Objective Twins have inconsistently shown poorer outcomes than singletons. Although a high proportion of twins are born very preterm, data are sparse on the long-term outcomes in very preterm twins. The objective of this study was to compare mortality and neurodevelopmental outcomes of very preterm singletons and twins and to study outcomes in relation to factors specific to twins.

Design Birth cohort study Etude Epidemiologique sur les Petits Ages Gestationnels (EPIPAGE).

Setting Nine regions in France.

Patients All very preterm live births occurring from 22 to 32 weeks of gestation in all maternity wards of nine French regions in 1997 (n=2773).

Main outcomes measures Neurodevelopmental status, including cerebral palsy, and a cognitive assessment with the Kaufman Assessment Battery for Children, with scores on the Mental Processing Composite (MPC) scale, was available for 1732 and 1473 children at 5 years of age, respectively.

Results Among live births, twins had higher hospital mortality than singletons (adjusted (a)OR: 1.4 (95% CI 1.1 to 1.9)). Among survivors, there was no crude difference at 5 years between twins and singletons in the prevalence of cerebral palsy (8.0% vs 9.1%, respectively), MPC <70 (9.5% vs 11.1%) and mean MPC (94.6 vs 94.4). However, after adjustment for sex, gestational age, intrauterine growth restriction and social factors, twins were more likely to have lower MPC scores (mean difference: −2.4 (95% CI–4.8 to 0.01)). Live born twins had a higher risk of mortality when birth weight discordance was present (aOR:2.9 (95% CI 1.7 to 4.8)), but there were no differences in long-term outcomes.

Conclusions Compared with very preterm singletons, twins had higher mortality, no difference with respect to severe deficiencies, but slightly lower MPC scores at 5 years.