Extra-axial cerebrospinal fluid volumes from 6 to 24 months of age are associated with poorer executive function at school-age in children with and without autism.

Abstract

BACKGROUND: Abnormally increased extra-axial cerebrospinal fluid (EA-CSF) volume is present as early as 6 months in infants later diagnosed with autism and is associated with symptom severity at the age of diagnosis, but it is unknown whether early EA-CSF enlargement has long-term impacts on other clinical domains. Executive function (EF) deficits are frequently observed in children with autism and are linked with worse academic outcomes, higher anxiety, and lower adaptive functioning. The current study examines the association between EA-CSF volume at infancy and EF at school age in a longitudinally phenotyped cohort of children with either high (HL) or low (LL) familial likelihood for autism. METHODS: In this prospective study, 239 infants underwent MRI scans during natural sleep at 6, 12, and 24 months of age. HL was defined as having an older sibling with autism. The sample was divided into three diagnostic groups: HL infants who were diagnosed with autism at 24 months (HL+, n = 34), HL infants not diagnosed with autism (HL-, n = 131), and LL infants without autism (LL-, n = 74). Two parent-rating scales of EF were collected at the school-age follow-up (Mage= 10.4 years ± 1.34): the Behavior Rating Inventory of Executive Function (BRIEF) and Conners Parent Short Form. ANOVA was used to test for diagnostic group differences in EF. Longitudinal mixed-effects models utilized EA-CSF volumes in infancy to predict EF at school-age, while controlling for IQ, sex, total cerebral volume, and diagnostic group. RESULTS: Consistent with previous literature, HL + participants had greater EF deficits than both HL- and LL- on the BRIEF (F = 18.46, p < .0001) and greater EF deficits than LL- on the Conners (F = 8.55, p < .001). Further, higher EA-CSF volumes during infancy were associated with poorer executive function eight years later on both BRIEF (ß=0.18, p < .001) and Conners (ß=0.18, p < .001). This association between EA-CSF and executive function was observed across familial likelihood and diagnostic groups. CONCLUSIONS: Our study indicates that elevations in EA-CSF volumes during infancy, even in those who do not have autism, have long-term associations beyond autism symptomatology. These findings underscore the potential link between infant CSF physiology and future behavioral domains such as executive function at school age.