Children with FXS were observed to have increased total brain, total tissue, total gray matter, total white matter, and temporal lobe white matter in comparison to controls. Cerebral cortex gray matter, cerebellar white matter, and frontal lobe gray matter volumes were noted to be slightly decreased in the FXS group compared to controls. We did not observe any group differences in the rate of brain growth between groups, meaning that the rate of brain growth across the interval from 2 to 5 years of age in children with FXS paralleled that seen in controls. This would suggest that, for the brain regions assessed with the methods described here, brain overgrowth had its onset prior to the first measurement at 2 years of age.
In comparison to a control group composed of both TYP and DD, and a contrast group consisting of children with iAUT, the children with FXS showed a unique pattern of brain volume measurements. Whereas, for the most part, children with FXS showed generalized enlargement of both gray and white matter volume in the cerebral cortical lobes in contrast to controls, enlargement was less robust than that observed in children with iAUT, with the exception of a more striking, disproportionate enlargement of temporal lobe white matter in children with FXS (as compared to both controls and those with iAUT). Temporal lobe white matter volume was the only cortical volume in individuals with FXS to show significant enlargement after adjustment for TBV, suggesting a specific neuroanatomical signature that goes beyond the generalized cerebral cortical volume enlargement seen in both FXS and iAUT. Temporal lobe white matter was also a region observed to be enlarged in the VBM study by Hoeft et al., (2011) (13). Specific differences were also observed in the cerebellum where FXS subjects showed significant enlargement of cerebellar gray matter volume in contrast to individuals with iAUT.
The pattern of subcortical volume changes also showed a specific pattern over time, consistent with previous reports by our group looking at brain volumes at age two years10
and voxel-based morphometric measurements throughout this early age interval.11,12
Specifically, here we report a stable increase in volume, maintained over this two year age interval, in basal ganglia structures, with a striking, persistent increase in caudate volume (34%), followed by globus pallidus (15%) and putamen (5%); along with a decreased volume of the amygdala, predominantly noted on the right side. In contrast to children with iAUT, where we have previously reported enlargement of the amygdala with a more modest but significant enlargement of caudate volume, children with FXS show a pattern of markedly increased caudate volume with significantly decreased volume of the amygdala that is present at 2 and maintained across early development. The robust caudate enlargement (by over 30%) replicates our earlier reports on this sample at age 2–310,13
and is consistent with observations in older individuals with FXS.26,27
Our findings of a specific pattern unique to FXS (in contrast to iAUT) are consistent with behavioral reports of distinct social and communicative profiles for children with FXS and ASD.28
Work examining 9–12 month old infants with FXS finds evidence for a specific pattern of motor-sensory deficits that can be distinguished from infants with DD and ASD (by diagnosis at later age).29
Specifically, problems with motor planning, repetitive movements and repetitive use of objects were observed, making these motor-sensory deficits some of the earliest detectable features of FXS. Our observation of significant basal ganglia enlargement is consistent with these behavioral findings and suggests these brain differences may be present from a very early age, if not prenatally.
Of note, reports about the molecular biology of FXS suggest a possible mechanism for brain growth patterns observed in this study. Harlow et al.30
has demonstrated that FMRP inhibits the generation of progenitor neurons from radial glia in mouse cerebral cortex, suggesting that lack of FMRP, as seen in FXS, might result in an increased proliferation of progenitor cells and subsequent cerebral cortical overgrowth. Alternatively, FMRP has been shown to act as a regulator of dendritic mRNA.31
There are a number of hypotheses about the role of FMRP in brain growth that could help explain specific pattern of brain development we find in FXS. Work with the Fmr1
knock-out mouse has identified FMRP as playing a specific role in neocortical development.32
FMRP has been shown to act as a regulator for dendritic mRNA33
and the lack of FMRP has been linked to defects in the differentiation and migration of neural progenitor cells in the neocortex.33
BDNF and trkB signaling in the neural progenitor cells plays a critical role in normal cortical development, and these signaling pathways function aberrantly in the neuroplast of the FXS knockout mouse.34
Grossman and colleagues35
proposed that FMRP acts as a general regulatory process over dendritic spine loss, maturation, and formation, and that this could explain why different brain regions may display different ‘phenotypes’. Our findings for specific patterns of enlargement (e.g., basal ganglia, cerebellum) as well as diminished size (amygdala) in young children with FXS would support this hypothesis.
These findings have significance for expanding our understanding about the neurodevelopmental mechanisms underlying FXS. The presence of early brain differences (evident by age 2) in young children with FXS points to aberrant early brain development in this condition. Examining brain growth in infants with FXS, and perhaps including both full and premutation cases may further illuminate the pathogenesis and provide additional clues for mechanisms to target for intervention during vulnerable periods of brain growth. This study also highlights the importance of longitudinal studies to help define the trajectory of brain development. Studying the trajectory of brain development as children enter school age would provide clues to better understand the effectiveness of targeted pharmacological interventions (e.g., mGluR5) on brain growth. Lastly, these data continue to point to a specific neuroanatomical signature for FXS that differs from that seen in iAUT, suggesting the importance of studying the neurobiological basis of autistic behavior in more etiologically-defined and etiologically-homogeneous disorders such as FXS, and support the notion that the field may be underestimating the neurobiological heterogeneity underlying autistic behavior.
One limitation of our study is that the brain volume focused on regions-of-interest defined by sulcal-gyral anatomy (e.g., cerebral lobes) or identifiable subcortical boundaries (e.g., hippocampus, caudate). This volumetric approach is complementary to, yet different from a voxel-level morphological approach for elucidating fine-grained neuroanatomical variations, which was used to analyze this FXS cohort in an earlier study from our group.11–13
In this previous study we detected FXS specific trajectories in various parts of the cerebral cortex, the thalamus and the basal forebrain. Another, limitation is the size of our subgroups of the controls (DD and TYP) was also modest and is a limitation, although the longitudinal design allows for increased power to detect significant differences, particularly with the effect sizes we observed in the basal ganglia. We only examined male children with the full mutation, and therefore our findings may not generalize to females or individuals with the premutation. Lastly, the measures we employed for the behavioral assessment (e.g., ADI-R, ADOS) were developed for categorical diagnosis of autistic disorder and we were therefore limited in the ability to look at dimensional qualities across time.