Objective. To examine the relationship between transient tachypnea of the newborn and asthma complicating pregnancy.
Design. Historical cohort analysis.
Setting. Singleton live deliveries in New Jersey hospitals during 1989 to 1992 (n = 447 963).
Patients. Mother-infant dyads were identified from linked birth certificate and maternal and infant hospital claims data. Women with an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis code (493) for asthma (n = 2289) were compared with a four-fold larger randomly selected control sample (n = 9156) from the remaining pool of women.
Main Outcome Measure. Transient tachypnea of the newborn.
Results. In the overall sample, after controlling for the confounding effects of important variables, infants of asthmatic mothers were more likely [odds ratio (OR), 1.79; 95% confidence interval (CI), 1.35–2.37] than infants of control mothers to exhibit transient tachypnea of the newborn. A stratified analysis by gestational age and sex revealed larger and statistically significant associations in term infants (OR, 2.02; 95% CI, 1.42–2.87) as opposed to preterm infants (OR, 1.51; 95% CI, 0.94–2.43) and in male infants (OR, 1.91; 95% CI, 1.35–2.71) as opposed to female infants (OR, 1.51; 95% CI, 0.92–2.47). On the other hand, after adjusting for important confounding variables, respiratory distress syndrome and maternal asthma were not found to be associated (OR, 1.14; 95% CI, 0.79–1.64).
Conclusion. The results of this study provide evidence that maternal asthma is a risk factor for transient tachypnea of the newborn and differences in gestational age and sex were apparent in this association. The mechanism for this association remains to be determined.
- OR =
- odds ratio •
- CI =
- confidence interval
Transient tachypnea of the newborn was first described by Avery and colleagues1 in 1966 as a distinct clinical condition causing neonatal respiratory distress. This syndrome is characterized chiefly by tachypnea shortly after birth, which clears quickly within 2 to 5 days.2 Neonates with this condition share similar roentgenographic findings of wet lung as described by Wesenberg et al.3 Transient tachypnea of the newborn is most commonly a benign and self-limited clinical condition.1 3 4 However, some studies have linked transient tachypnea of the newborn with substantial neonatal morbidity including the development of persistent fetal circulation5-8 as well as in childhood with symptoms and signs consistent with asthma and atopy.9
Transient tachypnea of the newborn has been reported to occur more frequently in preterm birth, cesarean delivery, and birth of a male infant.8 10 However, the etiology and pathogenesis of this condition is largely unknown. Delayed resorption of the lung fluid has been widely accepted as the central problem.1 2 7 11 The reason for the delayed absorption is unknown, but it has been suggested that it may be attributable to dysfunctional catecholamine regulation.12 13 Other theories include mild asphyxia resulting in a mild pulmonary capillary leak syndrome11 and myocardial dysfunction with elevated filling pressures.7
In a prospective study of 294 pregnant women with asthma, Schatz et al14 found an increased incidence of transient tachypnea of the newborn in the infants of asthmatic mothers as compared with controls. In an attempt to replicate this single report, we examined this issue among women delivering singleton births in a larger and more representative population.
The data for this historical cohort study were obtained from an administrative database that contains linked birth certificate, infant death certificate, and maternal and newborn hospital discharge claims data for New Jersey residents for all singleton live births in New Jersey hospitals during 1989 to 1992 (n = 447 963). The mother's and infant's hospital discharge data contain information extracted from discharge summaries of their delivery hospitalizations. Emergency room visits and hospitalization for asthma are not included in this database. The New Jersey Department of Health linked birth certificates and infant death certificates for each year, with only 43 death certificates remaining unmatched for 1989, and none being unmatched for 1990, 1991, and 1992. The Office of Technology and Information Systems then linked these data to the mother's and infant's hospital discharge data, which had been previously linked through the Medex system. The match rate for this database was 94.5% for 1989, 95.3% for 1990, 95.5% for 1991, and 95.8% for 1992.
