Origins and Predictors of Friendships in 6- to 8-Year-Old Children Born at Neonatal Risk.

Objective: To test effects of gestational age (GA), early social experiences, and child characteristics on children's friendships and perceived peer acceptance.

Study Design: As part of the prospective Bavarian Longitudinal Study (1147 children, 25-41 weeks GA), children's friendships (eg, number of friends, frequency of meeting friends) and perceived peer acceptance were assessed before school entry (6 years of age) and in second grade (8 years of age) using child and parent reports. The parent-infant relationship was evaluated during the 5 months after birth. Child characteristics (ie, height, motor impairment, cognitive ability, behavioral problems) were measured at 6 years of age. Multiple regressions estimated effects of GA, parent-infant relationship, and child characteristics.

Results: Overall, children with higher GA had more friends, spent more time with friends, and were more accepted by peers at 6 years of age. Better parent-infant relationships, higher cognitive abilities, and fewer motor and behavioral problems predicted more friendships and higher peer acceptance after adjusting for sex, socioeconomic status, multiples, siblings, and special schooling. Across all GA groups, number of friends (child report: mean change, 1.77; 95% CI, 1.57-1.96) and peer acceptance (child report: mean change, 0.14; 95% CI, 0.09-0.19; parent report: mean change, 0.14; 95% CI, 0.11-0.17) increased with age, but the increase in number of friends was higher among preterm children (ie, interaction effect age*GA group: P = .034).

Conclusions: Our results provide evidence of a dose-response effect of low GA on children's friendships and perceived peer acceptance. Improvements in early parenting and motor, cognitive, and behavioral development may facilitate friendships and peer acceptance for all children across the gestation spectrum.