Physical activity is widely believed to protect the aging brain, but recent evidence suggests the relationship might actually operate in the opposite direction. A new analysis of a large group of German adults found no statistically significant evidence that exercise leads to better cognitive performance or a younger-looking brain, indicating instead that having a healthy brain might make a person more likely to stay physically active. The findings were published in eLife.
The concept of a physically active lifestyle preserving cognitive health rests on a large foundation of prior research. Animal studies have frequently indicated that running and other movements boost the production of brain-fertilizing chemicals, helping neurons survive and form new connections. Based on these mechanisms, human studies often look at brain structures that are vulnerable to aging, such as the hippocampus. The hippocampus is a seahorse-shaped ridge of tissue deep in the brain that manages memory and learning.
Another common way to assess brain health is by calculating a Brain Age Gap Estimate, or BrainAGE. This measurement uses a machine learning algorithm to estimate a person’s age based purely on the physical structure of their brain. Subtracting their chronological age from this estimated age yields a score, where a positive score means the brain appears older than it actually is.
Despite the popularity of the idea that exercise builds a younger and larger brain, some recent large-scale population studies have found contradictory results. For instance, a 2023 study of middle-aged and older adults found that physical activity had almost no effect on cognitive test scores. This discrepancy raised the possibility of reverse causation, the idea that early, unnoticed declines in brain health might cause people to feel fatigued and stop exercising, rather than a lack of exercise causing the brain to decline.
To explore this dynamic, Polona Kalc and colleagues at Jena University Hospital analyzed data from the LIFE-Adult Study, a large project tracking the health of residents in Leipzig, Germany. The research team wanted to see if they could detect a positive effect of physical activity on brain structure and cognitive performance when looking at a broad slice of the general population, rather than relying on a small group of volunteers who signed up specifically for an exercise program.
The researchers evaluated baseline data from 2,576 participants. To measure physical activity, the team relied primarily on a standardized questionnaire that asked participants how much time they spent walking or engaging in moderate and vigorous physical tasks over the previous week. Because people are often inaccurate when recalling their own habits, the researchers also collected objective data from a subset of 227 participants who wore an armband accelerometer for a week to track their actual physical movements.
To measure brain structure, participants underwent magnetic resonance imaging scans, which use strong magnetic fields to create detailed pictures of the inside of the body. The researchers used these scans to calculate overall hippocampal volume and BrainAGE. To measure cognitive function, participants completed the Trail Making Test. In the first part of this test, participants draw a line connecting scattered numbers in chronological order as quickly as possible, providing a measure of basic mental processing speed. In the second part, they alternate between connecting numbers and letters, which tests executive function, a category of higher-level mental skills that includes cognitive flexibility and working memory.
A smaller subset of the original group returned for a follow-up visit several years later. About 340 participants completed the cognitive tests again, and about 250 provided a second magnetic resonance imaging scan. During this follow-up, participants also answered questions about their physical exercise capacity, estimating how much effort they could exert before becoming fatigued. In all of their statistical models, the researchers adjusted the data to account for the participants’ age, smoking history, alcohol intake, and socioeconomic status.
When analyzing the baseline data, the research team found no statistically significant positive effects of total or moderate-to-vigorous self-reported physical activity on hippocampal volume or BrainAGE. In fact, higher self-reported physical activity corresponded with slightly worse scores on the executive function test.
The objective accelerometer data yielded a similar outcome. The researchers found no statistically significant relationship between accelerometer-measured movement and the size of the hippocampus or the BrainAGE score in the general adult sample.
The analysis did reveal a notable age-related bias in how people report their own exercise habits. Older adults, particularly those over age 60, tended to overestimate how much physical activity they engaged in compared to what the armbands actually recorded. When the researchers isolated a small group of 70 participants between the ages of 30 and 60, they did find that higher accelerometer-measured activity was associated with a younger-appearing brain. This effect did not appear in adults over 60, who made up the majority of the sample.
The longitudinal data, which tracked participants over time, provided evidence for reverse causation. The amount of physical activity a person reported at the beginning of the study did not predict whether their brain would look younger or older years later. Instead, the participants who had an older BrainAGE at the start of the study tended to report a lower physical capacity at the follow-up visit. The researchers suggest that underlying age-related issues, such as cellular energy problems or low-grade inflammation, might silently diminish a person’s brain health and simultaneously cause the physical fatigue that makes them less active.
“The most important takeaway is that this study did not find consistent evidence that people who reported being more physically active had better brain structure or cognitive function, at least for the specific outcomes examined,” Ben Singh, a Research Fellow in Health at Adelaide University who was not involved in the research, told PsyPost. “That is important because physical activity and brain health are often discussed as though the causal direction is straightforward. This study suggests the relationship is likely more complicated and potentially bidirectional.”
These results stand in tension with several previous findings. For example, the findings contrast with research covered by PsyPost in 2025, which indicated that structured exercise interventions consistently boost general cognition and executive function. That prior review synthesized controlled exercise programs, whereas the current study assessed habitual, everyday physical activity in an observational cohort.
The new results also stand in tension with a 2025 study linking objectively measured exercise to larger hippocampal gray matter. That previous research focused specifically on isolated subregions of the hippocampus in young adults, which might behave differently than overall hippocampal volume across a broader, older population.
“I would describe it as challenging part of the existing observational literature rather than overturning the broader evidence for physical activity,” Singh noted. “Importantly, this study examines habitual physical activity in an observational cohort. That is not the same question as whether a structured exercise intervention can improve particular aspects of cognition or brain health. Randomized exercise trials remain especially important when we want to determine whether exercise itself is causing a change.”
As with all research, there are a few things to keep in mind when interpreting the results. The reliance on self-reported questionnaires for the bulk of the sample introduced measurement errors, as demonstrated by the older participants overestimating their activity levels. The objective accelerometer data bypassed this issue, but the subset of people who wore the devices was relatively small and generally exhibited low activity levels. It is possible that the participants simply did not exercise intensely or long enough to trigger the biological mechanisms that improve brain structure.
“The main thing I would caution against is interpreting this as ‘exercise does not benefit the brain.’ That would go well beyond what the study can establish,” Singh said. “It is also important to distinguish between no statistically detectable association and proof of no effect.”
The study also focused exclusively on hippocampal volume and BrainAGE. Physical activity might still benefit the brain in ways these specific measurements cannot capture, such as by improving the integrity of the brain’s white matter tracts or enhancing the functional connectivity between different brain regions. Additionally, the longitudinal analysis compared physical activity at baseline to physical capacity at follow-up. Because these two surveys measured slightly different concepts, tracking exact changes over time is difficult.
“The next step I would most like to see is a large, well-powered longitudinal study beginning in midlife that repeatedly measures physical activity objectively and repeatedly assesses both brain structure and cognitive function over time,” Singh explained. “Rather than asking simply whether ‘more activity equals a healthier brain,’ the more useful question may ultimately be what type of activity, how much, at what age, and for whom?”
The study, “The effect of physical activity on brain structure and cognitive function in the population-based cohort of LIFE-Adult Study,” was authored by Polona Kalc, Robert Dahnke, Christian Sanders, Frauke Beyer, Andrea Zülke, Steffi Riedel-Heller, A Veronica Witte, and Christian Gaser.
Leave a comment
You must be logged in to post a comment.