Abstract
While HIV is prevalent among adolescents and young adults, testing levels remain low and little is known about gender differences in HIV testing. The objectives of the study were to describe the prevalence of past-year HIV testing, and evaluate associations between HIV testing and individual- and partner-level factors by gender among heterosexually experienced youth (15-24 years) in Baltimore, Maryland (N=352). Past-year HIV testing was prevalent (60.1%) and differed by gender (69.4% among women vs. 49.6% among men, p=0.005). For women, African American race (AOR 3.09), and recent older partner by <=2 years (AOR 4.04) were significantly associated with increased testing. Among men, only African American race was associated with increased testing (OR 4.23), with no patterns identified based on risk behavior or perceived partner risk. HIV testing among adolescent and young adults was prevalent in this highly affected urban area and provide direction for optimizing engagement in HIV testing.
Keywords: Adolescent, HIV testing, gender differences
Introduction
Adolescents and young adults are disproportionately burdened by HIV(MMWR., 2012). Those ages 13-24 account for over one in four new infections (MMWR., 2012), yet only 12.9% of high school students have ever been tested (MMWR., 2012), with 15.7% of 18-24 year olds tested in the past year(CDC., 2013). More than half of the estimated 1.1 million HIV-infected adolescents in the United States are unaware of their status, the highest for any age group.(MMWR., 2012) In this era of effective treatment, it has never been more important to understand patterns and predictors of HIV testing, as it represents the entry point to the system of care.
Youth engagement in HIV testing, or lack thereof, often reflects low risk perception (Murphy, Mitchell, Vermund, Futterman, & Adolescent Medicine, 2002; Peralta, Deeds, Hipszer, & Ghalib, 2007), and has been associated with both individual (e.g., condom use, multiple sexual partners, injection drug use) and partner level risk behavior (e.g., injection drug use, concurrency) (Arrington-Sanders, Ellen, & Trent, 2008; Balaji et al., 2012; Goodman & Berecochea, 1994; Samet, Winter, Grant, & Hingson, 1997; Straub et al., 2011; Tolou-Shams et al., 2007). Gender differences exist in HIV testing, with more young women tested than men (CDC., 2009). This disparity is counter-intuitive given young men’s higher infection rates (32.7 per 100,000 vs. 6.5 per 100,000 women) (CDC., 2011). Gender differences in risk perception and healthcare utilization may play a role. For example, risk perception is associated with testing among both men and women; though men engage in greater sexual risk behavior, their perception of their risk is often lower than that of women (Stein & Nyamathi, 2000). Women appear to test based on perceived partner risk, whereas men are more often symptomatic when tested or diagnosed, suggesting gender differences in testing motivation (Siegel, Lekas, Olson, & VanDevanter, 2010).
In 2010, Baltimore-Towson, MD ranked fourth in the US for adult/adolescent HIV diagnoses (CDC, 2010), and third for the proportion of adults living with HIV (CDC, 2010). The majority of cases are concentrated in urban Baltimore City (Center for HIV Surveillance & Epidemiology., 2011). We describe the prevalence of past-year HIV testing, and evaluate individual- and partner-level factors associated with testing by gender, among a sample of heterosexually experienced youth in Baltimore, MD.
Methods
Sample
This cross-sectional study draws on baseline data from a household survey conducted between February 2011 and May 2013 among heterosexually experienced youth ages 15-24 in Baltimore City, MD (Sherman, Under development). Further details have been described.(Sherman, Under development). A representative sample of low and middle class African American and white participants was recruited from census block groups (CBGs), with oversampling of some racial and SES groupings (n= 352). Following eligibility determination and informed consent (participants under age 18 provided parental consent), participants completed data collection via audio computer-assisted self-interview (ACASI). Procedures were approved by the Johns Hopkins Bloomberg School of Public Health Institutional Review Board.
