The origin and outcome of preterm twin pregnancies

The Origin and Outcome Pregnancies

of Preterm

MICHAEL 0. GARDNER, MD, ROBERT L. GOLDENBERG, MD, SUZANNE P. CLIVER, BA, JAMES M. TUCKER, MD, KATHLEEN RACHEL L. COPPER, MSN, CRNP Objective: To define the etiology of preterm twin births and determine the contribution of twin births to preterm birth and related morbidity and mortality. Methods: The March of Dimes Multicenter Prematurity and Prevention Study included a total of 33,873 women who delivered between 1982-1986,432 (1.3%) of which delivered twins. Women were classified by reason for preterm birth and ethnicity. Neonates were classified as to stillbirth, neonatal death, and various short-term morbidities. A second data set from one center consisted of infants who weighed 1000 g or less, were born between 1979-1991, and survived to 1 year of age (n = 386, 15% twins); this was used to determine if twins and singletons born at comparable gestational ages have a similar risk for major developmental handicaps. Results: Of the deliveries in the data set, 54% of twins were preterm compared with 9.6% among singletons. Of those born preterm, twins were born at a significantly earlier gestational age than were singletons. Only 2.6% of all neonates born were twins, but they represented 12.2% of all preterm infants, 15.4% of all neonatal deaths, and 9.5% of all fetal deaths. Spontaneous labor accounted for 54% of twin births, premature rupture of membranes accounted for 22%, and indicated deliveries accounted for 23%. Of the indicated preterm births in twins, 44% were due to maternal hypertension, 33% to fetal distress or fetal growth restriction, 9% to placental abruption, and 7% to fetal death. Comparing infants of similar gestational age, twins weighed less, but had a mortality equivalent to that of singletons after 29 weeks. Between 26-28 weeks’ gestation, the risk of mortality for twins versus singletons was 1.6 (95% confidence interval 1.1-2.5). Preterm twins did not have significantly more respiratory distress syndrome, intraventricular hemorrhage, necrotizing enterocolitis, or other short-term morbidity than did preterm singletons. Twins who weighed 500-1000 g and

From the Center for Obstetric Research, Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Alabama at Birmingham, Birmingham, Alabama. Supported in part by the March of Dimes and the Agency for Health Care Policy and Research, contract no. DHHS 282-92-0055.

VOL.85,NO.4,Al'RIL

1995

Twin

G. NELSON, MD, AND

survived to 1 year had a 25% rate of major developmental handicaps. However, when gestational age was controlled, the rate of major handicaps was not higher in twins than in singletons. Conclusions: Twins accounted for a disproportional amount of preterm birth and associated morbidity and mortality. Also, when preterm twins were compared with preterm singletons and corrected for their gestational ages, the rates of morbidity were similar. Preterm twins weighing less than 1000 g did not have an increased prevalence of major handicaps at 1 year of age compared with preterm singletons. (Obstet Gynecol 2995;85:553-7)

Twin pregnancies have a higher rate of premature delivery than singleton pregnancies’ and a substantially higher perinatal mortality.2,3 The origins of preterm birth have been divided into three groups: 1) those after spontaneous labor, 2) those after premature rupture of membranes (PROM), and 3) indicated preterm births (ie, those after a medical decision that delivery is necessary for maternal or fetal concerns). Previous studies4,’ have evaluated these etiologies of preterm births in singleton pregnancies. The very low birth weight (VLBW) infant (less than 1500g) is the major contributor to perinatal mortality in twins and singletonszZ6Specialized prenatal care programs for twins have emphasized the prevention of preterm twin deliveries.7,sHowever, there is a disparity between twin and singleton preterm delivery rates. Before an effective prevention program is possible, an understanding of the causes of twin preterm birth is necessary. Similarly, knowledge of the effects of twin preterm birth, including short- and long-term morbidity, is needed when planning twin pregnancy interventions. In this study, we reviewed data collected prospectively to elucidate the incidence of preterm deliveries in twins. We then evaluated the causesof these preterm twin births and the contributions of twins to perinatal

0029.7844/95/$9.50 0029-7844(94)00455-M

553

mortality and morbidity. Finally, we compared VLBW twins and singletons for differences in rates of major developmental handicaps at 1 year of age. Percent Delivered