The sampling unit was the birth, not the claim. Mother-infant dyads were identified as singleton live births to New Jersey residents in New Jersey hospitals during 1989 to 1992. In each calender year (1989 to 1992), mothers with medical claims with an International Classification of Diseases, Ninth Revision, Clinical Modification(ICD-9-CM) diagnosis code of 493 (asthma) were first selected (n = 391 for 1989, n = 540 for 1990, n = 647 for 1991, and n = 711 for 1992). A control sample that was four times the count of cases (n = 1564 for 1989, n = 2160 for 1990, n = 2588 for 1991, and n = 2844 for 1992) was randomly chosen from the remaining pool of mothers without an ICD-9-CM asthma diagnosis code in each calender year. Cases from the 4 years (n = 2289) were then combined and compared with the combined controls (n = 9156) with respect to patient characteristics and transient tachypnea of the newborn.
An International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis code of 770.6 (transient tachypnea of the newborn) was used to define the outcome variable.
Other Patient Characteristics
This information was obtained from the birth certificate. Very low birth weight was defined as birth weight <1500 g, low birth weight as birth weight <2500 g, and high birth weight as birth weight >4000 g.
Gestational age was obtained from birth certificate data. To reduce error because of misreporting of gestational age, we limited the analysis to infants of the 22- to 45-week range and this resulted in the removal of 35 (1.53%) cases and 240 (2.62%) controls from the 4 years for analysis involving this variable. This restriction has been found to reduce the degree of error in reported gestational age. Balcazar15 found that the greatest degree of error in reporting was at the extremes of the distribution. Preterm birth was defined as gestational age <37 completed weeks.
Neonatal death was defined as death ≤28 days, and infant death as death in ≤1 year and the information was obtained from death certificate data.
Mode of Delivery
The mode of delivery codes used were: forceps (ICD-9-CM procedure code, 72.0 to 72.4); vacuum (ICD-9-CM procedure code, 72.7); and cesarean (ICD-9-CM procedure code, 74.0).
Medical Disorders Complicating Pregnancy
Medical disorders considered were: type-I or insulin-dependent diabetes mellitus (ICD-9-CM diagnosis code, 250.0 to 250.9 with a fifth digit code of 1); type-II or noninsulin-dependent diabetes mellitus (ICD-9-CM diagnosis code, 250.0 to 250.9 with a fifth digit code of 0); gestational diabetes (ICD-9-CM diagnosis code, 648.0); or abnormal glucose tolerance (ICD-9-CM diagnosis code, 648.8). ICD-9-CM diagnosis codes of 401.0 to 405.0 and 642.0 to 642.2 defined preexisting hypertension.
Prenatal care was defined using The Adequacy of Prenatal Care Utilization Index developed by Kotelchuck.16 This index considers both adequacy of initiation of prenatal care and adequacy of received services. Variables used to create this index were obtained from the birth certificate.
Information on payer (insurance type) was obtained from the mother's primary insurance that was coded on the hospital Unified Billing Patient Summary at the time of delivery. HealthStart is a state-sponsored program designed to improve the pregnancy outcomes of women who are on Medicaid.
The mother's age, race/ethnicity, level of education, marital status, and parity were derived from the birth certificate data.
The ICD-9-CM diagnosis codes 304 and 648.3 were used to define illicit drug use in pregnancy. Information on cigarette smoking and alcohol use during the index pregnancy was obtained from the birth certificate data.
Transient tachypnea of the newborn was the dependent variable. Maternal asthma was the explanatory variable of interest.
Transient tachypnea of the newborn in relation to maternal asthma was analyzed using unconditional multiple logistic regression before and after accounting for potential confounding variables.17 Selection and order of entry of confounding variables for the model were determined using a priori knowledge according to the method described by Greenland.18Statistical significance of regression coefficients was determined by the χ2 approximation to the likelihood ratio statistic.