Measures
All data were self-reported. Past-year HIV testing, was assessed via a single item, “Have you ever been tested for HIV, the virus that causes AIDS?” with a follow-up item to assess testing within the past year. Single items assessed demographics, including age and race; highest level of parental education approximated socio-economic status (SES). Participants described all, if any, of their sex partners in the past 6 months, age at first sex, and engagement in same-sex sexual behavior. Specifically, they described both individual risk behaviors and partner-level characteristics as follows for each sex partner: condom use consistency for vaginal and anal sex, respectively, enabling dichotomous variables for unprotected vaginal, and anal, sex. Index concurrency was defined as having had anal or vaginal sex with anyone else while in a relationship with a given partner. Casual partners were defined as “someone you’ve had sex with only once or a few times or you have sex with on an ongoing casual basis”. Participants self-reported their partners’ ages, enabling calculation of participant-partner age differences of >2 years, and their partners lifetime history of having been STD or HIV infected, ever having sold drugs, injection drug use during the time of their relationship, and ever having been gang-involved, arrested, or incarcerated, respectively.
Analysis
Prevalence of past-year HIV testing was calculated for the entire sample and by gender. Gender differences in HIV testing, demographics, participant sexual risks, and partner characteristics were assessed via logistic regression models. Within gender strata, cross-tabulations of past-year HIV testing based on demographics, participant sexual risks, and partner characteristics were calculated, and associations evaluated via logistic regression models. For females, a multivariate logistic regression model simultaneously evaluated factors significant at p<0.05 in univariate models. The sample size floated to accommodate small amounts of missing data; unweighted counts are provided in Table 1. All analyses were performed using SAS v9.3 (SAS institute INC., Cary, NC), and were weighted to estimate target population totals.
Table 1.
Sample characteristics and prevalence of past-year HIV testing N=352†
| Total (n=352) | Females (n=220) | Males (n=132) | p-value | |
|---|---|---|---|---|
| %†† (n) ††† | %†† (n) ††† | %†† (n) ††† | ||
| Past year HIV testing | 60.1 (216) | 69.4 (153) | 49.6 (63) | 0.005 |
| Obtained results | 97.2 (210) | 98.3 (150) | 95.4 (60) | 0.259 |
| Demographics | ||||
| Age | 0.043 | |||
| 15-18 | 24.0 (79) | 18.5 (46) | 30.4 (33) | |
| 19-24 | 76.0 (272) | 81.5 (174) | 69.6 (98) | |
| Race | 0.542 | |||
| White | 29.5 (122) | 31.3 (70) | 27.5 (52) | |
| African American | 70.5 (228) | 68.7 (148) | 72.4 (80) | |
| Parental education [SES proxy] | 0.077 | |||
| ≤ high school | 52.9 (145) | 58.9 (99) | 46.1 (46) | |
| > high school | 47.1 (203) | 41.1 (118) | 53.9 (85) | |
| Individual Sexual Behavior | ||||
| Sexual Behavior | 0.044 | |||
| Heterosexual | 79.9 (271) | 74.2 (159) | 86.4 (112) | |
| Same sex partner | 20.1 (74) | 25.8 (57) | 13.6 (17) | |
| Age at first vaginal sex | 0.012 | |||
| ≤15 yrs | 53.1 (171) | 44.5 (100) | 63.4 (71) | |
| >15 yrs | 46.9 (163) | 55.5 (113) | 36.6 (50) | |
