Birth weight is strongly and consistently associated with both short-term and long-term morbidity and mortality.1–4 Although physical activity during pregnancy has often been considered a potentially important determinant of birth weight, studies of the relationship between physical activity during pregnancy and birth weight have been equivocal.5–12 Some authors suggest that maternal participation in physical activity throughout pregnancy may enhance birth weight,5–8 whereas others indicate the opposite.9–12 A consistent limitation of these studies is imprecise measurement of physical activity.
We evaluated three physical activity measurement techniques in a cohort of women studied prospectively from 20 weeks of gestation.13,14 Twice in pregnancy (at 20 and 32 weeks), mothers were weighed, and their physical activity was assessed continuously for 48 hours by three different methods (accelerometry, heart rate monitoring, and physical activity recall) to obtain measures of all waking energy expenditure. From these analyses we found that the three physical activity measures were concordant with one another in both active and sedentary women and were sensitive to variations in physical activity intensity and duration among the active group.14
Maternal size measures, such as weight gain (during pregnancy) and height, are established determinants of birth weight.15,16 In this paper, we assess the separate and combined relationships of aerobic physical activity during pregnancy, maternal weight gain during pregnancy, and height to the fetal growth ratio, which is calculated as birth weight divided by the parity, race, gender, and gestational age-specific median birth weight of all live births in the United States in 1989.17 All reference made to physical activity in this sample is specific to aerobic physical activity and does not include other various forms of isometric exercise (ie, stretching, yoga, weight lifting).
MATERIALS AND METHODS
Fifty-one nonsmoking, mostly white and college-educated women with normal pregnancies were recruited from obstetric clinics in the Lansing, Michigan, area at approximately 15 weeks of pregnancy and were followed until 12 weeks postpartum. These are the same women who participated in our earlier methodological study.13,14 Women without obstetric complications were included in all analyses. The researchers had nothing to do with the obstetric care of the participants and thus made no attempt to blind the managing clinicians to the physical activity of their patients. All women received similar obstetric care by clinicians in the surrounding area where the research was completed. At the time of recruitment, each woman and her physician provided written informed consent. The informed consent process and all study protocols were approved by the Michigan State University Committee for Research Involving Human Subjects.
For two 48-hour periods, one in the second trimester (at 20 weeks, range 18–22 weeks) and one in the third trimester (at 32 weeks, range 30–34 weeks), subjects completed physical activity diaries and wore Caltrac accelerometers (Hemokinetics, Madison, WI) and heart rate monitors (Polar CIC Inc, Port Washington, NY) to record energy expenditure. During these 48-hour periods, women kept detailed physical activity diaries. On an hourly basis during all waking hours, they wrote down the amount of light, moderate, hard, or very hard physical activity performed during the previous hour. Moderate activity was defined as any activity intensity equivalent to that perceived during a brisk walk. Hard activity was defined as activity intensity equivalent to that perceived between walking briskly and running. Very hard activity was defined as activity intensity equivalent to the level of exertion the women felt during running. All other minutes of recorded activity were considered light, and all other waking minutes were considered inactive. Caltrac is a single-plane motion sensor that provides a digital readout of energy expenditure (estimated by using height, weight, and age), using a piezoelectric transducer to measure an individual’s movement and acceleration.18 Each participant was instructed to wear the sensor on her left hip, clipped to a waist band. The Caltrac is not waterproof, so energy expenditure during swimming was estimated from the values reported in the physical activity recall.
To adjust heart rate values for changes in resting heart rate and fitness during pregnancy, oxygen consumption by each woman was measured at rest and during a treadmill exercise protocol (three stages of increasing intensity) at 20 and 32 weeks.13 We averaged each woman’s heart rate on a minute-by-minute basis and determined her “flex” heart rate. The “flex” heart rate is the point between the highest heart rate recorded during the resting phase and the lowest heart rate recorded during the initial exercise phase. Individual heart rate/oxygen consumption regression lines were constructed for each woman by using steady-state values obtained during the three exercise intensities. Using the “flex” heart rate and the slope and y-intercept values from the individual heart rate/oxygen consumption regression lines, we determined an oxygen consumption (VO2) value for the most recorded heart rate during the 2-day period of activity monitoring. These adjusted values correlated well (r=0.37–0.90; average r=0.60) with both Caltrac accelerometer and physical activity recall data.14 For this report, our physical activity measures were those derived from results from the Caltrac accelerometer.
