- Email: [email protected]
Respiratory distress syndrome and maternal birth weight effects
Respiratory Distress Syndrome and Maternal Birth Weight Effects THOMAS P. STRANDJORD, MD, IRVIN EMANUEL, MD, MICHELLE A. WILLIAMS, ScD, WENDY M. LEISENRING, PhD, AND CHRISTY KIMPO, PhD Objective: To study traditional risk factors and the intergenerational risk factor maternal low birth weight (LBW) for respiratory distress syndrome (RDS) in infants in multiple ethnic groups. Methods: The population-based database consists of hospital records linked to Washington state maternal and infant vital records. Four racial-ethnic groups were studied, whites, blacks, Native Americans, and Hispanics. Poisson regression models were used to estimate relative risks of various factors for RDS. Results: Rates for RDS were whites 1.2%, blacks 1.9%, Native Americans 1.3%, and Hispanics 1.0%. Maternal LBW was associated with increased relative risk (RR) for RDS in whites (2.6, 95% confidence interval [CI] 1.6, 4.2) and blacks (3.3, 95% CI 1.9, 5.6) for infants born vaginally. Compared with mothers of normal infants, birth weights of mothers of infants with RDS and delivered vaginally were significantly lower in whites, blacks, and Native Americans. The association of maternal LBW with RDS persisted in blacks even when multiple risk factors were added to the model (RR 2.4; 95% CI 1.1, 5.1). Conclusion: The association of maternal LBW with RDS is probably due in part to the association of maternal LBW with infant LBW and preterm birth. The strong persistent association of maternal LBW with RDS in blacks suggests that improvement of perinatal outcomes in that group will require improvement of long-term birth weight distribution. (Obstet Gynecol 2000;95:174 –9. © 2000 by The American College of Obstetricians and Gynecologists.)
continues to be an important cause of infant morbidity and mortality. In 1996, RDS was the fourth leading cause of infant death in the United States, accounting for 4.8% of infant deaths.2 Knowledge of factors that predispose infants to immature lungs is essential for prevention. Factors previously shown to influence risk of RDS include prematurity,3,4 mode of delivery,5–7 maternal hypertension,8 prolonged rupture of amniotic membranes,9 and ethnicity.10 Maternal birth weight was found to be inversely related to several adverse pregnancy outcomes including low birth weight (LBW) and preterm birth, which are related to RDS, intrauterine growth retardation, and perinatal and infant mortality.11 A hospital-based study found that infants with RDS had significantly lower maternal birth weights than other infants.12 Most epidemiologic studies of RDS are hospital based and might lack generalizability. Relatively few population-based epidemiologic studies of RDS have been done, and few have considered more than one or two factors at a time.6,13 We report a study of RDS using a large population-based intergenerational database of births to women in four ethnic groups, whites, blacks, Native Americans, and Hispanics. We also studied the association of maternal birth weight with the subsequent risk for the development of RDS.
Despite dramatic progress in understanding the etiology of infant respiratory distress syndrome (RDS),1 it
Methods
From the Department of Pediatrics, School of Medicine, Department of Epidemiology, School of Public Health, Center on Human Development and Disability, and Maternal and Child Health Program, University of Washington; and the Fred Hutchinson Cancer Research Center, Seattle, Washington. Supported by grant MCJ-530837 from the Maternal and Child Health Bureau (Title V, Social Security Act), Health Resources and Services Administration, US Department of Health and Human Services.
174 0029-7844/00/$20.00 PII S0029-7844(99)00532-3
Since 1987, all nonfederal short-stay hospitals have been legally required to submit a discharge summary of all admissions. The Washington State Department of Health links the summaries of obstetric and neonatal admissions to all data from the live birth, infant death, and fetal death certificates, to form the Birth Event Records Database. Compared with a series of hospital records the file more accurately reports medical infor-
Obstetrics & Gynecology
mation than the birth certificates or hospital discharge summaries separately.14 A random number generator within the SAS for Windows statistical package (SAS Institute Inc., Cary, NC) was used to generate a 10% random sample, stratified by birth year, of all infants born to white mothers from 1987 to 1995. A full sample of births to black, Native American, and Hispanic mothers was made over the same period. We used a deterministic methodology to link the files to birth certificates of women identified on the infant vital records as born in Washington state. A success rate of 88.8% resulted in linking 46,246 births to birth certificates of 38,513 women born between 1949, when birth weight was first recorded, and 1979. Complete details of the database, the Washington State Intergenerational Study of Birth Outcomes, have been reported.15 We also linked the driver’s license database to obtain selfreported heights of the women. We compared selfreported heights from the driver’s license database to the measured heights of 463 women. The mean difference was 0.1 in. and the correlation was .95. All mothers and infants were singletons. Implausible combinations of birth weight and gestational age were detected and removed as described.15 A total of 1.24% of infant gestations were considered outliers. Infant birth weights ranged from 255 to 6350 g, and gestational ages ranged from 20 to 46 weeks. Maternal birth weights ranged from 765 to 6067 g. Diagnostic data in International Classification of Diseases, 9th Revision, codes16 from discharge summaries were combined with diagnostic data from vital records to yield the primary outcome variable, RDS, and other medical variables. The diagnosis of RDS was accepted if it appeared on the discharge summary, birth certificate, or both. Discharge summaries recorded 84.4% (n ⫽ 489) of RDS cases, compared with 32.6% (n ⫽ 189) on birth certificates, with 17.1% (n ⫽ 99) from both sources. Statistical analysis was done with Stata Release 5 software (Stata Corp., College Station, TX). Generalized linear models with log link functions and Poisson error structures were fit to the binary end point RDS.17 That model obtained unadjusted and adjusted relative risk (RR) estimates and accompanying confidence intervals (CIs). Individual women occasionally had more than one infant during the study, so the method of generalized estimating equations was used to adjust variance estimates to account for correlation between infants born to the same woman.18 Factors included in multivariate models were mode of delivery (vaginal or cesarean), infant LBW (under 2500 g), preterm birth (under 37 weeks’ gestation), maternal LBW (under 2500 g), maternal age (under 20 years, or 20 or more years), parity (0, 1, or more), type of health insurance (Medicaid or other), marital status (married or single),
VOL. 95, NO. 2, FEBRUARY 2000
smoked during pregnancy (yes or no), prolonged rupture of membranes (at least 12 hours or less than 12 hours), maternal hypertension (normotensive, or pregnancy induced or chronic hypertension), established diabetes (yes or no), gestational diabetes (yes or no), maternal height (inches), late prenatal care (first visit before 4 months, or first visit after 4 months or no prenatal care). Maternal education only became available in 1992 so models with that variable markedly reduced the sample and were not used.
