Men Who Have Sex With Men Have a 140-Fold Higher Risk for Newly Diagnosed HIV and Syphilis Compared With Heterosexual Men in New York City
Download the PDF here|
JAIDS Dec 1 2011
"Our analysis triangulated population-based behavioral surveys and HIV and STD surveillance registries and identified alarmingly high rates and rising trends in 2 epidemiologically linked health conditions among MSM.....MSM HIV and P&S syphilis rates were 2526.9/100,000 and 707.0/100,000, each of which was over 140 times MSW rates. Rates were highest among young and black MSM. Over 4 years, HIV rates more than doubled and P&S syphilis rates increased 6-fold among 18-year-old to 29-year-old MSM.We found that 5% or 1 in every 20 sexually active NYC men reported same-sex behavior in the last year.......This is almost double the national prevalence of male same-sex behavior estimated by the 2001-2006 NHANES and 2002 National Survey of Family Growth (both 2.9%).....Estimated HIV and syphilis case rates among NYC MSM and MSW were higher than estimated rates at the national level.3 .....Indeed, compared with NYC MSW, MSM were at least 140 times as likely to be diagnosed with either HIV or P&S syphilis. The much higher HIV diagnosis rates among MSM could be due, in part, to higher testing rates; For both HIV and P&S syphilis, case rates among younger NYC MSM (ages 18-29 years) were approximately double the overall MSM rates and increased dramatically between 2005 and 2008.....our data.....show a continued steep rise in syphilis rates in this group........Black MSM in NYC had disproportionately high rates of both new HIV and P&S syphilis diagnoses, more than 3 times the overall NYC MSM rates."
"Almost two-thirds (62.9%) of MSM reported using a condom at last sex act, compared with 38.3% of MSW......The number of new HIV diagnoses among MSM increased from 2370 in 2005 to 2499 in 2008 (a 5.4% increase), whereas diagnoses among MSW decreased steadily from 369 in 2005 to 258 in 2008 (a 30.1% decrease). During 2005-2008, there were 9571 new HIV cases among MSM and 1249 among MSW, resulting in an MSM HIV case rate that was 140.4 as high (95% CI: 132.1 to 148.7) as the rate among MSW (2526.9/100,000 vs. 18.0/100,000) (Table 3). MSM/MSW rate ratios varied substantially by age and by race/ethnicity. Among MSM, HIV rates decreased with increasing age; the rate for 18-year to 29-year olds (5956.4/100,000) was almost twice the rate among 30-year to 39-year olds, and 4.6 to 6.2 times as high as the rates among MSM aged 40 and older. Among MSW, 18-year to 29-year olds had the lowest HIV rate (12.1/100,000) of all age groups, leading to the highest MSM/MSW rate ratio (490.3). Among MSM, rates were highest for black MSM (8780.7/100,000), which were double that of Hispanic MSM (3520.7/100,000)."
Pathela, Preeti DrPH, MPH*; Braunstein, Sarah L. PhD, MPH; Schillinger, Julia A. MD, MSc*,; Shepard, Colin MD; Sweeney, Monica MD, MPH; Blank, Susan MD, MPH*,*Bureau of Sexually Transmitted Disease ControlBureau of HIV Prevention, New York City Department of Health and Mental HygieneDivision of Sexually Transmitted Disease Prevention, Centers for Disease Control and Prevention, New York City, NY.Correspondence to: Preeti Pathela, DrPH, MPH, New York City Department of Health and Mental Hygiene, Gotham Center, 42-09 28th Street, Queens, New York 11101-4132 (e-mail: email@example.com)
Objectives: To describe the population of men who have sex with men (MSM) in New York City, compare their demographics, risk behaviors, and new HIV and primary and secondary (P&S) syphilis rates with those of men who have sex with women (MSW), and examine trends in infection rates among MSM.
Design: Population denominators and demographic and behavioral data were obtained from population-based surveys during 2005-2008. Numbers of new HIV and P&S syphilis diagnoses were extracted from city-wide disease surveillance registries.
Methods: We calculated overall, age-specific and race/ethnicity-specific case rates and rate ratios for MSM and MSW and analyzed trends in MSM rates by age and race/ethnicity.
