Relationship Between Surface Area, Brain Volume, and Cortical Thickness in Young Males with Autism

Background: Increasing evidence indicates that brain growth in children with autism may be accelerated early in life and then slow, resulting in more significantly enlarged brain volumes in younger children (Courchesne, Neuron 2007).  Development of surface area and thickness of the cortex may be affected by different factors (Rakic, Science 1988), but the relationship of these aspects of cortical structure has not been previously described either in young children or in children with autism.

Objectives: Characterize cortical morphometry including surface area and cortical thickness in a cohort of young males with autism.

Methods: Subjects: 44 males meeting strict criteria for Autistic Disorder (age 4.8(s.d. 1.1)) and 31 healthy male age-matched controls (age 4.3(s.d. 1.3)).  Diagnoses were established by trained raters using the ADOS and ADI-R.  IQ was assessed on a subset of subjects using a hierarchy of tests, including the WASI, Mullen, WPPSI, DAS-II, or WISC-R. The mean IQ of the autistic subjects was 56.8 (s.e. 1.78, n=36); and for healthy controls the mean was 117.9 (s.e. 3.58, n=26).  Autistic subjects were sedated during MRI procedure; controls were not sedated.  All subjects were scanned using the same 1.5T scanner and protocol (Giedd, Cerebral Cortex 1996).  Measures of surface area, average cortical thickness, and gray and white matter volumes for total cerebrum and lobar regions were obtained using the fully automated CIVET pipeline developed by the Montreal Neurological Institute (Ad-Dab’bagh, Neuroimage, 2006).  Group comparisons were performed in SPSS using one-way ANOVA with age as a covariate.   Relationships between cortical features were analyzed using stepwise linear regression and partial correlation.

Results: Total brain volume (926.0 (s.e. 15.8) vs 1033.5 (s.e. 13.2), F=26.3, p<.001), gray matter in all regions and occipital white matter volumes were increased in autistic subjects.    Surface area and mean cortical thickness were increased in all regions in subjects with autism.  Both surface area and cortical thickness significantly predicted variation in gray matter volume, with the greater contribution from surface area.  Regression analysis of the relation between surface area and cortical thickness including age, diagnosis, cortical thickness, and interaction of age and cortical thickness found all terms to be significant, suggesting a different relationship between area and cortical thickness in autistic subjects than healthy controls.  Subsequent partial correlation analysis controlling for age found a significant correlation between area and cortical thickness in typical controls (pr=.509, p =0.005) but not autistic subjects (pr=.145, p=.346).  A similar pattern of results was also found across brain regions except for the occipital lobe, in which neither group showed significant correlation.

Conclusions: Increased brain volumes in young males with autism are associated with both increased surface area and increased cortical thickness.  We did not find the same correlation between surface area and cortical thickness in autistic subjects as in healthy controls, which may relate to early abnormal neurodevelopment.