Monday, July 30, 2012

Semen an antidepressant? Alternative explanations appear more plausible.

A decade ago a study made an extraordinarily bold claim: that semen has antidepressant properties in women (Gallup, Burch, & Platek, 2002). Although widely-reported, there seems to have been a lack of critical response to this study (although see The Straight Dope and this article for exceptions) and apparently no follow up studies have been done to test whether this claim is actually valid.[1] The study was correlational in nature and provided no direct biological evidence for the claim that semen has an antidepressant effect. A well-known research principle is that correlation does not imply causation, and there are plausible alternative explanations that the authors of the study did not take into consideration.
Could condom use be depressing for women? Or are depressed women more likely to use them?

What the study actually found was that women who did not use condoms during sex had lower levels of depressive symptoms compared to women who usually or always used them, and to women who abstained from sex altogether. The authors argued that vaginal exposure to semen was the causal mechanism underlying this effect, arguing that semen has components including various hormones, particularly prostaglandins, that are readily absorbed into the woman’s bloodstream and that these have an antidepressant effect. The authors’ source for the antidepressant hypothesis was a single case study (Ney, 1986) which found that evening primrose oil apparently alleviated depression in a child-abusing mother. Ney argued that evening primrose oil and semen have in common the fact that they contain prostaglandins, and claimed that the latter have an antidepressant effect. Ney even went so far as to argue that in this particular case disruption to the woman’s sex life ensuing from the birth of her child was a causative factor in her depression because of the resulting lack of exposure to her husband’s semen. A single case study seems like a pretty tenuous basis for proposing such a far-fetched hypothesis but the story becomes even stranger. Ney admitted that semen contains fairly minute quantities of prostaglandins but argued that even so “they have powerful direct effects”, presumably effects on mood if this theory is to make sense. His authority for this was a review of the psychiatric implications of prostaglandins (Gross et al., 1977). What seems truly bizarre to me is that Gross et al. reported that from the little research that had been done on the subject, one study found that depressed patients had slightly elevated levels of prostaglandins, whereas another study found depressed patients had normal levels of prostaglandins (p. 1195). Furthermore, Gross et al. noted that antidepressant medications had an inhibitory effect on prostaglandins, although they stated that whether or not this had any clinical implications was as yet unknown (p. 1194). Surely, if absorption of prostaglandins in semen had an antidepressant effect, one would reasonably expect this to mean that depressed patients had low levels of these hormones or that antidepressant medications would actually increase not decrease their production, yet neither of these things appear to be true. Hence the claim that semen has any antidepressant properties at all does not appear to rest on any biological evidence.
Returning to Gallup et al.’s (2002) findings, not only have they provided no evidence that their finding that women who do not use condoms were less depressed than condom users was due to semen exposure, they did not properly consider a range of alternative explanations. For example they did not adequately explore why some women choose not to use condoms in the first place and whether pre-existing differences between users and non-users might have affected their results. Their results suggest some intriguing possibilities. They noted that non-users had the highest frequency of intercourse and in fact those who never used condoms had sex almost twice as frequently as those who always used condoms. However, when statistically controlling for condom use, frequency of intercourse did not predict depressive symptoms. Nevertheless, the fact that non users were not only less depressed but also had the most frequent intercourse suggests that non-users may be different from users in some important way. Gallup et al. argued that the important difference is their exposure to semen but since this was a correlational study they are not justified in arguing that this is the underlying cause of the relationship between condom use and depression, particularly when there are other plausible causal factors. Subsequent research has found that frequency of intercourse is positively correlated with both satisfaction with mental health and satisfaction with life in general (Brody & Costa, 2009) but this would not explain why condom use would seem to be related to depression. Brody (2010) has argued that sex with condoms is not real intercourse but something ‘akin to mutual masturbation’. I confess to finding this statement rather baffling but it is possible that for some women at least, sex with a condom may be less satisfying than without. One survey found that 40% of women reported decreased sensation associated with condom use and that some women associate condoms with a number of ‘turn-offs’ such as discomfort (Crosby, Milhausen, Yarber, Sanders, & Graham, 2008). Therefore, it seems possible that sexual enjoyment has an antidepressant effect that may be reduced by condom usage.
Another possibility, although it may sound strange, is that it is depression itself that leads to condom usage. Evidence for this comes from a study examining safe sex practices over a three month period (Morrill, Ickovics, Golubchikov, Beren, & Rodin, 1996). In this study, women who were depressed when the study began or who became depressed later on were more likely to use condoms. The authors noted that the association between depression and safer sex was influenced by the inclusion of women who were not sexually active. They suggested that depression may inhibit sexual activity. Even in a study like this, it is still difficult to untangle causality, as it is not clear why the women were initially depressed and whether or not their depression was somehow caused by their sexual behaviour or by something unrelated.
The possibility that depression may precede condom usage rather than being a result also suggests the possibility that personality characteristics that influence a woman’s decision whether or not to use a condom may also play a role in depression. Gallup et al. claimed that sexual risk-taking is unrelated to depression, but their two references supporting this statement were a study on homosexual men and one on drug users receiving psychiatric treatment. Considering the special nature of these two population groups it seems fair to say that they are not representative of the women in their study. It is possible that women who choose not to use condoms might have “happy-go-lucky” personalities compared to their more cautious counterparts, or perhaps some other combination of personality traits that protects them from depression. A Portuguese study found that women who use condoms are more likely to have “immature psychological defense mechanisms” than women who do not use them (Costa & Brody, 2008). These immature defense mechanisms are apparently associated with poorer mental health and more depression and anxiety. An earlier study found that socially anxious women were also more likely to use condoms compared to less anxious women and have less frequent sex (Leary & Dobbins, 1983). This seems comparable to Gallup et al.’s results who found condom using women were not only more depressed but had less frequent sex.
In conclusion, the claim by Gallup et al. that semen has an antidepressant effect is not only lacking any direct evidence, there does not even appear to be any plausible biological reason to believe that the components of semen have a beneficial effect on mood. The authors acknowledged that more definitive and direct evidence is needed, e.g. manipulation of the presence of semen[2] or measures of seminal components in the bloodstream. However, no studies providing such evidence appear to have been done and it may be there is little justification for investigating such a far-fetched hypothesis. There are plausible psychological mechanisms that could explain the relationship between condom use and depression, such as pre-existing depression, personality differences, or “turn-offs” associated with condom usage, that have not been adequately explored and seem more likely to yield informative results.

