There were several interesting findings in this study. In controls, there was significantly stronger leftward planum temporale asymmetry in the older group than the younger group. This was also found in the ASD group. There was also a significant correlation between age and left planum temporale size, found only in controls. We also found larger frontal language region volume in the ASD group compared to controls and this was found across the age groups. There were significant correlations in the younger ASD group between frontal language regions and measures of language and autism severity.
The finding of increased leftward asymmetry of the planum temporale in controls partly supported our hypothesis of increased left volume and leftward asymmetry of language areas in older compared to younger individuals. This finding is consistent with previous studies of typically developing individuals, demonstrating increased asymmetry of temporal language regions in adolescents compared to adults (Sowell et al., 2001
) and differences in asymmetry of the Sylvian fissure in adolescents and adults compared to children (Sowell et al., 2002
). These developmental changes in anatomy may also be related to age-related changes in language functions. It was surprising that we did not find a significant interaction between diagnosis and age group, rather increases in leftward asymmetry with age were equivalent between the typically developing and ASD groups. This did not support our hypothesis, but is consistent with one previous study that found no developmental differences in the superior temporal gyrus between ASD and controls. One interesting difference, however, was in the correlations with age. In the control group, the left planar volume increased with age. In ASD, we did not find this correlation and when left-handers were removed, the right planum temporale decreased with age in this group. So, although both groups had increased leftward asymmetry, these correlations suggest that there may be different mechanisms. More studies, however, need to examine developmental changes in these critical language areas.
An unexpected finding in our study was increased frontal language region volume in the ASD group compared to controls. Broca's area has been shown to be involved in the organization of articulatory sequences and inner speech (Stuss & Benson, 1986, cited in Démonet et al., 1992
), in subvocal rehearsal (Paulesu et al., 1993
), and in the initiation of speech (Alexander et al., 1990
). The pars triangularis and pars opercularis are also involved in semantic encoding and retrieval and in on-line manipulation of semantic representations (Blumenfeld et al., 2006
). Functional neuroimaging studies have demonstrated that this region may also be important for phonological (Démonet et al., 1992
; Zatorre et al., 1996
) and syntactic processing (Dapretto & Bookheimer, 1999
; Embick et al., 2000
). Both anatomical (Abell et al., 1999
; de Fossé et al., 2004
; Herbert et al., 2002
; Herbert et al., 2005
; McAlonan et al., 2005
) and functional studies utilizing language tasks (Gaffrey et al., 2007
; Harris et al., 2006
; Just et al., 2004
; Kana et al., 2006
) have indicated differences in these frontal areas in individuals on the autism spectrum, which may be related to differences in language development. Findings, however, of anatomical studies have been inconsistent and increased pars triangularis and opercularis volume in ASD has not previously been reported. This, however, is the only study that has performed manual measurements of these critical frontal language regions in ASD, which may better take into account differences in individual anatomy than more automated methods. The sample size was also larger than many previous anatomical studies of ASD. More studies need to be conducted that focus on both anatomical and functional differences of these frontal areas.
Differences in pars triangularis and pars opercularis volume could be due to differences in language abilities between the groups, as controls had significantly higher CELF scores than the ASD group, however, significant correlations of these regions with behavioral measures were only found in the younger ASD group. Correlations in the control groups were probably not found due to the smaller range of CELF scores found in those groups. In the younger ASD group, larger right pars triangularis volume was associated with lower language scores on the CELF. This finding may be related to suggestions that a lack of pruning of tissue, especially in the right hemisphere, in individuals with neurodevelopmental language disorders may contribute to problems with language (Beaton, 1997
; Morgan & Hynd, 1998
). It is unclear, however, why this correlation would not be found in the older group of ASD individuals. In the younger group, larger left pars triangularis and left frontal language volumes were associated with more severe communication and social autism symptoms, indicating that these regions may be involved in language and social functions.
In summary, there were age-related changes in the planum temporale in both typically developing individuals and individuals on the autism spectrum, with both groups having increased leftward asymmetry in older than younger individuals. Group differences were found in frontal language regions, with the ASD group having larger volume than controls, with no age effects. There are several limitations to this study. One limitation is that this was a cross-sectional study, rather than a longitudinal study, which would be able to directly assess developmental changes in behavior and brain anatomy. Another limitation is that the ASD individuals in this study had a wide range of functioning levels and language abilities. Larger sample sizes would be necessary, so that these groups could be further broken down based on functioning level or language abilities or so that variability related to these could be examined within the group. Another limitation is that because the study involved manual tracings, only specific regions, with well-defined anatomical boundaries were examined and other potentially important language regions were not examined in this study. Future studies should use longitudinal designs to further explore the developmental trajectory of cortical language regions in ASD.