Average metabolic equivalent unit (1 metabolic equivalent unit=1 kcal×kg–1×h−1) intensity of physical activity performed above resting values was calculated, using the amount of time (h) the accelerometers were worn, kilocalorie (kcal) expenditure recorded during that time, and body weight (kg). Physical activity during pregnancy was defined as the average metabolic equivalent unit value obtained over 48 hours at both 20 and 32 weeks gestation. Maternal height was measured at the first laboratory visit. Maternal weight gain between 20 and 32 weeks of pregnancy was based on body weight measured in our laboratory at each time point. Birth weight, gestational age at delivery, and other birth measures were abstracted from medical records. Each participant’s obstetrician determined gestational age during a routine prenatal visit using ultrasonography and the date of the woman’s last menstrual period. The research team performed no independent assessment of gestational age. Fetal growth ratio was calculated for each infant by dividing actual birth weight by the median birth weight for the gestational week from curves established by Zhang and Bowes,17 adjusting for gender, race, and parity. The fetal growth ratio is a unit-free measure of growth relative to other neonates of similar sex, parity, and gestational age. Furthermore, it takes into account scaling, because 100 g deviation from the “norm” may be more meaningful for an infant of close to 2,500 g than for one of close to 4,000 g.
We used a series of linear regression analyses to estimate the independent and combined association of physical activity, maternal weight gain, and maternal height on fetal growth ratio. Because timing of events is crucial in determining causality, we ran regression analyses using physical activity measured at 20 weeks of gestation (which preceded our maternal weight gain measures) and at 32 weeks gestation. We divided physical activity levels during pregnancy into quartiles and compared the infants’ birth weights among the four groups using analysis of variance. We estimated the multiplicative interaction between maternal height and physical activity on fetal growth ratio after centering both these variables at their respective means. To visually determine the source of the interaction, we dichotomized maternal height at the population median of 1.65 m and estimated the association between physical activity during pregnancy and fetal growth ratio within height strata.
At baseline (20 weeks of gestation) women were (mean±standard deviation [SD]) 29.1±5.3 years of age, 1.64±0.07 m in height, 68.9±9.8 kg in weight, with a body mass index BMI of 25.4±3.5 (Table 1). Between 1997 and 1999, all women delivered healthy infants (birth weight range 2,743–4,943 g), with 29% of the women delivering infants weighing more than 4,000 g (n=15, 4,300±316 g). Gestational age at delivery was (mean±SD) 39.4±1.21, range 36–42 weeks, with 2% at less than 37 weeks (n=1, 36 weeks). No women were electively induced. One woman delivered before 32 weeks, and two women were on bed rest for premature contractions at the time of the 32-week visit; these three women were not included in our analyses. Among the women included in our analyses (N=51) 47 had vaginal delivery, and four had cesarean delivery (Table 1). No other complications (eg, gestational diabetes mellitus, preeclampsia, abruption) were reported. Women enrolled in the study sought obstetric care in the surrounding area of the institution where the study was completed, comprising 36 different physicians (obstetricians and family physicians). The approximate altitude above sea level in the area where the study was conducted is 262 meters.
Aerobic physical activity, estimated by Caltrac accelerometry, was (mean±SD, range) (1.53±0.27, 1.20–2.45 metabolic equivalent units) and (1.45±0.28, 1.1–2.3 metabolic equivalent units) at 20 and 32 weeks, respectively. The pairwise correlation between the two time points was r=0.73. Physical activity during pregnancy (averaged at both time points: 20 and 32 weeks of gestation) measured by Caltrac accelerometry was inversely related to fetal growth ratio (r=−0.42, P<.002, Table 2, Fig. 1) and unrelated to gestational age (r=0.06, not significant). The strength of this association was unchanged whether physical activity was assessed at 20 weeks of gestation (r=−0.39, P<.004) or 32 weeks of gestation (r=−0.39, P<.005, Table 2, Fig. 1). Fetal growth ratio decreased across quartiles of physical activity levels in pregnancy (df: 3, 47, F=3.60, P<.02), with a one-quartile change in maternal physical activity associated with an average 203 g change in birth weight (Table 3). Similar associations between physical activity and fetal growth ratio were observed for estimates derived using the heart rate monitor and physical activity diary (r=−0.40, r=−0.42), respectively.
