Abstract
One in five transgender women (TW) worldwide are living with HIV, yet little has been published about their health outcomes. We analyzed data from TW (n=37), cisgender women (CW, n=165), and cisgender men who have sex with men (MSM, n=151) in Thailand and Brazil. We hypothesized: (1) TW will have higher odds of depressive symptoms, lower odds of condom use and greater odds of a detectable viral load compared to MSM and CW; and (2) TW will have lower odds of condom use and higher odds of detectable viral load in multivariable analyses. We found that TW had higher odds of depression (OR 2.2, 95%CI: 1.0, 4.8, p=0.04), were less likely than MSM (22% v. 42%, p=0.01) to use condoms with partners of unknown serostatus, and less likely than CW to disclose to sexual partners (15% v. 62%, p<0.0001). In multivariable models, TW had lower odds than MSM of using condoms with partners with unknown serostatus (OR 0.38, 95%CI: 0.15, 0.90) and CW had lower odds than MSM of using condoms with HIV-negative partners (0.60 [0.38, 0.95], p=0.029). We found no significant differences in detectable viral load. Disaggregating data by gender is important to understand factors that contribute to viral suppression and HIV transmission risk among people living with HIV.
Keywords: transgender, HIV, depression, HPTN, Brazil, Thailand
Introduction
One in five (19%) TW are estimated to be living with HIV worldwide (Baral et al., 2013). In addition to HIV, prevalence estimates of lifetime depression among TW are also high, ranging from 35% to 63% (Bazargan & Galvan, 2012; Jefferson, Neilands, & Sevelius, 2013; Nuttbrock et al., 2013). Depression has been associated with HIV risk behavior as well as treatment non-adherence and lack of viral suppression among people living with HIV (PLHIV) (Brickman, Propert, Voytek, Metzger, & Gross, 2017; Castel et al., 2016; Do et al., 2014).
Despite high rates of HIV and depression among TW, data are limited on the mental health, sexual behavior, or HIV treatment outcomes of TW living with HIV (TWLHIV) (Tonia Poteat et al., 2019). TW are often conflated with MSM in pooled data that make it difficult to identify the unique factors relevant for TW (T. Poteat, German, & Flynn, 2016). Very few studies situate TWLHIV in the context of other women living with HIV who may share common experiences as women. The few published studies on TWLHIV are based in North America, and data from low and middle-income countries are scant (T. Poteat, Scheim, Xavier, Reisner, & Baral, 2016).
The HIV Prevention Trials Network (HPTN) 063 study was a one-year multi-site observational cohort of adults in HIV care in Chiang Mai, Thailand; Rio de Janeiro, Brazil; and Lusaka, Zambia. While not actively recruited, TW were not excluded from participation. The present analysis tests the hypotheses that TW will have higher odds of depressive symptoms, lower odds of condom use, and greater odds of a detectable viral load than MSM and CW.
Methods
We analyzed baseline data from 353 participants enrolled in Brazil and Thailand. The Zambia site did not include MSM or TW; therefore, those data were excluded. TW included participants assigned male at birth who reported their current gender as “female” or who responded “yes” to the question, “Do you consider yourself transgender?” The final sample included 37 TW, 151 MSM, and 165 CW. The study was approved by ethical review boards in both Brazil and Thailand.
Participants
Enrolled participants were 18 years of age or older; had documented HIV infection; had at least two visits within nine months of enrollment at a formal health care setting; and had a history of sexual risk behavior in the prior 12 months. Informed consent was obtained from all participants. Details about the parent study are published elsewhere (Magidson et al., 2016).
Data Collection
Interviewer-administered individual structured interviews were conducted with each participant. Sexual behaviors were assessed via Audio-Computerized Self-Interview (ACASI). All data are from the baseline enrollment visit.
Measures
Demographics included sex assigned a birth, current gender, age, education, employment, and country of residence. Sexual risk behaviors included: (1) insertive positioning with last penetrative sexual partner (MSM and TW only); (2) receptive positioning with last penetrative sexual partner (MSM and TW only); and (3) condom use with partners in the prior three months–dichotomized as “always” or “not always.” The Center for Epidemiologic Studies Depression Scale (CES-D) (Radloff, 1977) measured depressive symptoms. A score of 16 or higher was used as the cut-off for significant depressive symptoms. HIV viral load was abstracted from the medical record.
