good treatments for Alzheimer’s disease (AD). This may be because the drugs are being administered too late. Once people start showing symptoms of AD, the brain has already suffered substantial damage. Therefore, researchers are very interested in finding biomarkers that can provide a way to detect underlying AD before the affected person experiences any symptoms.
Especially promising biomarkers are tests that indicate that abnormal levels of AD proteins are present in the body. Because these biomarkers may be abnormal a decade or more before the appearance of dementia symptoms, understanding the relationships between abnormal biomarker levels and when someone will experience dementia symptoms is very important. Understanding these relationships will improve the design of clinical trials of AD drugs by including in the trials many individuals who are likely to experience dementia symptoms within the period of the trial. It will also help to identify the best time to begin administration of AD medications, and will prevent exposing healthy people to potential side effects of AD medications many years before these drugs are needed.
Little is currently known about how accurate these biomarkers are in predicting who will get AD in the future, how the accuracy of different biomarkers compares to one another, and whether the accuracy of biomarkers can be improved by combining them in predictive computer models with variables reflecting attributes of the individual. Some of our previous research suggests that the ability of some biomarkers to predict when a person will get AD (biomarkers obtained by imaging amyloid protein in the brain and by lumbar puncture to acquire cerebrospinal fluid [CSF]), can be improved by combining biomarker information with information about the size of an individual’s brain and about some of his or her life experiences, such as the amount of education received.
This project will follow over time 209 individuals, aged 50 years and older and with normal memory and thinking at the beginning of the study. We will (1) examine how individual attributes of people combine with imaging and CSF biomarkers to predict cognitive impairment and AD; and (2) we will build computer biomarker models that can be used by researchers and clinicians to predict who will develop cognitive impairment and AD. We will also determine which model is the best at predicting cognitive impairment and AD.
This project is directly relevant to the Longer Life Foundation’s mission to study factors that help in predicting sickness and death. Currently, 5.3 million U.S. residents live with AD. As the Baby Boomers age, the prevalence of AD in the U.S. is expected to increase to 7.7 million within the next 20 years, and to 13.5 million by 2050.
Through this project, we will understand more about risk factors for AD and how they combine with abnormal biomarker levels to produce dementia symptoms. We will then apply this knowledge to the real world by developing computer programs that can predict for a particular individual his or her risk of later getting AD. We will also approach other funding agencies to test these computer programs in other groups of people who may develop AD, but who may have slightly different characteristics than participants in our study.