10 Jul 2012: Wagenaar BH, Sullivan PS, Stephenson R (2012) Correction: HIV Knowledge and Associated Factors among Internet-Using Men Who Have Sex with Men (MSM) in South Africa and the United States. PLOS ONE 7(7): 10.1371/annotation/7910bc43-9371-4ad5-aa5d-900161daf089. https://doi.org/10.1371/annotation/7910bc43-9371-4ad5-aa5d-900161daf089 View correction
We compared factors associated with low HIV/AIDS knowledge among internet-using MSM in South Africa and the United States.
1,154 MSM in the US and 439 MSM in South Africa, recruited through Facebook.com, completed an online survey using a US-validated HIV knowledge scale (HIV-KQ-18). Separate multivariable logistic regression models were built, one for the US and one for South Africa, using a dichotomized variable of scoring less than and equal to 13/18 (“low knowledge”) on the HIV-KQ-18 as outcome.
Median knowledge scores were 16/18 for both groups of respondents. For South African MSM, factors associated with low knowledge were: a high school education or less (adjusted odds ratio [aOR]: 2.5, 95% confidence interval [CI]: 1.4–4.6), not using condom-compatible lubrication during last anal sex with another man (aOR: 1.9, CI: 1.0–3.5), number of gay or bisexual acquaintances (aOR: 0.89, CI: 0.81–0.99), being unemployed (aOR: 2.2, CI: 1.0–4.6), and testing HIV negative (aOR: 0.30, CI: 0.16–0.59) or testing HIV positive (aOR: 0.15, CI: 0.03–0.74) compared to those never HIV tested. For US MSM, associated factors were: a high school education or less (aOR: 2.7, CI: 1.9–3.8), low pride and acceptance of homosexuality (aOR: 1.3, CI: 1.2–1.5), age 18–24 (aOR: 2.3, CI: 1.3–3.8) or age 50+ (aOR: 3.2, CI: 1.6–6.3) compared to age 25–29, Hispanic ethnicity compared to white non-Hispanic (aOR: 1.9, CI: 1.1–3.2), and testing HIV positive (aOR: 0.34, CI: 0.16–0.69) or testing HIV negative (aOR: 0.59, CI: 0.39–0.89) compared to those tested.
Those developing programs for MSM in South Africa should weigh these data and other relevant factors, and might consider focusing education services towards MSM with limited education, less integration into gay/bisexual communities, no HIV testing history, limited use of condom-compatible lube, and the unemployed. In the United States, Hispanic MSM, those with limited education, no HIV testing history, low pride/acceptance of homosexuality, and those aged 18–24 or 50+ may be at risk for gaps in HIV knowledge.
Citation: Wagenaar BH, Sullivan PS, Stephenson R (2012) HIV Knowledge and Associated Factors among Internet-Using Men Who Have Sex with Men (MSM) in South Africa and the United States. PLoS ONE 7(3): e32915. https://doi.org/10.1371/journal.pone.0032915
Editor: Javier R. Lama, Asociacion Civil Impacta Salud y Educacion, Peru
Received: October 3, 2011; Accepted: February 7, 2012; Published: March 12, 2012
Copyright: © 2012 Wagenaar et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported, in part, by the National Institute of Allergy and Infectious Diseases (R01-AI094575) and the Emory Center for AIDS Research (P30 AI050409). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Since the emergence of HIV as a global pandemic in the 1980s, men who have sex with men (MSM) have shared a disproportionately large burden of infection in many high-income countries in Western and Central Europe, Australia, and North America . Due to this recognized high burden, MSM represent a large target population for resources on HIV/AIDS prevention, treatment, and research in these areas. By contrast, Africa's HIV/AIDS epidemic has long been understood primarily as a “heterosexual epidemic”, with an estimated 80% of HIV infections being tied to heterosexual transmission . This focus has led to HIV/AIDS prevention efforts in Africa being targeted primarily to heterosexuals.
