Fox, Nathan S. MD; Saltzman, Daniel H. MD; Kurtz, Howard MD; Rebarber, Andrei MD
OBJECTIVE: To estimate pregnancy outcomes in twin pregnancies with excessive weight gain, according to the 2009 Institute of Medicine (IOM) recommendations.
METHODS: A retrospective cohort of 170 patients from a single practice with twin pregnancies at 37 weeks or more of gestation from 2005 to 2010 were analyzed. We compared outcomes between three cohorts based on the 2009 IOM recommendations for weight gain during pregnancy: poor weight gain, normal weight gain, and excessive weight gain. One-way analysis of variance and χ2 for trend were used when appropriate.
RESULTS: Comparing poor, normal, and excessive weight gain, the proportion of women delivering both newborns weighing more than 2,500 g increased from 40% to 60.5% to 79.5% across the three groups (P<.001). Additionally, the mean birth weight of the larger (P<.001) and smaller (P=.002) twin increased significantly across the three groups. The proportion of women with gestational diabetes, gestational hypertension, or pre-eclampsia did not increase significantly across the three subgroups. Our results were similar when we only included women with a normal prepregnancy body mass index.
CONCLUSION: Excessive weight gain in term twin gestations, as defined by the IOM, does not appear to be associated with worse maternal outcomes and is associated with larger birth weights. Our results call into question the 2009 IOM definition of excessive weight gain for a twin pregnancy.
LEVEL OF EVIDENCE: II
In 2009, the Institute of Medicine (IOM) revised their recommendations for weight gain in twin pregnancies. The basis for the revised recommendations was previous research that demonstrated a positive association between gestational weight gain in twin pregnancies and birth weight,1–3 as well as the increasing knowledge that maternal body mass index (BMI) should be taken into account when defining the optimal weight gain in twin pregnancies.4 The revised guidelines for twin pregnancies are BMI-specific and assuming a term (37–42 weeks of gestation) delivery: 17–25 kg (37–54 lb) for normal weight women (BMI 18.5–24.9), 14–23 kg (31–50 lb) for overweight women (BMI 25–29.9), and 11–19 kg (25–42 lb) for obese women (BMI 30 or more). The IOM concluded that regarding underweight women (BMI less than 18.5) with twin pregnancies, there was insufficient evidence to make recommendations for weight gain during pregnancy. Although the IOM referred to these recommendations as “provisional,” a recent comprehensive review of nutrition in twin pregnancies advised adhering to the 2009 IOM recommendations for weight gain.5 In a recent retrospective cohort, we found that, in patients with twin pregnancies, weight gain that met or exceeded the 2009 IOM minimum weight gain recommendations was significantly associated with improved outcomes as compared with weight gain during pregnancy that did not meet the 2009 IOM minimum recommendations.6 Therefore, the IOM definition for minimum weight gain in twin pregnancies appears to correlate with pregnancy outcomes.
Regarding the definition of excessive weight gain, the 2009 IOM guidelines reflect the 75th percentile of weight gain among women who delivered newborns 2,500 g or larger at 37–42 weeks of gestation.4 The definition of “excessive” was therefore based on descriptive data, not outcome-based data. Therefore, it is unknown whether excessive weight gain, as defined by the IOM, is in fact associated with an increased likelihood of adverse outcomes such as gestational diabetes and hypertension. It is possible that additional weight gain may be associated with improved outcomes, such as less low birth weight (LBW) newborns. Finally, excessive weight gain could be associated with less LBW at the expense of gestational diabetes or hypertension. However, without data, it is difficult to counsel women with twin pregnancies regarding what constitutes excessive weight gain and, more importantly, what are the clinical consequences of this excessive weight gain. The objective of this study was to estimate pregnancy outcomes in women with excessive weight gain according to the 2009 IOM recommendations for twin pregnancies.
MATERIALS AND METHODS
After obtaining Institutional Review Board approval from the Biomedical Research Alliance of New York, we reviewed a historical cohort of patients with twin pregnancies in our private Maternal-Fetal Medicine practice between 2005 and 2010. Using our computerized medical record, baseline characteristics and pregnancy outcomes were obtained. All of our patients are asked to provide their prepregnancy weight at the initial prenatal visit, and maternal height was measured by a registered nurse or certified medical assistant. The prepregnancy BMI was calculated using these two values in kg/m2. At each prenatal visit, the maternal weight is measured on a scale by a registered nurse or certified medical assistant. The weight gain during pregnancy was calculated by subtracting the prepregnancy weight from the final weight measured in our office. Gestational age was determined by last menstrual period and confirmed by ultrasound examination in all patients. The pregnancy was redated if there was more than a 5-day discrepancy up to 14 weeks or more than a 7-day discrepancy after 14 weeks. If the pregnancy was the result of in vitro fertilization (IVF), gestational age was determined from IVF dating.
