Effects of Human Pregnancy and Aerobic Conditioning on Alveolar Gas Exchange During Exercise

625

Effects of human pregnancy and aerobic conditioning on alveolar gas exchange during exercise

Sonja E. McAuley, Dennis Jensen, Michael J. McGrath, and Larry A. Wolfe

Abstract: This study examined the effects of aerobic conditioning during the second and third trimesters of human pregnancy on ventilatory responses to graded cycling. Previously sedentary pregnant women were assigned randomly to an exercise group (n = 14) or a nonexercising control group (n = 14). Data were collected at 15–17 weeks, 25–27 weeks and 34–36 weeks of pregnancy. Testing involved 20 W·min–1 increases in work rate to a heart rate of 170 beats·min–1 and (or) volitional fatigue. Breath-by-breath ventilatory and alveolar gas exchange measurements were compared at rest, a standard submaximal V&O and peak exercise. Within both groups, resting VV& , & , and V /T increased signifi- 2 & & & EA T I & & cantly with advancing gestation. Peak work rate, O2 pulse (VOHR2/), VVEA, respiratory rate, VT/TI, VVOCO2 , 2, and the ventilatory threshold (Tvent) were increased after physical conditioning. Chronic maternal exercise has no significant effect on pregnancy-induced changes in ventilation and (or) alveolar gas exchange at rest or during standard submaximal exercise. Training-induced increases in Tvent and peak oxygen pulse support the efficacy of prenatal fitness programs to improve maternal work capacity. Key words: human gestation, respiration, chronic exercise. Résumé : La présente étude a examiné les effets de l’activité aérobique sur la réponse ventilatoire à un effort gradué sur bicyclette durant les deuxième et troisième trimestres de la grossesse. Des femmes enceintes initialement sédentai- res on été réparties aléatoirement dans un groupe entraîné (n = 14) ou dans un groupe témoin non entraîné (n = 14). Les données ont été recueillies entre les semaines 15–17, 25–27, et 34–36 de la grossesse. L’épreuve a consisté en des augmentations de 20 W·min-1 de l’intensité de l’exercice à une fréquence cardiaque de 170 battements·min-1 et (ou) fatigue volontaire. Les mesures cycle à cycle des échanges gazeux alvéolaires et ventilatoires ont été comparées au repos, & & & à une VO2 sous-maximale normale et à l’exercice maximal. Chez les deux groupes, les VVEA, ,etVT/VI ont augmenté significativement avec l’évolution de la gestation. L’intensité maximale, le pouls d’O (V&OF/),laV& ,laV& ,lafré- & & 2 2 C E A For personal use only. quence respiratoire, VT/TI,laVO2 ,laVCO2, et le seuil ventilatoire (Svent) ont augmenté après l’entraînement. L’exercice de longue durée n’a pas d’effet significatif sur les variations induites par la grossesse des échanges gazeux ventilatoires et/ou alvéolaires au repos ou durant un exercice sous-maximal normal. Les augmentations induites par l’entraînement du Svent et du pouls d’O2 maximal confirment l’efficacité des programmes d’exercice prénatal pour améliorer la capa- cité d’adaptation de la mère à l’effort. Mots clés : gestation humaine, respiration, exercice de longue durée. [Traduit par la Rédaction] McAuley et al. 633

Introduction Virtually all maternal physiological control systems are affected by pregnancy and the changes in these systems are Human pregnancy involves critical anatomical and physio- interactive. The ventilatory effects of pregnancy are particu- logical changes that must occur in a specific time sequence larly striking. These include changes in diaphragm and chest to support fetal growth and development while maintaining wall configuration (Crapo 1996), lung volumes and capaci- maternal homeostasis. These changes are mediated by gesta- ties (McAuliffe et al. 2002), and substantial increases in & tional hormones, initiated early in the first trimester, and tidal volume (V ), minute ventilation (V ), alveolar ventila- & T E they may alter maternal exercise tolerance (Wolfe et al. tion (VA), and the ventilatory equivalents for oxygen

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by MCGILL UNIVERSITY on 12/08/11 & & & & 1989; Wolfe et al. 2005). (VVE / O2) and carbon dioxide (VVE / CO2) at rest and during

Received 21 September 2004. Published on the NRC Research Press Web site at http://cjpp.nrc.ca on 03 August 2005. Reposted on the web site with correction on 05 August 2005. S.E. McAuley, and D. Jensen.1 School of Physical and Health Education, Queen’s University, Kingston, ON K7L 3N6, Canada. M.J. McGrath. Department of Obstetrics and Gynecology, Queen’s University, Kingston, ON K7L 3N6. Canada. L.A. Wolfe.2 School of Physical and Health Education, and Departments of Physiology and Obstetrics and Gynecology, Queen’s University, Kingston, ON K7l 3N6. Canada. 1Corresponding author (e-mail: [email protected]). 2Deceased.