Recent epidemiological evidence has shown that MSM in Africa share a disproportionate burden of HIV infection . Prevalence estimates of MSM in Africa range from 1–4% of the general population, but high levels of HIV infection and a high prevalence of MSM also engaging in sex with women has led MSM transmission to be linked to over 20% of all HIV cases in several countries of the Middle East, North Africa, and West Africa –. These data are at odds with the fact that most African countries have not dedicated any national HIV/AIDS funds to specifically target HIV/AIDS among MSM .
The 2009 UNAIDS report on universal access for MSM and transgender people highlights the global failure in addressing the needs of MSM regarding HIV/AIDS education, prevention, treatment, research, and care. One of the foci of this report is increasing access to HIV/AIDS prevention materials for MSM and transgender individuals . Although increasing HIV/AIDS knowledge alone is not sufficient to promote sustainable behavior change, accurate knowledge of transmission and prevention of HIV is necessary if MSM are to adopt risk reduction strategies.
Globally, reporting on HIV knowledge among MSM is sparse. Only 33 out of 147 low and middle income countries (LMIC) reported knowledge data through the 2008 United Nations General Assembly Special Session (UNGASS) . Only 2 of these 37 countries reporting UNGASS HIV knowledge data were in Africa, with Nigeria and Mauritius reporting that only 44% and 48% of MSM respectively could “correctly identify ways of preventing sexual transmission of HIV and could correctly reject major misconceptions about HIV transmission” . Additionally, across all low and middle-income countries reporting knowledge scores, less than half of MSM held correct HIV knowledge.
Other studies from Sudan and Kenya indicate that MSM in Africa may have low knowledge regarding HIV prevention and transmission. More than half (55%) of a sample of MSM in Sudan and 35% of respondents in Mombasa, Kenya did not understand the link between anal sex and HIV infection –. By contrast, over 90% of samples of MSM in Malawi, Botswana, and Namibia understood that HIV can be transmitted through anal sex with a man . However, of these samples, only 57%, 50%, and 85% respectively had ever received educational materials on preventing HIV transmission between men.
Data on levels and correlates of HIV/AIDS knowledge in Africa are essential to develop effective prevention and education strategies. Previous studies of HIV knowledge among MSM have focused on levels of HIV knowledge, but have not systematically examined factors associated with low knowledge. The present study aims to fill this gap by examining factors associated with low HIV/AIDS knowledge among MSM in South Africa and the United States using a validated HIV knowledge scale.
We chose to focus on South African and US MSM because: (1) the HIV-KQ-18 knowledge scale has only been previously validated and described among populations in the US, (2) South Africa has a large population of people living with HIV/AIDS, representing 25% of all those living in Sub-Saharan Africa, and (3) there is emerging evidence that MSM in South Africa share a disproportionate burden of HIV infection , . This information will support the development and scale up of educational programs to address the specific needs of MSM, and may aid in optimizing existing prevention programs for MSM.
1,154 internet-using MSM in the United States and 439 in South Africa were recruited through banner advertisements on Facebook.com targeted to men who stated they were interested in men on their Facebook profiles. Facebook ads were displayed in South Africa from June 1 to June 30, 2010 and in all US states from October 1 to November 30, 2010. Banner ads included a range of images, including groups of men, individual men of a range of ethnic and racial backgrounds, and rainbow-themed images. Facebook users who clicked on the banner ads were taken to an internet-based survey. To be eligible to begin the survey men had to report male-to-male sex in the past year.
The survey collected information on participant demographics (age, race, education, and employment), sexual orientation, the number of friends, colleagues, or acquaintances they felt identified as gay or bisexual, HIV testing behavior, knowledge of HIV transmission, condom and water-based lubrication use, and questions on gay identity. HIV knowledge was quantified using the brief HIV knowledge scale (HIV-KQ-18), an internally consistent and stable HIV knowledge scale shown to be appropriate for low-literacy populations . Questions in the HIV-KQ-18 focus on basic HIV transmission and prevention and are summed to form an index of overall HIV knowledge (0–18) with non-responses and “don't know” coded as incorrect.
