Conflict of interest The authors have stated explicitly that there are no conflicts of interest in connection with this article.
Please cite this article as: Koepp UMS, Andersen LF, Dahl-Joergensen K, Stigum H, Nass O, Nystad W. Maternal pre-pregnant body mass index, maternal weight change and offspring birthweight. Acta Obstet Gynecol Scand 2012; 91:243–249.
Unni Mette Stamnes Køpp, Research Department, Soerlandet Hospital HF, Serviceboks 416, 4604 Kristiansand, Norway. E-mail: email@example.com
Objective. To estimate the association between maternal pre-pregnant body mass index (BMI) and maternal weight change during pregnancy and offspring birthweight using the BMI classification developed by World Health Organization (WHO) and adopted by the Institute of Medicine (IOM) in 2009. Design. The Norwegian Mother and Child Cohort Study (MoBa) is a population-based pregnancy cohort study conducted by The Norwegian Institute of Public Health. Setting. Women were recruited from all geographic areas of Norway. Population. The study includes 58 383 pregnant women. Methods. Women were enrolled in 2000–2007 by a postal invitation offered to women in Norway at 17–18 weeks of gestation. Linear regression analyses are based on exposure data from two questionnaires during pregnancy and on birthweight data. Main outcome measures. Birthweight. Results. Mean pre-pregnancy BMI was 24kg/m2 (SD 4.3), mean maternal weight change in the first 30 weeks of gestation was 9.3kg (SD 4.4), mean birthweight was 3675g (SD 487) and mean age 30.3 years. Of the women, 65.2% had a normal pre-pregnancy weight, 2.9% were underweight, 22.3% overweight, and 9.5% obese (Classes 1–3). Linear regression analyses adjusted for potential confounders showed that offspring birthweight increased with increasing maternal pre-pregnant BMI, and with increasing maternal weight gain during pregnancy in all six categories of pre-pregnancy BMI. Women with the highest level of education had the highest offspring birthweight. Conclusion. Offspring birthweight increased with both increasing maternal pre-pregnant BMI and maternal weight gain during pregnancy in all six categories of maternal pre-pregnancy BMI.
The prevalence of overweight and obesity in young women has increased worldwide over the past 20 years (1). Obesity in pregnancy is associated with increased risk of gestational diabetes, hypertensive disorders, thromboembolic complications, stillbirths, caesarean section, and macrosomia and delivery complications (2). Both pre-pregnancy BMI and gestational weight gain are associated with the outcome of pregnancy (2,3). Maternal adiposity may lead to greater placental transfer of nutrients (glucose, free fatty acids and amino acids) during embryonic and fetal development (4). In offspring this may give rise to permanent changes in the hypothalamus, pancreatic islet cells, adipose tissue and the neuro-endocrine and biological system that regulate bodyweight (5). A ‘fetal overnutrition hypothesis’ mechanism has important public health implications as it may propel the obesity epidemic over several generations without further environmental or genetic influences (4,5). Consequently, weight gain in pregnancy has increasingly gained attention as a critical period for preventing obesity at the population level. Few studies have assessed in particular the association between maternal BMI and weight gained during pregnancy, and future risk of overweight/obesity in children (6–8), although there is epidemiological evidence that higher birthweight is associated with increased risk of adiposity in childhood and adulthood (6,8).
The optimal weight gain for the mother has periodically been questioned and revised and yet there is no general agreement on the recommended weight gain for pregnant women (3,9–11). The most widely adopted recommendations concerning weight gain in pregnancy are from the Institute of Medicine (IOM) published in 1990 and 2009 (12,13). The guidelines recommend weight gain ranges during pregnancy on the basis of four pre-pregnancy BMI groups (underweight, normal weight, overweight and obese), with smaller gains recommended for heavier women (12). Evidence was insufficient to construct specific guidelines for women with class II (BMI 35.0–39.9) or class III (BMI≥40.0) obesity (12). No large population-based studies have previously investigated the association between maternal pre-pregnant BMI and offspring birthweight across the six pre-pregnant BMI categories, taking the pattern of maternal weight gain during pregnancy into account.
The objective of the present study was thus to estimate the association between maternal pre-pregnant BMI and maternal weight change during pregnancy and offspring birthweight based on the BMI classification developed by WHO (1) and adopted by IOM in 2009 (12).