| # of Partners last 6 months | <.0001 | |||
| 0-1 | 59.9 (236) | 72.5 (160) | 45.6 (76) | |
| ≥2 | 40.1 (116) | 27.5 (60) | 54.4 (56) | |
| Of those in a relationship in the past 6 months (n=282) | ||||
| Index concurrency | 25.0 (64) | 22.5 (38) | 27.5 (26) | 0.486 |
| Unprotected vaginal sex | 73.5 (179) | 76.8 (118) | 69.8 (61) | 0.342 |
| Unprotected anal sex | 68.6 (57) | 72.5 (30) | 66.6 (27) | 0.643 |
| Partner characteristics | ||||
| Any “casual” partners | 51.1 (119) | 32.3 (56) | 68.9 (63) | <.0001 |
| Partner sexual risk | ||||
| Older by >2 yrs | 30.3 (81) | 34.4 (59) | 26.2 (22) | 0.292 |
| Partner concurrency | 26.9 (67) | 31.4 (47) | 22.6 (20) | 0.242 |
| Partner STD/HIV | 12.1 (31) | 12.1 (19) | 12.1 (12) | 0.988 |
| Partner drug-related risk | ||||
| Partner injection drugs | 2.0 (4) | 1.1 (2) | 3.0 (2) | 0.399 |
| Partner sells drugs | 21.1 (57) | 31.6 (47) | 10.5 (10) | 0.002 |
| Partner criminal involvement | ||||
| Partner gang involvement | 11.9 (29) | 13.5 (18) | 10.4 (11) | 0.551 |
| Partner arrest | 29.2 (79) | 41.9 (64) | 16.3 (15) | 0.0007 |
| Partner incarcerated | 18.4 (52) | 32.3 (48) | 4.4 (4) | 0.0002 |
Sample size floats to accommodate small amounts of missing data
Column %
unweighted counts and weighted percentages presented
Results
Overall, 60.1% reported past-year HIV testing, with testing more prevalent among females relative to males (69.4% vs. 49.6% respectively, p=0.005; Table 1). Across both genders, upwards of 95% received their test results. Young age at first sex (≤15) was more common among males (63.4%) than females (44.5%; p=0.012). Significant gender differences included the greater portion of men with recent casual partners (68.9% vs. 32.3%, p<0.001), and greater portions of women reporting partner drug selling (31.6% vs. 10.5%, p<0.002), arrest (41.9% vs. 16.3%, p<0.001), and incarcerated (32.3% vs. 4.4%, p<0.001).
Among females, past-year HIV testing was significantly more common among those aged 19-24 relative to their younger counterparts (72.9% vs. 53.7%, OR 2.32, 95% CI 1.05-5.14; Table 2), among African Americans relative to whites (76.5% vs. 55.7%, OR 2.59, 95% CI 1.22, 5.47), and among low relative to high SES as approximated by guardian education level (75.9% vs. 58.8%, OR 2.21, 95% CI 1.06, 4.61). Women whose recent sexual partner(s) were older by at least two years were significantly more likely to report past-year HIV testing (87.3% vs. 68.3%, OR 3.19, 95% CI 1.05, 9.71). Partner lifetime history of arrest was also associated with past-year HIV testing for women (87.1% vs. 65.2%, OR 3.61, 95% CI 1.31, 9.92). In the multivariate model, African American race (AOR 3.09, 95% CI 1.26, 7.57), low SES (AOR 3.28, 95% CI 1.18, 9.14), and having had a recent partner older by at least two years (AOR 4.04, 95% CI 1.18, 13.8) remained significantly associated with past-year HIV testing. In contrast, among men, African American race was the only factor significantly associated in univariate models with past-year HIV testing, with 58.9% engaging in testing relative to 25.3% of Whites (OR 4.23, 95% CI 1.61, 11.1), thus adjusted models were not pursued. Among males, HIV testing was most prevalent (70.4%) among those reporting same-sex behavior, though not statistically significant likely due to small cell sizes.
Table 2.
