Menstrual Pattern Among HIV-infected Women
Early in the AIDS epidemic, there were reports of increased menstrual irregularities among HIV-infected women. Most of the studies had no control groups with data based only on history, and included women with advanced AIDS. More recent studies have been published since. Chirgwin et al.1 used menstrual calendars to track menstrual cycles among 248 premenopausal HIV-positive women without AIDS and 82 HIV-negative women. HIV-positive women were more likely to experience menstrual intervals of more than 6 weeks (8 versus 0%, OR 10.8, 95% CI 1.8–1000), and have amenorrhea of more than 3 months (5 versus 0%, OR 7.1, 95% CI 1.1–1000). There were no differences in intermenstrual bleeding or irregular menstrual cycles. A past history of substance abuse was associated with an irregular menstrual cycle (adjusting for age, drug use, alcohol use, smoking, CD4 cell count and disease stage). Ellerbrock et al.2 performed a cross-sectional study of 197 HIV-positive and 189 HIV-negative women over a 12-month period. A total of 78% of the HIV-positive women and 80% of the HIV-negative women had 10–14 menstrual cycles (P = 0.74). There was also no difference in intermenstrual bleeding, postcoital bleeding, or amenorrhea. In this study, neither HIV infection nor CD4 cell count was associated with menstrual problems. In combined data from the HIV Epidemiology Research Study (HERS) and Women's Interagency Health Study (WIHS), Harlow et al.3 reported on the menstrual patterns of 802 HIV-positive and 273 HIV-negative women. They adjusted for demographic characteristics, body mass index and substance abuse. HIV infection increased the odds of having a very short cycle (< 18 days, OR 1.45) and very long cycles (> 90 days, OR 1.32). HIV serostatus had no effect on amenorrhea, menstrual cycle length or variability.
Hormonal Levels Among HIV-infected Women
There is a paucity of normative data on hormonal levels among HIV-infected women. The Women's Health Study (WHS) enrolled 55 HIV-positive and 10 high-risk HIV-negative women with self-reported regular menstrual cycles and determined progesterone and estradiol levels weekly for 8 weeks4. The median age was 35 years and the median CD4 cell count for the HIV-positive women was 210 cells/mm3. The median menstrual cycle length was 28 days for HIV-positive and 25 days for HIV-negative women. The maximum progesterone level during the luteal phase was normal (> 3.0 ng/ml) for 96% of the HIV-positive women and for 78% of the HIV-negative women (P = 0.09). There was no significant difference in the progesterone and estradiol levels by menstrual phase between HIV-positive and HIV-negative women. Among HIV-positive women, there was no significant difference in the progesterone and estradiol levels by antiretroviral therapy, baseline plasma viral load, or median CD4 cell count. They concluded that HIV-infected women with self-reported normal menstrual cycles have normal levels of progesterone and estradiol during the menstrual cycle.
HIV Shedding and Menstrual Cycle
There are conflicting data on the role of hormonal fluctuations and their effect on HIV viral load level in the blood and genital tract. Several studies have shown no significant effect of menstrual cycle on the detection of HIV in the genital tract and in the blood. It has been difficult to compare several studies because of differences in specimen collection techniques, assays used, the genital subcompartments that were sampled, characteristics of the women sampled, and differences in study design.
Reichelderfer et al.5 (WHS) assessed the variation of HIV-1 over the menstrual cycle, including HIV-1-RNA levels in the female genital tract, plasma HIV-1-RNA levels, CD4 cell counts, and culturable virus in a prospective study of 55 HIV-positive women. Seventy-four per cent of the women were on HAART, with a median CD4 cell count of 196 cells/mm3. The median plasma viral load at baseline was 3.83 log10. Median genital tract HIV-1 RNA was 3.9 log10 by Sno-Strip, < 1.9 log10 by cervicovaginal lavage (CVL), and < 2.6 log10 by cytobrush. Blood and genital tract specimens were collected weekly over 8 weeks spanning two complete menstrual cycles. Variations in viral RNA were compared in endocervical canal fluid and cells (collected by Sno-Strips and cytobrush) and ecto CVL fluid. Genital tract HIV-1 RNA from the CVL fluid and endocervical cytobrush were highest during menses and lowest immediately thereafter (P = 0.001 and P = 0.04). The HIV-1-RNA levels in the endocervical canal fluid was highest in the week preceding menses (P = 0.34). The menstrual cycle had no effect on blood levels of HIV-1 RNA (P = 0.62), culturable virus (P = 0.34), or CD4 cell counts (P = 0.55). They concluded that HIV-1-RNA levels vary with the menstrual cycle in the female genital tract but not in the blood compartment.