Materials and Methods Maternal demographic and obstetric data and neonatal outcome data were collected prospectively as part of a multicenter trial. The March of Dimes Multicenter Prematurity and Prevention Study data set included all nonreferral deliveries from November 1982 through October 1986 at the University of California, San Diego, Ohio State University, Vanderbilt University, and Northwestern University, and deliveries at the University of Alabama at Birmingham from November 1982 through April 1986.9 All centers had Institutional Review Board approval. We excluded from analysis those women who left the study before delivery, those who delivered before 20 weeks’ gestation, and multiple gestations of more than twins. Of a total of 33,873 deliveries, there were 432 sets of twins reviewed for this study. The protocol for gestational age dating in the March of Dimes data base has been described previously.” Gestational age was based on the last menstrual period, physical examination at the first prenatal clinic visit, and ultrasonography, if performed. A modified Dubowitz examination was performed to determine neonatal gestational age. The obstetric estimate was used if the obstetric and neonatal ages agreed within 3 weeks, but if there was a greater discrepancy or if the delivery was classified as preterm, the case was reviewed before assignment of gestational age. The preterm delivery rates of twins and singletons were compared. In addition, the preterm twin births were further categorized by the cause of delivery, and comparisons were then made to the causes of singleton preterm births. Fetal, perinatal, and neonatal mortality rates were calculated for twins and singletons. Also, the rates of neonatal morbidity, including respiratory distress syndrome, intraventricular hemorrhage, and necrotizing enterocolitis, were compared between the two groups. Univariate analyses used 2, with significant differences defined as P < .05. Relative risks (RR) and 95% confidence intervals (Cl) were calculated. All statistical analyses were performed with SAS software (SAS Inc., Cary, NC). We analyzed the twin infants asindividuals, although this may violate the assumption of independence among data points. The University of Alabama at Birmingham low birth weight (LBW) data set consists of all infants born from 1979-1991 and weighing 500-1000 g. This data set was used because there was no long-term follow-up data

554

Gardner

et al

Preterm

Tzuins

Weeks Gestation Figure 1. Distribution singleton deliveries.

of preterm

twin

deliveries

versus

preterm

available from the March of Dimes data set. Seventyeight percent of those infants who survived to 1 year of age (n = 386) received a comprehensive neurodevelopmental examination at or after 1 year of age (mean 38 months). Of these survivors, 57 (15%) were twins. These twins were compared with surviving singletons for the incidence of major developmental handicaps at 1 year of age. Major handicaps were defined as cerebral palsy, mental retardation, and deafness or blindness alone or in combination.

Results During the time of the March of Dimes Multicenter Prematurity and Prevention Study (1982-1986), the five participating medical centers had a total of 33,873 deliveries, excluding transfers. In this nonreferral population, the incidence of twin births was 1.3%. Of the 432 sets of twins, 233 (54%) delivered before 37 weeks’ gestation compared with a singleton preterm birth rate of 9.6%. Therefore, although the twins represented only 1.3% of the total population, they contributed 12.2% to the total preterm births. The preterm twins tended to be born earlier than the preterm singletons (P = .05) (Figure 1). In this population, the incidence of twin pregnancy was similar between black (1.44%) and white (1.32%) women. Furthermore, maternal race was not a significant factor in the incidence of preterm delivery in twins. For black twins, the preterm delivery rate was 55% compared with a rate of 52% for white twins, a nonsignificant difference. For black twins, spontaneous labor accounted for 54% of the preterm births, PROM for 2110, and indicated deliveries for 25%. Spontaneous labor accounted for 56% of the white preterm twin deliveries, PROM for 20%, and indicated deliveries for 22%. Of the preterm births, spontaneous labor was associated with a higher percentage of preterm deliveries in

Obstetrics

6 Gynecology

Table

1. Mortality

by Plurality

Table

Outcome

Singletons (N = 33,434)

Twins (N = 858)

Relative risk (95% CI)

Contribution of twins to specific outcome (%/o)

All births Stillbirth Neonatal death P&natal death

97.4 1.1 0.9 2.0

2.6 4.4 5.2 9.6

4.0 (2.9-5.6) 5.9 (4.4-8.0) 4.9 (3.9-6.1)

2.6 9.5 15.4 12.2

Specific

outcome

CI = confidence interval. All differences between singletons

by plurality

(a)

and twins significant

at P < ,001.

