Study Design

To assess program impact, the researchers recruited 780 young men, 14 to 25 years old, to participate in the study. The young men were recruited from local schools and community organizations in three similar low-income communities, or favelas, characterized by high levels of violence and drug trafficking. On average, the men in the sample were 17 years old and most had not completed primary school.

Study participants in two of the communities were exposed to different combinations of activities. In Bangu, adult male facilitators used the curriculum to conduct two-hour interactive group sessions once a week over a six-month period with the recruited young men in the community. Bangu was also the site where the lifestyle social marketing campaign was implemented. In a second community, Maré, only interpersonal group sessions were conducted.

A third community, Morro dos Macacos, did not receive any activities during the first six months of the study period and thus served as a control. This helped the researchers determine whether any changes detected among the young men in the other two communities were likely due to the interventions or to other factors or events.

In each community the young men recruited for the study were surveyed prior to the start of any activity (baseline) and after the intervention activities had been ongoing for six months. Young men in the two intervention sites also completed surveys after one year. To complement the survey data, the researchers conducted individual in-depth interviews with a sub-sample of young men and their sexual partners to ascertain their views about the interventions and the impact on their relationships.

Follow-up rates for this longitudinal study were good in the intervention sites: 85 percent of respondents recruited at baseline in Bangu and 76 percent in Maré completed follow-up surveys at one year. In the control community, Morro dos Macacos, the follow-up rate was less—66 percent at six months—reflecting a commonly observed phenomenon of a dip in response rates in sites where there is no intervention.

Other study limitations aside from the relatively lower response rate in the control group include the lack of biological markers (e.g., prevalence of STIs) to corroborate reported data, such as condom use and STI symptoms, and that the young men in each community were self-selected rather than being chosen at random.

The GEM Scale

Each survey used an innovative series of questions to find out respondents’ views about appropriate roles and behavior for men and women. The Gender Equitable Men or GEM Scale was developed using findings from qualitative and quantitative research with men in Rio de Janeiro and from a review of publications on studies about men and gender roles. The resulting scale used in the study consists of 17 items and is designed to measure the extent to which respondents support traditional notions about men’s and women’s characteristics and behaviors related to HIV and pregnancy prevention, violence, sexual relationships, domestic chores and caregiving, and homosexuality (see sidebar).

“The scale quantitatively measures changes in support for prevailing gender norms,” explained Dr. Julie Pulerwitz of Horizons/PATH, one of the study’s principal investigators. “These data, when combined with qualitative information, can give researchers a good idea whether norms about gender roles are shifting within a community or sub-population, such as young men.”

In addition to sociodemographic information, the surveys also collected data from respondents about their level of HIV-related risk by asking about STI symptoms, condom use, and number of sexual partners.

Risky Behaviors

The researchers analyzed behavioral data from respondents at baseline to determine the extent to which respondents were at risk of HIV prior to the implementation of the interventions.

They found that across all three study sites, the young men typically engage in a number of risky sexual behaviors. More than 70 percent of respondents were sexually experienced, with an average age of 13 for sexual initiation. Almost one-third (30 percent) of sexually experienced youth had more than one sexual partner in the previous month. Twenty-five percent of respondents reported at least one STI symptom (e.g., penile discharge, pain during urination, etc.) during the three months prior to the survey. About ten percent of the young men reported ever being physically violent with their current or most recent regular partner. Fewer than 10 percent had ever taken an HIV test.

Less Support for Traditional Norms

Analysis of the data indicates that the group intervention alone and in combination with the social marketing campaign had a positive impact on young men’s attitudes about gender roles.

In both intervention sites, young men were less likely to support traditional gender norms as measured by overall GEM scale scores at the six-month follow-up survey compared to baseline figures, and the differences were statistically significant. In the control community, no significant improvement in the GEM scale scores was detected for the same time period. Moreover, the positive changes were maintained at the one-year follow-up in both intervention sites.

“For many young men, the workshops were the first time they had the opportunity to discuss in depth the costs of being a man in our society,” said Marcio Segundo of Instituto PROMUNDO and the study’s research coordinator. “The activities are very creative and participatory and have helped young men think critically about issues that affect their lives, such as sex, HIV, relationships, violence, drugs, and fatherhood.”

