Weekly
August 12, 2011 / 60(31);1045-1049
In the United States, approximately one in three new human immunodeficiency virus (HIV) infections are transmitted via heterosexual contact (1). To monitor HIV risk behaviors and HIV prevalence among heterosexuals and other populations, CDC surveys persons in selected metropolitan statistical areas (MSAs), using the National HIV Behavioral Surveillance System (NHBS). This report summarizes data collected from heterosexuals in 24 MSAs with a high prevalence of acquired immunodeficiency syndrome (AIDS) that participated in NHBS during 2006--2007. Of 14,837 heterosexuals aged 18--50 years who were interviewed and tested, 2.0% were HIV infected. HIV prevalence was higher among those with lower socioeconomic status (SES). For example, HIV prevalence was 2.8% among participants with less than a high school education compared with 1.2% among those with more than a high school education, 2.6% among participants who were unemployed compared with 1.0% among those who were employed, and 2.3% among participants with annual household incomes at or below the poverty level compared with 1.0% among those with incomes above the poverty level. This association between HIV prevalence and SES could not be attributed to factors commonly associated with HIV infection risk in heterosexuals, such as using crack cocaine, exchanging sex for things such as money or drugs, or being diagnosed with a sexually transmitted disease (STD). Based on the association observed between HIV prevalence and SES, HIV prevention activities targeted at heterosexuals in urban areas with high AIDS prevalence should be focused on those with lower SES.
NHBS is an annual cross-sectional survey of three populations at high risk for HIV infection: men who have sex with men (MSM), injection-drug users (IDUs), and heterosexuals at increased risk for HIV infection. Data are collected in annual cycles from one risk group per year, with each population surveyed once every 3 years. This report describes the first NHBS survey among heterosexuals, conducted from September 2006 to October 2007. Twenty-five MSAs with high AIDS prevalence were selected for the survey. In each MSA, NHBS project staff members recruited participants using either respondent-driven sampling (15 MSAs) or venue-based sampling (10 MSAs) (2).* Recruitment efforts targeted residents of census tracts with high rates of poverty and HIV diagnoses, referred to as high-risk areas. For respondent-driven sampling, a small number of initial participants were recruited by project staff members or referred by community-based organizations. Initial and subsequent participants who lived in high-risk areas were then asked to recruit up to five other persons using a coded coupon to track their referrals. Recruitment continued for multiple waves of peer referral.
For venue-based sampling, project staff members from each MSA selected five to 10 high-risk areas in which they identified venues (e.g., retail businesses, social organizations, restaurants, bars, and parks) attended by local residents, as well as the days and times when the venues were frequented. Project staff members then randomly chose venues where they would recruit participants and the days and times when recruitment would occur. At the venues, persons who entered a designated area were approached and invited to participate in the survey. For both recruitment methods, persons were eligible to participate if they were aged 18--50 years, residents of the MSA, able to complete the survey in English or Spanish, and had sex with an opposite-sex partner during the 12 months before interview. Residency in a high-risk area was not an eligibility criterion. After participants provided informed consent, interviewers administered an anonymous survey using a handheld computer. All participants were offered anonymous HIV testing in accordance with CDC and local testing guidelines. Participants were compensated for their time taking the survey ($20--$30) and, when applicable, for taking the HIV test ($10--$25).
Final data were available from 24 MSAs.† Because outcomes did not differ between respondent-driven and venue-based sampling, data were combined and analyzed as a single sample for this report. Univariable and multivariable regression models§ were used to test associations with HIV prevalence and to calculate prevalence ratios, adjusted prevalence ratios,¶ and 95% confidence intervals.
Of 22,169 persons recruited to participate, 18,377 (83%) were eligible and completed the survey. To limit the analysis to non-IDU heterosexuals, persons were excluded if they acknowledged ever injecting drugs (2,224 persons), having male-male sex (413), both injecting drugs and having male-male sex (309), or if they refused to provide this risk information (five). Persons also were excluded if they did not consent to HIV testing (374), did not have a negative or confirmed positive HIV test result (210), or reported being HIV-positive but, when tested, were HIV-negative (five).
