GMAC: A Matlab toolbox for spectral Granger causality analysis of fMRI data

Investigation of causal interactions within brain networks using Granger causality analysis (GCA) is a key challenge in studying neural activity on the basis of functional magnetic resonance imaging (fMRI). The article describes an open-source software toolbox GMAC (Granger multivariate autoregressive connectivity) implementing multivariate spectral GCA. Available features are: fMRI data importing/exporting, network nodes definition, time series preprocessing, multivariate autoregressive modeling, spectral Granger causality indexes estimation, statistical significance assessment using surrogate data, network analysis and visualization of connectivity results. All functions have been integrated into a user-friendly graphical interface developed in the Matlab environment, easily accessible to both technical and clinical users.     

High-level language ability in healthy individuals and its relationship with verbal working memory

The aims of the study were to investigate healthy subjects’ performance on a clinical test of high-level language (HLL) and how it is related to demographic characteristics and verbal working memory (VWM). One hundred healthy subjects (20–79 years old) were assessed with the Swedish BeSS test (Laakso, Brunnegård, Hartelius, & Ahlsén, 2000) and two digit span tasks. Relationships between the demographic variables, VWM and BeSS were investigated both with bivariate correlations and multiple regression analysis. The results present the norms for BeSS. The correlations and multiple regression analysis show that demographic variables had limited influence on test performance. Measures of VWM were moderately related to total BeSS score and weakly to moderately correlated with five of the seven subtests. To conclude, education has an influence on the test as a whole but measures of VWM stood out as the most robust predictor of HLL.     

Migrant care workers as protective factor against caregiver burden: results from a longitudinal analysis of the EUROFAMCARE study in Italy

Objective: The aim of the analysis is to assess the impact of privately employed migrant care workers (MCWs) on the burden of Italian family members who care for a disabled older person.

Methods: EUROFAMCARE is a one-year prospective survey carried out to provide evidence on the availability and use of support services by family carers of older people in Europe. In Italy, 990 family caregivers were enrolled and successful follow-ups were completed for 863 subjects. The survey also assessed the level of caregiver burden using the COPE index, which has three sub-sections: ‘Positive Value’ (PV), ‘Quality of Support’ (QS) and ‘Negative Impact’ (NI). We used the one-year change of the COPE-NI as dependent variable and we realised multilevel regression models to estimate the longitudinal predictors of caregivers’ burden increase.

Results: At a cross-sectional level, most burdened caregivers are those caring for a demented relative (COPE-NI = 13.6), with no educational title (14.5) and looking after their own spouses (15.1). Longitudinally those employing an MCW are significantly protected against burden increase (regression coefficient: ?1.52; p < 0.01) while those who cannot rely on the support of other family members are exposed to the risk of burden increase (0.991; p < 0.05). Other formal services do not have any protective effect.

Conclusion: Our study suggests that employing an MCW, rather than using formal services, is associated with a reduction of caregiving burden. Further research should assess whether the shift in care responsibilities to the MCWs also implies a transfer of care burden, and understand how these workers can be better supported by existing formal services.