Analysis
Analyses were conducted using Stata 14 (StataCorp, College Station, Texas, USA). Due to previously identified country by risk group interactions (Safren et al., 2016), we conducted stratified analyses and adjusted results by country. Descriptive statistics, chi square tests of association, and multivariable logistic regression models were conducted to test relationships between gender and key outcome variables. Due to the small number of TW, we did not control for other covariates.
Results
Demographics
The mean age of participants was 37 (range 18–62) with no difference by gender (Table 1). TW had the lowest proportion of unemployment (16%) and the highest engagement in sex work (11%). None of the TW had undergone genital reconstruction surgery; and all retained a penis.
Table 1.
Demographic Characteristics
| Participant Characteristic | Transgender women (n=37) | Cisgender women (n=165) | Cisgender MSM (n=151) | TOTAL (n=353) |
|---|---|---|---|---|
| Age in years | ||||
| mean (range) | 37 (19–54) | 38 (18–54) | 36 (18–62) | 37 (18–62) |
| Country | ||||
| Thailand | 73.0% (n=27) | 40.0% (n=66) | 40.4% (n=61) | 43.6% (n=154) |
| Brazil | 27.0% (n=10) | 60.0% (n=99) | 59.6% (n=90) | 56.5% (n=199) |
| Education | ||||
| Primary or less | 37.8% (n=14) | 61.2% (n=101)* | 24.5% (n=37) | 43.1% (n=152) |
| Secondary | 46.0% (n=17) | 29.7% (n=49)* | 39.7% (n=60) | 35.7% (n=126) |
| Tech/Some college | 5.4% (n=2) | 6.1% (n=10) | 17.2% (n=26) | 10.8% (n=38) |
| College graduate | 10.8% (n=4) | 3.0% (n=5)* | 18.5% (n=28) | 10.4% (n=37) |
| Work Situation | ||||
| Unemployed | 16.2% (n=6) | 21.2% (n=35) | 18.5% (n=28) | 19.5%(n=69) |
| Sex work | 10.8% (n=4) | 0.0%* | 0.66% (n=1)* | 1.4% (n=5) |
| Marital Status | ||||
| Never married | 100% (n=37) | 43% (n=70)* | 92% (n=139) | 69.7% (n=246) |
| Currently Married | 0% | 41% (n=68)* | 5% (n=7) | 21.2% (n=75) |
| Separated/Divorced | 0% | 10% (n=17)* | 1% (n=2) | 5.4% (n=19) |
| Widowed | 0% | 6% (n=10)* | 2% (n=3) | 3.7% (n=13) |
TW = transgender women, CW = cisgender women, and MSM = cisgender men who have sex with men.
p<0.05 compared with TW
Depressive Symptoms
Depressive symptoms were common with mean scores consistent with significant depression in all gender groups (Table 2). Controlling for site, TW had twice the odds of significant depression compared with MSM (OR 2.24 [95%CI: 1.04, 4.80], p=0.039); however, differences between TW and CW did not reach statistical significance.
Table 2.
Depressive Symptom Scores by Gender
| TW | MSM | p value (TW v. MSM) | CW | p value (TW v CW) | |
|---|---|---|---|---|---|
| ALL | n=37 | n=151 | n=165 | ||
| Depression Score (0–60), mean, [95% CI] | 16.8 [13.3,20.4] | 15.8 [13.9, 17.7] | -- | 17.0 [15.1,19.0] | -- |
| Depression cutoff (>16) n (%) | 19 (51%) | 61 (40%) |
OR 2.24* [CI: 1.04, 4.80] p=0.039 |
78 (47%) | OR 1.67*
[CI: 0.79, 3.54] p=0.181 |
| THAI | n=27 | n=47 | n=41 | ||
| Depression Score (0–60), mean, [95% CI] | 13.6 | 11.3 | -- | 13 | -- |
| Depression cutoff (>16) n (%) | 11 (41%) | 14 (23%) | 0.797 | 25 (38%) | 0.092 |
| BRAZIL | n=10 | n=90 | n=99 | ||
| Depression Score (0–60), mean, [95% CI] | 25.6 | 19 | -- | 19.8 | -- |
| Depression cutoff (>16) n (%) | 8 (80%) | 47 (52%) | 0.113 | 53 (53%) | 0.127 |
adjusted for site
Sexual Position and Condom Use
The vast majority (97%) of TW had receptive anal sex with their last partner and had greater odds of doing so than MSM (78%) [OR 11.6; 95% CI: 1.5, 90.6]. TW had reduced odds of insertive anal sex with their last partner (27%) compared with MSM (56%) [OR 0.21; 95%CI: 0.09, 0.51]. (Table 3). TW (22%) had significantly lower odds [OR 0.33, 95%CI: 0.14, 0.80] than MSM (42%) of reporting condom use with partners whose HIV status was unknown; however, in site-stratified analyses, this difference was no longer statistically significant.
