Simple mediational analyses with bootstrapping procedures (n=5000 bootstrap resamples) were used to assess the indirect effect of the mediator on the outcome [42]. An indirect effect was considered significant in the case zero was not contained in the 95% confidence interval