exercise (Ohtake and Wolfe 1998; Wolfe et al. 1994). This Experimental design results in a partly compensated respiratory alkalosis with re- Qualified subjects entered the study on a staggered time ductions in arterial carbon dioxide tension (PaCO2 basis and were randomly assigned to either an exercise ~30− 34 mmHg) and plasma bicarbonate concentration, and group (EG) or control group (CG) using a randomized/block increases in arterial pH (7.42–7.49) and arterial oxygen ten- procedure with 4 subjects per block (Popcock 1984). Drop- sion (PaO2 ~100–106 mmHg) (Wolfe et al. 1998). outs for medical reasons or poor compliance from either An important theoretical consequence of the aforemen- group were replaced with the next available qualifying sub- tioned changes is a reduction in maternal ventilatory reserve. & & ject until a total of 14 subjects were accrued for each group. In this regard, values for (VVE / O2) are increased (Wolfe et al. At entry, each subject attended an information session to be- 2003; Ohtake and Wolfe 1998; Wolfe et al. 1994; Pivarnik et & come familiar with the laboratory, study procedures, and cy- al. 1993) and higher peak values are observed for VE at max- cle ergometer testing protocol. imal exercise (Wolfe et al. 2003; Lotgering et al. 1998; The EG participated in a closely monitored prenatal exer- Ohtake and Wolfe 1998; Wolfe et al. 1994) in late gestation. cise program that included both aerobic and moderate mus- Since maximum voluntary ventilation is either unchanged or cular conditioning components, whereas the CG only moderately reduced (Berry et al. 1989), the capacity to in- & performed moderate muscle conditioning exercises designed crease VE in the transition from rest to exercise is reduced. to improve muscular fitness without stressing or causing im- Controlled studies in nonpregnant subjects have reported & & provement in the aerobic energy system. Both conditioning reductions in ventilatory demand (VV/ O ), improvements in E 2 & programs were conducted during the second (TM2) and third ventilatory efficiency, reduced CO output (VCO ), and re- & 2 2 (TM3) pregnancy trimesters. Members of both groups partic- duced VE at a given work rate after physical conditioning ipated in physiological testing between 14 and 17 week ges- (Hoogeveen 2000; Taylor and Jones 1979). It is reasonable tation (entry), at the end of TM2 (25–27 weeks) and TM3 to assume that similar beneficial effects may be experienced (34–36 weeks). Procedures included maternal anthropometric by pregnant women. However, only limited information ex- measurements, assessment of forced vital capacity (FVC), ists regarding the effect of aerobic conditioning on respira- and a graded cycle ergometer test. Qualified medical person- tory and alveolar gas exchange responses to exercise in nel obtained data on pregnancy outcome during labor and human pregnancy (Ohtake and Wolfe 1998; Wolfe et al. delivery and neonatal morphometrics were measured within 1994; Pivarnik et al. 1993). 24 h of delivery. The present study employed a controlled randomized design to examine the effects of a closely monitored prenatal exercise program on maternal aerobic work capacity and Physical conditioning programs measures of ventilation and alveolar gas exchange during As described above, the EG participated in both aerobic standard submaximal exercise in healthy human pregnancy. and muscle conditioning (minimum of 2 days·week–1), We hypothesized that physical conditioning would improve whereas the CG performed only muscle conditioning (mini- For personal use only. aerobic working capacity, increase the ventilatory anaerobic mum 1 day·week–1) in accordance with guidelines from the threshold (Tvent), and attenuate respiratory responses during Society of Obstetricians and Gynaecologists of Canada standard submaximal exercise. (SOGC) and the Canadian Society for Exercise Physiology (Davies et al. 2003; Wolfe and Davies 2003; PARmed-X for Pregnancy 2002; Kochan–Vintinner 1999). For the EG, stair- stepping (Stairmaster 4000CT) was employed as modality Methods for aerobic exercise. Aerobic exercise intensity was prescribed and monitored on an individual basis by an experienced in- Subjects structor using modified pulse rate target zones for pregnant Pregnant subjects were recruited via posted announce- women and Borg’s (1982) rating of perceived exertion ments, newspaper advertisements, and contact with local ob- (RPE) scale (Kochan-Vintinner 1999). Exercise duration was –1 stetricians and health care providers in Kingston, Ontario. increased progressively from 15 to 30 min·session during –1 Prospective subjects were screened for medical contraindica- the TM2 and held constant at 30 min·session during the tions to exercise by the physician or midwife monitoring TM3 (Kochan-Vintinner 1999). Logs of each exercise ses- their pregnancies. A standard form developed in this labora- sion were kept by the instructor and included information on tory for the Canadian Society for Exercise Physiology was attendance, exercise heart rate (HR), RPE, and duration. employed for this purpose (PARmed-X for Pregnancy 2002). This information was averaged for each subject in both TMs Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by MCGILL UNIVERSITY on 12/08/11 Specific exclusion criteria also included the following: regu- to characterize the training stimulus. lar participation in strenuous occupational or recreational Muscle conditioning for both the EG and CG consisted of physical activity (> 2 days·week–1) verified by questionnaire, moderate resistance exercises in accordance with published smoking before or during pregnancy, presence of maternal prenatal fitness guidelines (Kochan-Vintinner 1999). Low re- obesity (BMI > 27 kg·m–2) or eating disorders, maternal age sistance elastic bands were utilized where appropriate to pro- > 40 y, parity > 2, twin or multiple pregnancy, or taking vide resistance and rest periods were used to maintain a HR medications other than prenatal vitamins. The study protocol less than 110 beats·min–1 (as verified by regular pulse rate and consent form were approved by the Research Ethics checks) to avoid aerobic conditioning effects. Warm-up and Board, Faculty of Health Sciences, Queen’s University, and cool-down procedures were incorporated as part of both aer- each subject provided written consent. obic and muscular conditioning sessions.