NB: Nothing in this post should be construed as advocating for or against condom usage. This is purely a discussion of scientific issues.


[1] Gallup et al. claim to have replicated their results with a larger sample, but their results have not as yet been published in a peer-review journal.
[2] I would imagine that trying to get such an experiment passed by an Ethics Committee would be far from easy!

Follow me on Facebook, Google Plus, or Twitter.

© Scott McGreal. Please do not reproduce without permission. Brief excerpts may be quoted as long as a link to the original article is provided. 

This article also appears on Psychology Today on my blog Unique - Like Everybody Else. 

Brody, S. (2010). The Relative Health Benefits of Different Sexual Activities. The Journal of Sexual Medicine, 7(4pt1), 1336-1361. doi: 10.1111/j.1743-6109.2009.01677.x
Brody, S., & Costa, R. M. (2009). Satisfaction (Sexual, Life, Relationship, and Mental Health) Is Associated Directly with Penile–Vaginal Intercourse, but Inversely with Other Sexual Behavior Frequencies. The Journal of Sexual Medicine, 6(7), 1947-1954. doi: 10.1111/j.1743-6109.2009.01303.x
Costa RM, & Brody S (2008). Condom use for penile-vaginal intercourse is associated with immature psychological defense mechanisms. The journal of sexual medicine, 5 (11), 2522-32 PMID: 18761591
Crosby, R., Milhausen, R., Yarber, W. L., Sanders, S. A., & Graham, C. A. (2008). Condom ‘turn offs’ among adults: an exploratory study. International Journal of STD & AIDS, 19(9), 590-594. doi: 10.1258/ijsa.2008.008120
Gallup GG Jr, Burch RL, & Platek SM (2002). Does semen have antidepressant properties? Archives of sexual behavior, 31 (3), 289-93 PMID: 12049024
Gross HA, Dunner DL, Lafleur D, Meltzer HL, Muhlbauer HL, & Fieve RR (1977). Prostaglandins. A review of neurophysiology and psychiatric implications. Archives of general psychiatry, 34 (10), 1189-96 PMID: 20867
Leary, Mark R., and Dobbins, Sharon E. Social Anxiety, Sexual Behavior, and Contraceptive Use. Journal of Personality and Social Psychology 45 (1983): 1347-54.
Morrill, A. C., Ickovics, J. R., Golubchikov, V. V., Beren, S. E., & Rodin, J. (1996). Safer Sex: Social and Psychological Predictors of Behavioral Maintenance and Change Among Heterosexual Women. Journal of Consulting and Clinical Psychology, 64(4), 819-828. doi: 10.1037/0022-006X.64.4.819
Ney PG (1986). The intravaginal absorption of male generated hormones and their possible effect on female behaviour. Medical hypotheses, 20 (2), 221-31 PMID: 3637620


  1. Didn't the Gallup study also fail to consider oral contraceptives?

    1. They did record oral contraceptive use and found that there were no differences in depressive symptoms between those using contraceptives and those who did not. Interestingly, 7 out of 10 women who never used condoms were on oral contraceptives. Strange that they thought the hormones contained in semen would affect mood when the hormones contained in oral contraceptives apparently do not.

  2. One should also consider the work on positive delusions. Women that are not depressed might see themselves as being untouchable or that pregnancy and STDs happen to others but not them. Where as women that are not depressed and at the same time are not influenced by positive delusions might be more aware of the risks of not having their partner wear a condom. That being said it would have been nice to have seen the participants views on condom use.

    1. Now there's a very interesting point. There is evidence that positive illusions help maintain self-esteem and that those who are depressed have fewer illusions. Perhaps this might have something to do with why women who were more depressed were more likely to use condoms, that they were more worried about realistic risks of unprotected sex.

  3. Great article! I appreciate your take on this issue and the thorough research you did to support your claims. I did a write-up on this particular study a few months ago and argued for a similar conclusion; however, I focused on a few different critiques. For one thing, Gallup et al. never actually measured "semen exposure"--the measured condom use and just assumed that translated to differential semen exposure. Moreover, their interpretation of the data may not be correct--why is it that semen is an "antidepressant" when an equally plausible effect is that condom use is a "depressant?" For more on my take, see here: Physician Loses Job After Suggesting Semen Is A Better Valentine’s Day Gift Than Chocolates

    1. I hadn't seen your article, so thanks for pointing me to it. I am in agreement with your conclusions. I think the Gallup et al. paper was interesting as a correlational study, but I'm amazed that they focused on such a far-fetched explanation as to the underlying cause when the evidence was so flimsy and alternative explanations plentiful. Seems like the idea that semen could be an antidepressant captured the imagination of quite a few people, including those who ought to know better, such as the illustrious Dr. Greenfield.