Secular trends in risk behaviour of Cape Town grade 8 students

To compare prevalence rates of selected risk behaviours and age of first intercourse of grade 8 students in Cape Town between 1997 and 2004.

Cross-sectional surveys in 1997 and 2004. Survival analysis was used to estimate the cumulative incidence of first intercourse. The log-rank statistic was used to compare the survival distributions. When comparing data from the two studies we used a logistic regression model with the factors year, race and age group to test the difference in reported risk behaviours between 1997 and 2004 within each gender.

Public high schools in Cape Town.

Multistage cluster samples of 1 437 and 6 266 grade 8 students in 1997 and 2004 respectively.

Ever having had sexual intercourse; for those that had, whether any method was used to prevent pregnancy or disease at last intercourse, and (if so) what was used; use of tobacco, alcohol and marijuana; violence-related behaviours; and suicidal behaviour.

There was a significant delay in first intercourse in 2004 compared with 1997. For males, levels of condom use were lower in 2004 than in 1997, while for females levels of injectable contraceptive use were lower. There were significant increases in past month use of cigarettes for males and marijuana for both genders. Rates of perpetration of violence behaviour remained stable or decreased from 1997 to 2004, while the rate of suicidal behaviour for males increased.

School-based interventions that address sexual risk behaviours should be expanded to include other risk behaviours.