METHODS

Baseline and six month follow up data collected as part of an HIV prevention intervention trial were used to create a prospective study design. The trial tested the efficacy of a culturally based and gender specific education programme on female teenagers sexual risk behaviours. Recruitment sites were located in low income neighbourhoods of Birmingham, Alabama. From 1997–1999 project recruiters screened 1130 female teenagers in diverse community locations (five schools and six health clinics). Adolescents were eligible to participate if they were African-American females, 14–18 years old, unmarried, and reported recent sexually activity (609 were eligible). The study achieved an 85.7% participation rate, with 92% returning at follow up (n=482). The University of Alabama Institutional Review Board approved the study protocol.

Data collection was conducted at the University of Alabama Family Medicine Clinic and included several assessments of sexual risk behaviour (table 1). Each measure shown in table 1 was collected at the six month follow up. Frequency of adolescents' communication about preventing STD/HIV and pregnancy with their sex partners and with their parents was assessed by two 5-item scales with adequate reliability (α=0.80 and 0.88, respectively). Scales assessed adolescents' frequency of discussing sex related issues such as STD, HIV, and pregnancy prevention during the past six months. Response alternatives were “never” (0 times), “sometimes” (1 to 3 times), “often” (4 to 6 times), and “a lot” (7 or more times). Obtained distributions were highly skewed (partner communication mean =8.9, SD=4.5; parent communication mean =17.3, SD=5.2) and subsequently dichotomised by median split.

The predictor variable (adolescents' activity in black social organisations) was assessed at baseline. Adolescents were asked to indicate agreement or disagreement with the following statement “I am active in black organisations or social groups.”

To control for possible confounding effects, we assessed numerous variables to determine whether they were statistically related to the predictor variable as well as any of the outcome variables; none were identified. To control for intervention effects, assignment to condition was entered into a hierarchal logistic regression as the first block. The second block contained the variable representing participation in community organisations. Outcome measures of sexual risk and communication behaviour were each regressed on these two blocks of variables.

RESULTS

Average age of the sample was 16.0 years. Fifty five per cent reported being active in black organisations or social groups. Table 1 displays the obtained prevalence ratios, adjusted odds ratios, and the percentage that reported each risk factor, stratified by participation in black community organisations. As shown, all but three of the outcome measures achieved significance. The exceptions were both measures of vaginal sex unprotected by condoms and inconsistent contraceptive use with casual partners.

COMMENT

These exploratory findings suggest that African-American adolescent females who do not participate in black community organisations may be more likely to have multiple sex partners than adolescents who participate in these organisations. Similarly, protective effects of participation may include a lower likelihood of: engaging in sex with casual partners, inconsistent contraceptive use with steady partners, and infrequent sexual communication with sex partners and parents. Findings warrant further research and suggest that community involvement may be an important intervention modality. Further research may benefit from investigating the effects of unstructured time on adolescents' level of sexual risk behaviour. For example, involvement in organised activities may be protective against engagement in sexual risk based on adolescents' reduced time in unchaperoned circumstances. Structured time may also possibly provide adolescents with a sense of belonging that may preclude seeking social satisfaction in the context of a sexual relationship.

Findings are limited by several factors. For example, adolescents were not provided with a definition of what “active” means in the context of being “active in black organisations or social groups.” Although an imposed definition could have provided greater clarity regarding adolescents' responses, we believe that adolescents' self defined perceptions of what it means to be “active” are important. In addition, it should be noted that our testing for statistical covariates did not yield any potentially confounding variables. Thus, one important aspect of subsequent research may be identification of factors that may predispose adolescents to activity in black organisations and also predispose them to the positive outcomes described in this study. Finally, it should be noted that the predictor variable was assessed using a single item measure. Based on findings from this exploratory study, further empirical investigation of this research question should be expanded to include scale measures that capture adolescents' activity levels in black organisations.