Objectives: To describe the distribution and changes in CD4 cell counts (both initial and subsequent) in HIV-infected persons over time and determine the factors influencing these counts.
Design: Reports were requested from laboratories measuring CD4 cell counts in England and Wales. Initial counts were analysed and median counts were followed over time.
Methods: Time trends and the relationship between initial CD4 cell count and age, sex, and HIV risk category were studied using quantile regression methods or chi-square tests.
Results: Between 1990 and 1998, 9553 adults were newly diagnosed with HIV infection and had a CD4 cell count within 6 months of HIV diagnosis. Over 50% of initial CD4 cell counts in each major risk category were below 350 cells/mm3. Older age (P < 0.001), male sex (P < 0.013) and heterosexual risk (P < 0.001) were independently associated with lower initial CD4 cell counts. For heterosexually infected adults, the median initial CD4 cell count was significantly negatively associated with the year of diagnosis (P = 0.03) and the median age increased through the time period examined (P < 0.001), whereas for men who have sex with men (MSM), there was no significant change in these values over time. For each year cohort of newly diagnosed individuals, the median CD4 cell count in subsequent years decreased until 1996 and then increased thereafter, consistent with a treatment effect.
Conclusion: Across all major risk groups, a large proportion of HIV-infected adults are being diagnosed late in the course of HIV disease. For the heterosexually infected, the data suggest an ageing cohort effect, whereas for MSM the data are consistent with continuing transmission.