Fetal macrosomia, often defined as birth weight above 4,000 or 4,500 g regardless of gestational length,1 is associated with both maternal and perinatal complications. When birth weight exceeds 4,000 g, both mother and newborn are at greater risk of morbidity including perineal lacerations, postpartum hemorrhage, caesarean delivery, shoulder dystocia, low Apgar score, birth trauma, and obesity.2–4 Several studies show that both mean birth weight and the proportion of newborns weighing more than 4,000 g and 4,500 g have increased during the past decades.5,6
Evidence-based guidelines indicate that regular exercise is an important component of a healthy pregnancy.7 However, recent studies show a decreasing trend of regular exercise during pregnancy.8,9 Both frequency and the intensity of exercise seem to decrease as pregnancy progresses,10,11 and most pregnant women shift from weight-bearing to non-weight-bearing exercises such as swimming and bicycling.12 Despite extensive literature on the relationship between regular exercise during pregnancy and mean birth weight, the results are ambiguous and lack consistency. Both a positive13–15 and negative association with newborn birth weight have been suggested.16–18 A few studies also report no difference in birth weight of neonates born to exercising and non-exercising mothers.19,20
The aim of the present study was to estimate, in a prospective cohort of pregnant women, the association of regular exercise, performed before and during pregnancy, with excessive newborn birth weight.
METHODS AND MATERIALS
The data used for this study are derived from the Norwegian Mother and Child Cohort Study (MoBa) conducted by the Norwegian Institute of Public Health.21 The Norwegian Mother and Child Cohort Study is a nationwide pregnancy cohort that aimed to include 100,000 pregnancies by 2008 and was designed to estimate the associations between some of the lifestyle variables to which pregnant women and their fetuses are exposed in addition to diseases.22 Pregnant women are recruited into the study through a postal invitation 2 weeks ahead of their routine ultrasound examination at gestational week 17 at their local hospital. Data are obtained from 50 of 52 maternity units in Norway.21 The overall participation rate for the present data file is 45%. However, the follow-up rate from inclusion to questionnaire 3 is 92%. The present study includes pregnancies enrolled between June 1, 2001, and May 31, 2005.
Participants receive three questionnaires during pregnancy weeks 17 and 30 (questionnaire 1, 2, and 3). Questionnaire 1 includes items of maternal health status, lifestyle behaviors, previous diseases, and medication covering both prepregnancy and the first weeks of pregnancy. Questionnaire 2 is a Food Frequency Questionnaire and is mailed with the invitation and questionnaire 1 in gestational week 17. Questionnaire 3, which is sent out in gestational week 30, focuses mainly on health outcomes during pregnancy and follows up some of the items from questionnaire 1. One reminder is sent by mail if the questionnaires have not been returned within 2 weeks. Linkage to the Medical Birth Registry of Norway was also provided. The questionnaires are available at www.fhi.no/morogbarn. Informed consent was obtained from each participant before inclusion. The study has received approval from the Regional Committees for Medical Research Ethics (S-95113) and The Norwegian Social Science Data Services (01/4325-6).
The second version of the quality-assured data file released for research in April 2006 provided data that were used in the present study. Both questionnaires 1 and 3 had to be answered in order for the women to be included (n=40,049). The record in the Medical Birth Registry of Norway23 from the present pregnancy and energy intake (MJ/d) from questionnaire 2 were also linked to the Norwegian Mother and Child Cohort Study data set. Pregnancies with missing information on year of birth were omitted from the analyses (n=142). We also excluded multiple pregnancies (n=723) and pregnancies ending before 37 weeks of gestation (n=2,315), leaving 36,869 pregnancies that constitute the study population.
The main outcome measure was excessive newborn birth weight as registered in the Medical Birth Registry of Norway. There is no widely agreed upon definition of fetal macrosomia or excessive newborn birth weight. To account for the increasing birth weight with increasing parity, we defined birth weight to be excessive if it was equal to or above the 90th percentile (ie, 4,170 g and 4,362 g for nulliparous and multiparous women, respectively).
