Bacterial vaginosis occurs among substantial proportions of healthy, nonpregnant young women, regardless of whether they have ever had intercourse.1 Almost three in 10 sexually experienced women in a sample tested in 1999-2000 and one in five of those who had never had sex received a diagnosis of bacterial vaginosis, a syndrome that is associated with pelvic inflammatory disease, poor pregnancy outcomes and other reproductive health problems. The odds of receiving this diagnosis varied significantly by race and ethnicity, were elevated among women with chlamydial infection and were reduced among hormonal contraceptive users.
The sample consisted of Marine recruits participating in a longitudinal study of an intervention designed to prevent pregnancy and sexually transmitted diseases (STDs). Participants" background data were collected from standard forms they completed before undergoing pelvic examinations at a naval hospital"s well-women clinic. Laboratory specimens were obtained from urine samples, participants" self-collected vaginal swabs and endocervical samples taken during the examination. A variety of diagnostic techniques were used; the analyses were based on the 1,938 women who had complete data for assessment of bacterial vaginosis according to Nugent Gram stain criteria.
Participants ranged in age from 17 to 33, but 75% were 17-19 years old. Fifty-six percent were white, 20% Hispanic, 16% black and the rest members of other racial or ethnic groups. Most (86%) had had vaginal intercourse; of these, 11% tested positive for chlamydia, and 2% each had gonorrhea and trichomoniasis. Overall, 20% reported vaginal discharge, and 7% said they had noticed a vaginal odor; during the pelvic examination, clinicians detected vaginal discharge in 5% of the women.
Bacterial vaginosis was detected in 27% of participants; the researchers used chi-square and Fisher exact tests to identify potential correlates of this diagnosis. Results showed that bacterial vaginosis was more prevalent among sexually experienced women than among their sexually inexperienced counterparts (28% vs. 18%), and that prevalence ranged from 11% among Asians and Pacific Islanders to 32-34% among blacks and Native Americans. Women who reported having vaginal discharge or noticing vaginal odor, those who had had more than one sex partner in the previous three months and those with chlamydial infection had rates of 32-41%—significantly higher than the rates among women without these characteristics. Users of hormonal contraceptives had a lower prevalence of bacterial vaginosis than nonusers (25% vs. 31%).
In a multiple logistic regression analysis controlling simultaneously for all of these factors, only self-reported vaginal discharge was not independently associated with bacterial vaginosis. Compared with Asians and Pacific Islanders, members of other racial and ethnic groups had significantly higher odds of diagnosis: The odds were more than doubled for white women (odds ratio, 2.4), roughly tripled for Hispanics and blacks (3.1 each), and nearly quadrupled for Native Americans (3.7). The odds also were elevated among women who were sexually experienced (1.7), who reported a vaginal odor (1.8), who had had multiple partners in the past three months (1.3) or who tested positive for chlamydia (1.6); they were reduced among hormonal contraceptive users (0.8).
The researchers also examined Pap smears and a recently developed colorimetric pH test, which take less time and require less specialized personnel than the Gram stain, as alternative tools for diagnosing bacterial vaginosis. Both tests performed "moderately well." (The pH test had a sensitivity of 72% and specificity of 67%; the Pap smear, 72% sensitivity and 79% specificity.) Thus, the researchers suggest that these techniques may be advantageous in some situations. In particular, the Pap smear "might be a reasonable "first-line" screening tool during routine gynecologic examinations of healthy populations, given...that it adds no cost."
Although the findings regarding correlates of bacterial vaginosis largely echo those of previous studies, the researchers note that their sample offered several advantages over those in earlier work. First, whereas most research on bacterial vaginosis has used clinic-based samples of pregnant or STD-infected women, these analyses were based on healthy, nonpregnant women in a nonclinic setting. Additionally, the sample had the largest group of sexually inexperienced women ever evaluated for bacterial vaginosis with the Nugent criteria. Finally, it was large enough to allow comparisons across racial and ethnic groups.
1. Yen S et al., Bacterial vaginosis in sexually experienced and non-sexually experienced young women entering the military, Obstetrics & Gynecology, 2003, 102(5, pt. 1):927-933.