Objective
The purpose of this study was to investigate the associations between gestational weight gain (GWG) and small- and large-for-gestational-age (SGA, LGA), cesarean delivery, child overweight, and maternal postpartum weight retention in a diverse sample of women in the Unites States.
Study Design
We estimated associations between GWG (continuous and within categories defined by the Institute of Medicine), maternal prepregnancy body mass index, and each outcome in 4496 births in the National Longitudinal Survey of Youth 1979, which was a prospective cohort.
Results
GWG (kilograms) was associated with decreased risk of SGA and increased risk of LGA, cesarean delivery, postpartum weight retention, and child overweight independent of maternal demographic and pregnancy characteristics. Gain above the Institute of Medicine guidelines was associated with decreased risk of SGA and increased risk of all other outcomes.
Conclusion
Excessive gain may have long-term consequences for maternal and child body size, but the benefits of lower gain must be balanced against risk of SGA.
In recent decades, maternal gestational weight gain (GWG) during pregnancy has become a highly controversial topic in the United States. The difficulty in balancing risks that are associated with inadequate gain (eg, small-for-gestational-age [SGA] birth) against risks that are associated with excessive gain (eg, large-for-gestational-age [LGA] birth and cesarean delivery) and growing concerns about the 40% of women who enter pregnancy overweight or obese are reflected in the changing guidelines that were released by the Institute of Medicine (IOM) in 1970, 1990, and 2009. Weight gain recommendations are now specific to prepregnancy body mass index (BMI) and include a range of “adequate” gain. For example, the 2009 guidelines recommend gain of 11.5-16 kg for normal weight women and 5-9 kg for obese women.
The 2009 IOM report discusses evidence that GWG may also affect long-term maternal and child health, which includes maternal postpartum weight retention and adiposity in childhood. Animal research suggests that maternal nutrition during pregnancy may affect fetal growth, development of adipose tissue, or hypothalamic control of appetite or metabolism in the offspring.
Most existing studies examine the impact of GWG on birth outcomes, postpartum weight retention, or child overweight in isolation. Two recent studies have examined both short- and long-term maternal and child outcomes with the use of data from relatively homogenous populations. Research is now needed that examines the impact of GWG on both short- and long-term maternal and child outcomes in a diverse sample of women from across the United States. The recent release of the new IOM guidelines also provides an opportunity to examine the impact of gain within vs outside the recommended ranges on birth and later-life outcomes.
In this study, we used data from a diverse sample of US women to examine the associations between GWG and SGA, LGA, cesarean delivery, childhood overweight, and maternal postpartum weight retention within maternal prepregnancy BMI categories. We also examined the associations between these outcomes and GWG below and above (compared with within) the recently revised 2009 IOM recommended ranges.
Methods
We used data from the National Longitudinal Survey of Youth 1979 (NLSY79) and the NLSY79 Children and Young Adult Survey that were weighted to account for nonresponse bias and oversampling of black and Hispanic women and below-poverty non-black/non-Hispanic women. The NLSY79 is a prospective cohort study of women and men enrolled in 1979 (aged 14-22 y) who were surveyed annually until 1994 and biannually from 1994-2006. Children who were born to women in NLSY79 have been surveyed biannually since 1986 (NLSY79 Children and Young Adult Survey). Here, we examined births from 1972-2000 with at least 1 follow-up survey on the child (n = 5704).
Our study sample consisted of live, singleton births to women ≥15 years old (n = 5586) and included multiple births to the same woman. We excluded records with missing or implausible values for the following variables: birthweight or gestational age ( implausible values were defined as gestational age <22 or >44 weeks and/or unlikely combinations of birthweight and gestational age as defined by Alexander et al ; 15.0%); maternal prepregnancy BMI (records missing for prepregnancy height or weight or for women <20 years old, with implausible values as defined by the Centers for Disease Control and Prevention [CDC] ; 2.1%); demographic information (records missing for income, educational attainment, marital status, or smoking during pregnancy; 1.6%), and GWG (records missing for prepregnancy weight or at delivery; 0.8%). Our final sample included 4496 births to 3733 mothers. The percentage of mothers with 1, 2, 3, and 4 births in the dataset was 80.9%, 17.7%, 1.3%, and 0.10%, respectively. Births that were included in our sample were more likely to be firstborn (45.8% vs 42.0%) but were otherwise comparable with our target dataset.
At each survey, participants in NLSY79 reported their income, education, marital status, weight, and height (in 1981-1983, 1985, and 2006 only). In the first survey after a pregnancy, women were asked about the pregnancy, including weight before and at delivery, birthweight, mode of delivery, and length of gestation (self-report based either on date of last menstrual period or clinical measure). Child height and weight were measured by the interviewer or reported by the mother. To correct for potential self-reporting bias, we adjusted all maternal and child self-reported weights and heights using a regression calibration that had been calculated with the use of measured and self-reported data from the Third National Health and Nutrition Examination Survey. Race/ethnicity was categorized by NLSY79 interviewers at baseline as black, Hispanic (includes any Hispanic participants, regardless of race), and non-Hispanic/non-black (includes Asian, Pacific Islander, Native American, white, and all other).
