3.2.6. Morris Water Maze

The performance of the mice was separated into initial, learning phase, and maximum performances (Figure 4). During the initial session, there was no effect of strain or treatment for females, however, in the males, all treatment improved performance but it only reached significance for Ex-Con (*p* = 0.044). During the learning phase, female mice had higher path lengths than the males (main effect of sex; *p* < 0.001). Treated GFAP-ApoE4 females had lower latencies than the sex- and strain-matched Sed-Con group (*p* = 0.064), however, it only reached significance with the Ex-Aox mice. Maximum performance was impaired in the GFAP-ApoE4 males, and was reversed by all treatments. However, an ANOVA did not reach significance for strain (*p* = 0.098) or strain x treatment (*p* = 0.096). Females swam slower than males, GFAP-ApoE4 mice seemed slightly faster than GFAP-ApoE3 mice, and none of the treatments affected the swimming speed (not shown). ANOVA yielded a main effect of sex (all *p* < 0.007), but the effect of strain did not reach significance (*p* = 0.057).

**Figure 4.** Exercise and antioxidants were associated with mild improvement in swim maze performance in male GFAP-ApoE4 mice but not in GFAP-ApoE3 mice. Each value represents mean ± SEM, *n* = 5–16. \* *p* < 0.05 vs. sex- and strain-matched Sed-Con groups; # *p* < 0.05 comparing sex-matched Sed-Con GFAP-ApoE3 and GFAP-ApoE4 mice; † *p* < 0.05 comparing strain-matched Sed-Con males and females.