Associations of demographics, individual sexual behavior, & partner-level risk with past-year HIV testing†
| Females (N=220) | Males (N=132) | ||||
|---|---|---|---|---|---|
|
| |||||
| % tested | OR (95% CI) | AOR (95% CI) | % tested | OR (95% CI) | |
|
| |||||
| Demographics | |||||
|
| |||||
| Age 15-18 | 53.7 | --- | 34.8 | --- | |
| Age 19-24 | 72.9 | 2.32 (1.05-5.14)* | 1.04 (0.33,3.26) | 57.8 | 2.57 (0.83,7.95) |
|
| |||||
| White | 55.7 | --- | 25.3 | --- | |
| African American | 76.5 | 2.59 (1.22,5.47)* | 3.09 (1.26,7.57)* | 58.9 | 4.23 (1.61,11.1)* |
|
| |||||
| Parental Education [SES proxy] | |||||
| > high school | 58.8 | --- | 47.0 | --- | |
| ≤ high school | 75.9 | 2.21 (1.06,4.61)* | 3.28 (1.18,9.14)* | 52.9 | 1.27 (0.51,3.18) |
|
| |||||
| Individual Sexual Behavior | |||||
|
| |||||
| Same sex partner | |||||
| No | 68.1 | --- | 47.5 | --- | |
| Yes | 76.3 | 1.51 (0.63,3.60) | 70.4 | 2.63 (0.66,10.5) | |
|
| |||||
| Age at first vaginal sex | |||||
| >15 | 73.0 | --- | 53.0 | --- | |
| ≥15 | 68.1 | 1.27 (0.61,2.65) | 44.3 | 1.42 (0.53,3.83) | |
|
| |||||
| No. of partners last 6 mo | |||||
| 0-1 | 67.3 | --- | 45.8 | --- | |
| ≥2 | 74.8 | 1.44 (0.62,3.35) | 52.8 | 1.33 (0.56,3.16) | |
|
| |||||
|
Of those in a relationship in the past 6 months (n=282)
| |||||
| Index concurrency | |||||
| No | 75.8 | --- | 46.7 | --- | |
| Yes | 71.4 | 0.80 (0.29,2.17) | 59.7 | 1.69 (0.54,5.28) | |
|
| |||||
| Unprotected vaginal sex | |||||
| No | 60.9 | --- | 64.2 | --- | |
| Yes | 80.8 | 1.33 (0.54,3.30) | 49.1 | 0.80 (0.30,2.19) | |
|
| |||||
| Unprotected anal sex | |||||
| No | 67.7 | --- | 36.9 | --- | |
| Yes | 80.2 | 1.54 (0.42,5.68) | 57.4 | 1.19 (0.36,3.92) | |
|
| |||||
| Partner characteristics | |||||
|
| |||||
| Any “casual” partners | |||||
| No | 77.8 | --- | 53.7 | --- | |
| Yes | 67.4 | 0.59 (0.24,1.43) | 48.7 | 0.82 (0.33,2.06) | |
|
| |||||
| Partner sexual risk | |||||
|
| |||||
| Older Partner by 2 years | |||||
| No | 68.3 | --- | 47.0 | --- | |
| Yes | 87.3 | 3.19 (1.05,9.71)* | 4.04 (1.18,13.8)* | 63.1 | 1.93 (0.60,6.25) |
|
| |||||
| Partner concurrency | |||||
| No | 73.3 | --- | 48.9 | --- | |
| Yes | 77.5 | 1.25 (0.46,3.38) | 43.4 | 0.80 (0.21,3.09) | |
|
| |||||
| Partner STD/HIV | |||||
| No | 71.0 | --- | 51.9 | --- | |
| Yes | 91.6 | 4.49 (0.55,36.9) | 20.6 | 0.24 (0.05,1.22) | |
|
| |||||
| Partner drug-related risk | |||||
|
| |||||
| Partner sells drugs | |||||
| No | 72.0 | --- | 51.7 | --- | |
| Yes | 79.1 | 1.47 (0.55,3.96) | 31.8 | 0.44 (0.09,2.03) | |
|
| |||||
| Partner criminal involvement | |||||
|
| |||||
| Partner gang involvement | --- | ||||
| No | 72.7 | 1.85 (0.38,9.02) | 49.2 | --- | |
| Yes | 83.1 | 46.5 | 0.90 (0.20,3.97) | ||
|
| |||||
| Partner arrest | |||||
| No | 65.2 | --- | 50.5 | --- | |
| Yes | 87.1 | 3.61 (1.31,9.92)* | 1.98 (0.60,6.51) | 42.1 | 0.71 (0.20,2.49) |
|
| |||||
| Partner incarcerated | |||||
| No | 69.8 | --- | 51.2 | --- | |
| Yes | 87.3 | 2.96 (0.91,9.69) | 0 | ||
weighted to estimate target population totals
p<.05
Conclusion
