Partnership-level correlates of sexual risk taking within the sexual partnerships of young transwomen in San Francisco, California

Abstract

Objectives

Little is known about the sexual networks of young transwomen, leaving a major gap in what we know about transmission dynamics and the elevated rates of HIV in this population. The objective of this study was to understand partnership-level factors associated with condomless anal sex among young transwomen.

Methods

A secondary data analysis of the sexual partnerships of young transwomen was conducted using baseline data from the SHINE study. Generalised estimating equation logistic regressions were used to assess for partnership-level associations between partnership type, age, injection drug use and racial concordance, HIV seroconcordance, sexual role and condomless receptive (CRAI) and insertive anal intercourse (CIAI).

Results

Our analysis included 187 young transwomen that reported a total of 464 sexual partnerships where they had at least one episode of anal sex in the past 6 months. We found casual (n=232 or 50%) and commercial partnerships (n=106 or 22.8%) to be significantly associated with a lower odds of CIAI (OR=0.53, 95% CI 0.32 to 0.86 and OR=0.39, 95% CI 0.18 to 0.82) and CRAI (OR=0.30, 95% CI 0.19 to 0.47 and OR=0.35, 95% CI 0.2 to 0.62) compared with main partnerships (n=126 or 27.2%). Additionally, HIV-positive seroconcordant (n=25 or 5.4%, OR=4.05, 95% CI 1.44 to 11.40) and injection-drug using partnerships (n=25 or 5.4%, OR=3.66, 95% CI 1.34 to 9.95) were found to be significantly associated with an increased odds of CIAI among participants compared with HIV-negative seroconcordant (n=330 or 71.1%) and non-using partnerships (n=338 or 72.8%), respectively.

Conclusion

Young transwomen, like other populations, engage in condomless sex more often with main than casual and commercial partners, suggesting a need for interventions that address sexual practices with steady main partners.

INTRODUCTION

Globally, the estimated HIV prevalence of transwomen is 19.1% and the odds of being infected with HIV for a transgender woman is 48.8 times higher than that of all adults of reproductive age.¹ Within the USA, there is a notable difference between the prevalence of HIV for young transwomen (4%–19%) compared with adult transgender women (27.7%–39.5%).^{2 3}

Sexual networks may, in part, explain the high prevalence of HIV among transwomen.^{4 5} Studies with adult transwomen and men who have sex with men (MSM) found that main partnerships, seroconcordance, racial concordance and injection drug use discordance to be associated with greater sexual risk taking behaviours.^5–8 Additionally, greater age discordance between sexual partners has been associated with increased HIV risk in cisgender heterosexual women and MSM.^{9 10 W1 W2} Sexual role has previously been documented to also influence condom use among the partnerships of MSM and adult transwomen, such that seropositioning may be practiced by individuals given the lower risk of HIV transmission associated with being the insertive partner.⁷ However, little is known specifically about the sexual partners and networks of young transwomen.^W3 Research is needed to determine if young transwomen face the same sexual network risks as their adult transwomen peers.

The goal of this analysis was to describe the sexual partnerships of young transwomen and to evaluate associations between partnership-level characteristics and sexual risk taking. We assessed if greater age discordance, main partnerships, seroconcordance, racial concordance and injection drug use discordance were associated with greater sexual risk taking. Last, we assessed if sexual roles (either insertive, receptive or both types of anal intercourse) affected engagement in condomless anal intercourse by the transwomen.

METHODS

The SHINE study was the first study of HIV risk and resilience among young transwomen ages 16–24 years old in the San Francisco Bay Area. This secondary analysis uses baseline data collected between 2012 and 2013.^W4 Participants were recruited using peer and direct referrals from community-based organisations, outreach at events and online outreach through social networks. Young transwomen were considered eligible for participation if they met the following criteria: (1) self-identified as any gender different from that typically associated with their male assigned sex at birth, (2) were 16–24 years of age, (3) reported living in the San Francisco Bay Area. All study procedures were reviewed and approved by the UCSF Institutional Review Board (IRB approval number 12–08875).

