The incidence of preterm delivery was 28.2% of all twin deliveries. Preterm delivery was common in monozygotic twins due to a high incidence of spontaneous rupture of the membranes. In addition, preterm delivery was associated more often with boy rather than girl infants.

During the last 33 years, the perinatal mortality of twins has decreased regularly in our department. When examined by duration of pregnancy, mortality was lowest at 37–38 weeks. Below 38 weeks, fetal loss was paradoxically higher in twins born from multiparae. When delivery occurred beyond 38 weeks, mortality rates were increased, but only differences for primiparous group showed statistical significance. Near-term fetal loss was predominantly due to antepartal intrauterine deaths. Analysis of individual cases indicated that early termination of pregnancy might help prevent near-term stillbirths in twins delivered by primiparae. Efforts should be directed at specific detection of twins threatened by antepartum death.

Energy and protein intake as measured by 24-hour urinary nitrogen values are similar in twin and singleton pregnancies. The relationship between urinary nitrogen and nitrogen intake is equally significant in twin and singleton pregnancies. Dietary zinc, copper, and iron are not different in women with twins, nor are the levels of these elements in plasma. These observations are surprising in view of the extra fetal demands on the mother and the different adaptation of twin pregnancies.

Five cases of twin pregnancy associated with elevated maternal serum alpha-fetoprotein levels are presented. In each instance an abnormality was detected by ultrasound examination—in four cases this represented anomalous development of one or both fetuses: discordant anencephaly; a monozygotic heterokaryotypic twin pregnancy with coexistent Turner's syndrome and intrauterine death; concordant body-stalk syndrome; and fetal exomphalos together with an acardiac monster. In the remaining case, there had been an intrauterine death of one of the twins. Ultrasound examination proved valuable in all instances; three pregnancies were terminated and single healthy infants were delivered from each of the other two pregnancies.

Serial ultrasonic measurement of the biparietal diameter is an unsatisfactory means of detecting the small-for-gestational age (SGA) fetus in twin pregnancies. A new two-stage ultrasound examination schedule, highly effective in detecting the SGA singleton fetus, has been evaluated prospectively in 31 twin pregnancies. The schedule comprises ultrasonic assessment of gestational age in early pregnancy, followed by measurement of the product of the crown–rump length and trunk area of both fetuses at 34–36 weeks. All Nineteen SGA twin fetuses were detected using this schedule; the technique offers several other advantages over serial biparietal cephalometry.

The maternal risk factors that correlate with small-for-dateness among twins have been analyzed using a sample of 659 twin pairs and a matched sample of singletons. Non-marital status, job involvement, and the previous delivery of a low-birth weight (<2,500g) infant present a negative interaction with twinning, as low gestational age-adjusted birth weight does not correlate significantly with these risk factors among twin gestations, while it does among singleton gestations. On the other hand, the effects of parity, habitual maternal weight, smoking during pregnancy, and twinning are additive on gestational age-adjusted birth weight. Indeed, the decrease in adjusted birth weight associated with these risk factors is of the same magnitude among twins and singletons and is statistically significant in both cases. These findings suggest that exposure of twin pregnancies to these latter risk factors, and particularly to smoking during pregnancy, can lead to the delivery of newborns with extremely low birth weights.

A literature review of the occurrence of multiple pregnancies associated with artificial induction of ovulation is reported. This report considers three treatment schedules: (1) clomiphene citrate; (2) human pituitary gonadotrophin with human chorionic gonadotrophin; and (3) human menopausal gonadotrophin with human chorionic gonadotrophin. The majority of the increase in twinning is related to hyperstimulation of the ovary by these medications, resulting in dizygotic twinning. The true incidence of twin pregnancy cannot be calculated because the vital statistics of all nations report live birth rates. Increased rates of fetal wastage, late abortion and prematurity associated with the occurrence of multiple pregnancies are overlooked by these statistics. The increased incidence of twinning appears to be related to the type and dosage of medication used, and the patient's underlying problem.

The lecithin/sphingomyelin ratio showed no difference between the two amniotic sacs in 24 multiple pregnancies examined before the onset of labor. Once labor started, however, the L/S ratio was higher in the first sac compared to the second in the six pregnancies where labor had commenced and was somewhat advanced at the time of amniocentesis. Dual-sac amniocentesis is essential during labor when assessment of pulmonary maturity is required, whereas single-sac sampling is adequate before the start of labor.

Twenty-two twin pregnancies each with one growth-retarded and one normal twin fetus were examined to determine the effects of fetal growth retardation on the maturation of the adrenal gland with respect to cortisol and dehydroepiandrosterone sulphate (DHEAS). The growth-retarded twin fetuses (IUGR) had lower umbilical arterial concentrations of DHEAS than their siblings (IUGR 5.25 ± 2.4; non-IUGR 6.51 ± 2.9 μmol/l; p < 0.01), whereas cortisol concentrations were not statistically different (no labor, IUGR 1,134 ± 751, non-IUGR 1,140 ± 958 μmol/l; labor IUGR 2,062 ± 929, non-IUGR 1,609 ± 469 μmol/l). These data suggest that while the definitive zone of the fetal adrenal is as well-developed as in non-growth-retarded twins, the fetal zone shows reduced secretory capacity of Δ5-steroids. This supports the hypothesis that in growth retardation the adrenal gland shows features of increased maturation.

The antenatal management and outcome of 31 higher-order multiple gestations, 24 triplets, six quadruplets, and one set of quintuplets were analyzed. Bedrest was advised as soon as the diagnosis was made; hospitalization along with the administration of betamethasone and phenobarbital was begun during the second trimester. Caesarian section was performed on 13 women (42%). The overall mortality rate was 14.8%; excluding neonates less than 28 weeks of age, however, it became 7.5%. The primary cause of death was respiratory distress syndrome.