Adjustment for maternal weight gain and height attenuated the estimate from −0.20 to −0.13 fetal growth ratio per metabolic equivalent unit/day (P<.02, Table 2). Maternal weight gain from weeks 20–32 was not related to physical activity at either point in pregnancy, and it did not modify the association of physical activity with fetal growth ratio. Maternal height modified the association of physical activity during pregnancy with fetal growth ratio (test for interaction, P<.001). The inverse association of activity with fetal growth was essentially limited to women in the upper half of the height distribution in our sample (Fig. 2). Taller women did not differ from shorter women with respect to physical activity during pregnancy or maternal weight gain (Table 4). Among shorter women, physical activity was not associated with fetal growth.
Aerobic physical activity during mid-to-late pregnancy accounted for 18% of the variance in fetal growth ratio in this relatively small and homogenous group of women completing healthy pregnancies. Both maternal weight gain and height were also related to fetal growth ratio. Collectively, maternal weight gain during pregnancy, maternal height, and physical activity during pregnancy explained 41% of the variance in fetal growth ratio. Maternal weight gain and height were not related to physical activity during pregnancy, but height modified the association between physical activity during pregnancy and fetal growth. The association of physical activity during pregnancy with fetal growth ratio was restricted to the taller (more than 1.65 m) women in the sample.
One possible interpretation of this interaction is that fetal growth may be sufficiently constrained by maternal size in shorter women, so that physical activity during pregnancy has less opportunity to impact on fetal growth. In taller women, with fewer constraints on fetal growth, physical activity during pregnancy may be more important than maternal size in setting limits on birth weight. This concept is supported by Figure 2, which shows that offspring of shorter women, whether sedentary or active, and offspring of physically active, taller women have fetal growth ratios of about 1.0, whereas the offspring of sedentary, taller women have fetal growth ratios averaging about 1.20. Physical inactivity may thus be a risk factor for excessive fetal growth in taller women. Lim and colleagues19 found that high birth weight was related to several complications during labor (eg, shoulder dystocia, perineal trauma, increased labor duration, emergency cesarean delivery, postpartum hemorrhage), indicating that physical inactivity in tall women may be a potential cause of labor difficulties.
The correlation (r=−0.42) between physical activity during pregnancy and the infant’s fetal growth ratio is higher than others have reported for either cigarette smoking (approximately −0.25) or for maternal weight gain (approximately 0.16),20–23 both of which are considered important, potentially modifiable determinants of birth weight.24 Our ability to demonstrate a strong and significant relationship between daily physical activity during pregnancy and fetal growth ratio may reflect the precision and accuracy with which we were able to measure maternal physical activity in our cohort, in contrast to many previous studies.25,26
The infants born to our middle-class study participants were healthy and had birth weights at and above the population average. Consequently, our results cannot be generalized to the problem of low birth weight. Gestational age, maternal smoking, alcohol consumption, nutrition, parity, maternal height, paternal height, third-trimester hemoglobin concentration, sex of the infant, marital status, race, and familial factors have been shown to affect birth weight.15,27–31 Our sample comprised almost exclusively white, college-educated, and nonsmoking women, and because we controlled for parity, gender, race, and gestational age through the calculation of fetal growth ratio, it is unlikely that the relationship between physical activity during pregnancy and fetal growth ratio was confounded by these variables. We did not ascertain maternal alcohol intake or diet. Maternal undernutrition appears to have a restrictive effect on birth weight only when caloric intake is severely reduced,32–34 which was unlikely in our sample. Paternal height and third-trimester hemoglobin may also contribute to birth weight,27,28 but we have neither measurement available to us. To explain our results, these two variables would have to be associated both with maternal height and with physical activity. We have no reason to suspect that women who may be married to taller men differ in physical activity from women married to shorter men. Although Nahum and Stanislaw28 found that lower third-trimester hemoglobin was associated with a small increase in birth weight, other research35 found no difference in hemoglobin between very physically active and sedentary pregnant women, even when the active women were slightly taller. Lastly, we were unable to determine the role of prepregnant obesity on maternal weight gain and fetal growth ratio. Cogswell et al36 found that the prevalence of high birth weight almost doubled among obese women who gained 13.7 kg or more compared with those who gained between 6.8 and 8.6 kg during pregnancy. Future studies that examine the role of maternal height, weight gain, and physical activity on birth outcome should account for prepregnancy weight. Finally, our study focused on aerobic physical activity (with the most commonly recorded forms being walking and running); therefore, the results cannot be applied to other forms of isometric exercise during pregnancy (eg, stretching, yoga, weight training).
