The objective of this study is to understand the relationships between prepregnancy obesity and excessive gestational weight gain (GWG) and adverse maternal and fetal outcomes. Pregnancy risk assessment monitoring system (PRAMS) data from Maine for 2000–2010 were used to determine associations between demographic, socioeconomic, and health behavioral variables and maternal and infant outcomes. Multivariate logistic regression analysis was performed on the independent variables of age, race, smoking, previous live births, marital status, education, BMI, income, rurality, alcohol use, and GWG. Dependent variables included maternal hypertension, premature birth, birth weight, infant admission to the intensive care unit (ICU), and length of hospital stay of the infant. Excessive prepregnancy BMI and excessive GWG independently predicted maternal hypertension. A high prepregnancy BMI increased the risk of the infant being born prematurely, having a longer hospital stay, and having an excessive birth weight. Excessive GWG predicted a longer infant hospital stay and excessive birth weight. A low pregnancy BMI and a lower than recommended GWG were also associated with poor outcomes: prematurity, low birth weight, and an increased risk of the infant admitted to ICU. These findings support the importance of preconception care that promotes achievement of a healthy weight to enhance optimal reproductive outcomes.
While the prevalence of obesity has recently stabilized in the United States, approximately 60 percent of women of reproductive age are either overweight or obese [
The association between maternal obesity and adverse fetal outcomes such as preterm labor, congenital abnormalities, macrosomia, and shoulder dystocia is also well established [
While entering pregnancy overweight or obese increases the risk for pregnancy complications, excessive gestational weight gain (GWG) also increases the risk of adverse outcomes for the mother and infant. Current Institute of Medicine (IOM) guidelines recommend a total weight gain of 15–25 lbs during pregnancy for overweight women with a BMI between 25 and 29.9 [
Given the correlations between prepregnancy obesity and excessive maternal weight gain during pregnancy on the one hand and health challenges for both the pregnant woman and her infant on the other, it is important to determine how prepregnancy obesity and excessive gestational weight gain interact with other factors to place pregnancies at the greatest risks. To help address this question, we performed multivariable analysis on data for the State of Maine obtained from the PRAMS (Pregnancy Risk Assessment Monitoring System) project, a national database collected and maintained by the Centers for Disease Control and Prevention [
This study analyzes data from the Maine Pregnancy Risk Assessment Monitoring System (PRAMS) questionnaires for the years 2000–2010, which we obtained from the Maine Department of Health and Human Services [
To compile this dataset, the Maine Department of Health and Human Services uses birth certificates to identify women who gave birth to a live infant within the previous 2–4 months. A stratified sample of 125 women per month is selected and mailed PRAMS questionnaires. Up to 2 follow-up questionnaires are mailed to nonresponders with attempted telephone follow-up during the subsequent 9 months of the postpartum period. Members of high-risk groups (e.g., women with low birth weight infants and Medicare recipients) were oversampled.
To determine the maternal factors related to poor maternal and infant outcomes, we analyzed a broad range of variables from the PRAMS dataset. Maternal prepregnancy height and weight were used to calculate BMI, which was analyzed as a continuous variable. Maternal age as well as the gestational age when each woman was sure she was pregnant and at her first prenatal visit (in weeks) was also analyzed as continuous variables. The following variables were analyzed as dichotomous variables: previous live birth (yes or no), marital status (married or not married), educational attainment (≤12th grade or >12th grade), household income (<
We performed multivariable logistic regression analysis using the survey procedures in the Statistical Analysis Software v9.3 [
The protocol was approved by the University of Southern Maine Institutional Review Board.
During the study period (2000 to 2010), Maine PRAMS questionnaires were obtained from 12,600 women who gave birth to live infants. The response rate in Maine is consistently >70%. A total of 39 questionnaires were excluded from analysis due to unknown birth weights resulting in 12,561 questionnaires for analysis. For each logistic regression analysis we considered only pregnancy records for which all variables were available. In all cases this resulted in analysis of >10,400 pregnancies.
We have previously published detailed mean demographic values, health behavior indicators, and outcomes for this dataset [
Predictors of the negative infant outcomes in this study including admission to an intensive care unit, longer hospital stay, low and excessive birth weight, and premature birth are shown in Tables
Maternal predictors of infant admission to ICU.
