This is an archive page. The links are no longer being updated.

Date: December 27, 1995
For Release: 5 p.m., E.S.T.
Contact: NICHD (301) 496-5133 or NIAID (301) 402-1663

Pregnant women who were diagnosed with BV during the second trimester were 40 percent more likely to give birth to a premature infant with low birth weight (i.e., an infant born before 37 weeks' gestation and weighing less than 5 pounds) than were women who did not have the vaginal infection. This increased risk remained after adjusting for other variables, including smoking, race, previous delivery of a low-birth-weight infant, previous pregnancy loss, number of previous live births, maternal age, antibiotic use, and other vaginal infections. The research, part of a larger NICHD/NIAID clinical study called the Vaginal Infections and Prematurity (VIP) Study, will appear in the December 28 issue of the New England Journal of Medicine.

"Preterm delivery and low-birth-weight delivery remain two of the most difficult unsolved problems in our country," said principal investigator Dr. Sharon Hillier, now at the University of Pittsburgh/Magee Women's Hospital. "If we find that treating this very common vaginal condition can prevent preterm birth, this would be a really important step in preventing the long-term sequelae that some preterm, low-birth-weight infants face."

BV is caused by an imbalance among the bacteria that are normally found in the vagina. In healthy women, the predominant strain of bacteria found is Lactobacillus. With BV, increased numbers of anaerobic organisms are found, including Gardnerella vaginalis, Mycoplasma hominis, and bacteroides. Although these bacteria are also found in healthy women, they signal an abnormality when they outnumber the normal Lactobacillus flora.

Previously called Gardnerella vaginitis or nonspecific vaginitis, BV is the most common vaginal infection in reproductive-aged women. It is also one of the most common vaginal infections in pregnancy, affecting from 12 to 22 percent of pregnant women.

While BV has been linked to preterm birth before, this study is the first large enough in size to document the association between the infection and the premature delivery of a low-birth-weight infant after controlling for other variables, such as smoking and obstetrical history.

Between 1984 and 1989, the VIP research team enrolled 10,397 pregnant women from seven medical centers in five cities into this cohort study. None of the women had known medical risk factors for preterm delivery.

At 23-26 weeks of gestation, all women were screened for BV using laboratory staining for bacterial strains associated with BV. Since vaginal pH tends to be less acidic with BV, pH testing was also used to determine if infection was present.

The women were then followed until delivery to identify birth outcome involving prematurity and low birth weight. Of 504 women who delivered premature, low-birth-weight infants, 20 percent had been diagnosed with BV. Further statistical analysis indicated that women with BV were 40 percent more likely than women without the condition to have such a birth. At highest risk were women with two particular strains of bacteria, Mycoplasma hominis and bacteroides.

In addition to BV, the primary risk factors for preterm delivery of a low-birth-weight infant were smoking, which increased the risk of preterm delivery by about 40 percent; having already had a preterm infant, which increased the risk about six fold; and being of African American race, which increased the risk by about 40 percent. Women with a history of bladder infections or antibiotic use prior to the study were also more likely to have a premature, low-birth-weight infant.

Although BV is more common among women with new or multiple sexual partners, the organisms that cause it have been isolated in young women who are not sexually active. The use of intrauterine devices has been linked to an increased risk of becoming infected.

The main symptom associated with BV is an abnormal vaginal discharge with a characteristic "fishy" odor. As many as 50 percent of women, however, have no symptoms at all, and are only diagnosed after microscopic examination of a sample of vaginal discharge under a microscope. The infection can be treated with antibiotics. "Pregnant women who have genital symptoms should be screened for this and treated appropriately if they're found to have bacterial vaginosis," Dr. Hillier said.

The mechanisms underlying the increased risk of prematurity and low birth weight associated with BV are unclear. The investigators theorize that, in addition to causing an infection in the vagina, BV may cause an infection in the uterus, which somehow triggers preterm birth.

Although premature delivery and low birth weight are major contributors to perinatal mortality, there is no effective way to prevent them at this time. The findings from this study, however, may be the first step toward a prevention strategy that would involve screening and treatment of high-risk women.