Gay identity was quantified using a modified version of the US-validated “Gay Identity Questionnaire” consisting of 20 questions summed to form an index of decreasing pride and acceptance of homosexuality as one moves from 0 to 80 . Race in the US was classified into white non-Hispanic, black non-Hispanic, Hispanic, and other (including Asian/Pacific Islander, Native American/Alaska Native, Multi-Racial, and other). Race in South Africa was classified into black African, white/European/African, and other (including Asian, colored, and other). For both groups of respondents, individuals responding “don't know” for previous HIV testing were coded as not being tested and individuals who indicated receiving “indeterminate” HIV results were coded as testing HIV negative.
Statistical analyses were conducted using SAS 9.2 (Cary, NC, USA). Statistical significance was assessed using an alpha value of .05 and two-tailed tests. Individuals were excluded from analyses if they did not report being male and having had male-to-male sex in the past year (n = 15 US; n = 80 South Africa) or if no knowledge questions were answered (n = 883 US; n = 85 South Africa). Individuals excluded for not answering any knowledge questions did not have significantly different age or gay/bisexual peer network distributions using t-tests.
Separate multivariable logistic regression models were built, one for the US and one for South Africa, using the dichotomous “low knowledge” variable as our outcome. Knowledge scores were dichotomized using a data-derived cutoff of greater than or less than and equal to 13/18 correct, with 19.9% and 15.3% scoring at this level or below in the US and South Africa respectively (Figure 1; Figure 2). We dichotomized knowledge this way because the proportional odds assumption was grossly violated for use of ordinal logistic regression with raw scores, and because there is no standard cutoff for a critical level of knowledge for the HIV-KQ-18. As a sensitivity analysis, changing this “critical knowledge” cutoff to other possible cut points of 14, 15, or 16 out of 18 correct on the knowledge scale did not meaningfully change factors associated or their magnitudes.
Variables considered for inclusion into each model were age, race, education, sexual orientation, HIV testing behavior, employment status, number of gay or bisexual friends known, having and/or had a female sex partner, relationship status (male or female), condom and lubrication use, and score on the gay identity scale. Backward selection procedures (α = .05) were used to arrive at the final models. Wald chi-square tests were used to establish significance of individual predictors, whereas likelihood ratio tests (LRT) were used to evaluate significance of groups of predictors (race, HIV-testing behavior). Hosmer and Lemeshow's goodness of fit test was used to determine if final models were adequate.
Basic demographics showed both groups of respondents to be primarily of white race, homosexual sexual orientation, and having had sex with only men (Table 1). Median knowledge scores were 16/18 correct for both cohorts, with 13.4% (59) and 17.1% (197) respondents in South Africa and the United States correctly responding to all 18 knowledge questions, respectively.
For South African MSM, less than 70% of respondents correctly answered that all pregnant women infected with HIV will not have babies born with AIDS, that a natural skin condom does not work better against HIV than a latex condom, and that a person can get HIV from oral sex (Table 2). For US MSM, less than 70% of respondents correctly answered that there is a female condom that can help decrease a woman's chance of getting HIV, and that a natural skin condom does not work better against HIV than a latex condom.
Hosmer and Lemeshow goodness-of-fit tests for each final model revealed no significant lack-of-fit (SA, p = .60; US, p = .95). Sexual orientation, having/and or had a female sex partner, sexual relationship status, and condom use were eliminated from both final models through backward elimination.
Factors Associated with Low Knowledge among Both Cohorts
Controlling for all other factors in the final models, compared to MSM with greater than 12 years of education, men with less than 12 years of education were 2.5 and 2.7 times as likely to score “low” on HIV knowledge in South Africa and the US, respectively (SA, p = .003; US, p<.001; Table 3). Additionally, previous HIV testing as a group predictor was significantly associated with higher HIV knowledge scores (SA, p<.001; US, p = .006). South African MSM testing HIV positive were 85% (p = .02) less likely to score low on HIV knowledge, and men who tested HIV negative were 70% (p<.001) less likely, both compared to men never HIV tested. US MSM testing HIV positive were 66% (p = .003) less likely to score in the lowest quintile of knowledge scores, and men who tested HIV negative were 41% (p = .012) less likely, both compared to men never HIV tested.