Results Table 1 presents the characteristics of successfully linked mothers and infants of all ethnic groups. White mothers were at least sociodemographic risk, whereas black mothers were at highest risk. The risk for RDS was inversely related to infant birth weight and gestational age (Tables 2 and 3). Over 40% of infants in all ethnic groups with birth weights under 1250 g were affected by RDS (Table 2). The incidence of RDS did not appear to be associated with birth weight in infants over 2500 g, so all subsequent analyses dichotomized infant birth weight to under 2500 g and at least 2500 g. Over half of infants of less than 28 weeks’ gestation had RDS, and incidence of RDS did not appear to be associated with gestational age in infants of greater than 37 weeks’ gestation (Table 3), so all subsequent analyses dichotomized infant gestation to under 37 weeks and at least 37 weeks. Frequency of RDS also was related to race. The overall frequency of RDS was 1.2% in singleton infants of white mothers, 1.9% in blacks, 1.3% in Native Americans, and 1.0% in Hispanics. Those rates were not significantly different from infants whose records could not be linked to mothers’ birth certificates (1.0% in whites, 2.0% in blacks, 1.2% in Native Americans, and 1.3% in Hispanics). Compared with white infants, the RR for RDS was 1.6 (95% CI 1.3, 2.0) for blacks, 1.1 (95% CI 0.9, 1.4) for Native Americans, and 0.9 (95% CI 0.7, 1.1) for Hispanics. In LBW infants the frequency of RDS was 16.1% in whites, 13.8% in blacks, 18.5% in Native Americans, and 13.2% in Hispanics. White infants delivered by cesarean were 3.7 times (95% CI 2.9, 4.7) more likely to develop RDS compared with infants delivered vaginally. The corresponding RR (95% CI) for the other ethnic groups were blacks 3.3 (2.4, 4.6), Native Americans 3.9 (2.6, 5.7), and Hispanics 3.2 (2.0, 5.2). Infants delivered by cesarean were more likely to be LBW. For white mothers 6.5% of infants delivered by cesarean were LBW compared with 3.5% of those delivered vaginally, for blacks 15.3% versus 10.2%, for Native Americans 9.3% versus 4.5%, and for Hispanics 7.2% versus 4.9%. A less pronounced trend was seen between gestational age and mode of delivery. For
Strandjord et al
Maternal Birth Weight
175
Table 1. Maternal and Infant Characteristics* Characteristic
White†
Black†
Native American†
Hispanic†
Maternal age (mean) Maternal age ⬍20 y (%) Maternal height (in.) Primigravida (%) Maternal LBW (%) Unmarried (%) Medicaid (%) Smoked during pregnancy (%) Late prenatal care (%) Prolonged rupture of membranes (%) Cesarean delivery Maternal hypertension Established diabetes (%) Gestational diabetes (%) Infant LBW (%) Preterm birth (%)
26.5 (23,177) 11.5 (23,177) 65.0 (17,324) 43.9 (22,940) 5.4 (22,381) 24.3 (23,154) 30.3 (20,889) 24.0 (21,842) 16.1 (22,378) 6.6 (23,187) 19.4 (23,187) 5.1 (23,187) 0.5 (22,381) 2.7 (23,187) 4.1 (23,150) 6.9 (20,329)
23.6 (7128) 27.8 (7128) 64.8 (3978) 44.4 (7058) 11.1 (5138) 70.9 (7122) 63.2 (4934) 27.4 (6133) 32.4 (6398) 9.0 (7133) 22.6 (7133) 6.4 (7133) 0.5 (5138) 2.4 (7133) 11.3 (7109) 13.3 (5937)
24.0 (7883) 24.5 (7883) 63.8 (4118) 35.3 (7833) 6.1 (5354) 65.0 (7866) 62.6 (5181) 34.2 (7395) 34.5 (7474) 6.0 (7884) 15.9 (7884) 4.3 (7884) 0.4 (5354) 2.5 (7884) 5.3 (7869) 11.8 (6611)
22.9 (6789) 31.5 (6789) 63.1 (3957) 45.7 (6745) 5.8 (4956) 50.1 (6781) 64.6 (4779) 13.7 (6457) 29.8 (6491) 5.5 (6790) 18.1 (6790) 5.2 (6790) 0.7 (4956) 3.0 (6790) 5.3 (6780) 10.4 (5747)
LBW ⫽ low birth weight. * Live-born singleton infants of singleton mothers. † Number of subjects for each variable in parentheses.
white mothers, 9.2% of infants delivered by cesarean were preterm compared with 6.3% delivered vaginally, for blacks 14.8% versus 12.8%, for Native Americans 12.9% versus 11.6%, and for Hispanics 11.1% versus 10.2%. Despite association of cesarean delivery with LBW and preterm birth, when those three factors were considered together in a multiple Poisson regression for risk of RDS, cesarean delivery was still a strong risk factor. The RR (95% CI) for RDS in infants delivered by cesarean delivery, taking into account the effects of LBW and preterm birth was 2.5 (2.0, 3.3) for whites, 2.2 (1.5, 3.1) for blacks, 2.2 (1.5, 3.4) for Native Americans, and 2.4 (1.5, 4.1) for Hispanics. Given the potent effect of delivery mode on risk of RDS, we evaluated maternal birth weight effects separately in infants delivered by cesarean and vaginally. In infants delivered by cesar-
ean, maternal LBW was not significantly related to RDS in any ethnic group (Table 4), which justified further separate analyses for infants delivered by the two modes (Tables 4 and 5). In addition to maternal birth weight, the strongest risk factors for RDS were markedly attenuated in infants delivered by cesarean, infant LBW, preterm birth, and prolonged rupture of membranes. Established diabetes was a strong risk factor in vaginally delivered Native Americans and Hispanics, and statistically insignificant in whites. Insufficient numbers in some ethnic groups prevented complete comparisons. White and black infants delivered vaginally were at increased risk of RDS if their mothers LBW: RR 2.6 (95% CI 1.6, 4.2) and 3.3 (95% CI 1.9, 5.6), respectively. Maternal LBW was associated with an almost doubled
Table 2. Incidence of Respiratory Distress Syndrome by Birth Weight* % RDS (n) Birth weight (g) ⬍1000 1000 –1249 1250 –1499 1500 –1749 1750 –1999 2000 –2249 2250 –2499 2500 –2749 2750 –2999 3000 –3249 3250 –3499 ⱖ3500 g Total
White
Black
Native American
Hispanic
54.6% (55) 46.7% (30) 60.5% (38) 24.7% (81) 19.6% (97) 11.5% (192) 5.3% (454) 3.5% (877) 1.3% (1995) 0.6% (3378) 0.4% (4591) 0.3% (11,362) 1.2% (23,150)
63.1% (65) 42.1% (38) 31.6% (38) 19.7% (76) 15.0% (100) 5.1% (158) 1.5% (334) 1.0% (590) 0.5% (981) 0.4% (1330) 0.4% (1397) 0.3% (2026) 1.9% (7109)
56.7% (30) 41.2% (17) 52.2% (23) 50.0% (38) 16.7% (42) 8.0% (75) 4.7% (191) 1.6% (314) 0.8% (671) 0.4% (1164) 0.4% (1610) 0.1% (3740) 1.3% (7869)
43.3% (30) 63.6% (11) 31.3% (16) 28.0% (25) 17.1% (35) 7.4% (81) 2.4% (165) 2.0% (351) 1.0% (726) 0.0% (1220) 0.1% (1397) 0.2% (2765) 1.0% (6780)
RDS ⫽ respiratory distress syndrome. * Includes only infants with known birth weights.
