Genes influence infants' preference to look at faces

A new study shows that whether 5-month-old infants look at faces or non-social objects like cars or mobile phones is heritable. This indicates that there is a genetic basis for how babies curate their unique learning experiences.
Genes influence infants' preference to look at faces


A 5-month-old baby looking at their mother’s face.
A 5-month-old baby looking at their mother’s face. 

The way babies discover the world with their eyes shapes what they learn from it and their development. Faces are known to attract babies’ attention uniquely, already at birth. It is quite a powerful feeling and heartwarming experience to witness a baby locking their gaze on our face, with their big and bright eyes in admiration. Naturally, babies tend to look at faces with passionate interest; they are a rich source of learning for their social and emotional development. Their look also serves as a strong signal to other people and can, for example, foster interaction with their parents and caregivers. However, the exploration of non-face objects is equally valuable in their learning journey, so a baby needs to adaptively prioritize visual focus.   

Looking preferences are more similar in identical twins.

While babies differ in their looking preferences, the factors contributing to this variation remain unclear. In this research, conducted at Uppsala University and Karolinska Institutet, we recorded face preference from more than 500 twins at 5 months of age using eye tracking technology. We showed that an infant’s preference for faces was largely explained by their genetic differences. In contrast, the family environment, that is, the environmental characteristics that siblings are exposed to equally (for example, the number of family members in the home) did not explain preferences for social versus non-social information at this early age. This suggests that, already before infants can express their preferences by pointing or walking, they curate their own learning experiences by systematically looking more at social or non-social objects, based partly on their unique genetic predispositions.

Illustration of the testing session and visual stimuli. An eye tracker underneath the screen recorded where on the screen the infant looked at a moment-to-moment basis while they watched several images containing faces together with several other objects – a car, a bird, a phone, and a scrambled face.
Illustration of the testing session and visual stimuli. An eye tracker underneath the screen recorded where on the screen the infant looked at on a moment-to-moment basis while they watched several images containing faces together with other objects – a car, a bird, a phone, and a scrambled face.

Part of the broader Babytwins Study Sweden (BATSS) project, our research compared identical and fraternal twins at the Karolinska Institutet Center of Neurodevelopmental Disorders (KIND) in Stockholm, Sweden, using various infant-friendly methods. We saw that identical twins were more similar than fraternal twins in their visual preferences for faces versus non-face objects. For example, if one identical twin in the pair looked mostly at the non-face objects, chances were that their co-twin showed the same preference. In contrast to identical twins, fraternal twins share on average only 50% of their genes, and their looking preferences tended to be less similar within pairs.

To check if the heritability was specifically linked to face looking, we measured infants’ preference for the car object, which was the most salient non-face object in our study. Unlike face preference, we did not find that genetics explained the variance in infant car preferences, which gives support to the idea that faces have something unique in capturing our attention which is rooted in our biology. Our results were obtained when infants watched a computer screen with static, photograph-like, stimuli; future research is needed to confirm that the results generalize to real-life social interaction.

Moreover, we showed that the genetic influences explaining face preference differed from the genetic influences involved in eyes preference (relative to mouth preference). Eyes preference is an important social communication behaviour, for example, to establish eye contact, to estimate another person’s direction of gaze, and to infer emotional states. It is curious that two foundational social behaviours like face looking and eye looking have a distinct genetic bases.

While our study had a large sample which covered roughly 30% of the population in the area, families generally had a higher socioeconomic status compared to the rest of Stockholm’s population.  Estimates of genetic and, importantly, shared environment effects, can vary in populations with a wider socioeconomic distribution, and presumably in populations with different sociocultural habits and policies as well (e.g., the availability of caregivers at home due to parental leave). This means that, although our study did not find shared environment effects in our population of twins living in Sweden, it is possible that the family environment would play a role in looking preferences if we included families with more diverse backgrounds.

 Looking preferences are important for learning. 

We also showed that looking more at faces than at non-social objects at 5 months was associated with having a larger vocabulary in the beginning of their second year, supporting the idea that looking preferences early in life may be consequential for later development. We also checked if this visual preference could predict whether the infants, also in the second year of life, showed behaviours associated with autism, a neurodevelopmental condition defined by difficulties with social communication, or whether boys and girls differed in face preference. We did not find support for these ideas. A limitation of these results is that in contrast to the objective assessment of looking preferences at 5 months via eye tracking technology, later development was only assessed via questionnaires to parents. Further, autism characteristics often become evident much later in development, which might explain the lack of associations early in life.


Our study indicates that genetics contribute to how an infant curates their visual environment early in life, in terms of their selection of perceptual input for social versus non-social information. The infant's active choice to look at different targets leads to different learning opportunities and, because their looking might signal something to others, it may evoke different reactions from other people, especially caregivers. Selection by means of viewing behaviours that enrich or constrain their visual experience and social interactions precedes other exploratory behaviours such as pointing, grasping, or crawling to targets or partners, and marks a baby's developmental journey.

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in