Background. Infection with human immunodeficiency virus type 1 (HIV-1)
causes progressive immune deficiency, the acquired immunodeficiency s
yndrome (AIDS), and death. Mortality, however, particularly with cause
s other than AIDS, deserves further study. A retrospective cohort stud
y among drug users in Italy was performed to estimated absolute and pr
oportional mortality rates due to AIDS and other causes, with or witho
ut HIV-1 infection. Methods. All subjects who enrolled between January
1980 and July 1990 in the drug treatment programme in the Province of
Bologna, Italy, were included in the cohort. Each subject was categor
ized for HIV-1 antibody status (positive, negative, untested), vital s
tatus (in 1990 by national surveillance), and causes of death (by deat
h certificate). Data were analysed with actuarial and time-dependent c
ovariate methods. Results. There were 332 deaths among 4962 drug users
who were followed for 21 130 person-years. This mortality rate (1.57
per 100 person-years) was increased 18-fold compared to the general po
pulation. Actuarial 10-year mortality estimates were 28.2% for the 204
0 HIV-1 positive subjects, 12.1% for the 1859 HIV-1 untested subjects,
and 2.5% for the 1063 HIV-1 negative subjects. AIDS contributed to 15
0 deaths, followed by drug overdose (64 deaths) and trauma (39 deaths)
. Compared to others in the cohort, mortality with AIDS and non-AIDS c
auses was reduced for HIV-1 negative subjects. In contrast, mortality
for HIV-1 positive subjects was increased with AIDS, trauma, overdose,
various bacterial infections, hepatitis, and cirrhosis. Conclusions.
Mortality with HIV-1 infection was associated not only with opportunis
tic infections and malignancies but also with competing causes of deat
h, particularly hepatic disease. Further investigation is needed to cl
arify whether alcohol, analgesics, hepatitis viruses, or other agents
have enhanced hepatotoxicity for HIV-1 infected patients.