Basic measurements culated from PETCO2 using the equation of Jones et al. Anthropometric data collected at each measurement time (1979) agrees closely with simultaneously collected included body height (cm) and body mass (kg). The sum of arterialized PCO2 measurements at rest and during exercise in healthy pregnant and non-pregnant women. Both work 7 skinfolds (triceps, biceps, subscapular, costal, suprailiac, & front thigh, and suprapatellar) was recorded to monitor the rate and O2 pulse (VOHR2 /)at peak exercise were used as effects of advancing gestation and aerobic conditioning on indices of maternal aerobic working capacity. subcutaneous body fat (Taggart et al. 1967). FVC was also measured in the sitting posture (Pneumoscan, model S-301) Minimum sample size calculation & to verify that none of the subjects had obstructive lung dis- VO2 at Tvent was identified as the most important variable ease and to monitor possible changes in vital capacity during since it is determined by both metabolic and respiratory the study. functions and is an important indicator of aerobic working capacity in healthy subjects (Brooks et al. 1996). Minimum α Exercise testing sample size was calculated based on 80% power and level Exercise tests were performed 1 hour after subjects con- of 0.05 using a paired subject formula for comparison of sumed a standard meal (350 kcal; 20% protein, 20% fat, means (Keppel and Wickens 2004; Press et al. 1992). An es- 60% carbohydrate) to ensure that the results were not altered timate of Tvent (± SD) in healthy pregnant women was ob- by variations in maternal nutritional status. In addition, tests tained from an earlier study from this laboratory (Wolfe et al. 1994): 1.25 ± 0.20 L·min–1. Using this formula, the criti- for each subject were conducted at approximately the same –1 time of day to minimize circadian effects. cal sample size needed to detect a 12% (0.15 L·min ) All tests were conducted using a constant work rate cycle change in Tvent was identified as n = 9. Therefore, a sample ergometer (Sensor Medics, model 800s, Yorba Linda, Calif.). size of 14 subjects per group was adequate for this study. The protocol involved a 5-min period of resting (pre- exercise) data collection with subjects seated quietly on the Statistical analyses cycle ergometer; subjects then cycled for 4 min at a constant The effects of aerobic conditioning and advancing gesta- work rate of 20 W with a pedaling cadence between 60 and tion on physical characteristics, as well as respiratory, meta- 80 revolutions· min–1. This was followed by a ramp increase bolic, and alveolar gas exchange responses during rest, in work rate of 20 W·min–1 until a HR of 170 beats·min–1 exercise below Tvent, and at peak exercise were analyzed us- and (or) volitional fatigue. Heart rate was recorded both ing a 2-way ANOVA (group vs. measurement time) for re- electrocardiographically and with Polar Electro Vantage XL peated measures. When significant F ratios were observed, heart rate monitors (Polar, Finland). pairwise comparisons were made using Tukey’s honestly Breath-by-breath respiratory gas exchange was measured significant difference post-hoc test. Student’s t statistics for continuously during the protocol described above and for 15 dependent samples were used to compare physical condi- min immediately following exercise. For this purpose, a tioning variables during TM2 and TM3. For personal use only. computerized system (First Breath Inc., St, Agatha, Ont.) A Mann-Whitney U test for independent samples was em- that incorporates a respiratory mass spectrometer (Perkin- ployed to determine significant differences between the EG Elmer MGA 1100, Marquette Electronics, Milwaukee, Wis.), and CG for parity at entry, as it was not considered a contin- and a bi-directional volume turbine was used (Interface As- uous variable. Values are expressed as mean ± SE. Results sociates, VMM 110, Laguna Niguel, Calif.). The mass spec- for all statistical tests were considered significant if p < trometer and volume turbine were calibrated before each 0.05. test. Electrical signals from this equipment were converted from analog-to-digital and stored on a microcomputer. As Results described by Hughson et al. (1991) breath-by-breath gas exchange variables were calculated using the algorithm of Bea- Subjects & ver et al. (1981). Oxygen uptake (VO2) at Tvent was A total of 47 women volunteered to participate in the calculated using the V-slope method (Beaver et al. 1986). study and were assigned randomly to participate as members Resting values for metabolic, ventilatory, and alveolar gas of the EG or CG as described above. Reasons for dropout in- exchange variables were taken as the average values for 2–3 cluded lack of time or poor compliance (EG, n =7;CG:n = min of data collection. Similarly, 30-s averages of breath-by- 3), delivery before TM3 testing (EG, = 1), miscarriage (EG, breath data obtained during and following exercise were n=1), development of mild hypertension (EG, n=2; CG, computed for each test. Data were compared within and be- n=2), anemia (EG, n = 1), and spotting (EG, n = 2). Thus, Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by MCGILL UNIVERSITY on 12/08/11 & –1 tween groups at rest, a standard VO2 100 mL·min below data were available from 14 women in both the EG and CG, Tvent identified for each subject at entry and at peak exercise. respectively. Variables examined at each measurement time included: HR, Mean values for age (EG, 28.9 ± 0.8 vs. CG, 30.9 ± 0.9 & & & & VVOCO, , respiratory exchange ratio (RER), VV, respira- yr), gestational age (EG, 17.1 ± 0.6 vs. CG, 17.0 ± 22 & & & & EA tory rate (f), VT, VVEE/ OO22, VV/ , inspiratory time (TI), mean 3.5 week), height (EG, 163 ± 2 vs. CG, 164 ± 1 cm), and inspiratory flow (VT/TI), and end-tidal oxygen (PETO2) and pre-pregnant body mass (EG, 66.1 ± 3.1 vs. CG, 66.0 ± carbon dioxide (PETCO2) tensions. Arterial carbon dioxide 2.4 kg) were not significant different between groups. Mean tension (PaCO2) was calculated from PETCO2 using the gestational ages of the EG vs. CG at the time of exercise equation of Jones et al. (1979). Studies in our laboratory tests were not significantly different: entry, 18.8 ± 0.8 vs. (unpublished observations) have confirmed that PaCO2 cal- 19.0 ± 0.9 weeks; TM2, 28.2 ± 0.4 vs. 28.2 ± 0.3 weeks;

