The reason why we lose the ability to distinguish faces as we age might be because 'our eye movements become less stable.'

It is known that the ability to distinguish faces declines with age, but the mechanism behind this decline is not fully understood. A research team from the Hong Kong University of Science and Technology and Hong Kong Polytechnic University, among others, has pointed out from facial recognition experiments that instability in how the eyes move when viewing faces may be one of the causes of the age-related decline in facial recognition ability.

The Role of Eye Movement Consistency in Aging-Related Decline in Face Recognition | npj Science of Learning
https://www.nature.com/articles/s41539-026-00437-3

Face recognition declines with age may lie in how your eyes move
https://www.news-medical.net/news/20260714/Face-recognition-declines-with-age-may-lie-in-how-your-eyes-move.aspx

As people grow, they learn how to move their eyes when looking at faces. In adulthood, although there are individual differences such as 'people who focus on the eyes' or 'people who focus on the nose,' each person's eye movements remain relatively stable even when looking at different faces. On the other hand, older people have lower accuracy in facial recognition than younger people, and their gaze moves more frequently to different parts of the face, but it was not known whether the way they move their eyes becomes unstable with age.

Therefore, a research team led by Janet Shao of the Hong Kong University of Science and Technology had 301 Asian adults aged 40 to 81 participate in a task in which they had to 'memorize the faces of 20 people they had never met before, and then distinguish between the 20 memorized faces and 20 new faces.' In addition to eye movements during the task, they also measured attention and the ability to control behavior, and investigated whether age-related changes in abilities could explain the differences in face recognition performance and eye movements.

To model where the gaze lingers on a face and where it moves next, the research team analyzed gaze data using a ' hidden Markov model ' that analyzes data that changes over time. To investigate the initial movements, which are considered important for face recognition, they focused on how well they could predict the location to be looked at for the third time based on the location where the gaze stopped for the second time. The more difficult it is to predict the location to be looked at for the third time, the more varied the gaze paths are for each face, and the less consistent the way the eyes move is judged to be.

When the research team analyzed the experimental results, they found that older people performed worse at recognizing faces and were more likely to mistakenly identify new faces as 'previously seen.' Furthermore, older people had more difficulty predicting eye movements from the second to the third time they saw a face, and their eye movements were more unstable. The research team stated that 'cognitive abilities that change with age, such as attention, cannot explain the decline in face recognition performance and the variability in eye movements.'

The research team also reported that people whose eye movements varied more from the second to the third time performed worse in face recognition. In addition, they confirmed an indirect relationship: older people tended to have more varied eye movements, and the more varied their eye movements, the worse their face recognition performance. However, even after accounting for the effects of eye movement variability, the relationship between age and face recognition performance remained, so the research team explained that 'eye movement alone cannot fully explain the decline due to aging.'

Furthermore, it was confirmed that those who struggled with the task of 'ignoring surrounding shapes and reacting only to the central shape' had more unstable eye movements. According to the research team, rather than older people forgetting how to look at faces as they did when they were younger, it is possible that their ability to focus attention on necessary areas and ignore surrounding information has declined, making it more difficult to move their eyes in a fixed order. They also suggest that the difficulty in accurately moving the eyes to a target location due to aging may be one of the causes of the variability in eye movements.

However, this study is a cross-sectional study comparing people of different ages at the same point in time, and does not prove a causal relationship between unstable eye movements and impaired facial recognition ability. It is also important to note that the participants and their facial photographs were limited to people of Asian descent, and long-term memory was not measured.

The research team states that in addition to long-term studies tracking the same individuals, it will be necessary to investigate whether training eye movements and attention can mitigate the decline in facial recognition abilities in older adults.