| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the Perinatal Epidemiology Research Unit, Department of Obstetrics and Gynecology and Pediatrics, the Division of Neonatal-Perinatal Medicine, Department of Pediatrics, and the Division of Maternal/Fetal Medicine, Department of Obstetrics and Gynecology, Dalhousie University Faculty of Medicine, Halifax, Nova Scotia, Canada.
Address reprint requests to: K. S. Joseph, MD, PhD, Division of Neonatal Pediatrics, IWK Grace Health Centre, 5980 University Avenue, Halifax, Nova Scotia, B3H 4N1, Canada; E-mail: ks.joseph{at}np.iwkgrace.ns.ca
 |
ABSTRACT |
|---|
TOP
 ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: We studied all twin births among residents of the province of Nova Scotia, Canada, between 1988 and 1997. Rates of preterm birth, preterm labor induction, preterm cesarean, small-for-gestational age (SGA), respiratory distress syndrome (RDS), stillbirth, perinatal mortality, and infant mortality were compared between past and more recent years. Changes in perinatal mortality were examined using logistic regression to adjust for the effects of other determinants.
RESULTS: The study included 2516 twin births (73 stillbirths and 2443 live births). The rate of preterm birth increased from 42.3% in 1988–1992 to 48.2% of twin live births in 1993–1997 (14% increase, P = .04). Twin live births born after preterm labor induction increased from 3.5% in 1988–1989 to 8.6% in 1996–1997 (P for trend = .007). Of live births between 34 and 36 weeks’ gestation, the proportion born SGA decreased from 17.5% in 1988–1992 to 9.2% in 1993–1997 (P = .005). Over the same period, rates of prophylactic maternal steroid therapy increased substantially and rates of RDS declined. Perinatal mortality rates among pregnancies reaching 34 weeks decreased from 12.9 per 1000 total births in 1988–1992 to 4.2 per 1000 total births in 1993–1997 (P = .05).
CONCLUSION: Increases in preterm labor induction appear to be responsible for the recent increase in preterm birth among twins. These changes have been accompanied by decreases in perinatal morbidity and mortality among twin pregnancies that reach 34 weeks’ gestation.
Multiple births have become a focus of perinatal interest, especially because of a dramatic increase in frequency in recent years. In Canada, multiple births increased from 18.2 per 1000 total births in 1974 to 25 per 1000 total births in 1997.1,2 Similarly, in the United States twin live births increased from 18.9 per 1000 live births in 1980 to 26.9 per 1000 live births in 1997.3 The rate of triplet and higher-order multiple births in the United States increased over four-fold from 37 per 100,000 live births in 1980 to 174 per 100,000 live births in 1997.3 These changes are a consequence of increasing maternal age and greater use of ovarian stimulation and in vitro fertilization.4–7
The increase in the frequency of multiple births has been matched by an equally substantial, though less well understood, increased rate of preterm birth (before 37 weeks’ gestation) among multiple births.8 In Canada, the rate of preterm birth among multiple births increased from 33% in 1974 to 40% in 1981–1983, to 50% in 1992–1994 and to 53% in 1997.1,2,8 The mean gestational age of twins declined from 36.5 weeks in 1981–1983 to 35.8 weeks in 1992–1994, while the mean gestational age of triplet and higher-order multiple births decreased from 32.9 weeks to 32.2 weeks across the same period.8 These findings are similar to observations in other industrialized countries; in the United States, preterm birth rates among twins have increased from 41% in 1981 to 55% in 1997.9
Of particular interest in this context is the contrast between multiple and singleton births. Modest increases in preterm singleton birth rates have been associated with statistically significant declines in fetal death rates, suggesting that obstetric intervention (in concert with advances in neonatal care) is saving compromised fetuses through early preterm delivery.8 However, rates of preterm multiple births have increased substantially without an associated decline in fetal death rates.8 We carried out a study to ascertain the causes and consequences of the recent increase in preterm birth rates among twin births.