We identified prevalent past-year HIV testing among young males (49.6%) and females (69.4%) in this high-prevalence urban environment of Baltimore, MD. Consistent with national patterns(CDC., 2009; Chandra A. et al., 2012), women were significantly more likely to have been tested than young men. The low levels of testing among men ages 15-18 in particular suggests the potential value of school-based messaging. Across both men and women alike, testing was significantly more prevalent among African Americans (61% and 75.7%, respectively). Among women, both race and SES were independently associated with testing in the final adjusted model. In contrast, for men, African American race emerged as the sole predictor of HIV testing, with testing levels comparable across SES groupings. Among women, having a sexual partner older by two years or more was also significantly associated with HIV testing, corroborating past research (Siegel et al., 2010). Findings suggest the role of risk perception in women’s HIV testing patterns; in contrast, no individual or partner-level risk behaviors were associated with testing among men.
Current estimates of past-year HIV testing are high relative to national estimates for youth, (CDC., 2009; Chandra A., Billioux V.G., Copen, Balaji, & E, 2012), indicative of modest success in engaging youth in HIV testing in this high-prevalence setting, particularly among African Americans. This higher testing prevalence may reflect recognition of racial disparities in the HIV epidemic or targeted outreach messages. Additional demographic and partner-related risk factors were associated with testing among women but not men, suggesting potential gender differences in HIV testing patterns. Findings suggest the utility of a gendered lens, including gender-stratified analyses, for HIV testing patterns and potentially other dimensions of access, utilization and uptake across the cascade of care. Several additional data elements would aid in interpretation, including testing-related media exposure, and reasons and locations for testing. Participants may lack knowledge concerning their partners’ HIV risk behavior, thus partner-level assessments are considered perceptions. While the household-based sample enables inferences to the target population, generalizability to other settings is unclear. Findings illustrate high engagement in HIV testing among youth in a high-prevalence urban environment, and, simultaneously, significant work remaining in realizing the US Preventive Services Task Force recommendation for HIV screening for all persons ages 15 and over.(Moyer & U. S. Preventive Services Task Force, 2013)
Acknowledgments
This study was supported by the National Institute of Child Health and Human Development (NICHD R01HD057789 ; PI Sherman) and the Johns Hopkins Center for AIDS Research (JHU CFAR; NIAID 1P30AI094189; PI Chaisson).
Footnotes
Competing interests: none declared
References
- Arrington-Sanders R, Ellen J, Trent M. HIV testing in adolescents and young adults receiving STI testing in an urban primary care setting. Sex Transm Dis. 2008;35(7):686–688. doi: 10.1097/OLQ.0b013e31816b1f9b. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Balaji AB, Eaton DK, Voetsch AC, Wiegand RE, Miller KS, Doshi SR. Association between HIV-related risk behaviors and HIV testing among high school students in the United States, 2009. Arch Pediatr Adolesc Med. 2012;166(4):331–336. doi: 10.1001/archpediatrics.2011.1131. [DOI] [PubMed] [Google Scholar]
- CDC. HIV Surveillance Report, 2010. Vol. 22. Atlanta: CDC; 2010. [Google Scholar]
- CDC. HIV testing among high school students—United States, 2007. MMWR. 2009;58:665–668. [PubMed] [Google Scholar]
- CDC. NCHHSTP Atlas. 2011 Retrieved July 1, 2013, from http://www.cdc.gov/nchhstp/atlas/