(54%) than in singletons (44%) (P = .OOl). Although preterm PROM occurred in 12% of all twin pregnancies versus 3% of all singleton pregnancies, PROM contributed significantly less to twin preterm births (22%) than to singleton preterm births (31%) (P = .004). Likewise, obstetric factors leading to indicated preterm birth occurred in 13% of all twins versus 2% of all singletons. Nevertheless, the contribution of indicated preterm deliveries to preterm twins and preterm singletons was the same, 23%. Of the indicated preterm births in twins, 44% were associated with maternal hypertension, 33% with fetal distress or fetal growth restriction (FGR), and 9 and 7%, respectively, with placental abruption and fetal death. Twin and singleton mortality is summarized in Table 1. Although only 2.6% of all fetuses were twins, they accounted for 9.5% of all stillbirths, 15.4% of all neonatal deaths, and 12.2% of perinatal deaths. After 29 weeks, twins had a mortality equivalent to singletons. Between 26-28 weeks’ gestation, the risk of mortality of twins versus singletons was 1.6 (95% CI 1.1-2.5). We examined the mean gestational age in this age group, and there was no significant difference (twins 27.3 weeks, singletons 27.1 weeks). Birth weights were also similar; the mean for twins born between 26-28 weeks was 1008 g, and the mean for singletons was 979 g. Compared with singletons, twins contributed a disproportionate share to neonatal morbidity in our study population. Twin rates of low Apgar scores, respiratory distress syndrome, severe intraventricular hemorrhage, and other morbidity were significantly higher than singleton rates. Table 2 summarizes the rates of neonatal morbidity and the contribution of twin infants to each outcome. We next compared neonatal morbidity in preterm twins and preterm singletons and found no significant differences in the morbidities evaluated, except for the prevalence of respiratory distress syndrome (Table 3). Preterm infants were then stratified by gestational age. Twins and singletons born at 24-28 weeks, 29-34 weeks, and 35-36 weeks were compared. The incidence of respiratory distress syndrome, severe intraventricular hemorrhage, and necrotizing enterocolitwins

VOL.

85,

NO.

4, APRIL

1995

2. Morbidity

by Plurality Specific

Outcome All births I-min Apgar score <7 5-min Apgar

outcome

by plurality

(%I

Contribution of twins to specific outcome (%)

Singletons (N = 33,070)

Twins (N = 820)

Relative risk (95% CI)

97.6 13.7

2.4 27.9

2.0 (1.8-2.3)

2.4 4.8

2.1

7.2

3.4 (2.7-4.4)

7.9

Score <7 IVH (grades

0.2

0.9

5.2 (2.5-10.8)

11.4

3 & 4) Sepsis NEC RDS

0.9 0.4 2.4

3.1 1.6 15.8

3.3 (2.2-5.0) 4.5 (2.57.9) 6.4 (5.4-7.7)

7.6 9.9 13.8

CI = confidence interval; IVH = intraventricular hemorrhage; NEC = necrotizing enterocolitis; RDS = respiratory distress syndrome. All differences between singletons and twins significant at P 4 ,001.

tis was not significantly different between twin and singleton infants born at 24-28 weeks or 29-34 weeks (Table 4). At 35-36 weeks, twins had significantly more necrotizing enterocolitis. The University of Alabama at Birmingham LBW data set was reviewed, and the incidence of major handicaps in twins was compared with that in singleton survivors. One of the 57 (1.75%) twins versus six of the 329 (1.82%) singletons were blind (RR 0.96, CI 0.12-7.84). Five of the twins (8.77%) versus ten of the singletons (3.04%) were deaf (RR 2.89, CI 1.02-8.13). Of the twins, 10.5% suffered from cerebral palsy compared with 13.4% of the singletons (RR 0.79, CI 0.35-1.76). The overall incidence of infants with major handicaps was similar: 15 (26.3%) of the twins and 86 (26.1%) of the singletons (RR 1.10, CI 0.63-1.61).

Discussion Twin gestations have long been recognized as high-risk pregnancies. Fetal mortality rates are known to be substantially higher for twins than for singleton preg-

Table

3. Morbidity

Outcome I-min Apgar score <7 5-min Apgar score <7 IVH (grades 3 & 4) Sepsis NEC RDS

in Preterm

Infants

by Plurality

Preterm singletons (N = 2978)

Preterm twins (N = 420)

P

33.6 13.0 1.9 5.3 3.4 22.3

36.4 11.9 1.9 5.5 3.4 29.0

NS NS NS NS NS ,002

IVH = intraventricular hemorrhage; NEC = necrotizing enterocolitis; RDS = respiratory distress syndrome; NS = not significant. Data are presented as %.