Of the 14,837 survey participants who met the analysis criteria, 57% were women, and 48% were aged ≤29 years (Table). The majority of participants were black (72%) or Hispanic** (18%); the remainder were white (5%) or of other races (4%). SES among participants was low; 31% had less than a high school education, 36% were unemployed, 73% had annual household incomes at or below the poverty level,†† and 19% were homeless. In the 12 months before their interview, 11% had used crack cocaine, 12% had exchanged sex for things such as money or drugs, and 14% had received an STD diagnosis.
Overall, 294 (2.0%) of the 14,387 participants tested positive for HIV infection, and HIV prevalence was similar among men (1.9%) and women (2.1%) (Table). HIV prevalence was higher in the Northeast (3.1%) and South (2.7%) compared with the Midwest (0.9%), West (0.8%), and Territories (0.7%). By race/ethnicity, HIV prevalence was highest among blacks (2.1%), followed by Hispanics (1.8%), persons of other races (1.4%), and whites (1.1%). Only the difference between blacks and whites was statistically significant, but after controlling for all other characteristics in the analysis, this difference was no longer significant. Moreover, among the 10,451 (73%) participants who lived in high-poverty areas (i.e., census tracts in which ≥20% of residents had an annual household income below the U.S. poverty level), no significant differences in HIV prevalence by race/ethnicity were observed: Hispanics (2.4%), persons of other races (2.4%), blacks (2.3%), and whites (1.8%) (chi-square, p=0.89).
HIV prevalence was associated with SES. For example, HIV prevalence was higher among participants with less than a high school education (2.8%) compared with high school graduates (1.9%) and those with more than a high school education (1.2%), higher among participants who were unemployed (2.6%) than those who were employed (1.0%), higher among participants with annual household incomes at or below the poverty level (2.3%) compared with those with incomes above the poverty level (1.0%), and higher among participants who were homeless (3.1%) than those who were not (1.7%) (Table). After controlling for the other characteristics in the analysis, HIV prevalence was significantly higher among persons who had less than a high school education (compared with those who had more than a high school education), were unemployed (compared with those who were employed), and had annual household incomes ≤$9,999 (compared with those with incomes of $10,000--$49,999).
By HIV risk factor, HIV prevalence was higher among participants who used crack cocaine (4.5%) compared with those who did not (1.7%), participants who exchanged sex for things such as money or drugs (3.4%) compared with those who did not (1.8%), and participants who had received an STD diagnosis (4.0%) compared with those who had not (1.7%) (Table). However, among these three common HIV risk factors, only an STD diagnosis was associated with higher HIV prevalence after controlling for the other characteristics in the analysis.
Reported by
Paul H. Denning, MD, Elizabeth A. DiNenno, PhD, Ryan E. Wiegand, MS, Div of HIV/AIDS Prevention, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, CDC. Corresponding contributor: Paul H. Denning, pdenning@cdc.gov, 404-639-2963.
Editorial Note
For the first NHBS survey of heterosexuals, described in this report, a high percentage of participants with low SES and high HIV prevalence were enrolled from 24 MSAs. The overall 2.0% HIV prevalence among survey participants is 10 to 20 times the 0.1%--0.2% estimated for all non-IDU heterosexuals in the United States (CDC, unpublished data, 2011). HIV prevalence was higher among those participants with lower SES. Low SES and other adverse social conditions can increase the risk for HIV infection through sexual exploitation, marital instability, unstable sexual partnerships, poor mental health, substance abuse, and limited access to health care and preventive services (4,5). In addition, socioeconomic segregation confines low-SES persons to sexual networks with high underlying rates of HIV and other STDs, thereby further increasing their risk for HIV infection (6).