Odds ratio (95% CI) |
| |
---|---|---|
Smoked before pregnancy only/never smoked | 0.992 (0.790–1.247) | 0.9477 |
Smoked before and during pregnancy/never smoked | 0.923 (0.743–1.146) | 0.4656 |
Age | 1.001 (0.986–1.016) | 0.8741 |
First live birth/previous live birth | 1.454 (1.249–1.692) |
|
Not married/married | 1.135 (0.939–1.373) | 0.1908 |
Education ≤ 12 yrs/> 12 yrs | 1.227 (1.034–1.457) |
|
Prepregnancy BMI | 0.974 (0.965–0.983) |
|
Annual HH income ≤ |
1.313 (1.074–1.606) |
|
Urban or suburban/rural town or isolated rural | 1.540 (1.333–1.780) |
|
Nonwhite/white | 1.327 (0.863–2.039) | 0.1973 |
Drank alcohol prior to pregnancy/did not drink alcohol | 1.149 (0.987–1.338) | 0.0736 |
Drank alcohol in last 3 months of pregnancy/did not drink | 0.952 (0.707–1.281) | 0.7449 |
Gestational age when being sure she is pregnant | 0.989 (0.970–1.008) | 0.2634 |
Gestational age at first prenatal visit | 1.010 (0.987–1.034) | 0.3768 |
Pregnancy weight gain < recommended/recommended | 1.261 (1.064–1.495) |
|
Pregnancy weight gain > recommended/recommended | 0.764 (0.645–0.905) |
|
Logistic regression results with infant admission to an intensive care unit as the dependent variable: the infant is more likely to be admitted to an ICU if the mother was having her first birth, had no education past high school, lived in a household with an annual income <
Maternal predictors of longer hospital stay by newborn.
Comparison | Odds ratio (95% CI) |
|
---|---|---|
Smoked before pregnancy only/never smoked | 1.141 (0.991–1.313) | 0.0668 |
Smoked before and during pregnancy/never smoked | 1.137 (0.987–1.308) | 0.0745 |
Age | 1.033 (1.023–1.043) |
|
First live birth/previous live birth | 1.626 (1.477–1.789) |
|
Not married/married | 1.105 (0.979–1.248) | 0.1069 |
Education ≤ 12 yrs/> 12 yrs | 0.923 (0.827–1.031) | 0.1560 |
Prepregnancy BMI | 1.035 (1.028–1.042) |
|
Annual HH income ≤ |
1.307 (1.151–1.486) |
|
Urban or suburban/rural town or isolated rural | 0.962 (0.879–1.054) | 0.4074 |
Nonwhite/white | 0.787 (0.596–1.040) | 0.0921 |
Drank alcohol prior to pregnancy/did not drink alcohol | 0.909 (0.824–1.004) | 0.0593 |
Drank alcohol in last 3 months of pregnancy/did not drink | 0.961 (0.800–1.155) | 0.6737 |
Gestational age when being sure she is pregnant | 0.999 (0.984–1.014) | 0.9104 |
Gestational age at first prenatal visit | 0.995 (0.981–1.009) | 0.4786 |
Pregnancy weight gain < recommended/recommended | 1.066 (0.943–1.205) | 0.3046 |
Pregnancy weight gain > recommended/recommended | 1.124 (1.015–1.107) |
|
Logistic regression results with length of infant hospitalization as the dependent variable: infants were more likely to spend longer time in the hospital if mother was older, was having her first birth, had a higher prepregnancy BMI, lived in a household with an annual income <
Maternal predictors of low (<2500 gms) versus normal birth weight.