Statistical significance was defined as a two-tailed, P< .05. Statistical analysis was carried out using SAS statistical software.19
Descriptive Characteristics of the Study Population
There were a total of 447 963 singleton, live births to New Jersey residents in New Jersey hospitals during the 4 year period (1989 to 1992). Approximately 9% of the births were to mothers under the age of 20 years, and approximately 12% of the births were to mothers 35 years and older. Close to 15% of the mothers had less than a high school education and approximately 22% had a government assistance program as their primary source of medical insurance. Forty-four percent of the births were to primiparous mothers, and the parity of 55% ranged between two and five. Less than 1% had a parity of six and higher. According to the Kotelcheck Prenatal Care Utilization Index, approximately two thirds of the mothers received adequate or more than adequate prenatal care services and 14% received inadequate care.
The rates of most infant and maternal outcomes in our study population were comparable to published results whereas the reported use of cigarettes and illicit drugs during pregnancy was lower. For example, in our population, the rate of transient tachypnea of the newborn was 2%, of preterm infant was 11.9%, of small-for-gestational age was 11.9%, of large-for-gestational age was 2.9%, of cesarean delivery was 24.7%, and of placenta previa was 0.6%. These figures were in agreement with those reported in the literature.14 20-24
Descriptive Characteristics of the Sample Population
Comparison of the asthmatic mothers with the sample of nonasthmatic controls is displayed in Table1. Higher percentages of asthmatic mothers than of control mothers were represented at the extremes of the reproductive life (<20 and >35 years of age).
Asthmatic mothers were more likely than controls to be African-American or Hispanic, and to have Medicaid as their primary payer. They were less likely to be currently married and to have completed a high school or university education. Asthmatic mothers were also less likely than controls to have received adequate prenatal care and more likely to have received adequate plus prenatal care.
Asthmatic mothers also were more likely to have smoked cigarettes and used illicit drugs during the index pregnancy than control mothers, but no meaningful difference was seen in terms of the frequency of alcohol consumption during pregnancy. The rates of type I, type II, and gestational diabetes, abnormal glucose tolerance, preexisting hypertension, preterm birth, low birth weight, and cesarean delivery were higher among asthmatic than control mothers.
Transient Tachypnea of the Newborn in the Pregnancies of Asthmatic Women
In the overall sample, infants of asthmatic mothers were more likely than infants of control mothers to have transient tachypnea of the newborn before and after accounting for potential confounding variables (Table 2).
We examined the relationship between transient tachypnea of the newborn and maternal asthma after stratifying the births by gestational age (term or preterm infant). Before adjustment for important confounding variables, maternal asthma was significantly associated with transient tachypnea of the newborn in both term and preterm infants. However, the odds ratio (OR) was higher in term as compared with preterm infants (Table 3). After complete adjustment for important confounding variables, maternal asthma continued to be significantly associated with transient tachypnea of the newborn in term infants. The association did not achieve statistical significance in preterm infants (Table 3).
Stratified analysis by sex also revealed a sex difference in the strength of the association between maternal asthma and transient tachypnea of the newborn (Table 4). After complete adjustment for important confounding variables, maternal asthma continued to be associated with transient tachypnea of the newborn in the male infant. In the female infant this association did not achieve statistical significance.
In the overall sample, our analyses also reaffirmed previously recognized risk factors for transient tachypnea of the newborn8 10 25 such as cesarean delivery [OR, 2.16; 95% confidence interval (CI), 1.65–2.82], infant male sex (OR, 1.86; 95% CI, 1.41–2.45), and preterm infant (OR, 3.63; 95% CI, 2.74–4.81).
In our analysis, we did not find an association between respiratory distress syndrome and maternal asthma after adjusting for important confounding variables (OR for maternal asthma, 1.14; 95% CI, 0.79–1.64).
