| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
*From the Departments of Maternal and Child Health and Epidemiology, University of Alabama at Birmingham, Birmingham, Alabama.
Address reprint requests to: Hamisu M. Salihu, MD, PhD, Department of Maternal and Child Health, University of Alabama at Birmingham, 1665 University Boulevard, Room 320, Birmingham, AL 35294; e-mail: hsalihu{at}uab.edu.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: This was a retrospective cohort study on live births of extremely preterm twins delivered to teenaged mothers (aged 15–19 years) in the United States within the period 1995 through 1998. Overall neonatal and early and late neonatal mortality in this category was compared with that of a similar group of twins born to young adult mothers (aged 20–29 years). We used the generalized estimating equation framework in computing relative risks after adjusting for intracluster correlations.
RESULTS: Analysis involved 2,290 extremely preterm liveborn twins of teenaged mothers and 8,709 born to young adult mothers. Overall, neonatal mortality was 29% higher among the extremely preterm twins born to teenaged mothers (adjusted odds ratio [OR] 1.29; 95% confidence interval [CI] 1.04%, 1.59%). The disparity in neonatal survival was chiefly in the early neonatal period (adjusted OR 1.34; 95% CI 1.07%, 1.67%), while late neonatal mortality was comparable (adjusted OR 0.91; 95% CI 0.58%, 1.42%). In addition, twins of teenaged mothers had significantly higher level of mortality, except for the birth weight category of 1,000–1,499 g.
CONCLUSION: Low maternal age was found to be associated with elevated risk of neonatal death among extremely preterm twins. The preponderance of deaths among extremely preterm twins of teenaged mothers in the early neonatal period appeared to be responsible for the disparity in survival. This information may be useful for targeted interventions aimed at enhancing survival of extremely preterm twins born to teenagers, as well as for instituting optimal management options in the clinical setting.
LEVEL OF EVIDENCE: II-2
|
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The NCHS undertook a 3-stage matching algorithm to match records for deliveries involving multiple gestations. The first stage involved the building of an algorithm consisting of uniquely identifying variables from live birth and fetal death records, followed by identification of records with identical values for these variables. If the number of records with identical information equaled the reported plurality (eg, 2 records reported as twins) these records were considered members of the same multiple and assigned a unique set identification number. In those instances in which matching failed using this procedure, visual review was conducted and matching done as appropriate. All other records were considered unmatched and included in subsequent matching procedures that involved the use of additional variables, a composite of algorithmic combinations and manual identification. Perfect matching was achieved for 98.8% of the records. The whole process has been adequately validated and found to be very accurate.13
In this study, the sociodemographic group of interest was teenaged gravidas, defined as mothers aged 15–19 years. The control group consisted of young adult gravidas, defined as mothers within the age bracket of 20–29 years. Previous studies have justified this categorization.15,16 We further selected extremely preterm live births (20–28 weeks of gestation) with birth weight less than 2,000 g. We compared the following sociodemographic characteristics between the 2 groups of mothers with twin gestations: maternal race or ethnicity, marital status, educational level, prenatal smoking, parity, and adequacy of prenatal care. Adequacy of prenatal care was determined by using the revised graduated index algorithm17,18 and was categorized into adequate and inadequate prenatal care use. This index of prenatal care has been found to be more accurate than several others, especially in describing the level of prenatal care use among groups that are high-risk and therefore exposed to intense care (eg, multiple pregnancies).19 The revised graduated index algorithm assesses the adequacy of care based on 3 variables (trimester prenatal care began, number of visits, and the gestational age of the infant at birth). In this study, inadequate prenatal care use refers to patients who had either missing prenatal care information or prenatal care but the level was considered suboptimal or to mothers who had no prenatal care at all. The accuracy of all these aforementioned sociodemographic variables on the birth certificate has been validated in previous studies.20,21 The information coded in these variables was reported accurately.20,21
We also compared the 2 maternal age cohorts with respect to the occurrence of selected maternal complications of pregnancy (anemia, diabetes, genital herpes, pregnancy-associated hypertension, eclampsia, and abruptio placentae) as well as mode of delivery (vaginal, cesarean, and vacuum and forceps extraction).
