may be masked by a comparable amount of volume gain from inflamed astrocytes), or where GM loss is offset by a neuroadaptive response to the need for greater cognitive control over substance use (e.g., see Koehler et al., 2013). The discrepancies between regional GM volume findings from our PSU participants and those of the previous studies could relate to the differences in age (our participants were on average 12–15 years older), the duration and type of drugs used, and/or the duration of abstinence: i.e., the younger cohorts in the previous studies could have been less prone to brain inflammation from chronic substance use, brain inflammation may not be present in active users (as in Ersche et al., 2011) or in polysubstance abusers abstinent for a few weeks (as in Liu et al., 1998), or brain inflammation could have subsided after 2 years of abstinence (as in Tanabe et al., 2009). However, this is speculative and can only be tested in longitudinal studies of abstinent polysubstance users: If our speculation/interpretation were correct, the observed GM differences – and particularly the larger WM volumes – should diminish with duration of abstinence as the substance withdrawal-related inflammation subsides over time. Such a dynamic