Table 1. Effects of advancing gestational age and maternal aerobic conditioning on physical characteristics during the study. Entry TM2 TM3 Variable Group (14–17 weeks) (25–27 weeks) (34–36 weeks) Body mass (kg) EG 70.4±3.2 76.3±3.0* 80.2±2.8*,† CG 71.0±2.5 77.3±2.5* 80.8±2.7*,† Sum of 7 skinfolds (mm) EG 148±13 167±112* 155±12 CG 163±11 174±12 177±11 FVC (L) EG 3.9±0.2 4.0±0.3 4.0±0.2 CG 4.0±0.2 4.1±0.2 4.1±0.1 –1 ,§ Peak O2 pulse (mL·beat ) EG 10.4±0.5 12.0±0.5* 12.5±0.5* CG 10.3±0.5 10.8±0.5 10.9±0.5 Peak work rate (W) EG 144±7 164±7* 173±8*,†,§ CG 149±6 149±6 148±7 & –1 ,§ ,†,§ VO2 at Tvent (L·min ) EG 1.30±0.3 1.47±0.4* 1.57±0.2* CG 1.22±0.4 1.30±0.4* 1.31±0.3* Note: Values are mean ± SE. EG, exercise group (n = 14); CG, control group (n = 14); TM, trimes- & & ter; FVC, forced vital capacity; VO2/HR, Peak O2 Pulse; VO2 at Tvent, oxygen uptake at the ventilatory threshold.*p < 0.05 vs. entry;†p < 0.05 vs. TM2; §p <0.05 vs. CG.

TM3, 36.0 ± 0.3 vs. 36.3 ± 0.2 weeks. All pregnant subjects Metabolic and heart rate responses & & delivered normal healthy infants. None of the babies were As expected, resting VO2 and VCO2 increased with ad- premature (< 37 week), or had low birth weights (< 2500 g). vancing gestation in both the EG and CG (Table 2). By de- & Mean birth weights were not significantly different between sign, the standard VO below T was kept constant in both & 2 vent& the EG and CG (3502 ± 96 vs. 3464 ± 158 g). groups. VCO2 at the standard VO2 below Tvent was signifi- As expected, maternal body mass increased significantly cantly higher in the EG vs. CG at entry, TM2, and TM3. & with advancing gestational age (Table 1). Significant training- Within the EG, VCO decreased significantly from entry to 2 & induced increases in O2 pulse and the work rate achieved at TM3 and from TM2 to TM3. RER at the standard VO2 be- peak exercise were observed between entry and both TM2 low Tvent decreased significantly from entry to TM3 in both and TM3 and from TM2 to TM3, respectively (Table 1). groups and from entry to TM2 in the EG. Within the EG, & & Mean values for both O pulse and work rate at peak exer- VO and VCO at peak exercise increased from entry to both 2 2 2 & & cise were significantly greater in the EG vs. CG at TM3. TM2 and TM3 and from TM2 to TM3. Peak VO and VCO & 2 2 For personal use only. Significant increases in VO2 at Tvent were observed within were also significantly higher in the EG vs. CG at TM3. A both the EG and CG between entry and TM2 and TM3, re- significant reduction in RER at peak exercise was observed & spectively (Table 1). Significant increases in VO2 at Tvent between entry and TM3 within the EG. were observed within the EG between TM2 and TM3. Mean Neither advancing gestation nor physical conditioning had & & values of VO2 at Tvent were significantly higher at both TM2 an effect on resting HR (Table 2). HR at the standard VO2 (EG, 13% vs. CG, 6.5%) and TM3 (EG, 21% vs. CG, 7.5%) below Tvent decreased significantly from entry to TM3 in the in the EG vs. CG. Furthermore, the magnitude of change in EG and peak HR was significantly higher in the EG vs. CG & VO2 at Tvent was greater in the EG vs. CG between entry and at TM3. TM2 (EG, 13% vs. CG, 6.5%) and TM3 (EG, 21% vs. CG, Within both groups at entry, TM2, and TM3, the HR 7.5%), respectively. achieved at end-exercise was slightly less than the 170 beats· min–1 end-point used to define “peak exercise” (Table 2). Aerobic Conditioning Records This effect was attributed to a number of individual subjects Members of the EG participated in aerobic/muscular con- within both groups that stopped because of volitional fatigue ditioning for a mean duration of approximately 17 weeks and (or) poor motivation. In this regard, 57% (n = 8), 64% (9.2 ± 0.6 week in TM2; 7.8 ± 0.6 week in TM3). Heart rate (n = 9), and 50% (n = 7) of all subjects within the EG; and during steady-state stair-stepping exercise corresponded to 71% (n = 10), 71% (n = 10), and 93% (n = 13) of all subjects within the CG stopped exercise at a HR < 170 approximately 75% of age-predicted HRmax and were similar –1 Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by MCGILL UNIVERSITY on 12/08/11 in TM2 and TM3 (146 ± 2 vs. 150 ± 2 beats·min–1). How- beats·min at entry, TM2, and TM3. However, peak O2 ever, steady-state RPE values were significantly higher in pulse, which corrects for small differences in the peak HR TM3 than TM2 (3.5 ± 0.1 vs. 3.1 ± 0.1). By design, the achieved (Wolfe et al. 1994), was calculated and compared mean duration of aerobic exercise was significantly greater within and between groups to evaluate the effects of advanc- in TM3 vs. TM2 (29.4 ± 0.3 vs. 22.4 ± 0.7 min). ing gestation and chronic aerobic conditioning on maternal aerobic working capacity. Members of the EG participated in an average of 3.0 ± 0.1 aerobic/muscular conditioning sessions·week–1 in TM2 and 2.7 ± 0.1 sessions·week–1 in TM3. Members of the CG partici- Respiratory measurements at rest –1 & pated in 1.7 ± 0.3 muscular conditioning sessions·week in In both groups, VE increased significantly between entry –1 TM2 and 1.5 ± 0.4 sessions·week in TM3, and all were in- and TM3 and from TM2 to TM3; VT/TI increased signifi- volved in an average of at least 1 session·week–1 in both TMs. cantly from entry to both TM2 and TM3; f was significantly

