Background: Abstinence-only interventions promote sexual abstinence as the only means of preventing sexual acquisition of HIV; they do not promote safer-sex strategies (e.g., condom use). Although abstinence-only programs are widespread, there has been no internationally focused review of their effectiveness for HIV prevention in high-income countries.
Objectives: To assess the effects of abstinence-only programs for HIV prevention in high-income countries.
Search strategy: We searched 30 electronic databases (e.g., CENTRAL, PubMed, EMBASE, AIDSLINE, PsycINFO) ending February 2007. Cross-referencing, handsearching, and contacting experts yielded additional citations through April 2007.
Selection criteria: We included randomized and quasi-randomized controlled trials evaluating abstinence-only interventions in high-income countries (defined by the World Bank). Interventions were any efforts to encourage sexual abstinence for HIV prevention; programs that also promoted safer-sex strategies were excluded. Results were biological and behavioral outcomes.
Data collection and analysis: Three reviewers independently appraised 20,070 records and 326 full-text papers for inclusion and methodological quality; 13 evaluations were included. Due to heterogeneity and data unavailability, we presented the results of individual studies instead of conducting a meta-analysis.
Main results: Studies involved 15,940 United States youth; participants were ethnically diverse. Seven programs were school-based, two were community-based, and one was delivered in family homes. Median final follow-up occurred 17 months after baseline. Results showed no indications that abstinence-only programs can reduce HIV risk as indicated by self-reported biological and behavioral outcomes. Compared to various controls, the evaluated programs consistently did not affect incidence of unprotected vaginal sex, frequency of vaginal sex, number of partners, sexual initiation, or condom use. One study found a significantly protective effect for incidence of recent vaginal sex (n=839), but this was limited to short-term follow-up, countered by measurement error, and offset by six studies with non-significant results (n=2615). One study found significantly harmful effects for STI incidence (n=2711), pregnancy incidence (n=1548), and frequency of vaginal sex (n=338); these effects were also offset by studies with non-significant findings. Methodological strengths included large samples, efforts to improve self-report, and analyses controlling for baseline values. Weaknesses included underutilization of relevant outcomes, underreporting of key data, self-report bias, and analyses neglecting attrition and clustered randomization.
Authors' conclusions: Evidence does not indicate that abstinence-only interventions effectively decrease or exacerbate HIV risk among participants in high-income countries; trials suggest that the programs are ineffective, but generalizability may be limited to US youth. Should funding continue, additional resources could support rigorous evaluations with behavioral or biological outcomes. More trials comparing abstinence-only and abstinence-plus interventions are needed.