|
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Triplet and higher-order multiple births were not included in the study because their numbers were small (none to three triplet sets per year). The gestational age and birth weight distributions, rates of preterm labor induction, preterm cesarean, small for gestational age (SGA) and large for gestational age (LGA) (below the tenth percentile and above 90th percentile of standard for twins, respectively11), respiratory distress syndrome (RDS), and infant mortality of twin live births were contrasted between past and more recent years. Idiopathic RDS was diagnosed if grunting, retractions, and decreased air entry occurred before 3 hours of age, persisted beyond 6 hours, were associated with typical clinical and radiologic signs, and were not explained by any other disease. Changes in rates of moderate (requiring at least 35% oxygen or continuous positive airway pressure) and severe (requiring ventilation) RDS and in the rates of other forms of respiratory distress including transient respiratory distress (duration up to 6 hours), benign respiratory distress (duration more than 6 hours), and transient tachypnea of the newborn (respiratory rate above 80 per minute for more than 24 hours, having a benign course and never requiring more than 40% oxygen) were also examined. Stillbirth and perinatal mortality rates were compared across the same period. Because the primary focus of the study was on fetal and infant outcomes, we analyzed the data with fetuses or infants (rather than pregnancies) as the unit of observation. As per convention, preterm birth rates, low birth weight rates, and infant mortality rates were calculated among live births, whereas total births served as the denominator for stillbirth and perinatal mortality rates.
For the purpose of examining trends over time, the 10-year study period was arbitrarily divided into five 2-year periods (1988–1989, 1990–1991, etc) for the more common outcomes and into two 5-year periods (1988–1992 and 1993–1997) for the less common outcomes. Information on gestational age in the database included gestational age calculated from the last normal menstrual period and a clinical assessment that was based on a physician’s physical examination of the infant at birth. Gestational age for this study was based on that calculated from the last normal menstrual period, except for instances when this information was missing or differed from the clinical estimate of gestational age by more than 2 weeks. In the latter instance, the clinical estimate of gestational age was used.
Inference was guided by computing relative risks, excess relative risks, 95% confidence intervals (CI), and two-tailed P values. Comparisons of the changes in maternal characteristics (such as age) over time were made using 
2 tests. Because fetal–infant outcomes in a twin pregnancy tend to be correlated (leading to a violation of the statistical assumption of independence between observations), all such statistical comparisons (between two proportions or a trend over the five 2-year periods) were made using the procedure of generalized estimating equations.12 This procedure adjusts the variance estimates for the correlation in outcomes between the births of a single twin pregnancy. The statistical significance of temporal changes in mean birth weight for gestational age was also assessed after correcting the variance using the generalized estimating equations procedure with appropriate specification. SAS software (proc genmod) was used for this analysis (SAS 6.12, SAS Institute, Cary, NC).
Temporal changes in stillbirth and perinatal mortality were examined using logistic regression to control for relevant, potentially confounding influences to ascertain if changes over time could be explained by other factors. Variables considered as potential confounders in the logistic regression models included maternal age, parity, prepregnancy weight, weight gain during pregnancy, maternal smoking, method of delivery [spontaneous vaginal (cephalic), breech, forceps or vacuum, or cesarean], and birth order. Only those variables found to be associated with perinatal mortality (P < .10) in univariate analyses were included in the final logistic model.
|
RESULTS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Prepregnancy weight increased and the proportion of women who smoked decreased between the two periods.
|
, P value for linear trend = .09). Most of the increase in preterm birth rates occurred as a result of increases in twin live births at 34–36 weeks’ gestation, which increased from 27.0% of twin live births in 1988–1992 to 30.9% in 1993–1997. The gestational age distribution of twin live births in the two periods showed a "shift to the left" in recent years, especially at later gestational ages (Figure 1). Rates of low birth weight fluctuated across the study period (Table 2, P for trend = .43). The rate of low birth weight was 47.1% in 1988–1992 and 47.7% in 1993–1997 (P = .82).
|
|
, P for trend = .007). Among twin live births at 34–36 weeks’ gestation, labor was induced in 12.3% in 1988–1992 and in 18.5% 1993–1997 (51% increase, P = .13). This change in labor induction among live births between 34 and 36 weeks resulted from increases in the frequency of indications such as hypertension (from 2.9% in 1988–1992 to 5.0% in 1993–1997), premature rupture of membranes (from 1.8% to 3.6%), oligohydramnios (from 0.0% to 1.4%), abnormal biophysical profile (from 0.6% to 2.0%), and miscellaneous indications (from 2.9% to 3.7%). Labor induction for fetal growth restriction decreased from 4.1% in 1988–1992 to 2.8% in 1993–1997.
|
, P for trend = .55). The overall cesarean rate also did not change significantly and was 42.6% in 1988–1992 and 44.7% in 1993–1997 (P = .46).