The main exposure was regular exercise before and during pregnancy weeks 17 and 30, defined in terms of frequency. In both questionnaires 1 and 3, the participants were asked how often they engaged in the following exercises: strolling, brisk walking, running (jogging or orienteering), bicycling, fitness training in training centers, swimming, aerobic classes (low or high impact), prenatal aerobic classes, dancing (swing, rock, folkdance), skiing, ball games, horseback riding, and other. For all exercises, the respondents were asked to report frequency with the following categories: “never,” “one to three times per month,” “once a week,” “twice a week,” and “three or more times a week.” Strolling was excluded from the analysis due to its very low energy expenditure.24 Regular exercise participation before pregnancy was collected retrospectively in pregnancy week 17 (questionnaire 1). The respondents were asked to recall the type and frequency of exercises performed during the last 3 months before the present pregnancy. The questions on recreational exercise have shown moderate correlations with motion sensor measurements.25
Potential confounders of excessive birth weight were selected by cross-tabulations and literature review.26 The following confounders of excessive birth weight were evaluated: maternal age, maternal education, parity, hypertension, diabetes, gestational weight gain, body mass index (BMI) prepregnancy (both as a continuous and categorical variable), preeclampsia, smoking habits, and maternal height.5,27,28 Diabetes was defined as either preexisting diabetes or gestational diabetes of any kind. Hypertension was defined as any pregestational or gestational hypertensive disorder complicating pregnancy. Preeclampsia was defined as any diagnosis of preeclampsia. All diagnoses were based on ICD-9 codes from the Medical Birth Registry of Norway records. Parity was collected from the Medical Birth Registry of Norway and was defined in terms of earlier pregnancies lasting more than 20 weeks.29 Gestational length was also retrieved from the Medical Birth Registry of Norway and was based on a combination of ultrasound scanning and last menstrual period. Body mass index was calculated from self-reported body weight (questionnaire 1) and height (questionnaire 1) and categorized according to the World Health Organization: less than 18.5, 18.5–24.9, 25–29.9, 30–34.9, and 35 or higher. Total gestational weight change was calculated as the difference between the last pregnancy weight before 30 weeks of gestation and the self-reported weight when pregnancy started. Energy intake (MJ/d) was assessed using a Food Frequency Questionnaire (questionnaire 2), and the cutoff intervals for energy intake presented by Meltzer et al30 were used.
All analysis was carried out in the statistical software program, SPSS 15.0 for Windows (SPSS, Chicago, IL). Three logistic regression models were used to investigate the association between regular exercise before (Model A) and during pregnancy (Model B and C) and excessive newborn birth weight. All models adjusted for maternal age, education, BMI prepregnancy, and current smoking habits. Model B, which assessed the association between regular exercise in week 17 and excessive newborn birth weight, additionally adjusted for exercise prepregnancy, gestational weight change, energy intake (MJ/d), and preexisting diabetes/gestational diabetes mellitus. Lastly, the association between regular exercise in week 30 and excessive newborn birth weight was assessed in Model C, additionally adjusting for exercise prepregnancy, exercise in week 17, total gestational weight change, energy intake (MJ/d), preeclampsia, and preexisting diabetes/gestational diabetes mellitus. Further, to investigate which types of exercises were associated with excessive newborn birth weight, we used stepwise logistic regression adjusting for the same covariates as in Models A through C.
To evaluate the hypothesis that the odds of giving birth to newborns with an excessive birth weight continues to increase with further increases in regular exercise (frequency), we conducted tests for trends by treating the category numbers of regular exercise as an interval-scale variable in the logistic regression models (Wald test).
The possible interaction between maternal height and regular exercise on excessive newborn birth weight was estimated using stratification and multiplicative interaction term. Maternal height was dichotomized at the population median of 1.68 m, and regular exercise was dichotomized at a frequency of three or more times per week, before estimating the association between regular exercise before and during pregnancy and excessive newborn birth weight. However, we did not detect an interaction between maternal height and regular exercise before or during pregnancy on excessive newborn birth weight. Furthermore, we explored whether parity might modify the association between regular exercise and excessive newborn birth weight using stratification. This was done due to the observation that nulliparous women exercise more frequently than their multiparous counterparts.11,12 Hence, the results are presented separately for nulliparous and multiparous women.
