| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the Department of Epidemiology, School of Public Health, Carolina Population Center, Department of Obstetrics and Gynecology, School of Medicine, and Division of Infectious Diseases, Department of Medicine, School of Medicine, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina; Department of Obstetrics and Gynecology, Duke University School of Medicine, Durham, North Carolina; and Section of Epidemiology and Biostatistics, Department of Obstetrics, Gynecology and Reproductive Sciences, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, Piscataway, New Jersey.
Address reprint requests to: David A. Savitz, PhD, The University of North Carolina at Chapel Hill, Department of Epidemiology, CB #7435 McGavran Greenberg Building, Chapel Hill, NC 27599-7400; E-mail: david_savitz{at}unc.edu.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: Vaginal fluid specimens were obtained at 24–29 weeks’ gestation from 242 cases with preterm birth and 507 noncases sampled using a case–cohort study design. We tested for bacterial vaginosis by Gram staining and Nugent scores and assayed for neutrophil defensins by enzyme-linked immunosorbent assay. Bacterial vaginosis was studied as a categoric variable (negative, intermediate, and positive), whereas defensins were studied as a continuous, categoric (based on percentiles), and dichotomous measure (presence versus absence). Three gestational age cut points were used to define preterm birth. Modified Cox proportional hazard models were used to evaluate the associations between bacterial vaginosis, defensins, and degree (less than 32, less than 34, and less than 37 weeks) and type (premature rupture of membranes, preterm labor) of pre-term birth.
RESULTS: Elevated vaginal fluid neutrophil defensins were not associated with birth before 37 weeks. Compared with women who did not have measurable vaginal fluid defensins, women with higher defensin levels (0–2.8 µg/mL, 2.8–8.2 µg/mL, and greater than 8.2 µg/mL) had a greater risk of delivering before 32 weeks. Hazard ratios adjusted for maternal race and vaginal bleeding during pregnancy and 95% confidence intervals for these defensin levels were 1.7 (0.4, 6.9), 2.4 (0.7, 7.9), and 3.1 (1.0, 9.8), respectively. Bacterial vaginosis status did not influence the association between defensins and preterm birth.
CONCLUSION: Elevated concentrations of vaginal fluid neutrophil defensins at 24–29 weeks’ gestation might predict preterm birth before 32 weeks.
Preterm birth or birth before the completion of 37 weeks’ gestation is a leading cause of infant mortality worldwide. Prematurity is also associated with neonatal complications, including periventricular leukomalacia,1–3 cerebral palsy, and bronchopulmonary dysplasia.4 Several clinical and epidemiologic studies have suggested an association between reproductive tract infections, including bacterial vaginosis, during pregnancy and subsequent preterm birth.5–9 Bacterial vaginosis is a clinical condition in which normal vaginal flora, characterized by high concentrations of lactobacilli, are replaced by high concentrations of anaerobic bacteria, such as Bacteroides sp, Mobiluncus sp, Gardnerella vaginalis, and Mycoplasma hominis.10 Most observational studies and some randomized trials suggest an association,11,12 but others do not support a relationship between changes in vaginal flora and preterm birth. Most notably, a recent study examining five intervention trials reported that treatment for bacterial vaginosis did not reduce the incidence of preterm birth.13
Defensins are endogenous antimicrobial peptides synthesized by epithelial cells and neutrophils in humans. Human neutrophil defensins are specific to azurophilic granules in neutrophils and are compartmentally elevated during inflammatory processes like intraamniotic infection and pediatric meningitis.14,15 Functionally, they are capable of making microbial cells and normal human and tumor cells permeable by integrating into their membrane structures.16,17 Because defensins are released in response to infection and can be measured easily, they may have significant diagnostic properties for generalized or specific infections.