Table 2. Effects of advancing gestational age and maternal aerobic conditioning on metabolic variables at rest, below Tvent and peak exercise. Entry TM2 TM3 Variable Group (14–17 weeks) (25–27 weeks) (34–36 weeks) Rest HR (beats·min–1) EG 93±2 92±1 96±3 CG 89±3 92±3 93±3 & –1 ,†,§ VO2 (L·min ) EG 0.366±0.014 0.391±0.013 0.445±0.018* CG 0.340±0.011 0.375±0.019* 0.400±0.013* & –1 ,† VCO2 (L·min ) EG 0.316±0.012 0.341±0.012 0.394±0.014* CG 0.305±0.014 0.328±0.014 0.361±0.014*,† RER EG 0.861±0.014 0.896±0.01 0.894±0.014 CG 0.897±0.019 0.883±0.012 0.909±0.007 Below Tvent HR (beats·min–1) EG 137±2 134±2 129±3* CG 132±3 135±3 130±2 & –1 VO2 (L·min ) EG 1.20±0.05 1.21±0.05 1.21±0.05 CG 1.12±0.06 1.13±0.07 1.13±0.06 & –1 § § ,†,§ VCO2 (L·min ) EG 1.22±0.05 1.16±0.05 1.11±0.05* CG 1.14±0.07 1.13±0.07 1.09±0.07 RER EG 1.02±0.01 0.97±0.02* 0.93±0.02* CG 1.01±0.02 1.00±0.01 0.96±0.02* Peak exercise HR (beats·min–1) EG 168±1 168±1 170±1* CG 166±1 167±1 164±2 & –1 ,†,§ VO2 (L·min ) EG 1.73±0.09 1.97±0.08* 2.12±0.08* CG 1.70±0.09 1.80±0.09 1.79±0.08 & –1 ,†,§ VCO2 (L·min ) EG 2.10±0.10 2.30±0.11* 2.46±0.11* CG 2.04±0.11 2.14±0.11 2.11±0.09 RER EG 1.20±0.02 1.17±0.02 1.16±0.02* CG 1.20±0.01 1.20±0.02 1.18±0.02 Note: Values are mean ± SE. EG, exercise group (n = 14); CG, control group (n = 14); TM, trimes-

For personal use only. & & ter; HR, heart rate; VO2, oxygen uptake; VCO2, carbon dioxide output; RER, respiratory exchange ratio.*p < 0.05 vs. entry, †p < 0.05 vs. TM2, §p < 0.05 vs. CG.

Table 3. Effects of advancing gestational age and maternal aerobic conditioning on respiratory responses at rest. Entry TM2 TM3 Variable Group (14–17 weeks) (25–27 weeks) (34–36 weeks) & –1 ,† VE (L·min ) EG 10.4±0.5 11.5±0.4* 13.2±0.4* CG 10.4±0.6 11.0±0.4 12.2±0.5*,† f (breaths ·min–1) EG 16.6±0.9 19.4±1.0* 19.6±0.8* CG 16.1±1.3 16.9±1.1 18.4±1.3* VT (L) EG 0.69±0.04 0.65±0.05 0.74±0.04 CG 0.73±0.07 0.70±0.03 0.73±0.4 & & VVE / O2 EG 30.3±1.6 30.2±0.7 30.9±0.7 CG 30.6±1.2 30.4±0.9 30.7±0.8 VV& / &CO EG 37.4±4.0 33.2±0.7 37.0±2.3 Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by MCGILL UNIVERSITY on 12/08/11 E 2 CG 33.8±1.0 36.2±2.2 33.9±1.1 TI (s) EG 1.68±0.15 1.33±0.09* 1.33±0.08* CG 1.78±0.24 1.48±0.11 1.39±0.10* –1 ,† VT/TI (L·s ) EG 0.43±0.02 0.49±0.02* 0.56±0.02* CG 0.44±0.03 0.49±0.02* 0.54±0.02* Note: Values are means ± SE. EG, exercise group (n = 14); CG, control group (n = 14); TM, trimester; V& , minute ventilation; f, respiratory rate; V , tidal volume; VV& / &O , ventilatory equivalent for oxy- E& & T E 2 gen; VVE / CO2, ventilatory equivalent for carbon dioxide; TI, inspiratory time; VT/TI, mean inspiratory flow. *p < 0.05 vs. entry, †p < 0.05 vs. TM2.