Mean birth weight in this cohort was 3,682 g (standard deviation 488). Among the 36,869 pregnancies included, 4,033 (10.9%) newborns had a birth weight equal to or above the 90th percentile. A higher number of newborns with an excessive birth weight were born to multiparous women (n=2,263) compared with nulliparous women (n=1,770).
The distribution of maternal characteristics by parity is given in Table 1 and shows that nulliparous and multiparous women did not differ significantly in height, education, smoking habits, or diabetes. Nevertheless, nulliparous women were younger, had a lower energy intake (−0.23 MJ/d) (P<.001), gained more weight during pregnancy (P<.001), and their offspring had a lower mean birth weight compared with offspring of multiparous women (P<.001). The highest proportion of overweight women (BMI greater than 24.9), non-exercisers, and excessive newborn birth weight was seen in multiparous women.
Regular exercise performed 3 months before the present pregnancy did not affect the probability of delivering a high birth weight newborn in nulliparous or multiparous women (Table 2, Model A). A moderate protective effect of regular exercise during pregnancy was observed in nulliparous women, irrespective of time of exposure (gestational week 17 or 30) (Table 2, Models B and C).
Nulliparous women exercising at least three times a week in pregnancy week 17 were less likely to give birth to an newborn with an excessive birth weight (P for trend .008) (Table 2, Model B). Adjustment for hypertension and preeclampsia did not change the observed association between regular exercise in pregnancy week 17 and excessive newborn birth weight.
In week 30, nulliparous women exercising one to two times a week were less likely to deliver newborns with an excessive birth weight compared with non-exercisers, but this association was attenuated when we adjusted for gestational weight change independent of diabetes. The adjusted association reached significance only for nulliparous women exercising at least three times a week in pregnancy week 30 (Table 2, Model C).
Walking (adjusted odds ratio [aOR] 0.86, 95% confidence interval [CI] 0.75–0.99) and running (aOR 0.63, 95% CI 0.45–0.89) in pregnancy week 17 were negatively associated with excessive newborn birth weight in nulliparous women. Walking in pregnancy week 30 was also negatively associated with the outcome (aOR 0.84, 95% CI 0.73–0.96) (data not shown).
Multiparous women who participated in dancing in pregnancy week 17 were less likely to deliver newborns with an excessive birth weight (aOR 0.75, 95% CI 0.63–0.90), whereas training in fitness centers in pregnancy week 17 was positively associated with excessive newborn birth weight (aOR 1.16, 95% CI 1.00–1.35). In pregnancy week 30, low impact aerobics (aOR 0.68, 95% CI 0.47–0.97) and dancing (aOR 0.69, 95% CI 0.53–0.88) were negatively associated with excessive newborn birth weight. Multiparous women participating in swimming in pregnancy week 30 were more likely to give birth to an newborn with an excessive birth weight (aOR 1.16, 95% CI 1.04–1.30) compared with those who did not swim (data not shown).
In this large prospective pregnancy cohort study, nulliparous women performing a high level of exercise during pregnancy were less likely to give birth to newborns with an excessive birth weight. The highest number of newborns with excessive birth weight was observed in multiparous women. Interestingly, independent of parity, there seems to be an increasing trend of a protective effect with increasing frequency of regular exercise during pregnancy.
The results indicate that regular exercise during pregnancy may have a protective effect on excessive newborn birth weight, and this association tends to be different with parity. Excluding women with preexisting diabetes/gestational diabetes or preeclampsia from the analysis did not change the estimates substantially. As expected, regular exercise performed during pregnancy seems to have a greater influence on the upper extreme of the birth weight distribution compared with regular exercise performed before pregnancy. Nonetheless, women exercising regularly before pregnancy are also more likely to continue their exercise programs during pregnancy. Based on this study, we cannot rule out that exercising regularly before pregnancy may also affect the upper extreme of the birth weight distribution.
The strengths of this study are the prospective design, study size and that the outcome was obtained from an external source, the Medical Birth Registry of Norway.23 We therefore consider it unlikely that any misclassification due to imprecise measurements of the outcome influenced the results.
