Prepregnancy weight and the risk of adverse pregnancy outcome

August 6, 2009

Obesity before pregnancy is associated with an increased risk of several adverse outcomes of pregnancy. The risk profiles among lean, normal, or mildly overweight women are not, however, well established.

Title: Prepregnancy weight and the risk of adverse pregnancy outcome
Reference: NEJM 1998;338:147-52
Authors: Sven Cnattingius, M.D., Ph.D., Reinhold Bergstrom, Ph.D., Loren Lipworth, Sc.D., and Michael S. Kramer, M.D.
Institutions: Karolinska Institute, Stockholm Sweden, Upsala University, Upsala Sweden, and McGill University, Montreal Canada
Study Design: Statistical analysis of Birth Registry Data


Background: Obesity before pregnancy is associated with an increased risk of several adverse outcomes of pregnancy. The risk profiles among lean, normal, or mildly overweight women are not, however, well established.

Methods: We studied the associations between prepregnancy body-mass index (defined as the weight in kilograms divided by the square of the height in meters) and the frequency of late fetal death, early neonatal death, preterm delivery, and delivery of a small-for-gestational-age infant in a population based cohort of 167,750 women in Sweden in 1992 and 1993. The women were categorized as follows, according to body-mass index: lean, less than 20.0; normal, 20 through 24.9; overweight, 25 through 29.9; and obese, 30 or more. The estimates were adjusted for maternal age, parity, smoking, education, whether the mother was living with the father, and maternal height.

Results: Among nulliparous women, the odds ratio for late fetal death were increased among women with higher body-mass indexes as compared with lean women, as follows: normal women, 2.2 (95 percent confidence interval, 1.2 to 4.1); overweight women, 3.2 (95 percent confidence interval, 1.6 to 6.2); and obese women, 4.3 (95 percent confidence interval, 2.0 to 9.3). Among parous women, only obese women had a significant increase in the rate of late fetal death (odds ratio, 2.0; 95 percent confidence interval, 1.2 to 3.3). Among nulliparous women, the risk of very preterm delivery (at <32.9 weeks' gestation) was significantly increased among obese as compared with lean women (odds ratio, 1.6 95 percent confidence interval, 1.1 to 2.3), whereas among parous women, the risk was highest among those who were lean. The risk of delivering a small-for-gestational-age infant decreased more with increasing body-mass index among parous than among nulliparous women.

Conclusions: Higher maternal weight before pregnancy increases the risk of late fetal death, although it protects against the delivery of a small-for-gestational-age infant.


While it is not news that obese women have increased risk of poor outcome with their pregnancies, it is news that their fetuses have an increased risk of late fetal death. But, hold on, maybe not. In the paper, the authors do mention that there was a higher incidence of diabetes and gestational diabetes among the obese women. They did not look at the increased late fetal death rate in relationship to diabetes however. This is the only mechanism by which I can explain an increase in the late fetal death rate among obese women.

Now that I have said that, onwards to look at the rest of the paper. This study suffers from the same ills as all studies based on large data bases. The investigator is at the mercy of a large number of people who have no interest whatsoever in the study. Over the two year interval in which data were collected, there were 204,555 singleton births. However, prepregnancy body-mass index could only be calculated on 167,750 or 82%. Actually this is probably better than most large data bases. It still excluded nearly 40,000 deliveries and depending on the demographics and outcomes of those pregnancies, might have altered the results.

The author's classification of weight groups based on BMI (Body-Mass Index) is slightly different from the one I am familiar with. They classified "Lean" as BMI <20, "Normal" as BMI = 20 through 24.9, "Overweight" as BMI = 25 through 29.9, and "Obese" as BMI > 29.9. I was taught that "Lean" was < 24, "Normal" was 24 through 26.9, "Overweight" was 27 through 29.9, "Obese" was 30 through 34.9, and "Morbidly Obese" was > 34.9. This probably makes little difference except that we have to be aware of the definitions being used in the study. Other definitions of interest are as follows: "Late Fetal Death" = Stillbirth occurring at 28 or more completed weeks of gestation' [The authors do not state whether this was the fetus dying at > 28 weeks or the fetus being delivered at > 28 weeks]; "Early Neonatal Death" = death occurring during the first week after birth; "Preterm Delivery" = Delivery at < 37 weeks; "Moderately Preterm" = 33 through 36 weeks; and "Very Preterm" as Delivery at < 33 weeks.

Keeping these definitions in mind, their significant results for
association with Late Fetal Death are as follows:


CharacteristicNo. of Births N=167,550LFD N=466O.R.95% C.I.
BMI < 2022634461.0 
BMI 20-24.91012662571.20.9 to 1.7
BMI 25-29.9334381091.61.1 to 2.3
BMI > 29.910412542.61.7 to 3.8
Para 0680672051.21.0 to 1.4
Para > 0996832611.0 
0 cig/ day1263853201.0 
1-9 cig/day22700751.31.0 to 1.7
>9 cig/day14265621.71.3 to 2.3

The association with Late Fetal Death when broken down by parity follows:


BMI < 2.61.0 
BMI 20-
BMI 25-

The association with delivery at < 33 weeks and at 33-36 weeks follow:


CharacteristicDel < 33 (N=1332)Del 33-36 (N=6165)
 O.R.95% C.I.O.R.95% C.I.
BMI < 201.0 1.0 
BMI 20-
BMI 25-

The association with delivery of a Small for Gestational Age infant follows:


BMI<204.5%1.0 2.7%1.0 
BMI 20-24.93.5%0.80.7-0.91.7%0.60.5-0.7
BMI 25-29.93.2%0.70.6-0.81.5%0.50.4-0.5
BMI >29.94.2%0.90.7-1.01.4%0.40.3-0.5

I interpret these data to show that there is a statistically significant increase in the Late Fetal Death Rate as the BMI increases. This is particularly true in the nulliparous population. In the parous women, this did not become significant until after the BMI was > 29.9. How can we utilize this clinically? Well, I don't think we can go out and get all women to lower their BMI to < 20 prior to achieving pregnancy for the first time, nor do we necessarily want to because this increased BMI was also associated with a decreased risk of delivery of a small for gestational age infant. The 4.3 fold increase in LFD was from 1.4/1000 to 5.7/1000. This is from less than 1% to less than 1 %. This comes about from the fact that there were so many patients in the study. The higher the "N", the more likely that any diference found will be significant statistically.

The protective effect of an increased BMI in parous women is about 2.5 fold and actually decreases the rate of SGA from 2.7% to 1.4%. This could be clinically significant. But again how can we use it? Can we have all of our parous patients go out and increase their BMI prior to getting pregnant. I doubt it.

The other finding of interest to me was the fact that there was a slight decrease in the incidence of very preterm and moderately preterm deliveries in those parous patients whose BMI resided in the middle range. Those women who were on the extremes. Lean and obese, had a higher incidence of preterm delivery than did those who were normal and overweight.

Another item to consider when evaluating this paper is the population that was studied. There was a 4.5% incidence of preterm delivery. As I write this, I sit in a hospital which has a 15.9% incidence of preterm delivery. In other words what is a small clinically insignificant change in this group of patients could be a clinically significant change in a group that has a higher incidence of the problem. Looking at this same data on another population might be well worth the effort.

I frequently catch students and residents with a trick question. I ask them what is the number one strongest association with birth weight. They usually respond with smoking. Then I get to smile and say no it is not. It is maternal prepregnant weight. 1 The reference listed below is one of many that can be found in the literature which will attest to this.

One of the soap boxes, that I like to stand on is about the complications of pregnancy that occur as a result of obesity. I find it interesting that the authors called a BMI of 30, obese. In the population I deal with, approximately 10% have a BMI >35. This leads to a higher cesarean rate, anywhere from 1.5 to 4 times higher in the obese, depending on which study you read and what definition of obese was used. Most papers looking at this used definition of a weight greater than 250 pounds at any time during the pregnancy. With this definition, the cesarean section rate, the rate of hypertensive disease of pregnancy, the incidence of gestational diabetes, and the incidence of fetal macrosomia are all increased. Yet very few laypeople and only a few more physicians consider obesity to be a serious complication of pregnancy. In fact, I am aware of patient with a weight of greater than 300 pounds who have had ovulation induction.

In summation, I find this to be a well done study, that raises some interesting issues that should be studied further.



Eastman,NJ and Jackson, E; Weight relationships in pregnancy. I. The bearing of maternal weight gain and prepregnancy weight on birth weight in full term pregnancies. Obstet Gynecol Surv 1968;23:1003-25.

NEJM 1998;338:147-52