Patients were included if they had a recorded prepregnancy weight, maternal height, and maternal weight measurements during pregnancy. Because the IOM definition for excessive weight gain are for women who deliver at 37 weeks of gestation or more, we only included patients who delivered at 37 weeks of gestation or more. All patients received prenatal care in our practice; patients who transferred to our practice after 20 weeks of gestation were excluded from analysis. We also excluded monoamniotic twins and pregnancies with major fetal anomalies.
We categorized patients into three groups based on the 2009 IOM recommendations: poor weight gain, normal weight gain, and excessive weight gain, taking into account prepregnancy BMI (Table 1). We compared pregnancy outcomes across the three groups. To account for the differences in prepregnancy BMI, we then performed the same analysis in women with a normal starting BMI only (18.5–24.9). Because the IOM did not make recommendations for underweight women (BMI less than 18.5), these patients were excluded from all analyses.
In our practice, we typically deliver uncomplicated twin pregnancies by 38–39 weeks of gestation. The diagnosis of gestational hypertension or pre-eclampsia was made using standard criteria.7 One-way analysis of variance and χ2 test for trend8 were used when appropriate (SPSS for Windows 16.0). A P<0.05 was considered significant. A multivariable regression analysis controlling for potential confounding variables was performed as well.
We delivered 439 patients with twin pregnancies during the study period. We excluded seven patients with monoamniotic–monochorionic placentation, six patients who transferred to our practice after 20 weeks of gestation, and one patient with didelphys uterus. Of the remaining 425 patients, 178 (41.9%) delivered at 37 weeks of gestation or more (the latest gestational age at delivery was 39 0/7 weeks). Eight additional patients were excluded because their prepregnancy BMI was less than 18.5 (underweight). This left 170 patients for analysis. No patients were lost to follow-up.
Fifty five (32.4%) patients had poor weight gain, 76 (44.7%) patients had normal weight gain, and 39 (22.9%) patients had excessive weight gain, according to the 2009 IOM guidelines. Baseline characteristics of the three groups are shown in Table 2. The mean age and proportion of IVF pregnancies increased across the three weight gain groups. Pregnancy outcomes are shown in Table 3. Comparing poor, normal, and excessive weight gain, the proportion of women delivering both newborns weighing more than 2,500 g increased from 40% to 60.5% to 79.5% across the three groups (P<.001). Additionally, the mean birth weight of the larger (P<.001) and smaller (P=.002) twins increased significantly across the three groups. We did not see any significant increase in the likelihood of gestational hypertension, pre-eclampsia, gestational diabetes, or neonatal intensive care unit admission across the three groups. After controlling for potential confounding variables (maternal age, starting BMI, IVF pregnancy, multifetal reduction, chorionicity, gestational age at delivery, and race), excessive weight gain remained significantly associated with larger birth weights and was not significantly associated with measured adverse outcomes (Table 3).
We then performed the same analysis only including the 117 women with a normal prepregnancy BMI, excluding all overweight and obese women. In this cohort, 43 (36.8%) patients had poor weight gain, 49 (41.9%) patients had normal weight gain, and 25 (21.4%) patients had excessive weight gain, according to the 2009 IOM guidelines. The baseline characteristics and pregnancy outcomes for this cohort are shown in Tables 4 and 5. In the cohort of women with a normal prepregnancy BMI, we found similar results as with the entire cohort. Comparing poor, normal, and excessive weight gain, the proportion of women delivering both newborns more than 2,500 g (ie, not LBW) increased from 32.6% to 56.1% to 76% across the three groups (P=.001). Additionally, the mean birth weight of the larger (P<.001) and smaller (P=.004) twins increased significantly across the three groups. We also did not see any significant increase in the likelihood of gestational hypertension, pre-eclampsia, gestational diabetes, or neonatal intensive care unit admission across the three groups in this analysis. After controlling for potential confounding variables (maternal age, starting BMI, IVF pregnancy, multifetal reduction, chorionicity, gestational age at delivery, and race) in this cohort of women with normal prepregnancy BMI, excessive weight gain remained significantly associated with larger birth weights and was not significantly associated with measured adverse outcomes (Table 5).