For this study, the main birth outcomes of interest were neonatal mortality (death of the newborn within the first 28 days of life), which we further subdivided into early (from day 1 to day 7 postdelivery) and late (from day 8 to day 28 postdelivery) neonatal mortality. We computed gestational age and birth weight–specific mortality rates stratified by maternal age category. For gestational age–specific mortality rates, unit intervals were used except for gestational ages 24 weeks or less, which were grouped together, whereas birth weight was divided into 500-g strata.
We estimated regression parameters by taking into account the presence of intracluster correlation using the methodology of generalized estimating equations.22 The generalized estimating equations method considers 2 sources of variance in computing effect estimates. One is the intracluster variation (variation between individual twins) and the other is intercluster variation (variation between mothers). The generalized estimating equations model applied was based on the following assumptions:
We assessed goodness-of-fit of models using the –2 log likelihood ratio test,24 and we estimated the significance of main effects using the Wald test.24 The results of the tests for goodness-of-fit demonstrated that the retained model fitted the data adequately. The generalized model procedure in Statistical Analysis Systems (SAS Institute, Inc, Cary, NC) was used to conduct the generalized estimating equations analysis.
We calculated adjusted excess overall neonatal and early and late neonatal mortality. This represents the mortality burden that could have been averted among twins of teenaged mothers if pregnancy had been delayed until young adult age (20–29 years). This was calculated by using the following formula25:
|
| (1) |
where RR denotes adjusted relative risk, which in our study was approximated by the odds ratios.
All tests of hypotheses were 2-tailed with a type 1 error rate fixed at 5%. This study was approved by the Institutional Review Board of the University of Alabama at Birmingham.
|
RESULTS |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Table 1 compares the 2 maternal age categories in relation to selected maternal sociodemographic characteristics. Teenaged mothers with extremely preterm twin live births were more likely to be of black race, uneducated, unmarried, and nulliparous, with a lower level of adequate prenatal care compared with their adult counterparts.
|
Table 2 compares the distribution of relatively common medical complications experienced during pregnancy between teenaged and young adult mothers. The rates of occurrence of these complications were similar for the 2 maternal age cohorts except for diabetes and chronic hypertension. A higher proportion of young adult mothers had diabetes mellitus compared with teenaged mothers (1.5% versus 0.6%, P = .002).
|
Slightly more than one half of twins in both maternal age categories were delivered by the vaginal route (Table 3). There were no differences between the 2 groups in terms of route of delivery or complications of labor, except for repeated cesarean delivery. Extremely preterm twins of young adult mothers with previous history of abdominal delivery were about 4 times as likely as those of teenaged gravidas with the same history to be delivered by cesarean.
|
A total of 4,412 neonatal deaths was recorded among extremely preterm twins within the study period, yielding an overall crude neonatal mortality rate of 401.1 per 1,000. Of this count, 3,842 deaths (87.1%) occurred in the early neonatal period, and the remaining were late neonatal deaths (12.9%). Figure 1 compares the 3 crude mortality indices by maternal age subgroups.
|
Twins of the 2 maternal age groups did not differ in birth weight (mean birth weight ± standard error for the teenaged group was 747.2 ± 6.8 g compared with 755.3 ± 3.3 g for twins of young adult mothers). Birth weight–specific neonatal mortality rates by maternal age category are presented in Table 4. In both maternal age groups, neonatal mortality rates were extremely high at the lowest birth weights but then plummeted markedly with increasing birth weight. When twins of the 2 maternal age groups were compared relative to birth weight–specific neonatal mortality rates, twins of teenaged mothers had significantly higher level of mortality, except for the birth weight category of 1,000–1,499 g, for which the neonatal mortality rates were comparable.