176 Strandjord et al
Maternal Birth Weight
Obstetrics & Gynecology
Table 3. Incidence of Respiratory Distress Syndrome by Gestational Age* % RDS (n) Gestational age (wk) ⬍28 28 –29 30 –31 32–33 34 –35 36 –37 38 –39 ⱖ40 Total
White
Black
Native American
Hispanic
51.1% (45) 61.8% (34) 37.0% (54) 18.0% (128) 10.8% (463) 2.0% (1980) 0.4% (7086) 0.2% (10,539) 1.1% (20,329)
52.5% (59) 63.2% (19) 37.5% (48) 11.9% (84) 2.74% (292) 0.9% (781) 0.6% (2066) 0.3% (2588) 1.8% (5937)
68.2% (22) 50.0% (20) 33.3% (27) 12.5% (88) 6.1% (312) 0.5% (927) 0.2% (2450) 0.2% (2765) 1.2% (6611)
53.3% (30) 33.3% (15) 30.0% (20) 7.8% (64) 4.5% (246) 0.9% (673) 0.3% (2040) 0.0% (2659) 0.9% (5747)
Abbreviation as in Table 2. * Includes only infants with known gestational ages.
risk of RDS (RR 1.9, 95% CI 0.7, 5.4) among Hispanics, although not statistically significant (Table 5). Maternal LBW was not associated with RDS among Native Americans. Several risk factors were considered confounders of the association of maternal LBW with RDS because each altered the RR of maternal LBW by 10% or more in each of the ethnic groups, infant LBW, preterm birth, maternal smoking, and late or no prenatal care. Full models with multiple Poisson regression for the effects of those adjusting variables on maternal LBW were constructed for each ethnic group, delivered vaginally (Table 6) or by cesarean. For infants delivered by cesarean, the addition of covariates did not alter the relation of maternal birth weight and RDS (data not shown). In Native American and Hispanic infants delivered vaginally, addition of covariates also failed to make maternal LBW a significant risk factor for RDS. With adjustment, maternal LBW was no longer significantly associated with RDS in whites delivered vaginally, RR 1.5 (95% CI 0.9, 2.6). In blacks, with adjustment the association was reduced but remained statistically significant, RR 2.4 (95% CI 1.1, 5.1, P ⫽ .024). In all ethnic
groups infant LBW was strongest in reducing the association of maternal LBW when infants were delivered vaginally. Maternal adult height was not related to RDS in infants.
Discussion There was a pattern of increased risk among vaginally delivered infants, but it was statistically significant only in whites and blacks. Adjustment, particularly for infant LBW and prematurity, reduced the association, which remained statistically significant for blacks, indicating that one pathway by which maternal LBW related to RDS was through its effect on infant birth weight or gestation. After adjustment, statistical significance persisted in black infants, indicating there might be other mechanisms besides effect of maternal birth weight on infant size which might explain the association of maternal birth weight with RDS. Cesarean delivery was a potent risk factor for RDS in previous studies.3,5–7,13 In this study, cesarean delivery was associated with a higher risk for RDS in all ethnic groups. The association of cesarean delivery with RDS
Table 4. Relative Risk for Respiratory Distress Syndrome Among Infants Delivered by Cesarean* RR (95% confidence interval) Risk factor
White
Black
Native American
Hispanic
Maternal LBW Infant LBW Preterm birth Medicaid Primigravida Prolonged rupture of membranes Maternal hypertension Gestational diabetes Late prenatal care
0.4 (0.1, 1.2) 19.2 (13.8, 26.8) 24.7 (16.6, 36.9) 1.6 (1.1, 2.3) 0.7 (0.5, 1.1) 3.6 (2.4, 5.3) 1.4 (0.8, 2.4) 1.5 (0.7, 2.9) 1.7 (1.1, 2.6)
1.1 (0.6, 2.3) 25.8 (14.1, 47.3) 14.2 (8.0, 25.4) 1.3 (0.8, 2.2) 1.3 (0.6, 1.6) 3.80 (2.3, 6.2) 2.1 (1.2, 3.8)
2.0 (0.8, 4.8) 31.2 (15.8, 61.5) 37.9 (14.9, 96.2) 1.2 (0.6, 2.1) 0.6 (0.3, 1.2) 2.95 (1.5, 6.0) 1.9 (0.8, 4.4) 1.1 (0.3, 4.3) 1.5 (0.8, 2.8)
1.5 (0.5, 4.7) 21.2 (10.4, 43.3) 27.3 (10.3, 72.5) 1.8 (0.8, 4.1) 0.4 (0.2, 0.9) 1.2 (0.4, 4.0) 3.1 (1.5, 6.9) 3.2 (1.1, 8.9) 1.3 (0.6, 2.9)
†
0.9 (0.5, 1.6)
RR ⫽ relative risk; LBW ⫽ low birth weight. * Each factor is considered individually; only factors with a significant RR in at least one of the ethnic groups are shown. † Missing cells due to insufficient cases.
VOL. 95, NO. 2, FEBRUARY 2000
Strandjord et al
Maternal Birth Weight
177
Table 5. Relative Risk for Respiratory Distress Syndrome Among Infants Delivered Vaginally* RR (95% confidence interval) Risk factor
White
Black
Native American
Hispanic
Maternal LBW Infant LBW Premature infant Unmarried Prolonged rupture of membranes Established diabetes Teenage mother
2.6 (1.6, 4.2) 33.6 (24.4, 46.2) 25.6 (17.7, 37.1) 0.6 (0.4, 0.9) 6.1 (4.3, 8.8) 1.9 (0.3, 13.3) 1.0 (0.6, 1.7)
3.3 (1.9, 5.6) 35.3 (20.1, 62.1) 27.2 (14.6, 52.4) 1.3 (0.7, 2.2) 7.9 (4.9, 12.6)
1.1 (0.4, 3.0) 68.4 (38.0, 123.0) 77.9 (28.0, 216.4) 1.5 (0.9, 2.8) 8.6 (5.1, 14.5) 11.7 (3.0, 44.9) 2.5 (1.5, 4.1)
1.9 (0.7, 5.4) 52.8 (26.6, 104.6) 85.0 (25.9, 278.7) 3.4 (1.5, 7.5) 10.2 (5.5, 19.0) 9.9 (1.5, 65.2) 1.9 (1.0, 3.5)
†
1.3 (0.8, 2.1)
Abbreviations as in Table 4. * Each factor is considered individually; only factors with a significant RR in at least one of the ethnic groups are shown. † Missing cells due to insufficient cases.