The neonatal and infant death rates were comparable in the asthmatic and control mothers: 0.3% (7/2289) versus 0.3% (32/9156) for neonatal deaths and 0.6% (14/2289) versus 0.5% (43/9156) for infant deaths. We further examined the rates of neonatal and infant deaths in those infants who exhibited transient tachypnea of the newborn and in those who did not exhibit transient tachypnea of the newborn at birth. Among infants of asthmatic mothers, the rates of neonatal and infant deaths in those infants who exhibited transient tachypnea of the newborn as compared with those infants who did not exhibit transient tachypnea of the newborn, respectively, were 0.0% (0/88) versus 0.32% (7/2201) for neonatal deaths and 1.14% (1/88) versus 0.59% (13/2201) for infant deaths. The corresponding figures among infants of the control mothers were 0.0% (0/172) versus 0.36% (32/8984) for neonatal deaths and 0.0% (0/172) versus 0.48% (43/8984) for infant deaths.
In this historical cohort study, we found that infants of asthmatic women, as compared with nonasthmatic control women, were more likely to exhibit transient tachypnea of the newborn. We also observed that the association between maternal asthma and transient tachypnea of the newborn was more pronounced in term births as opposed to preterm births and in infants of male sex as opposed to infants of female sex.
Strengths of the present study included: 1) large sample size; 2) historical cohort design; 3) extensive control for potentially confounding variables; 4) stratified analysis by gestational age and sex.
Several limitations of our study must be kept in mind. Administrative databases are prone to some degree of coding errors.26Because the analyses reported here depend on the patient data recorded at the time of obstetrical delivery, it is possible that only mothers with severe or active asthma would have been identified as such and mothers with less severe or inactive asthma may have been included in the control group. Likewise, pregnant women who had asthma but did not receive prenatal care in the preceding year could also be misclassified. However, this type of misclassification if anything, would have resulted in attenuating the association between transient tachypnea of the newborn and maternal asthma (ie, the misclassification would be random).
One surprising finding in our study population is that asthmatic mothers were more likely than controls to be smoking during the index pregnancy. We would have expected asthmatic mothers to quit smoking during pregnancy. This finding could result from more preferential reporting of smoking for the asthmatics than controls. However, this bias, if present and if smoking is associated with transient tachypnea of the newborn, would result in the ORs being biased toward the null and would not explain our findings. Further, smoking is not a recognized risk factor for transient tachypnea of the newborn. Although smoking is known to be underrepresented on birth certificates, in our analysis, we did not find reported smoking to be associated with transient tachypnea of the newborn (OR, 0.95; 95% CI, 0.65 to 1.39). A further limitation of our study is the likelihood of misdiagnosis in diseases such as transient tachypnea of the newborn. Many preterm infants with respiratory distress may have received a diagnosis of respiratory distress syndrome when the disease process was in fact, transient tachypnea of the newborn.
As demonstrated in our study population, as well as by other investigators, cesarean delivery has been found to have the most compelling association with transient tachypnea of the newborn, especially when the cesarean delivery is unlabored. In our sample the rate of cesarean delivery was higher among asthmatic than control mothers. This could have explained some of the association between transient tachypnea of the newborn and maternal asthma. In our analysis, adjustment for cesarean delivery diminished the associations (OR) between maternal asthma and transient tachypnea of the newborn but still maternal asthma was a significant predictor of transient tachypnea of the newborn independent of the effects of cesarean delivery. In our data set, information whether the cesarean delivery was labored or unlabored was not available.
In our study population, the rate of maternal asthma has almost doubled from 1989 (0.3%) to 1992 (0.6%). Although this increase might have been related to diagnostic fashion, the rate of increase is in congruence with the increase in mortality and hospitalizations attributable to asthma that has been observed in the last two decades in children and adults.27 28 To be sure that diagnostic fashion is not the explanation of our findings, we reran the analysis for each calender year separately. Other than fluctuations in precision of the results because of sampling variations, we did not observe a pattern (trend) during the 4 years in the ORs (95% CI) associating maternal asthma and transient tachypnea of the newborn. For example, the ORs (95% CI) associating maternal asthma and transient tachypnea of the newborn after adjusting for important confounding variables were 1.49 (0.66–3.39) for 1989, 1.85 (1.08–3.16) for 1990, 1.70 (0.97–2.97) for 1991, and 1.87 (1.22–2.86) for 1992.