Material and methods
This study is based on the Norwegian Mother and Child Cohort Study (MoBa) conducted by the Norwegian Institute of Public Health (14,15). In brief, MoBa is a pregnancy cohort consisting of more than 100 000 pregnancies from 1999 to 2009. The majority of all pregnant women in Norway were invited to participate, and the participation rate was 44%. Participants were recruited to the study through a postal invitation in connection with a routine ultrasound examination offered to all pregnant women in Norway at 17–18 weeks of gestation (http://www.fhi.no/morogbarn). The study including women during 2000–2007 is based on version four of the quality-assured data files released for research in February 2009. MoBa has been approved by the Norwegian Regional Committee for Ethics in Medical Research and the Data Inspectorate, and informed consent was obtained from each participant before inclusion. The approval from the Data Inspectorate is from September 2003.
For the present purpose we used data from two postal questionnaires. The first (Q1) was responded to around gestational week 17–18 of pregnancy, and the third around gestational week 30 (Q3). Records from the Medical Birth Registry of Norway (MBRN) (16) from the present pregnancy, are included as part of the dataset. The response rate among participating women answering the questionnaires distributed during pregnancy was about 93%. This present study includes 63 491 singleton pregnancies. It was restricted to pregnancies with singleton, full-term births born live with a gestational age of 37 weeks or more (≥259 days), no Morbus Down, women with no weight reduction of more than 26kg and no weight gain of more than 40kg until gestational age 30 of pregnancy. We excluded 170 women with unrealistic (length <130cm, >195cm, weight <25kg) information on height, weight and 27 with missing information on birthweight. This gave a total number of 58 383 pregnant women and their offspring. According to this criteria, a total of 5.4% were excluded due to missing information on maternal weight or height pre-pregnancy (n=2935) or on maternal education (n=219) included in the regression analyses.
The main outcome variable was birthweight measured in grams, as registered in the MBRN (16). The main exposure was maternal pre-pregnant BMI based on self-reported weight and height from conception reported in Q1 (17). BMI was treated as a categorical variable divided into six categories based on World Health Organization recommendations: underweight (BMI <18.5kg/m2), normal range (BMI 18.5–24.9kg/m2), overweight (BMI 25.0–29.9kg/m2), obese class I (BMI 30.0–34.9kg/m2), obese class II (BMI 35.0–39.9), obese class III (BM I≥40.0) (1,12). Maternal weight change was calculated from maternal self-reported weight at gestational age 30 reported in Q3 and pre-pregnancy weight from conception reported in Q1. Weight change was coded into five categories: 1=decrease 26kg until ≤0kg, 2=increase>0 to ≤5kg, 3=increase >5 to ≤13kg, 4=increase >13 to ≤20kg, 5=increase more than 20kg. Based on a review of previous studies and an assumed possible underlying causal mechanism, we assessed covariates including plausible confounding factors using directed acyclic graphs (DAGs) (18). Other variables included in the model were: (1) maternal education classified according to length of education; ≤9 years (primary and lower secondary), ≥10–12 years (high school, vocational school), ≥13–16 years (college, higher education), and ≥17 years (university degree or equivalent), and other education were education levels not corresponding to the four above-mentioned categories; (2) parity (number of previous pregnancies of >20 weeks’ duration) based on MBRN classified into three categories: 0, 1, 3 or more; (3) smoking obtained from Q1 and coded ‘yes’ (smoking more than zero cigarettes per day or reporting occasional smoking) or ‘no’ (never smoked and not smoking); and (3) maternal age (<25 years, 25–30 years, >30 years). Depending on the analysis we treated maternal BMI, maternal weight change, parity and maternal age as either a continuous or a categorized variable.
We estimated the association between maternal pre-pregnant BMI and maternal weight change during pregnancy and offspring birthweight based on Figure 1. Three linear regression models were used. We regressed (1) pre-pregnant maternal BMI on birthweight, (2) pre-pregnant BMI on maternal weight change, and (3) pre-pregnant maternal BMI on birthweight mediated through maternal weight change. From these models we estimated the total effect of pre-pregnant BMI on offspring birthweight, the effect of pre-pregnant maternal BMI on maternal weight change, and the effect of pre-pregnant BMI on offspring birthweight, taking maternal weight change during pregnancy into account. In all models we adjusted for the confounders (maternal education, age, parity and smoking). For each regression model we checked assumptions (linearity and constant variance) and looked for points with large influence (plotting delta-beta), so the results were deemed robust. The analyses were conducted in SPSS version 18.0 (SPSS Inc. Chicago, IL, USA).