Comparison | Odds ratio (95% CI) |
|
---|---|---|
Smoked before pregnancy only/never smoked | 1.113 (0.961–1.289) | 0.1529 |
Smoked before and during pregnancy/never smoked | 1.624 (1.424–1.853) |
|
Age | 1.033 (1.023–1.044) |
|
First live birth/previous live birth | 1.832 (1.658–2.024) |
|
Not married/married | 1.198 (1.058–1.356) |
|
Education ≤ 12 yrs/> 12 yrs | 1.176 (1.055–1.311) |
|
Prepregnancy BMI | 1.004 (0.997–1.011) | 0.2993 |
Annual HH income ≤ |
0.895 (0.787–1.017) | 0.0882 |
Urban or suburban/rural town or isolated rural | 0.970 (0.885–1.063) | 0.5119 |
Nonwhite/white | 1.089 (0.824–1.439) | 0.5490 |
Drank alcohol prior to pregnancy/did not drink alcohol | 0.838 (0.758–0.926) |
|
Drank alcohol in last 3 months of pregnancy/did not drink | 0.799 (0.657–0.972) |
|
Gestational age when being sure she is pregnant | 1.020 (1.004–1.036) |
|
Gestational age at first prenatal visit | 0.972 (0.957–0.987) |
|
Pregnancy weight gain < recommended/recommended | 2.161 (1.935–2.413) |
|
Pregnancy weight gain > recommended/recommended | 0.721 (0.646–0.804) |
|
Logistic regression results with infant birth weight < 2500 gms as the dependent variable: compared to normal weight infants, infants are more likely to be born weighing < 2500 gms if their mother was older, was having her first child, was not married, had no education past high school, was not sure she was pregnant until later in gestation, had a weight gain < recommended range (compared to within recommended range), or smoked before and during pregnancy. Infants were less likely to be underweight if mother had a gestational weight gain > the recommended amount of weight (compared to within recommended range), had her first prenatal visit later in gestation, or drank alcohol before or during pregnancy.
Maternal predictors of excessive (≥4000 gms) versus normal birth weight.
Comparison | Odds ratio (95% CI) |
|
---|---|---|
Smoked before pregnancy only/never smoked | 0.933 (0.752–1.159) | 0.5335 |
Smoked before and during pregnancy/never smoked | 0.445 (0.338–0.586) |
|
Age | 0.999 (0.984–1.015) | 0.9480 |
First live birth/previous live birth | 0.677 (0.583–0.786) |
|
Not married/married | 1.089 (0.892–1.329) | 0.4008 |
Education ≤ 12 yrs/> 12 yrs | 0.815 (0.684–0.971) |
|
Prepregnancy BMI | 1.030 (1.020–1.041) |
|
Annual HH income ≤ |
1.161 (0.934–1.444) | 0.1783 |
Urban or suburban/rural town or isolated rural | 0.974 (0.846–1.121) | 0.7097 |
Nonwhite/white | 1.026 (0.626–1.681) | 0.9204 |
Drank alcohol prior to pregnancy/did not drink alcohol | 1.021 (0.876–1.189) | 0.7902 |
Drank alcohol in last 3 months of pregnancy/did not drink | 1.150 (0.883–1.496) | 0.2995 |
Gestational age when being sure she is pregnant | 1.005 (0.981–1.030) | 0.6675 |
Gestational age at first prenatal visit | 1.011 (0.990–1.032) | 0.3152 |
Pregnancy weight gain < recommended/recommended | 0.723 (0.568–0.921) |
|
Pregnancy weight gain > recommended/recommended | 2.210 (1.886–2.589) |
|
Logistic regression results with infant birth weight > 4000 gms as the dependent variable: compared to normal weight infants, infants are more likely to be born weighing > 4000 gms if their mother had a higher BMI or a gestational weight gain > recommended range. Babies were less likely to have a birth weight > 4000 gms if their mother smoked before and during pregnancy, was having her first live birth, had no education past high school, or had a gestational weight gain < recommended range.
Maternal predictors of infant born prematurely (<37 wks of gestation).
