Objective: The authors investigated sexual behaviors related to HIV tr
ansmission among homeless mentally ill men in a New York City shelter.
A previous study of a similar population found HIV prevalence to be 1
9%. Method: In standardized interviews with 122 melt, data on sexual b
ehaviors for the previous 6 months were collected. The frequency and n
ature of sexual episodes that may facilitate HIV transmission were exa
mined. In addition, sexual risk behaviors among broadly defined diagno
stic groups were compared. Results: Of the 222 men, 65 (53%) had sex,
56 with women and 20 with men (11 who had sex with both women and men
are counted in each group). The sexually active men, in most cases, ha
d only occasional sex (once per month or less). The majority of sexual
ly active men-29 (52%) of those who had sex with women and 12 (60%) of
those who had sex with men-had sex without a condom and with nonmonog
amous partners. Comorbid cocaine abuse or dependence was significantly
associated with high-risk sexual behaviors. Conclusions: The majority
of these men had sex occasionally or not at all. Nonetheless, because
many of them had unprotected sex with nonmonogamous partners, the few
sexual episodes may have carried an appreciable risk of HIV transmiss
ion. Moreover, men with a comorbid cocaine dependence may represent a
group with an especially high risk for sexual HIV transmission. The au
thors propose that in this population, preventive interventions could
modify the nature, if not the frequency, of sexual episodes.