If I run a multiple regression analysis in SPM8 with multiple scans at different time points (covariate 'Time') for each subject (covariate 'Subject'), and other covariates as 'Gender', 'Motivation', etc., is it correct to talk about areas correlating positively/negatively with one of my covariates (e.g., Time) when I set this covariate was set to +1/-1 and all other covariates to 0?

The question arises from the fact that the analysis is called a regression analysis, however, I found in the literature the results being described as correlations.

Many thanks in advance!
Iris