Among participants in this NHBS survey, racial/ethnic disparities in HIV prevalence were not as great as those found in the overall U.S. population. Nationally, HIV prevalence among blacks (1.7%) is more than eight times that among whites (0.2%), and HIV prevalence among Hispanics (0.6%) is three times that among whites (7). The findings in this report suggest that poverty-related factors might account for some of the racial/ethnic disparities in HIV prevalence observed nationally. Compared with whites, blacks and Hispanics are approximately four times as likely to live in low-income areas such as the ones in the NHBS survey that were shown to have high HIV prevalence (8). When whites live in low-income communities and are exposed to the same socioeconomic conditions and sexual networks as blacks and Hispanics, their risk for HIV infection might be similar to that of blacks and Hispanics.
The findings in this report are subject to at least three limitations. First, because NHBS participants were recruited from 24 urban MSAs with high AIDS prevalence, participants likely are not representative of all low-income heterosexuals in the United States. Second, because the survey targeted census tracts with high rates of HIV diagnoses in addition to high rates of poverty, the former might have led to an overestimation of HIV prevalence in the 24 MSAs. Finally, because of fear of stigma, some participants who said they had not engaged in injection-drug use or male-male sex might actually have done so. Inclusion of IDUs and MSM, who are known to have high HIV prevalence, could have resulted in an overestimation of HIV prevalence. However, of the 18,377 persons who were initially eligible and completed the survey, a large proportion were excluded after acknowledging injection-drug use (14%) or male-male sex (9% of men), making it unlikely that these stigmatized behaviors were markedly underreported.
Based on the association observed between HIV prevalence and SES in the NHBS survey, HIV prevention activities targeted at heterosexuals in urban areas with high AIDS prevalence should focus on those in low-income communities. To reduce new HIV infections, the National HIV/AIDS Strategy§§ calls for intensifying HIV prevention efforts in communities where HIV is most heavily concentrated. The strategy also advocates adopting community-level approaches to prevention in high-risk communities. Structural interventions, which address adverse social, economic, policy, and environmental conditions within communities, have been shown to be effective public health interventions (9,10). The association between HIV prevalence and low SES in the NHBS survey suggests that improvements in educational and employment opportunities in low-income communities, along with concomitant reductions in poverty, could reduce new HIV infections. Without effective approaches to HIV prevention in low-income communities, new HIV infections will continue among these most vulnerable populations.
Acknowledgments
Local National HIV Behavioral Surveillance System staff members Luke Shouse, Laura Salazar, Atlanta, Georgia; Colin Flynn, Frangiscos Sifakis, Baltimore, Maryland; Debbie Isenberg, Maura Driscoll, Elizabeth Hurwitz, Boston, Massachusetts; Carol Cieselski, Nikhil Prachand, Nanette Benbow, Chicago, Illinois; Sharon Melville, Richard Yeager, Jim Dyer, Nandita Chaudhuri, Alicia Novoa, Dallas, Texas; Mark Thrun, Doug Richardson, Beth Dillon, Denver, Colorado; Renee McCoy, Vivian Griffin, Eve Mokotoff, Detroit, Michigan; Marcia Wolverton, Jan Risser, Hafeez Rehman, Paige Padgett, Houston, Texas; Bob Salcido, Jay DiCotignano, SaBrina Hagan-Finks, Las Vegas, Nevada; Trista Bingham, Ekow Kwa Sey, Los Angeles, California; Marlene LaLota, Dano Beck, Stefanie White, Lisa Metsch, David Forrest, Fort Lauderdale and Miami, Florida; Chris Nemeth, Carol-Ann Watson, Nassau-Suffolk, New York; Aaron Roome, Margaret Weeks, New Haven, Connecticut; William Robinson, DeAnn Gruber, New Orleans, Louisiana; Chris Murrill, Samuel Jenness, Holly Hagan, Travis Wendel, New York, New York; Helene Cross, Barbara Bolden, Sally D'Errico, Henry Godette, Newark, New Jersey; Dena Bensen, Judith Bradford, Norfolk, Virginia; Kathleen Brady, Althea Kirkland, Philadelphia, Pennsylvania; Vanessa Miguelino, Al Velasco, Rosana Scolari, San Diego, California; Henry Raymond, Willi McFarland, San Francisco, California; Sandra Miranda De León, Yadira Rolón-Colón, San Juan, Puerto Rico; Maria Courogen, Hanne Thiede, Nadine Snyder, Richard Burt, Seattle, Washington; Yelena Friedberg, Dean Klinkenberg, LaBraunna Friend, St. Louis, Missouri; Tiffany West-Ojo, Manya Magnus, Irene Kuo, Washington, DC.