Between 1988–1992 and 1993–1997, birth weight for gestational age increased among twin live births at most gestational ages from 33 to 37 weeks, although not all these increases were statistically significant (Table 4). There were no statistically significant increases in birth weight for gestational age at 38 weeks of gestational age and beyond. The proportion of twin live births that were SGA decreased nonsignificantly between 1988–1989 and 1996–1997 (Table 5, P for trend = .11). Small for gestational age twin live births at 34–36 weeks’ gestation decreased from 17.5% in 1988–1992 to 9.2% in 1993–1997 (P = .005). The proportion of twin live births that were LGA did not change significantly (Table 5, P for trend = .70).
|
|
, P for trend = .81). The proportion of twin live births with mild, moderate, or severe idiopathic RDS decreased nonsignificantly between 1988–1989 and 1996–1997 (Table 5, P for trend = .13). Moderate and severe forms of RDS also decreased nonsignificantly from 10.5% in 1988–1989 to 6.5% in 1996–1997 (P for trend = .10). Other forms of respiratory distress such as transient respiratory distress, transient tachypnea of the newborn, and benign respiratory distress did not show a consistent trend, decreasing from 6.3% in 1988–1989 to 4.7% in 1994–1995, then increasing to 9.4% in 1996–1997 (P for trend = .15). Among twin live births at 34–36 weeks, the rate of mild, moderate, or severe RDS decreased from 17.3% in 1988–1992 to 9.8% in 1993–1997 (P = .02); the rate of moderate or severe RDS decreased from 9.9% in 1988–1992 to 4.8% in 1993–1997 (P = .02); and the rate of transient respiratory distress, transient tachypnea of the newborn, and benign respiratory distress did not change significantly (9.9% in 1988–1992 and 11.2% in 1993–1997, P = .61).
During the same period, rates of prophylactic maternal steroid therapy (first dose of dexamethasone or beta-methasone given at least 48 hours before delivery) increased dramatically (Table 5, P for trend <.001). The proportion of twin live births delivered between 34 and 36 weeks’ gestation, whose mothers received the first dose of prophylactic steroids at least 48 hours before delivery, increased from 9.7% in 1988–1992 to 21.8% in 1993–1997 (P < .001).
The stillbirth rate among twin pregnancies was 33.9 per 1000 total births in 1988–1992 and 23.6 per 1000 total births in 1993–1997 (P = .21). Among pregnancies reaching 34 weeks’ gestation, stillbirth rates decreased from 8.3 per 1000 in 1988–1992 to 3.1 per 1000 total births in 1993–1997, although this difference was also not statistically significant (P = .15). Perinatal mortality rates registered a small nonsignificant decrease from 60.2 per 1000 total births to 50.5 per 1000 total births (P = .41). Among pregnancies reaching 34 weeks’ gestation, perinatal mortality rates decreased from 12.9 per 1000 total births to 4.2 per 1000 total births (P = .05). The gestational age-dependent nature of the declines in perinatal mortality rates is illustrated in Figure 2. Examination of the primary causes of perinatal death among pregnancies of at least 34 weeks’ gestation showed that in recent years there were fewer deaths resulting from cord prolapse and related problems (three in 1988–1992 compared with none in 1993–1997), unexplained stillbirths (four and two, respectively), placental abruption (one and none, respectively), congenital anomalies (two and one, respectively), hyaline membrane disease (one and none, respectively), and twin-to-twin transfusion (one and none, respectively). There was one death resulting from fetal growth restriction in each period. Infant mortality rates were stable at 35.1 per 1000 live births in 1988–1992 and 33.6 per 1000 live births in 1993–1997. Among live births at or after 34 weeks’ gestation, infant mortality rates were 7.5 per 1000 in 1988–1992 and 3.1 per 1000 live births in 1993–1997 (P = .20).
