BACKGROUND. Late preterm infant mortality is higher than that for term newborns. The association between weight for gestational age (WGA) category and late preterm mortality has not been well described.
OBJECTIVES. Our objectives for this research were as follows: (1) to compare neonatal and infant mortality rates of SGA, AGA, and LGA late preterm, early term, and term newborns; (2) to determine the relative risk of neonatal and infant death for each WGA category; and (3) to examine causes of neonatal and infant death.
METHODS. We reviewed linked birth and death certificate data for all infants from Utah born between 1999 and 2005 with a GA ≥34 weeks. We calculated neonatal and infant mortality rates for each GA/birth weight stratum and estimated mortality rate ratios using AGA term infants as the reference. International Classification of Diseases, Ninth Revision, codes were used to classify cause of death.
RESULTS. There were 343322 newborns with GA ≥34 weeks from 1999 to 2005. Late preterm SGA infants were ∼44 times more likely than term AGA newborns to die in their first month and 22 times more likely to die in their first year. When infants dying from congenital conditions were excluded, the differences in mortality rate ratios persisted for SGA infants, especially those born in the late preterm period.
CONCLUSIONS. Being SGA substantially increases the already higher mortality of late preterm and early term newborns. This increased risk cannot be fully explained by an increased prevalence of lethal congenital conditions among SGA late preterm newborns. Clinicians caring for late preterm and early term newborns should be cognizant of their WGA category.
Infants born before 37 weeks of gestation are at greater risk for morbidity and mortality compared with term newborns.1–10 Preterm births have increased 30% since 1981 and composed 12.5% of all US births in 2004.11 Nearly two thirds of the increase in preterm births is attributable to the increase in “late preterm” newborns, those born between 34 and 36 weeks of gestation.12,13 Late preterm newborns account for ∼75% of all singleton premature births in the United States.13 Because of their large numbers, deaths of infants born in the late preterm window represent a larger “etiologic fraction” of overall infant mortality than do more premature newborns.4 As a result, the mortality of late preterm newborns is of growing interest.5,6,14 Our group has previously reported substantially higher neonatal and infant mortality rates for late preterm newborns compared with term infants born in Utah.9
Gestational age (GA), however, is just 1 factor in newborn and infant mortality. In the 1960s, Battaglia and Lubchenco15 developed a classification system that combined GA with birth weight to categorize newborns as preterm, term, or postterm and small for GA (SGA), appropriate for GA (AGA), or large for GA (LGA). Using this classification scheme to compare outcomes among different weight-for-GA (WGA) groups, the same authors found substantial variation in the mortality rates at each week of GA for the different WGA groups.16 Much has changed in neonatal care, because these findings were published 35 years ago, and the current impact of WGA category on neonatal and infant mortality of late preterm newborns has not been reported. Using a large Utah database, we studied these issues with the hypothesis that late preterm newborns who were SGA would have higher mortality than newborns in other WGA categories.
PATIENTS AND METHODS
Setting and Human Subjects Protection
The study was conducted using records from the Utah Department of Health Office of Vital Records and Statistics. Utah statute 26-2-22 and administrative rule R436-17 allow the release of birth and death data for qualified research studies. The institutional review board of the University of Utah determined that this study was exempt from the federal regulations governing human research.
Study Design and Subject Selection
This retrospective cohort study population included all of the infants born in Utah between 1999 and 2005 with a GA of ≥34 weeks. GA was assigned by clinicians caring for the infant on the basis of the best estimate using available information from the last menstrual period, prenatal ultrasound, and/or newborn examination.
Data abstracted from the probabilistically linked database included birth day/month/year, birth weight, GA, death during the first 364 days, age at death, and cause of death.
Subjects were categorized by the exposure of interest, GA as related to birth weight category (eg, SGA/AGA/LGA at 34–42 weeks). We plotted the birth weights of the newborns for each week of GA to create WGA curves. The 10th percentile and 90th percentile cutoffs for each week of GA were identified and used to divide the newborns into SGA, AGA, and LGA categories. Newborns were grouped into late preterm (34–36 weeks), early term (37–38 weeks), and term (39–42 weeks) categories, thus creating 9 possible strata. The primary outcomes were neonatal and infant mortality (death occurring before 28 days and 1 year of age) for each stratum.
Frequencies of birth weights were determined and used to establish weight parameters for SGA (<10%), AGA (>90%), and LGA (>90%) infants at each week of gestation (34–42 weeks) for this population. Infant and neonatal mortality rates were calculated for each GA/birth weight stratum. Modified Poisson regression was used to estimate incident rate ratios for death for each stratum using AGA infants born at 40 weeks as the reference.17 All of the statistical comparisons were done using Stata 10 statistical software (Stata Corp, College Station, TX).