However, regular exercise was assessed indirectly by two self-administered questionnaires. Despite its limited accuracy and imprecision when it comes to assessing exercise duration and intensity, postal questionnaires are considered the most feasible method for assessing frequency of physical activity in large epidemiological studies.31 Because of the prospective data collection, misclassification of regular exercise in our study is most likely to be nondifferential and would most likely have biased the association toward the null. The questions used to assess regular exercise in our study have recently been compared with position and motion sensor measurements of physical activity. A positive association between self-reported frequency of recreational exercise and objectively measured physical activity was observed, indicating that the questions used in the Norwegian Mother and Child Cohort Study can be useful for ranking pregnant women according to their exercise level.25
Another limitation is the low response rate in the Norwegian Mother and Child Cohort Study. When comparing participants with nonparticipants using the Medical Birth Registry of Norway, some differences are indicated.21 Participating women seem to have a slightly different age distribution and to have a lower parity than nonparticipating women. They also smoke less and tend to have lower rates of preterm birth and low birth weight newborns compared with women from the source population.21 However, not all characteristics or exposures differ between participants and nonparticipants. And even though women in lower socioeconomic classes were underrepresented and may have influenced the prevalence estimates, we believe that estimates of associations will not necessarily be biased as long as reporting of outcomes and exposures is nondifferential and confounding is handled properly. This study estimates the association between regular exercise both before and during pregnancy and excessive newborn birth weight, which is believed to be an effect dependent on biological mechanisms, and therefore valid for participants as well as nonparticipants.
In the adjusted analysis, we strived to control adequately for possible confounding factors. Well-known predictors of birth weight, such as gestational diabetes and smoking, did not change the estimates substantially. Only a few women with preexisting or gestational diabetes mellitus were identified in our study, and excluding these women did not change the observed association between regular exercise and excessive newborn birth weight. We therefore consider it unlikely that the effect estimates are confounded by these factors.
The literature available on the relationship between physical activity during pregnancy and mean birth weight has been inconsistent.13,16,17,32,33 Nevertheless, a shift in mean birth weight may be of little relevance to the practicing obstetrician, whose main concern is directed toward the two extremes of the birth weight range where maternal and perinatal complications are increasing. If, for instance, a shift in mean birth weight is due to a factor exerting more, or all, of its influence at one extreme and little or none at the other, extrapolation from effects on mean values to other parts of the distribution can be misleading. Furthermore, a factor which only affects the spread of the birth weight distribution will make no difference to the mean but would increase (or decrease) the proportion at both extremes.34 Regular exercise may be an example of such a factor, rendering physical inactivity a risk factor for excessive newborn birth weight. To date, data relating regular exercise before and during pregnancy to the risk of excessive newborn birth weight are sparse. A moderate protective effect of regular exercise during pregnancy on excessive birth weight was observed in our study, which is in agreement with a case-control study by Alderman et al (1998),18 albeit a stronger protective effect was observed in their study. On the contrary, a recent study by Voldner et al35 in 2008 did not observe an association between level of physical activity during pregnancy and macrosomia risk. However, in contrast to these studies, our study is large and population based with a comprehensive prospective data collection. The discrepancy in findings between studies may be due to study design and size of study population in addition to different methods in defining type, intensity and frequency of regular exercise performed during pregnancy.
A possible mechanism behind our findings is the effect of aerobic exercise on glucose tolerance.36 Our observation that running, walking, dancing, and low-impact aerobics were negatively associated with excessive newborn birth weight supports this hypothesis. Both randomized trials37,38 and a prospective observational study39 have shown that light-to-moderate physical activity during pregnancy may reduce glucose levels both in women with gestational diabetes mellitus and in nondiabetic pregnant women. Given the adverse maternal and prenatal complications associated with excessive newborn birth weight, clinicians should promote regular exercise during pregnancy for the purpose of prevention.7 Nevertheless, neither a Cochrane review40 nor search on PubMed revealed randomized controlled trials evaluating the effect of regular exercise during pregnancy on excessive newborn birth weight. Although our results indicate a protective effect of regular exercise during pregnancy, there seems to be an urgent need for randomized controlled trials with high methodological and interventional quality to be carried out to study the causal relationship between regular exercise in pregnancy and excessive newborn birth weight.
