| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the Division of Epidemiology, School of Public Health, University of Medicine and Dentistry of New Jersey, Piscataway, New Jersey; Environmental and Community Medicine, and Section of Epidemiology and Biostatistics and Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology and Reproductive Sciences, University of Medicine and Dentistry of New Jersey–Robert Wood Johnson Medical School, Piscataway, New Jersey; and Division of Vital Statistics, National Center for Health Statistics, Centers for Disease Control and Prevention, Hyattsville, Maryland.
Address reprint requests to: Kitaw Demissie, MD, PhD, UMDNJ School of Public Health, Division of Epidemiology, 675 Hoes Lane, Piscataway, NJ 08854; E-mail: demisski{at}umdnj.edu.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: We used the United States (1995–1997) Matched Multiple Birth File (n = 297,155).
RESULTS: Among twin live births and stillborn fetuses, 29.9% had less than 5% birth weight discordance, 24.2% had 5–9%, 29.6% had 10–19%, 11.1% had 20–29%, 3.4% had 30–39%, and 1.8% had 40% or more. The stillborn fetus rate increased progressively with increasing birth weight discordance for smaller and larger twins of the same sex. Compared with the less than 5% birth weight discordance category, the adjusted odds ratios (OR) (95% confidence intervals [CIs]) for stillborn fetus associated with 5–9%, 10–19%, 20–29%, 30–39%, and 40% or more birth weight discordance, respectively, were 0.81 (95% CI 0.58, 1.11), 1.41 (95% CI 1.07, 1.84), 1.74 (95% CI 1.28, 2.35), 3.06 (95% CI 2.21, 4.24), and 4.29 (95% CI 3.05, 6.04) for smaller twins. The corresponding ORs (95% CIs) for larger twins were 0.78 (95% CI 0.57, 1.08), 1.26 (95% CI 0.96, 1.66), 1.77 (95% CI 1.27, 2.46), 3.38 (95% CI 2.33, 4.92), and 2.91 (95% CI 1.89, 4.47). Similar associations were observed among smaller but not larger twins of opposite sex. Among larger but not smaller twins of the same sex, increasing birth weight discordance was associated with overall neonatal deaths. This association was not apparent among smaller and larger twins of opposite sex. However, increasing birth weight discordance was associated with neonatal deaths related to congenital malformations among smaller and larger twins.
CONCLUSION: The results provide evidence that increased twin birth weight discordance was associated with increased risk of intrauterine death and malformation-related neonatal deaths.
Intrapair size differences in twin members have been a cause for concern in the management of twin pregnancies. Often, serial ultrasound examinations are performed to detect a significant intrauterine growth difference within a twin pair. The reported frequency of this phenomenon varies greatly with the definitions used,1–3 and a clinically important threshold that correlates well with perinatal morbidity and mortality has not been established. The lack of a standard definition for a clinically significant growth difference within a twin pair is partly due to conflicting data on the associated adverse perinatal outcomes.
Although some studies3–5 have observed an association between birth weight discordance and increased risk of perinatal mortality and disability in later life, others2,6,7 have not. The latter studies concluded that the actual birth weight of each twin within the pair, rather than the difference between the twins, is the most important determinant of outcome. The handful of studies reporting on this subject share several methodological problems, including failure to standardize for the actual birth weight of the infant. The significance of the extent of birth weight discordance may differ (ie, be modified) by gestational age and may be confounded by the birth weight of the smaller twin. However, previous studies have not estimated risk separately in the larger and smaller twins, nor controlled for confounding variables such as infant sex and birth order. The aim of the present study was to examine the clinical and public health relevance of intrapair birth weight discordance in a large population of twin births.
|
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The main outcomes of interest in our analyses were fetal (20 weeks or more), and neonatal (0–28 days) death. The International Classification of Diseases, Ninth Revision codes9 were used to identify the underlying cause of death recorded on death certificates of live-born infants. Underlying cause of death was not recorded for stillborn fetuses. The following causes of death were examined: congenital malformation (codes 2700–2759, 2770–2799, 2820–2829, 2840–2849, 2860–2889, 3300–3309, 3350–3359, 3430–3439, 3590–3599, 3940–4110, 4140–4179, 4240–4269, 5500–5539, 5600–5609, 5710–5719, 5720–5729, 7400–7599, 7771), neonatal septicemia (7611, 7627, 7718), respiratory distress syndrome (769), asphyxia related (7616, 7617, 7620–7622, 7624–7626, 7630–7639, 7660–7689, 7701, 7722, 7790, 7792), and intraventricular hemorrhage (7721). The cause-of-death groupings were developed based on those used by the International Collaborative Effort on Infant Mortality,9 as well as standard National Center for Health Statistics categories and other sources.
