Predicting risk of preterm birth: the roles of stress, clinical risk factors, and corticotropin-releasing hormone.

The relationships and predictive abilities of perceived stress, selected clinical risk factors, and corticotropin-releasing hormone (CRH) levels in maternal plasma were investigated for their association with preterm labor (PTL), preterm birth, and gestational age at delivery. An exploratory, prospective, longitudinal research design was used to measure CRH 4 times over pregnancy, perceived stress at 24 and 32 weeks of pregnancy, clinical risk factors, and genitourinary infections in low-income women. Multiple regression analyses revealed that a model of measurement ofperceived stress at 24 or 32 weeks, CRH at 24 or 32 weeks, and PTL (indicated by a diagnosis by the physicians on the medical record and greater than 5 contractions per hour on the fetal monitor) was predictive of as much as 0.23 to 0.27 of the variance in gestational age at birth. Entering ethnicity as a variable into a model did not improve the predictive value. An analysis of variance between Caucasian and Hispanic ethnic groups revealed differences in CRH levels. However, simple regression analysis of ethnicity as a predictor showed a weak r2 with no significance for prediction. There was some evidence of an association of smoking with stress and both PTL and preterm birth. The measurement of stress combined with the measurement of CRHfrom maternal plasma may improve the prediction of which pregnant women are at riskfor pretern birth. The measurement of CRH has potential as an early biological marker of preterm birth.