The Brain Connectome: A Multimodal Study of Discordant Monozygotic Autism Twins

Background: Unraveling the relative contributions of genetics and environment remains difficult in Autism Spectrum Disorders (ASD). Comparison of monozygotic (MZ) twins discordant for ASD appears to be a promising lead. However, studies investigating such a population are still scarce, limited to small samples and/or to isolated brain measures. According to this, the sensitivity of such investigations could be improved in two ways. First, control samples have to be extended from typically developed MZ twins to pairs concordant for ASD and discordant for other mental disorders. Second, as proposed in the Human Connectome Project, the cerebral pathways of ASD could be highlighted while combining anatomical and functional measures of brain connectivity.

Objectives: As a part of a comprehensive twin project (Roots of Autism Twin Study Sweden “RATSS”) including (epi-)genetics, immunology, environmental, and behavioral assessments, to explore brain anatomical and functional connectivity within and between the different categories of MZ twins listed above.

Methods: The project presented here has started in the late summer 2011 and is still in progress. Up to 140 pairs of MZ twins aged from 9 to 19 years old will be recruited for this project. All of them have been screened for ASD and comorbidities in a previous study and are categorized according to the followings: 50 pairs discordant for ASD and three control samples matched for age, sex, IQ and socioeconomics - 30 pairs concordant for ASD, 30 pairs concordant for other mental disorders and 30 pairs typically developed of MZ twins. As for the neuroimaging assessment, twins undergo MRI anatomical and functional examinations using a 3T GE scanner. This includes one structural T1, one resting state functional MRI (rsfMRI; cross fixing) and two High Angular Diffusion Imaging (HARDI) sequences with 8 b0 images, 60 directions and b=1000 or b=3000 s/mm². As recommended when following a brain connectome analysis pipeline, datasets will be preprocessed to get covariance matrices for each twin in both modalities, i.e. HARDI and rsfMRI. Based on statistical inferences between measures in parcellated regions of the brain, these matrices provide an exhaustive and easy-understanding quantification of interregional brain connectivity. These covariance measures will be not only compared between twins but also between groups.

Results: To date, 8 pairs of MZ twins have been included, among which one is discordant for ASD. Taking into account the low statistical power allowed by this small sample size, no quantitative results can be provided here. However, 20 more pairs will be included until May 2012, which should be enough for the authors to show preliminary results during the IMFAR congress.

Conclusions: Unraveling the respective influence of genetics and environment is currently challenging ASD research. For the first time, MZ twins discordant for these disorders will be compared to different control samples using a multimodal approach. While mapping exhaustive brain connectivity in these groups, we should be able to access complex information about ASD mechanisms that have not been highlighted yet.