Our results indicate that taller women who engage in regular physical activity during pregnancy may deliver lighter babies within the lower bound of normal birth weight range. Our findings may have implications for studies evaluating birth weight and chronic disease risk during adulthood because physical activity during pregnancy has not been considered in these investigations. Birth weight differences of the magnitude of those between the lowest and highest physical activity quartile groups in our study (approximately 600 g) have been associated with differences in blood pressure.2 Our findings make plausible the idea that physical activity and maternal height may influence the association between lower birth weight and higher cardiovascular disease risk. To the extent that childhood obesity is associated with both greater maternal height and lesser propensity to physical activity in adulthood, we might anticipate seeing more women in the group of tall, inactive women whose babies have fetal growth ratios of 1.2, possibly increasing the risks of difficult labors and cesarean deliveries.
1. Barker DJ, Osmond C, Golding J, Kuh D, Wadsworth ME. Growth in utero, blood pressure in childhood and adult life, and mortality from cardiovascular disease. BMJ 1989;298:564–7.
2. Barker DJ. Fetal origins of coronary heart disease. BMJ 1995;311:171–4.
3. Bibby E, Stewart A. The epidemiology of preterm birth. Neuro Endocrinol Lett 2004;25 suppl:43–7.
4. Kramer MS. The epidemiology of adverse pregnancy outcomes: an overview. J Nutr 2003;133 suppl:1592S–6S.
5. 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.
6. Hatch MC, Shu XO, McLean DE, Levin B, Begg M, Ruess L, et al. Maternal exercise during pregnancy, physical fitness, and fetal growth. Am J Epidemiol 1993;137:1105–14.
7. Lewis RD, Yates CY, Driskell JA. Riboflavin and thiamin status and birth outcome as a function of maternal aerobic exercise. Am J Clin Nutr 1988;48:110–16.
8. Magann EF, Evans SF, Newnham JP. Employment, exertion, and pregnancy outcome: assessment by kilocalories expended each day. Am J Obstet Gynecol 1996;175:182–7.
9. Clapp JF 3rd, Capeless EL. Neonatal morphometrics after endurance exercise during pregnancy. Am J Obstet Gynecol 1990;163:1805–11.
10. Clapp JF 3rd, Lopez B, Harcar-Sevci R. Neonatal behavioral profile of the offspring of women who continued to exercise regularly throughout pregnancy. Am J Obstet Gynecol 1999;180:91–4.
11. Collings CA, Curet LB, Mullin JP. Maternal and fetal responses to a maternal aerobic exercise program. Am J Obstet Gynecol 1983;145:702–7.
12. Marquez-Sterling S, Perry AC, Kaplan TA, Halberstein RA, Signorile JF. Physical and psychological changes with vigorous exercise in sedentary primigravidae. Med Sci Sports Exerc 2000;32:58–62.
13. Pivarnik JM, Stein AD, Rivera JM. Effect of pregnancy on heart rate/oxygen consumption calibration curves. Med Sci Sports Exerc 2002;34:750–5.
14. Stein AD, Rivera JM, Pivarnik JM. Measuring energy expenditure in habitually active and sedentary pregnant women. Med Sci Sports Exerc 2003;35:1441–6.
