Number of sexual partners, condom use and risk of human immunodeficiency virus infection

Background. To analyse the relation between number of sexual partners, selected sexual habits and the risk of human immunodeficiency virus (HIV) infection. Methods. We conducted a case-control study nested in a cross-sectional survey conducted among subjects attending sexually transmitted diseases (STD) clinics in Northern Italy. Eligible for the study were 1711 subjects (1259 males, 452 females) who referred themselves for the first time between September 1988 and March 1993 to three STD clinics in Northern Italy for suspected STD or STD treatment. A total of 145 subjects (113 males and 32 females) were HIV positive. Results. In comparison with subjects reporting no or one sexual partner over the 3 years before the interview, the estimated odds ratios (OR) of HIV serum positivity were 1.2 (95% confidence interval [CI] : 0.6-2.3), 0.8 (95% CI : 0.4-1.8) and 0.3 (95% CI : 0.4-2.5) in subjects reporting 2-3, 4-5, and greater than or equal to 6 partners, respectively. The results were similar considering separately males and females and in men reporting only homosexual partners. Regular condom use decreased the risk of HIV infection: in comparison with subjects reporting no or occasional use of condoms, the OR of HIV infection was 0.5 (95% CI : 0.4-0.8) for regular users. Considering men only, compared with men with no homosexual intercourse, the OR of HIV infection was 2.3 (95% CI : 1.4-3.9) in those reporting bisexual intercourse and 2.2 (95% CI : 1.2-4.2) in men reporting only homosexual intercourse (among homosexuals), There was no relation between HIV infection risk and receptive anal sex. Conclusions. The risk of HIV infection does not increase linearly with the number of sexual partners in this population. This is reasonable, as the prevalence of HIV infection in this population is essentially determined by drug use. Caution is needed in the interpretation of these results since the analysis of role of number of sexual partners in male intravenous drug users is impaired by low statistical power