Birth weight discordance was the independent variable and was calculated as the intrapair difference in birth weight expressed as a percentage of the larger twin’s weight. Gestational age was derived from the dates of child’s birth and the mother’s last menstrual period (LMP). Records missing the date of the LMP are imputed when there is a valid month and year. Imputation procedures are described in detail elsewhere.10 Small for gestational age (SGA) was defined as birth weight below the tenth percentile for gestational age. This normogram was based on all twin births delivered in the United States during the period 1995–1997.
Factors considered as potential confounders included the gestational age and birth weight of each twin within the pair, infant sex, maternal age (younger than 20, 20–24, 25–29, 30–34, and 35 years or older), birth order within the set, maternal race (white, black, or other; Hispanicity was not used in race/ethnic classification), maternal education (less than 12, 12, 13–15, 16, and 17 years or more of schooling), marital status, trimester of prenatal care initiation, and live-birth parity (0, 1, 2, or 3 or more). Disorders complicating pregnancy included diabetes (juvenile onset, adult onset, and gestational), chronic hypertension, pregnancy-induced hypertension, eclampsia, placental abruption, placenta previa, and anemia. Data on characteristics of labor and delivery include delivery method (cesarean delivery, forceps, or vacuum), prolonged labor (lasting more than 20 hours), premature rupture of membranes, and breech presentation.
For this article, the unit of analysis was the individual twin and not the twin pair. We examined the distribution of potential confounders according to birth weight discordance. Logistic regression analysis was used to examine the relationship between fetal and neonatal mortality and birth weight discordance before and after accounting for important confounding factors. Birth weight discordance was first modeled as a continuous variable by incorporating a restricted cubic spline transformation.11 This approach avoids assumptions about the shape of the distribution of birth weight discordance and is a nonparametric flexible smoothing procedure. As an alternative approach we categorized birth weight discordance into six groups (less than 5%, 5–9%, 10–19%, 20–29%, 30–39%, and at least 40%) and performed similar analyses. Birth weight and gestational age were used as adjusting variables in the regression models and were treated as continuous variables using cubic spline transformations with five and four knots, respectively. Separate models were constructed for the smaller and larger twin of same and opposite sex. This is because of unavailability of information on chorionicity; opposite sex was used as a surrogate for dichorionic twins.
We also examined if the association between birth weight discordance and fetal and neonatal mortality differed (ie, modified) when the twin pairs were both appropriately grown for their gestational age, when one was SGA and the other appropriately grown, and when both were SGA.
The study was approved by the University of Medicine and Dentistry of New Jersey–Robert Wood Johnson Medical School’s institutional review board.
|
RESULTS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The distribution of selected variables by birth weight discordance categories is provided in Table 1
. Among live birth and stillborn fetus twin pairs, 29.9% had less than 5% birth weight discordance, 24.2% had 5–9%, 29.6% had 10–19%, 11.1% had 20–29%, 3.4% had 30–39%, and 1.8% had at least 40%. The proportion of teenage mothers was slightly higher among the most severe birth weight discordance category. Older mothers (35 or older) were also more likely to give birth to infants with higher birth weight differences. The proportion of nulliparous mothers increased with increasing birth weight discordance. The proportion of mothers with chronic hypertension, placental abruption, and cesarean delivery increased with increasing birth weight discordance, and this pattern generally also holds for pregnancy-induced hypertension and gestational age at the lower end of the distribution.
|
. In general, the stillborn fetus rate increased progressively with increasing birth weight discordance, although the increase was stronger among the smaller twins. Adjustment for important confounding variables including birth weight and gestational age did not alter the results meaningfully. Furthermore, a discordance of 10% or more conferred a significant increase in the stillborn fetus rate for smaller twins, whereas this threshold was 20% or more for larger twins. Similarly, the risk of neonatal death increased with increasing birth weight discordance for both smaller and larger twins. After adjustment for the effects of important confounding factors, there was no association between birth weight discordance and neonatal mortality among smaller twins. Among larger twins, however, the unadjusted association between neonatal mortality and birth weight discordance persisted even after adjustment was made for important confounding variables. Interaction (product) terms between birth weight discordance and birth weight as well as gestational age were not significant.