persisted even when infant birth weights and gestational ages were included in a multiple regression model with mode of delivery. Also, cesarean delivery attenuated associations with most other risk factors for RDS, including maternal LBW. There was an association of maternal LBW with cesarean delivery; 21.7% of LBW mothers were delivered by cesarean compared with 18.8% of normal birth weight mothers (P ⬍ .001). The association of cesarean delivery with maternal LBW might partially confound the association of maternal LBW with RDS. The strong association of mode of delivery with risk of RDS suggests that in future investigations of RDS, separate analyses should be made for infants delivered by each mode. Diagnoses of RDS were made at multiple institutions by many observers, so it is reassuring that the association of RDS with infant birth weight and gestational age in our study was similar to previous work.3,19,20 In our population-based study, the incidence of RDS ranged from 1.02% in infants of Hispanic mothers to 1.93% in infants of black mothers. Previously reported frequency estimates of RDS ranged from 0.33% to 0.96% in population-based studies in Sweden and England,19,20 respectively, and 1.3% in vaginally born infants in a hospital-based study in Canada.3 Low birth weight infants in our study had frequencies of RDS ranging from 13.2% in Hispanics to 18.5% in Native Americans, remarkably similar to those reported in previous stud-
ies (10 –16%).4,21 The overall incidences and associations of RDS are similar to those reported by others, which suggests that our findings are a valid representation of the general population. In this study, prolonged rupture of membranes greater than 12 hours was consistently associated with an increased risk for RDS in infants delivered vaginally and by cesarean. The association of prolonged rupture of membranes with RDS was attenuated after adjustment for preterm delivery or infant LBW, but continued to be a significant risk factor for RDS even when preterm delivery and infant LBW were included in the analysis (data not shown). That finding is in marked contrast with at least one hospital-based study that reported a significant decrease in risk for RDS in mothers who had prolonged rupture of membranes for greater than 24 hours and no signs of infection.9 A possible explanation for that difference might be infection in some of the mothers in our study. The association of prolonged rupture of membranes with maternal infection in our study was not known. Prolonged rupture of membranes should be considered with gestational age, infant birth weight, and mode of delivery when assessing the risk of RDS in a newborn. Our study is further documentation that blacks are at high risk for suboptimal perinatal outcomes and that intergenerational effects are partially responsible because of the inverse association of maternal birth weight
Table 6. Multifactorial Relative Risks for Respiratory Distress Syndrome Among Infants Delivered Vaginally* RR (95% confidence interval) Risk factor
White
Black
Native American
Hispanic
Maternal LBW Infant LBW Premature infant Smoked Late prenatal care
1.5 (0.9, 2.6) 8.5 (5.1, 14.4) 9.3 (5.4, 15.9) 0.6 (0.4, 0.9) 0.9 (0.5, 1.6)
2.4 (1.1, 5.1) 21.6 (7.7, 60.5) 5.7 (2.4, 13.6) 0.6 (0.3, 1.4) 0.9 (0.4, 1.8)
1.1 (0.4, 3.1) 18.8 (8.8, 40.1) 20.1 (7.2, 56.6) 0.7 (0.4, 1.2) 0.5 (0.2, 1.0)
1.0 (0.3, 3.6) 11.8 (4.4, 31.5) 23.3 (5.8, 92.8) 1.2 (0.5, 3.2) 1.2 (0.6, 2.5)
Abbreviations as in Table 4. * Each factor is adjusted for all factors in the model.
178 Strandjord et al
Maternal Birth Weight
Obstetrics & Gynecology
with infant LBW and preterm birth. Maternal birth weight distribution of blacks was displaced downward compared with the other three ethnic groups in this study,15 which is expected because the birth weight distribution of black infants has been similarly displaced downward for as long as it has been recorded. Therefore, improvement in perinatal outcomes for that ethnic group will require affecting the improvement of black birth weight distribution on a long-term basis, which so far has defied solution. It is suggested that historical factors be considered and addressed in the efforts to improve black birth weight.15 Human prenatal and postnatal growth is strongly influenced by maternal childhood and adult socioeconomic circumstances, so improvement of socioeconomic conditions of black families needs to be addressed on a long-term basis.11
11.
12.
13.
14.
15.
16.
References 1. Jobe AH. Preterm factors influencing surfactant deficiency. Int J Technol Assess Health Care 1991;7 Suppl:16 –20. 2. Guyer B, Martin JA, MacDorman MF, Anderson RN, Strobino DM. Annual summary of vital statistics—1996. Pediatrics 1997;100:905– 18. 3. Usher RH, Allen AC, McLean FH. Risk of respiratory distress syndrome related to gestational age, route of delivery, and maternal diabetes. Am J Obstet Gynecol 1971;111:826 –32. 4. Miller HC, Futrakul P. Birth weight, gestational age, and sex as determining factors in the incidence of respiratory distress syndrome of prematurely born infants. J Pediatr 1968;72:628 –35. 5. Morrison JJ, Rennie JM, Milton PJ. Neonatal respiratory morbidity and mode of delivery at term: Influence of timing of elective caesarean section. Br J Obstet Gynaecol 1995;102:101– 6. 6. Krantz ME, Wennergren M, Bengston LGW, Hjalmarson O, Karlsson K, Sellgren U. Epidemiological analysis of the increased risk of disturbed neonatal respiratory adaptation after caesarean section. Acta Paediatrica Scand 1986;75:832–9. 7. Hales KA, Morgan MA, Thurnau GR. Influence of labor and route of delivery on the frequency of respiratory morbidity in term neonates. Int J Gynaecol Obstet 1993;43:35– 40. 8. Tubman TR, Rollins MD, Patterson C, Halliday HL. Increased incidence of respiratory distress syndrome in babies of hypertensive mothers. Arch Dis Child 1991;66:52– 4. 9. Bryan H, Hawrylyshyn P, Hogg JS, Inwood S, Finley A, D’Costa M, et al. Perinatal factors associated with the respiratory distress syndrome. Am J Obstet Gynecol 1990;162:476 – 81. 10. Kavvadia V, Greenough A, Dimitriou G, Hooper R. Influence of
VOL. 95, NO. 2, FEBRUARY 2000
17. 18. 19. 20. 21.
ethnic origin on respiratory distress syndrome in very premature infants. Arch Dis Child Fetal Neonatal Ed 1998;78:F25– 8. Emanuel I. Invited commentary: An assessment of maternal intergenerational factors in pregnancy outcome. Am J Epidemiol 1997; 146:820 –5. Hackman E, Emanuel I, van Belle G, Daling J. Maternal birth weight and subsequent pregnancy outcome. JAMA 1983;250: 2016 –9. Stanley FJ, Alberman ED. Infants of very low birthweight. II: Perinatal factors in and conditions associated with respiratory distress syndrome. Dev Med Child Neurol 1978;20:313–22. Parrish KM, Holt VL, Connell FA, Williams B, LoGerfo JP. Variations in the accuracy of obstetric procedures and diagnoses on birth records in Washington State, 1989. Am J Epidemiol 1993;138: 119 –27. Emanuel I, Leisenring W, Williams M, Kimpo C, Estee S, O’Brien W, et al. The Washington State intergenerational study of birth outcomes: Methodology and some comparisons of maternal birthweight and infant birthweight and gestation in four ethnic groups. Paediatr Perinat Epidemiol 1999;13:352–71. Commission on Professional and Hospital Activities. The International Classification of Diseases. 9th Revision, Clinical Modification. 1989. Armitage P, Berry G. Statistical methods in medical research. 2nd ed. Oxford: Blackwell Scientific, 1987. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13–22. Hjalmarson O. Epidemiology and classification of acute, neonatal respiratory disorders. Acta Paediatrica Scand 1981;70:773– 83. Field DJ, Milner AD, Hopkin IE, Madeley RJ. Changing patterns in neonatal respiratory diseases. Pediatr Pulmonol 1987;3:231–5. Gairdner D. Recent advances in pediatrics. 3rd ed. London, United Kingdom: Churchill, 1965.
