Infant attentional behaviours as prognostic indicators in Cornelia-de-Lange syndrome.

Sarimski (2007) watched five babies who have Cornelia-de-Lange syndrome. The team counted how often each baby looked at a caregiver’s eyes during short play moments. They filmed the infants at home when they were about one year old. Four years later they checked the same children’s social skills.

Babies who rarely met Mom’s eyes stayed behind in later social games. Eye-contact levels stayed almost the same for each child. One glance per minute or less was the danger line. The kids who passed that line still struggled to share toys or chat at age five.

Wang et al. (2022) extend the idea to autism. They show that poor sleep plus odd gaze during face viewing predicts worse core symptoms. Hempkin et al. (2025) flip the gaze idea into a teaching tool. They add eye-gaze cues while naming objects and preschoolers learn words faster. Matson et al. (2013) look almost contradictory at first. Autistic teens stared less at faces yet scored normal on emotion tests. The difference is age and task: Sarimski (2007) tracks live social play, while L et al. used still photos. The teens may have built work-arounds that babies have not yet learned.

You can spot social risk in Cornelia-de-Lange before the child can talk. Count eye-contact for one minute during a favorite song or toy play. If looks fall below one per minute, plan extra joint-attention trials now. Pair the gaze check with parent coaching so Mom knows when to enter the baby’s line of sight. Early action may bend the social curve upward.