Benki et al. also reported that cervical virus levels were significantly correlated with the number of days from the mid-cycle surge in luteinizing hormone (LH) (P = 0.008). The lowest level of cervical HIV-1 RNA were present at the LH surge, and this nadir was followed by an increase in virus levels that reached a maximum before the start of the menses. They concluded that this data support the hypothesis that the level of HIV-1 RNA in cervical secretions is influenced by the menstrual cycle, and they suggested that the risk of heterosexual transmission of HIV-1 may increase as menses is approached6.
Villanueva et al.7, however, did not find that hormonal changes during the menstrual cycle had any significant effect on HIV-1-RNA levels in the vaginal secretions. Vaginal lavage samples were collected during 33 menstrual cycles in 25 women at 7, 14, and 21 days after the initiation of menses. Twenty per cent of the women were on antiretroviral drugs and the median CD4 cell count was 262 cells/mm3. The median plasma viral load was 4.9 log10. They reported that HIV-1-RNA levels in the vaginal secretions ranged from less than 1000 to 5.3 × 107 copies/lavage, and weekly changes ranged from less than 0.5 to 2.5 log10 copies/lavage. HIV-1-RNA levels in the lavage samples from days 7, 14, and 21 were not significantly different. No discernible pattern was found in changes in the vaginal virus load during the menstrual cycle. Vaginal virus loads were not correlated with plasma estradiol or progesterone levels (P > 0.05).
A study of 17 women from Mombasa, Kenya8, with normal menstrual cycles who underwent daily sampling of cervical and vaginal secretions over the course of a menstrual cycle also showed no discernible pattern of genital tract HIV shedding with the phase of the menstrual cycle and no significant association with serum estradiol and progesterone levels. There was no clustering of HIV-1-positive DNA results during menses. The median age was 31 years. None of the women were on antiretroviral therapy. The median plasma viral load was 3.7 log10 and the median CD4 cell count was 377 cells/mm3. Ten women had subtype A virus. In that study, 46% of endocervical swabs and 16% of vaginal swabs were positive for HIV-infected cells. There was considerable variability in the percentage of positive swabs from each woman ranging from 4 to 100% of endocervical swabs and 0 to 71% of vaginal swabs. In multivariate analysis, plasma HIV-1 RNA was significantly associated with the shedding of HIV-infected cells.
In a smaller study of six HIV-infected women who were studied weekly over 8 weeks9, plasma and cervical HIV-1 RNA fluctuated two to threefold in each woman over 8 weeks. Cervical HIV-1 RNA was consistently detected in two women with plasma viral loads greater than 100,000 copies/ml but not in two women with plasma viral loads less than 10,000 copies/ml regardless of the menstrual cycle status.
There is no definite consensus as to the effect of hormonal changes during the menstrual cycle on HIV shedding in the genital tract. As previously mentioned, it is difficult to compare currently available studies because of their varied design, differences in study population, differences in collection methods and the assays used. There is a need to continue to assess the effect of hormonal changes on HIV shedding in the genital tract. Other variables such as the microbiological and immunological environment in the genital tract may affect HIV shedding aside from hormonal factors.
This study was supported in part by R01 AI40350, from the National Institute of Allergy and Infectious Diseases.
1. Chirgwin KD, Feldman J, Muneyyirci-Delale O, et al
. Menstrual function in human immunodeficiency virus-infected women without acquired immunodeficiency syndrome. J AIDS Human Retrovirol 1996; 12:489–494.
2. Ellerbrock TV, Wright TC, Bush TJ, et al
. Characteristics of menstruation I women infected with human immunodeficiency virus. Obstet Gynecol 1996; 87:1030–1034.
3. Harlow SD, Schuman P, Cohen M, et al
. Effect of HIV infection on menstrual cycle length. J AIDS 2000; 24:68–75.
4. Cu-Uvin S, Wright D, Anderson D, et al
. Hormonal levels among HIV-1 seropositive women compared with high-risk HIV-seronegative women during the menstrual cycle. J Women's Health 2000; 9:857–863.
5. Reichelderfer PS, Coombs RW, Wright D, et al
. Effect of menstrual cycle on HIV-1 levels in the peripheral blood and genital tract. AIDS 2000; 14:2101–2107.
6. Benki S, Mostad SB, Mandaliya K, et al
. Cyclic shedding of HIV-1 RNA in cervical secretions during the menstrual cycle. J Infect Dis 2004; 189:2192–2201.
7. Villanueva JM, Ellerbrock TV, Lennox JL, et al
. The menstrual cycle does not affect human immunodeficiency virus type 1 levels in vaginal secretions. J Infect Dis 2002; 185:170–177.
8. Mostad S, Jackson S, Overbaugh J, et al
. Cervical and vaginal shedding of human immunodeficiency virus type 1-infected cells throughout the menstrual cycle. J Infect Dis 1998; 178:983–991.
9. Goulston C, Stevens E, Gallo D, et al
. Human immunodeficiency virus in plasma and genital secretions during the menstrual cycle. J Infect Dis 1996; 174:858–861.