Already, a related study by Hauth, et al, has demonstrated that antibiotic treatment is an effective way to reduce the risk of premature delivery among women with BV and other noninfectious risk factors for preterm birth. In this study, also appearing in the December 28 New England Journal of Medicine, investigators at the University of Alabama at Birmingham found that they were able to reduce the rate of premature delivery in such women by treating them with the antibiotics metronidazole and erythromycin.

Currently, the NICHD's Maternal/Fetal Medicine Unit Network is conducting a large clinical trial involving approximately 1,900 pregnant women to determine whether metronidazole alone will reduce the risk of preterm delivery in low- risk women with asymptomatic bacterial vaginosis. In addition to Dr. Hillier, the research team for the study on the association between BV and preterm delivery included Drs. Robert Nugent, Donald McNellis, and Mark Klebanoff of the NICHD; Drs. David Eschenbach and Marijane Krohn (now at the University of Pittsburgh) of the University of Washington, Seattle; Dr. Ronald Gibbs (now at the University of Colorado) of the University of Texas, San Antonio; Drs. David Martin and Joseph Pastorek II of Louisiana State University, New Orleans; Drs. Mary Frances Cotch (now at the Research Triangle Institute) and Robert Edelman (now at the University of Maryland) of the NIAID; Dr. A. Vijaya Rao of the Research Triangle Institute, North Carolina; Dr. Joan Regan of Columbia University, New York; and Dr. J. Christopher Carey of the University of Oklahoma, Oklahoma City.

The NICHD and the NIAID are part of the National Institutes of Health, the biomedical research arm of the Federal government. The NICHD was established in 1962. Since then, the Institute has become a world leader in promoting research on reproductive biology, including fertility regulation, and population issues; development before and after birth; maternal, child, and family health; and medical rehabilitation. The NIAID conducts and supports research aimed at preventing, diagnosing, and treating illnesses such as AIDS and other sexually transmitted diseases, tuberculosis, asthma, and allergies.

"This study is the first very large, cross-sectional study of women in the United States to document that bacterial vaginosis is associated with having a preterm, low-birth-weight infant.

"This very common vaginal syndrome is not routinely diagnosed in pregnant women currently because many physicians have not really appreciated that this very common vaginal complaint could be a cause of complications in pregnancy.

"This was the first study that was large enough in size to really document the association between bacterial vaginosis and delivery of a preterm, low-birth-weight infant after accounting for other risk factors such as smoking and obstetrical history.

"The data that were derived from this study certainly indicate that women who have symptoms of vaginal infections who are pregnant should be screened for this and treated if they're found to be symptomatic. I think we need to await the outcome of large, randomized treatment trials to determine whether or not women who are asymptomatic should be screened and treated as well.

"About half of women with bacterial vaginosis will be asymptomatic, and so it will be important to ascertain whether just symptomatic women should be treated or all women.

"Unfortunately, we did not routinely record symptoms in this study so we do not know if the association was stronger in women with symptoms than in those without symptoms.

"The primary risk factors in this study for preterm delivery were smoking, which increased the risk of preterm delivery by about 40 percent, having had another infant born preterm in an earlier pregnancy, which increased the risk about six fold, and being of African American race, which increased the risk about 40 percent.

(What percentage of the 504 women who delivered premature, low birth weight infants had BV?) "It was 20 percent.

"All of the organisms which cause bacterial vaginosis are found normally in the vagina. The difference is that in bacterial vaginosis, the organisms become the dominant species in the vagina, and therefore increase by about 100 to a 1,000 fold over what they would normally be present at. It's this change in their relative concentrations that makes them a problem.

"Preterm delivery and low-birth-weight delivery remain two of the most unsolvable problems in our country, and despite massive amounts of effort in trying to prevent preterm birth, a lot of efforts have failed. I think that this study is exciting because this is something we can potentially do something about. If we find that treating this very common vaginal infection can preterm birth, this would be a really important step in preventing the long-term sequelae that preterm, low-birth-weight infants face."