Factors affecting decidual IGFBP-1 levels in the vagina and cervix in the first and mid-second trimester of pregnancy.

Abstract

The aim of this study was to determine the concentrations of and factors associated with decidual insulin-like growth factor-binding protein-1 (IGFBP-1) in the lower genital tract in early- and mid-gestation in singleton pregnancies. Prospective population-based cohort study. Maternity Clinic, Department of Obstetrics and Gynaecology, University Central Hospital, Helsinki, Finland. A total of 1702 unselected pregnant women undergoing the first- and the second-trimester ultrasound screening between April 2005 and December 2006. The vaginal and cervical swab samples for assay of decidual IGFBP-1 and vaginal pH measurement were taken before transvaginal ultrasonography in the first trimester and in the mid-second trimester. Use of antibiotics, history of vaginal bleeding, and the history of sexual intercourse were questioned on both occasions. The concentration of IGFBP-1 was measured by a quantitative immunoenzymometric assay, which detects the decidual phosphoisoforms of IGFBP-1 (phIGFBP-1). The concentration of 10 micrograms/l was used as a cutoff when factors influencing phIGFBP-1 levels were analysed. The phIGFBP-1 concentrations in the vagina and the cervix and associations between the levels of > or =10 micrograms/l and selected factors. In the first trimester, the median (range) concentrations of phIGFBP-1 in vaginal and cervical samples were <0.3 micrograms/l (<0.3-176 micrograms/l) and 4.8 micrograms/l (<0.3-174 micrograms/l), respectively. During the second trimester, the corresponding values were <0.3 micrograms/l (<0.3-55 micrograms/l) in the vagina and 3.6 micrograms/l (<0.3-126 micrograms/l) in the cervix. In the vaginal samples, the frequency of phIGFBP-1 concentrations > or =10 micrograms/l was 5.8% in the first trimester and 1.5% in the second trimester (P < 0.001). In the cervical samples, the corresponding rates were 34.3 and 28.4%, respectively (P < 0.001). Of the factors studied, nulliparity (P < 0.001) and history of vaginal bleeding (P < 0.001) were independently associated with cervical phIGFBP-1 concentrations > or =10 micrograms/l during both trimesters. In addition, short cervical length (<30 mm) was associated with phIGFBP-1 concentration > or =10 micrograms/l in both vaginal and cervical samples in the second trimester in multivariate analysis. The rate of phIGFBP-1 concentrations > or =10 micrograms/l, both in the vagina and in the cervix, was significantly lower during the second trimester compared with the first trimester. The low rate of levels > or =10 micrograms/l in vaginal samples compared with cervical samples during both trimesters indicates that the exact site of sampling is important when phIGFBP-1 is used as a decidual marker. Nulliparity and history of vaginal bleeding were independently associated with phIGFBP-1 concentrations > or =10 micrograms/l in cervical samples during both trimesters.