Although bacterial vaginosis and other reproductive tract infections are common in pregnancy, not all women who develop these conditions deliver preterm, implying that certain subgroups of women with such infections are at a higher risk for preterm birth. Given the adverse consequences of preterm birth, accurate identification of women who are at a higher risk for delivering preterm is crucial for improvement of pregnancy care and for better understanding the etiologic processes leading to preterm birth. In a previous study, we examined the association between vaginal fluid defensins and bacterial vaginosis status (negative, intermediate, and positive, defined by Nugent scores of 0–3, 4–6, and 7–10, respectively). We found a moderately strong association between intermediate bacterial vaginosis and elevated vaginal fluid defensins.18 Given the important role that defensins could play during pregnancy, we examined the independent as well as joint effects of vaginal fluid defensins and bacterial vaginosis on the risk of preterm birth.
|
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Basic demographic data were obtained through a screener questionnaire. Telephone interviewers administered a more detailed questionnaire within 2 weeks of recruitment into the study. The study subjects were asked to provide information on the current pregnancy, health behaviors, reproductive history, employment history, and the presence of sexually transmitted diseases. At the time of recruitment (24–29 weeks’ gestation), vaginal fluid samples were collected by inserting a fresh Dacron-tipped swab in the posterior vaginal fornix and wiping the fornix to saturate the swab. For this study, samples were retrieved from storage freezers and sent to Biex Inc. (Dublin, CA). The samples were thawed and eluted out of the swabs and then assayed for the concentration of vaginal fluid defensins by double-sandwich enzyme-linked immunosorbent assay, as previously described by Heine et al.20 Standard curves were generated with purified anti-human neutrophil peptide type 2 and the concentration of human neutrophil peptide type 2 defensins in vaginal fluid samples was expressed as µg/mL. The samples were originally diluted with 1 mL of Tris-P-T80 buffer. After elution, the samples were diluted 1:1000 times, by first adding 10 µL of vaginal fluid solution to 990 µL Tris-cetyl ethyl ammonium bromide buffer and then further diluting this solution ten times. Unless otherwise stated, all references to vaginal fluid defensins in this article imply human neutrophil defensin 2.
To assess bacterial vaginosis, a Dacron-tipped swab was wiped on the lateral side of the vaginal wall, and the pH was recorded by touching the swab on a pH indicator strip. A microscope slide was prepared by rolling the swab on the surface of a glass slide and air-drying before storage. The smears were Gram stained and scanned under the microscope. Lactobacillus spp, Gardnerella vaginalis, and Mobiluncus spp were identified and counted. Based on the Nugent method, each morphotype was quantitated according to the number per oil immersion field: no morphotypes, less than 1 morphotype, 1–4 morphotypes, 5–30 morphotypes, and more than 30 morphotypes.21 For every slide, ten nonadjacent oil immersion fields were scored on a scale of 0 to 10, and scores were then averaged across the fields. Bacterial vaginosis status was classified as negative, intermediate, and positive, based on average score of 0–3, 4–6, and 7 or higher, respectively.
Gestational age was assessed from last menstrual period (LMP) information, which was corroborated by ultrasounds performed before 24 weeks’ gestation for the women in the Pregnancy, Infection, and Nutrition study, on which our study sample is based. For women who had LMP dates and ultrasound information, LMP dates were used to calculate gestational age when both estimates were within 14 days of one another. If the dates disagreed by more than 14 days, the ultrasound estimate was used. Preterm birth was classified as spontaneous or medically induced. Spontaneous preterm birth was further classified as premature rupture of membranes (PROM) and preterm labor. To ensure that the underlying reason for membrane rupture was not premature labor, rupture occurring at least 4 hours before the onset of labor was classified as PROM. We did not evaluate the association between neutrophil defensins and medically induced preterm birth; this was not of interest, owing to the diverse pathways they reflect that may not be associated with defensin effects. Women with medically induced preterm birth were excluded from the analyses when we evaluated the association between markers of infection and PROM and preterm labor.
