IMPACT ON SEXUALLY-TRANSMITTED DISEASE SPREAD OF INCREASED CONDOM USEBY YOUNG FEMALES, 1987-1992

Background Sexual behaviour data are crucial for understanding the spr ead of sexually transmitted diseases (STD). Over a period (1987-1992) in which the HIV epidemic increased public awareness of safe sexual pr actices, we describe predictors of condom use, changes in condom use o ver time, and the estimated effects of these changes on the spread of STD. Methods. Condom use reported by females aged 18-35 years with non -cohabiting partners was analysed using data from two cross-sectional postal surveys performed 5 years apart (1987 and 1992) on two separate representative samples of 10 000 subjects aged 18-60 years living in Norway. A simple mathematical model was used to assess the effects of selection bias. A more complicated model was used to predict the effec ts of condom use on the prevalence of STD in a population which includ es a core group of highly sexually active subjects. Results. We found an increase in the prevalence of condom use in the latest intercourse from 14% to 20% with non-foreign partners and from 10% to 38% with for eign partners, from 1987 to 1992. In a logistic regression model, low frequency of intercourse, high education, one lifetime partner, and la te sexual debut were predictors for condom use. Controlled for these v ariables, the odds ratio (OR) for condom use in 1999 Versus 1987 was 1 .4 (95% confidence interval [Cl]: 0.9-2.0) ii the partner was non-fore ign, and 7.1 (95% CI:2.5-20.5) if the partner was foreign. Not using o ther contraceptive methods was a strong predictor for condom use; OR = 17.4 (95% CI :8.0-38.0). Condom use in the first intercourse with the last partner was a strong predictor for condom use in the last interc ourse; OR = 19.2 (95% CI :8.2-45.3). It appeared unlikely that the inc rease in condom use could be explained by response bias. The predicted reductions in STD prevalence due to the increase in condom use ranged from zero to 30% depending on consistency of use, and on the agent th at is transmitted. Conclusions. Condom use among 18-35 year old women has increased over the period, particularly with foreign partners. Con doms are used primarily as contraception. The prevalences of STD with high transmission rates are not reduced by inconsistent condom use, wh ile the prevalences of STD with low transmission rates are reduced by both consistent and inconsistent condom use. Condom use in a core grou p is more effective for reducing the STD prevalence than condom use in the non-core group for gonorrhoea and HIV. For chlamydial infection, condom use in the non-core group is more effective.