Are gays more suicidal than straights?
Warren Throckmorton, PhD
A reader wrote and asked if I would review another new study from the Archives of Sexual Behavior regarding suicidality and sexual orientation. The reference is:
de Graaf, R., Sandfort, T.G.M., & ten Have, M. (2006). Suicidality and sexual orientation: Differences between men and women in a general population-based sample from the Netherlands. Archives of Sexual Behavior, 35, 253-262.
I don't have time for a full review but in looking over the study, it is noteworthy and important the authors controlled for prior psychiatric conditions and they asked participants about their feelings of discrimination.
There were significant differences between straight and gay men, after controlling for psychiatric status on death ideation, death wishes, suicidal contemplation and deliberate self-harm. Controlling for psychiatric status left only one statistical difference between straight women and lesbians: suicide contemplation.
Some points to consider:
1. Although the majority of gay men were free of suicidality, there is a risk for one or more measure of suicidality based on sexual orientation.
2. The effect sizes of the risk are small to modest. In other words, there is much overlap between gay and straight groups. The differences require much else besides sexual orientation to account for the variation.
- Death ideation - .7% (trivial)
- Death wish - 2%
- Suicide contemplation - 4%
- Deliberate self-harm - 2%
3. Discrimination moderated the first two factors but even with this variable considered, suicide contemplation and self-harm were association with sexual orientation.
These data are neutral with regard to ideology. Caution is warranted and one cannot say that homosexual orientation leads to suicidality but for reasons that are unclear, there is a small to modest risk. There are other factors that should have been controlled as well but apparently were not: child abuse, current sexual behavior, substance abuse, etc. The remainder of the variation might be accounted for by factors not assessed.
(Thanks to Dr. Gary Welton for the effect size calculations).