Objectives: To determine whether rate of development of AIDS is affected by category of exposure to HIV and whether the more rapid development found in older subjects persists for each exposure category.
Design: Longitudinal study of people with known date of seroconversion to HIV.
Setting: 16 HIV treatment centres throughout Italy.
Subjects: 1199 people infected with HIV through use of injected drugs, homosexual sex, or heterosexual sex.
Main outcome measures: AIDS as defined by 1987 definition of Centers for Disease Control (including and excluding neoplasms) and by 1993 European definition.
Results: 225 subjects (18.8%) progressed to AIDS (Centers for Disease Control 1987 definition) during median follow up of 5.8 years. Univariate analyses showed more rapid progression to AIDS for older subjects compared with younger subjects and for homosexual men compared with other exposure categories. The age effect was of similar size in each exposure category and in men and women. In a bivariate model with age and exposure categories simultaneously included as covariates, differences by exposure category disappeared for use of injected drugs and heterosexual sex compared with homosexual sex (relative hazards 1.02 (95% confidence interval 0.71 to 1.45) and 1.07 (0.70 to 1.64) respectively), while the age effect remained (relative hazard 1.55 (1.32 to 1.83) for 10 year increase in age). Analyses using the other definitions for AIDS did not appreciably change these results.
Conclusions: There was no evidence of differences in rate of development of AIDS by exposure category, while there was a strong tendency for more rapid development in older subjects for all three groups. This supports the view that external cofactors do not play major role in AIDS pathogenesis but that age is of fundamental importance.