|
|
DISCUSSION |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The increased rate of obstetric intervention appears to have led to a substantial change in birth weight for gestational age among twin live births at some gestational ages. We speculate that twin births that would have delivered spontaneously after a period of faltering growth are being increasingly delivered at preterm gestation before growth restriction has occurred. This practice has led to declines in the proportion born SGA at 34–36 weeks’ gestation. This finding appears to conflict with findings from other recent twin studies, however.9 Potential explanations include study differences in populations and in sources of gestational age information. Our study also showed that increases in the use of prophylactic maternal steroid therapy have been accompanied by simultaneous declines in rates of RDS. Several clinical trials have demonstrated the protective efficacy of such prophylactic therapy.16
The finding that overall rates of stillbirth did not decline significantly over the 10-year study period is similar to what has been observed in some studies8,17,18 but not others.19 Our finding regarding unchanging rates of overall stillbirth is probably related to the higher rate of second trimester fetal loss among twin pregnancies (relative to singleton pregnancies). Compared with singleton pregnancies, twin pregnancies have higher rates of congenital anomalies, chromosomal abnormalities, and growth restriction, which are risk factors for fetal death.20–23 Other mechanisms responsible for higher rates of early fetal mortality among twins include cervical incompetence24,25 and transplacental communications in twin circulation (eg, twin-twin transfusion syndrome26).
Although our study showed that recent increases in preterm twin births have been accompanied by improvements in mortality among twins reaching 34 weeks of gestational age, this assessment is indirect. We used logistic regression to study the effects of other factors (such as maternal age), which could potentially explain changes in mortality between earlier and later periods. However, adjustment for risk factors significantly associated with perinatal death in pregnancies reaching 34 weeks’ gestation could not explain the temporal decrease in perinatal mortality. This suggests that increased labor induction and higher rates of iatrogenic preterm birth are probably responsible for the observed decrease in perinatal mortality among pregnancies reaching 34 weeks’ gestation.
We were not able to identify the effect of temporal changes in health services such as improvements in access to obstetric and neonatal specialist care in tertiary/ regional centers and increased use of other obstetric interventions. Other limitations of our study include an inability to assess the contribution of changing modalities of gestational age ascertainment to increases in pre-term birth among twins. Increasing use of ultrasonography in determining gestational age is known to have resulted in some increase in preterm birth.27–29 Although the assessment of gestational age in this study was based on menstrual dates or the physical examination of the infant at birth, routine ultrasound information on gestational age may have had some influence on gestational age assessment in recent years (especially on clinical estimates of gestation). This phenomenon is unlikely to account for more than a small fraction of the increase in preterm birth rates in the study, however.9
Although improvements in obstetric and neonatal care have made early delivery a safer proposition than in previous years, mild and moderate preterm birth infants are responsible for an important fraction of infant deaths.14 The need for early delivery in twin pregnancy, given evidence suggestive of compromised fetal well being, must always be balanced against the higher risk of morbidity and mortality that accompanies birth at pre-term gestation.
|
Footnotes |
|---|
We are grateful to the Reproductive Care Program of Nova Scotia for providing access to the data.
Received October 19, 2000. Received in revised form February 13, 2001. Accepted March 14, 2001.
|
REFERENCES |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Canadian Perinatal Health Report, 2000. Health Canada. Ottawa: Minister of Public Works and Government Services Canada, 2000.
3. Ventura SJ, Martin JA, Curtin SC, Mathews TJ. Births: Final data for 1997. National vital statistics reports; Vol. 47, No. 18. Hyattsville, Maryland: National Center for Health Statistics, 1999.
4. Luke B. The changing pattern of multiple births in the United States: Maternal and infant characteristics, 1973 and 1990. Obstet Gynecol 1994;84:101–6.
5. Kiely JL. What is the population-based risk of preterm birth among twins and other multiples? Clin Obstet Gynecol 1998;41:3–11.[Medline]
6. Keith L, Oleszczuk JJ. Iatrogenic multiple birth, multiple pregnancy and assisted reproductive technologies. Int J Gynaecol Obstet 1999;64:11–25.[Medline]
7. Bergh T, Ericson A, Hillensjö T, Nygren K-G, Wennerholm U-B. Deliveries and children born after in-vitro fertilization in Sweden 1982–95: A retrospective cohort study. Lancet 1999;354:1579–85.[Medline]
8. Joseph KS, Kramer MS, Marcoux S, Ohlsson A, Wen SW, Allen A, et al. Determinants of preterm birth rates in Canada from 1981 through 1983 and from 1992 through 1994. N Engl J Med 1998;339:1434–9.
9. Kogan MD, Alexander GR, Kotelchuck M, Macdorman MF, Buekens P, Martin JA, et al. Trends in twin birth outcomes and prenatal care utilization in the United States, 1981–1997. JAMA 2000;284:335–41.
10. Fair M, Cyr M, Allen AC, Wen SW, Guyon G, Macdonald RC. Validation study for a record linkage of births and infant deaths in Canada. Statistics Canada, catalogue No. 84F0013XIE. Ottawa, 1999.