Results: The average prevalence of male same-sex behavior during 2005-2008 (5.0%; 95% CI: 4.5 to 5.6) differed by both age and race/ethnicity (2.3% among non-Hispanic black men; 7.4% among non-Hispanic white men). Compared with MSW, MSM differed significantly on all demographics and reported a higher prevalence of condom use at last sex (62.9% vs. 38.3%) and of past-year HIV testing (53.6% vs. 27.2%) but also more past-year sex partners. MSM HIV and P&S syphilis rates were 2526.9/100,000 and 707.0/100,000, each of which was over 140 times MSW rates. Rates were highest among young and black MSM. Over 4 years, HIV rates more than doubled and P&S syphilis rates increased 6-fold among 18-year-old to 29-year-old MSM.
Conclusions: The substantial population of MSM in New York City is at high risk for acquisition of sexually transmitted infections given high rates of newly diagnosed infections and ongoing risk behaviors. Intensified and innovative efforts to implement and evaluate prevention programs are required.
The successful targeting of resources for the prevention and treatment of sexually transmitted diseases (STD), including HIV, benefits from knowledge of the population size and demographic and behavioral characteristics of those at highest risk for infections. Although national and local data have shown that men who have sex with men (MSM) comprise the majority of new HIV and new syphilis cases in the United States, understanding the full burden of disease among the MSM population has been challenging given that, until recently, direct estimates of MSM numbers in the general population were unavailable.
Several recent population-based studies using MSM denominator estimates from behavioral surveillance have quantified point prevalence of HIV1,2 or primary and secondary (P&S) syphilis and HIV rates among MSM.3 Our analysis adds to this body of work by examining trends in newly diagnosed HIV and P&S syphilis among sexually active MSM in New York City (NYC), an epicenter of the US HIV epidemic. In NYC, the proportion of reported male HIV diagnoses that were among MSM increased by 19% in just 5 years.4 MSM have also been disproportionately affected by P&S syphilis since the outbreak began in 1999. In 2008, 87% of male P&S syphilis cases in NYC reported sex with other men.5 To effectively plan, implement, and evaluate programs aimed at preventing transmission of HIV and other STDs, we describe the population of MSM in NYC, compare demographic and behavioral characteristics of MSM and men who have sex with women (MSW), estimate rates of disease in both groups, and examine disparities among MSM by race/ethnicity and age using 3 population-based data sources.
The population of MSM in NYC is large, diverse, and has high rates of risky sexual behaviors. There are substantial differences in the characteristics and newly diagnosed infection rates of MSM, as compared with MSW. Our analysis triangulated population-based behavioral surveys and HIV and STD surveillance registries and identified alarmingly high rates and rising trends in 2 epidemiologically linked health conditions among MSM. The planning, implementation, and evaluation of programs and policy to reduce disease burden among MSM will require continued and innovative efforts.
We found that 5% or 1 in every 20 sexually active NYC men reported same-sex behavior in the last year. This is almost double the national prevalence of male same-sex behavior estimated by the 2001-2006 NHANES and 2002 National Survey of Family Growth (both 2.9%), which included respondents of narrower age ranges but used comparable survey questions.1 Given that previous research has indicated that same-sex behavior is more prevalent in large cities,9 the higher prevalence in NYC is not surprising. With the backdrop of a mature HIV epidemic, decade-long syphilis epidemic, and evidence of ongoing sexual risk behaviors, however, the substantial size of this high-risk population means that high rates of disease could persist for some time.
Consistent with NHANES, we found that among major racial/ethnic groups, the prevalence of same-sex behavior was highest among non-Hispanic white men and lowest among non-Hispanic black men. It has been suggested that relative to white MSM, black MSM may not disclose same-sex behavior due to other sensitizing experiences with prejudice (eg, racism) or concerns about losing connections to their communities, and underreporting may be one explanation for observed racial/ethnic distributions among urban MSM.10,11 Alternatively, the higher proportion of white MSM in urban settings could be related to the phenomenon of seeking out large cities for early-life career opportunities that complement the social advantages of living in gay-friendly environments.12
Estimated HIV and syphilis case rates among NYC MSM and MSW were higher than estimated rates at the national level.3 Rates among NYC MSM were an order of magnitude higher; they had HIV and P&S syphilis case rates that were at least 2.5 and 4 times the national rates, respectively. Although reducing racial disparities in STD/HIV infection has been a national health priority, investigators examining MSM and MSW chlamydia and gonorrhea rates13 have proposed that the disparities paradigm be expanded to include sexual behavior. Indeed, compared with NYC MSW, MSM were at least 140 times as likely to be diagnosed with either HIV or P&S syphilis. The much higher HIV diagnosis rates among MSM could be due, in part, to higher testing rates; CHS data showed that 54% of MSM versus 27% of MSW tested for HIV in the past year. Syphilis diagnoses among MSM could be prompting some additional HIV testing and diagnosis. Disparities by sexual behavior could be driven somewhat by differential testing rates, but clearly MSM in NYC are significantly burdened with STD,14-16 and these disparities warrant extra public health attention and resources.