Causes of death were categorized into 1 of the 8 groups proposed by the international collaborative effort (ICE) on birth weight, plurality, perinatal, and infant mortality, which uses International Classification of Diseases, Ninth Revision, codes to classify deaths as resulting from congenital conditions, immaturity-related conditions, asphyxia-related conditions, infections, other specific conditions, sudden infant death syndrome, external causes, and remaining causes.18
There were 343322 live newborns with GA ≥34 weeks born in Utah between 1999 and 2005. Of these live newborns, 365 died in the first 28 days (neonatal mortality rate of 1.1 deaths per 1000 live births), and 827 died in the first year (infant mortality rate of 2.4 deaths per 1000 live births).
Neonatal Mortality by WGA and Gender
Table 1 shows the neonatal mortality rates for each WGA category by gender. As expected, mortality decreased as GA increased for both boys and girls. Of note, however, is that SGA infants had the highest mortality rates at each week of GA.
Figure 1 displays the neonatal mortality rate ratios for the WGA categories for late preterm, early term, and term newborns (with term AGA infants serving as the referent group). Of the 9 categories, late preterm SGA infants had the highest mortality, being ∼44 times more likely to die in the first month than the term AGA referent group. Compared with term AGA newborns, those in 7 of the WGA/GA strata categories, including the early term category (37 and 38 weeks), had a significantly higher risk of dying in the first month. There was no significant difference in neonatal mortality between LGA and AGA term newborns.
Infant Mortality by WGA and Gender
Table 2 shows the infant mortality rates for each WGA category by gender. Similar to the neonatal mortality findings, infant mortality decreased as GA increased for both boys and girls. Again, there was a strikingly higher mortality rate among the SGA infants at each week of GA.
Figure 2 displays the infant mortality rate ratios for the WGA categories for late preterm, early term, and term newborns (with term AGA infants serving as the referent group). Of the 9 categories, late preterm SGA infants had the highest mortality, being ∼21 times more likely to die in the first year than the term AGA referent group. Compared with term AGA newborns, those in 7 of the WGA/GA strata categories, including those born at early term (37 and 38 weeks), had a significantly higher risk of dying in the first year. There was no significant difference in infant mortality between LGA and AGA term newborns.
Neonatal and Infant Causes of Death
Causes of death were grouped according to the ICE.18 Congenital conditions accounted for 69% of the neonatal deaths (those occurring in the first 28 days), with immaturity, asphyxia, and infection collectively accounting for another 25%. Congenital conditions accounted for 47% of deaths in the first year (infant mortality), with sudden infant death syndrome the cause in 15%. Immaturity, asphyxia, and infection together accounted for 15%. Other specific causes and external causes each accounted for 5% of infant deaths in this population.
The causes of deaths in late preterm and term infants for each WGA category are shown in Table 3. Of the 67 deaths in the SGA late preterm infants, 75% died of congenital conditions compared with 37% of the 194 infant deaths among term AGA newborns.
Neonatal and Infant Mortality Excluding Congenital Anomalies
Because lethal congenital conditions may be more common among late preterm newborns, we re-estimated the neonatal and infant mortality rate ratios for each WGA category after excluding infants who died from congenital conditions. Term AGA female infants served as the referent group for these calculations. After excluding deaths from congenital conditions, late preterm SGA male and female newborns were 16.6 and 14.2 times more likely than the referent group to die in the first 28 days, respectively, as shown in Table 4. Neonatal mortality rate ratios remained lowest for AGA infants within each GA category (late preterm, early term, or term).
Similarly, after removing those infants who died in the first year from congenital conditions, late preterm SGA male and female newborns were 14.2 and 5.8 times more likely than term girls born AGA (the referent group) to die during the first year, respectively (Table 5). Infant mortality rate ratios were lowest for AGA infants in every GA category.
Over the last several years there has been increasing interest in the outcomes of late preterm newborns. As a group, late preterm newborns have substantially higher neonatal and infant mortality rates.4,9 Our finding that, within this group of newborns, additional risk is associated with being small or large for GA should alert clinicians to pay particular attention to late preterm newborns whose intrauterine growth pattern has been abnormal. Our study supports the findings of many previous investigations that have demonstrated the significantly increased neonatal and infant mortality risks for infants born SGA and emphasizes the particular impact of SGA on mortality for late preterms.15,19–22
Using Utah-specific WGA curves, we categorized the late preterm, early term, and term newborns as SGA, AGA, or LGA. We found, as have several other studies, that, overall, late-preterm newborns had the greatest risk of dying.4,6 We also found that even newborns who we consider to be “early term” (37–38 weeks) had an increased risk of death compared with those born at ≥39 weeks of age and that this risk increased significantly if they were also SGA, a finding that has not been reported previously. These findings point to substantial heterogeneity in the risk of mortality even among term newborns, that is, those of 37 to 42 weeks' gestation.