We calculated GWG (kilograms) as the difference between maternal prepregnancy weight and weight at delivery. To calculate maternal prepregnancy BMI, we used the woman’s height (meters) that was reported closest to the birth of the child and prepregnancy weight. For women ≤20 years old at the birth of the child, we calculated age- and sex-specific BMI percentiles and defined BMI categories according to the CDC definitions for children. For women >20 years old, we used the standard formula (kilograms/square meter) and defined BMI categories according to the World Health Organization. GWG was categorized as “inadequate” if it was below the IOM’s recommended range for the woman’s prepregnancy BMI, “adequate” if it was within the range, and “excessive” if it was above the range.
We calculated birthweight for gestational age z-values based on 1999-2000 US national reference data and used these values to identify SGA ([<10 th percentile] and LGA ([LGA], > 90 th percentile] births. We calculated age- and sex-specific BMI percentiles for children at each age for whom height and weight were available between 2 and 20 years and defined child overweight as BMI ≥85th percentile (includes both overweight [85th-94th percentile] and obesity [≥95th percentile] as defined by the CDC ). Two percent of births in our sample did not have adequate follow-up data to calculate a child BMI at any age. Cesarean deliveries (22.7% of all deliveries) were reported as either primary (65.4%) or repeat (35.5%). Because previous cesarean delivery predicts subsequent cesarean deliveries, we defined primary cesarean delivery (referred to as cesarean) as our outcome of interest.
Postpartum weight retention was calculated as the difference between a woman’s prepregnancy weight and the first postpartum weight reported within 12-24 months (mean, 15.8 months) after delivery. No standard threshold has been established for “unhealthy” postpartum weight retention; we chose ≥2.5 kg (≥5.5 lbs) as our outcome because we believe that retaining this amount of weight at ≥1 year after delivery may be detrimental to women who were not underweight before pregnancy. Postpartum weights were not used if a woman was pregnant at the survey or between the survey and the delivery of the child in question. Fifty-two percent of first reported postpartum weight was outside of our time frame, and postpartum weight was not reported for 1%. We performed bivariate t -tests and multivariate logistic regressions to determine whether the measured covariates in our dataset predicted missing data for either birthweight for gestational age or postpartum weight retention. This exploration demonstrated that missing data for both outcomes was dependent on measured covariates (ie, was “missing at random”) and indicated that an analysis with the use of only records with available data would not induce bias.
We used separate generalized estimating equations to examine the association between GWG (kilograms) and each outcome (SGA, LGA, primary cesarean delivery, child overweight, and postpartum weight retention) with clustering by mothers with >1 birth and children with >1 BMI percentile observation. We also used generalized estimating equations to examine the associations between the same outcomes and weight gain below and above, compared with within the 2009 IOM guidelines. The residuals of these models were constant across the observed range of GWG.
Covariates were chosen a priori based on existing literature, and all models were adjusted for maternal age, race/ethnicity, poverty status, educational attainment, marital status, parity, smoking during pregnancy, maternal prepregnancy BMI, child sex, and year of birth and included an interaction term between maternal prepregnancy BMI and GWG to produce effect estimates for GWG with categories of maternal BMI. Models for SGA, LGA, and cesarean delivery were adjusted additionally for maternal height, child overweight and cesarean models for length of gestation, and postpartum weight retention models for number of weeks since birth. We also examined models that included covariates that may be on the causal pathway between GWG and cesarean delivery (birthweight), child overweight (birthweight and breastfeeding), postpartum weight retention (breastfeeding), and models that excluded women who gained <5 or >20 kg. These analyses demonstrated slightly attenuated effect estimates but did not change our results.
To examine the relative contribution of GWG to the risk of each outcome, we graphed the multivariate adjusted predicted probability of each outcome across the range of GWG that was observed in our sample for each category of maternal prepregnancy BMI, using the sample mean or proportion for each covariate within that BMI category. All analyses were performed with SAS software (version 9.1; SAS Institute, Cary, NC) and were weighted for sampling proportions. The Office for the Protection of Human Subjects at the University of California, Berkeley, waived the requirement for formal review of this research.
Results
Table 1 shows the distribution of key variables in our sample overall and by maternal prepregnancy BMI categories. Our sample was 18% black, 8% Hispanic, and 74% non-Hispanic/non-black. Before pregnancy, 7% of the women were underweight; 69% of the women were normal weight; 16% of the women were overweight, and 8% of the women were obese. Mean (±SD) total maternal GWG was 14.2 ± 6.9 kg. Approximately 30% of the women gained below or within the 2009 IOM guidelines, and 40% of the women gained above.
| Maternal-child characteristic | Analytic sample (4496 cases) | Maternal prepregnancy body mass index categories, n | |||
|---|---|---|---|---|---|
| Low: <18.5 kg/m 2 (306 cases [6.8%]) | Normal: 18.5-24.9 kg/m 2 (3108 cases [69.1%]) | Overweight: 25.0-29.9 kg/m 2 706 cases [15.7%]) | Obese: >30 kg/m 2 (376 cases [8.4%]) | ||
| Race/ethnicity, n (%) | |||||
| Non-Hispanic/non-black | 3336 (74.2) | 234 (76.5) | 2333 (75.1) | 502 (71.1) | 267 (71) |
| Black | 788 (17.5) | 49 (16.0) | 524 (16.8) | 131 (18.6) | 84 (22.4) |
| Hispanic | 372 (8.3) | 23 (7.5) | 251 (8.1) | 73 (10.3) | 25 (6.5) |
| Family below poverty level, n (%) | 829 (18.4) | 74 (24.1) | 563 (18.1) | 121 (17.1) | 72 (19.2) |
| Maternal educational attainment, n (%) | |||||
| Some high school or less (0-11 y) | 880 (19.6) | 76 (24.8) | 616 (19.8) | 123 (17.5) | 64 (17.1) |
| High school (12 y) | 1937 (43.1) | 140 (45.9) | 1338 (43.1) | 292 (41.4) | 167 (44.3) |
| Some college or more (>12 y) | 1679 (37.3) | 90 (29.4) | 1153 (37.1) | 291 (41.2) | 145 (38.5) |
| Married, spouse present vs never married or other, n (%) | 3156 (70.2) | 201 (65.8) | 2163 (69.6) | 521 (73.7) | 271 (72.1) |
| Child first born vs other, n (%) | 2060 (45.8) | 132 (43.3) | 1540 (49.5) | 266 (37.8) | 121 (32.2) |
| Mother smoked within 1 year before birth, n (%) | 1491 (33.2) | 136 (44.4) | 1052 (33.8) | 199 (28.2) | 104 (27.6) |
| Maternal age at birth, y b | 26.1 ± 5.5 | 24.9 ± 5.1 | 25.5 ± 5.5 | 27.5 ± 5.3 | 28.9 ± 5.2 |
| Length gestation, wk b | 38.6 ± 2 | 38.3 ± 2.1 | 38.7 ± 2.0 | 38.7 ± 2.0 | 38.6 ± 2.0 |
| Maternal gestational weight gain, kg b | 14.2 ± 6.9 | 15.5 ± 6.8 | 14.8 ± 6.1 | 13.2 ± 7.2 | 9.7 ± 9.9 |
| Maternal gestational weight gain categories, n (%) | |||||
| <0 kg | 62 (1.4) | 0 (0) | 11 (0.3) | 17 (2.4) | 34 (9.1) |
| 0-5 kg | 235 (5.2) | 8 (2.5) | 103 (3.3) | 53 (7.5) | 71 (19.0) |
| 5-10 kg | 828 (18.4) | 56 (18.3) | 530 (17.1) | 150 (21.3) | 92 (24.4) |
| 10-15 kg | 1589 (35.4) | 110 (35.9) | 1158 (37.3) | 230 (32.6) | 91 (24.3) |
| 15-20 kg | 1035 (23) | 69 (22.7) | 755 (24.3) | 163 (23.0) | 47 (12.6) |
| 20-25 kg | 501 (11.1) | 43 (14.0) | 371 (11.9) | 62 (8.8) | 25 (6.6) |
| >25 kg | 246 (5.5) | 21 (6.8) | 180 (5.8) | 30 (4.3) | 15 (3.9) |
| Institute of Medicine category, n (%) | |||||
| Inadequate (below recommended range) | 1334 (29.7) | 114 (37.2) | 1005 (32.3) | 109 (15.4) | 106 (28.1) |
| Adequate (within recommended range) | 1369 (30.4) | 97 (31.6) | 1032 (33.2) | 183 (25.9) | 57 (15.1) |
| Excessive (above recommended range) | 1793 (39.9) | 95 (31.2) | 1070 (34.4) | 414 (58.7) | 213 (56.8) |
| Female (vs male) child, n (%) | 2179 (48.5) | 167 (54.6) | 1495 (48.1) | 316 (44.8) | 201 (53.3) |
| Outcomes of interest, n (%) | |||||
| Small for gestational age (<10th percentile) | 500 (11.1) | 50 (16.5) | 358 (11.5) | 58 (8.3) | 33 (8.7) |
| Large for gestational age (>90th percentile) | 552 (12.3) | 23 (7.6) | 350 (11.2) | 107 (15.2) | 72 (19.1) |
| Primary cesarean delivery (vs vaginal) c | 647 (15.7) | 32 (11.1) | 427 (14.8) | 111 (17.6) | 77 (24.2) |
| Child overweight (body mass index, ≥85th percentile) d | 28.8% | 16.9% | 25.4% | 38.9 | 49.8 |
| Retained ≥2.5 kg from 12-24 mo, n (%) | 831 (40.3) | 66 (44.2) | 548 (39.2) | 143 (40.9) | 74 (44.7) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
