unsplash.com
Research on aging has reached a major milestone. A study published in the prestigious journal Psychological Science in 2025 reveals that our ability to use words is the best cognitive indicator of our longevity. Led by Paolo Ghisletta, a psychologist at the University of Geneva, this research shows that verbal agility outperforms memory or cognitive speed in predicting life expectancy at old age.
Human aging is often viewed through the lens of physical health. However, research by Paolo Ghisletta, a psychologist at the University of Geneva and a member of the LIVES Centre, suggests that our mental agility is an equally powerful indicator of our vitality. To reach these conclusions, the researchers analyzed data from the Berlin Aging Study (BASE), launched in 1989, just before the fall of the Berlin Wall. This study followed 516 people aged 70 to over 100 until their deaths.
The study examined nine different cognitive abilities, ranging from short-term memory to processing speed. The finding was clear: “Only verbal fluency can uniquely predict the time of death,” emphasizes Paolo Ghisletta. Participants who scored highest on verbal fluency tests had a median survival rate nine years longer than those with weaker performance. While the link between intelligence and survival was already known, the Geneva-based researcher’s study provides a significant clarification: once a person’s verbal fluency is known, the other tests (such as memory or perceptual speed) add no additional predictive information. Verbal fluency thus appears to “sum up” on its own the state of health of the nervous system affecting cognitive performance.
In practical terms, verbal fluency is measured using two exercises that seem simple but are quite challenging for our neurons. The first involves naming as many animals as possible in 90 seconds. The second involves listing as many words as possible that begin with the letter “S.” These two tasks engage very different mechanisms. Naming animals involves semantic fluency: the brain navigates through logical categories (farm animals, zoo animals, forest animals) to organize its search. Conversely, finding words starting with “S” is a task of phonemic fluency that is just as demanding. It forces the brain to break its language habits to focus solely on the sound, an exercise that recruits specific frontal brain areas highly vulnerable to decline. To succeed at both tasks, the brain must search its long-term memory while remaining extremely vigilant: “You have to think fast, remember the words you’ve already named, and avoid repeating yourself,” explains Paolo Ghisletta. It is this combination of speed, search strategy, and control that reflects the integrity of brain health.
Unique predictive power
The originality of this study lies in the use of “multivariate joint models,” statistical tools capable of handling data of rare complexity. This method achieves a dual feat: it simultaneously analyzes the evolution of mental abilities as well as the risk and timing of death (taking into account the gradual loss of participants over the years), while integrating the nine cognitive tasks together within a single statistical model.
Previously, researchers often studied one or two abilities in isolation. Paolo Ghisletta’s approach is a game-changer: “It’s a holistic view of cognitive abilities... they are tested against one another simultaneously within the same model to see which one stands out,” he explains. This statistical competition is crucial because the various faculties of intelligence are typically correlated with one another, creating significant redundancy. By pitting them against one another in this way, the researcher was able to demonstrate that only verbal fluency possesses unique predictive power.
This method of analysis also helps avoid a major statistical pitfall. In traditional analyses, the gradual loss of the most fragile participants creates a bias: ultimately, only the “healthy” survivors are observed, which falsifies the results. By analyzing cognitive decline and mortality as a single, interconnected process, researchers obtain an “error-free” picture of the cognitive trajectory. These cutting-edge analyses were unfeasible just a few years ago: “There wasn’t statistical software capable of handling this,” the researcher points out.
Hasty interpretation
These models confirm that psychological indicators are sometimes more revealing than traditional medical markers. In previous research, Paolo Ghisletta had already noted that “self-reported” health (feeling fit) or mental decline were predictors of mortality “stronger than the number of cigarettes smoked or the amount of medication taken.”
However, the researcher cautions against hasty interpretation in everyday life. “This is always an observation made about a group of people, and we can’t necessarily draw conclusions about individuals,” he cautions. “A veterinarian won’t necessarily live longer just because they know dozens of animals, as I’ve heard before; the study describes an overall statistical trend, not a fixed destiny.”
A tool for the medicine of tomorrow
The impact of this study is primarily clinical. While these 90-second tests are already used intermittently to assess the potential consequences of concussions or strokes, Paolo Ghisletta’s work “innovatively solidifies” their systematic use in geriatrics. They could become essential routine tools for the early detection of generalized decline in the nervous system.
In the future, the researcher hopes to compare this data with biological markers such as chronic diseases or lifestyle factors to further refine this compass for longevity. After all, our ability to find the right words must also depend on biological factors that determine our inner vitality.
By Kalina Anguelova
Ghisletta, P., Aichele, S., Gerstorf, D., Carollo, A., & Lindenberger, U. (2025). Verbal Fluency Selectively Predicts Survival in Old and Very Old Age. Psychological Science, 36(2), 87-101.
https://doi.org/10.1177/09567976241311923