A new study published in Psychological Science finds that not all aspects of intelligence are equally tied to how long people live. While it has long been known that smarter individuals tend to live longer, researchers using data from the Berlin Aging Study found that one specific ability—verbal fluency—stood out. People who could quickly name animals or words beginning with a specific letter tended to live significantly longer than those who struggled with these tasks. In fact, higher scores on these tasks predicted nearly a nine-year difference in median survival time.
The study was designed to address a longstanding question in aging research: Do different types of cognitive abilities relate differently to how long a person lives? Past studies showed that general intelligence is linked to survival, but they often relied on a single snapshot of cognitive performance, rather than tracking changes over time. Researchers have also debated whether it is a person’s average level of ability or their rate of cognitive decline that better predicts longevity.
To better understand these relationships, the authors behind the new study analyzed long-term data from very old adults, aiming to compare various mental abilities side by side while also tracking how they changed over time.
“The association between cognitive performance and mortality has long been known in the scientific literature. Lately, evidence has also accumulated in favor of an association between change in cognitive performance (as opposed to merely one-time estimation of general performance) and survival,” said study author Paolo Ghisletta, a full professor of psychology at the University of Geneva.
“This latter question requires longitudinal data and mortality status follow-up updates, both of which were rare half a century ago, but have become more easily available in recent years. My colleagues at the Max Planck Institute for Human Development in Berlin, Germany, have very rich data on cognitive performance assessed over up to 18 years, as many as 14 times, available for a sample of individuals, initially aged between 70 and 104 years.”
“Moreover, they obtained reliable mortality status on all participants. Lastly, a sophisticated statistical model, that simultaneously analyzes how people change on cognitive performance and how both their level of performance and change therein relate to mortality, has become robust and easy to use. We simply had to put all the pieces together to arrive at our final results.”
Ghisletta and his colleagues used data from 516 adults who participated in the Berlin Aging Study. These individuals were between 70 and 103 years old at the start of the study, with an average age of about 85. All participants had passed away by the time of analysis, allowing researchers to precisely examine the link between their cognitive performance and the timing of their deaths.
The participants completed nine different mental tasks that assessed four broad abilities: perceptual speed, episodic memory, verbal fluency, and verbal knowledge. Additionally, researchers calculated a general intelligence score based on performance across all tasks. The team used a sophisticated method called a joint multivariate longitudinal survival model, which allowed them to simultaneously examine how each person’s cognitive abilities changed over time and how those changes were related to the person’s survival. This method is especially powerful because it takes into account that people who decline faster may be more likely to die sooner, which can skew results if not properly addressed.
The most striking result was that only verbal fluency—specifically, the ability to quickly list animals and words beginning with the letter “s”—predicted how long participants lived. People who performed better on these tasks tended to live longer, regardless of how they performed on other tasks like memory or vocabulary. The effect was sizable. On average, participants who could name many animals or words lived up to nine years longer than those who struggled with these tasks. To put it simply, naming more animals in 90 seconds was associated with living longer.
“At the population level, intelligence predicts survival, in that people capable of greater performance survive longer,” Ghisletta told PsyPost. “Moreover, we all naturally decline, at different rates, in cognition. The rate of this decline also predicts survival, in that those who decline more and at a faster pace survive less. Verbal fluency, the capacity to generate words from well-learned categories appears to hold a prominent association with survival among different cognitive abilities.”
“However, although tempting, it is ill-advised to infer from a population-level finding to a single individual, because we do not know how such an individual would fare with respect to the population on which the finding was based. It is thus erroneous to conclude that ‘Ah, you cannot recall another type of fruit, so you must be on your terminal slope!’”
Importantly, it wasn’t just general intelligence or other types of cognitive ability that mattered. When all nine tasks and the general intelligence score were considered together in the same model, only the fluency tasks added unique predictive value. This suggests that verbal fluency is not just a proxy for being generally smart. It may reflect something distinct about brain health or how cognitive systems are aging.
“I expected that another cognitive task would relate as strongly as did verbal fluency, namely perceptual speed,” Ghisletta explained. “Such tasks require scanning and comparing simple stimuli, such as figures, letters, and numbers, to match them according to some template. Typically, most individuals manage these tasks, but do so at different speed. These tasks are timed, so that in the end, a successful performance consists of many, correct matches within a very short period (60 to 90 seconds).”
“Much research concluded that such tasks are among the most sensitive in detecting differences among people in their cognitive decline, so I expected that this greater sensitivity, compared to other cognitive tasks, would also relate to shorter survival. Yet, perceptual speed tasks were inferior to verbal fluency tasks in that regard.”
There are several possible reasons why verbal fluency might be such a strong predictor of survival. Unlike many intelligence tests that measure a single skill, verbal fluency draws on multiple brain functions at once. It involves memory, vocabulary, attention, and the ability to quickly switch between mental categories. This makes it especially sensitive to changes in brain health. Previous studies have shown that verbal fluency is often one of the first abilities to decline in conditions like dementia or Parkinson’s disease. The task also relies heavily on the prefrontal cortex, a brain area that is especially vulnerable to aging and disease.
Another reason verbal fluency may outperform other cognitive measures is because of the way it links to both fluid and crystallized abilities. Fluid intelligence refers to the ability to think quickly and solve new problems, while crystallized intelligence reflects accumulated knowledge. Verbal fluency appears to tap into both. This dual nature may make it especially informative when evaluating overall brain health in older adults.
The study also underscores the benefits of using joint statistical models over traditional two-stage approaches. In older research, cognitive decline and survival were often analyzed in separate steps, which can introduce bias. The joint approach used here allowed the researchers to simultaneously model how people’s abilities changed over time and how those changes related to their risk of death, offering a more accurate picture.
There were some limitations to the study. The sample was not fully representative of the general German population, as it was intentionally stratified by age and sex, and focused on very old adults. All participants were from Berlin, and most were born between 1887 and 1922. It remains unclear whether similar results would be found in younger people or in more diverse populations.
“Any study has to admit caveats, simply because it is impossible to test every form of an association and further hope that this remains stable across time, in different cultures, and varying social contexts,” Ghisletta noted. “Although our sample has been conceived with the greatest care at study inception (in 1989), it has particular features, which surely limit our capacity to generalize to any group of 70-plus year-old individuals.”
“More importantly, the study lacked solid evidence about mechanisms that may relate to both decline in cognition and shortened survival. Multiple scientific hypotheses have been phrased, some relating to biological indicators of senescence, others to neuroscientific evidence cerebral structures and functioning, all of which were not available in this study.”
Nonetheless, the findings provide evidence that verbal fluency may offer a window into a person’s overall cognitive and biological resilience. Future research could examine whether interventions that improve or maintain verbal fluency in older adults might also impact health or longevity. The researchers also hope their methods can be applied to other datasets to confirm whether verbal fluency consistently outperforms other measures of cognitive ability in predicting how long people live.
“Ideally, studies as ours are followed up by other, more powerful studies capable of investigating more directly possible mechanisms relating intelligence to longevity,” Ghisletta said.
“Studies as ours are possible thanks to collaborative efforts by many specialists from different fields who are motivated to share their knowledge and experience. It would be impossible to carry out such a large and long study without a solid scientific infrastructure responsible for collecting, organizing, and hosting the data and colleagues who complement each other’s expertise and perspectives.”
The study, “Verbal Fluency Selectively Predicts Survival in Old and Very Old Age,” was authored by Paolo Ghisletta, Stephen Aichele, Denis Gerstorf, Angela Carollo, and Ulman Lindenberger.
Leave a comment
You must be logged in to post a comment.