Table 3.
Sexual Behavior by Gender
| TW | MSM | OR [95%CI] (TW v. MSM) | CW | OR [95%CI] (CW v. TW) | |
|---|---|---|---|---|---|
| ALL* | n=37 | n=151 | n=165 | ||
| Insertive anal sex with last partner | 27% | 56% |
0.21 [0.09,
0.51] p=0.001 |
-- | -- |
| Receptive anal sex with last partner | 97% | 78% |
11.6 [1.5,
90.6] p=0.019 |
-- | -- |
| Condom use (or no sex) with HIV-negative | 46% | 50% | 0.89 [0.42, 1.85] p=0.748 |
37% | 0.56 [0.26, 1.21] p=0.138 |
| Condom use (or no sex) HIV-unknown | 22% | 42% |
0.33 [0.14,
0.80] p=0.014 |
33% | 1.77 [0.74, 4.27] p=0.201 |
| Disclose to sex partners (last 3 months) | 15% | 19% | 0.88 [0.30, 2.59] p=0.813 |
62% |
9.25 [3.32,
25.79] p<0.0001 |
| THAI | n=27 | n=61 | n=66 | ||
| Insertive anal sex with last partner | 39% | 66% |
0.33 [0.12,
0.90] p=0.032 |
-- | -- |
| Receptive anal sex with last partner | 96% | 71% |
9.3 [1.2,
75.4] p=0.036 |
-- | -- |
| Condom use (or no sex) with HIV-negative | 41% | 57% | 0.52 [0.21, 1.32] p=0.172 |
29% | 0.60 [0.23, 1.53] p=0.281 |
| Condom use (or no sex) HIV-unknown | 27% | 49% | 0.38 [0.14, 1.05] p=0.060 |
32% | 1.29 [0.47, 3.57] p=0.621 |
| Disclose to sex partners (last 3 months) | 22% | 12% | 2.08 [0.56, 7.70] P=0.271 |
60% |
5.4 [1.77,
16.5] p=0.003 |
| BRAZIL | n=10 | n=90 | n=99 | ||
| Insertive anal sex with last partner | 0 | 51% |
0.07 [0,
0.47] p=0.003 |
-- | -- |
| Receptive anal sex with last partner | 100% | 82% | 2.87 [0.42, infinity] p=0.326 |
-- | -- |
| Condom use (or no sex) with HIV-negative | 60% | 45% | 1.81 [0.48, 6.87] P=0.385 |
43% | 0.50 [0.13, 1.88] p=0.302 |
| Condom use (or no sex) HIV-unknown | 10% | 36% | 0.19 [0.20, 1.61] p=0.128 |
34% | 4.57 [0.55, 37.6] p=0.158 |
| Disclose to sex partners (last 3 months) | 0 | 23% | 0.25 [0, 1.70] p=0.183 |
63% |
23.2 [3.56,
infinity] p=0.0002 |
TW = transgender women, CW = cisgender women, and MSM = cisgender men who have sex with men.
OR = odds ratio; [95% CI] = confidence interval
= adjusted for site
Adherence and Detectable Viral Load
Adherence to ART was suboptimal with 50%, 54%, and 70% of CW, MSM, and TW reporting no missed doses in the prior 6 months (Table 4). A greater proportion of TW reported no missed doses compared to other gender groups, but this difference did not reach statistical significance. There were no significant differences in detectable viral load by gender within each site nor overall when adjusted for site (Table 5).
Table 4.