Gardner

et al

Pveterm

Twirls

555

Table

4. Neonatal

Morbidity

by Plurality

and

Gestational

Age

24 -28 Weeks

RDS IVH (grades NEC

Gardner

et al

35-36 Weeks

Singletons (N = 254)

P

Twins (N = 208)

Singletons (N = 1195)

P

Twins (N = 156)

Singletons (N = 1497)

P

76.7 14.0 4.7

77.6 17.3 11.6

NS NS NS

35.6 0.5 3.4

32.1 1.0 5.2

NS NS NS

7.0 0.6 2.6

5.0 0.1 0.7

NS NS .04

distress as %.

svndrome; ’

IVH

= intraventricular

hemorrhage;

nancies, and this was corroborated by the fourfold increased incidence of stillbirth in our study. Preterm delivery is believed to be the major factor in the increased risk of neonatal death in twins.“,” The neonatal mortality rate for twins in our study was significantly higher than that of singletons, with a greater than sixfold increase. However, when the timing of these deaths was examined, this difference was only significant between 26-28 weeks’ gestation. The etiology of preterm twin births was different than preterm singleton births. Spontaneous labor was significantly more likely to be the cause of preterm deliveries in twins than in singletons. Conversely, PROM contributed more to singleton preterm births than to twin preterm births. Although controversial, it has been suggested that uterine overdistention is a cause of preterm labor in twin gestations.13-I5 Subclinical chorioamnionitis secondary to ascending infection associated with premature cervical ripening has also been suggested as an etiology of preterm labor and PROM, but it is not clear if this phenomenon is more common in twins than singletons.‘6,‘7 Preeclampsia and placental abruption are reputed to be more common in twin pregnancies.l* Our study confirmed these associations. Because of the high percentage of preterm birth in twins, indicated preterm delivery accounted for only 23% of preterm births in twins, a rate similar to that in singletons. Still, 12.4% of all twin pregnancies ended with an indicated preterm delivery compared with 2.3% of singleton pregnancies. In our population, maternal hypertension (44%) and fetal distress and FGR (33%) were the main indications for delivery of preterm twins. Bronsteen et alI9 reported that FGR may be the primary cause of stillbirth in twins. Their data also suggest that FGR in individual fetuses is a better predictor of neonatal morbidity than discordant growth between both fetuses. In our population, preeclampsia, FGR, and placental abruption were more common in twins than in singleton pregnancies, contributing to the increased rate of indicated preterm births in twins.” The preterm birth rate in black women is twice that of whites.20 However, our results agree with Medearis et

556

Weeks

Twins (N = 43) 3 & 4)

RDS = respiratow Data are piesenteh

29-34

Preterrn

Twins

NEC = necrotizing

enterocolitis;

NS = not significant.

al2 that maternal race does not appear to increase the rate of preterm delivery in twin gestations. This suggests that the mechanisms leading to preterm twin births may be different than those leading to singleton preterm births. These mechanisms in twins may override the factors leading to the black-white disparity in singleton preterm delivery. Cerebral palsy, mental retardation, blindness, and deafness are all well-known, long-term sequelae of premature birth. Cerebral palsy occurs more frequently in twins than singletons.*’ In our study, twins with a birth weight between 500-1000 g had a 25% incidence of major handicaps, a rate similar to that seen in 500-1000 g singletons. Scheller and Nelson2’ have shown that the higher rate of cerebral palsy in twins may be due to preterm delivery. Petterson et a123 reviewed cerebral palsy in twins in Western Australia. Although the rate of cerebral palsy in twins was seven times that of singletons, the rate of cerebral palsy was similar in VLBW twins and singletons. Grether et a124 studied the prevalence of cerebral palsy in northern California and reported that the risk for twins was 12 times that of singletons, but found an equivalent risk in VLBW twins and singletons. Preterm delivery occurs in approximately half of twin pregnancies with spontaneous preterm labor as the major contributing factor. The high rate of preterm births in twins leads to disproportionately high morbidity and mortality. For the most part, when preterm twins are compared with preterm singletons and corrected for their gestational ages, the morbidity and mortality rates are similar. Continuing study of the causes of preterm delivery, particularly spontaneous preterm labor in twins, is necessary if twin morbidity and mortality is to be reduced.

