Good hypothesis.
For my part, I'm saying that the trend would virtually disappear if the stats were adjusted for marital status.
(I.e., women have become more liberal and consume more anti-depressants because they are "riding the carousel.")
Regards,