Antiretroviral Therapy (ART) Use and Adherence by Gender
| TW | MSM | OR [95%CI] (TW v. MSM) | CW | OR [95%CI] (CW v. TW) | |
|---|---|---|---|---|---|
| ALL* | |||||
| Ever taken ART | 34/37 (92%) | 111/148 (75%) | 2.02 [0.53, 7.75] p=0.30 |
139/163 (85%) | 0.80 [0.22,3.00] p=0.75 |
| ART in last 3 mo. | 33/34 (97%) | 111/111 (100%) | 0.15 [0.0, 6.0] p=0.27 |
131/139 (94%) | 1.23 [0.02,11.9] p=1.00 |
| No missed doses, 3 mo. | 23/33 (70%) | 59/110(54%) | 1.56 [0.66, 3.72] p=0.31 |
66/131 (50%) | 0.77 [0.31,1.94] p=0.585 |
| Adherence (1–6)* | 4.8 [4.5, 5.0] | 4.5 [4.3, 4.7] | -- | 4.4 [4.2, 4.6] | -- |
| THAI | |||||
| Ever taken ART | 26/27 (96%) | 59/61 (97%) | 0.88 [0.08,10.16] p=0.92 | 62/66 (94%) | 0.60 [0.06,5.59] p=0.65 |
| ART in last 3 mo. | 26/26 (100%) | 59/59 (100%) | *** | 62/62 (100%) | *** |
| No missed doses in 3 mo. | 20/26 (77%) | 38/59 (59%) | 1.84 [0.64,5.30] p=0.26 |
47/62 (76%) | 0.94 [0.32,2.77] p=0.91 |
| Adherence (1–6)* | 4.7[4.4,5.0] | 4.4[4.2,4.5] | -- | 4.3[4.2,4.6] | -- |
| BRAZIL | |||||
| Ever taken ART | 8/10 (80%) | 52/87 (60%) | 2.69 [0.53,13.43] p=0.23 |
77/97 (79%) | 0.96 [0.19,4.89] p=0.96 |
| ART in last 3 mo. | 7/8 (88%) | 52/52 (100%) | 1.22 [0.02, 11.9] p=1.00 |
69/77 (90%) | 1.23 [0.13,11.34] p=0.85 |
| No missed doses in 3 mo. | 3/7 (43%) | 21/51 (41%) | 1.07 [0.22,5.29] p=0.93 |
19/69 (28%) | 0.51 [0.10,2.48] p=0.40 |
| Adherence (1–6)* | 4.9[3.7,6.0] | 4.7[4.4,5.0] | -- | 4.5[4.2,4.8] | -- |
TW transgender women, CW=cisgender women, and MSM=cisgender men who have sex with men. Mo.=months
OR = odds ratio; [95% CI] = confidence interval
Adherence scores range from very poor (1) to excellent (6).
= adjusted for site
Table 5.
Detectable Viral Load (VL) by Gender and Site
| Both Sites | Brazil | Thailand | OR [95%CI] Brazil v. Thailand | |
|---|---|---|---|---|
| Transgender Women | n=30 | n=8 | n=22 | |
| Current Detectable VL (%) | 5 (17%) | 4 (50%) | 1 (5%) |
21.0 [1.8,
240.5] p=0.014 |
| Cisgender MSM | n=129 | n=77 | n=52 | |
| Current Detectable VL (%) | 44 (34%) | 36 (47%) | 8 (15%) |
4.8 [2.01,
11.60] p<0.001 |
| Cisgender Women | n=135 | n=88 | n=52 | |
| Current Detectable VL (%) | 39 (28%) | 31 (35%) | 8 (15%) |
2.99 [1.24,
7.15] p=0.014 |
| ALL GENDERS | ||||
| Current Detectable VL (%) | 88/299 (29%) | 71/173 (41%) | 17/126 (13%) |
4.45 [2.46,
8.08] p<0.001 |
OR = odds ratio; = mean score; [95% CI] = confidence interval
Predictors of Condom Use and Viral Load
On bivariate analysis, depression was not significantly associated with condom use. However, participants with significant depression had 1.6 times the odds [95% CI: 1.00, 2.71] (p=0.05) of having a detectable viral load (Table 6). Among TW, higher depression scores were associated with a detectable viral load [OR 10.4, 95% CI: 1.24, infinity; p=0.03]. The relationship between depressive symptoms and detectable viral load was not significant for CW nor MSM.
Table 6.