Mean pre-pregnancy BMI was 24.0kg/m2[standard deviation (SD) 4.3], mean maternal weight change in the first 30 weeks of gestation was 9.3kg (SD 4.4), mean birthweight was 3675 (SD 487), and mean age 30.3 years. Table 1 presents the distribution of offspring birthweight by maternal pre-pregnant BMI, maternal weight change and maternal educational level. Offspring birthweight increased with increasing pre-pregnant maternal BMI and maternal weight change during pregnancy. Offspring of women with 9 years of education had the lowest birthweight. There was no interaction between educational level and maternal BMI (data not shown).
Table 1. Offspring birthweight by maternal pre-pregnant body mass index (BMI), maternal weight change until week 30 of gestation and education in the Norwegian Mother and Child Cohort Study 2000–2007 (n=58 383).
Offspring mean birthweight (g)
p* value for mean birthweight
** Weight at 30 weeks of gestation minus maternal pre-pregnancy weight.
*** Incomplete data about weight, height or education.
Results for maternal pre-pregnancy BMI, maternal weight change and education are presented as percentage and mean birthweight and calculated by ANOVA trend analysis.
Maternal pre-pregnancy BMI
18.5–24.9, normal range
30.0–34.9, obese class I
35.0–39.9, obese class II
>40.0, obese class III
Maternal weight change**
−26 to 0kg
20kg or more
Completed maternal education (years)
incomplete data ***
Table 2 presents the adjusted analyses. Mean adjusted birthweight was 3277g if the mother had a pre-pregnancy BMI of 20kg/m2, a weight gain during the first 30 weeks of 0–5kg, the lowest level of education (≤9 years), was a non-smoker, nullipara and 20 years old. Using this as a reference category, Table 2 describes that every increase in one unit (1kg/m2) of pre-pregnancy BMI increased birthweight by 25.9g (95%CI: 25.0, 26.9). Furthermore, every maternal weight gain of 1kg increased birthweight by 22.4g (95%CI: 21.5, 23.3). An increase in weight gain of 10kg increased birthweight with therefore 224g. Offspring birthweight in women with 17 years or more of education was 79.2g (95%CI: 54.4, 104.0) higher than those with ≤9 years of education. Smoking decreased birthweight by −167.6g (95%CI: −181.6, −153.7) and being primiparous added 165.5g (95%CI: 156.5, 174.5). Every increase in age decreased birthweight by –6.1g (95%CI: −7.1, −5.1) for every year, giving a decrease of 61g at age 30.
Table 2. An example of the contribution of pre-pregnancy BMI, maternal weight change, education, smoking, parity and age on offspring birthweight based on adjusted analyses of 58 383 participants in the Norwegian Mother and Child Cohort Study (MoBa).
* Adjusted for maternal education, smoking, parity and age.
Figure 2 shows that offspring birthweight increased with increasing maternal weight in all six categories of maternal pre-pregnant BMI based on the BMI classification developed by WHO (1) and adopted by IOM (12). As there were only two mothers in the obesity class II and one mother in the obesity class III with weight gain of more than 20kg, they are not included in Figure 2.
In our study, a maternal weight gain of >13kg during the first 30 weeks of pregnancy was common in the lowest BMI categories, ranging from 16.7% in underweight mothers to 15.1% for normal weight and 14.3% overweight women, whereas it was, respectively, 6.3, 5.3 and 2.4% in obese class I, II and III. Obese women reduced weight more frequently during pregnancy (obese class 1; 8.7%, obese class II; 19.4%, obese class III; 34.2%) compared with overweight, normal weight and underweight women, respectively 2.5, 0.5 and 0.4% (data not shown).
The results of the statistical analysis guided by the assumed association between maternal pre-pregnancy BMI, maternal weight change during pregnancy and offspring birthweight shown in Figure 1 are illustrated in Figure 3. The adjusted regression estimates showed that the total effect of maternal pre-pregnant BMI on birthweight was a 25.9g increase in birthweight for a one unit increase in maternal pre-pregnant BMI. The expected weight change during pregnancy was 10.5kg for a mother with a BMI of 20. Mothers with higher BMI had slightly less weight change; −0.28g per unit BMI. Weight change was also associated with birthweight; +22.4g increase in birthweight per kg increase in weight change. This leads to an indirect effect of –0.28*22.4=−6.3g on birthweight per unit increase in BMI. Finally, the effect of maternal pre-pregnant BMI on offspring birthweight was 25.9−6.3 = 19.7g per unit increase in maternal pre-pregnancy BMI. These analyses were adjusted for maternal education, age, parity and smoking.
This population-based study showed that offspring birthweight increased with both increasing maternal pre-pregnant BMI and maternal weight gain during the first 30 weeks of pregnancy in all six categories of pre-pregnancy BMI.
