Longitudinal risk of std acquisition in adolescent girls using a generalized estimating equations model

Abstract

Background: To date, few studies have used a longitudinal approach in examining risk factors for STD acquisition in teenage girls. In the present study, we examined risk of STD acquisition in adolescent girls over a three-year period using longitudinal analyses.Methods: One hundred and seventy-four girls between the ages of 12 and 16 (mean age = 14.5) participated in a longitudinal study of adolescent romantic relationships. The racial composition of the subject pool was 80% African-American and 20% Caucasian. These girls were followed for three years at six-month intervals. By the end of the study, 127 of the girls were sexually experienced, and thus, were included in the analyses regarding risk for STDs. The risk for history of STD was evaluated using a generalized estimating equations (GEE) model, which allows for the inclusion of data from subjects with missing visits. The following independent variables were entered into the model: age, race, screening IQ score, qualitative cognitive functioning, perceptions of STD prevalence among female friends, and whether a condom was used at last intercourse. Additional independent variables included number of lifetime partners (4 or more lifetime partners, 2 to 3 lifetime partners, only one partner) and age of sexual debut (less than 14 years, 14-16 years, and 17 years and older), five factors from the Family Environment Scale, and three types of parental monitoring (direct, direct when with peers, and indirect).Results: The results of the longitudinal (GEE) model indicated that having a history of an STD was significantly related to a younger age of sexual debut (p <.01), having more partners (p <.0001), being African-American (p =. 01), and having a lower screening IQ score (p <.01).Conclusions: These results suggest that risk of STD is associated with both modifiable and non-modifiable variables. Race and intellectual functioning independently contributed to STD risk, stressing the importance of utilizing prevention programs that are culturally and developmentally specific. For example, less intelligent girls may need to have information presented in a more concrete manner, with a focus on the present rather than the future. The targets of these prevention programs should continue to emphasize delaying initiation, as well as partner selection. Finally, the fact that variables for STD acquisition in this longitudinal model are comparable to previous cross-sectional studies of STD acquisition suggests that these are potent predictors of risk.