Generative Artificial Intelligence, Health Risk Perception Bias, and Sports Participation Among Older Adults —An Empirical Investigation Grounded in the Search-as-Learning Paradigm
Drawing on the search-as-learning paradigm and panel data from 2020–2023, this study empirically demonstrates that generative artificial intelligence significantly promotes sports participation among older adults in China by reducing health risk perception bias, with effects varying by knowledge level, task type, and motivation.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
As the global population ages, the question of how to keep older adults active and healthy has moved from a personal concern to a public priority. In many places, including China, the number of people over sixty is growing rapidly, creating a pressing need for strategies that support "active aging." One of the most effective ways to maintain health in later life is regular physical exercise, yet many older adults hesitate to start or continue due to fear. This fear often stems from a gap between how they feel and what is actually true about their bodies; they may believe they are too frail to move, or that any exertion will lead to injury, even when medical evidence suggests otherwise. This disconnect is known as a bias in how health risks are perceived. At the same time, technology is advancing at a breakneck pace. A new generation of artificial intelligence, capable of holding natural conversations and generating personalized advice, is beginning to change how people search for information. Unlike older search engines that simply list links, these new tools can answer specific questions, clarify doubts, and guide users through complex topics. The central question for researchers has become whether this new technology can do more than just answer questions: can it actually help older adults overcome their fears and start moving?
A team of researchers from Wuhan Sport University and Yunnan Normal University set out to answer this by looking at real-world data from thousands of older adults in China. They analyzed information collected over two years, from 2020 to 2023, using a large national survey that tracks the lives of older people. The study focused on a specific group of smart devices, such as voice-activated home assistants, which serve as a gateway to generative artificial intelligence. By comparing the exercise habits of those who used these devices with those who did not, the researchers could see if the technology made a difference. They also examined the mental processes behind the decision to exercise, specifically looking at how people judge their own health risks compared to their actual physical condition. The goal was to understand not just if the technology worked, but how it worked.
The findings were clear and statistically significant: older adults who used generative artificial intelligence were much more likely to participate in physical activity than those who did not. The data showed that using these tools was associated with a substantial increase in the likelihood of exercising regularly. This result held true even after the researchers accounted for other factors that might influence exercise, such as income, education, where a person lives, and their general health status. To ensure the results were not a fluke, the team tested their conclusions in several different ways. They checked if the findings changed when they looked at different years, when they excluded data from the pandemic period, and even when they used different statistical methods to analyze the numbers. In every case, the positive link between the technology and physical activity remained strong. The researchers also used a specific statistical technique to rule out the possibility that the relationship was caused by something else entirely, confirming that the technology itself was a driving force.
The study went deeper to uncover the mechanism behind this change. It turned out that the technology worked by correcting a specific type of mental error: the bias in how older adults perceive their health risks. Many older people tend to overestimate the dangers of exercise and underestimate their own capabilities. They might feel a minor ache and assume it is a sign of serious decline, or they might believe that their age makes them too fragile for movement. The research showed that generative artificial intelligence helped reduce this pessimistic bias. When older adults used these tools to ask about their health, the technology provided clear, evidence-based answers that helped them see their condition more accurately. It acted as a guide, helping them distinguish between normal aging and actual medical risks. Interestingly, the technology was particularly effective at correcting biases based on how people viewed themselves compared to their own past, rather than how they compared themselves to others. This suggests that the tool was successful in helping individuals align their self-perception with objective reality, removing the psychological barrier that often stops them from exercising.
However, the impact of the technology was not the same for everyone. The researchers found that the benefits depended on three main factors: what the person already knew, what kind of exercise they were interested in, and why they wanted to exercise. For those with a higher level of prior knowledge about health and exercise, the technology was even more effective. These individuals could ask better questions and make better use of the answers they received, turning the information into action more quickly. For those with less experience, the technology still helped, but the process was slower and required more guidance. The type of exercise also mattered. The technology was particularly helpful for activities that required some skill or planning, such as cycling, where it could offer advice on safety and technique. For very simple activities like walking, the technology was less critical because the rules are already well known. Finally, the motivation behind the exercise played a role. The technology was most effective for those whose primary goal was to prevent disease or manage a health condition. For these individuals, the tool provided the precise, actionable information they needed to feel safe and confident. In contrast, for those motivated by a desire for personal growth or self-discovery, the technology was sometimes less effective, as these individuals often prefer to explore and learn on their own without a system dictating the path.
The implications of these findings are significant for how societies support an aging population. The study suggests that simply making technology available is not enough; the tools must be designed to address the specific fears and cognitive biases that older adults face. The researchers propose that future systems should be built to actively correct these misconceptions, perhaps by providing clear, positive feedback based on objective health data rather than just listing risks. They also suggest that these tools should be tailored to different users, offering more guided support for beginners and more detailed, technical advice for those with experience. By integrating these intelligent systems into community centers and health clinics, governments and organizations can help older adults overcome the mental hurdles that keep them inactive. The research demonstrates that when digital tools are used to bridge the gap between perception and reality, they can become powerful allies in promoting a healthier, more active life for older adults. The path forward involves not just adopting new technology, but understanding the human mind well enough to use that technology in a way that truly empowers people to move.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.