By SRINITYA ARASANIPALAI
Memory loss is a normal part of the aging process, but memory loss is also one of the main symptoms used by clinicians to determine if an individual has dementia. Prof. Charles Brainerd, human development, Prof. Valerie Reyna, human development, and Carlos Gomes grad have created a mathematical model that can help discern normal memory loss from memory loss due to a neurological disease.
Memory loss, whether or not due to disease, is characterized by episodic memory loss. Episodic memory loss is when a person cannot remember specific things that happen at a particular time at a particular location.
“It’s remembering what a hot dog is but not remembering that you ate a hot dog in a baseball game last Saturday evening,” Brainerd said.
A specific component of episodic memory loss is called recollection in which the brain uses a particular method to remember the when and where components of a memory. Recollection is characterized by vivid memories of what happened which one can see in their mind’s eye and hear in their mind’s ear. According to Brainerd, recollection is what is expected to decline in a healthy normal person in their late middle age.
Contrary to popular belief, this decline does not hold steady throughout adulthood. It has been found that steady decline in memory loss starts at late middle age, declines until about the age of 70 and then stays constant as long as the individual stays healthy, Brainerd said.
Brainerd focuses on differentiating healthy elderly people with normal memory loss from people who have memory loss due to neurological diseases. Within the segment who have neurological diseases, Brainerd also seeks to separate individuals with diseases such as dementia from people who have cognitive impairment, an intermediate stage between age-related memory loss and a neurological disease.
According to Brainerd, people with neurological diseases will show a different pattern of memory loss as compared to individuals with normal memory loss.
Current clinical tests used to detect episodic memory loss in individuals include recall and recognition tests, Brainerd said. In a recall test, a person is given a list of words to study and asked to recall as many words as they can. In a recognition test, one is given a list of words and then given a second list of test words. The person is then asked to identify which of the test words were on the first list and which were not.
According to Brainerd, these tests are not consistent and do not rise to the gold standard of sensitivity and specificity. Sensitivity is being able to positively identify a person in the disease category while specificity is being able to reject a person as not belonging to the disease category. Eighty-five percent of specificity and sensitivity are considered gold standards.
When individuals remember things or words in recall and recognition tests, different, independent component memory processes are involved. Different individuals can also use different processes to remember the same word. When an individual takes the currently used recall or recognition tests, however, they are just repeating words with nothing to pinpoint the processes used, according to Brainerd.
Brainerd, Reyna and Gomes created a theory-driven mathematical model that measures the separate memory processes used on the recall and recognition tests instead of just raw performance.The model seeks to detect error patterns by analyzing the performance on these tests, which results in a higher level of both sensitivity and specificity.
By using this new test, researchers found that individuals with dementia or cognitive impairment experience memory loss during recognition processes. This differs from normal memory loss which loses recall of words or events instead of recognition, Brainerd said.
In addition to the diagnostic component, Brainerd also focuses on trying to predict which healthy people will be likely to develop cognitive impairment or dementia.
“It is really important to do this because treatments for dementia and cognitive impairment are not very good,” Brainerd said. “In order to develop treatments to prevent neurological diseases, we have to be able to identify people who are still healthy but will, in the future, develop these conditions so as to test the current treatments.”
According to Brainerd, over a period of one to six years, increased decline in recognition memory indicates a transition in healthy individuals from normal memory loss to cognitive impairment or dementia.
These mathematical results are significant because they are better predictors than genetic markers in identifying neurological diseases, Brainerd said. Generally, genetic markers are only good in detecting diseases during their later stages. According to Brainerd, the Apolipoprotein E genotype is the best known genetic marker in predicting future cognitive impairment or dementia. However, it is only 50 percent accurate.
“We are at a tipping point where we are really at the ground floor of things with so much fundamental work and advancement to be made at this point,” Brainerd said. “No matter how many good ideas we have about treatments, we cannot cure the diseases if we cannot identify [them].”