|
. After accounting for confounders, stillborn fetus risk increased with increasing birth weight discordance among smaller twins of opposite sex. A discordance of 20% or more was associated with a significant stillborn fetus rate for the smaller twin of opposite sex. No consistent association was observed between birth weight discordance and stillborn fetus among larger twins of opposite sex, nor with neonatal mortality among smaller and larger twins.
|
presents the risk of stillborn fetus and neonatal mortality when the twin pairs were categorized simultaneously by fetal growth and birth weight discordance. For this particular subanalysis, discordance was defined as 20% or more based on earlier results (Tables 2
and 3). SGA was associated with stillborn fetus for same and opposite sex concordant and discordant twins, except for opposite sex concordant twins, where one twin was not SGA and no relationship was observed. It was also evident that birth weight discordance was associated with stillborn fetus regardless of SGA status. However, twins who were SGA and also discordant were at the highest risk of dying in utero. This pattern was not apparent for neonatal mortality.
|
We also performed similar analyses after excluding fetuses and infants with known congenital anomalies. The results were fairly similar to the ones reported without excluding anomalies (data not shown).
Risk factors for stillborn fetus before 24 weeks’ gestation may differ from those occurring after 24 weeks’ gestation. To be sure that such confounding did not affect our results, we repeated the analyses after restricting the data to those with 24 or more weeks’ gestation. The results of these analyses were also similar to the overall study results.
|
DISCUSSION |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Birth weight discordance in twins may contribute to increased twin mortality, which is significantly higher than in singleton births.12 The existing scientific evidence on the relationship between birth weight discordance and perinatal mortality is conflicting. The study of Erkkola et al3 was among the earlier ones that showed a 6.5-fold risk of intrauterine death and a 2.5-fold risk of perinatal death associated with 25% or more discordance. Subsequently, other studies4,13–15 reported an association between birth weight discordance and fetal and infant death. In contrast to these findings, several studies challenged the clinical importance of birth weight discordance and argued that the actual birth weight of each twin within the pair, rather than the difference between the twins, is the most important determinant of outcome.2,6,7,16
The results of the present study were based on a large database (the largest of all reporting on this issue) and includes all twin births in the United States between 1995 and 1997 (adding generalizability). Our analyses estimated risks separately for the smaller and larger twin after stratifying the data for same and opposite sex twins. Our study also established a clinically important threshold at which clinicians should start to be concerned. These cutoff values were based on birth weight discordance levels that were associated with significant risk of stillborn fetus. The actual birth weight and gestational age of the twin pair were used as adjusting variables in the regression models when estimating the risk of fetal and neonatal mortality associated with birth weight discordance. A cubic spline transformation that allowed us to model birth weight and gestational age as continuous variables without loss of information and maximizing statistical efficiency was used for this purpose. Similarly, our analyses also extensively controlled for the confounding effects of sociodemographics, pregnancy complications, and other important pregnancy-related factors such as parity and birth order.
Yalcin et al14 speculated that the effect of birth weight discordance may differ by gestational age, though they did not provide data supporting their premise. Because of this, several investigators have restricted their analyses of discordance to either preterm15,17 or term gestations.7 We performed analyses after stratifying the data by gestational age categories (17–31, 32–36, and 37–41 weeks), and the results across gestational age categories were consistent and similar to the overall analyses’ results (data not shown).
Our finding that birth weight discordance was not associated with neonatal death from respiratory distress syndrome, asphyxia, and neonatal septicemia is not surprising given the adequate adjustment for birth weight and gestational age of each twin pair. If these variables had not been accounted for, an increase in neonatal death due to preterm gestations and low birth weight might have been observed with birth weight discordance. However, our aim was to estimate the risk of neonatal death associated with birth weight discordance independent of birth weight and gestational age of each twin pair.
