This looks epidemiologically hopeless as a study. Try as they might, there is no way to rule out other factors based on the way they did the study. The least depressed women may have been in the most
stable relationships, in which condoms might have been even undesirable for one reason or another. This could
easily account for the lower depression scores.
What the investigators would need to do is a long-term study in which they switched the non-condom-users to make them condom users. If the depression scores worsened for these women, it would be potentially significant in suggesting the hormonal effect. If the results didn't change soon, then the original conclusion was wrong.
Shhhhhh!!! Don't question the results when we are WINNING!!!
You're only supposed to question the results when we are LOSING, LOL.
Ruck