The outcome variable for our analyses was preterm birth, including severity of prematurity and types of preterm birth. Bacterial vaginosis status (negative, intermediate, and positive) and vaginal fluid defensin concentration were the primary predictor variables. We examined defensins as a cubic spline–transformed, continuous measure to flexibly describe the pattern of association.22 Regression models with three, four, five, and six knots for defensins were fit and compared, based on the likelihood–ratio test.22 The model with three knots for defensins (at locations 0, 2.8, and 32.9 µg/mL) provided the best fit. Based on our findings in a previous study, defensins were also examined as a categoric variable, based on percentile values among noncases, with categories at 2.8 µg/mL (50th percentile) and 8.2 µg/mL (75th percentile).18 As a first step in data analysis, descriptive statistics were generated for maternal age, race, education, marital status, parity, self-reported reproductive tract infection, working history, smoking status, alcohol consumption during pregnancy, and vaginal bleeding. These distributions were stratified on categories of gestational age at birth.
To study the relationships between defensins and pre-term birth, modified Cox proportional hazards models were fit to the data.23 Time was measured as weeks of gestation, and the outcome of interest was preterm birth (birth before completing 37 weeks), severity of preterm birth (birth before completing 32 or 34 weeks), and types of preterm birth (PROM and preterm labor leading to delivery prior to completing 37 weeks’ gestation). Deliveries occurring after 32, 34, or 37 weeks were censored, depending on the severity of preterm birth being analyzed.
Maternal race, age, education, employment history, marital status, parity, vaginal bleeding during pregnancy, the presence of self-reported reproductive tract infection, and previous preterm birth, abortion, or low birth weight were examined as potential confounders and effect modifiers, based on a possible association with preterm birth or reproductive tract infection. Because we did not have a priori information on possible confounders for vaginal fluid defensins, we studied factors that could be associated with reproductive tract infections.
The case–cohort study design was taken into account, and robust variance measures were incorporated into the models to account for cases originating within and outside of the subcohort. Because of the complicated sampling method used, a standard Cox proportional hazard model would underestimate the true variances.23 Therefore, we used a modified Cox model to get a more accurate estimate of the standard error. Stratified hazard ratios were generated for gradations of bacterial vaginosis status to assess the joint effect of elevated defensins and bacterial vaginosis status on preterm birth. Confounders were retained in the model if they altered the hazard ratio by more than 10% when compared with unadjusted models. Likelihood ratio tests were also conducted with a 5% level of significance.
Analyses were performed using SAS for Windows (SAS Institute, Cary, NC), based on a modified Cox proportional hazard model implemented as an SAS macro.23
|
RESULTS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Multiparous women and women who had vaginal bleeding were at a higher risk for delivery before 32 weeks. Women who did not have a high-school level of education were more likely to deliver before 34 weeks, compared with women who had higher levels of education.
|
); increasing defensin levels were not associated with the adjusted hazard ratios for delivery before 37 weeks.
|
). Women with measurable vaginal fluid defensins (0–2.8 µg/mL, 2.8–8.2 µg/mL, and greater than 8.2 µg/mL) did not have an increased risk of delivery before 37 weeks, compared with women who did not have detectable defensins. Compared with women who did not have measurable vaginal fluid defensins, women with measurable defensins had a higher risk of delivering before 32 weeks, with evidence of a dose–response gradient. This association persisted after adjustment for maternal race and vaginal bleeding during pregnancy.
|
). However, because of small sample size, these results were imprecise. Among bacterial vaginosis–negative women, the adjusted hazard ratio for having preterm birth before 32 weeks for women with detectable defensins was 2.8 (95% confidence interval [CI] 0.8, 9.4) compared with women with undetectable vaginal fluid defensins. In this same group of women, there were weaker associations for defensins above 2.8 µg/mL or 8.2 µg/mL (50th and 75th percentiles, respectively) as compared with lower defensin concentrations.
|
|
DISCUSSION |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Because neutrophil defensins are cytotoxic, their synthesis and release in large concentrations could potentially damage fetal membranes and cause PROM. Therefore, we had expected that elevated neutrophil defensins would be more strongly associated with PROM, but they were not. Defensins have been found to act like pro-inflammatory cytokines; at submicromolar concentrations, some ß-defensins have been found to attract both immature dendritic cells and memory T cells, initiating primary and recall immune response, respectively.24 Defensins also induce the synthesis of interleukin-8 and other cytokines. Although human neutrophil peptide type 2 and other alpha defensins have not been studied in relation to these immune functions, we speculated that the human neutrophil defensins we measured could play an important role in the immunologic balance during pregnancy and bring about preterm birth by mechanisms other than membrane rupture. However, our study did not show an association between human neutrophil peptide type 2 and preterm labor before 37 weeks. The small number of births before 32 weeks’ gestation precluded analysis of subtypes in the interval.