Comparison | Odds ratio (95% CI) |
|
---|---|---|
Smoked before pregnancy only/never smoked | 0.925 (0.736–1.161) | 0.5008 |
Smoked before and during pregnancy/never smoked | 1.010 (0.823–1.239) | 0.9261 |
Age | 1.017 (1.002–1.032) |
|
First live birth/previous live birth | 1.425 (1.230–1.650) |
|
Not married/married | 1.046 (0.873–1.254) | 0.6238 |
Education ≤ 12 yrs/> 12 yrs | 0.911 (0.769–1.079) | 0.2802 |
Prepregnancy BMI | 1.019 (1.009–1.028) |
|
Annual HH income ≤ |
1.255 (1.032–1.524) |
|
Urban or suburban/rural town or isolated rural | 0.884 (0.772–1.012) | 0.0740 |
Nonwhite/white | 1.341 (0.825–2.178) | 0.2360 |
Drank alcohol prior to pregnancy/did not drink alcohol | 0.787 (0.680–0.911) |
|
Did not drink alcohol in last 3 months of pregnancy/drank | 0.832 (0.621–1.116) | 0.2192 |
Gestational age when being sure she is pregnant | 1.025 (1.006–1.045) |
|
Gestational age at first prenatal visit | 0.972 (0.949–0.996) |
|
Pregnancy weight gain < recommended/recommended | 1.645 (1.401–1.931) |
|
Pregnancy weight gain > recommended/recommended | 0.721 (0.610–0.853) |
|
Logistic regression results with infant being born at < 37 weeks of gestation as the dependent variable. Infants were more likely to be born at < 37 weeks of gestational age which is greater for mothers who were older, were having their first birth, had a higher prepregnancy BMI, lived in a household with an annual income <
Maternal predictors of hypertension during pregnancy.
Comparison | Odds ratio (95% CI) |
|
---|---|---|
Smoked before pregnancy only/never smoked | 1.205 (1.022–1.420) |
|
Smoked before and during pregnancy/never smoked | 1.028 (0.866–1.220) | 0.7558 |
Age | 0.999 (0.987–1.011) | 0.8272 |
First live birth/previous live birth | 0.657 (0.586–0.736) |
|
Not married/married | 0.976 (0.840–1.133) | 0.7458 |
Education ≤ 12 yrs/> 12 yrs | 0.976 (0.855–1.113) | 0.7140 |
Prepregnancy BMI | 1.046 (1.037–1.055) |
|
Annual HH income ≤ |
1.018 (0.872–1.189) | 0.8223 |
Urban or suburban/rural town or isolated rural | 0.989 (0.888–1.102) | 0.8456 |
Nonwhite/white | 0.943 (0.673–1.320) | 0.7307 |
Drank alcohol prior to pregnancy/did not drink alcohol | 1.010 (0.898–1.135) | 0.8742 |
Drank alcohol in last 3 months of pregnancy/did not drink | 1.060 (0.849–1.323) | 0.6070 |
Gestational age when being sure she is pregnant | 0.993 (0.974–1.012) | 0.4463 |
Gestational age at first prenatal visit | 1.012 (0.995–1.029) | 0.1582 |
Pregnancy weight gain < recommended/recommended | 0.925 (0.792–1.081) | 0.3293 |
Pregnancy weight gain > recommended/recommended | 1.359 (1.205–1.534) |
|
Logistic regression results with maternal hypertension as the dependent variable: mothers were more likely to be hypertensive during pregnancy if they smoke before pregnancy, had a higher prepregnancy BMI, or had a gestational weight gain > recommended as compared to within the recommended range. Mothers were less likely to be hypertensive if they were having their first live birth.