Factors Associated with Low Knowledge among South African MSM
First, unemployed South African MSM were 2.2 times as likely to score low on knowledge scores, compared to those who were employed (p = .042). Second, MSM not using condom-compatible lubrication during last anal sex with another man were 1.9 times as likely to score low on HIV knowledge (p = .036). Last, for each ten point increase in the number of gay or bisexual friends known, the odds of scoring low on HIV knowledge decreased 11% (p = .024).
Factors Associated with Low Knowledge among MSM in the United States
For each ten point increase on the gay identity scale (decreasing pride and acceptance of homosexuality) the odds of scoring in the lowest quintile on knowledge scores increased by 30% for US MSM (p<.001). Age as a group predictor was significantly associated with knowledge scores (p = .002). Compared to US MSM age 25–29, those age 18–24 and 50+ were 2.3 (p = .002) and 3.2 (p<.001) times as likely to score in the lowest quintile on knowledge scores, respectively. Finally, race as a construct was significantly associated with the odds of scoring low on HIV knowledge (p = .049). Compared to white non-Hispanics, Hispanics were 1.9 times as likely to score low on HIV knowledge (p = .018).
Among Facebook-using MSM in the US and South Africa, HIV/AIDS knowledge levels were high: the median respondent in both groups of respondents missed only 2 out of 18 questions. In both US and South African MSM, men with less than a high school education had significantly lower HIV/AIDS knowledge. This is in corroboration with a large body of research around the world showing that as general education level increases, so does knowledge of HIV/AIDS prevention and transmission –. Efforts to increase general educational opportunities for MSM in South Africa and the US, would be helpful inherently, and might also support higher HIV/AIDS knowledge.
Second, previous HIV testing was associated with having higher knowledge for MSM in the US and South Africa. Although causality is unclear from our study design, we hypothesize that MSM who have tested for HIV and received negative results gain some knowledge through contact with testing facilities, while individuals testing positive show a trend towards further increased knowledge due to continued contact with healthcare providers, educational materials, or peer groups. According to our data, traditional HIV/AIDS education and counseling associated with voluntary counseling and testing centers (VCT), hospitals, or other care centers is associated with higher levels of HIV knowledge among MSM.
Among South African MSM, having fewer acquaintances gay or bisexual was associated with lower HIV/AIDS knowledge, which could indicate that, in South Africa, peer networks are key avenues where HIV/AIDS information is shared. Additionally, after accounting for HIV testing behavior, not using condom-compatible lubrication during anal sex was associated with lower HIV knowledge in South Africa. South African MSM who receive condom-compatible lubrication from a community based organization or an NGO may also receive educational materials, making use of lubrication use a marker for access to outreach or comfort with service utilization.
For US MSM, there was a trend towards lower HIV knowledge among some racial and age groups. Hispanic MSM had significantly lower HIV/AIDS knowledge and black non-Hispanic MSM showed a trend towards lower knowledge, compared to White non-Hispanic MSM. Thirty years after HIV was first propelled into global consciousness, US MSM of color are at risk for gaps in HIV knowledge at a time when new HIV infections among these men are increasing.
US MSM age 18–24 or 50+ were at significantly increased risk for low HIV knowledge compared to men age 25–29. That these two age groups have the lowest HIV knowledge is troubling because sex with older partners has been shown to be associated with HIV infection, and US MSM age 50+ have an estimated HIV prevalence of 25% compared to 10.5% for men age 18–24 –. Tailored educational interventions, perhaps using social media for young MSM, and other creative methods to reach older MSM are urgently needed.
A novel finding of the current study is that low pride and acceptance of homosexuality was one of the factors most strongly associated with low HIV/AIDS knowledge among US MSM. Low pride and acceptance of homosexuality may correlate with willingness or ability of US MSM to access prevention services or with the extent of peer networks through which HIV education is shared. Additional studies are needed to characterize how gay identity formation in the United States may be related to other HIV-associated factors, especially in light of the recently questioned relationship between internalized homophobia and risky sexual behaviors .