Address reprint requests to:
Thomas P. Strandjord, MD Department of Pediatrics University of Washington Box 356320 Seattle, WA 98195-6320 E-mail: [email protected]
Received March 30, 1999. Received in revised form July 28, 1999. Accepted August 12, 1999. Copyright © 2000 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc.
Strandjord et al
Maternal Birth Weight
179
continues to be an important cause of infant morbidity and mortality. In 1996, RDS was the fourth leading cause of infant death in the United States, accounting for 4.8% of infant deaths.2 Knowledge of factors that predispose infants to immature lungs is essential for prevention. Factors previously shown to influence risk of RDS include prematurity,3,4 mode of delivery,5–7 maternal hypertension,8 prolonged rupture of amniotic membranes,9 and ethnicity.10 Maternal birth weight was found to be inversely related to several adverse pregnancy outcomes including low birth weight (LBW) and preterm birth, which are related to RDS, intrauterine growth retardation, and perinatal and infant mortality.11 A hospital-based study found that infants with RDS had significantly lower maternal birth weights than other infants.12 Most epidemiologic studies of RDS are hospital based and might lack generalizability. Relatively few population-based epidemiologic studies of RDS have been done, and few have considered more than one or two factors at a time.6,13 We report a study of RDS using a large population-based intergenerational database of births to women in four ethnic groups, whites, blacks, Native Americans, and Hispanics. We also studied the association of maternal birth weight with the subsequent risk for the development of RDS.
Despite dramatic progress in understanding the etiology of infant respiratory distress syndrome (RDS),1 it
Methods
From the Department of Pediatrics, School of Medicine, Department of Epidemiology, School of Public Health, Center on Human Development and Disability, and Maternal and Child Health Program, University of Washington; and the Fred Hutchinson Cancer Research Center, Seattle, Washington. Supported by grant MCJ-530837 from the Maternal and Child Health Bureau (Title V, Social Security Act), Health Resources and Services Administration, US Department of Health and Human Services.
174 0029-7844/00/$20.00 PII S0029-7844(99)00532-3
Since 1987, all nonfederal short-stay hospitals have been legally required to submit a discharge summary of all admissions. The Washington State Department of Health links the summaries of obstetric and neonatal admissions to all data from the live birth, infant death, and fetal death certificates, to form the Birth Event Records Database. Compared with a series of hospital records the file more accurately reports medical infor-
Obstetrics & Gynecology
mation than the birth certificates or hospital discharge summaries separately.14 A random number generator within the SAS for Windows statistical package (SAS Institute Inc., Cary, NC) was used to generate a 10% random sample, stratified by birth year, of all infants born to white mothers from 1987 to 1995. A full sample of births to black, Native American, and Hispanic mothers was made over the same period. We used a deterministic methodology to link the files to birth certificates of women identified on the infant vital records as born in Washington state. A success rate of 88.8% resulted in linking 46,246 births to birth certificates of 38,513 women born between 1949, when birth weight was first recorded, and 1979. Complete details of the database, the Washington State Intergenerational Study of Birth Outcomes, have been reported.15 We also linked the driver’s license database to obtain selfreported heights of the women. We compared selfreported heights from the driver’s license database to the measured heights of 463 women. The mean difference was 0.1 in. and the correlation was .95. All mothers and infants were singletons. Implausible combinations of birth weight and gestational age were detected and removed as described.15 A total of 1.24% of infant gestations were considered outliers. Infant birth weights ranged from 255 to 6350 g, and gestational ages ranged from 20 to 46 weeks. Maternal birth weights ranged from 765 to 6067 g. Diagnostic data in International Classification of Diseases, 9th Revision, codes16 from discharge summaries were combined with diagnostic data from vital records to yield the primary outcome variable, RDS, and other medical variables. The diagnosis of RDS was accepted if it appeared on the discharge summary, birth certificate, or both. Discharge summaries recorded 84.4% (n ⫽ 489) of RDS cases, compared with 32.6% (n ⫽ 189) on birth certificates, with 17.1% (n ⫽ 99) from both sources. Statistical analysis was done with Stata Release 5 software (Stata Corp., College Station, TX). Generalized linear models with log link functions and Poisson error structures were fit to the binary end point RDS.17 That model obtained unadjusted and adjusted relative risk (RR) estimates and accompanying confidence intervals (CIs). Individual women occasionally had more than one infant during the study, so the method of generalized estimating equations was used to adjust variance estimates to account for correlation between infants born to the same woman.18 Factors included in multivariate models were mode of delivery (vaginal or cesarean), infant LBW (under 2500 g), preterm birth (under 37 weeks’ gestation), maternal LBW (under 2500 g), maternal age (under 20 years, or 20 or more years), parity (0, 1, or more), type of health insurance (Medicaid or other), marital status (married or single),
VOL. 95, NO. 2, FEBRUARY 2000
smoked during pregnancy (yes or no), prolonged rupture of membranes (at least 12 hours or less than 12 hours), maternal hypertension (normotensive, or pregnancy induced or chronic hypertension), established diabetes (yes or no), gestational diabetes (yes or no), maternal height (inches), late prenatal care (first visit before 4 months, or first visit after 4 months or no prenatal care). Maternal education only became available in 1992 so models with that variable markedly reduced the sample and were not used.
