Background:  The predictive value of HIV viral load on disease progression and mortality has been established in a number of studies in both men and women with HIV infection.  However, current recommendations for using HIV RNA levels to determine timing of HAART initiation and to assess therapeutic efficiency have been based on studies performed largely in men.  Various studies have shown that HIV viral loads are lower in women than men at similar stages of infection, with other studies disputing that finding.  We thus performed a critical epidemiologic review of the evidence regarding effects of gender on viral loads in HIV infection. 

Methods:  English-language publications examining the relationship between gender and viral load were identified by searching MEDLINE, AIDSLINE, Dissertation Abstracts Online (1990-2001), and related bibliographies and proceedings of annual HIV/AIDS meetings from 1994 to 2001.  13 studies were identified comparing viral load measurements in HIV-infected men and women at one point in time (cross sectional) or over time (longitudinal), with some adjustment made for relative stage of infection in the 2 groups.   Differences in log₁₀ viral load measurements between the women and men from each study were plotted, along with 95% confidence intervals (calculated if not provided). 

Results:    7 of the 9 cross-sectional studies demonstrated that women had 0.13-0.35 log₁₀ lower HIV RNA levels than men, with women having approximately half the HIV RNA concentrations of men, even upon controlling for CD4 count.  In the 4 longitudinal studies reviewed, women had 0.33-0.78 log₁₀ (2- to-6-fold) lower HIV RNA levels than men at similar stages of disease, despite controlling for time since seroconversion.  Adjustment for other possible confounders of the relationship between sex and viral load, including age, race and use of antiretroviral therapy, did not change the outcome of lower HIV viral load values in women compared to men in most of the studies.

Conclusions:  Women consistently have lower viral RNA loads than men at similar stages of HIV infection, an effect most marked early in the course of infection following HIV seroconversion.  This finding should serve as an impetus for further research on viral pathogenesis and its interplay with sex-related factors, inclusion of an adequate number of women in HIV-related studies, and a consideration of adjusting treatment guidelines for HAART initiation by gender.