By Beth Krane
The level of esteem that Whites hold toward African Americans is linked to the impact of behavioral HIV interventions for African Americans, according to a recent study co-authored by CHIP Principal Investigator (PI) Blair T. Johnson.
While previous research has shown that people who perceive more bias experience poorer mental and physical health, access healthcare less often, and have higher mortality rates, this study breaks new ground because it evaluates the health effects of actual attitudes of Whites’ toward African Americans. The study, published in the journal Social Science & Medicine (DOI: 10.1016/j.socscimed.2013.06.028), also is novel because it examines how the interplay between individual and social variables affects the success of HIV interventions.
“The present research provides support for an influence of structural stigma on the efficacy of interventions for bringing about behavior change, highlighting the role of stigma in the environment in perpetuating health disparities,” wrote Johnson, a UConn professor of social psychology.
Johnson’s team also examined the effects of the level of racial residential segregation in communities and found HIV prevention trials had more success in communities where Whites and African Americans live less segregated from each other and less success with greater segregation.
The study took advantage of a meta-analysis Johnson and his team had published in 2009 in the Journal of Acquired Immune Deficiency Syndromes (JAIDS). This earlier effort used features of the studies themselves in order to try to understand why interventions sometimes succeed and other times fail, but did not examine any community-level variables. The newer effort added information drawn from benchmark polls and economic databases and enabled Johnson’s research team to look at how Whites’ attitudes toward African Americans and levels of racial segregation jointly influenced outcomes in 70 studies conducted between 1988 and 2005. The studies took place at 50 different sites and involved a total of 40,000 participants, the majority of whom were African American.
The main results of the study were:
- Interventions to increase African Americans’ condom use were more successful when Whites’ attitudes toward African Americans were more positive or when racial segregation was low. Conversely, such interventions failed when attitudes were negative or segregation was high.
- The two social or structural factors interacted: Interventions improved condom use only when Whites’ attitudes were positive and racial segregation was low.
- In communities where Whites had more positive attitudes toward African Americans, intervention participants improved their condom use over time, compared to individuals in control groups, but in communities with more negative attitudes, there was no improvement over time, on average.
The negative effects of Whites’ attitudes remained even controlling for variables examined in the original meta-analysis, such as study participant characteristics (e.g., gender and HIV status) and intervention characteristics (e.g., provision of skills training). The effects also held up controlling for community-level characteristics such as levels of education and poverty.
“The stress experienced by African Americans in more negative environments appears to hinder intervention participants’ ability to adopt new, safer, health behaviors,” said CHIP Affiliate Allecia Reid, an assistant professor of psychology at Colby College and lead author of the study.
One approach that buffered study participants from the effects of negative attitudes and high levels of segregation was tailoring intervention content to the values and needs of study participants, perhaps because tailoring helps to reduce African Americans’ mistrust of intervention providers or because tailored interventions are more likely to provide needed skills to cope with the stresses caused by experiencing prejudice and discrimination, Johnson said.
The detrimental effects of Whites’ attitudes on the efficacy of interventions was even more pronounced for African American adolescents, with HIV interventions for this target population failing, on average, in communities with negative attitudes.
“Adolescents may feel the effects of discrimination more acutely because they are still forming their sense of identity and integrating external evaluations into their self-identity,” Johnson said.
For this study, information about Whites’ attitudes toward African Americans was drawn from the American National Election Studies (ANES). The researchers matched the location and year of the studies included in the meta-analysis to the location and year of the ANES studies. Residential segregation was approximated using dissimilarity scores, which reflect the proportion of African Americans who would have to move into different census tracts to obtain even distribution of African Americans across a metropolitan area.
The current study uses a multi-level or structural approach developed by Johnson’s Systematic Health Action Research Program (SHARP). The Network Individual Resource model, a 2010 publication that resulted from the work of 14 scholars led by Johnson, situates behavior in the web of interactions between individuals and their networks, highlighting how the exchange of resources may increase or decrease risk across the lifespans of the individuals and the networks. The model has fared well in empirical tests and was highlighted in a review of multi-level approaches in an article published in July in AIDS.
These research findings may have implications for the success of other behavior change interventions focused on reducing health disparities experienced by African Americans, such as those focused on obesity and diabetes, the authors note.
“Members of stigmatized networks must cope day to day with other networks that often present challenges due to prejudice and discrimination; therefore, it is logical that members of networks living in communities marked by greater stigma will have more difficulty incorporating new health habits whether it relates to HIV risk, physical activity, or nutrition,” Johnson noted.
“It would not surprise me if the effects we found with HIV risk are even larger for some other health behaviors.”
Along with Johnson and Reid, John F. Dovidio, and Estrellita Ballester are co-authors of the article; Johnson took the senior, last, author position. Dovidio, of Yale University, is a CHIP affiliate. Ballester and Reid are both baccalaureate graduates of the psychology department at UConn; they are also former CHIP full-time employees working for SHARP prior to pursuing graduate studies elsewhere.