Oral presentation to the 9th annual Canadian Conference on HIV/AIDS Research, in April 2000, in Montreal.

Determinants of Bisexuality Among Young Men Who Have Sex with Men

Arn Schilder, Keith Chan, Steve Martindale, Mary Lou Miller, Amy E.Weber, Michael R. Botnick, Kevin JP Craib, Martin T. Schechter, Robert S. Hogg.

ABSTRACT

Objective: To compare socio-demographic, psychosocial and behavioural characteristics of gay and bisexual men (MSM) enrolled in an open cohort.

Methods: MSM aged 15-30 living in Vancouver who had not previously tested HIV+ were recruited by the Vanguard Project through physicians, clinics and outreach. On an annual basis, men undergo an HIV test and complete a self-administered questionnaire. Data collected include demographics, sexual behaviours, history of forced and paid sex, sexual orientation, use of illicit drugs, and validated measures of depression, social support, alcohol dependency and self-esteem. Comparisons between those who reported only having sex with men, and those who reported having sex with men and women in the previous year were carried out using contingency table analysis. Persons who reported any sex-trade involvement and injection drug use in the previous year were excluded from the comparison. Step-wise logistic regression was used to identify independent predictors of bisexuality. All reported p-values are two-sided.

Results: A total of 605 men completed baseline test results and questionnaires as of December 1999. Of these, 71 (11.7%) had sex with both men and women in the previous year. Bisexual men were less likely to have a high school education (p<0.001), live in stable housing (p<0.001), to be employed (p<0.001), and have a regular sex partner (p=0.002) and more likely to be younger (p<0.001), have paid for sex (p<0.008), have a high depression score (p=0.002) and have a casual sex partner (p=0.022) than gay men. In multivariate analysis no high school education (AOR=0.40; 95%CI 0.20, 0.77), younger age (AOR 0.86; 95%CI: 0.79, 093), high depression score (AOR=1.77; 95%CI 1.02, 3.07), and having a casual sex partner (AOR=2.72; 95%CI 1.03, 7.15), were associated with bisexuality.

Conclusion: Our data indicate that bisexual men enrolled in the Vanguard Project are marginalized and more likely to engage in sex with a casual partner.

SLIDE PRESENTATION

• To compare socio-demographic, psychosocial and behavioural characteristics of gay and bisexual men (MSM) enrolled in an open cohort.

The Vanguard Project

• Ongoing prospective study of over 850 MSM in Vancouver, British Columbia
• 15 to 30 years of age at baseline
• No previous HIV-positive test result
• Reside in Greater Vancouver region at baseline
• Recruited through outreach, publicity, medical clinics and physicians

Data Collection

• Annual self-administered questionnaire: Including demographics, sexual behaviours, history of forced and paid sex, sexual orientation, substance use, and measures of depression, social support, alcohol dependency and self-esteem
• HIV testing and stored blood samples

Eligibility Criteria

• Completed baseline HIV test and questionnaire between May 1995 and December 1999
• Had reported having sex with men and/or women in the previous year
• Did not report any sex-trade involvement or injection drug use in the previous year

Definitions

Gay:

• Had reported having sex with only men in the previous year

Bisexual:

• Had reported having sex with both men and women in the previous year

Statistical Analysis

• Contingency table analysis using non-parametric methods
• Multivariate analysis using step-wise logistic regression
• All reported p-values are two-sided

• As of December 1999, 605 men had completed test results and questionnaires and met the eligibility criteria for this analysis
• Of these men, 534 (88.3%) were gay and 71 (11.7%) were bisexual

Table 1: HIV Prevalence and Incidence

	Gay men Rate (95% CI)	Bisexual Rate (95% CI)	p-value
HIV prevalence	1.31% (0.35, 2.27)	1.43% (0.00, 4.21)	0.933
HIV incidence	0.76% (0.26, 1.25)	0.96% (0.00, 2.84)	0.722

Table 2: Bivariate Predictors of Bisexuality

	Bisexual n (%)	Gay n (%)	p-value
Median age (IQR)	23 (20-26)	26 (23-28)	p < 0.001
High school education	50 (71.4)	484 (91.5)	p < 0.001
Stable housing	63 (88.7)	529 (99.1)	p < 0.001
Employed	42 (60.9)	424 (81.1)	p < 0.001
Regular sex partner	46 (65.7)	434 (81.4)	p = 0.002
Casual sex partner	65 (92.9)	433 (82.0)	p = 0.022
Paid for sex	5 (7.0)	10 (1.9)	p = 0.008
High depression score (> CESD median)	41 (57.7)	206 (38.6)	p = 0.002

Table 3: Multivariate Predictors of Bisexuality

	Odds Ratio	95% CI	p-value
High school education (Yes versus no)	0.40	0.20, 0.77	p = 0.007
Age (in years)	0.86	0.79, 0.93	p < 0.001
High depression score (> CESD median	1.77	1.02, 3.07	p = 0.041
Casual sex partner (Yes versus no)	2.72	1.03, 7.15	p = 0.043

• Self-reported behaviours
• Low literacy levels (including ESL)
• Self-reported behaviours were assessed at baseline, not longitudinally

• Our data indicate that bisexual men in this cohort are younger, less likely to have a high school education, more likely to be depressed and more likely to engage in sex with a casual partner.