One-way ANOVA of normalized DCX-IR values in the SVZ showed a biphasic variation as a function of time (Fig. 4; main effect: F4,37=57.2, p<0.00001), with chronic alcohol reducing DCX-IR at the end of exposure (p<0.001), followed by a doubling of DCX-IR at 3 d (p<0.001), with an increase still present on day 7 (p<0.001). In contrast, DXC-IR was reduced to levels below those of controls on day 21 (p<0.001, Fig. 4). Because this analysis used normalized optical densities to address batch variation in staining intensity, we also carried out a two-way ANOVA with time following exposure, and treatment as factors, on raw optical density data not subjected to normalization. Using this approach, there was also a significant effect of time (F3,34=18.5, p<0.00001), and a significant time×treatment interaction (F3,34=32.0, p<0.00001). Once again, post-hoc analysis showed that DCX-IR levels in alcohol-exposed animals were lower than their respective controls upon termination of exposure (p<0.001), higher than controls on day 3 (p<0.001) and day 7 (p<0.01), and finally decreased below control values on day 21 (p<0.01). Thus, analysis of normalized and raw optical density data yielded essentially identical results.
