Preterm Twins Are Born Smaller but Catch Up Faster Than Singletons, Global Study Finds

Preterm twins enter the world smaller than singletons, but most of that apparent disadvantage in early postnatal growth turns out to be inherited from the womb rather than earned in the nursery. That is the central finding of a large, prospective, multinational analysis from the INTERPRACTICE-21st study, published in the Journal of Perinatology, which followed 3,245 preterm newborns—2,148 singletons and 1,097 twins—across 14 countries on five continents. By classifying infants according to the underlying cause of their early delivery rather than gestational age alone, the researchers uncovered a far more nuanced picture of twin growth than previous comparisons had allowed, one that could reshape how clinicians monitor the smallest and most vulnerable multiple-birth babies.

The scale of the twinning phenomenon makes the question clinically urgent. Between 2010 and 2015, the global twinning rate stood at 12.0 twin deliveries per 1,000 deliveries, ranging from 9.2 per 1,000 in Asia to 17.1 per 1,000 in Africa, with the variation driven almost entirely by differences in dizygotic—fraternal—twinning, since the monozygotic rate remains remarkably stable at roughly 4 per 1,000 births worldwide. Since the 1980s, the overall rate has climbed by about a third, from 9.1 per 1,000 births, a rise attributed largely to older maternal age at conception and the expanding use of medically assisted reproduction. With more multiple pregnancies than ever before, parents and healthcare providers face growing concern about the complications that accompany them.

Twin pregnancies carry a substantially elevated risk profile. Compared with singletons, twins face higher rates of congenital abnormalities, miscarriage, stillbirth and impaired fetal growth, along with greater neonatal morbidity and mortality—driven chiefly by their dramatically higher preterm birth rate, which a population-based registry of nearly 500,000 pregnancies put at 54.7 percent for twins versus 6.1 percent for singletons. Fetal growth in twins generally mirrors that of singletons through the first and second trimesters, but from about 32 weeks’ gestation onward twins grow more slowly, a pattern researchers describe as relative growth restriction. This divergence has prompted some groups to develop twin-specific fetal growth charts and newborn size references.

What remained largely unknown until now was how preterm twins grow after birth, and whether that growth depends on why they were born early. The INTERPRACTICE-21st study, conducted from January 2018 to September 2020 in city hospitals and maternity institutions in Argentina, Brazil, France, Ghana, Italy, India, Kenya, North Macedonia, Mexico, Pakistan, Russia, Spain, Turkey and the United Kingdom, was designed to answer precisely that. All infants were born alive between 23+0 and 36+6 weeks’ gestation, and gestational age was estimated by ultrasound measurement of crown-rump length before 14 weeks or head circumference before 24 weeks, using the INTERGROWTH-21st standards.

The methodological rigor of the anthropometry is a defining strength of the study. Newborn weight, length and head circumference were measured within 12 to 24 hours of birth and then repeatedly until hospital discharge, using identical equipment at every site: an electronic scale accurate to 0.1 grams for daily weight measurements, a Harpenden infantometer for weekly recumbent length, and a metallic non-extendable tape precise to 1 millimeter for weekly head circumference. Two trained study anthropometrists independently took every measurement twice, with maximum allowable differences of 50 grams for weight, 7 millimeters for length and 4 millimeters for head circumference; any discrepancy beyond those thresholds triggered a third measurement. After excluding 130 triplets and four quadruplets, the analysis included 2,320 singletons and 1,180 twins, of whom 3,245 had at least one postnatal measurement.

The researchers then applied a two-step cluster analysis to assign each newborn to one of nine etiologically defined preterm birth subgroups: pre-eclampsia, extrauterine infections, chorioamnionitis, perinatal sepsis, early bleeding, mid-late bleeding, suspected intrauterine growth restriction, fetal distress and congenital anomaly, plus a no-main-condition group. Infants with congenital anomalies were excluded from further analysis. Growth was expressed as z-scores—standard deviations from the sex- and age-adjusted mean—calculated against the INTERGROWTH-21st Newborn Size Standards at birth and the Preterm Postnatal Growth Standards afterward. Multilevel mixed-effects linear regression models accounted for repeated measurements per infant and per mother, adjusting for maternal age, parity, mode of delivery, fertility drug use, feeding type and age at discharge, with additional models adding the birth z-score to separate intrauterine from postnatal effects.

