Levels and patterns of HIV RNA viral load in untreated pregnant women

Objective: To assess pregnancy levels and patterns of HIV RNA in the absence of antiretroviral therapy, while appropriately adjusting for potential confounders, including maternal immune status and race. Methods: Data on >= 1 antenatal HIV RNA measurements were available for 333 untreated HIV-infected pregnant women enrolled in the European Collaborative Study. CD4 counts and HIV RNA measurements were routinely collected from 1992 and 1998, respectively. Linear mixed effects models based on 246 women for whom complete data were available examined changes in HIV RNA levels over pregnancy, with a nested random effects term accounting for measurement variability within women and period of sample collection. Results: The change in HIV RNA over pregnancy varied significantly by race (p = 0.005): from the second trimester until delivery, HIV RNA decreased significantly by an estimated 0.019 log(10) copies/ml/week in white women (95% Cl -0.03, -0.007); in black women the estimated 0.016 log(10) copies/ml/week increase (95% Cl -0.005, 0.037) was not statistically significant. At delivery, HIV RNA levels in black women were 0.45 log(10) copies/ml higher (95% Cl 0.08, 0.83) than in white women. Conclusions: Our findings suggest that HIV RNA dynamics over pregnancy differ by race, although other interpretations cannot be excluded, due to potential for unmeasured confounding. (C) 2008 International Society for Infectious Diseases