|
Extremely preterm twins of teenaged mothers had a significantly lower mean gestational age at delivery than their counterparts born to young adult mothers (24.4 ± 0.04 weeks versus 24.6 ± 0.02 weeks). For both maternal age categories, neonatal and early neonatal mortality rates were extremely high at lower gestational ages, with a steep decline until 25 weeks of gestation, at which point the rate of decrease diminished sharply. Between 25 weeks and 27 weeks of gestation, there was an apparent gap in both neonatal (12%) and early neonatal (15%) mortality to the disadvantage of twins of teenaged mothers, although the discrepancy was only marginally significant (P = .08 and 0.06 for neonatal and early neonatal death, respectively; Figure 2). For late neonatal mortality, the gestational age–specific trajectory was entirely different from the other 2 mortality parameters. From an initial slight rise, mortality rates decreased progressively after 25 gestational weeks. Apart from a modest initial gap before the 25th week, the 2 mortality trajectories were indistinguishable (Figure 2).
|
Table 5 shows crude and adjusted estimates as well as levels of excess mortality among extremely preterm neonates of teenaged mothers by using the generalized estimating equation framework. It is noteworthy that the crude odds ratios obtained by using the generalized estimating equations differed from the crude mortality ratios mentioned above because the generalized estimating equations framework factored out the effects of intracluster correlations within twin pairs. For both neonatal and early neonatal mortality indices, twins of teenagers had significantly lower survival likelihood. Late neonatal mortality estimates were, however, comparable.
|
|
DISCUSSION |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Another important finding in this study is that the survival disadvantage experienced by extremely preterm twins in the neonatal period was in the early neonatal period only, with late neonatal survival probabilities being comparable to those of young adult mothers. For programs targeted at reducing excess mortality among this subpopulation of twins born to teenagers, it might be more cost-effective to focus resources in the first 7 days of life, a time when disparity is most pronounced. This is also the period during which extra attention may be needed in the clinical setting to optimize survival of these infants born to teenagers.
A limitation in this study is the absence of information on the use of assisted reproductive technology (ART), a procedure that has been linked to the epidemic of multiples in the United States.30,31 However, twin pregnancies among teenagers are spontaneous events,32,33 and to expect ART usage among adolescents is unrealistic.32,33 It has also been documented that ART-related pregnancies have generally poor outcomes compared with spontaneous conceptions.34 Given this premise, one would logically expect twins of young adult mothers to experience higher mortality than those of adolescent mothers, because the former are more likely to be associated with ART than the latter. This implies that our results presumably underestimated the magnitude of the disparity because of our inability to take into account the influence of ART. However, currently, only 4.8% of twins born to young adult mothers are iatrogenic,34 so that it is highly unlikely that this small percentage of infants would have influenced the results of this study. Nevertheless, assuming that they did, our reported findings would represent conservative estimates.
Perhaps one would question the accuracy of gestational age as documented on the birth certificate, because more than 95% of these estimates are based on the last menstrual period as remembered by the mother, information that depends on the mother's recall ability. It is, therefore, reasonable to expect some degree of underestimation or overestimation of gestational age based on the last menstrual period. It is also reasonable to question the accuracy of gestational age estimation as low as 20 weeks. However, reliability tests have demonstrated that gestational age as documented on birth certificates is accurately reported when compared with medical records.21 In addition, estimates of birth outcomes with gestational age information on the birth certificates have been found to be valid.19 With regard to the reliability of results based on low gestational age ranges, recent published data on multiples from vital records assembled by the NCHS have reported valid and reliable results based on analysis of outcomes on fetuses of multiple gestations with gestational age as low as 20 weeks.4
This analysis did not attempt to delineate the causes of death among very preterm twin neonates. Factors such as fetal anomalies, perinatal infections, and lack of supportive family environment after discharge could have contributed to early death. Although information such as family support is not routinely collected on the birth certificate, causes of death as documented on birth certificates are generally not reliable.37 For instance, fetal anomalies are reported for only 14% of cases38 (Kirby RS. The quality of data reported on birth certificates [letter]. Am J Public Health 1997;87:301.), so that analysis based on this very low level of reporting could be misleading. To avert erroneous conclusions, we avoided analyzing causes of death in this study. Another limitation is the absence of information that could be used to match levels of neonatal intensive care units to mortality outcomes in this study.
Because this study was population-based, influences attributable to selection biases (eg, highly selected or unique populations) are unlikely to distort our results. The large sample size allowed for adequate adjustment for the effects of known as well as potential confounders, especially those related to maternal characteristics. The use of the generalized estimating equations framework to model the contributions of intracluster and intercluster sources of variation or heterogeneity averted spurious associations being made.
