The sample consisted of 164 men and 286 women. The majority of respondents reported never having been married: 78% of men and 77% of women. Only 8% of men and 6% of women reported being currently married. The majority of respondents were black (over 80% of men and women). The sample was predominately lower-income although displaying a range of SEP characteristics. In 2007, the median household income in San Francisco County was $67,333 and approximately 11% of residents had a household income below the Federal Poverty Line (which was $10,210 for a single adult). Within our sample, 45% of men and 55% of women reported household income below $10,000, and 18 and 12% of men and women, respectively, reported an income above $30,000. Roughly 42% of men and 38% of women reported full or part time employment, and 49 and 48% of men and women, respectively, reported being unemployed or homemakers. Most (76% of men and 73% of women) respondents resided in HRAs. The distribution of education was similar across sexes: 21% reported having less than a high school degree, 48% had a high school diploma or GED, and 31% had at least some postsecondary education.
In general, none of our measures of socioeconomic position—household income, per capita income and employment status—produced significant results for any of the HIV risk-relevant measures. With respect to total UVI partners in the prior 12 months, men reported a mean of four and women reported a mean of five. As evident in Fig. , the mean values are substantially influenced by the presence of outliers for both men and women. Men and women reported a median of two UVI partners in a 12 month period. The incidence rate ratio (IRR) for the mean number of partners (our inferential model) was roughly one regardless of income, per capita income, or employment status for both men and women (Table ). Exceptions were evident among male respondents when income categories were used. Although the pattern is suggestive of a modest gradient decline in UVI partners as income increases the only significant income effect was evident for men with an annual income of $15,000–$29,000 compared to men with income less than $5,000 (IRR 0.52, P < .01). For both men and women respondents who identified as having ‘other’ employment had significantly higher rates of UVI partners compared to employed persons (IRR 2.0, P < .05) (‘Other’ employment respondents are those who were students or did not consider themselves employed, unemployed or a homemaker. Disabled and retired persons were excluded.). Owing to lack of significant findings, we do not present relative risk tables for subsequent analyses.
Fig. 1 Distribution of total UVI partners in 12 months among respondents, by sex and marital status. Note: Values in parentheses reflect the total number of respondents. Boxes represent the interquartile range, and the dark horizontal line indicates (more ...)
Incidence rate ratios (IRR) of number of unprotected vaginal intercourse partners in 12 months by sex
Roughly 15% of men and 21% of women reported having at least one exchange sex partner of the opposite sex in the prior 12 months. A slightly larger proportion of women who had ever married reported having exchange sex in the prior 12 months compared to women who had never married (Fig. ). There were no differences in the relative odds of exchange sex based on SEP, with the exception that men with an income of $30,000 or more had lower odds of exchange sex compared to other men (ROR 0.09, P < .05).
Proportion of respondents reporting exchange sex in the prior 12 months, by sex and marital status. Note: Values in parentheses reflect the total number of respondents
The concurrency analysis consisted of 984 respondent-month observations on 164 men and 1,716 observations on 286 women. Concurrency—having overlapping partners in a given month—was identified in 37% of male observations and 35% of the female observations. Differences were evident by marital status for both men and women, but the patterns differed between men and women (Fig. ). Whereas, the proportion of concurrent observations declines for men who were ever married (29%) compared to men who were never married (39%), it increases among women who were ever married (40%) compared to women who were never married (34%). These unadjusted differences reflect different gender distributions in marital status. In the adjusted model, married men and women had significantly lower odds of concurrent partners compared to those who never married (ROR 0.04–0.07, P < .01). There were no significant differences in the relative odds ratios of concurrency by any of our SEP measures.
Proportion of respondent-month observations with two or more overlapping partners in a month (6 month recall period), by sex and marital status. Note: Values in parentheses reflect the total number of respondent-month observations
Analyses of per partner UVI episodes consisted of 475 partner observations on 161 men, and 738 observations on 283 women (Excluded respondents did not report sexual intercourse with any partner of the opposite sex during the prior 6 months.). Men reported 19 mean UVI episodes per partner (SD 56) compared to 25 mean UVI episodes reported by women (SD 52). Partner type and outliers exert substantial influence on the mean number of episodes as evident in Fig. . With casual partners the median UVI per partner episodes was one for both men and women, in contrast to a median of 20–25 UVI episodes with main partners. Compared to sexual episodes with casual partners the IRR of UVI with main partners was significantly greater for men (IRR 7.2, P < .001) and women (IRR 7.7, P < .001) in the multivariable Poisson regression models. Men and women reported significantly fewer UVI episodes with partners of undisclosed HIV-infection status compared with partners of known HIV status (IRR 0.6 and 0.5, respectively, P < .001). No significant associations were evident between the measures of SEP and UVI episodes in the models. Findings were similar with respect to per partner episodes of UVI while high or drunk. Men reported a mean of 17 UVI episodes while high or drunk (SD 59) and women reported a mean of 20 UVI episodes while high or drunk (SD 44). In multivariable models no SEP measures were significantly associated with UVI episodes in the presence of substances.
Fig. 4 Distribution of per partner UVI episodes in the prior 6 months among respondent-partner observations, by sex and partner type. Note: Values in parentheses reflect the total number of respondent-partner observations. Boxes represent the interquartile (more ...)