Objectives: To describe the early manifestations of HIV-1 infection before the development of AIDS, in a rural Ugandan population.
Methods: Three monthly follow-up of HIV-1-infected and uninfected participants in an HIV-1 natural history cohort from the start of the cohort in 1990 to the end of 1996.
Results: A total of 107 persons with prevalent infection and 104 persons with incident infection were enrolled. Eighty (75%) prevalent and 89 (86%) incident individuals were asymptomatic on enrollment. Of the 91 persons with incident infection seen within 2 years of their estimated date of seroconversion, 51% [95% confidence interval (CI), 40-61] were still asymptomatic 2 years after seroconversion. At 4 and 5 years after seroconversion, only 26% (95% CI, 17-36) and 11% (95% CI, 4-22), respectively, remained asymptomatic. A total of 89 participants entered World Health Organization (WHO) stage 2, and their main stage-defining conditions were weight loss <10% (and > 5%) and minor mucocutaneous manifestations. The median CD4 lymphocyte count for participants entering WHO stage 2 was 516 x 10(6) cells/l (interquartile range, 360-884 x 10(6)/l). A total of 94 participants entered WHO stage 3 and the main reasons were weight loss >10%, unexplained chronic diarrhoea, fever for more than 1 month, and severe bacterial infection. The median CD4 lymphocyte count for participants entering WHO stage 3 was 428 x 10(6) cells/l (interquartile range, 276-736 x 10(6)/l). The rates of all conditions reported under the WHO staging system were significantly more frequent in HIV-positive persons than HIV-negative controls with the exception of fever for more than 1 month and oral hairy leukoplakia (which was seen in only three individuals).
Conclusion: These are the first data from a non-selected African population describing the early manifestations of HIV infection. The main early manifestations were weight loss, minor mucocutaneous features, chronic diarrhoea, chronic fever and severe bacterial infections. The apparent rapid development of HIV-related signs and symptoms is probably indicative of the high background level of these conditions in our study area, as reflected by the rates of these conditions in the HIV-negative controls.