Because of the increased incidence of preterm and SGA births with increasing birth weight discordance, most studies concluded that the excess stillborn fetus rate and neonatal death observed in discordant twin pairs were mediated through these variables. However, the birth weight of the smaller twin may increase the effect of birth weight discordance. In our analyses, the higher stillborn fetus rate among discordant twins persisted even after the effects of gestational age and birth weight were fully accounted for. The increased risk of neonatal death among discordant twins is likely to be due to deaths from congenital anomalies. These results suggest that the association between stillborn fetus and weight difference may be due to a common cause for both. The possibility of other mediating factors in addition to preterm birth and SGA for the higher stillborn fetus rate among discordant twins was further supported by our analyses of risk of fetal and neonatal mortality when the twin pairs were categorized simultaneously by fetal growth and birth weight discordance. Although SGA twins were at increased risk of stillborn fetus, the risk increase was even higher for those SGA twins who were at the same time discordant.
Our finding that birth weight discordance and stillborn fetus were associated in both same and opposite sex twins was supported by the study of Cheung et al,15 who studied placental histology among 118 twin pairs (29.7% monochorionic and 70.3% dichorionic). The authors found no difference among the different categories of birth weight discordance for the distribution of placental chorionicity. The slightly increased risk in neonatal mortality for the larger twin of same sex births may be related to the twin-twin transfusion syndrome and possibly to hydrops.
The results of the relationship between birth weight discordance and stillborn fetus should be interpreted in light of potential differences between fetal weight at death and fetal weight at delivery (or birth weight) as the result of fetal maceration. The use of birth weight as a surrogate for fetal weight may introduce errors in the association between fetal weight discordance and the risk of stillborn fetus. A study that estimated fetal weight during the fortnight before delivery and compared it with values of the actual birth weight18 did not find a significant difference between the mean fetal weight and birth weight. However, if the time elapsed since the death of one twin to the delivery of the co-twin is long, the birth weight of the dead twin may be unreliable. A well-designed and executed study of an expectant management of twin pregnancy with a single stillborn fetus19 showed that 69% of mothers went into spontaneous labor within 7 days after diagnosis of a stillborn fetus, 86% within 35 days, and 100% within 56 days. Another study limitation is the accuracy and completeness of gestational age based on the LMP date as reported on the birth certificate. This measure is subject to error, but has been demonstrated to be the most accurate available on a population-wide basis.20 Similarly, under-reporting of stillborn fetuses and pregnancy complications is likely to occur in birth certificate and stillborn fetus data.21 Potential bias may also have been introduced in the definition of discordancy among twins with twin-twin transfusion syndrome, as the larger twin may not be an appropriate standard for judging the extent of discordancy (eg, hydrops). Furthermore, in a monozygotic twin gestation, once there is death of one twin, the death of the second may no longer be an independent event, thus affecting the precision of our estimates.
In conclusion, the results provide evidence that increased twin birth weight discordance was associated with increased intrauterine death and malformation-related neonatal deaths.
|
Footnotes |
|---|
Received August 28, 2001. Received in revised form December 20, 2001. Accepted January 17, 2002.
|
REFERENCES |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Patterson RM, Wood RC. What is twin birthweight discordance? Am J Perinatol 1990;7:217–9.[Medline]
3. Erkkola R, Ala-Mello S, Piiroinen O, Kero P, Sillanp M. Growth discordancy in twin pregnancies: A risk factor not detected by measurement of biparietal diameter. Obstet Gynecol 1985;66:203–6.[Medline]
4. Hollier LM, McIntire DD, Leveno KJ. Outcome of twin pregnancies according to intrapair birthweight differences. Obstet Gynecol 1999;94:1006–10.
5. Rydhstroem H. The relationship of birthweight and birth-weight discordance to cerebral palsy or mental retardation later in life for twins weighing less than 2500 grams. Am J Obstet Gynecol 1995;173:680–6.[Medline]
6. Fraser D, Picard R, Picard E, Leiberman JR. Birthweight discordance, intrauterine growth retardation and perinatal outcomes in twins. J Reprod Med 1994;39:504–8.[Medline]
7. Hsieh T, Chang T, Chiu T, Hsu J, Chao A. Growth discordancy, birthweight and neonatal adverse events in third trimester twin gestations. Gynecol Obstet Invest 1994;38:36–40.[Medline]
8. National Center for Health Statistics. 1995–97 Matched Multiple Birth data set. NCHS CD-ROM Series 21, No. 12. Hyattsville, Maryland: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, 2000.