Our finding that neonatal respiratory distress syndrome was not associated with maternal asthma coded in the hospital record at delivery is in agreement with a previous report14 that found no association between respiratory distress syndrome and maternal asthma. On the other hand, Bertrand and colleagues29 have shown an association of airway hyperreactivity and respiratory distress syndrome. In that study, children born preterm with the respiratory distress syndrome were found to have long-term pulmonary sequelae consisting of airway obstruction and airway hyperresponsiveness that are related to both the severity of the initial pulmonary insult and its treatment, as well as to a familial predisposition to airway hyperreactivity. Children born preterm, but who did not develop neonatal respiratory distress syndrome also had evidence of airflow limitation that was related to familial airway hyperreactivity and not to their preterm delivery. Based on their findings, Bertrand et al29 speculated a common pathway for idiopathic preterm labor and airway lability.
The lack of association between neonatal or infant deaths and transient tachypnea of the newborn or maternal asthma could be attributable to the small number of events and therefore we could not rule out the possibility of falsely declaring no association in which there might be one.
A potential proposed mechanism for the association between transient tachypnea of the newborn and maternal asthma has been that both the mothers and infants of asthmatic mothers have a genetic predisposition to β-adrenergic hyporesponsiveness.14 30 This hypothesis is further supported by the findings of a study9 that evaluated 58 children at ages of 4 to 5 years who were diagnosed as having transient tachypnea of the newborn at birth. In that study, infants with transient tachypnea of the newborn had a significantly higher incidence of recurrent episodes of wheezy breathlessness, symptoms consistent with asthma, and signs of atopy. Experiments have demonstrated that the lung epithelium secretes Cl− and fluid throughout gestation and develops the ability to actively reabsorb Na+ (fluid) only during late gestation.13 It is at birth that the mature lung switches from fluid secretion to Na+ (fluid) absorption in response to circulating catecholamine, steroids, and vasopressin.31-35 Changes in oxygen tension are thought to augment the Na+ transporting capacity of the epithelium and increase gene expression for the epithelial Na+channel.13 Biochemical studies of human infants' nasal epithelia also showed defective amiloride-sensitive Na+transport in transient tachypnea of the newborn.12 Gowen and colleagues12 studied the change in ion transport function by measuring the basal transepithelial potential difference across the ciliated epithelium of the nose in 85 term neonates during the first 72 hours of life. Basal potential differences during the first 24 hours of life were higher in neonates delivered by cesarean section without prior labor and in those with transient tachypnea of the newborn than in neonates born during normal spontaneous vaginal delivery or cesarean section with prior labor or in those with respiratory insufficiency. The percentage inhibition of potential difference by amiloride superfusion (<24 hours) was significantly lower in infants with transient tachypnea of the newborn and after cesarean section without prior labor than in other groups. By 48 hours, nasal potential differences after cesarean section without prior labor and in neonates with transient tachypnea of the newborn had declined, and by 72 hours, values were similar to those in other groups; respiratory rate paralleled the decline in potential difference. Amiloride sensitivity was similar for all groups by 72 hours. The findings of Gowen et al12 indicate that transient tachypnea of the newborn is associated with abnormal epithelial ion transport.
Our finding of an association between transient tachypnea of the newborn and maternal asthma has been described by another single study.14 To our knowledge, the sex and gestational age differences in the association between transient tachypnea of the newborn and maternal asthma are new.
The significant sex difference in the association of maternal asthma and transient tachypnea of the newborn, may be attributable to differences in the growth and maturation of the lungs in males and females, resulting in differing susceptibilities. Fetal lung structure has been shown to exhibit a sexual dimorphism in animal models.36 Early differentiation and maturation of female as compared with male fetal lung predates surfactant differences and are first observed during a period of epithelial proliferation.36 It is interesting to point out that males are more likely to have respiratory distress syndrome and airway hyperresponsiveness as neonates and asthma as adults, as well, suggesting a common association.37-39
Respiratory distress syndrome and transient tachypnea of the newborn may actually share some overlapping pathophysiologic mechanisms.31 In the study of Lawson et al,31 a marked decrease in fetal tracheal fluid flow and an increase in pulmonary surfactant efflux were found after infusion of epinephrine into fetal sheep suggesting that endogenous catecholamine may play a role in regulating both the surfactant level and fluid absorption.