This study has several limitations. First, our findings are not generalizable to all MSM, or all Facebook-using MSM, in either the United States or South Africa. Nevertheless, according to publically available data, 58.6% (136,937,240) of over-18 individuals in the US and 8.5% (3,446,000) of over-20 South Africans have Facebook accounts –. Compared to the general population, our samples over-represent individuals who identify as white in South Africa (91.1% versus 9.6%) and individuals who identify as black in the United States (33% versus 12.6%) , . This indicates that even though our sample is not representative of the population as a whole, internet-based surveys can achieve good penetration into at-risk populations in both the US and South Africa.
Second, since our Facebook advertisements were only targeted to men who explicitly state that they are “interested in” men, our sample likely consists of men who are more open with their sexuality than are MSM in general. Third, recall bias, sample selection bias, and/or social desirability bias could cause underreporting of HIV status, male sex partners, or unprotected sex and could result in misclassification. However, we judge that these biases, if they exist, are not strongly heterogeneous among those scoring high/low on HIV knowledge or between cohorts. Fourth, our choice to dichotomize the knowledge data was empirical; further research is needed to characterize epidemiologically significant critical levels of HIV/AIDS knowledge. Last, the HIV-KQ-18 knowledge scale was not developed or validated for use with MSM or with South Africans, and some of the items, such as the item about natural skin condoms, are somewhat outdated.
Those developing programs for MSM in South Africa should weigh these data and other relevant information, and might consider focusing educational services towards MSM who have lower education levels, less integration into gay or bisexual communities, no HIV testing history, limited use of condom-compatible lube, and who are unemployed. In the United States, Hispanic MSM, those who have low pride and acceptance of homosexuality, those who have not tested for HIV, those with less than a high school education, and those aged 18–24 or over 50 may be at risk for gaps in HIV knowledge.
As we begin to work with recently acknowledged at risk populations, such as MSM in South Africa, or strive to strengthen service delivery in the US, programs must be tailored to populations most at risk. As further research is conducted, our study shows that online surveys are one appropriate way to reach populations of MSM in the US and South Africa.
Conceived and designed the experiments: PSS RS BHW. Performed the experiments: PSS RS BHW. Analyzed the data: BHW. Contributed reagents/materials/analysis tools: PSS RS BHW. Wrote the paper: BHW.
- 1. UNAIDS (2008) 2008 report on the global AIDS epidemic. Geneva, Switzerland. UNAIDS20082008 report on the global AIDS epidemic. Geneva, Switzerland.Available: http://www.unaids.org/en/dataanalysis/epidemiology/2008reportontheglobalaidsepidemic. Accessed 2011 March 29. Available: http://www.unaids.org/en/dataanalysis/epidemiology/2008reportontheglobalaidsepidemic. Accessed 2011 March 29.
- 2. Hunter DJ (1993) AIDS in sub-Saharan Africa: the epidemiology of heterosexual transmission and the prospects for prevention. Epidemiology 4(1): 63–72.DJ Hunter1993AIDS in sub-Saharan Africa: the epidemiology of heterosexual transmission and the prospects for prevention.Epidemiology416372
- 3. Baral S, Sifakis F, Cleghorn F, Beyrer C (2007) Elevated risk for HIV infection among men who have sex with men in low- and middle-income countries 2000–2006: a systematic review. PLoS Med 4: e339.S. BaralF. SifakisF. CleghornC. Beyrer2007Elevated risk for HIV infection among men who have sex with men in low- and middle-income countries 2000–2006: a systematic review.PLoS Med4e339
- 4. Mumtaz G, Hilmi N, McFarland W, Kaplan RL, Akala FA, et al. (2011) Are HIV Epidemics among Men Who Have Sex with Men Emerging in the Middle East and North Africa?: A Systematic Review and Data Synthesis. PLoS Med 8(8): e1000444.G. MumtazN. HilmiW. McFarlandRL KaplanFA Akala2011Are HIV Epidemics among Men Who Have Sex with Men Emerging in the Middle East and North Africa?: A Systematic Review and Data Synthesis.PLoS Med88e1000444