It is noticeable that with adjustment, the odds ratios went up for neonatal mortality (from 20% to 29%) and for early mortality (from 15% to 34%). This is probably contrary to expectations, because one would normally expect that after controlling for confounding factors that were disadvantageous to adolescent mothers, the risk of mortality should decline. A possible explanation is that certain biologic factors might have been masked by the presence of sociodemographic characteristics. However, the moment the influences of these sociodemographic factors are removed (eg, through adjustment), the effects of the biologic characteristics become more manifest. However, this is mere speculation because we do not have any evidence to support this hypothesis. A more plausible reason is the use of the generalized estimating equations method of estimation, which takes into account intracluster correlations. Apart from the effects of the confounding variables, the magnitude of captured intracluster correlations also contributes to the nature of the final adjusted estimates. However, detailed statistical discussion as to how this occurs and the iterations involved is far beyond the intended purpose and scope of this paper.
In summary, we observed lower neonatal survival among extremely preterm twins born to teenaged mothers. The survival disadvantage was noted only in the early neonatal period. This information could be considered by care providers for intervention purposes among this group of infants born to teenagers.
|
Footnotes |
|---|
Received January 13, 2004. Received in revised form March 4, 2004. Accepted March 11, 2004.
10.1097/01.AOG.0000126724.91988.fa
|
REFERENCES |
|---|
|
TOP
ABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Phipps MG, Sowers M. Defining early adolescent childbearing. Am J Public Health 2002;92:125–8.
3. Rees JM, Lederman SA, Kiely JL. Birth weight associated with lowest neonatal mortality: Infants of adolescent and adult mothers. Pediatrics 1996;98:1161–6.
4. Fraser AM, Brockert JE, Ward RH. Association of young maternal age with adverse reproductive outcomes. N Engl J Med 1995;332:1113–7.
5. Olausson PO, Cnattingius S, Haglund B. Teenage pregnancies and risk of late fetal death and infant mortality. Br J Obstet Gynaecol 1999;106:116–21.[Medline]
6. Misra DP, Ananth CV. Infant mortality among singletons and twins in the United States during 2 decades: effects of maternal age. Pediatrics 2002;110:1163–8.
7. Luke B, Keith LG. The contribution of singletons, twins and triplets to low birth weight, infant mortality and handicap in the United States. J Reprod Med 1992;37:661–6.[Medline]
8. 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;41:114–25.[Medline]
9. Hack M, Fanaroff AA. Outcomes of extremely-low-birth-weight infants between 1982 and 1988. N Engl J Med 1989;321:1642–7.[Abstract]
10. Allen MC, Donohue PK, Dusman AE. The limit of viability—neonatal outcome of infants born at 22 to 25 weeks’ gestation. N Engl J Med 1993;329:1597–601.
11. Tyson JE, Younes N, Verter J, Wright LL. Viability, morbidity, and resource use among newborns of 501- to 800-g birth weight. National Institute of Child Health and Human Development Neonatal Research Network. JAMA 1996;276:1645–51.[Abstract]
12. Stevenson DK, Wright LL, Lemons JA, Oh W, Korones SB, Papile LA, et al. Very low birth weight outcomes of the National Institute of Child Health and Human Development Neonatal Research Network, January 1993 through December 1994. Am J Obstet Gynecol 1998;179:1632–9.[Medline]
13. Martin J, Curtin S, Saulnier M, Mousavi J. Development of the Matched Multiple Birth File. In: 1995–1998 Matched Multiple Birth Dataset. NCHS CD-ROM series 21, no. 13a. Hyattsville, MD: National Center for Health Statistics; 2003.
14. National Center for Health Statistics. 1995–1998 linked birth/infant death data set. Vital Statistics of the United States: Quality Control Procedures. Hyattsville, MD: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2000.
15. Shumpert MN, Salihu HM, Kirby RS. Impact of maternal anemia on birth outcomes of teen twin pregnancies: a comparative analysis with mature young mothers. J Obstet Gynaecol 2004;24:16–21.[Medline]
16. Salihu HM, Shumpert MN, Slay M, Kirby RS, Alexander GR. Childbearing beyond maternal age 50 and fetal outcomes in the United States. Obstet Gynecol 2003;102:1006–14.
