• eating disorders;
  • preterm infants;
  • pregnancy and perinatal complications;
  • brain ultrasounds;
  • adolescents



We attempt to resolve the question of whether pregnancy complications and perinatal trauma, including brain insults, in premature infants increase the susceptibility to eating disorder symptomatology during the adolescent years.


This study uses a historical, prospective methodology to investigate the 84 members of a cohort of infants born prematurely (<33 weeks gestation) at Thomas Jefferson University Hospital during a 25-month period, from 1979 to 1981. We extracted the following information from their neonatal intensive care unit (NICU) records: ultrasound examination findings (graded on intraventricular hemorrhage [IVH] and periventricular leukomalacia [PVL]), records of pregnancy complications and perinatal trauma, and medical problems during the NICU stay. This method eliminated recall bias, a problem with previous studies. We followed up the members of this cohort, obtaining data on 53 (63%). We correlated the NICU data with the following outcome measures: physical measurements and psychosocial, psychological, and eating disorder symptomatology data obtained by self-report questionnaires. Pregnancy and perinatal complications were combined into one composite variable. We used the method of multiple discriminant function analysis to determine statistical significance between groups.


There were no statistically significant differences between the low (0–2) and high (3–7) composite variable of pregnancy/perinatal complications and outcome variables.


These results indicate that traumatic episodes early in life, including brain insults, do not appear to increase the susceptibility of developing eating disorder symptomatology, depression, deficiency of self-esteem, or distortion of body shape during late adolescence. © 2002 by Wiley Periodicals, Inc. Int J Eat Disord 31: 202–209, 2002; DOI. 10.1002/eat.10021