The Predictive Value of Early Temperament and Sensory Features for Cognitive Development and Early Symptoms of Autism Spectrum Disorder in Preterm Children

Background: There is accumulating evidence that temperament and sensory processing differs between children at high risk for autism spectrum disorder (ASD) who do and do not develop ASD. A study by Garon et al. (2018) suggests that temperament, particularly Effortful Control at 24 months, plays a role in predicting later ASD symptoms. In addition, Wolff et al. (2018) reported cross-sectional associations between sensory experiences and restricted and repetitive behavior (RRB). However, research on this matter focused mainly on high-risk siblings. Yet, preterm children are another important high-risk group for developing ASD (Agrawal et al., 2018). It could be hypothesized that also in these children, temperament and sensory features may be of predictive value for cognitive development or symptoms of ASD. Pérez-Pereira et al. (2015) for example, reported that extroverted temperament predicted better language outcome in a group of preterm children.

Objectives: This study explored the predictive value of early temperament and sensory features for cognitive development and ASD symptomatology at 36 months in a group of preterm children.

Methods: Participants were 45 children born before 30 weeks of gestation (21 girls). At the corrected ages of 10, 18 and 24 months, sensory features and temperament were assessed using the Infant/Toddler Sensory Profile (ITSP) and the Infant Behavior Questionnaire (IBQ) or Early Childhood Behavior Questionnaire (ECBQ). Mullen Scales of Early Learning, ADOS-2 and the Social Responsiveness Scale (SRS) were administered at 36 months.

Results: Linear regressions were performed to predict ADOS_RRB, ADOS_SA (social affect), SRS_RRB, SRS_SCI (social communication and interaction) and Developmental Index (DI) of the Mullen at 36 months, using the measures of temperament and sensory-related features at 10, 18 and 24 months. More sensory avoiding as measured with the ITSP_24 predicted lower Mullen_DI scores (F(1,40) = 4.172, p = .048, R² = 0.094) and higher SRS_RRB scores (F(1,43) = 8.361, p = .006, R² = 0.163). Higher ADOS_RRB scores were predicted by less sensory seeking as assessed with the ITSP_24 (F(1,43) = 6.527, p = .014, R² = 0.132). A lower score on low registration on the ITSP_10 was predictive for more social problems on the SRS_SCI (F(1,43) = 12.508, p = .001, R² = 0.225). Temperament and sensory features did not have any predictive effect on ADOS_SA scores.

Conclusions: This study indicates that sensory processing is predictive for cognitive development and early ASD symptoms in preterm children, especially at 24 months. Concerning RRB, the ITSP was predictive for both parent-report (SRS) and observational measures (ADOS-2). It is possible that children avoid stimuli because of oversensitivity and develop RRB’s to cope with this oversensitivity, or that children who already have specific interests, are less tempted to process other stimuli. As for the social-communicative aspects of ASD, only parent-report (SRS) was predicted by sensory features. Children who registrate less of their environment, possibly miss attempts to social interaction from others, have less learning experiences, and may develop less optimal social skills. In this study, early temperament does not seem to be of predictive value for ASD symptomatology or cognitive development.