Verbal fluency in old age is an indicator of higher life expectancy

Longitudinal data of the Berlin Aging Study show that verbal fluency is a strong predictor

Older people who can easily list animals or different words beginning with the same letter are more likely to live longer. In a new analysis of the Berlin Aging Study an apparently simple task proves to be a particularly predictive measure of life expectancy in old age.

The findings published in the journal Psychological Science are based on longitudinal data of the Berlin Aging Study. Headed by Paolo Ghisletta (University of Geneva), an international research team–including researchers from the Center of Lifespan Psychology at the Max Planck Institute for Human Development–examined the cognitive development of 258 women and 258 men who were aged between 70 and 103 years at the beginning of the study. The sample was drawn randomly from the city registry and can therefore be considered representative. The participants’ cognitive abilities were regularly assessed. Twenty-two people took part in the final assessment that took place 18 years after the beginning of the study. At the time of the data analysis all of the participants were deceased. The dates of their death are known, but generally not the causes.

516 participants, nine tests, including two on verbal fluency

Nine cognitive tasks were at the heart of the study. Beyond two memory tasks, two of verbal knowledge, and three of perceptual speed, there were two tests on verbal fluency. In one of the fluency tasks, the participants were asked to name as many different animals as possible within 90 seconds. In the other, they were prompted to list as many different words beginning with “S” as possible within 90 seconds.

The two verbal fluency tasks predicted the number of remaining years of life much better than the other seven tests did. “People whose verbal fluency was within the upper 25 percent of the sample lived almost nine years longer than those whose fluency was within the lower 25 percent,” Ghisletta explains. Specifically, the risk of death was reduced by five percent with each additionally named animal and by three percent with each additional word beginning with “S.” Variables such as age, gender, and educational level are also related to life expectancy. However, verbal fluency showed the strongest correlation, even when all these variables were considered simultaneously.

Why is verbal fluency such a powerful predictor?

Fluency tasks require the use of several cognitive abilities. First is the goal-directed access to long-term memory, in this case to animals one knows or to words beginning with the letter “S.” Second, access needs to be quick because as many words as possible are to be listed within 90 seconds. And third, one needs to remember the words one has already named as only different words are considered correct responses. Therefore, the predictive validity did not come as a surprise to the researchers. “Verbal fluency is a particularly complex cognitive ability that requires the smooth interplay of different brain functions,” explains Ulman Lindenberger, Director at the Max Planck Institute for Human Development. “If this interplay declines strongly in the course of aging, it is associated with a loss of life expectancy.” In extreme cases of age-related dementia, this can lead to the patient repeating only two words, “dog, cat,” Lindenberger says.

Particularly precise statistical method

One of the study’s strengths consists in the applied statistical method. Examining the research question at hand requires two submodels. The first refers to the estimation of individual differences in performance levels and their changes, the second to the estimation of individual differences in life expectancy. To date, these submodels have usually been estimated separately in a first step and then combined in a second step. In contrast, the authors applied a unified estimation process, referred to as „joint multivariate longitudinal and survival model” (JMLSM) in the scientific literature. With the JMLSM, performance levels, performance changes as well as the survival probabilities are all simultaneously estimated in one step. The authors show that this unified method is more precise and less biased than the two-step method.

Statements of probability need to be handled with care

In conclusion, the authors emphasize that statistical statements about the association between test performance and life expectancy cannot be directly transferred to individuals. “The association we found only expresses a probability. It is not possible to draw specific conclusions about the remaining lifetime of an individual person from such test results,” says Ghisletta. Co-author Lindenberger considers the result an important starting point for further work. “Research into the strong correlation between word fluency and life expectancy will help us better understand the causes of cognitive performance maintenance and decline in old age.”

Lindenberger is not particularly convinced by the idea of training animals to name things in order to live longer. “Then the test would no longer measure what it measured before, and life expectancy would probably not change much.” However, he says there is good evidence that an intellectually stimulating, socially engaged, and physically active life is linked to the maintenance of cognitive performance and a longer life expectancy. So a walk with friends at the zoo is definitely recommended.

About the Berlin Aging Study:

The Berlin Aging Study is a multidisciplinary investigation of old people aged 70 to over 100 years who lived in former West Berlin. In the main study (1990–1993), a core sample of 516 individuals was closely examined in 14 sessions covering their mental and physical health, their psychological functioning, and their social and economic situation. Since then, the study has been continued as a longitudinal study, and surviving participants have been reexamined seven times.

In brief

• A new study: Older people’s fluency is a better predictor of life expectancy than other measures of cognitive functioning.

• Life expectancy in old age: Research has established a link between longevity and cognitive performance. Numerous tests are used to measure this.

• 516 participants: The analysis is based on data of the Berlin Aging Study (BASE) with participants who were at least 70 years old.

Original publication: