Morisaki, Naho MD, MPH; Esplin, M. Sean MD; Varner, Michael W. MD; Henry, Erick MPH; Oken, Emily MD, MPH
Birth weights and fetal growth decreased from 2000 to 2008 despite a successful initiative to minimize elective early-term deliveries and to maintain gestational length.
From the mid to late 20th century, average birth weight increased in many countries, including the United States.1 However, more recent data suggest a reversal in this trend.2 In the United States, mean birth weight among term (37–41 weeks of gestation) singleton newborns decreased from 3,441 to 3,389 g from 1990 to 2005, and the entire distribution of birth weight shifted to the left.3,4 Similar decreases in birth weight have been seen in many other countries.5–10
Both gestational age at birth and fetal growth contribute to birth weight. Over the past decade, mean gestational age among term births in the United States has decreased by 3 days,11,12 and changes in obstetric practice that have permitted delivery “on demand” via scheduled induction of labor or cesarean delivery are likely to have played a substantial role.13–16 The extent to which fetal growth trends are contributing to the decline in birth weight is less clear.
Two recent studies—both using U.S. birth certificate data but with different analytic approaches—reached different conclusions. Zhang et al4 found that although gestational age decreased from 1995 to 2003, fetal growth was still increasing. In contrast, Donahue et al3 conclude that gestational age and fetal growth had both declined from 1990 to 2005. Inconsistencies in these studies show that observing trends in fetal growth is difficult in a population in which gestational age, which has a large effect on birth weight, is changing as well.
One way to isolate the contribution of fetal growth to birth weight trends is to study a population in which gestational age remained stable. We therefore examined trends in birth weight and fetal growth within the Intermountain Healthcare system in Utah, where in January 2001, clinical leaders developed and implemented guidelines to discourage elective early-term deliveries.17,18 We hypothesized that despite the decrease in early elective inductions, we would observe an ongoing decrease in birth weight and fetal growth.
MATERIALS AND METHODS
We used data extracted from the electronic medical records of 21 hospitals affiliated with Intermountain Healthcare, a vertically integrated health care system in Utah and Southeast Idaho. All hospitals used the StorkBytes perinatal data program, an obstetric and delivery database application that captures maternal history, labor progression, delivery, and postpartum data.
The initiative at Intermountain Healthcare has been described in detail elsewhere.17,18 Briefly, beginning in January 2001, hospital leaders launched a quality-improvement program with the intent of limiting elective deliveries before 39 completed weeks of gestation.
We obtained information of 219,757 singleton term newborns born from July 1, 2000 through December 31, 2008, which included all data currently available in the StorkBytes system. We limited our analyses to singleton newborns born at 37–41 completed weeks of gestation to ensure our findings would not be driven by trends in multiple births and preterm deliveries, although we included the small number of postterm deliveries (n=406) in sensitivity analyses. We excluded births with missing birth weight or with birth weight inconsistent with gestational age based on the method of Alexander et al,19 or that had missing data for gestational age (n=63). Thus, we retained information for 219,694 newborns and their mothers. We performed all analyses on this population.
We based our definition of gestational age on the clinical gestational age estimate, which incorporates information from the date of the last menstrual period as well as other factors, including ultrasound dating.20 We calculated z scores, defined small for gestational age (SGA) as fetal growth less than the 10th percentile at each completed week of gestation, and defined large for gestational age (LGA) as greater than 90th percentile, according to reference data based on all U.S. births that occurred in 1999–2000.21 We also performed sensitivity analyses using the last menstrual period–based gestational age and gestational age–specific percentile references by last menstrual period, although in this study population only 0.2% (n=473) of births had discordant last menstrual period–based and clinical-based gestational ages.
For all analyses, we also categorized maternal characteristics as presented in Table 1. We calculated maternal body mass index (BMI, calculated as weight (kg)/[height (m)]2) from self-reported prepregnancy weight and height and set categories as follows: obese (BMI 30 or more), overweight (BMI 25 to less than 30), normal (BMI 18.5 to less than 25), and underweight (BMI less than 18.5). We based gestational weight gain categories on 2009 U.S. Institute of Medicine guidelines.22
We classified induction status, route of delivery, and delivery status by their indications. We categorized cesarean deliveries that could have been performed at a different time if chosen as “scheduled cesarean deliveries.” This category included all elective cesarean deliveries, cesarean deliveries after elective inductions, and cesarean deliveries with “breech position” or “repeated cesarean delivery” being the indication. Similarly, “scheduled vaginal delivery” included all vaginal deliveries initiated by elective inductions, as well as inductions with “postterm pregnancy” being the indication. We defined “spontaneous vaginal deliveries” as vaginal deliveries that did not start by induction or end in cesarean delivery. We categorized the remaining births, ie, deliveries that were neither spontaneous nor scheduled and were considered to have had an indication for urgent delivery, as “indicated cesarean deliveries” or “indicated vaginal deliveries.”
Several maternal characteristics known to contribute to fetal growth had some missing data. Maternal smoking during pregnancy was added to the database later; therefore, smoking status is missing in 38% of the data, including nearly all of 2000–2002 data and most of 2003 data. Data also were missing in small numbers of mothers for race (n=513), marital status (n=1,679), age (n=79), weight or height (n=4,225), gestational weight gain (n=4,868), and indication for cesarean delivery (n=66). Because these variables were most likely missing at random and not associated with the missing data itself, other maternal factors, or birth weight, we used multiple imputation using a Markov Chain Monte Carlo algorithm to impute missing values. We generated 10 imputed data sets and combined results from each complete data set to produce inferential results. The Harvard Pilgrim Health Care Institutional Review Board determined this study was exempt from review because the previously collected dataset did not contain any personal health information.
We calculated the distributions of maternal and newborn characteristics to observe their changes over the 8.5-year period. Next, because we expected change in indication as well as timing of birth because of the intervention, we plotted the percentages of births according to delivery route and indication for early-term, full-term, and all term newborns, as well as the distribution of gestational age in completed weeks for every 6-month period. To estimate trends in birth outcomes over time, we performed linear regression analyses for birth weight and z score, and multinomial logistic regression for SGA and LGA, using separate models for each outcome. We included each half-year of birth as a categorical variable to generate estimates for every 6-month period compared with the baseline time period of July–December 2000, and we also included month of birth as a continuous predictor to examine the overall change in each outcome over the course of the study period.
Finally, we used multivariable-adjusted regression models to account for trends in maternal and neonatal characteristics as well as delivery methods, which also evolved over the 8.5-year study period. All models were adjusted for available maternal and newborn characteristics shown in Table 1. We again fit separate models including period of birth (in 6-month increments) as a categorical exposure and also as a continuous exposure.
We performed our primary analyses using multiple imputations to address the problem of missing data, which were especially notable for smoking status. We also performed secondary analyses using four other methods to address missing data: a complete case analysis restricted to births in 2004–2008 (n=130,538); a complete case analysis for all years 2000–2008 (n=131,591); a complete case analysis excluding smoking status (n=211,780); and a complete case analysis including missing smoking status as a separate category (n=211,780). Additionally, we performed stratified analyses of early-term or full-term births, as well as among scheduled or nonscheduled deliveries.
We also performed several additional sensitivity analyses. Because the intervention could have led to an increased number of postterm deliveries, we repeated our analysis including the small number of births at 42 weeks or more (n=406). We also repeated our analysis of birth weight including gestational length in days (instead of completed weeks) as a categorical variable, and including time of birth by each month (instead of by each half-year) as a continuous variable. We performed all analyses using SAS 9.3.
In Table 1, we present information on selected maternal and neonatal characteristics for three representative time periods (the last 6 months of 2000, 2004, and 2008). Over the period, mean birth weight (3,410–3,383 g) and LGA (9.0–7.4%) both decreased, whereas SGA increased (7.5–8.2%). An increasing proportion of newborns were born to mothers with characteristics associated with lower fetal growth, such as: nonwhite race, unmarried, and had preeclampsia or other hypertensive disorders of pregnancy. However, an increasing proportion of neonates also were born to mothers with characteristics associated with higher fetal growth, such as: older age, higher parity, diabetes before or during pregnancy, higher BMI, and greater gestational weight gain. Average gestational age at birth stayed stable at approximately 39 weeks over the 8.5 years.
In Figure 1, we present trends in delivery type separately for early-term and full-term newborns. The intervention to reduce early-term elective deliveries began in January 2001.
Trends over time in ...Image Tools
The proportion of births during early term decreased from a high of 36.2% in July–December 2000 to a low of 30.1% in January–June 2003; thereafter, it remained in the range of 32–33% (Fig. 1A). Among all term births, scheduled deliveries slightly increased from 29% to 34%. Scheduled early-term births before 39 weeks of gestation decreased more than half, from 9.7% to 4.4%, whereas scheduled births after 39 weeks of gestation increased from 19% to 29%. Cesarean delivery among term births increased from 14% to 20%. Cesarean delivery performed during early term increased minimally from 5.5% to 6.9%, it but increased more markedly among full-term births from 8% to 13%, with increases in both indicated and elective cesarean deliveries. The decrease in early-term births and increase in late-term births is attributable mostly to the shift of births at 38 weeks to 39 weeks of gestation rather than any change in the proportion of births at 37, 40, or 41 weeks of gestation.
In Figure 2, we present the change in birth weight for each period of birth, adjusted for maternal and newborn characteristics, compared with births during July–December 2000. Birth weight decreased in all subgroups by duration of gestation and indication for delivery; ie, both early-term and full-term newborns as well as scheduled, spontaneous, and medically indicated urgent deliveries.
Change in birth weig...Image Tools
In Table 2, we present the estimated change in birth weight from July 2000 to December 2008, adjusted for delivery method and maternal characteristics. Over the 8.5 years, birth weight decreased by 36 g (95% confidence interval [CI] −31 to −42) and z score decreased by standard deviation of 0.08 (95% CI −0.07 to −0.09). The estimated odds ratio for SGA was 1.12 (95% CI 1.06–1.19) and for LGA was 0.77 (95% CI 0.73-0.82) for deliveries in July–December 2008 compared with delivery in July–December 2000.
In subgroup analyses, early-term newborns showed a larger decrease in birth weight (−40 g; 95% CI −30 to −49) and z score (−0.09 standard deviation; 95% CI −0.06 to −0.11) than did full-term newborns. Births with indications for urgent deliveries showed larger decrease in birth weight (−48 g; 95% CI −34 to −63) and fetal growth (−0.11 units; 95% CI −0.07 to −0.14) compared with those born via scheduled deliveries or spontaneous vaginal deliveries.
As expected, in each model, all higher levels of maternal age, parity, prepregnancy BMI, or gestational weight gain, as well as taller height, longer gestational length (within term births), and diabetes before or during pregnancy, were significantly associated with greater fetal growth. Also, nonwhite race and Hispanic ethnicity, smoking before or during pregnancy, preeclampsia, eclampsia, prepregnancy hypertension, hypertensive disorders of pregnancy, and being unmarried were all significantly associated with less fetal growth (data not shown). The observed decreases in birth weight and LGA and increase in SGA persisted in all sensitivity analyses, whether altering the population or categorization of covariates.
We observed a recent decline in birth weight among singleton term births in a hospital system that implemented a successful intervention to minimize early-term elective deliveries. Previous analyses of birth weight trends have been complicated by the fact that gestational length decreased in parallel with birth weight, which may complicate efforts to isolate trends in fetal growth. For the present analysis, we studied a population in which mean gestational age did not change. Thus, any decrease in birth weight must result from reduced fetal growth.
Gestational length is notoriously difficult to exactly determine. We repeated our analyses using both last menstrual period–based and clinical-based last menstrual period, and also by using two reference datasets based on either clinical or last menstrual period dating, for our calculations of fetal growth. All methods showed similar results. Our results suggest that newborns are becoming smaller independent of any trend in gestational length.
The cause of this decline in fetal growth remains unknown. Maternal and newborn characteristics in our dataset that are known to contribute to fetal growth did not explain the observed decline in fetal growth, because adjustment for these variables did not eliminate the observed declines. This is concordant with the fact that trends of most maternal factors have not changed direction from the 1990s, when birth weight was increasing. We also included maternal prepregnancy BMI, which was not available on U.S. birth certificates until recently and thus was not included in previous studies.1 However, obesity rates have continued to increase over recent decades, and maternal BMI is directly associated with fetal growth. We found, as expected, that adjusting for maternal BMI resulted in even stronger temporal decreases in birth weight.
Environmental, behavioral, or nutritional factors not addressed on birth certificates may be contributing to the observed decrease in fetal growth. Unmeasured factors such as environmental toxicants, including lead, mercury, and persistent organic pollutants, and psychosocial stressors may influence fetal growth.23,24 Fetal growth also has been related to maternal prenatal intake of fatty acids,25–27 such as the omega-3 fatty acids found in fish, and trans fatty acids. Even though maternal BMI and presumably total energy intake are increasing, diet quality may declining, because studies show that those who are obese have poorer micronutrient intake despite greater intake of total energy.28
Our findings that recorded maternal and newborn characteristics and trends in obstetric care did not explain the decrease in fetal growth indicates that urgent research to identify contributing factors is necessary. Although a 36-g decrease in a single newborn may seem minimal, a recent study found that a 70-g decrease in birth weight among girls born during an economic depression was associated with higher blood glucose and greater odds of obesity in adulthood.29 Furthermore, 10% increase in risk of being born SGA is a concern from public health perspectives. Numerous studies show SGA newborns have increased risks for not only neonatal morbidity and mortality30 but also obesity, coronary heart disease, stroke, hypertension, and type 2 diabetes31–34 throughout the lifespan. Without elucidating the contributors to declining fetal growth, it is difficult to understand its implications and plan to intervene.
In addition, we note that although the Intermountain Healthcare initiative dramatically reduced early-term elective deliveries, overall elective deliveries and use of cesarean delivery still increased. The American College of Obstetricians and Gynecologists recommended in 2007 that labor should be electively induced only after 39 weeks of gestation have been completed.14,35 Our study suggests that although implementation of these guidelines may reduce early-term elective deliveries, it is unlikely to reverse the increasing rates of elective deliveries in general.
There are several limitations to our study. Because we used a preexisting database and did not validate measures by chart review, there may have been misclassification or measurement error. However, we anticipate any errors would have been consistent over time. Missing data were minimal, except for smoking status. However, we found similar results using a number of approaches to account for these missing data. Our study was based on a multicenter health care system covering 23 hospitals in Utah and Southeast Idaho, and a greater proportion of mothers were white compared with many other regions. Therefore, findings may not be generalizable to populations elsewhere. However, because decrease in fetal growth was consistent over racial and ethnic subgroups in the national data analysis by Donahue et al,3 we anticipate that this observed decrease is not unique to our population.
Increasing elective inductions and subsequently shorter gestational length most likely do not fully explain the widely observed decrease in birth weight in term newborns. No commonly measured maternal or newborn factors can account for this decrease in fetal growth.
1. Kramer MS, Morin I, Yang H, Platt RW, Usher R, McNamara H, et al.. Why are babies getting bigger? Temporal trends in fetal growth and its determinants. J Pediatr 2002;141:538–42.
2. Oken E. Secular trends in birth weight. In: Gillman MW, Gluckman PD, Rosenfeld RG, editors. Recent Advances in Growth Research: Nutritional, Molecular and Endocrine Perspectives: 71st Nestle Nutrition Institute Workshop, Vienna, October 2011. Basel, Switzerland: S Karger Ag; 2011. p. 103–14.
3. Donahue SM, Kleinman KP, Gillman MW, Oken E. Trends in birth weight and gestational length among singleton term births in the United States: 1990-2005. Obstet Gynecol 2010;115:357–64.
4. Zhang X, Joseph KS, Kramer MS. Decreased term and postterm birthweight in the United States: impact of labor induction. Am J Obstet Gynecol 2010;203:124 e1–7.
5. Edwards NM, Audas RP. Trends of abnormal birthweight among full-term infants in Newfoundland and Labrador. Can J Public Health 2010;101:138–42.
6. Diouf I, Charles MA, Blondel B, Heude B, Kaminski M. Discordant time trends in maternal body size and offspring birthweight of term deliveries in France between 1972 and 2003: data from the French National Perinatal Surveys. Paediatr Perinat Epidemiol 2011;25:210–7.
7. Lahmann PH, Wills RA, Coory M. Trends in birth size and macrosomia in Queensland, Australia, from 1988 to 2005. Paediatr Perinat Epidemiol 2009;23:533–41.
8. Schack-Nielsen L, Molgaard C, Sorensen TI, Greisen G, Michaelsen KF. Secular change in size at birth from 1973 to 2003: national data from Denmark. Obesity (Silver Spring) 2006;14:1257–63.
9. Lu Y, Zhang J, Lu X, Xi W, Li Z. Secular trends of macrosomia in southeast China, 1994-2005. BMC Public Health 2010;11:818.
10. Ohmi H, Hirooka K, Hata A, Mochizuki Y. Recent trend of increase in proportion of low birthweight infants in Japan. Int J Epidemiol 2001;30:1269–71.
11. Martin JA, Kirmeyer S, Osterman M, Shepherd RA. Born a bit too early: recent trends in late preterm births. NCHS Data Brief 2009;24:1–8.
12. Davidoff MJ, Dias T, Damus K, Russell R, Bettegowda VR, Dolan S, et al.. Changes in the gestational age distribution among U.S. singleton births: impact on rates of late preterm birth, 1992 to 2002. Semin Perinatol 2006;30:8–15.
13. Murthy K, Grobman WA, Lee TA, Holl JL. Trends in induction of labor at early-term gestation. Am J Obstet Gynecol 2011;204:435 e1–6.
14. Cesarean delivery on maternal request. ACOG Committee Opinion No. 394. American College of Obstetricians and Gynecologists. Obstet Gynecol 2007;110:1501.
15. Reddy UM, Ko CW, Raju TN, Willinger M. Delivery indications at late-preterm gestations and infant mortality rates in the United States. Pediatrics 2009;124:234–40.
16. Goldenberg RL, McClure EM, Bhattacharya A, Groat TD, Stahl PJ. Women's perceptions regarding the safety of births at various gestational ages. Obstet Gynecol 2009;114:1254–8.
17. Oshiro BT, Henry E, Wilson J, Branch DW, Varner MW. Decreasing elective deliveries before 39 weeks of gestation in an integrated health care system. Obstet Gynecol 2009;113:804–11.
18. Oshiro BT, Berns SD. Quality improvement opportunities to prevent preterm births. Clin Perinatol 2011;38:565–78.
19. Alexander GR, Himes JH, Kaufman RB, Mor J, Kogan M. A United States national reference for fetal growth. Obstet Gynecol 1996;87:163–8.
20. Mathews TJ, Minino AM, Osterman MJ, Strobino DM, Guyer B. Annual summary of vital statistics: 2008. Pediatrics 2010;127:146–57.
21. Oken E, Kleinman KP, Rich-Edwards J, Gillman MW. A nearly continuous measure of birth weight for gestational age using a United States national reference. BMC Pediatr 2003;3:6.
22. Rasmussen KM, Yaktine AL, editors. Weight gain during pregnancy: reexamining the guidelines. Washington, DC: National Academies Collection; 2009.
23. Karagas MR, Choi AL, Oken E, Horvat M, Schoeny R, Kamai E, et al.. Evidence on the human health effects of low-level methylmercury exposure. Environ Health Perspect 2012;120:799–806.
24. McDiarmid MA, Gehle K. Preconception brief: occupational/environmental exposures. Matern Child Health J 2006;10:S123–8.
25. Al MD, van Houwelingen AC, Hornstra G. Long-chain polyunsaturated fatty acids, pregnancy, and pregnancy outcome. Am J Clin Nutr 2000;71:285–91S.
26. Olsen SF, Hansen HS, Sorensen TI, Jensen B, Secher NJ, Sommer S, et al.. Intake of marine fat, rich in (n-3)-polyunsaturated fatty acids, may increase birthweight by prolonging gestation. Lancet 1986;2:367–9.
27. Cohen JF, Rifas-Shiman SL, Rimm EB, Oken E, Gillman MW. Maternal trans fatty acid intake and fetal growth. Am J Clin Nutr 2011;94:1241–7.
28. Rifas-Shiman SL, Rich-Edwards JW, Kleinman KP, Oken E, Gillman MW. Dietary quality during pregnancy varies by maternal characteristics in Project Viva: a US cohort. J Am Diet Assoc 2009;109:1004–11.
29. Imai CM, Halldorsson TI, Gunnarsdottir I, Gudnason V, Aspelund T, Jonsson G, et al.. Effect of birth year on birth weight and obesity in adulthood: comparison between subjects born prior to and during the great depression in Iceland. PLoS One 2012;7:e44551.
30. McIntire DD, Bloom SL, Casey BM, Leveno KJ. Birth weight in relation to morbidity and mortality among newborn infants. N Engl J Med 1999;340:1234–8.
31. Forsén T, Eriksson JG, Tuomilehto J, Osmond C, Barker DJ. Growth in utero and during childhood among women who develop coronary heart disease: longitudinal study. BMJ 1999;319:1403–7.
32. Singhal A, Lucas A. Early origins of cardiovascular disease: is there a unifying hypothesis? Lancet 2004;363:1642–5.
33. Barker DJ. Adult consequences of fetal growth restriction. Clin Obstet Gynecol 2006;49:270–83.
34. Oken E, Gillman MW. Fetal origins of obesity. Obes Res 2003;11:496–506.
35. Induction of labor. ACOG Practice Bulletin No. 107.American College of Obstetricians and Gynecologists. Obstet Gynecol 2009;114:386–97.
Figure. No available...Image Tools