Results Table 1 presents the characteristics of successfully linked mothers and infants of all ethnic groups. White mothers were at least sociodemographic risk, whereas black mothers were at highest risk. The risk for RDS was inversely related to infant birth weight and gestational age (Tables 2 and 3). Over 40% of infants in all ethnic groups with birth weights under 1250 g were affected by RDS (Table 2). The incidence of RDS did not appear to be associated with birth weight in infants over 2500 g, so all subsequent analyses dichotomized infant birth weight to under 2500 g and at least 2500 g. Over half of infants of less than 28 weeks’ gestation had RDS, and incidence of RDS did not appear to be associated with gestational age in infants of greater than 37 weeks’ gestation (Table 3), so all subsequent analyses dichotomized infant gestation to under 37 weeks and at least 37 weeks. Frequency of RDS also was related to race. The overall frequency of RDS was 1.2% in singleton infants of white mothers, 1.9% in blacks, 1.3% in Native Americans, and 1.0% in Hispanics. Those rates were not significantly different from infants whose records could not be linked to mothers’ birth certificates (1.0% in whites, 2.0% in blacks, 1.2% in Native Americans, and 1.3% in Hispanics). Compared with white infants, the RR for RDS was 1.6 (95% CI 1.3, 2.0) for blacks, 1.1 (95% CI 0.9, 1.4) for Native Americans, and 0.9 (95% CI 0.7, 1.1) for Hispanics. In LBW infants the frequency of RDS was 16.1% in whites, 13.8% in blacks, 18.5% in Native Americans, and 13.2% in Hispanics. White infants delivered by cesarean were 3.7 times (95% CI 2.9, 4.7) more likely to develop RDS compared with infants delivered vaginally. The corresponding RR (95% CI) for the other ethnic groups were blacks 3.3 (2.4, 4.6), Native Americans 3.9 (2.6, 5.7), and Hispanics 3.2 (2.0, 5.2). Infants delivered by cesarean were more likely to be LBW. For white mothers 6.5% of infants delivered by cesarean were LBW compared with 3.5% of those delivered vaginally, for blacks 15.3% versus 10.2%, for Native Americans 9.3% versus 4.5%, and for Hispanics 7.2% versus 4.9%. A less pronounced trend was seen between gestational age and mode of delivery. For
Strandjord et al
Maternal Birth Weight
175
Table 1. Maternal and Infant Characteristics* Characteristic
White†
Black†
Native American†
Hispanic†
Maternal age (mean) Maternal age ⬍20 y (%) Maternal height (in.) Primigravida (%) Maternal LBW (%) Unmarried (%) Medicaid (%) Smoked during pregnancy (%) Late prenatal care (%) Prolonged rupture of membranes (%) Cesarean delivery Maternal hypertension Established diabetes (%) Gestational diabetes (%) Infant LBW (%) Preterm birth (%)
26.5 (23,177) 11.5 (23,177) 65.0 (17,324) 43.9 (22,940) 5.4 (22,381) 24.3 (23,154) 30.3 (20,889) 24.0 (21,842) 16.1 (22,378) 6.6 (23,187) 19.4 (23,187) 5.1 (23,187) 0.5 (22,381) 2.7 (23,187) 4.1 (23,150) 6.9 (20,329)
23.6 (7128) 27.8 (7128) 64.8 (3978) 44.4 (7058) 11.1 (5138) 70.9 (7122) 63.2 (4934) 27.4 (6133) 32.4 (6398) 9.0 (7133) 22.6 (7133) 6.4 (7133) 0.5 (5138) 2.4 (7133) 11.3 (7109) 13.3 (5937)
24.0 (7883) 24.5 (7883) 63.8 (4118) 35.3 (7833) 6.1 (5354) 65.0 (7866) 62.6 (5181) 34.2 (7395) 34.5 (7474) 6.0 (7884) 15.9 (7884) 4.3 (7884) 0.4 (5354) 2.5 (7884) 5.3 (7869) 11.8 (6611)
22.9 (6789) 31.5 (6789) 63.1 (3957) 45.7 (6745) 5.8 (4956) 50.1 (6781) 64.6 (4779) 13.7 (6457) 29.8 (6491) 5.5 (6790) 18.1 (6790) 5.2 (6790) 0.7 (4956) 3.0 (6790) 5.3 (6780) 10.4 (5747)
LBW ⫽ low birth weight. * Live-born singleton infants of singleton mothers. † Number of subjects for each variable in parentheses.
white mothers, 9.2% of infants delivered by cesarean were preterm compared with 6.3% delivered vaginally, for blacks 14.8% versus 12.8%, for Native Americans 12.9% versus 11.6%, and for Hispanics 11.1% versus 10.2%. Despite association of cesarean delivery with LBW and preterm birth, when those three factors were considered together in a multiple Poisson regression for risk of RDS, cesarean delivery was still a strong risk factor. The RR (95% CI) for RDS in infants delivered by cesarean delivery, taking into account the effects of LBW and preterm birth was 2.5 (2.0, 3.3) for whites, 2.2 (1.5, 3.1) for blacks, 2.2 (1.5, 3.4) for Native Americans, and 2.4 (1.5, 4.1) for Hispanics. Given the potent effect of delivery mode on risk of RDS, we evaluated maternal birth weight effects separately in infants delivered by cesarean and vaginally. In infants delivered by cesar-
ean, maternal LBW was not significantly related to RDS in any ethnic group (Table 4), which justified further separate analyses for infants delivered by the two modes (Tables 4 and 5). In addition to maternal birth weight, the strongest risk factors for RDS were markedly attenuated in infants delivered by cesarean, infant LBW, preterm birth, and prolonged rupture of membranes. Established diabetes was a strong risk factor in vaginally delivered Native Americans and Hispanics, and statistically insignificant in whites. Insufficient numbers in some ethnic groups prevented complete comparisons. White and black infants delivered vaginally were at increased risk of RDS if their mothers LBW: RR 2.6 (95% CI 1.6, 4.2) and 3.3 (95% CI 1.9, 5.6), respectively. Maternal LBW was associated with an almost doubled
Table 2. Incidence of Respiratory Distress Syndrome by Birth Weight* % RDS (n) Birth weight (g) ⬍1000 1000 –1249 1250 –1499 1500 –1749 1750 –1999 2000 –2249 2250 –2499 2500 –2749 2750 –2999 3000 –3249 3250 –3499 ⱖ3500 g Total
White
Black
Native American
Hispanic
54.6% (55) 46.7% (30) 60.5% (38) 24.7% (81) 19.6% (97) 11.5% (192) 5.3% (454) 3.5% (877) 1.3% (1995) 0.6% (3378) 0.4% (4591) 0.3% (11,362) 1.2% (23,150)
63.1% (65) 42.1% (38) 31.6% (38) 19.7% (76) 15.0% (100) 5.1% (158) 1.5% (334) 1.0% (590) 0.5% (981) 0.4% (1330) 0.4% (1397) 0.3% (2026) 1.9% (7109)
56.7% (30) 41.2% (17) 52.2% (23) 50.0% (38) 16.7% (42) 8.0% (75) 4.7% (191) 1.6% (314) 0.8% (671) 0.4% (1164) 0.4% (1610) 0.1% (3740) 1.3% (7869)
43.3% (30) 63.6% (11) 31.3% (16) 28.0% (25) 17.1% (35) 7.4% (81) 2.4% (165) 2.0% (351) 1.0% (726) 0.0% (1220) 0.1% (1397) 0.2% (2765) 1.0% (6780)
RDS ⫽ respiratory distress syndrome. * Includes only infants with known birth weights.