Excessive prepregnancy weight and gestational weight gain predicted a range of negative outcomes for both mother and infant. Unsurprisingly, both a higher prepregnancy BMI and a gestational weight gain greater than recommended predicted maternal hypertension during pregnancy, as does smoking before pregnancy (Table
A low prepregnancy BMI or a gestational weight gain less than the recommended range also correlated with some risks. The risk of an infant being admitted to an ICU is greater if the mother has a low prepregnancy BMI or a level of gestational weight gain less than recommended (Table
This study included 2 measures of socioeconomic status, household income, and educational attainment, as independent variables; both were important predictors of negative outcomes. Mothers with no education past high school were more likely to give birth to infants who were admitted to an ICU (Table
The findings reported here showing that increasing prepregnancy BMI and excessive GWG are risks to both mother and newborn infant are in agreement with multiple previous studies [
The US Department of Health and Human Services encourages primary care providers to talk to all of their patients about overweight/obesity [
Once a woman becomes pregnant, providers of prenatal care should endeavor to meet the guidelines developed by the Institute of Medicine (IOM) and endorsed by the American College of Obstetricians and Gynecologists (ACOG) for nutrition and weight gain counseling [
Prior to pregnancy, obese women who are not achieving optimal weight through lifestyle management may benefit from medical interventions including behavioral therapy, pharmacotherapy, and bariatric surgery. Weight loss medications can be useful adjuncts to lifestyle changes in patients with a BMI greater than 30 kg/m2 who have failed to achieve weight loss goals. However, the role of weight loss medications in obesity management remains controversial because of the side effects of these medications and concerns about long-term efficacy. There are five drugs approved for weight loss management in the United States; however none of these drugs are approved for use during pregnancy. Current guidelines recommend that weight loss drugs be discontinued if the patient exhibits any adverse effects or does not achieve adequate weight loss. The definition of adequate weight loss varies among medications but is generally defined as 5 percent or more over baseline within 3 to 6 months [
Pregnant women who have no contraindications should be advised to participate in physical activity, as recommended by the US Department of Health and Human Services guidelines, of at least 150 minutes of moderate-intensity physical activity, such as brisk walking, hiking, or bicycling, per week [
This study also found what appear to be benefits of having a higher prepregnancy BMI or GWG. Both a higher prepregnancy BMI and a GWG > recommended decrease the risk that the infant will be admitted to an ICU (Table
The issue of how BMI impacts health is complex. A meta-analysis of international scope found that obesity is associated with increased all-cause mortality. It also found that people who are overweight but not obese exhibit decreased mortality compared to those of “normal” weight [
Previous studies have also found that, along with its positive impacts, stringent control of GWG in obese women may increase risk of prematurity, low birth weight, and infant admission to an ICU [
Our results show that, overall, having a GWG < IOM recommended amounts increased the risk of poor outcomes. When compared to women with GWG within the recommended range, women who had a GWG < recommended were more likely to give birth to infants who were admitted to the ICU, weighed <2500 gms, and were born prematurely (Tables
A pilot study of obese women with type 2 diabetes mellitus conducted in Denmark compared women who had a GWG < 5 kg (<than the IOM recommended) to women who had GWG > 5 kg. Women with GWG < 5 kg in this study were less likely to give birth to large for gestational age infants, although not less likely to give birth to infants weighing >4000 gms [
SES can be defined by occupation, education, or income, and marital status is a related variable [
The correlation between increased obesity risk and lower socioeconomic status in developed countries is well documented [
Lower SES, obesity, tobacco exposure, and alcohol consumption can all impact pregnancy outcomes and interact in complex ways. Previous studies have shown that women who have lower SES and/or live in poor neighborhoods are at increased risk for giving birth to infants who are premature or low birth weight [
This study has the limitations inherent in the PRAMS dataset. PRAMS questionnaires collect self-reports, which can be unreliable in matters such as weight, weight gain, and health behaviors. Furthermore PRAMS data include only women who have delivered a live-born infant and thus do not capture women whose pregnancies ended in a miscarriage, fetal death, or stillbirth.
This study also has limitations that arise from decisions made during data analysis. In our data analysis we could have excluded women with gestational diabetes mellitus (GDM) but chose not to do so. Women with GDM commonly have other important risk factors such as older age, higher prepregnancy BMI, and greater GWG [
There are also multiple methods of categorizing birth weight; birth weight can be categorized into fixed weight categories or, taking gestational age into account, can be categorized as large for gestational age (LGA) or small for gestational age (SGA). We used the fixed birth weight categories of <2500 gms, 2500–3999 gms, and ≥4000 gms because this is the system used by the National Institutes of Health [
Nonetheless, our results report strong correlations between prepregnancy BMI, gestational weight gain, and SES (among other variables) and negative pregnancy outcomes infant admission to an ICU, longer hospital stays by infants, prematurity, and both low and excessive birth weight. These results highlight which women are at risk to have negative pregnancy outcomes. They underscore the importance of high-quality preconception care to ensure that a woman is in optimal health prior to becoming pregnant. Because not all of the risk variables identified in this study are directly modifiable by the healthcare system (e.g., lower SES) healthcare providers would do well to focus on patient education around modifiable risk factors such as eating habits, physical activity, and smoking cessation in all women of childbearing age who plan to become pregnant. Once a woman is pregnant education around GWG should also be addressed.
The authors declare that they have no competing interests.