Our finding of a more pronounced association between maternal asthma and transient tachypnea of the newborn in term infants as opposed to preterm infants may best be explained by the following mechanism. There may be two distinct contributing pathophysiologic principles operating in transient tachypnea of the newborn: 1) decreased surfactant, more prominent in preterm infants, and 2) β-adrenergic hyporesponsiveness, more significant in term infants. This may explain why there is an increased risk of transient tachypnea of the newborn in preterm infants while at the same time there is a closer association of maternal asthma with transient tachypnea of the newborn in term infants. With increasing gestational age, surfactant deficiency may be less of a problem, but a genetic predisposition to β-adrenergic hyporesponsiveness may become dominant.
Another issue that could theoretically play a role in the link between maternal asthma and transient tachypnea of the newborn is the use of antiasthmatic drugs by the mother. Several studies have shown that antiasthmatic medications such as β2-agonists40 41 and methylzanthines40 may increase surfactant secretion at first, followed by surfactant depletion in sustained high levels of these drugs. As well, β-adrenergic agonists have been shown to raise intracellular concentrations of cyclic adenosine monophosphate in excised human tissue and stimulate chloride secretion.12 On the other hand, steroids may enhance the maturity of the fetal lungs and are believed to decrease transient tachypnea of the newborn.40 42 However, we were unable to assess the role of antiasthmatic drugs in the association of transient tachypnea of the newborn and maternal asthma in our study population because of the unavailability of information on drug use.
The results of this study provide evidence that maternal asthma is a risk factor for transient tachypnea of the newborn and differences in sex and gestational age were apparent in this association. The mechanism for this association remains to be determined.
The study was supported by Grant 5-T32-PE10011 from the Division of Medicine of the Health Resources and Services Administration (HRSA). Its contents are solely the responsibility of the authors and do not necessarily represent the views of HRSA.
We wish to acknowledge the support of Maryanne Florio and Virginia Dato of the New Jersey Department of Health and Senior Services and the cooperation of the staff of the New Jersey Office of Technology and Information Systems in the provision of data.
- Received September 25, 1997.
- Accepted January 19, 1998.
Reprint requests to (K.D.) Department of Family Medicine, Robert Wood Johnson Medical School, 1 Robert Wood Johnson Place—CN19, New Brunswick, NJ 08903-0019.
- Buccirelli RL,
- Egan EA,
- Gessner IH,
- Eitzman DV
- Halliday HL,
- McClure G,
- McCreid M
- Shohat M,
- Levy G,
- Levy I,
- Schonfeld T,
- Merlob P
- Gowen CW,
- Lawson EE,
- Gingras J,
- Boucher RC,
- Gatzy JT,
- Knowles MR
- ↵Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic Research. Principles and Quantitative Methods. Belmont, CA: Lifetime Learning; 1982;419–446
- ↵SAS Institute Inc. Statistical Analysis System. Version 6. Cary, NC: SAS Institute Inc; 1989
- Ventura SJ,
- Martin JA,
- Tafel SM,
- et al.
- ↵Graham ADM. Pre-term labor and premature rupture of membranes. In: Hacker NF, Moore JG, eds. Essentials of Obstetrics and Gynecology. 2nd ed. Philadelphia, PA: Saunders Company; 1992:270–289
- Evans R III,
- Mullally DI,
- Wilson RW
- Kitterman JA,
- Ballard PL,
- Clements JA,
- Mescher EJ,
- Tooley WH
- Sears MR,
- Holdaway MD,
- Flannery EM,
- Herbison GP,
- Silva PA
- Taeusch HW,
- Frigoletto F,
- Kitzmiller J,
- et al.
- Copyright © 1998 American Academy of Pediatrics