Table 4. Effects of advancing gestational age and maternal aerobic conditioning on alveolar gas exchange at rest, below Tvent and peak exercise. Entry TM2 TM3 Variable Group (14–17 weeks) (25–27 weeks) (34–36 weeks) Rest & –1 ,† VA (L·min ) EG 8.1±0.4 9.0±0.3* 10.5±0.4* CG 8.1±0.5 8.5±0.3 9.6±0.4*,† PETCO2 (mmHg) EG 31.±0.7 31.7±0.4 31.0±0.4 CG 31.±0.8 31.3±0.6 30.8±0.4 PaCO2 (mmHg) EG 32.±0.6 32.7±0.4 31.9±0.4 CG 32.±0.7 32.2±0.5 31.7±0.3 PETO2 (mmHg) EG 1118±0.8 112.7±0.6 113.1±0.7 CG 1130±1.2 113.2±0.7 114.6±0.7 Below Tvent & –1 VA (L·min ) EG 29.2±1.5 29.6±1.7 28.3±1.8 CG 28.0±2.2 28.8±2.0 27.7±2.0 PETCO2 (mmHg) EG 36.2±0.7 35.7±0.6 34.6±0.6 CG 36.5±1.0 34.6±0.8 34.0±0.7 PaCO2 (mmHg) EG 35.3±0.6 34.8±0.5 33.9±0.6* CG 35.5±0.9 33.8±0.7* 33.4±0.5* PETO2 (mmHg) EG 111.7±0.8 109.9±0.8 109.8±1.2 CG 110.8±1.3 112.5±0.8 112.4±1.0 Peak exercise

& –1 VA (L·min ) EG 54.2±3.2 60.2±3.8 64.5±4.2* CG 54.0±4.5 61.2±4.2* 59.7±3.1 PETCO2 (mmHg) EG 35.2±1.0 35.7±1.0 35.2±1.0 CG 35.0±1.2 33.1±1.2* 33.1±0.9* PaCO2 (mmHg) EG 33.5±0.9 33.9±1.0 33.3±1.1 CG 32.9±1.1 31.2±1.0 31.2±0.8 § PETO2 (mmHg) EG 116.6±1.1 115.2±1.2 115.3±1.4 CG 116.6±1.1 118.2±1.2 118.6±0.8

For personal use only. Note: Values are mean ± SE. EG, exercise group (n = 14); CG, control group (n = 14); TM, trimes- & ter; VA, alveolar ventilation; PETCO2, end-tidal carbon dioxide tension; PaCO2, arterial carbon dioxide † § tension; PETO2, end-tidal oxygen tension. *p < 0.05 vs. entry, p < 0.05 vs. TM2, p < 0.05 vs. CG. & greater, and T significantly lower between entry and TM3 Similar changes in V and V /T were observed between en- I & & A T I (Table 3). Within the EG, resting VVEA, , f,TI, and VT/TI try and TM2 in the CG. Mean TI values in the TM3 were were significantly greater at both TM2 and TM3 vs. entry. significantly less in the EG vs. CG. & Within the CG, VT/TI increased significantly from entry to Significant increases in VA were observed between entry both TM2 and TM3. VT/TI was significantly greater at TM3 and TM2 and entry and TM3 within the EG and CG, respec- vs. TM2 in the EG. Both groups demonstrated a significant tively (Table 4). A significant decrease in P CO was ob- & ET 2 increase in V between entry and TM3 and from TM2 to served between entry and both TM2 and TM3 in the CG. A & TM3 (Table 4). Within the EG, VA was significantly greater Significant between-group differences were found for PETO2 at both TM2 and TM3 vs. entry. at TM3.

Respiratory measurements below T Discussion & & & vent Measures of V , f,V, VV/ O , and V /V were not signifi- E T E 2 T I This study tested the hypothesis that maternal aerobic cantly different within or between the EG and CG (Table 5). Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by MCGILL UNIVERSITY on 12/08/11 & & conditioning would increase aerobic working capacity and However, VV/ CO increased significantly from entry to E 2 attenuate pregnancy-induced respiratory adaptations to pro- TM2 within the EG, with no appreciable change thereafter. gressive exercise. Strengths of the study include its con- P CO and calculated P CO decreased significantly from ET 2 a 2 trolled longitudinal design, quantification of the exercise entry to TM3 within both groups, and from entry to TM2 in training program, and verification of an aerobic conditioning the CG (Table 4). effect within the EG. The exercise program was conducted in accordance with Respiratory measurements at peak exercise clinical practice guidelines published by the Society of Ob- & Significant increases in VE were observed between entry stetricians and Gynaecologists of Canada (SOGC) and the and both TM2 and TM3 within the EG (Table 6). Within the Canadian Society for Exercise Physiology (Davies et al. EG, f, and VT/TI increased significantly from entry to TM3. 2003; Wolfe and Davies 2003; PARmed-X for Pregnancy

