Since I'm in grad school, this will be good practice for me at analyzing papers and validating research methods. So the fact that my response is so lengthy and critical is really more because I'm having fun doing this, and not an attempt to beat up on you for bringing this.
This article was published in 2002 and has only been cited 14 times in 10 years. Not especially influential.
From the paper itself
It is important to acknowledge that these data are preliminary
and correlational in nature, and as such are only
suggestive. More definitive evidence for antidepressant
effects of semen would require more direct manipulation
of the presence of semen in the reproductive tract and,
ideally, the measurement of seminal components in the
They also acknowledged that their study found a correlation between being in a relationship and not being depressed. They decided that these correlations were insignificant. They also did not further sub-divide between condom use in a relationship and condom use out of a relationship
Those who "always" use condoms were actually less depressed than those who "sometimes" use condoms and those who have "no sexual intercourse," and they were only slightly more depressed than those who "sometimes" use condoms.
AND then, just to round it off, their standard deviations on all counts were ginormous. In other words, the values at the top and bottom were with 1 standard deviation of the bottom and top, respectively. For those who don't know statistics, that means that these findings are pretty much statistically insignificant.
Then they blew their entire argument out of the water with this confession:
Although the overall effect was modest (R2 = 0.076), condom
use accounted for more variance in depression than
any of the other predictors, and the proportion of variance
(.25) due to condom use was greater than that of all of the
other predictor variables combined.
An R2 value of 1 suggests perfect correlation. This correlation is 0.076, which to me is much closer to "no correlation" than anything.
This was their attempt to factor in promiscuity:
A third possibility might be that having sex without
condoms could be an indicator of high-risk behavior,
because the prospect of contracting a sexually transmitted
disease is greater among those who do not use condoms.
Thus, one could argue that individual differences
in risk taking behavior may be confounded with condom
use. However, several studies have shown that various instances
of sexual risk taking behavior do not correlate
with scores on the Beck Depression Inventory (Dilley,
McFarland, Sullivan, & Discepola, 1998; McCusker,
Goldstein, Bigelow, & Zorn, 1995).
In my experience, most people use condoms when they're sleeping around. AIDS is terrifying and word has gotten out. I was in University in 2002, and there were free condoms everywhere. People were taking them by the handful. Most college-age women are more likely to use condoms if they're sleeping around, and only if they're in a stable committed relationship do they feel safe to ride bareback. Coincidentally, being in a stable committed relationship is good for your mood.
Among sexually active females who never used condoms
or only used condoms some of the time, there was
a significant correlation between depression scores and
how long it had been since they had sexual intercourse.
I don't know about you, but when I'm feeling bummed out, I lose interest in sex. Chicken or the egg? Perhaps. But more likely, me being bummed out (depressed) has a negative effect on my sex drive, which in turn increases the time since my last sexual intercourse.
In other words, depressive symptoms among females who
did not typically use condoms increased as a function of
the elapsed time since their last sexual encounter.
Or, "time since their last sexual encounter increased as a function of depressive symptoms among females who did not typically use condoms."
Researchers at my own Alma Mater wrote an excellent analysis
of this study and found that there were serious problems with both their sampling methods and analysis techniques.
Edit: verified back at home, those links only work from a University campus that subscribes to Springer's publications.