References 1. Ho SK, Wu LYK. Perinatal factors and neonatal morbidity in twin pregnancies. Am J Obstet Gynecol 1975;122:979-87. 2. Medearis AL, Jonas HS, Stockbauer JW, Danke HR. Perinatal deaths in twin pregnancy. Am J Obstet Gynecol 1979;134:413-21. 3. Fowler MG, Kleinman JC, Kiely JL, Kessel SS. Double jeopardy:

Obstetrics

G Gynecology

Twin infant mortality in the United States, 1983 and 1984. Am J Obstet Gynecol 1991;165:15-22. 4. Tucker JM, Goldenberg RL, Davis RO, Copper RL, Winkler CL, Hauth JC. Etiologies of preterm birth in an indigent population: Is prevention a logical expectation? Obstet Gynecol 1991;77:343-9. 5. Main DM, Gabbe SG, Richardson D, Strong S. Can preterm deliveries be prevented? Am J Obstet Gynecol 1985;151:892-8. 6. McCarthy BJ, Sachs BP, Layde PM, Burton A, Terry JS, Rochat R. The epidemiology of neonatal death in twins. Am J Obstet Gynecol 1981;141:252-6. 7. Gardner MO, Amaya MA, Sakakini J. Effects of prenatal care on twin gestations. J Reprod Med 1990;35:519-21. JM, Newman RB, Hulsey TC, Bivins HA, Keenan A. 8. Ellings Reduction in very low birth weight deliveries and perinatal mortality in a specialized multidisciplinary twin clinic. Obstet Gynecol 1993;81:387-91. 9. Collaborative Group on I’reterm Birth Prevention. Multicenter randomized, controlled trial of a preterm birth prevention program. Am J Obstet Gynecol 1993;169:352-66. 10. Robertson PA, Sniderman SH, Laros RK, et al. Neonatal morbidity according to gestational age and birth weight from five tertiary care centers in the United States, 1983 through 1986. Am J Obstet Gynecol 1992;166:1629-41. 11. Ferguson WF. I’erinatal mortality in multiple gestations: A review of perinatal deaths from 1609 multiple gestations. Obstet Gynecol 1964;23:861-70. 12. Spellacy WN, Handler A, Ferre CD. A case-control study of 1253 twin pregnancies from a 1982-1987 perinatal data base. Obstet Gynecol 1990;75:168-71. 13. Quilligan E. Pathologic causes of preterm labor. In: Elder M, Hendricks C, eds. Preterm labor. London: Butterworth, 1981:68. 14. Neilson J, Verkuyl D, Bannerman C. Tape measurement of symphysis-fundal height in twin pregnancies. Br J Obstet Gynaecol 1988;95:1054-9. 15. Rouse DJ, Skopec GS, Zlatnik FJ. Fundal height as a predictor of preterm twin delivery. Obstet Gynecol 1993;81:211-4. 16. Romero R, Shamma F, Avila C, et al. Infection and labor. VI. Prevalence, microbiology, and clinical significance of intraamniotic infection in twin gestations with preterm labor. Am J Obstet Gynecol 1990;163:757-61.

VOL.

85, NO.

4,

APRIL

1995

17. Neilson JR, Verkuyl DAA, Crowther CA, Bannerman C. Preterm labor in twin pregnancies: Prediction by cervical assessment. Obstet Gynecol 1988;72:219-23. 18. Cunningham FG, MacDonald PC, Gant NF, Leveno KJ, Gilstrap LC. Williams Obstetrics. 19th edition. Norwalk, Connecticut: Appleton & Lange, 1993:891. 19. Bronsteen R, Goyert G, Bottoms S. Classification of twins and neonatal morbidity. Obstet Gynecol 1989;74:98-101. 20. McCormick MC. The contribution of low birth weight to infant mortality and childhood morbidity. N Engl J Med 1985;312:82-90. 21. Bejar R, Vigliocco G, Gramajo H, et al. Antenatal origin of neurologic damage in newborn infants. 11. Multiple gestations. Am J Obstet Gynecol 1990;162:1230-6. 22. Scheller JM, Nelson KB. Twinning and neurologic morbidity. Am J Dis Child 1992;146:1110-3. 23. Petterson B, Nelson KB, Watson L, Stanley F. Twins, triplets and cerebral palsy in births in Western Australia in the 1980s. BMJ 1993;307:1239-43. 24. Grether JK, Nelson KB, Cummins SK. Twinning and cerebral palsy: Experience in four northern California counties, births 1983 through 1985. Pediatrics 1993;92:854-8.

Address

reprint

requests

to:

Michael 0. Gardner, MD The University of Alabama at Birmingham Department of Obstetrics and Gynecology Division of Maternal-Fetal Medicine 618 South 20th Street, Room 458 OHB Birmingham, AL 35233-7333

Received May 6, 1994. Received in revised form December Accepted December 15, 1994. Copyright 0 1995 by The Gynecologists.

American

Gardner

7, 1994.

College

et al

of Obstetricians

and

Twins

557

Preterm