There needn't even be a confounding variable, the direction of causation may simply be reversed: not personality traits -> superager, but superager -> personality traits.

Sounds quite plausible to me that having a good memory at old age leads to a more active and engaged lifestyle leading to happy people that appear less neurotic and more extrovert.