- CDC. HIV Testing Trends in the United States, 2000-2011. Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention; 2013. [Google Scholar]
- Center for HIV Surveillance & Epidemiology. HIV/AIDS in the Baltimore/Toswon Metropolitan Area: An Epidemiological Profile. Baltimore, MD: Maryland Department of Health & Mental Hygiene; 2011. [Google Scholar]
- Chandra A, Billioux VG, Copen CE, Balaji A, E D. HIV testing in the US household population aged 15-44: data from the National Survey of Family Growth, 2006-2010. Vol. 58. National Center for Health Statistics; 2012. [PubMed] [Google Scholar]
- Goodman E, Berecochea JE. Predictors of HIV testing among runaway and homeless adolescents. J Adolesc Health. 1994;15(7):566–572. doi: 10.1016/1054-139x(94)90140-x. [DOI] [PubMed] [Google Scholar]
- MMWR. Vital signs: HIV infection, testing and risk behaviors among youths - United States. MMWR. 2012;61(47):971–976. [PubMed] [Google Scholar]
- Moyer VA U. S. Preventive Services Task Force. Screening for HIV: U.S. Preventive Services Task Force Recommendation Statement. Ann Intern Med. 2013;159(1):51–60. doi: 10.7326/0003-4819-159-1-201307020-00645. [DOI] [PubMed] [Google Scholar]
- Murphy DA, Mitchell R, Vermund SH, Futterman D Adolescent Medicine, H. I. V. A. R. N. Factors associated with HIV testing among HIV-positive and HIV-negative high-risk adolescents: the REACH Study. Reaching for Excellence in Adolescent Care and Health. Pediatrics. 2002;110(3):e36. doi: 10.1542/peds.110.3.e36. [DOI] [PubMed] [Google Scholar]
- Peralta L, Deeds BG, Hipszer S, Ghalib K. Barriers and facilitators to adolescent HIV testing. AIDS Patient Care STDS. 2007;21(6):400–408. doi: 10.1089/apc.2006.0112. [DOI] [PubMed] [Google Scholar]
- Samet JH, Winter MR, Grant L, Hingson R. Factors associated with HIV testing among sexually active adolescents: a Massachusetts survey. Pediatrics. 1997;100(3 Pt 1):371–377. doi: 10.1542/peds.100.3.371. [DOI] [PubMed] [Google Scholar]
- Sherman SG. The role of ethnicity, class, and gender norms in STI risks among adolescents in Baltimore. Maryland: Under development. [Google Scholar]
- Siegel K, Lekas HM, Olson K, VanDevanter N. Gender, sexual orientation, and adolescent HIV testing: a qualitative analysis. J Assoc Nurses AIDS Care. 2010;21(4):314–326. doi: 10.1016/j.jana.2009.12.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stein JA, Nyamathi A. Gender differences in behavioural and psychosocial predictors of HIV testing and return for test results in a high-risk population. AIDS Care. 2000;12(3):343–356. doi: 10.1080/09540120050043007. [DOI] [PubMed] [Google Scholar]
- Straub DM, Arrington-Sanders R, Harris DR, Willard N, Kapogiannis B, Emmanuel P, et al. Adolescent Trials Network for, H. I. V. A. I. Correlates of HIV testing history among urban youth recruited through venue-based testing in 15 US cities. Sex Transm Dis. 2011;38(8):691–696. doi: 10.1097/OLQ.0b013e318214bb70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tolou-Shams M, Payne N, Houck C, Pugatch D, Beausoleil N, Brown LK Project, S. S. G. HIV testing among at-risk adolescents and young adults: a prospective analysis of a community sample. J Adolesc Health. 2007;41(6):586–593. doi: 10.1016/j.jadohealth.2007.06.004. [DOI] [PubMed] [Google Scholar]
- Turner CF, Rogers SM, Miller HG, et al. Untreated gonococcal and chlamydial infection in a probability sample of adults. Journal of the American Medical Association. 2002;287:726–733. doi: 10.1001/jama.287.6.726. [DOI] [PubMed] [Google Scholar]