A woman's risk of having a preterm delivery is influenced largely by conditions that occur during pregnancy, but a clinic-based longitudinal study in the western United States demonstrates that her health status and health behaviors prior to conception also may play a substantial role.1 In analyses controlling for demographic characteristics and risk factors both before and during pregnancy, the odds of preterm delivery were nearly doubled for women whose level of physical function before conception was poor and for those who had chronic hypertension before becoming pregnant. Factors that predated pregnancy accounted for 40% of the variability in risk of preterm delivery.

The study cohort comprised women who received prenatal care at a site affiliated with one of six hospitals in the San Francisco Bay area of California. To be eligible, women had to be at least 18 years old at recruitment (between May 2001 and July 2002), had to have begun prenatal care before 16 weeks' gestation and had to be planning to deliver at one of the six hospitals.

Participants were asked to complete four telephone interviews: before 20 weeks' gestation, at 24–28 weeks, at 32–36 weeks and at 8–12 weeks after delivery. Each interview included questions from a standard instrument assessing physical, mental and emotional health, as well as screening for depressive symptoms. Questions in the baseline interview referred to the month before conception, and those in later interviews referred to the previous four weeks. The baseline interview also collected detailed information about participants' demographic characteristics and their medical conditions and health-related behaviors before pregnancy; subsequent interviews included questions about pregnancy complications. Additional data about the pregnancy and delivery were obtained from women's medical records.

The researchers restricted their analyses to the 1,619 women who had a singleton delivery at one of the participating hospitals. Most of these women were married or living with a partner, had been born in the United States and had given birth before. The cohort was racially, ethnically and socioeconomically diverse.

Eight percent of the women delivered pre-term (i.e., at less than 37 weeks' gestation). Initial analyses suggested that a wide variety of factors were associated with the risk of this outcome; the researchers conducted a series of multivariate logistic regression analyses to determine which of these were independently associated with the odds of preterm delivery.

When only demographic characteristics were considered, the odds of preterm delivery were significantly elevated for black women (odds ratio, 1.9) and for women who completed high school but not college (1.7). With the addition of conditions and behaviors in the month before conception, demographic characteristics were no longer significant; the odds were elevated for women who had been underweight (2.4), those whose physical function had been poor (2.3), those who had suffered from chronic hypertension (3.1) and those who had smoked (2.2). In analyses that included pregnancy-related factors, most of these associations remained significant (the exception was being underweight), although the odds ratios were reduced. Pregnancy-associated hypertension and other complications were associated with increased odds of preterm birth (3.2 and 2.2, respectively). The researchers estimate that 13% of the variation in the risk of preterm delivery was attributable to demographic factors, 40% to conditions and behaviors that preceded pregnancy and 47% to factors occurring during pregnancy.

The researchers caution that the associations they found are not causal and relate only to factors that were present immediately before conception; both a woman's health status before pregnancy and her risk of delivering preterm may be associated with conditions that occurred earlier in her life. They conclude that interventions and policies directed at improving access to care during pregnancy may fall short of the goal of reducing preterm delivery because they cannot address [a] legacy of poor health status and health behaviors. Reducing the incidence of preterm delivery, they maintain, may require attention to the health status of women before pregnancy.—D. Hollander


1. Haas JS et al., Prepregnancy health status and the risk of preterm delivery, Archives of Pediatrics and Adolescent Medicine, 2005, 159(1):58–63.