Neuroanatomical Correlates in Bilinguals: The Case of Children and Elderly

Abstract

How bilingualism modulates brain areas beyond the language regions is still controversial. Through a comprehensive set of analyses on brain structure, we investigated brain differences between Basque-Spanish bilinguals and monolinguals in children and the elderly, the most sensitive target groups to detect potential brain differences. In particular, we employed Diffusion MRI in combination with T1-MRI, network-based statistics and a graphtheoretical approach to investigate differences between bilinguals and monolinguals in structural connectivity and topological properties of brain networks. Additionally, regional grey and white matter structural differences between groups were examined. The findings suggest that the effects of bilingualism on brain structure are not solid but unstable. However, lifetime experience of active bilingualism may lead to increased neural reserve in ageing, since better global network graph-efficiency has been observed in the elderly lifelong bilinguals compared to monolinguals.3

A meta-analysis of gray matter volume differences between bilinguals and monolinguals

“…It is important to note that we did not exclude studies for age range of participants or L1/L2 language (other than bimodal bilinguals as noted above), as there are not enough studies fitting the criteria to investigate consistency within specific age ranges (see below). It is also worth stating that we included García-Pentón et al (2019), which at the time was a preprint available on bioRxiv (https://www.biorxiv.org/). The methodology was determined to be of the quality of the other studies found, and as it met all of the inclusion criteria it was determined appropriate to include in the final meta-analyses conducted here.…”

“…Foci from monolingual > bilingual contrasts were only reported for one of the eight studies (Olulade et al, 2016), so only the bilingual > monolingual contrast analyses were run and reported here. Out of the ten eligible comparisons, two included separate age groups with unique contrasts eligible for inclusion: older and younger adults (Gold et al, 2013), and adults and children (García-Pentón et al, 2019). The later study was the only study included that contained children.…”

“…Methodological variability included software package, where two studies used FSL (Pliatsikas et al, 2014;García-Pentón et al, 2019), and the rest used SPM (Mechelli et al, 2004;Ressel et al, 2012;Gold et al, 2013;Abutalebi et al, 2014Abutalebi et al, , 2015bOlulade et al, 2016).…”

“…Multiple comparison corrections also varied, with FSL-based studies using threshold free cluster enhancement (TFCE; Pliatsikas et al, 2014;García-Pentón et al, 2019), and SPM-based analyses using variable corrections including family wise error (FWE) correction at the voxel level (Mechelli et al, 2004;Ressel et al, 2012;Gold et al, 2013), FWE correction at the cluster level (Abutalebi et al, 2014(Abutalebi et al, , 2015b, non-stationary cluster correction (Olulade et al, 2016), and some studies additionally reported coordinates for clusters not corrected for multiple comparisons (Mechelli et al, 2004;Ressel et al, 2012). We include all the reported foci in the first analysis (to get a full picture from the limited overall number of eligible studies), and subsequently included only foci corrected for multiple comparisons for a more stringent analysis (described below).…”

“…To date, to our knowledge there have been ten experiments across eight studies that have employed a whole-brain comparison of bilinguals and monolinguals using VBM for GMV (Mechelli et al, 2004;Ressel et al, 2012;Gold et al, 2013;Abutalebi et al, 2014Abutalebi et al, , 2015bPliatsikas et al, 2014;Olulade et al, 2016;García-Pentón et al, 2019). These studies vary considerably in several aspects of the sample characteristics, including first (L1) and second (L2) language, age of acquisition of L2 for the bilingual group, language proficiency, sample size, covariates of no interest included in the analyses, and in overall reporting of these sample details.…”