Condom exposure risk and risks of HIV infection in Spain

Barthelemy D. Kuate, Université de Montréal
Teresa Castro Martin, Spanish National Research Council (CSIC)

Nationally representative data from the Health and Sexual Behaviour Survey of Spain fielded in 2003 among 10838 men and women aged 18-49 years are used to assess whether ‘condom exposure risk’ predicts HIV infection. While over 93% of the sample has ever had sexual intercourse, only about half of respondents used condom during the first sexual intercourse. Our study incorporates the context of behaviour in the construction of ‘condom exposure risk’. Our constructed condom exposure risk variables defines higher risk as not using condoms at first sexual intercourse, not using condom at last sexual intercourse, or not using condom with multiple sexual partners in the last 12 months. Our data have a wealth of information on knowledge of modes of transmission and means of protection, risk perception, and attitudes towards prevention and information sources about HIV infection and AIDS. We have also imputed a series of community-level variables into these dataset, making them more complete for carrying out multivariate multilevel analyses that will consider the role of contextual factors on studied relationships. Using multilevel logistic regression models, we test (1) the imprinting hypothesis that condom use at sexual debut predicts subsequent condom use, after controlling for putative factors of condom use and (2) the hypothesis that the composite variable ‘condom exposure risk’ predicts HIV infection risks. We use information on (1) ever being diagnosed with sexually transmitted infections (11 STI items), (2) efficiency of preventive measures against HIV (9 items), and (3) being at risk of getting infected with HIV (7 items) as proxies for HIV infection risks. We will construct three ‘HIV infection risk’ outcomes based on the score from each of the three variables, and will use multilevel analyses to account for individual-level and community-level variations in the influences of condom exposure risk on these outcomes.