Young transwomen’s individual-level demographic characteristics—age, race/ethnicity, income, education, HIV status and injection drug use—were self- reported. Participants were surveyed about their six most recent sexual partners. Only participants who had anal sex in the last 6 months were included in this analysis. Participants were asked the age, race, sexual orientation, HIV status of their sexual partners, if their partner had sex with other transwomen, had a penis, if they injected drugs and whether the partner was a main or casual partner. Partnerships where a participant reported she had been paid or had paid her partner were reclassified as a commercial partnership, thereby creating three exclusive categories of partnership type. Concordance variables (HIV-status, injection drug use and race/ethnicity) were all similarly operationalised and considered concordant if the participant and her sexual partner’s respective characteristics were the same (eg, participant identified as African American and identified her sexual partner as African American). Age discordance was a continuous variable equal to the absolute difference in age between the young transwomen and their sexual partner and was consequently broken into five ordinal ranges to avoid assumptions of a linear trend in risk. Last, the sexual role of participants within the partnership was categorised as being exclusively the insertive partner, exclusively the receptive partner or having been both types of roles during anal sex within the sexual partnership, depending on what sexual behaviour participants reported. The primary outcomes of our analysis were condomless receptive (CRAI) and insertive (CIAI) anal intercourse given their higher HIV transmission risks compared with other types of intercourse. For each partnership, participants were asked the number of times they had receptive and insertive anal sex with their partner and how many times they did not use a condom in the past 6 months for each sexual position. We constructed dichotomous variables using this count information that were indicative of at least one episode of CRAI and CIAI within the partnership.

Using generalised estimating equation logistic regressions to account for the clustered (by the young transwomen) nature of our dataset, we examined bivariate associations between partner-level characteristics and condomless insertive and condomless receptive anal intercourse. All analysis were conducted in STATA 14 (Stata 2015. Stata Statistical Software: Release 14. College Station, Texas, USA).

RESULTS

Our sample included 187 young transwomen that reported a total of 464 sexual partnerships where they had at least one episode of anal sex. Most partnerships were HIV-negative concordant (71.1%), casual (50.2%), had an age discordance of 0–2 years of age (36.6%), non-injection drug use concordant (72.8%) and were racially discordant (58.2%) (table 1). Among young transwomen who reported having a main sexual partner, 51.9% had at least one additional outside casual and/or commercial partner. The prevalence of condomless insertive and condomless receptive anal intercourse within partnerships was 13.4% and 35.1%, respectively. The sexual roles of young transwomen during anal sex were diverse: 13.3% had insertive anal sex exclusively, 67.5% had receptive anal sex exclusively and 19.2% had both receptive and insertive anal sex. The majority of young transwomen and their partners were HIV-negative with 80.8% of young transwomen participants testing negative for HIV and 82.3% of partners being reported by participants as HIV negative.

Table 1.

Descriptives of young transwomen’s sexual partnerships in San Francisco, California (n=464)

Characteristics of the transwomen	N (total=187)	%
Race/Ethnicity
White, Non-Hispanic	56	29.95
Asian, Non-Hispanic	8	4.28
Black, Non-Hispanic	34	18.18
Hispanic or Latino/a	57	30.48
Multiple races, Non-Hispanic	22	11.76
Other	10	5.35
Poverty line
At or above	41	21.93
Below	129	68.98
Unknown	17	9.09
Education
Less than High School	38	20.32
High School or General Education Development (GED)	69	36.9
Some college & beyond	80	42.78
HIV status
Negative	151	80.75
Positive (or unknown)	36	19.25
Age (mean, SD)	21.36	0.15
Characteristics of their partners	N	%
Race
White	216	46.55
Asian	11	2.37
Black	76	16.38
Latino/a	111	23.92
Multiple races	22	4.74
Other	25	5.39
Donť know	3	0.65
Sexual orientation
Straight	234	50.43
Gay/Lesbian	80	17.24
Bisexual	59	12.72
Queer	42	9.05
Donť know	18	3.88
Previously had sex with other transwomen
Yes	210	45.26
No	179	38.58
Donť know	75	16.16
Partner had a penis
Yes	412	88.79
No	29	6.25
HIV status
Negative	382	82.33
Positive (or unknown)	82	17.67
Age (mean, SD)	28.75	0.49
Partnership characteristics	N	%
Serodiscordance
Concordant (Both HIV−)	330	71.12
Concordant (Both HIV+)	25	5.39
Discordant (Transwomen HIV+)	52	11.21
Discordant (Partner HIV+)	57	12.28
Partnership type
Main partner	126	27.16
Casual partner	232	50
Commercial partner	106	22.84
Age discordance
0–2 years	170	36.64
3–4 years	61	13.15
5–9 years	108	23.28
10–14 years	50	10.78
15 or more years	71	15.3
Sexual role (of transwomen)
Insertive	62	13.36
Receptive	313	67.46
Both insertive and receptive	89	19.18
Injection drug use concordance
Neither use (concordant)	338	72.84
Both use (concordant)	25	5.39
Partner uses (discordant)	45	9.7
Transwoman uses (discordant)	56	12.07
Racial concordance
Racially discordant	270	58.19
Racially concordant	191	41.16
Condomless Insertive anal Intercourse
Yes	62	13.36
No	401	86.42
Condomless receptive anal Intercourse
Yes	163	35.13
No	294	63.36