One of the limitations of our study was the inability to standardize the quantity of vaginal fluids. We recognize that there may have been variations in the volume of vaginal fluids collected across women in the study. However, by following standard protocols of sample collection, dilution, and assay, measurement error for defensins would likely be random error. Further, by controlling for such variables as maternal race, age, parity, self-reported reproductive tract infections, and the presence of bacterial vaginosis, we attempted to account for interparticipant variation. We recommend that future studies be conducted to standardize vaginal fluid volume collected using swabs, as this would facilitate measurement of various markers present in these specimens. At the current time, there is no available test for quantifying human defensin-5 and human ß-defensin-1, which are types of defensins found in female genital epithelia.15,16 Although testing for all subtypes of defensins that could be present in vaginal secretions would be ideal, doing so was not possible in this study. Thus, there is a potential for underestimating the concentration of vaginal fluid defensins. Another drawback was the unavailability of defensin measurements for 180 women in the case cohort study. These women were not found to differ from women who were included in the analyses in terms of demographic and clinical variables. However, approximately 49% of women missing information on defensins delivered prior to 37 weeks, compared with 31% of women in the study sample. Therefore, the availability of additional defensin measures may have given us better power for many of the analyses.
Identifying the factors that predispose women to deliver before 30–37 weeks’ gestation is essential. The observed association between elevated defensins and early preterm birth encourages further studies to examine these associations. Subsequent studies could investigate the predictive value for preterm birth of a combination of factors, such as cervical length, bacterial vaginosis status, vaginal fluid defensins, and parity.25
Future studies should incorporate multiple tests for defensins at various points during gestation. This would help to explain whether the predictive value of defensins in preterm birth is truly restricted to a limited window of time, extending 3–8 weeks after measurement of defensins. If true, defensins would predict preterm birth in a manner similar to fetal fibronectin.26 However, based on the function of defensins in the human body and the potential role that they might play in the immunologic balances of the reproductive tract, defensins could be involved in the etiology of preterm birth rather than being just markers for increased risk of early birth.
|
Footnotes |
|---|
The authors thank the project manager, Jude F. Williams; the clinic site coordinators, Barbara Eucker and Anne Carter; the obstetrics clinics and their staffs for their cooperation, particularly Dr. Peter Morris, Ida Dawson, Dr. Cathi Weatherly-Jones, Dr. Juan Granados, Dr. Thad McDonald, and Sara Caviness; and the many obstetric care providers in these clinics who assisted in the collection of specimens.
doi:10.1016/S0029-7844(03)00042-5
Received June 10, 2002. Received in revised form September 2, 2002. Accepted November 7, 2002.
|
REFERENCES |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Resch B, Vollaard E, Maurer U, Haas J, Rosegger H, Muller W. Risk factors and determinants of neurodevelopmental outcome in cystic periventricular leucomalacia. Eur J Pediatr 2000;159:663–70.[Medline]
3. Okumura A, Hayakawa F, Kato T, Kuno K, Watanabe K. MRI findings in patients with spastic cerebral palsy. I: Correlation with gestational age at birth. Dev Med Child Neurol 1997;39:363–8.[Medline]
4. Speer CP. New insights into the pathogenesis of pulmonary inflammation in preterm infants. Biol Neonate 2001; 79:205–9.[Medline]
5. Rieder RF, Thomas L. Studies on the mechanisms involved in the production of abortion by endotoxins. J Immunol 1960;84:189–93.