15. Nahum GG, Stanislaw H, Huffaker BJ. Accurate prediction of term birth weight from prospectively measurable maternal characteristics. J Reprod Med 1999;44:705–12.
16. Rush D, Davis H, Susser M. Antecedents of low birth weight in Harlem, New York City. Int J Epidemiol 1972;1:375–87.
17. Zhang J, Bowes WA 3rd, Birth-weight-for-gestational-age patterns by race, sex, and parity in the United States population. Obstet Gynecol 1995;86:200–8.
18. Klesges RC, Klesges LM, Swenson AM, Pheley AM. A validation of two motion sensors in the prediction of child and adult physical activity levels. Am J Epidemiol 1985;122:400–10.
19. Lim JH, Tan BC, Jammal AE, Symonds EM. Delivery of macrosomic babies: management and outcomes of 330 cases. J Obstet Gynecol 2002;22:370–4.
20. Catalano PM, Thomas AJ, Huston LP, Fung CM. Effect of maternal metabolism on fetal growth and body composition. Diabetes Care 1998;21 suppl:B85–90.
21. Kariniemi V, Rosti J. Maternal smoking and alcohol consumption as determinants of birth weight in an unselected study population. J Perinat Med 1988;16:249–52.
22. Haddow JE, Knight GJ, Palomaki GE, Kloza EM, Wald NJ. Cigarette consumption and serum cotinine in relation to birthweight. Br J Obstet Gynaecol 1987;94:678–81.
23. Secker-Walker RH, Vacek PM, Flynn BS, Mead PB. Estimated gains in birth weight associated with reductions in smoking during pregnancy. J Reprod Med 1998;43:967–74.
24. Institute of Medicine. Energy requirements, energy intake, and associated weight gain during pregnancy. In: Nutrition during pregnancy: Part I. Weight gain, Pt. II. Nutrient supplements. Washington, DC: National Academies Press; 1990. p. 137–75
25. Hatch MC, Stein ZA. Work and exercise during pregnancy: epidemiological studies. In: Artal-Mittelmark R, Wiswell RA, Drinkwater BL, editors. Exercise in pregnancy. 2nd ed. Baltimore (MD): Williams & Wilkins; 1991. p. 279–91.
26. Pivarnik JM. Potential effects of maternal physical activity on birth weight: brief review. Med Sci Sports Exerc 1998;30:400–6.
27. Nahum GG, Stanislaw H. Relationship of paternal factors to birth weight. J Reprod Med 2003;48:963–8.
28. Nahum GG, Stanislaw H. Hemoglobin, altitude and birth weight: does maternal anemia during pregnancy influence fetal growth? J Reprod Med 2004;49:297–305.
29. Dougherty CR, Jones AD. The determinants of birth weight. Am J Obstet Gynecol 1982;144:190–200.
30. Kramer MS. Intrauterine growth and gestational duration determinants. Pediatrics 1987;80:502–11.
31. Briend A. Maternal physical activity, birth weight and perinatal mortality. Med Hypotheses 1980;6:1157–70.
32. Stein Z, Susser M, Rush D. Prenatal nutrition and birth weight: experiments and quasi-experiments in the past decade. J Reprod Med 1978;21:287–99.
33. Stein AD, Zybert PA, van de Bor M, Lumey LH. Intrauterine famine exposure and body proportions at birth: the Dutch Hunger Winter. Int J Epidemiol 2004;33:831–6.
34. Tafari N, Naeye RL, Gobezie A. Effects of maternal undernutrition and heavy physical work during pregnancy on birth weight. Br J Obstet Gynaecol 1980;87:222–6.
35. Pivarnik JM, Mauer MB, Ayres NA, Kirshon B, Dildy GA, Cotton DB. Effects of chronic exercise on blood volume expansion and hematologic indices during pregnancy. Obstet Gynecol 1994;83:265–9.
36. Cogswell ME, Serdula MK, Hungerford DW, Yip R. Gestational weight gain among average-weight and overweight women: what is excessive? Am J Obstet Gynecol 1995;172:705–12.
© 2007 The American College of Obstetricians and Gynecologists
Figure. No caption available.