A Prospective Case Series of Premature Infants Who Developed Autism Spectrum Disorder

Background:  Prematurity has been identified as a significant perinatal risk factor for Autism Spectrum Disorder (ASD), but no study has used standardized measures to follow a cohort of premature infants prospectively to examine the emergence of autistic symptomatology.

Objectives:  To document the emergence of ASD in a case series of premature infants.

Methods:  We recruited 49 participants (2 sets of triplets, 11 sets of twins, 21 singletons) at 12 months corrected age for our prospective study. All participants completed the Mullen Scales of Early Learning at ages 1, 1.5, 2, and 3 years. The Autism Observation Scale for Infants (AOSI) was completed at the 1- and 1.5-year visits, and the ADOS was completed at the 2- and 3-year visits. Independent diagnoses were made/confirmed at age 3 by a clinician blind to previous study data, and were based on the Autism Diagnostic Interview-Revised, Autism Diagnostic Observation Schedule (ADOS), and clinical judgment using DSM-IV criteria.

Results:  Five out of the 17 cases seen to date for their 3-year visit received an ASD diagnosis. Cases 1 and 2 were fraternal twin girls, cases 3 and 4 were identical twin girls, and case 5 was a fraternal twin boy (born at 32, 26, and 28 weeks gestation, respectively). Cases 1, 3, and 5 had average cognitive functioning at 12 months; Case 5 remained in the average range at age 3, whereas Cases 1 and 3 showed a dramatic decrease in cognitive functioning through age 3. Cases 2 and 4 had below average cognitive functioning at 12 months and showed a further decrease through age 3. All five cases showed an increase in ADOS severity scores between ages 2 and 3 (Case 5 had the lowest ADOS severity score at age 3). Cases 1 and 5 had few signs of ASD on the 12-month AOSI, and then showed more ASD signs at each subsequent visit through age 3. Cases 3 and 4 showed fewer ASD signs on the AOSI at 18 months compared to 12 months, but then scored in the Autism range on the ADOS at ages 2 and 3. In Cases 1 through 4, one particular ASD sign, lack of social babbling, was present at 12 months, and another, lack of social smiling, was present at 18 months.

Conclusions:  Preliminary examination of the first five cases of ASD in our premature sample indicated that signs of ASD were evident as early as 12 months of age, but that there was heterogeneity in the particular signs manifested across children. There was a pattern of cognitive decline in the majority of cases, although none had a measured or reported language regression. These results suggest that premature infants should be monitored more closely in the second year of life for both developmental delay and ASD signs, particularly if they are multiples. It will be important to assess ASD outcome for the remainder of the sample to determine whether the patterns in cognitive development and ASD emergence seen in the first five cases are consistent findings.