The birth-size results were strikingly dependent on both gestational age and etiology. Twins were significantly smaller than singletons at birth in the late preterm category—34+0 to 36+6 weeks—across weight, length and head circumference, and lower in birth weight in the moderate preterm category, but showed no significant differences in the very preterm category. Across etiological subgroups, twin measures were lower in the no-main-condition, chorioamnionitis and mid-late bleeding groups, with birth weight and length lower in the perinatal sepsis subgroup. Yet in the pre-eclampsia and fetal distress subgroups, twin and singleton birth measures did not differ at all. Most intriguingly, in the suspected IUGR subgroup, all birth measures were lower in singletons, suggesting that in twins, apparent growth restriction often reflects the mechanical and nutritional constraints of sharing a womb, whereas in singletons it signals genuine pathology.

Postnatal growth told an equally nuanced story. In unadjusted models, late preterm twins grew more slowly than singletons, and twins in the no-main-condition, chorioamnionitis, perinatal sepsis and mid-late bleeding subgroups showed lower postnatal size on one or more measures. But once the models were adjusted for birth size, most of these differences attenuated or vanished entirely. After adjustment, only weight z-scores in the very preterm category remained significantly lower in twins, along with weight in the perinatal sepsis subgroup and head circumference in the mid-late bleeding subgroup. Meanwhile, twins in the pre-eclampsia subgroup actually exceeded singletons in adjusted postnatal weight, and twins in the suspected IUGR subgroup—larger than singletons even before adjustment—showed compensatory catch-up growth, likely reflecting greater fat mass accumulation, particularly among late preterm infants.

The perinatal sepsis finding stands out as the most clinically consequential. That twins in this subgroup exhibited persistently poorer growth even after accounting for birth size indicates that infection or associated unclassified morbidities may uniquely impair postnatal growth in multiple-birth infants, marking them out for targeted interventions and closer nutritional and developmental follow-up. Conversely, the accelerated weight gain observed in twins recovering from suspected intrauterine growth restriction warrants long-term monitoring: population-based studies have linked rapid infant weight gain to increased adiposity in childhood, although the specific long-term consequences for preterm infants remain uncertain. Notably, fertility drug use, though far more common among the twins’ mothers at 26.2 percent versus 7.3 percent, did not meaningfully alter growth patterns in sensitivity analyses.

Beyond its immediate clinical implications, the study delivers a strong endorsement of the international Preterm Postnatal Growth Standards, produced in healthy, adequately nourished populations with optimal perinatal care. The authors conclude that these standards constitute a valid, universal reference for monitoring postnatal growth in preterm twins across all etiological subgroups and gestational age categories—with particular relevance for the subgroups, such as perinatal sepsis and suspected IUGR, that deviate most markedly from the healthy reference trajectory. The work also carries caveats: no long-term follow-up, potential residual confounding from unmeasured factors such as placental pathology and zygosity, limited granularity in feeding data, and some heterogeneity within etiological subgroups due to overlapping conditions. Even so, the message is clear: when evaluating a preterm twin’s growth, what happened in the womb—and why delivery came early—matters as much as what happens in the incubator.

Subject of Research: Birth size and postnatal growth patterns of preterm twins compared with singletons by etiology of preterm birth

Article Title: Distinct birth size and postnatal growth patterns in preterm twins and singletons: an etiology-based analysis from the INTERPRACTICE-21st study

Article References: Distinct birth size and postnatal growth patterns in preterm twins and singletons: an etiology-based analysis from the INTERPRACTICE-21st study. (n.d.). https://doi.org/10.1038/s41372-026-02912-6

Image Credits: AI Generated

DOI: 10.1038/s41372-026-02912-6

Keywords: preterm birth, twins, postnatal growth, INTERGROWTH-21st, birth weight, intrauterine growth restriction, perinatal sepsis, pre-eclampsia, neonatology, growth standards, multiple pregnancy, pediatric research