Older people may retain visual strategies learned early in life, but less consistent eye movements may make it difficult to execute those routines, which helps explain why recognizing familiar faces becomes harder with age.
Study: The role of eye movement consistency in age-related decline in face recognition. Image credit: CGN089/Shutterstock.com
Recent research published in journals npj science of learning Researchers found that increasing age is associated with less consistent eye movements when viewing faces, which in turn is associated with age-related declines in face recognition performance.
How does facial recognition change with age?
As the world’s population ages rapidly, understanding cognitive aging has become an important public health priority, as age-related cognitive decline can reduce the quality of life of older adults. One important aspect of this decline is impairment in facial recognition, which can make everyday social interactions more difficult and therefore an important target for intervention.
Facial recognition develops gradually over the lifespan. During childhood, individuals learn a consistent visual routine for scanning faces, but by early adulthood, more individualized or idiosyncratic eye movement patterns emerge. In adults, a more eye-focused scanning pattern is associated with better face recognition, the authors suggest, reflecting the higher diagnostic value of the information contained in the eye region. However, this relationship is not observed in children. This may be because children have not yet learned how to efficiently extract information from this part of the face.
Once these visual routines are well established in young adulthood, differences in the consistency of eye movements become less informative in predicting individual differences in face recognition, whereas individual scanning patterns become more important. In contrast, previous studies have shown that older adults generally perform worse on face recognition tasks, with lower identification accuracy and slower response times.
Previous research has also suggested that older adults tend to make more transitions between facial features and adopt a scanning strategy that focuses less on the eyes than younger adults. However, it remains unclear whether aging is also associated with a decline in eye movement coherence and whether this contributes to declines in face recognition independent of age-related cognitive decline.
Research investigating cognition, eye movements, and recognition
Cognitive abilities such as working memory, executive function, and visual attention are clearly related to the pattern and consistency of eye movements. Age-related declines in these cognitive abilities may underlie the increased eye movement discrepancies and variability in eye movement patterns that affect face recognition. Conversely, other studies suggest that these parameters themselves contribute to poor face recognition performance, independent of cognitive impairment.
The authors therefore used the Dalhousie Computer Attention Battery (DalCAB) to test cognitive abilities along with eye tracking and face recognition. A cross-sectional study using Hidden Eye Markov Model (EMHMM) among 301 Asian adults aged 40-81 years. The objective was to assess the relative roles of each of these components.
In previous studies, higher eye movement entropy, as measured using electroencephalography (EEG), was associated with poorer face recognition and less efficient neural processing of faces in both children and adults. Based on these findings, the authors hypothesized that increasing age might be associated with poorer face recognition, poorer eye-focused scan patterns, and less consistent eye movements.
To measure this consistency, they assessed both global eye movement entropy, which reflects the overall variation in eye movements, and C3 entropy, which measures how consistently the eyes move from second to third fixation when looking at a face.
Main findings
Aging is associated with decline in facial recognition
The researchers first used the Dalhousie Computerized Attention Battery (DalCAB) to examine how aging affects cognitive performance. Older age is consistently associated with declines in reaction times across all cognitive tasks, indicating general psychomotor slowing, as well as changes in inhibitory control and verbal working memory. This finding also suggested that spatial working memory and spatial orientation may be impaired, consistent with previous research.
These age-related changes were reflected in face recognition performance. Older participants discriminated poorly between previously seen and new faces, produced more false recognitions, and took longer to respond. However, even after accounting for measured cognitive ability, age was still associated with poorer discrimination ability and increased false alarm rates, suggesting that cognitive decline alone cannot fully explain poorer face recognition performance.
As we get older, our eye movements become less consistent.
The researchers then investigated whether aging was associated with changes in the way participants visually scanned faces. Contrary to their hypothesis, age was not associated with changes in eye movement patterns, such as whether participants focused on the eyes or nose. Rather, aging is associated with a decrease in the consistency of eye movements, meaning that participants follow an unpredictable visual scanning routine when observing faces.
Considering the decline in cognitive function associated with aging, C3 entropy, a measure of the predictability of third eye fixations based on second fixations, remained significantly associated with age. Furthermore, exploratory analyzes suggested that decreased eye movement coherence was associated with decreased selective attention and inhibitory control, suggesting that age-related declines in these cognitive processes may make it more difficult to consistently execute established visual routines.
Predicting facial recognition from eye movement behavior
Although aging was not associated with changes in global eye movement patterns in this cohort, participants who adopted more eye-focused scanning patterns achieved better face recognition performance and produced fewer false recognitions. Similarly, more consistent eye movements result in higher recognition accuracy, fewer false alarms, and faster recognition responses.
Further analysis suggested that decreased coherence of eye movements may partially explain the relationship between aging and decreased face recognition. Specifically, decreased consistency significantly affected the association between age and false alarm rate, but its effect on discrimination sensitivity was weak after correction for multiple comparisons.
Hierarchical regression analysis showed that a more nasally focused scanning pattern and lower eye movement consistency explained additional variance in poor face recognition performance, beyond the effects of age and measured cognitive ability. These factors also contribute to further variation in false alarm rates.
Taken together, these findings suggest that eye movement consistency plays a variety of roles across the lifespan. In childhood, increased coherence supports the development of face recognition, but in adolescence, highly personalized scanning routines make coherence less useful as a predictor of recognition ability. However, in later adulthood, eye movement consistency appears to become important again, alongside scanning patterns.
This finding suggests that older adults do not completely forget the eye movement patterns and visual routines they learned early in life. Rather, reduced selective attention and inhibitory control may make these routines more difficult to perform, leading to less consistent eye movements.
Building on previous computational modeling studies, the authors suggest that the increased eye movement entropy observed in older adults may reflect increased fixation noise associated with decreased sensorimotor function, which may further impair face recognition.
Study results may not be generalizable to all populations
This study had several limitations that should be considered when interpreting the results. Only young adult faces were used in the face recognition task. That is, it remains unclear whether the same relationship would be observed when older adults recognize faces closer to their own age. Additionally, this study only included Asian participants viewing Asian faces. Previous research has shown that eye movement patterns during face recognition vary across cultures, so this result may not be generalizable to other populations.
The study was also cross-sectional, meaning that it is not possible to determine whether decreased coherence of eye movements causes decreased face recognition or simply occurs in parallel with decreased face recognition. Additionally, participants ranged in age from 40 to 81 years, although most were between 50 and 70 years, which may limit the generalizability of the study findings across the broader adult lifespan. Finally, the cognitive assessment primarily focused on attention and executive function, and other potentially relevant abilities such as long-term memory were not tested.
Consistency of eye movements may help maintain facial recognition
The findings suggest that age-related declines in facial recognition are associated not only with changes in attention and executive function, but also with less coherence in the visual routines with which people scan faces. Although older adults appear to retain distinct eye movement patterns formed early in life, age-related declines in selective attention and inhibitory control may make it more difficult to consistently perform these routines, contributing to poorer cognitive performance.
Although the cross-sectional design means that causal relationships cannot be established, the results highlight eye movement consistency as a potential target for future interventions. Longitudinal studies will be needed to determine how these visual routines change with age and whether training designed to enhance eye movement coherence and attentional control can help maintain facial recognition later in life.
Download your PDF copy now.
Reference magazines:
-
Zheng, Y., Lo, TWS, Lau, EYY, et al. (2026). The role of eye movement consistency in age-related decline in face recognition. Npj Science of Learning. Toi: https://doi.org/10.1038/s41539-026-00437-3. https://www.nature.com/articles/s41539-026-00437-3
