The recent article by Paul et al. examines the lifetime prevalence and antecedents of suicidal ideation among gay and bisexual men.1 Their work confirms various ideas, extrapolates trends quite effectively, and reveals some new and interesting insights. However, I would like to make a few comments.
The first of these concerns Table 2. I feel that the title, “Logistic Regression Differentiating ‘Never Attempters’ from Those Who Attempted Suicide after Age 17” should be changed or appended. Perhaps a subtitle such as “Odds of Being an ‘Attempter’” would make more sense. This would reflect the true meaning of the data and make the table less confusing.
More important, the content of the table is troubling. The article claims that subjects whose parents abused alcohol or drugs are more likely to attempt suicide, but the confidence interval for that odds ratio clearly overlaps 1, making the claim erroneous. The same is true of the birth cohort category ‘25 y in 1971–1980 vs before 1971.’ The text states that those who reached 25 years of age before 1970 were more likely to have attempted suicide. Therefore, the odds ratios for all birth cohorts who turned 25 after 1970 should show a protective effect compared with the pre-1970 cohort. It does, but, once again, the confidence interval for 1971–1980 clearly overlaps 1, making the claim false.
In addition, 2 of the 4 questions stated in the introduction were not adequately addressed. Both concerned suicide ideation in relation to age. No investigation of suicide ideation was made of attempts that occurred after the age of 25 years. Clearly, a first attempt is more likely to occur before the age of 25, as shown by the data, but the data also show that men are likely to attempt suicide more than once. The regression to differentiate between the suicide rates of the two groups of men (younger and older) focuses only on first attempts. However, the introduction of the article makes an interesting point concerning the proximal and distal relations to suicidality of antigay stresses and developmental life transitions. In light of this point, I feel that suicide attempts made after the age of 25 should be investigated further.