For both HIV and P&S syphilis, case rates among younger NYC MSM (ages 18-29 years) were approximately double the overall MSM rates and increased dramatically between 2005 and 2008. Although our data indicate a possible leveling off of HIV, they show a continued steep rise in syphilis rates in this group. Differing patterns by disease among young MSM could be explained by the fact that approximately 40% of 18-year-old to 29-year-old MSM with newly diagnosed P&S syphilis have already been diagnosed with HIV and are therefore not at risk for incident HIV infection. This hypothesis is supported by the pronounced divergence in HIV and syphilis trends among non-Hispanic blacks, a group in which HIV prevalence is particularly high. The difference could also be attributable to seroadaptive behaviors (eg, serosorting, serodisclosure) used by some MSM to reduce HIV risk17 or to differential effectiveness of prevention programs on the 2 diseases. Nonetheless, increases in syphilis diagnoses indicate ongoing unprotected sex and signal the potential for increases in HIV acquisition among HIV-uninfected young MSM.
Black MSM in NYC had disproportionately high rates of both new HIV and P&S syphilis diagnoses, more than 3 times the overall NYC MSM rates. The elevated HIV rate may reflect higher HIV testing rates among this subgroup; among black MSM CHS respondents, 65% reported HIV testing in the previous year. However, despite high testing rates, there remains a high risk of ongoing transmission. Young black MSM have reported onset of sexual behaviors with other men at younger ages and have been more likely to engage in unprotected receptive anal intercourse with casual partners than their white counterparts, which are possible contributing factors to the disproportionate rise in new diagnoses.18 Black MSM have been found to be diagnosed relatively late with HIV19-21; indeed, in NYC in 2008, a higher proportion of young black MSM (aged 20-29) were diagnosed with AIDS within 1 month of an HIV diagnosis than young white MSM (11% vs. 7%). Furthermore, once HIV infected, black MSM may be less likely to seek or have access to quality care and treatment than white MSM.22-24
Several features of this analysis may have had an impact on denominator estimates and/or case numbers. First, the 2005-2008 CHS sampled only landline telephones. Exclusion of cell phone-only users could introduce a bias if there were meaningful differences between them and the sampled landline users. However, when the 2009 CHS modified its sampling strategy to include cell phone users, there were no major differences in nearly all CHS indicators between 2009 and previous survey years.25 Second, the calculation of rates requires a precise estimation of the size of the relevant population. Socially stigmatized behaviors such as male-to-male sexual contact may be prone to underreporting in surveys, which could lead to underestimation of the MSM population size and consequent overestimation of HIV and syphilis case rates. A sensitivity analysis in which we used the upper limit of the 2005 CHS estimate for prevalence of male same-sex behavior (7.5%), rather than the average 4-year prevalence (5%), increased our cumulative 4-year MSM population from ~379,000 to ~550,000 and decreased our MSW population to ~6,750,000. MSM and MSW new HIV diagnosis rates were recalculated as 1747.7/100,000 and 18.5/100,000, respectively (rate ratio: 94.5); MSM and MSW P&S syphilis rates were 489.0/100,000 and 4.9/100,000, respectively (rate ratio: 98.9). Thus, MSM still had a substantially higher risk of new infections compared with MSW. Third, per surveillance practices at NYC DOHMH, HIV diagnoses are attributed to MSM if the patient reports any history of sex with men before diagnosis; this diverges from practices in syphilis surveillance and the CHS, which ascertain recent reported sexual behavior. Although the comparatively longer period for classifying sexual behavior used in HIV surveillance could overestimate MSM HIV case numbers relative to MSM denominator estimates, data from field interviews of newly diagnosed HIV cases have demonstrated good correlation between ever and past-year MSM behavior (Chi-Chi Udeagu, personal communication, December 4, 2010). Fourth, this analysis was done using separate databases, with HIV rates calculated using one registry and syphilis rates another. This allowed for comparison of rates by only the crudest demographics. Finally, our MSM population estimate would have included some number of men who had been HIV infected before our study period and inclusion of such nonsusceptible persons would have inflated our denominator, resulting in underestimated new HIV diagnosis rates. This may be especially true for older MSM.