6. Kass EH, McCormack WM, Lin JS, Rosner B, Munoz A. Genital mycoplasmas as a cause of excess premature birth. Trans Assoc Am Physicians 1981;94:261–6.[Medline]
7. Naeye RL. Factors that predispose to premature rupture of the fetal membranes. Obstet Gynecol 1982;60:93–8.
8. Gibbs RS, Blanco JD, St. Clair PJ, Castaneda YS. Quantitative bacteriology of amniotic fluid from women with clinical intraamniotic infection at term. J Infect Dis 1982; 145:1–8.[Medline]
9. Romero R, Roslansky P, Oyarzun E, Wan M, Emamian M, Novitsky TJ, et al. Labor and infection. II. Bacterial endotoxin in amniotic fluid and its relationship to the onset of preterm labor. Am J Obstet Gynecol 1988;158:1044–9.[Medline]
10. Hillier SL, Krohn MA, Nugent RP, Gibbs RS. Characteristics of three vaginal flora patterns assessed by gram stain among pregnant women. Vaginal Infections and Prematurity Study Group. Am J Obstet Gynecol 1992;166:938–44.[Medline]
11. Hill GB. Preterm birth: Associations with genital and possibly oral microflora. Ann Periodontol 1998;3:222–32.[Medline]
12. Kimberlin DF, Andrews WW. Bacterial vaginosis: Association with adverse pregnancy outcome. Semin Perinatol 1998;22:242–50.[Medline]
13. Brocklehurst P, Hannah M, McDonald H. Interventions for treating bacterial vaginosis in pregnancy. Cochrane Database Syst Rev 2000:CD000262.
14. Ganz T. Defensins in the urinary tract and other tissues. J Infect Dis 2001;183 Suppl:S41–2.
15. Thomas NJ, Carcillo JA, Doughty LA, Sasser H, Heine RP. Plasma concentrations of defensins and lactoferrin in children with severe sepsis. Pediatr Infect Dis J 2002;21: 34–8.[Medline]
16. Zasloff M. Antibiotic peptides as mediators of innate immunity. Curr Opin Immunol 1992;4:3–7.[Medline]
17. Lehrer RI, Lichtenstein AK, Ganz T. Defensin: Antimicrobial and cytotoxic peptides of mammalian cells. Annu Rev Immunol 1993;11:105–28.[Medline]
18. Balu RB, Savitz DA, Ananth CV, Hartmann KE, Miller WC, Thorp JM, et al. Bacterial vaginosis and vaginal fluid defensins during pregnancy. Am J Obstet Gynecol 2002; 187:1267–71.[Medline]
19. Savitz DA, Dole N, Williams J, Thorp JM, McDonald T, Carter AC, et al. Determinants of participation in an epidemiological study of preterm delivery. Paediatr Perinat Epidemiol 1999;13:114–25.[Medline]
20. Heine RP, Wiesenfeld H, Mortimer L, Greig PC. Amniotic fluid defensin: Potential markers of subclinical intrauterine infection. Clin Infect Dis 1998;27:513–8.[Medline]
21. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol 1991;29:297–301.
22. Harrell FE Jr, Lee KL, Pollock BG. Regression models in clinical studies: Determining relationships between predictors and response. J Natl Cancer Inst 1998;80:1198–202.
23. Barlow WE, Ichikawa L, Rosner D, Izumi S. Analysis of case-cohort designs. J Clin Epidemiol 1999;52:1165–72.[Medline]
24. Yang D, Chertov O, Oppenheim JJ. The role of mammalian antimicrobial peptides and proteins in awakening of innate host defenses and adaptive immunity. Cell Mol Life Sci 2001;58:978–89.[Medline]
25. Cook CM, Ellwood DA. A longitudinal study of the cervix in pregnancy using transvaginal ultrasound. Br J Obstet Gynaecol 1996;103:16–8.
26. Leitich H, Egarter C, Kaider A, Hohlagschwandtner M, Berghammer P, Husslein P. Cervicovaginal fetal fibronectin as a marker for preterm delivery: A meta-analysis. Am J Obstet Gynecol 1999;180:1169–76.[Medline]
This article has been cited by other articles:
|
|
 |
|
  J. Xu, C. B. Holzman, C. G. Arvidson, H. Chung, and A. R. Goepfert Midpregnancy Vaginal Fluid Defensins, Bacterial Vaginosis, and Risk of Preterm Delivery Obstet. Gynecol., September 1, 2008; 112(3): 524 - 531. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
