Family, cultural and gender role aspects in the context of HIV risk among African American women of unidentified HIV status: An exploratory qualitative study

Abstract

This was an exploratory, qualitative study of contextual cultural and social realities of the sexual interactions of a representative sample of African American women of unidentified HIV status. The study expanded our understanding of family and gender role variables by exploring influences of family of origin and idealistic perceptions of roles on sexual relationships. Data was collected on 51 African American women who were recruited through probability sampling. Between 39% and 70% of study participants reported at least one of the following HIV risk factors: low condom use, substance use during sex, partner's incarceration and history of abuse. Nonetheless, all women in our study perceived their chances of HIV infection to be almost non-existent, despite a fairly good knowledge of HIV/AIDS modes of transmission including that anyone could become HIV infected, knowing somebody with HIV/AIDS and acknowledgment, among some, of their partner's infidelity and risk behaviors. Our analysis revealed that parental communications about sexuality in relationships focused largely on trust (being mistrustful of men) and women's control of their sexual impulses. Trust was also emphasized (desired) by women in the discussions of gender roles. Women reported a strong reliance on God and made frequent references to the role of the church in HIV prevention. Our findings offer suggestions for HIV prevention for the general population of African American women. HIV-prevention messages that consider their views of relationships, gender roles, sexual abuse history and the role of the church are suggested.

Acknowledgments

Bradford, Young, and Honnold are members of the research staff at the Survey and Evaluation Research Laboratory (SERL) of Virginia Commonwealth University; at the time of the study, Jarama was also with SERL. Te Survey and Evaluation Research Laboratory conducted this research in conjunction with the Virginia HIV Community Planning Group, with funding from the Centers for Disease Control and Prevention to the Virginia Department of Health. We are grateful to the participants and acknowledge the support and assistance provided by the members of the Virginia HIVCPG and the staff of the Virginia Department of Health, Division of HIV/STD in the planning and implementation of this project. We also acknowledge the efforts of Madge Young and Iva Brown, the study's data collection coordinators. We extend our acknowledgement to Joan Koss-Chioino for her insightful comments to an earlier version of this article.

Notes

1. Women who had been tested for the HIV virus with negative test results were still considered of ‘unknown’ HIV status because: (a) information on a woman's HIV status was through self-report and was not confirmed, (b) date of last HIV-test was not asked and (c) the period between becoming infected with HIV and the point at which antibodies to HIV can be detected in the blood (seroconversion period) can range from two weeks to six months.

2. Microdata allow users to prepare their own customized tabulations and cross tabulations of most population and housing subjects, using specially prepared microdata files. These files are the actual responses to census questionnaires, but with names or addresses removed and the geography sufficiently broad to protect confidentiality. (http://www.census.gov/prod/2001pubs/mso-01icdp.pdf

3. If respondent was an African American woman in the targeted age group, the recruitment process continued. If the respondent was not an African American woman, the interviewer would ask for the oldest (or youngest, depending on sampling quota) woman currently available in the household.

4. Of the 7017 phone numbers in the initial pool, 5,363 were ineligible households/non-working numbers and 1274 were undetermined households.

5. This study is based on semi-structured interviews in which interviewers had latitude to make decisions in how questions were asked and if some questions were appropriate to be asked in a determined situation. The semi-structured nature of the interview can partly explain the small sample size for some variables given that not all questions in the interview instrument were asked to each participant, not all questions were asked in exactly the same way, and not all questions were asked in the same order.