In-depth interviews with participants echoed this observation. Many young men highlighted how the group sessions helped them question traditional views about manhood and made them more sensitive about women’s needs. According to one young man, “I learned to talk more with my girlfriend. Now I worry more about her… it’s important to know what the other person wants, listen to them. Before [the workshops], I just worried about myself.”

Reduced Risk

The researchers found improvements in key HIV-related outcomes among the intervention groups, while the control group showed either smaller improvements or changes in the wrong direction. For example, condom use at last sex with a primary partner increased over a one-year period in both intervention sites; however it decreased slightly in the control group at the six-month follow up (see figure 1).

Reported prevalence of one or more STI symptoms decreased among all groups; however the reductions were statistically significant between baseline and six months only in Bangu, the site of the combined intervention, and between baseline and one year in Bangu and Maré, the other intervention site (see figure 2).

The interventions did not seem to have an effect on condom use with casual partners, perhaps because levels were already high at baseline. There was also little impact on number of sexual partners, although there was a downward trend among the intervention groups, whereas a slightly greater proportion of respondents in the control group reported multiple partners in the previous month.

Gender Matters

The study also sought to find out whether attitudes about gender influence young men’s risk of HIV and whether Program H, an intervention that focuses on gender dynamics, could reduce their risk.

To answer these questions, the researchers first examined the association between support of traditional gender norms and HIV risk variables at baseline. They found that indeed there is a relationship. For example, young men who scored in the lowest third of the GEM scale (less support for traditional norms) were significantly less likely to report STI symptoms than those who scored in the highest third (greater support for traditional norms).

“For those young men who buy into rigid views about manhood,” explained Dr. Gary Barker, director of PROMUNDO and co-principal investigator on the study, “we found that they have a higher level of HIV risk.”

A similar association between attitudes toward gender norms and HIV risk variables was also found over time. For the group in Bangu (the site of the combined intervention), those who became less supportive of traditional norms were more than four times less likely to report STI symptoms (p < .05) at one-year follow up compared to baseline; in Maré, they were more than 8 times less likely to report STI symptoms over the same time period (p < .05).

Findings from the in-depth interviews support survey results on the importance of addressing gender norms in HIV programming. Young men spoke about how the group sessions changed how they thought about sex and relationships, which can affect their risk of HIV infection. “Used to be when I went out with a girl, if we didn’t have sex within two weeks of going out, I would leave her. But now [after the workshops], I think differently. I want to construct something [a relationship] with her,” said one respondent.

“Young men don’t learn behaviors in isolation,” said Dr. Barker. “Social norms play an important role, and from the study we learned that individual reflection can help change their views, which is a first step in changing what is appropriate and expected behavior for men.”

Conclusions and Next Steps

For policymakers and program implementers concerned about gender inequality and HIV prevention among young men, the study’s findings have important implications. First, the gender-based interventions for young men examined in this study were shown to be feasible and acceptable. Second, the interventions fostered diminished support for gender inequitable norms among the young men in the sample. And third, that addressing HIV-related themes from a gender perspective can reduce young men’s level of HIV risk.

The researchers now are conducting further analysis of the data that will provide more information about participation in the group sessions, the key issues raised during the group discussions, and the costs and relative merits of the different intervention combinations. The final report on the study will be available in early 2005.

Instituto PROMUNDO and Program H partners currently are working on developing new manuals and cartoon videos to complement this work. One set focuses on reducing homophobia among heterosexual youth, and the other focuses on young women’s empowerment, including reflection about gender roles and manhood.

Instituto PROMUNDO also is working with Salud y Género to evaluate Program H in Mexico and with JSI Research and Training Institute to test Program H in the U.S. with Hispanic populations. In addition, Instituto PROMUNDO is working with Horizons and others to adapt Program H in India (see page 10).

This article was written by Sherry Hutchinson and Ellen Weiss in conjunction with the research team, which includes Dr. Gary Barker and Marcio Segundo of Instituto PROMUNDO, and Dr. Julie Pulerwitz of Horizons/PATH. Information about Program H can be found on Instituto PROMUNDO’s website www.promundo.org.br. For more information about this study, contact Dr. Julie Pulerwitz (jpulerwitz@popcouncil.org) or Dr. Gary Barker (g.barker@promundo.org.br). To read more about this study go to www.popcouncil.org/pdfs/horizons/brgndrnrmsru.pdf. To receive e-mail notification when reports or summaries about this and other studies become available, sign up at www.popcouncil.org/horizons/signup.html.