176 Strandjord et al
Maternal Birth Weight
Obstetrics & Gynecology
Table 3. Incidence of Respiratory Distress Syndrome by Gestational Age* % RDS (n) Gestational age (wk) ⬍28 28 –29 30 –31 32–33 34 –35 36 –37 38 –39 ⱖ40 Total
White
Black
Native American
Hispanic
51.1% (45) 61.8% (34) 37.0% (54) 18.0% (128) 10.8% (463) 2.0% (1980) 0.4% (7086) 0.2% (10,539) 1.1% (20,329)
52.5% (59) 63.2% (19) 37.5% (48) 11.9% (84) 2.74% (292) 0.9% (781) 0.6% (2066) 0.3% (2588) 1.8% (5937)
68.2% (22) 50.0% (20) 33.3% (27) 12.5% (88) 6.1% (312) 0.5% (927) 0.2% (2450) 0.2% (2765) 1.2% (6611)
53.3% (30) 33.3% (15) 30.0% (20) 7.8% (64) 4.5% (246) 0.9% (673) 0.3% (2040) 0.0% (2659) 0.9% (5747)
Abbreviation as in Table 2. * Includes only infants with known gestational ages.
risk of RDS (RR 1.9, 95% CI 0.7, 5.4) among Hispanics, although not statistically significant (Table 5). Maternal LBW was not associated with RDS among Native Americans. Several risk factors were considered confounders of the association of maternal LBW with RDS because each altered the RR of maternal LBW by 10% or more in each of the ethnic groups, infant LBW, preterm birth, maternal smoking, and late or no prenatal care. Full models with multiple Poisson regression for the effects of those adjusting variables on maternal LBW were constructed for each ethnic group, delivered vaginally (Table 6) or by cesarean. For infants delivered by cesarean, the addition of covariates did not alter the relation of maternal birth weight and RDS (data not shown). In Native American and Hispanic infants delivered vaginally, addition of covariates also failed to make maternal LBW a significant risk factor for RDS. With adjustment, maternal LBW was no longer significantly associated with RDS in whites delivered vaginally, RR 1.5 (95% CI 0.9, 2.6). In blacks, with adjustment the association was reduced but remained statistically significant, RR 2.4 (95% CI 1.1, 5.1, P ⫽ .024). In all ethnic
groups infant LBW was strongest in reducing the association of maternal LBW when infants were delivered vaginally. Maternal adult height was not related to RDS in infants.
Discussion There was a pattern of increased risk among vaginally delivered infants, but it was statistically significant only in whites and blacks. Adjustment, particularly for infant LBW and prematurity, reduced the association, which remained statistically significant for blacks, indicating that one pathway by which maternal LBW related to RDS was through its effect on infant birth weight or gestation. After adjustment, statistical significance persisted in black infants, indicating there might be other mechanisms besides effect of maternal birth weight on infant size which might explain the association of maternal birth weight with RDS. Cesarean delivery was a potent risk factor for RDS in previous studies.3,5–7,13 In this study, cesarean delivery was associated with a higher risk for RDS in all ethnic groups. The association of cesarean delivery with RDS
Table 4. Relative Risk for Respiratory Distress Syndrome Among Infants Delivered by Cesarean* RR (95% confidence interval) Risk factor
White
Black
Native American
Hispanic
Maternal LBW Infant LBW Preterm birth Medicaid Primigravida Prolonged rupture of membranes Maternal hypertension Gestational diabetes Late prenatal care
0.4 (0.1, 1.2) 19.2 (13.8, 26.8) 24.7 (16.6, 36.9) 1.6 (1.1, 2.3) 0.7 (0.5, 1.1) 3.6 (2.4, 5.3) 1.4 (0.8, 2.4) 1.5 (0.7, 2.9) 1.7 (1.1, 2.6)
1.1 (0.6, 2.3) 25.8 (14.1, 47.3) 14.2 (8.0, 25.4) 1.3 (0.8, 2.2) 1.3 (0.6, 1.6) 3.80 (2.3, 6.2) 2.1 (1.2, 3.8)
2.0 (0.8, 4.8) 31.2 (15.8, 61.5) 37.9 (14.9, 96.2) 1.2 (0.6, 2.1) 0.6 (0.3, 1.2) 2.95 (1.5, 6.0) 1.9 (0.8, 4.4) 1.1 (0.3, 4.3) 1.5 (0.8, 2.8)
1.5 (0.5, 4.7) 21.2 (10.4, 43.3) 27.3 (10.3, 72.5) 1.8 (0.8, 4.1) 0.4 (0.2, 0.9) 1.2 (0.4, 4.0) 3.1 (1.5, 6.9) 3.2 (1.1, 8.9) 1.3 (0.6, 2.9)
†
0.9 (0.5, 1.6)
RR ⫽ relative risk; LBW ⫽ low birth weight. * Each factor is considered individually; only factors with a significant RR in at least one of the ethnic groups are shown. † Missing cells due to insufficient cases.
VOL. 95, NO. 2, FEBRUARY 2000
Strandjord et al
Maternal Birth Weight
177
Table 5. Relative Risk for Respiratory Distress Syndrome Among Infants Delivered Vaginally* RR (95% confidence interval) Risk factor
White
Black
Native American
Hispanic
Maternal LBW Infant LBW Premature infant Unmarried Prolonged rupture of membranes Established diabetes Teenage mother
2.6 (1.6, 4.2) 33.6 (24.4, 46.2) 25.6 (17.7, 37.1) 0.6 (0.4, 0.9) 6.1 (4.3, 8.8) 1.9 (0.3, 13.3) 1.0 (0.6, 1.7)
3.3 (1.9, 5.6) 35.3 (20.1, 62.1) 27.2 (14.6, 52.4) 1.3 (0.7, 2.2) 7.9 (4.9, 12.6)
1.1 (0.4, 3.0) 68.4 (38.0, 123.0) 77.9 (28.0, 216.4) 1.5 (0.9, 2.8) 8.6 (5.1, 14.5) 11.7 (3.0, 44.9) 2.5 (1.5, 4.1)
1.9 (0.7, 5.4) 52.8 (26.6, 104.6) 85.0 (25.9, 278.7) 3.4 (1.5, 7.5) 10.2 (5.5, 19.0) 9.9 (1.5, 65.2) 1.9 (1.0, 3.5)
†
1.3 (0.8, 2.1)
Abbreviations as in Table 4. * Each factor is considered individually; only factors with a significant RR in at least one of the ethnic groups are shown. † Missing cells due to insufficient cases.