9. World Health Organization. Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death; based on the recommendations of the Ninth Revision Conference, 1975. Geneva: World Health Organization, 1975.
10. Taffel S, Johnson D, Heuser R. A method for imputing length of gestation on birth certificates. Vital Health Stat 2 1982;93:1–11.
11. Durrelman S, Simon R. Flexible regression models with cubic splines. Stat Med 1989;8:551–61.[Medline]
12. Kiely JL. The epidemiology of perinatal mortality in multiple births. Bull N Y Acad Med 1990;66:618–37.[Medline]
13. Rydhstrom H. Discordant birthweight and late fetal death in like-sexed and unlike-sexed twin pairs: A population-based study. Br J Obstet Gynecol 1994;101:765–9.[Medline]
14. Yalcin HR, Zorlu G, Lembet A, Ozden S, Gokmen O. The significance of birthweight difference in discordant twins: A level to standardize? Acta Obstet Gynecol Scand 1998; 77:28–31.[Medline]
15. Cheung VYT, Bocking AD, Dasilva OP. Obstetrics: Pre-term discordant twins: What birthweight difference is significant? Am J Obstet Gynecol 1995;172:955–9.[Medline]
16. Talbot GT, Goldstein RD, Nesbirt T, Johnson JL, Kay HH. Is size discordancy an indication for delivery of preterm twins? Am J Obstet Gynecol 1997;177:1050–4.[Medline]
17. Salafia CM, Minior VK, Pezzullo JC, Popek EJ, Rosenkrantz TS, Vintzileos AM. Intrauterine growth restriction in infants of less than thirty-two weeks’ gestation: Associated placental pathologic features. Am J Obstet Gynecol 1995;173:1049–57.[Medline]
18. Nzeh DA, Rimmer S, Moore WMO, Hunt L. Prediction of birthweight by fetal ultrasound biometry. Br J Radiol 1992;66:987–9.
19. Santema JG, Swaak AM, Wallenburg HCS. Expectant management of twin pregnancy with single fetal death. Br J Obstet Gynaecol 1995;102:26–30.[Medline]
20. Alexander GR, Tomkins ME, Petersen DJ, Hulsey TC, Mor J. Discordance between LMP-based and clinically estimated gestational age: Implications for research, programs and policy. Public Health Rep 1995;110:395–402.[Medline]
21. Gaudino JA Jr, Blackmore-Prince C, Yip R, Rochat RW. Quality assessment of fetal death records in Georgia: A method for improvement. Am J Public Health 1997;87: 1323–7.
This article has been cited by other articles:
|
|
 |
|
  R. Feldman and A. I. Eidelman Does a Triplet Birth Pose a Special Risk for Infant Development? Assessing Cognitive Development in Relation to Intrauterine Growth and Mother-Infant Interaction Across the First 2 Years Pediatrics, February 1, 2005; 115(2): 443 - 452. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Yinon, R. Mazkereth, N. Rosentzweig, A. Jarus-Hakak, E. Schiff, and M. J. Simchen Growth Restriction as a Determinant of Outcome in Preterm Discordant Twins Obstet. Gynecol., January 1, 2005; 105(1): 80 - 84. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. C. Amaru, M. C. Bush, R. L. Berkowitz, R. H. Lapinski, and S. Gaddipati Is Discordant Growth in Twins an Independent Risk Factor for Adverse Neonatal Outcome? Obstet. Gynecol., January 1, 2004; 103(1): 71 - 76. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. R. Jacobs, K. Demissie, N. J. Jain, and W. L. Kinzler Birth Weight Discordance and Adverse Fetal and Neonatal Outcomes Among Triplets in the United States Obstet. Gynecol., May 1, 2003; 101(5): 909 - 914. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Branum and K. C. Schoendorf The Effect of Birth Weight Discordance on Twin Neonatal Mortality Obstet. Gynecol., March 1, 2003; 101(3): 570 - 574. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. W. Wen, K. F. K. Fung, L. Oppenheimer, K. Demissie, Q. Yang, and M. Walker Occurrence and Predictors of Cesarean Delivery for the Second Twin After Vaginal Delivery of the First Twin Obstet. Gynecol., January 1, 1998; 103(3): 413 - 419. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
20%) and Small for Gestational Age (SGA) Status