1. Scifres CM, Stamilio D, Allsworth J, Shanks A, Lewkowski B, Shroff R. Perinatal consequences of fetal macrosomia: Zhang et al. Am J Obstet Gynecol 2008;198:603–4.
2. Zhang X, Decker A, Platt RW, Kramer MS. How big is too big? The perinatal consequences of fetal macrosomia. Am J Obstet Gynecol 2008;198:517e1–6.
3. Heiskanen N, Raatikainen K, Heinonen S. Fetal macrosomia–a continuing obstetric challenge. Biol Neonate 2006;90:98–103.
4. Hardy DS. A multiethnic study of the predictors of macrosomia. Diabetes Educ 1999;25:925–33.
5. Orskou J, Henriksen TB, Kesmodel U, Secher NJ. Maternal characteristics and lifestyle factors and the risk of delivering high birth weight infants. Obstet Gynecol 2003;102:115–20.
6. Bell R. Trends in birthweight in the north of England. Hum Fertil (Camb) 2008;11:1–8.
7. Exercise during pregnancy and the postpartum period. ACOG Committee Opinion No. 267. American College of Obstetricians and Gynecologists. Obstet Gynecol 2002;99:171–3.
8. Pereira MA, Rifas-Shiman SL, Kleinman KP, Rich-Edwards JW, Peterson KE, Gillman MW. Predictors of change in physical activity during and after pregnancy: project viva. Am J Prev Med 2007;32:312–9.
9. Petersen AM, Leet TL, Brownson RC. Correlates of physical activity among pregnant women in the United States. Med Sci Sports Exerc 2005;37:1748–53.
10. Borodulin KM, Evenson KR, Wen F, Herring AH, Benson AM. Physical activity patterns during pregnancy. Med Sci Sports Exerc 2008;40:1901–8.
11. Haakstad LA, Voldner N, Henriksen T, Bo K. Physical activity level and weight gain in a cohort of pregnant Norwegian women. Acta Obstet Gynecol Scand 2007;86:559–64.
12. Owe KM, Nystad W, Bo K. Correlates of regular exercise during pregnancy: the Norwegian Mother and Child Cohort Study. Scand J Med Sci Sports 2008 [Epub ahead of print].
13. Clapp JF 3rd, Kim H, Burciu B, Lopez B. Beginning regular exercise in early pregnancy: effect on fetoplacental growth. Am J Obstet Gynecol 2000;183:1484–8.
14. Hatch MC, Shu XO, McLean DE, Levin B, Begg M, Reuss L, et al. Maternal exercise during pregnancy, physical fitness, and fetal growth. Am J Epidemiol 1993;137:1105–14.
15. Campbell MK, Mottola MF. Recreational exercise and occupational activity during pregnancy and birth weight: a case-control study. Am J Obstet Gynecol 2001;184:403–8.
16. Clapp JF 3rd, Capeless EL. Neonatal morphometrics after endurance exercise during pregnancy. Am J Obstet Gynecol 1990;163:1805–11.
17. Clapp JF 3rd, Kim H, Burciu B, Schmidt S, Petry K, Lopez B. Continuing regular exercise during pregnancy: effect of exercise volume on fetoplacental growth. Am J Obstet Gynecol 2002;186:142–7.
18. Alderman BW, Zhao H, Holt VL, Watts DH, Beresford SA. Maternal physical activity in pregnancy and infant size for gestational age. Ann Epidemiol 1998;8:513–9.
19. Collings CA, Curet LB, Mullin JP. Maternal and fetal responses to a maternal aerobic exercise program. Am J Obstet Gynecol 1983;145:702–7.
20. Bell R, Palma S. Antenatal exercise and birthweight. Aust N Z J Obstet Gynaecol 2000;40:70–3.
21. Magnus P, Irgens LM, Haug K, Nystad W, Skjaerven R, Stoltenberg C. Cohort profile: the Norwegian Mother and Child Cohort Study (MoBa). Int J Epidemiol 2006;35:1146–50.
