I guess this is my day for mea culpas: I've heard via e-mail from Professor Lee Ellis, whose research suggests that women who take diet pills during pregnancy are more likely to have gay children. (See this story from the Telegraph.) I speculated about his views without knowing much about them; he wrote to tell me,

I am not a biological determinist, but I do think biological variables are very important for just about every behavioral trait we humans exhibit. For a trait like sexual orientation, my guess is that most of the variation is due to an interaction of genetic factors and perinatal environmental factors. I've seen little evidence that social environmental factors during childhood or later in life are significant except in terms of the degree to which one expresses his/her sexual orientation.

I also looked at an earlier study and, finding some of his wording confusing, concluded that he might believe biology leads gay people to reject Christianity. He writes:

As I recall my thinking at the time, it was that gay children are more likely to reject Christianity because Christian teachings are in general opposition to their sexual behavior. I can imagine other possibilities, but such an explanation seems most reasonable to me currently.

So I stand corrected.