No large population-based pregnancy cohorts have previously investigated the effect of different patterns of weight gain based on the new BMI classification developed by WHO (1) and adopted by IOM in 2009 (12). Several studies have demonstrated a relation between high maternal BMI and large offspring (13,19) and also a consistent association between maternal weight gain and birthweight (5,20). The recommended gestational weight gain from IOM (12,13) within each pre-pregnancy BMI category is associated with more favorable birth outcomes (12,21) than if weight gain is above or below the suggested range (12,22). Similar weight gain recommendations do not exist in Scandinavia yet. The knowledge about factors that may influence women to gain weight within the recommended ranges are limited and reports indicate that only 30–40% of women actually gain weight within these ranges (23).
The optimal weight gain among obese women remains unclear (12,21). Evidence has been insufficient to construct guidelines for weight gain among women with obesity class II and III (12,21). This study indicates that even among women in obesity class II or class III, both BMI and weight gain up to 20kg until week 30 of gestation influences birthweight. Previous results from smaller studies suggest that weight gain influences birthweight more in underweight and normal weight mothers (24–26) and are less apparent in overweight and obese women. In contrast, our results are more in agreement with Shapiro et al. (27), who showed that maternal weight gain (>35lb, or 15.9kg) resulted in higher birthweight in all categories on the basis of four pre-pregnancy BMI groups, also in obese women.
The determinants of birthweight are likely to be socially patterned (28,29). Social inequalities in health are a public health concern, and birthweight as a person's health trajectory over the life-course is one of the health indicators that have been associated with a host of diseases in adulthood (28,30). In our study, women with the highest level of education had a higher offspring birthweight compared with those with the lowest level of education, and education had a minor effect on offspring birthweight. A recent Danish study showed that maternal smoking and BMI are important intermediate determinants of the educational gradient, and as the prevalence of obesity is increasing and smoking is decreasing, the relative impact of these might cancel each other out and could imply that later effects of the educational gradient is not reflected in birthweight (28).
The strength of the present study is that it is a large population-based study of pregnant women during early pregnancy (14). The study includes information on previous pregnancies, smoking habits and other life-style factors, items of health and different exposures during pregnancy. The data are linked to the the Medical Birth Registry of Norway (16), and the size of the study enables us to assess separately the effect of maternal pre-pregnancy BMI and weight change during pregnancy on birthweight in different categories of pre-pregnancy BMI, including obesity class I–III.
However, the present study has some limitations. Misclassification due to under- or over-reporting of maternal weight and height represents a potential bias. However, most studies on maternal pre-pregnancy weight and maternal weight change are based on self-reported weight, and validation studies have showed that self-reported weight may be adequate to use in large population-based epidemiological studies (31). Maternal weight change during pregnancy includes only maternal weight gain until week 30 of gestation. By limiting weight gain to 30 weeks of gestation, our findings are not distorted by the weight of edema, which frequently develops in the last trimester. Also, suboptimal weight gain contributing to preterm delivery is excluded. However, we have complete data on birthweight for all full-term infants. Maternal health problems during pregnancy may influence birthweight and we attempted to address this by performing analyses adjusting for gestational diabetes, diabetes and preeclampsia, with the findings remaining essentially unchanged.
The MoBa population tends to have a higher level of education compared with the Norwegian source population of pregnant women (14) and our results therefore may underestimate the effect of education. Furthermore, we can not exclude that uncontrolled confounding may have biased our results. Selection bias may also be an issue. We excluded approximately 5.4% of the participants due to incomplete information. The distribution of missing data was, however, approximately the same (about 5%) in each BMI category for maternal weight and height, and for education it was only 0.4%, suggesting no major selection bias that may influence our results.
The implication of this study is that pre-pregnant BMI alone is an important predictor of birthweight. Furthermore, weight gain during pregnancy has an effect on offspring birthweight independent of pre-pregnancy BMI. An obvious goal should be to encourage overweight and obese woman to attain healthy weight before conception and keep a moderate weight gain during pregnancy.
The donation of questionnaire data from participants of the Norwegian Mother and Child Cohort Study is gratefully appreciated.
The Norwegian Mother and Child Cohort Study is supported by the Norwegian Ministry of Health, NIH/NIEHS (grant no NO1-ES-85433), NIH/NINDS (grant no.1 UO1 NS 047537-01), and the Norwegian Research Council/FUGE (grant no. 151918/S10). Research grant from South-Eastern Norway Regional Health Authority and Sorlandet Hospital HF.