Bivariate Associations with Depression (CES-D Score >16) by Gender, OR [95%CI], p-value
| Outcome Variable | Transgender women | Cisgender women | Cisgender MSM | ALL |
|---|---|---|---|---|
| Condom with HIV negative partner | 1.13 [0.31, 4.10] p=0.858 |
0.95 [0.50,1.82] p=0.901 |
0.80 [0.41,1.55] p=0.499 |
0.87 [0.57,1.34] p=0.53 |
| Condom with HIV unknown partner | 0.52 [0.10, 2.61] p-0.427 |
1.18 [0.61,2.31] p=0612 |
0.59 [0.30,1.16] p=0.127 |
0.77 [0.50,1.22] p=0.28 |
| Detectable Viral Load |
10.36 [1.24,
infinity] p=0.028 |
1.23 [0.59, 2.58] p=0.583 |
1.80 [0.85,3.78] p=0.127 |
1.64
[1.00,2.71] p=0.052 |
CES-D = Center for Epidemiology Studies – Depression Scale.
In multivariable logistic regression models, TW had 62% lower odds than MSM of using a condom with a partner whose HIV status was unknown [OR 0.38, 95%CI: 0.16, 0.90; p=0.03] (Table 7). Study site was the sole significant predictor of detectable viral load with a 4-fold higher odds of a detectable viral load at the Brazil site [OR 4.1, 95%CI: 2.2,7.6; p<0.001].
Table 7.
Multivariable Logistic Regression Models for Condom Use and Detectable Viral Load, OR [95%CI]
| Dependent Variables | Condom Use: HIV-negative Partner | Condom Use: HIV-unknown partner | Detectable Viral Load |
|---|---|---|---|
| Gender | |||
| Transgender woman | 0.91 [0.43, 1.92], p=0.807 | 0.38 [0.16, 0.90], p=0.028 | 0.56 [0.19, 1.67], p=0.301 |
| Cisgender woman | 0.60 [0.38, 0.95], p=0.029 | 0.71 [0.44, 1.13], p=0.150 | 0.68 [0.40, 1.18], p=0.169 |
| Cisgender MSM | ref | ref | ref |
| Depression Score | |||
| ≥16 | 0.87 [0.55,1.36], p=0.531 | 0.85 [0.53, 1.36], p=0.504 | 1.33 [0.78, 2.27], p=0.299 |
| <16 | ref | ref | ref |
| Site | |||
| Brazil | 1.18 [0.75, 1.87], p=0.476 | 0.78 [0.49, 1.25], p=0.296 | 4.12 [2.22, 7.62], p<0.001 |
| Thailand | ref | ref | ref |
Discussion
We analyzed depression, sexual risk behavior, and detectable viral load among TW, CW, and MSM living with HIV in Thailand and Brazil. We found that TW had significantly greater odds of high depression scores compared with MSM but not with CW. TW had lower odds than MSM, but not CW, of using condoms with partners whose HIV status was unknown. The proportion with a detectable viral load did not vary significantly by gender. In multivariable regression models that accounted for geographic location, gender, and depressive symptoms, location was the only predictor of a detectable viral load that remained statistically significant.
This analysis was limited by the substantial differences between sites and the small number of TW at each site. Results should therefore be considered hypothesis-generating for future multi-national studies with TW. The finding that 10% of participants in a study designed to recruit MSM actually identified as TW supports the feasibility of recruiting larger samples of TW with tailored strategies. More research is needed to address the understudied HIV outcomes and risk behavior among TW outside of North America.
Acknowledgements
The authors gratefully acknowledge the staff at the HPTN 063 sites for their contributions to the study. We appreciate the HTPN Scholars Program for their generous support of this analysis and all the study participants without whom this study would have been impossible.
Funding: HPTN 063 was funded by the Division of AIDS (DAIDS), National Institute of Allergy and Infectious Disease (NIAID), National Institute on Drug Abuse (NIDA), and the National Institute of Mental Health (NIMH) under Cooperative Agreement # UM1AI068619. The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or The HIV Prevention Trials Network. Dr. Poteat’s work on this manuscript was supported by the Johns Hopkins Center for AIDS Research (JHU CFAR) NIH P30AI094189. Dr. Mayer was also supported by the Harvard University Center for AIDS Research (HU CFAR) NIH P30AI060354 and Dr. Safren by grant 9K24DA040489.
