Placental infarcts in the collaborative perinatal project: Variable associations infer variable constructs.

Purpose: Reproducible diagnoses of placental infarcts may permit more accurate assessment of their clinical significance. Using data across the 12 study sites of the National Collaborative Perinatal Project, we investigated the consistency of associations between infarct features with birthweight, placental weight and measures of placental "efficiency."

Methods: All delivered infants, live or stillborn, single or multiple, regardless of gestational age, were included. Pathologists scored infarcts by color (tan-white or "old" or pink-red "more recent"), size (cm), location (marginal or central), and total number.

Results: Incidence of any infarcts and distributions of specific features such as size, color (indicating age), locations and total numbers of infarcts were highly variable across sites, as were their associations with birthweight and placental efficiency. The most stable associations (consistent results across sites) of placental infarct scores were with placental size and/or other placental shape variables and with birthweight, but the number of significant associations ranged from 13 to 1.

Conclusion: Given the extremes of infarct incidence within each site plus the variable correlations of infarct features with other placental and birth outcome measures, CPP infarct scores cannot be used as indicative of an underlying shared pathophysiologic construct. However, given the accumulating evidence that intrauterine stressors have the potential for lifelong impact on health, we propose that the infarct features and distinctions proposed are neither complex nor should they be jettisoned. Rather these measures should be clarified and refined. Only then can we understand the reported associations of placental infarcts with child and adult health outcomes.