Table 5. Effects of advancing gestational age and maternal aerobic conditioning on respiratory responses during exercise below Tvent. Entry TM2 TM3 Variable Group (14–17 weeks) (25–27 weeks) (34–36 weeks) & –1 VE (L·min ) EG 34.2±1.5 34.7±1.8 33.3±1.9 CG 32.7±2.5 33.6±2.3 32.8±2.4 f (breaths·min–1) EG 27.2±1.1 27.5±1.3 27.3±0.9 CG 25.4±1.9 25.9±1.7 26.8±2.0 VT (L) EG 1.31±0.08 1.33±0.11 1.28±0.10 CG 1.36±0.10 1.37±0.08 1.28±0.08 & & VVE / O2 EG 28.8±0.9 30.6±2.0 28.4±1.2 CG 29.1±1.1 30.0±1.1 29.0±1.1 & & VVE / CO2 EG 28.0±0.6 30.9±1.4* 30.1±0.9 CG 28.7±1.0 29.8±1.0 30.1±0.9 TI (s) EG 1.03±0.05 1.03±0.04 0.99±0.05 CG 1.09±0.08 1.09±0.06 1.03±0.07 –1 VT/TI (L·s ) EG 1.27±0.06 1.30±0.09 1.30±0.09 CG 1.28±0.09 1.27±0.08 1.29±0.09 Note: Values are mean ± SE. EG, exercise group (n = 14); CG, control group (n = 14); TM, trimester; V& , minute ventilation; f, respiratory rate; V , tidal volume; VV& / &O , ventilatory equivalent for oxy- E& & T E 2 gen; VVE / CO2, ventilatory equivalent for carbon dioxide; TI, inspiratory time; VT/TI, mean inspiratory flow. *p < 0.05 vs. entry.

Table 6. Effects of advancing gestational age and maternal aerobic conditioning on respiratory responses at peak exercise. Entry TM2 TM3 Variable Group (14–17 weeks) (25–27 weeks) (34–36 weeks) & –1 VE (L·min ) EG 61.9±3.3 68.7±4.2* 72.9±4.3* CG 63.0±4.4 68.8±4.7 67.3±3.5 f (breaths·min–1) EG 35.6±1.3 38.2±1.7 40.0±1.6* CG 33.2±1.7 36.6±2.2 35.7±1.7

For personal use only. VT (L) EG 1.76±0.08 1.83±0.11 1.85±0.11 CG 1.92±0.10 1.92±0.08 1.92±0.08 & & VVE / O2 EG 35.6±1.2 35.0±1.3 34.4±1.4 CG 37.1±1.6 38.7±1.8 37.7±1.3 & & VVE / CO2 EG 29.6±0.7 29.8±0.8 29.6±1.0 CG 30.9±1.2 32.2±1.7 31.8±1.0 § TI (s) EG 0.83±0.02 0.78±0.03 0.75±0.03 CG 0.89±0.04 0.84±0.05 0.86±0.05 –1 VT/TI (L·s ) EG 2.14±0.10 2.35±0.12 2.48±0.13* CG 2.20±0.14 2.38±0.16* 2.29±0.12 Note: Values are mean ± SE. EG, exercise group (n = 14); CG, control group (n = 14); TM, trimester; V& , minute ventilation; f, respiratory rate; V , tidal volume; VV& / &O , ventilatory equivalent for oxy- E& & T E 2 gen; VVE / CO2, ventilatory equivalent for carbon dioxide; TI, inspiratory time; VT/TI, mean inspiratory flow. *p < 0.05 vs. entry, §p < 0.05 vs. CG.

2002; Kochan-Vintinner 1999) and recently endorsed by the changes was much smaller. In this regard, the increase in & Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by MCGILL UNIVERSITY on 12/08/11 American College of Sports Medicine (2004). Training- VO at T was greater in the EG vs. CG between entry and & 2 vent induced increases in peak O2 pulse, peak work rate, and VO2 both TM2 (13% vs. 6.5%, respectively) and TM3 (21% vs. at Tvent confirm the effectiveness of these exercise guidelines 7.5%, respectively). In addition, Tvent was 13% and 20% to improve maternal aerobic working capacity. greater (p < 0.05) in the EG vs. CG at TM2 and TM3, re- The use of breath-by-breath technology permits detailed spectively. It is unlikely that the increase in Tvent observed examination of changes in alveolar gas exchange as well as within the CG is the result of the moderate muscular condi- & –1 an unbiased estimate of VO2 at Tvent. In agreement with our tioning program since exercising HRs < 110 beats·min earlier findings (Wolfe et al. 1994), significant training- were maintained for all subjects. As observed previously by induced increases in Tvent were observed between entry and others (Ohtake and Wolfe 1998; Webb et al. 1994; Wolfe et TM2 and TM3 within the EG. Parallel changes were also al. 1994; Pivarnik et al. 1991) it appears that pregnancy- observed within the CG; however, the magnitude of these induced weight gain stresses the aerobic energy system