Similar to Tomashek et al,6 we found that congenital anomalies accounted for a higher proportion of the deaths in late preterm newborns compared with term newborns. However, when these infants were excluded, we found that the differences in mortality rate ratios persisted for SGA infants, especially those born in the late preterm period, indicating that the increased risk cannot be fully explained by an increased prevalence of lethal congenital conditions among SGA late preterm newborns.
Our study had several limitations. Because we relied on birth and death certificate data, our data may include instances of misclassification of GA, birth weight, and cause of death that we were unable to identify or correct. In addition, data regarding Utah newborns may not be generalizable to other populations. Utah's population is predominantly white, ∼11% of whom are Latino.23 Late-preterm newborns from populations with large numbers of blacks, for example, may have different mortality risks.24
Our results suggest several areas for future research. Information is needed about specific maternal and fetal factors associated with delivery in the preterm and early term periods. Research identifying possible associations between various reasons for late-preterm delivery and infant outcomes would be extremely useful to obstetricians and other clinicians as they weigh the risks and benefits to mother and child when making decisions regarding the ideal timing of delivery. Similarly, we need a better understanding of the pathophysiology of the conditions leading to an infant being SGA and the optimal timing of delivery. Evidence of this sort could help obstetric providers identify the infants most at risk and optimize their outcomes. Finally, more information about morbidities in late-preterm and SGA populations would be helpful in guiding those clinicians faced with management of these at-risk infants.
We thankfully acknowledge the assistance of Jeff Duncan (Director, Office of Vital Records and Statistics, Utah Department of Health).
- Accepted February 12, 2009.
- Address correspondence to Laurie S. Pulver, MD, MPH, 295 Chipeta Way, Salt Lake City, UT 84158. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject
Late preterm infants have higher morbidity and mortality than term newborns.
What This Study Adds
Weight-for-gestational-age classification provides insight into the heterogeneity that exists among late preterm and early term newborns. Even when newborns dying of congenital anomalies are excluded, SGA late preterm and early term infants have higher mortality.
- ↵Bhutani VK, Johnson L. Kernicterus in late preterm infants cared for as term healthy infants. Semin Perinatol. 2006;30 (2):89– 97
- Wang ML, Dorer DJ, Fleming MP, Catlin EA. Clinical outcomes of near-term infants. Pediatrics.2004;114 (2):372– 376
- ↵Young PC, Glasgow TS, Li X, Guest-Warnick G, Stoddard G. Mortality of late-preterm (near-term) newborns in Utah. Pediatrics.2007;119 (3). Available at: www.pediatrics.org/cgi/content/full/119/3/e659
- ↵Khashu M, Narayanan M, Bhargava S, Osiovich H. Perinatal outcomes associated with preterm birth at 33 to 36 weeks' gestation: a population-based cohort study. Pediatrics.2009;123 (1):109– 113
- ↵Institute of Medicine. Preterm Birth: Causes, Consequences, and Prevention. Brief report. Washington, DC: Institute of Medicine; 2006
- ↵Raju TNK, Higgins RD, Stark AR, Leveno KJ. Optimizing care and outcome for late-preterm (near-term) infants: a summary of the workshop sponsored by the National Institute of Child Health and Human Development. Pediatrics.2006;118 (3):1207– 1214
- ↵Zou G. A modified Poisson regression approach to prospective studies with binary data. Am J Epidemiol.2004;159 (7):702– 706
- ↵Cole S, Hartford RB, Bergsjo P, McCarthy B. International collaborative effort (ICE) on birth weight, plurality, perinatal, and infant mortality: III–a method of grouping underlying causes of infant death to aid international comparisons. Acta Obstet Gynecol Scand.1989;68 (2):113– 117
- Malloy MH. Size for gestational age at birth: impact on risk for sudden infant death and other causes of death, USA 2002. Arch Dis Child. 2007;92 (6):F473– F478
- ↵US Census Bureau. Annual State Population Estimates With Sex, 6 Race Groups (5 Race Alone Groups and One Group With Two or More Race Groups) and Hispanic Origin: April 1, 2000 to July 1, 2006. US Census Bureau; 2007 Available at: www.census.gov/popest/states/asrh/files/SCEST20066RACEALL.CSV. Accessed October 13, 2008
- ↵Alexander GR, Kogan M, Bader D, Carlo W, Allen M, Mor J. US birth weight/gestational age-specific neonatal mortality: 1995–1997 rates for whites, Hispanics, and blacks. Pediatrics.2003;111 (1). Available at: www.pediatrics.org/cgi/content/full/111/1/e61
- Copyright © 2009 by the American Academy of Pediatrics