Thanks, Alfonso.

The procedure worked well, but just to confirm, the QA_timeseries first-level covariate #1 reflects mean global BOLD signal changes (z-score) below a certain threshold relative all other individuals or just that person's scan data?  

What would you recommend if one were interested in the difference in mean BOLD signal over time (follow-up minus baseline)?  I'm seeking to control for potential systematic BOLD differences between baseline and follow-up that might be due to changes in brain perfusion.  Are the QA z-scores, once parsed by condition (baseline, follow-up), relative to mean BOLD signal change within condition or over all conditions?

Warm regards,
Jeff