Our approach to redistributing the 42% of P&S cases and 29% of HIV cases without reported transmission risk assumed a similar distribution of cases among those with known and unknown risk. In exploratory analyses, we found that white syphilis cases and black and older HIV cases were more likely than other subgroups to be missing risk information. If the proportion of MSM among groups missing this information were in fact different from those for whom it is known, case rates based on our redistribution methods could be overestimated or underestimated. Case rates could be conservative if same-sex behavior was underreported by cases with known risk and some proportion of true MSM cases was misclassified as MSW. For instance, in sensitivity analyses in which we reallocated the proportion of male cases with unknown risk so that all of them were classified as MSM, the new HIV diagnosis rates among MSM and MSW became 2622.7/100,000 and 12.9/100,000, respectively (rate ratio: 203.0); MSM and MSW P&S syphilis case rates changed to 745.1/100,000 and 2.8/100,000, respectively (rate ratio: 267.4). Alternatively, adjusting the proportion of cases with unknown risk so that only 50% were classified as MSM and 50% as MSW yielded new HIV case rates of 2204.7/100,000 and 36.0/100,000 among MSM and MSW, respectively (rate ratio: 61.2), and MSM and MSW P&S syphilis rates of 575.4/100,000 and 12.2/100,000 (rate ratio: 47.3). Results of these sensitivity analyses represent a plausible range within which the true numerators likely lie.
We examined HIV and P&S syphilis epidemics in NYC and found significant overlap between them, with young and non-Hispanic black MSM disproportionately affected by both conditions. It is critical to revise current data systems to enable integration of HIV and STD case data; this would allow for better identification and characterization of the population affected by these synergistic epidemics and would facilitate analyses to drive additional research, programming, and policy. Reducing health disparities associated with sexual behavior and STD is a complex task. A critical step is improving STD/HIV screening and testing coverage to increase the number of persons who are aware of their infections, which can lead to behavior change, decrease secondary transmission, and link infected individuals to care, treatment, and prevention services.26 Centers for Disease Control and Prevention's 2010 STD Treatment Guidelines call for annual STD/HIV screening for sexually active MSM and more frequent screening for MSM who report high-risk behaviors.27 To achieve this, it is imperative that providers are skilled in identifying MSM. A substantial proportion of MSM do not identify as gay. Therefore, providers need to be competent at taking a nonjudgmental sexual history that ascertains sex of sex partners and specific sexual practices with all partners.
Prevalence and Population Size of MSM
The estimated numbers of males aged 18-64 years living in NYC were 2,510,677 in 2005; 2,533,183 in 2006; 2,583,187 in 2007; and 2,607,031 in 2008.8 Based on an MSM prevalence of 6.1% (95% CI: 4.9 to 7.5) in 2005, 4.7% (95% CI: 3.8 to 5.8) in 2006, 5.0% (95% CI: 4.0 to 6.2) in 2007, and 4.6% (95% CI: 3.5 to 6.0) in 2008, there were an estimated 112,000 MSM in 2005; 88,000 MSM in 2006; 89,000 MSM in 2007; and 90,000 MSM in 2008. The average prevalence of male same-sex behavior for years 2005-2008 (5.0%; 95% CI: 4.5 to 5.6) was highest among men aged 40-49 years (8.0%) and lowest among men aged 18-29 years (3.9%) (Table 1). Prevalence of male same-sex behavior also differed by race/ethnicity, ranging from 2.2% and 2.3% among non-Hispanic Asian and black men, respectively, to 7.4% among non-Hispanic white men.
Demographic and Behavioral Characteristics
MSM were significantly different from MSW with regard to all demographic variables examined (Table 2). Unlike MSW, MSM had a skewed age distribution, with 38.0% aged 40-49 years. Among MSM, the majority (58.8%) were white. MSM were more likely than MSW to be US-born, have a higher education, and lower poverty level (all P < 0.001). A sizeable proportion of MSM (14.6%) reported a heterosexual/straight sexual identity. Although the subpopulation of black MSM was relatively small and yielded an unstable estimate for straight identity, more black MSM (24.3%; 95% CI: 12.1 to 42.7) and Hispanic MSM (18.8%; 95% CI: 11.3 to 29.7) reported being straight, compared with white MSM (9.7%; 95% CI: 5.6 to 16.2).
MSM were more likely than MSW to report good to excellent general health but also more likely to report poor mental health status and diagnoses of depression (26.0% vs. 10.5%, P < 0.001) (Table 2). MSM were significantly more likely than MSW to have a primary health care provider and to have visited their provider in the past year.