To assess whether the exclusion of preterm births in our study introduced a possible ascertainment bias, we compared the mean weight gain per week in the 170 term patients (1.13±0.39 lb) included in our study to the 228 patients who delivered at less than 37 weeks of gestation and were excluded (1.07±0.41 lb). There was no significant difference found (P=.106). We also performed two other comparisons, including all women with twin pregnancies (term and preterm). We found no significant difference when we compared the mean weight gain per week in all twin pregnancies with gestational diabetes to all twin pregnancies without gestational diabetes (1.01±0.36 lb compared with 1.10±0.41 lb; P=.273), as well as all twin pregnancies with gestational hypertension or pre-eclampsia to all twin pregnancies without gestational hypertension or pre-eclampsia (1.14±0.41 lb compared with 1.08±0.41; P=.315).
In this study, women with twin pregnancies at 37 weeks of gestation or more whose weight gain during pregnancy would be considered excessive according to revised 2009 IOM guidelines had significantly improved pregnancy outcomes, including larger newborns and a lower incidence of LBW. This did not come at the expense of a higher incidence of gestational diabetes, gestational hypertension, or pre-eclampsia. This was true in the entire cohort of 170 women, as well as the subgroup of 117 women with a normal prepregnancy BMI. We only included women who delivered at 37 weeks of gestation or more to control for the expected differences in weight gain based on gestational age at delivery.
These findings call into question the clinical validity of the 2009 IOM definition of excessive weight gain for a twin pregnancy. Because the definition of excessive weight gain was based solely on the upper quartile in a normal population, and was not based on data suggesting increased adverse maternal or fetal outcomes, we believe that further research is warranted to better define the true upper limit of normal weight gain during a twin pregnancy in relation to adverse pregnancy outcomes. Otherwise, women with twin pregnancies may be encouraged to gain less weight or decrease their caloric intake, which could lead to poorer outcomes, such as an increased likelihood of delivering LBW newborns. Until there are data supporting an upper limit of normal weight gain, we believe that obstetricians and other providers should only be focusing their efforts on encouraging women with twin pregnancies to meet or exceed the minimum 2009 IOM guidelines, which have been associated with improved outcomes.6
Because our cohort was primarily women with a normal prepregnancy BMI (68.8% of our cohort), it may not be appropriate to extrapolate these findings to all patients with twin pregnancies. Future research is warranted in overweight and obese women with twin pregnancies to determine their optimal weight gain. Additionally, because 100% of our patients had private health insurance, our results also may not be applicable to other populations with different socioeconomic backgrounds because dietary habits may be quite varied.
In our cohort, the women who gained excessive weight were not significantly more likely to have development of gestational hypertension, pre-eclampsia, or gestational diabetes. However, it is unknown whether any weight gain threshold exists above which the risk of these conditions outweighs the benefits. We were limited by our sample size to make group-specific comparisons and we did not have enough patients with an overweight or obese prepregnancy BMI to determine whether women in these weight categories who have excessive weight gain are more susceptible to these conditions. However, for women with twin pregnancies and a normal prepregnancy BMI, it does not appear that excessive weight gain increases their risk of development of gestational hypertension, pre-eclampsia, or gestational diabetes.
We chose not to include preterm deliveries because the gestational age at delivery is definitely associated with gestational weight gain and because the IOM recommendations were made for deliveries at 37–42 weeks of gestation. However, the relevant potential adverse outcomes related to gestational weight gain include complications that may have required or resulted in early delivery, and thus our decision to exclude preterm deliveries may have introduced an element of ascertainment bias because they would not have made it for inclusion in the study cohort. We did not find any evidence of ascertainment bias based on our analyses of weight gain in preterm compared with term deliveries, gestational diabetes compared with nongestational diabetes, and gestational hypertension compared with normotensive patients. Although this does not exclude the possibility of ascertainment bias, it does limit it.
Strengths of our study include the uniform and supervised measurements of height and weight as compared with the use of birth certificate data, which can be flawed. The prepregnancy weight was based on patient reporting, which may introduce bias. Like all retrospective studies, we are limited by our study design. It cannot be determined from our data if increased weight gain led to larger newborns or if larger fetuses lead to increased weight gain. Ideally, a prospective trial in twin pregnancies comparing various caloric regimens would help determine optimal weight gain in twins as they correlate to specific maternal and fetal outcomes.
© 2011 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.