Twin pregnancy is a rare event. The typical twin pregnancy rates in Europe and north America range between 9 and 20 per thousand, making it difficult to accumulate a sufficient sample size.

Some twin studies are based in university teaching hospitals, giving rise to a doubt of referral bias.

Twins and singletons have different causes of per-term delivery. It seems that pre-term delivery is more pathological in singletons than in twins. Statistical control for or matching by gestational age may introduce confounding due to underlying pathology while attempting to adjust for confounding due to a difference in gestation age.