M. Marmor et al., WEIGHT-LOSS ASSOCIATED WITH HIV SEROCONVERSION AMONG INJECTION-DRUG USERS, Journal of acquired immune deficiency syndromes and human retrovirology, 12(5), 1996, pp. 514-518
To describe symptoms associated with human immunodeficiency virus (HIV
) seroconversion, we studied a cohort of 366 injection-drug users (IDU
s) with a study design that included recall every 3 months to collect
symptom histories using a structured questionnaire. Eleven HIV serocon
versions were observed in 621.5 person years at risk (PYAR), equivalen
t to 1.8 seroconversions/100 PYAR. Cox regression analysis showed age
less than or equal to 35 years to be a significant risk factor for HIV
seroconversion after controlling for gender, race, and the frequency
of drug injection. An embedded case-control analysis then compared sym
ptom histories of HIV seroconverters with those of age-(+/- 5 years) a
nd visit number-matched controls who remained HIV seronegative for gre
ater than or equal to 3 months longer than the HIV-seroconverters, Mul
tivariate case-control analysis adjusted for injection frequency yield
ed significant associations of HIV seroconversion with histories of we
ight loss greater than or equal to 4.5 kg (seven of 11 cases; odds rat
io [OR] = 11.6, 95% confidence interval [CI] 3.1, 43.1) and oral ulcer
s (three of 11 cases; OR = 7.6, 95% CI = 1.2, 48.2) in the 3 months be
fore the subjects' first HIV-seropositive study visit. We conclude tha
t histories of recent symptoms reported by HIV-seroconverting IDUs dif
fer from those reported by non-HIV-seroconverting IDUs, and weight los
s may be particularly common among IDUs experiencing primary HIV infec
tion.