There were significantly lower odds of condomless insertive anal intercourse (CIAI) among participants in main partnerships compared with those in casual (OR=0.53, 95% CI 0.32 to 0.86) and commercial (OR=0.39, 95% CI 0.18 to 0.82) partnerships. The odds of condomless receptive anal intercourse (CRAI) were significantly lower in casual (OR=0.30, 95% CI 0.19 to 0.47) and commercial (OR=0.35, 95% CI 0.2 to 0.62) partnerships compared with main partnerships. Participants in HIV-positive concordant partnerships had significantly higher odds of having CIAI (OR=4.05, 95% CI 1.44 to 11.40) compared with those in concordant HIV-negative partnerships. However, those in HIV serodiscordant partnerships did not have significantly greater odds of engaging in CIAI or CRAI. Participants in partnerships with concordant injection drug use had significantly higher odds of CIAI compared with those in concordant partnerships where neither partner used injection drugs (OR=3.66, 95% CI 1.34 to 9.95).

DISCUSSION

Partnership type was the only partnership-level characteristic found to be significantly associated with both condomless receptive anal intercourse (CRAI) and condomless insertive anal intercourse (CIAI), with participants reporting significantly lower odds of CRAI and CIAI in casual and commercial partnerships compared with main partnerships. Similar to studies of adult transwomen, MSM and a previous study of young transwomen,^{6 7 W3 W5} we found a significantly higher risk of having condomless anal intercourse with main partners. Among MSM, main partners are believed to be the principal source of HIV transmission because of a higher number of sex acts, a higher frequency of receptive roles during anal sex and lower condom use.^W6 Among the young transwomen in our study that reported having a main partner, 51.9% had at least one additional casual and/or commercial partner outside of their primary relationship. Thus, we also observed a potential risk for onward transmission as more than half of participants had a recent casual and/or commercial partner in addition to a main partner during the 6-month period.

Overall, young transwomen in this analysis exhibited high sexual risk behaviour. Condomless receptive intercourse was reported by almost one third of participants, and young transwomen were most likely to be the receptive partner, which is the partner most at risk of HIV and other sexually transmitted infections. Most of the sexual partners of participants were cisgender males and almost half of the sexual partners of young transwomen had other transwomen sexual partners. The high prevalence of HIV within the transwomen population, which is San Francisco, is estimated to be almost half of transwomen,² and our data showing many sexual partners of transwomen have other transwomen sexual partners, and HIV risk is very high within this sexual network.

The most important limitation of our analysis is that we used egocentric data from young transwomen to determine their partnership level of risk (ie, we relied on data collection only from index participants whom may not necessarily have accurate knowledge of their partners characteristics). This limitation speaks to the need for studies that include sexual partners. Additionally, our findings may not be generalised to all young transwomen because of non-probability-based sampling and the time period our data were collected. Our data were collected prior to the promotion of biomedical HIV prevention methods (ie, PrEP), which may affect condom use and should be considered in future studies of sexual behaviours.

CONCLUSION

Individual behavioural differences do not adequately explain the disproportionate HIV burden experienced by adult and young transwomen. Our data point to important sexual risks within main partnerships and elevated risk within the sexual networks of young transwomen. Engaging young transgender women and their main partners in couples-based HIV prevention programmes may be particularly important for reducing the burden of HIV among transwomen given the evidence of increased sexual risk taking. Research and interventions addressing the partners of young transwomen are vital to curbing the HIV epidemic in this population. Future studies should also consider including partners of young and adult transwomen to improve our understanding of their sexual networks.

Additional references can be found in the online supplementary file 1.