- 5. Caceres CF, Konda K, Segura ER, Lyerla R (2008) Epidemiology of male same-sex behavior and associated sexual health indicators in low- and middle-income countries: 2003–2007 estimates. Sexually Transmitted Infection 84: Suppl 1i49–i56.CF CaceresK. KondaER SeguraR. Lyerla2008Epidemiology of male same-sex behavior and associated sexual health indicators in low- and middle-income countries: 2003–2007 estimates.Sexually Transmitted Infection84Suppl 1i49i56
- 6. Smith AD, Tapsoba P, Peshu N, Sanders EJ, Jaffe HW (2009) Men who have sex with men and HIV/AIDS in sub-Saharan Africa. Lancet 374: 416–22.AD SmithP. TapsobaN. PeshuEJ SandersHW Jaffe2009Men who have sex with men and HIV/AIDS in sub-Saharan Africa.Lancet37441622
- 7. UNAIDS (2009) UNAIDS Action Framework: universal access for men who have sex with men, and transgender people. Geneva, Switzerland: UNAIDS/UNDP. UNAIDS2009UNAIDS Action Framework: universal access for men who have sex with men, and transgender peopleGeneva, SwitzerlandUNAIDS/UNDP
- 8. Adam PCG, de Wit JBF, Toskin I, Mathers BM, Nashkhoev M, et al. (2009) Estimating levels of HIV testing, HIV prevention coverage, HIV knowledge, and condom use among men who have sex with men (MSM) in low-income and middle-income countries. J Acquir Immune Defic Syndr 52(2): S143–S151.PCG AdamJBF de WitI. ToskinBM MathersM. Nashkhoev2009Estimating levels of HIV testing, HIV prevention coverage, HIV knowledge, and condom use among men who have sex with men (MSM) in low-income and middle-income countries.J Acquir Immune Defic Syndr522S143S151
- 9. Geibel S, Luchters S, King'Ola N, Esu-Williams E, Rinyiru A, et al. (2008) Factors associated with self-reported unprotected anal sex among male sex workers in Mombasa, Kenya. Sex Trans Dis 35: 746–52.S. GeibelS. LuchtersN. King'OlaE. Esu-WilliamsA. Rinyiru2008Factors associated with self-reported unprotected anal sex among male sex workers in Mombasa, Kenya.Sex Trans Dis3574652
- 10. Elrashied S (2006) Prevalence, knowledge and related risky sexual behaviors of HIV/AIDS among receptive men who have sex with men (MSM) in Khartoum State, Sudan, 2005. S. Elrashied2006Prevalence, knowledge and related risky sexual behaviors of HIV/AIDS among receptive men who have sex with men (MSM) in Khartoum State, Sudan, 2005.In: AIDS 2006—XVI International AIDS Conference. Toronto, Canada.
- 11. Fay H, Baral SD, Trapence G, Motimedi F, Umar E, et al. (2011) Stigma, health care access, and HIV knowledge among men who have sex with men in Malawi, Namibia, and Botswana. AIDS and Behavior 15: 1088–1097.H. FaySD BaralG. TrapenceF. MotimediE. Umar2011Stigma, health care access, and HIV knowledge among men who have sex with men in Malawi, Namibia, and Botswana.AIDS and Behavior1510881097
- 12. Baral S, Burrell E, Scheibe A, Brown B, Beyrer C, et al. (2011) HIV risk and associations of HIV infection among men who have sex with men in Peri-Urban Cape Town, South Africa. BMC Public Health 11: 766.S. BaralE. BurrellA. ScheibeB. BrownC. Beyrer2011HIV risk and associations of HIV infection among men who have sex with men in Peri-Urban Cape Town, South Africa.BMC Public Health11766
- 13. Carey PM, Schroder KEE (2002) Development and psychometric evaluation of the brief HIV knowledge questionnaire. AIDS Educ Prev 14(2): 172–182.PM CareyKEE Schroder2002Development and psychometric evaluation of the brief HIV knowledge questionnaire.AIDS Educ Prev142172182
- 14. Brady S, Busse WJ (1994) The Gay Identity Questionnaire: a brief measure of Homosexual Identity Formation. J Homosex 26(4): 1–22.S. BradyWJ Busse1994The Gay Identity Questionnaire: a brief measure of Homosexual Identity Formation.J Homosex264122
- 15. Noden BH, Gomes A, Ferreira A (2010) Influence of religious affiliation and education on HIV knowledge and HIV-related sexual behaviors among unmarried youth in rural central Mozambique. AIDS Care 22(10): 1285–1294.BH NodenA. GomesA. Ferreira2010Influence of religious affiliation and education on HIV knowledge and HIV-related sexual behaviors among unmarried youth in rural central Mozambique.AIDS Care221012851294
- 16. Haile BJ, Chambers JW, Garrison JL (2007) Correlates of HIV knowledge and testing: Results of a 2003 South African survey. Journal of Black Studies 38(2): 194–208.BJ HaileJW ChambersJL Garrison2007Correlates of HIV knowledge and testing: Results of a 2003 South African survey.Journal of Black Studies382194208
- 17. UNICEF (2004) Girls, HIV/AIDS and education. UNICEF2004Girls, HIV/AIDS and education.Available: http://www.unicef.org/publications/files/Girls_HIV_AIDS_and_Education_(English)_rev.pdf. Accessed 2011 July 15. Available: http://www.unicef.org/publications/files/Girls_HIV_AIDS_and_Education_(English)_rev.pdf. Accessed 2011 July 15.