11. Alexander GR, Kogan M, Martin J, Papiernik E. What are the fetal growth patterns of singletons, twins and triplets in the United States? Clin Obstet Gynecol 1998;41:115–25.
12. Zeger SL, Liang K-Y. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986 42:121–30.[Medline]
13. Luke B. Reducing fetal deaths in multiple births: Optimal birthweights and gestational ages for infants of twin and triplet births. Acta Genet Med Gemellol 1996;45:333–48.[Medline]
14. Kramer MS, Demissie K, Yang H, Platt RW, Sauvé R, Liston R, for the Fetal and Infant Health study Group of the Canadian Perinatal Surveillance System. The contribution of mild and moderate preterm birth to infant mortality. JAMA 2000;284:843–9.
15. Joseph KS, Kramer MS, Allen AC, Cyr M, Fair M, Ohlsson A, et al. Gestational age- and birthweight-specific declines in infant mortality in Canada, 1985–94. Fetal and Infant Health Study Group of the Canadian Perinatal Surveillance System. Paediatr Perinat Epidemiol 2000;14: 332–9.[Medline]
16. Crowley P. Prophylactic corticosteroids for preterm birth (Cochrane Review). In: The Cochrane Library, Issue 2.Oxford, UK: Update Software, 2000.
17. Moreault L, Marcoux S, Fabia J, Tennina S. Time trends in characteristics and outcomes of twin pregnancies. Acta Genet Med Gemellol 1991;40:181–92.[Medline]
18. Hartikainen-Sorri A-L, Rantakallio P, Sipila P. Changes in prognosis of twin births over 20 years. Ann Med 1990;22: 131–5.[Medline]
19. Glinianaia SV, Pharoah P, Sturgiss SN. Comparative trends in cause-specific fetal and neonatal mortality in twin and singleton births in the north of England, 1982–1994.Br J Obstet Gynecol 2000;107:452–60.
20. Kallen B. Congenital malformations in twins: A population study. Acta Genet Med Gemellol 1986;35:167–78.[Medline]
21. Rodis JF, Egan JF, Craffey A, Ciarleglio L, Greenstein RM, Scorza WE. Calculated risk of chromosomal abnormalities in twin gestations. Obstet Gynecol 1990;76: 1037–41.
22. Bryan EM. The intrauterine hazards of twins. Arch Dis Child 1986;61:1044–5.[Medline]
23. Grobman WA, Peaceman AM. What are the rates and mechanisms of first and second trimester pregnancy loss in twins? Clin Obstet Gynecol 1998;41:37–45.
24. Guzman ER, Houlihan C, Vintzileos A. Sonography and transfundal pressure in the evaluation of the cervix during pregnancy. Obstet Gynecol Surv 1995;50:395–403.[Medline]
25. Souka AP, Heath V, Flint S, Sevatopoulou I, Nicolaides KH. Cervical length at 23 weeks in twins in predicting spontaneous preterm delivery. Obstet Gynecol 1999;94: 450–4.
26. Blickstein I. The twin-twin transfusion syndrome. Obstet Gynecol 1990;76:714–21.
27. Kramer MS, McLean FH, Boyd ME, Usher RH. The validity of gestational age estimation by menstrual dating in term, preterm and postterm gestations. JAMA 1988; 260:3306–8.[Abstract]
28. Goldenberg RL, Davis RO, Cutter GR, Hoffman HJ, Brumfield CG, Foster JM. Prematurity, postdates, and growth retardation: The influence of use of ultrasonography on reported gestational age. Am J Obstet Gynecol 1989;160:462–70.[Medline]
29. Henriksen TB, Wilcox AJ, Hedegaard M, Secher NJ. Bias in studies of preterm and postterm delivery due to ultrasound assessment of gestational age. Epidemiol 1995;6: 533–7.[Medline]
This article has been cited by other articles:
|
|
 |
|
  W. A. Engle, K. M. Tomashek, C. Wallman, and and the Committee on Fetus and Newborn "Late-Preterm" Infants: A Population at Risk Pediatrics, December 1, 2007; 120(6): 1390 - 1401. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. V. Ananth, K. S. Joseph, Y. Oyelese, K. Demissie, and A. M. Vintzileos Trends in Preterm Birth and Perinatal Mortality Among Singletons: United States, 1989 Through 2000 Obstet. Gynecol., May 1, 2005; 105(5): 1084 - 1091. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
