Background: Previous studies found that increased birthweight discordance among twins was associated with increased risk of perinatal mortality and morbidity, but there is disagreement regarding the relationship between birthweight discordance and the risk of preterm birth.
Aims: To examine the relationship between intratwin birthweight discordance and preterm birth, and the influence of total birthweight as a confounding factor.
Methods: We conducted a historical cohort study to assess the relationship between intratwin birthweight discordance and preterm birth, using the 1995–1997 twin registry data of the USA. Parallel analyses with and without adjustment for total birthweight were performed.
Results: A total of 147 262 eligible twin sets were included in the analysis. In the non-stratified analysis, preterm birth (< 37 weeks) rates were 64.3 and 51.8%, respectively, in subjects with a birthweight discordance > 25% and in those ≤ 25%. In the analysis stratified by total birthweight, the rates of preterm birth were higher in subjects with a birthweight discordance > 25% than in those ≤ 25% at every total birthweight stratum. Changing the thresholds to define preterm birth or treating gestational age as a continuously distributed variable yielded similar results. Using other methods to adjust for total birthweight also reversed the direction of this relationship.
Conclusions: Increased intratwin birthweight discordance is associated with increased risk of preterm birth, and adjustment for total birthweight biases the relationship to the wrong direction. Clinicians should be aware of the possibility of a preterm birth when there are signs of discordance in birthweight between the two twins.