References
1. CDC. Subpopulation estimates from the HIV incidence surveillance system---United States, 2006. MMWR 2008;57:985--9.
2. DiNenno EA, Oster AM, Sionean C, et al. Piloting a system for behavioral surveillance among heterosexuals at increased risk of HIV in the United States. Open AIDS J. In press.
3. Mancl LA, DeRouen TA. A covariance estimator for GEE with improved small-sample properties. Biometrics 2001;57:126--34.
4. Adimora AA, Schoenbach VJ. Social context, sexual networks, and racial disparities in rates of sexually transmitted infections. J Infect Dis 2005;191(Suppl 1):S115--22.
5. Silver E, Mulvey EP, Swanson JW. Neighborhood structural characteristics and mental disorder: Faris and Dunham revisited. Soc Sci Med 2002;55:1457--70.
6. Poundstone KE, Strathdee SA, Celentano DD. The social epidemiology of human immunodeficiency virus/acquired immunodeficiency syndrome. Epidemiol Rev 2004;26:22--35.
7. CDC. HIV prevalence estimates---United States, 2006. MMWR 2008;57:1073--6.
8. US Census Bureau. Areas with concentrated poverty: 1999. Census 2000 special reports. Washington, DC: US Census Bureau; 2005. Available at http://www.census.gov/prod/2005pubs/censr-16.pdf Adobe PDF fileExternal Web Site Icon. Accessed August 5, 2011.
9. Sumartojo E. Structural factors in HIV prevention: concepts, examples, and implications for research. AIDS 2000;14(Suppl 1):S3--10.
10. Blankenship KM, Bray SJ, Merson MH. Structural interventions in public health. AIDS 2000;14(Suppl 1):S11--21.
* Respondent-driven sampling: Boston, Massachusetts; Dallas, Texas; Denver, Colorado; Detroit, Michigan; Houston, Texas; Los Angeles, California; Nassau/Suffolk Counties, New York; New Haven, Connecticut; New Orleans, Louisiana; New York, New York; Norfolk, Virginia; St. Louis, Missouri; San Diego, California; San Francisco, California; and Washington, DC. Venue-based sampling: Atlanta, Georgia; Baltimore, Maryland; Chicago, Illinois; Fort Lauderdale, Florida; Las Vegas, Nevada; Miami, Florida; Newark, New Jersey; Philadelphia, Pennsylvania; San Juan, Puerto Rico; and Seattle, Washington.
† Data from Norfolk, Virginia could not be analyzed because of a malfunction in the project area's data collection software.
§ Models used marginal Poisson regression and generalized estimating equations. In addition, a variance correction was employed to account for the small number of MSAs in the sample (3).
¶ Controlling for MSA, sex, race/ethnicity, age group, education level, employment status, annual household income, homeless status, crack cocaine use, exchange sex partner, and STD diagnosis.
** All persons who reported Hispanic ethnicity were classified as Hispanic and might be of any race.
†† Additional information available at http://www.census.gov/hhes/www/poverty/data/threshld/thresh07.htmlExternal Web Site Icon.
§§ Available at http://www.whitehouse.gov/administration/eop/onap/nhasExternal Web Site Icon.
- Enviado mediante la barra Google"
.png)
No hay comentarios:
Publicar un comentario