Studies into the "causes of homosexuality" are extremely controversial. They should not be controversial because people don't like what they imply. Rejecting science because you don't like what it tells you is similar to believing in creationism. But, scientific studies of the sort that show a "cause" of a complex behavior are always suspect. Many have later been discredited or found to be flawed. But, as a scientist I do find it surprising how complex behavior DOES seem to be linked to physical aspects of our selves, including to brain structure and genetics.
Some studies claim that homosexual men have a tiny part of their brain that looks more like the equivalent brain region in women than the male version. This work was done quite some time ago and I know that back then people considered it suggestive but flawed.
More recently, studies have linked homosexuality to having older male siblings. On the one hand such studies are very hard to do well and should always be considered with skepticism. On the other hand, a more recent study not only confirms that correlation, but also suggests that the mechainism for it is in mom.
From BBC News [1]:
A man's sexual orientation may be determined by conditions in the womb, according to a study.
Previous research had revealed the more older brothers a boy has, the more likely he is to be gay, but the reason for this phenomenon was unknown.
But a Canadian study has shown that the effect is most likely due to biological rather than social factors...
Professor Anthony Bogaert from Brock University in Ontario, Canada, studied 944 heterosexual and homosexual men with either "biological" brothers, in this case those who share the same mother, or "non-biological" brothers, that is, adopted, step or half siblings.
He found the link between the number of older brothers and homosexuality only existed when the siblings shared the same mother.
The amount of time the individual spent being raised with older brothers did not affect their sexual orientation.
Hmmmm...you'd think that they could come up with a better way of referring to half-brothers than "non-biological," but the result is still interesting.
The researchers go on to speculate a kind of immune response mechanism to explain the result, which I think is going a bit beyond the data, but might be a testable theory. The idea is that males have proteins that are foriegn to the mother and that a mother who previously bore a son might produce antibodies to the male proteins she is exposed to. This could be seen as a more subtle version of what happens when a mother who is Rh- has an Rh+ child. The first child is unaffected, but does produce and immune response in the mother. That immune response will attack a second Rh+ child. In this case the researchers think that is is a male specific protein rather than Rh that is triggering the immune response and that the immune response, rather than harming the child, actually affects brain development. That hypothesis assumes the differences in brain structure previously observed between homosexual and heterosexual men hold up.
I think the observation may well be good. The total sample size isn't bad, though once divided into the different categories (full sib, half sib, adopted) the sample sizes are going to be less impressive. The immune hypothesis to explain it sounds more dubious since an immune response tends not to simply benignly alter development but rather damage tissues. I suspect a more hormonal response is at play, but both are valid hypotheses.
[2] | 
[3] | 
[4] | 
[5] | 
[6] | 
[7]