persisted even when infant birth weights and gestational ages were included in a multiple regression model with mode of delivery. Also, cesarean delivery attenuated associations with most other risk factors for RDS, including maternal LBW. There was an association of maternal LBW with cesarean delivery; 21.7% of LBW mothers were delivered by cesarean compared with 18.8% of normal birth weight mothers (P ⬍ .001). The association of cesarean delivery with maternal LBW might partially confound the association of maternal LBW with RDS. The strong association of mode of delivery with risk of RDS suggests that in future investigations of RDS, separate analyses should be made for infants delivered by each mode. Diagnoses of RDS were made at multiple institutions by many observers, so it is reassuring that the association of RDS with infant birth weight and gestational age in our study was similar to previous work.3,19,20 In our population-based study, the incidence of RDS ranged from 1.02% in infants of Hispanic mothers to 1.93% in infants of black mothers. Previously reported frequency estimates of RDS ranged from 0.33% to 0.96% in population-based studies in Sweden and England,19,20 respectively, and 1.3% in vaginally born infants in a hospital-based study in Canada.3 Low birth weight infants in our study had frequencies of RDS ranging from 13.2% in Hispanics to 18.5% in Native Americans, remarkably similar to those reported in previous stud-
ies (10 –16%).4,21 The overall incidences and associations of RDS are similar to those reported by others, which suggests that our findings are a valid representation of the general population. In this study, prolonged rupture of membranes greater than 12 hours was consistently associated with an increased risk for RDS in infants delivered vaginally and by cesarean. The association of prolonged rupture of membranes with RDS was attenuated after adjustment for preterm delivery or infant LBW, but continued to be a significant risk factor for RDS even when preterm delivery and infant LBW were included in the analysis (data not shown). That finding is in marked contrast with at least one hospital-based study that reported a significant decrease in risk for RDS in mothers who had prolonged rupture of membranes for greater than 24 hours and no signs of infection.9 A possible explanation for that difference might be infection in some of the mothers in our study. The association of prolonged rupture of membranes with maternal infection in our study was not known. Prolonged rupture of membranes should be considered with gestational age, infant birth weight, and mode of delivery when assessing the risk of RDS in a newborn. Our study is further documentation that blacks are at high risk for suboptimal perinatal outcomes and that intergenerational effects are partially responsible because of the inverse association of maternal birth weight
Table 6. Multifactorial Relative Risks for Respiratory Distress Syndrome Among Infants Delivered Vaginally* RR (95% confidence interval) Risk factor
White
Black
Native American
Hispanic
Maternal LBW Infant LBW Premature infant Smoked Late prenatal care
1.5 (0.9, 2.6) 8.5 (5.1, 14.4) 9.3 (5.4, 15.9) 0.6 (0.4, 0.9) 0.9 (0.5, 1.6)
2.4 (1.1, 5.1) 21.6 (7.7, 60.5) 5.7 (2.4, 13.6) 0.6 (0.3, 1.4) 0.9 (0.4, 1.8)
1.1 (0.4, 3.1) 18.8 (8.8, 40.1) 20.1 (7.2, 56.6) 0.7 (0.4, 1.2) 0.5 (0.2, 1.0)
1.0 (0.3, 3.6) 11.8 (4.4, 31.5) 23.3 (5.8, 92.8) 1.2 (0.5, 3.2) 1.2 (0.6, 2.5)
Abbreviations as in Table 4. * Each factor is adjusted for all factors in the model.
178 Strandjord et al
Maternal Birth Weight
Obstetrics & Gynecology
with infant LBW and preterm birth. Maternal birth weight distribution of blacks was displaced downward compared with the other three ethnic groups in this study,15 which is expected because the birth weight distribution of black infants has been similarly displaced downward for as long as it has been recorded. Therefore, improvement in perinatal outcomes for that ethnic group will require affecting the improvement of black birth weight distribution on a long-term basis, which so far has defied solution. It is suggested that historical factors be considered and addressed in the efforts to improve black birth weight.15 Human prenatal and postnatal growth is strongly influenced by maternal childhood and adult socioeconomic circumstances, so improvement of socioeconomic conditions of black families needs to be addressed on a long-term basis.11
11.
12.
13.
14.
15.
16.
References 1. Jobe AH. Preterm factors influencing surfactant deficiency. Int J Technol Assess Health Care 1991;7 Suppl:16 –20. 2. Guyer B, Martin JA, MacDorman MF, Anderson RN, Strobino DM. Annual summary of vital statistics—1996. Pediatrics 1997;100:905– 18. 3. Usher RH, Allen AC, McLean FH. Risk of respiratory distress syndrome related to gestational age, route of delivery, and maternal diabetes. Am J Obstet Gynecol 1971;111:826 –32. 4. Miller HC, Futrakul P. Birth weight, gestational age, and sex as determining factors in the incidence of respiratory distress syndrome of prematurely born infants. J Pediatr 1968;72:628 –35. 5. Morrison JJ, Rennie JM, Milton PJ. Neonatal respiratory morbidity and mode of delivery at term: Influence of timing of elective caesarean section. Br J Obstet Gynaecol 1995;102:101– 6. 6. Krantz ME, Wennergren M, Bengston LGW, Hjalmarson O, Karlsson K, Sellgren U. Epidemiological analysis of the increased risk of disturbed neonatal respiratory adaptation after caesarean section. Acta Paediatrica Scand 1986;75:832–9. 7. Hales KA, Morgan MA, Thurnau GR. Influence of labor and route of delivery on the frequency of respiratory morbidity in term neonates. Int J Gynaecol Obstet 1993;43:35– 40. 8. Tubman TR, Rollins MD, Patterson C, Halliday HL. Increased incidence of respiratory distress syndrome in babies of hypertensive mothers. Arch Dis Child 1991;66:52– 4. 9. Bryan H, Hawrylyshyn P, Hogg JS, Inwood S, Finley A, D’Costa M, et al. Perinatal factors associated with the respiratory distress syndrome. Am J Obstet Gynecol 1990;162:476 – 81. 10. Kavvadia V, Greenough A, Dimitriou G, Hooper R. Influence of
VOL. 95, NO. 2, FEBRUARY 2000
17. 18. 19. 20. 21.
ethnic origin on respiratory distress syndrome in very premature infants. Arch Dis Child Fetal Neonatal Ed 1998;78:F25– 8. Emanuel I. Invited commentary: An assessment of maternal intergenerational factors in pregnancy outcome. Am J Epidemiol 1997; 146:820 –5. Hackman E, Emanuel I, van Belle G, Daling J. Maternal birth weight and subsequent pregnancy outcome. JAMA 1983;250: 2016 –9. Stanley FJ, Alberman ED. Infants of very low birthweight. II: Perinatal factors in and conditions associated with respiratory distress syndrome. Dev Med Child Neurol 1978;20:313–22. Parrish KM, Holt VL, Connell FA, Williams B, LoGerfo JP. Variations in the accuracy of obstetric procedures and diagnoses on birth records in Washington State, 1989. Am J Epidemiol 1993;138: 119 –27. Emanuel I, Leisenring W, Williams M, Kimpo C, Estee S, O’Brien W, et al. The Washington State intergenerational study of birth outcomes: Methodology and some comparisons of maternal birthweight and infant birthweight and gestation in four ethnic groups. Paediatr Perinat Epidemiol 1999;13:352–71. Commission on Professional and Hospital Activities. The International Classification of Diseases. 9th Revision, Clinical Modification. 1989. Armitage P, Berry G. Statistical methods in medical research. 2nd ed. Oxford: Blackwell Scientific, 1987. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13–22. Hjalmarson O. Epidemiology and classification of acute, neonatal respiratory disorders. Acta Paediatrica Scand 1981;70:773– 83. Field DJ, Milner AD, Hopkin IE, Madeley RJ. Changing patterns in neonatal respiratory diseases. Pediatr Pulmonol 1987;3:231–5. Gairdner D. Recent advances in pediatrics. 3rd ed. London, United Kingdom: Churchill, 1965.
Address reprint requests to:
Thomas P. Strandjord, MD Department of Pediatrics University of Washington Box 356320 Seattle, WA 98195-6320 E-mail: [email protected]
Received March 30, 1999. Received in revised form July 28, 1999. Accepted August 12, 1999. Copyright © 2000 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc.
Strandjord et al
Maternal Birth Weight
179