- 18. Upreti D, Regmi P, Pant P, Simkhada P (2009) Young people's knowledge, attitude, and behavior on STI/HIV/AIDS in the context of Nepal: A systematic review. Kathmandu University Medical Journal 7(28): 383–391.D. UpretiP. RegmiP. PantP. Simkhada2009Young people's knowledge, attitude, and behavior on STI/HIV/AIDS in the context of Nepal: A systematic review.Kathmandu University Medical Journal728383391
- 19. Morbidity and mortality weekly report (2010) Prevalence and awareness of HIV infection among men who have sex with men – 21 cities, United States, 2008. 59(37): 1201–1207.Morbidity and mortality weekly report2010Prevalence and awareness of HIV infection among men who have sex with men – 21 cities, United States, 2008.593712011207
- 20. Hurt CB, Matthews DD, Calabria MS, Green KA, Adimora AA, et al. (2010) Sex with older partners is associated with primary HIV infection among men who have sex with men in North Carolina. J Acquir Immune Defic Syndr 54(2): 185–190.CB HurtDD MatthewsMS CalabriaKA GreenAA Adimora2010Sex with older partners is associated with primary HIV infection among men who have sex with men in North Carolina.J Acquir Immune Defic Syndr542185190
- 21. Newcomb ME, Mustankski B (2011) Moderators of the relationship between internalized homophobia and risky sexual behavior in men who have sex with men: a meta-analysis. Archives of Sexual Behavior 40(1): 189–99.ME NewcombB. Mustankski2011Moderators of the relationship between internalized homophobia and risky sexual behavior in men who have sex with men: a meta-analysis.Archives of Sexual Behavior40118999
- 22. US Census Bureau (2010) State and county quickfacts. US Census Bureau2010State and county quickfacts.Available: http://quickfacts.census.gov/qfd/states/00000.html. Accessed 2011 July 22. Available: http://quickfacts.census.gov/qfd/states/00000.html. Accessed 2011 July 22.
- 23. Facebook (2011) Targeting your ad. Facebook2011Targeting your ad.Available: https://www.facebook.com/ads/create/. Accessed 2011 July 22. Available: https://www.facebook.com/ads/create/. Accessed 2011 July 22.
- 24. Statistics South Africa (2010) Mid-year population estimates. Statistical release P0302. Statistics South Africa2010Mid-year population estimates. Statistical release P0302.Available: http://www.statssa.gov.za/publications/P0302/P03022010.pdf. Accessed 2011 July 22. Available: http://www.statssa.gov.za/publications/P0302/P03022010.pdf. Accessed 2011 July 22.
- 25. Central Intelligence Agency (2011) The world factbook: South Africa. Central Intelligence Agency2011The world factbook: South Africa.Available: https://www.cia.gov/library/publications/the-world-factbook/geos/sf.html#. Accessed 2012 January 5. Available: https://www.cia.gov/library/publications/the-world-factbook/geos/sf.html#. Accessed 2012 January 5.