17. Alexander GR, Kotelchuck M. Quantifying the adequacy of prenatal care: a comparison of indices. Public Health Rep 1996;111:408–18.[Medline]
18. Alexander GR, Cornely DA. Prenatal care utilization: its measurement and relationship to pregnancy outcome. Am J Prev Med 1987;3:243–53.[Medline]
19. Kogan MD, Martin JA, Alexander GR, Kotelchuck M, Ventura SJ, Frigoletto FD. The changing pattern of prenatal care utilization in the United States, 1981–1995, using different prenatal care indices. JAMA 1998;279:1623–8.
20. Buescher PA, Taylor KP, Davis MH, Bowling JM. The quality of the new birth certificate data: a validation study in North Carolina. Am J Public Health 1993;83:1163–5.
21. DiGiuseppe DL, Aron DC, Ranbom L, Harper DL, Rosenthal GE. Reliability of birth certificate data: a multi-hospital comparison to medical records information. Matern Child Health J 2002;6:169–79.[Medline]
22. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986;42:121–30.[Medline]
23. White HA. A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica 1980;48:817–38.
24. Clayton D, Hills M. Statistical models in epidemiology. New York (NY): Oxford University Press; 1993.
25. Hennekens CH, Buring JE. Measures of disease frequency and association. In: Mayrent SL, editor. Epidemiology in medicine. 1st ed. Boston (MA): Little, Brown;1987. p. 88.
26. Rees JM, Lederman SA, Kiely JL. Birth weight associated with lowest neonatal mortality: infants of adolescent and adult mothers. Pediatrics 1996;98:1161–6.
27. Blake DM, Lee MI. Twin pregnancy in adolescents. Obstet Gynecol 1990;75:172–4.
28. Yoder BA, Young MK. Neonatal outcomes of teenage pregnancy in a military population. Obstet Gynecol 1997;90:500–6.[Abstract]
29. Smith GC, Pell JP. Teenage Pregnancy and risk of adverse perinatal outcomes associated with first and second births: population based retrospective cohort study. BMJ 2001;323:476.
30. Kiely JL, Kiely M. Epidemiological trends in multiple births in the United States, 1997–1998. Twin Res 2001;4:131–3.[Medline]
31. Blickstein I, Goldman RD, Mazkereth R. Incidence and birth weight characteristics of twins born to mothers aged 40 years or more compared with 35–39 years old mothers: a population study. J Perinat Med 2001;29:128–32.[Medline]
32. Keith L, Oleszczuk JJ. Iatrogenic multiple birth, multiple pregnancy and assisted reproductive technologies. Int J Gynaecol Obstet 1999;64:11–25.[Medline]
33. Schieve LA, Peterson HB, Meickle SF, Jeng G, Danel I, Burnett NM, et al. Live-birth rates and multiple-birth risk using in vitro fertilization. JAMA 1999;282:1832–8.
34. Schieve LA, Meikle SF, Ferre C, Peterson HB, Jeng G, Wilcox LS. Low and very low birth weight in infants conceived with use of assisted reproductive technology. N Engl J Med 2002;346:731–7.
35. Reynolds MA, Schieve LA, Martin JA, Jeng G, Macaluso M. Trends in multiple births conceived using assisted reproductive technology, United States, 1997–2000. Pediatrics 2003;111:1159–62.
36. Salihu HM, Aliyu MH, Rouse DJ, Kirby RS, Alexander GR. Potentially preventable excess mortality among higher-order multiples. Obstet Gynecol 2003;102:679–84.
37. Salihu HM, Williams AT, McCainey TN, Kirby RS, Alexander GR. Early mortality among triplets in the United States: black-white disparity. Am J Obstet Gynecol 2004;190:477–84.[Medline]
38. Watkins ML, Edmonds L, McClean A, Mullins L, Mulinare J, Khoury M. The surveillance of birth defects: the usefulness of the revised US standard birth certificate. Am J Public Health 1996;86:731–4.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
