Objectives: To examine voxel-wise gray matter (GM) volumetry within a pair of MZ twins discordant for ASD traits versus typically developed controls, as a part of the Roots of Autism Twin Study Sweden (“RATSS”).
Methods: The ASD traits discordant twins pair were 16;5 years old right-handed males, sharing several psychiatric comorbidities (dyslexia, dyscalculia, ADHD) and scoring 98/88 (Co-Twin1) and 99/84 (Co-Twin2) to verbal/perceptual IQ testing. Additionally, Co-Twin1 presented with marked autistic traits as assessed by both the Autism Diagnostic Observation Schedule (ADOS) and Autism Quotient (AQ), as well as tics. Six pairs of right-handed MZ twins were included as typically developed controls, with a median age of 16;4 (range: 12;6-18;3), median verbal/perceptual IQ of 106/109 and a male/female ratio of 4:8.
3T T1-SPGR data were processed in SPM8 as follows: i) unified segmentation (prior: NIHPD 13-18.5); ii) DARTEL with modulation and spatial normalization to MNI; iii) quantitative normalization by the total intracranial volume (TIV); iv) smooth: 12mm FWHM. Additionally, an explicit mask was created from the mean of the modulated normalized GM maps. Thus, Co-twin1 and Co-twin2 final maps were subtracted from the controls’ in order to obtain statistical maps reflecting the inter-individual differences. SnPM8b was used for statistical inferences, i.e. for One Sample t-tests testing for positive and negative effects in Co-twin1 and Co-twin2, using the GM mask and controlling for differences in sex and verbal/perceptual IQ. Results were considered as significant at p FWE <0.05, k >150 voxels. Finally, individual values were extracted at each of the four most significant peak voxel pointed out by SnPM8b, to feed within-pairs Intraclass Coefficient (ICC) analyses.
Results: As compared to controls, both twins were characterized by lower TIVs, as well as GM volume i) increases in right medial (posterior cingulate) and lateral (supramarginal) parietal, and left temporal pole regions; ii) decreases in the left primary visual area. Co-Twin1 specifically showed increases in extended bilateral parietal regions, right inferior occipital, superior temporal and (orbito)frontal areas, and decreases in right inferior temporal, left orbitofrontal and middle occipital cortices. Right-lateralized increases (medial frontal, middle temporal, crus cerebellum) and decreases in right superior temporal, insula, visual associative cortex, and left primary somatosensory, medial and inferior temporal cortices were exclusively reported in Co-Twin2. ICC analyses highlighted the right superior temporal and medial frontal areas as being the regions showing the most significant differences between both twins, i.e. 72% and 47% increase when the discordant pair was excluded from the analyses, respectively.
Conclusions: For the first time, voxel-wise GM volumetric approaches are used to explore MZ twins discordant for ASD traits. Though they have to be cautiously considered, the findings pointed to differences in right superior temporal and medial frontal regions as potential neural substrates for differences in ASD traits.
See more of: Brain Imaging
See more of: Brain Structure & Function