22. The Norwegian Mother and Child Cohort Study (MoBa). Available at: http://www.fhi.no/morogbarn/
. Accessed April 12, 2009.
23. Irgens LM. The Medical Birth Registry of Norway. Epidemiological research and surveillance throughout 30 years. Acta Obstet Gynecol Scand 2000;79:435–9.
24. Ainsworth BE, Haskell WL, Leon AS, Jacobs DR Jr, Montoye HJ, Sallis JF, et al. Compendium of physical activities: classification of energy costs of human physical activities. Med Sci Sports Exerc 1993;25:71–80.
25. Brantsaeter AL, Owe KM, Haugen M, Alexander J, Meltzer HM, Longnecker MP. Validation of self-reported recreational exercise in pregnant women in the Norwegian Mother and Child Cohort Study. Scand J Med Sci Sports 2009 [Epub ahead of print].
26. Brooke OG, Anderson HR, Bland JM, Peacock JL, Stewart CM. Effects on birth weight of smoking, alcohol, caffeine, socioeconomic factors, and psychosocial stress. BMJ 1989;298:795–801.
27. Boulet SL, Alexander GR, Salihu HM, Pass M. Macrosomic births in the United States: determinants, outcomes, and proposed grades of risk. Am J Obstet Gynecol 2003;188:1372–8.
28. Scott A, Moar V, Ounsted M. The relative contribution of different maternal factors in large-for-gestational-age pregnancies. Eur J Obstet Gynecol Reprod Biol 1982;13:269–77.
29. Venes D, Taber CW. Taber’s cyclopedic medical dictionary. 20 ed. Philadelphia: Davis; 2005.
30. Meltzer HM, Brantsaeter AL, Ydersbond TA, Alexander J, Haugen M. Methodological challenges when monitoring the diet of pregnant women in a large study: experiences from the Norwegian Mother and Child Cohort Study (MoBa). Matern Child Nutr 2008;4:14–27.
31. Sallis JF, Saelens BE. Assessment of physical activity by self-report: status, limitations, and future directions [published erratum appears in: Res Q Exerc Sport 2000;71:409]. Res Q Exerc Sport 2000;71:S1–14.
32. Magann EF, Evans SF, Weitz B, Newnham J. Antepartum, intrapartum, and neonatal significance of exercise on healthy low-risk pregnant working women. Obstet Gynecol 2002;99:466–72.
33. Kardel KR, Kase T. Training in pregnant women: effects on fetal development and birth. Am J Obstet Gynecol 1998;178:280–6.
34. Ounsted M, Moar VA, Scott A. Risk factors associated with small-for-dates and large-for-dates infants. Br J Obstet Gynaecol 1985;92:226–32.
35. Voldner N, Froslie KF, Bo K, Haakstad L, Hoff C, Godang K, et al. Modifiable determinants of fetal macrosomia: role of lifestyle-related factors. Acta Obstet Gynecol Scand 2008;87:423–9.
36. Bung P, Artal R, Khodiguian N, Kjos S. Exercise in gestational diabetes. An optional therapeutic approach? Diabetes 1991;40:182–5.
37. Jovanovic-Peterson L, Durak EP, Peterson CM. Randomized trial of diet versus diet plus cardiovascular conditioning on glucose levels in gestational diabetes. Am J Obstet Gynecol 1989;161:415–9.
38. Garcia-Patterson A, Martin E, Ubeda J, Maria MA, de Leiva A, Corcoy R. Evaluation of light exercise in the treatment of gestational diabetes. Diabetes Care 2001;24:2006–7.
39. Oken E, Ning Y, Rifas-Shiman SL, Radesky JS, Rich-Edwards JW, Gillman MW. Associations of physical activity and inactivity before and during pregnancy with glucose tolerance. Obstet Gynecol 2006;108:1200–7.
40. Kramer MS, McDonald SW. Aerobic exercise for women during pregnancy. Cochrane Database of Systematic Reviews 2006, Issue 3. Art. No.: CD000180. DOI: 10.1002/14651858.CD000180.pub2.