POSSIBLY In this retrospective cohort study from Sweden, when the expected date of delivery was postponed more than 7 days as a result of early (16 weeks) ultrasonographic dating, the risk of a small-forgestational-age infant increased (odds ratio [OR], 1.77; 95% confidence interval [CI], 1.13–2.78 and OR 2.09; 95% CI, 1.59–2.73). This study involved a population of 28,776 singleton pregnancies dated between 1998 and 2004.
Conventional perinatal wisdom, since the inception of obstetric ultrasonography (US), has been that disordered growth in utero occurs only in the second half of pregnancy; growth in the first half of pregnancy is believed to be uniform, with little variation among individuals. This assumption of uniform growth at the beginning of gestation allows us to create growth curves for populations and generate estimates of gestational age for individual fetuses from their growth parameters. Utilization of US for dating has pushed the mean gestational age at delivery back a few days, tightened the distribution around the mean, and lowered the prevalence of postdatism.
In this new study, Thorsell and colleagues question conventional wisdom and introduce a new notion that disordered intrauterine growth may be present in the first half of pregnancy as early as the first trimester. Women whose US evaluation at 16–18 weeks moved their due date forward more than 6 days were at increased risk of intrauterine growth restriction, preterm birth, and preeclampsia. Those whose due date was moved forward more than 6 days as a result of US dating at 12–14 weeks were at increased risk of growth restriction, but not preterm birth or preeclampsia. The authors call for increased surveillance for growth restriction in pregnancies in which US evaluation changes dates.
Weaknesses of the study
These findings are intriguing, but take them with a grain of salt. “Intendedness” of conception can, of course, be a marker of higher social status and resources, thereby linking “unintendedness” to poor dates (dates that need to be adjusted by US) and poor pregnancy outcomes. To prove their point, Thorsell and associates would have to repeat the study in women using ovulation-prediction methods or assisted reproduction (which would be confounded by subfertility and its link to poor perinatal outcomes). Such a study would not be feasible, given that a sample size of more than 27,000 women was required to demonstrate very mild effects in this investigation (risk ratios from 1.1 to 1.5).
This provocative study challenges convention but is not ready for incorporation into clinical practice routines. However, it may be prudent to monitor pregnancies in which US dating significantly changes the due date, keeping in mind a potential for intrauterine growth restriction.—JOHN M. THORP JR, MD