U.Va. Psychologist Finds Aging Mental Decline May Not Be as Steep as Thought

September 22, 2010 — Though mental or cognitive abilities appear to generally decline as people age, a new University of Virginia study indicates that the declines may not be as steep as some studies of aging seem to indicate.

Cross-sectional studies, which compare the performance of different age groups, tend to show a steeper decline than do longitudinal studies, which test individuals at intervals over time during the course of their lives.

The criticism though of longitudinal studies is that individuals tend to develop improved skills for test-taking, which may provide an inaccurate view of how their mental abilities are actually changing.

The U.Va. researchers have developed a method to factor out what they call "practice effects" during cognitive tests that apparently skew results after repeat test-taking by individual longitudinal test-takers.

The purpose of both cross-sectional and longitudinal studies is to gain understanding of how the brain changes over lengthy periods of time.

Longitudinal studies retest the same study participants at intervals during the course of their lives, providing a progression of scores that present a big-picture view of how cognitive skills change as people age. For most there is a perceptible, though gradual decline in mental abilities during aging. The decline appears to accelerate after the age of 60.

But researchers know that the "practice effects" of repeat testing makes the study participants familiar with test-taking strategies and thereby better skilled in the act of test-taking – thereby falsely boosting their early scores, at least early on. However, the effect diminishes as time goes on, but because of the early practice-effect inflation, the declines in cognition appear to be steeper.

Such score inflation is the reason for criticism of the longitudinal method and a belief that cross-sectional studies, which compare age groups simultaneously rather than measure what occurs within the same individuals over time, may be a better method for determining whatever declines may be occurring during aging.

"We set out to find a way to separate out those practice effects from what might be the real effects of aging," said Timothy Salthouse, a psychologist in U.Va.'s College of Arts and Sciences, who led the study.

They did that by comparing people who were taking the test for a second time with people of the same age who were taking it for the first time. The difference between those scores, after adjusting for the various comparabilities as to age, gender and other factors, provided an estimate of the level of practice effect.

Salthouse found that, once the separation was made, mental abilities still appear to generally decline with age, but the drops are smaller than cross-sectional studies indicate, because the practice effect inflated the earlier scores.

"I think we're now getting a clearer picture of exactly what's happening with aging independent of some factors that might complicate the interpretations," he said.

The finding was reported in the journal Neuropsychology.

Longitudinal tests measure reasoning, spatial visualization, episodic memory, perceptual speed and vocabulary. These cognitive skills, required for everyday life at any age, are our ability to make rapid comparisons, remember unrelated information and detect relationships between objects, people and events. They are measures of abstract reasoning, brain speed and puzzle-solving.

Study participants are asked to solve various puzzles, remember words and details from stories, and identify patterns in an assortment of letters and symbols.
Many of the participants in Salthouse's study were tested several times during the course of years, allowing researchers to detect subtle declines in cognitive ability.
But the practice effects were found to occur at all ages, a byproduct of repeat testing.

One of the unique features of Salthouse's project in U.Va.'s Cognitive Aging Laboratory is that many of the participants return to the laboratory for repeated assessments after intervals of one to seven years. They range in age from 16 to 98.

"It's important to point out that the trends we see are trends of averages, and that there is a lot of variation around those averages from person to person," Salthouse said. "Most people function at a highly effective level throughout their lives, even into old age."