Data Availability Statement
The data that support the findings of this study are available the HIV Prevention Trials Network (HPTN at https://www.hptn.org/research/studies). Restrictions apply to the availability of these data, which were used after approval from the HPTN leadership for this study.
Footnotes
Conflict of Interest: The authors declare that they have no conflict of interest.
References
- Baral SD, Poteat T, Stromdahl S, Wirtz AL, Guadamuz TE, & Beyrer C (2013). Worldwide burden of HIV in transgender women: a systematic review and meta-analysis. Lancet Infect Dis, 13(3), 214–222. doi: 10.1016/S1473-3099(12)70315-8 [DOI] [PubMed] [Google Scholar]
- Bazargan M, & Galvan F (2012). Perceived discrimination and depression among low-income Latina male-to-female transgender women. BMC Public Health, 12, 663. doi: 10.1186/1471-2458-12-663 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brickman C, Propert KJ, Voytek C, Metzger D, & Gross R (2017). Association Between Depression and Condom Use Differs by Sexual Behavior Group in Patients with HIV. AIDS Behav, 21(6), 1676–1683. doi: 10.1007/s10461-016-1610-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Castel AD, Kalmin MM, Hart RL, Young HA, Hays H, Benator D, … Greenberg AE (2016). Disparities in achieving and sustaining viral suppression among a large cohort of HIV-infected persons in care - Washington, DC. AIDS Care, 28(11), 1355–1364. doi: 10.1080/09540121.2016.1189496 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Do AN, Rosenberg ES, Sullivan PS, Beer L, Strine TW, Schulden JD, … Skarbinski J (2014). Excess burden of depression among HIV-infected persons receiving medical care in the united states: data from the medical monitoring project and the behavioral risk factor surveillance system. PLoS One, 9(3), e92842. doi: 10.1371/journal.pone.0092842 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jefferson K, Neilands TB, & Sevelius J (2013). Transgender women of color: discrimination and depression symptoms. Ethn Inequal Health Soc Care, 6(4), 121–136. doi: 10.1108/eihsc-08-2013-0013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Magidson JF, Li X, Mimiaga MJ, Moore AT, Srithanaviboonchai K, Friedman RK, … Safren SA (2016). Antiretroviral Medication Adherence and Amplified HIV Transmission Risk Among Sexually Active HIV-Infected Individuals in Three Diverse International Settings. AIDS Behav, 20(4), 699–709. doi: 10.1007/s10461-015-1142-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nuttbrock L, Bockting W, Rosenblum A, Hwahng S, Mason M, Macri M, & Becker J (2013). Gender abuse, depressive symptoms, and HIV and other sexually transmitted infections among male-to-female transgender persons: a three-year prospective study. Am J Public Health, 103(2), 300–307. doi: 10.2105/ajph.2011.300568 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Poteat T, German D, & Flynn C (2016). The conflation of gender and sex: Gaps and opportunities in HIV data among transgender women and MSM. Glob Public Health, 11(7–8), 835–848. doi: 10.1080/17441692.2015.1134615 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Poteat T, Hanna DB, Rebeiro PF, Klein M, Silverberg MJ, Eron JJ, … Althoff KN (2019). Characterizing the HIV Care Continuum among Transgender Women and Cisgender Women and Men in Clinical Care: A Retrospective Time-Series Analysis. Clinical Infectious Diseases doi: 10.1093/cid/ciz322 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Poteat T, Scheim A, Xavier J, Reisner S, & Baral S (2016). Global Epidemiology of HIV Infection and Related Syndemics Affecting Transgender People. J Acquir Immune Defic Syndr, 72 Suppl 3, S210–219. doi: 10.1097/qai.0000000000001087 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Radloff LS (1977). The CES-D Scale: A Self-Report Depression Scale for Research in the General Population. Applied Psychological Measurement, 1(3), 385–401. doi: 10.1177/014662167700100306 [DOI] [Google Scholar]
- Safren SA, Hughes JP, Mimiaga MJ, Moore AT, Friedman RK, Srithanaviboonchai K, … Mayer KH (2016). Frequency and predictors of estimated HIV transmissions and bacterial STI acquisition among HIV-positive patients in HIV care across three continents. J Int AIDS Soc, 19(1), 21096. doi: 10.7448/ias.19.1.21096 [DOI] [PMC free article] [PubMed] [Google Scholar]