MSM were more likely than MSW to have been HIV tested, both in the last year (53.6% vs. 27.2%, P < 0.001) and ever (84.8% vs. 59.7%, P < 0.001) (Table 2). Among MSM, 18-year to 29-year olds (66.0%) were more likely than 30-year to 39-year olds (47.4%, p=0.033) and 50-year to 64-year olds (46.8%, P = 0.022) to have HIV tested in the last year. Self-reported HIV testing in the last year was 65.5% among black MSM (95% CI: 51.6 to 77.1), 61.2% among Hispanic MSM (95% CI: 50.3 to 71.1), and 51.1% among white MSM (95% CI: 42.9 to 59.3).
MSM reported more sex partners in the past year than MSW; 29.1% of MSM and 8.7% of MSW reported ≥4 partners (Table 2). Likelihood of reporting multiple partners among MSM did not vary by age or race/ethnicity. Almost two-thirds (62.9%) of MSM reported using a condom at last sex act, compared with 38.3% of MSW (Table 2). Among MSM, condom use at last sex was more prevalent among 18-year to 29-year olds (65.0%) than among 30-year to 39-year olds (54.9%, P = 0.040) and 50-year to 64-year olds (52.3%, P = 0.017), but did not vary among white, black, and Hispanic MSM.
HIV and Primary and Secondary Syphilis Rates
The number of new HIV diagnoses among MSM increased from 2370 in 2005 to 2499 in 2008 (a 5.4% increase), whereas diagnoses among MSW decreased steadily from 369 in 2005 to 258 in 2008 (a 30.1% decrease). During 2005-2008, there were 9571 new HIV cases among MSM and 1249 among MSW, resulting in an MSM HIV case rate that was 140.4 as high (95% CI: 132.1 to 148.7) as the rate among MSW (2526.9/100,000 vs. 18.0/100,000) (Table 3). MSM/MSW rate ratios varied substantially by age and by race/ethnicity. Among MSM, HIV rates decreased with increasing age; the rate for 18-year to 29-year olds (5956.4/100,000) was almost twice the rate among 30-year to 39-year olds, and 4.6 to 6.2 times as high as the rates among MSM aged 40 and older. Among MSW, 18-year to 29-year olds had the lowest HIV rate (12.1/100,000) of all age groups, leading to the highest MSM/MSW rate ratio (490.3). Among MSM, rates were highest for black MSM (8780.7/100,000), which were double that of Hispanic MSM (3520.7/100,000).
P&S syphilis cases among both MSM and MSW increased steadily, from 496 in 2005 to 895 in 2008 among MSM (an 80.4% increase), and from 84 in 2005 to 114 in 2008 among MSW (a 35.7% increase). The total number of P&S syphilis cases over 4 years was 2678 among MSM and 334 among MSW, resulting in an MSM P&S syphilis case rate that was 147.3 times as high (95% CI: 130.5 to 163.2) as the rate among MSW (707.0/100,000 vs. 4.8/100,000) (Table 3). P&S syphilis rates among MSM followed the same pattern as HIV rates, wherein they decreased with increasing age and were highest among black (2,375.9/100,000), followed by Hispanic MSM (754.1/100,000), who had approximately one-third the rate of black MSM. In every age group, MSM had more than 100 times the rate of P&S syphilis compared with MSW. MSM/MSW rate ratios by race/ethnicity ranged from 183.0 (black) to 255.7 (white).
MSM HIV and Syphilis Trends
Figure 1 shows rates among MSM by age group from 2005 through 2008. There were marked increases in HIV diagnoses among 18-year-old to 29-year-old MSM, with case rates more than doubling from 3078.7/100,000 (95% CI: 2875.7 to 3281.6 per 100,000) in 2005 to 8870.0/100,000 in 2008 (95% CI: 8342.7 to 9397.3 per 100,000) (Fig. 1A). Although the P&S syphilis rate among 18-year-old to 29-year-old MSM in 2005 was lower than the HIV rate that year, it increased much more dramatically, increasing 6-fold to reach 2900.4/100,000 in 2008 (95% CI: 2598.6 to 3202.1 per 100,000) (Fig. 1B). During this period, both HIV and P&S syphilis case rates among 30-year-old to 39-year-old MSM peaked in 2007.
Figure 2 shows rates by race/ethnicity category from 2005 through 2008. After peaks in 2006 and 2007 for Hispanic and black MSM, respectively, HIV rates remained relatively steady (Fig. 2A). P&S syphilis rates increased dramatically for black and Hispanic MSM from 2005 to 2007, with the rate among Hispanic MSM increasing 6-fold, whereas the rate rose steadily for white MSM (Fig. 2B).