causing a moderate improvement in maternal aerobic work- The present study employed an aerobic conditioning pro- ing capacity which is further accentuated by chronic mater- gram similar to that of Ohtake and Wolfe (1998). However, nal aerobic exercise. this study used a nonsteady-state testing protocol and sub- & Pregnancy influences ventilation both at rest and during maximal exercise responses were studied at a standard VO2 approximately 100 mL·min–1 below T at entry. Failure to exercise. The most notable changes include a significant in- vent & & observe a training-induced attenuation of VV/ O in the crease in VT with little or no change in f (Contreras et al. E 2 1991; Field et al. 1991). This results in an increase in resting present study may be the consequence of the nonsteady-state & protocol employed. In this regard, Casaburi et al. (1987) re- VE (Wolfe et al. 1998). In keeping with these results, we ob- & & & ported that post-conditioning reductions in VV/ O in served significant increases in resting VE with advancing ges- E 2 tation. However, this effect was the result of increased f with healthy non-pregnant subjects were mediated by reduced blood lactate responses to exercise. Consequently, the atten- little or no change in VT. Changes in f were accompanied by & & uating effect of physical conditioning on VV/ O is more significant reductions in TI such that VT/TI, an index of neu- E 2 prominent at higher work rates. Thus, training-induced re- ral inspiratory drive, increased between entry and TM2 and & & TM3 at rest within both the EG and CG. These findings, in ductions in VVE / O2 may not be expected during exercise at a & non-steady state work rate below T in which the appear- combination with the observation that resting VA increased vent with advancing gestation in both groups support the hypoth- ance of lactate in the blood may be delayed. Moreover, physical conditioning had no significant effect on either esis that pregnancy significantly increases ventilatory drive. & VCO or RER during both standard submaximal and peak A recent report from this laboratory suggests that preg- 2 exercise, respectively. These findings are consistent with a nancy-induced increases in ventilatory drive may be due, at previous investigation from this laboratory (Wolfe et al. least in part, to the effects of gestational hormones (namely 1994) that employed a similar physical conditioning pro- progesterone and estrogen) on both chemoreflex and non- gram and nonsteady-state testing protocol. chemoreflex drives to breathe (Jensen et al. 2005). Jensen et al. (2005) provided evidence that pregnancy-induced in- In summary, both pregnancy and advancing gestation re- creases in gestational hormone levels may increase resting sult in significant changes in ventilation and alveolar gas & exchange at rest and during exercise. However, physical con- VE via direct stimulation of non-chemoreflex drives to breathe thereby altering maternal blood gases and acid-base ditioning has no significant effect on ventilatory or alveolar gas exchange responses to standard submaximal exercise. status. As a result, the central ventilatory chemoreflex re- Nonetheless, moderate aerobic conditioning during the TM2 models to a higher sensitivity and lower threshold for CO , 2 and TM3 significantly increases maternal aerobic work tol- which further contributes to the control of breathing as preg- erance as reflected by significant increases in peak O pulse, nancy advances to term. & 2 peak work rate and VO2 at Tvent. These data support the effi- Jennings and associates (1994) have also highlighted the cacy and validity of the Society of Obstetricians and Gynae- importance of changes in plasma osmolality, the strong ion cologists of Canada and the Canadian Society for Exercise difference, angiotensin II, and arginine vasopressin in the For personal use only. Physiology prenatal fitness guidelines (Davies et al. 2003; chemical control of breathing. Since all of these factors are Wolfe and Davies 2003; PARmed-X for Pregnancy 2002; altered at rest and during exercise in pregnant women in di- & Kochan-Vintinner 1999) to improve maternal aerobic fitness. rections that would stimulate VE, we have hypothesized that Future studies should examine the effectiveness of regular these factors may also contribute to pregnancy-induced in- prenatal exercise to prevent and (or) treat harmful maternal- creases in ventilatory drive (Wolfe et al. 1998). fetal diseases such as gestational diabetes mellitus and pre- Previous studies have examined the effects of physical & & & & eclampsia. conditioning on VVE / CO2 and VVE / O2 relationships during standard submaximal exercise in pregnancy (Ohtake and Wolfe 1998; Wolfe et al. 1994; Pivarnik et al. 1993). Find- Acknowledgements ings from the present and past investigations from this laboratory (Ohtake and Wolfe 1998; Wolfe et al. 1994) confirm Financial support from the Canadian Fitness and Lifestyle that physical conditioning has no significant effect on Research Institute, Ontario Thoracic Society, and Natural & & VVE / CO2 relationships during standard submaximal and (or) Sciences and Engineering Research Council of Canada is peak exercise. gratefully acknowledged. A cross-sectional study by Pivarnik et al. (1993) reported & & significantly lower VVE / O2 values during submaximal cy-

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by MCGILL UNIVERSITY on 12/08/11 cling in a group of pregnant recreational athletes compared References with healthy sedentary pregnant controls. From this labora- American College of Sports Medicine. 2004. Sports medicine bulle- tory, Ohtake and Wolfe (1998) reported a significant post- tin: September/October [online]. Available from http://www.acsm. conditioning attenuation of pregnancy-induced increases in & & org/sports_medicine_bulletin_master/SMB_395/smb395.html VVE / O2 during exercise in late gestation. Subjects were stud- [cited 14 September 2004]. ied at standard work rates corresponding to maternal HR tar- Beaver, W.L., Lamarra, N., and Wasserman, K. 1981. Breath-by- –1 gets of 110, 130, and 150 beats·min during initial testing at breath measurement of true alveolar gas exchange. J. Appl. approximately 17–18 weeks gestation. Work rates used in Physiol. 51: 1662–1675. these initial pretraining tests were repeated during post- Beaver, W.L., Wasserman, K., and Whipp, B.J. 1986. A new training evaluations conducted at 26–27 and 37 weeks gesta- method for detecting anaerobic threshold by gas exchange. tion. J. Appl. Physiol. 60: 2020–2027.

Berry, M.J., McMurray, R.G., and Katz, V.L. 1989. Pulmonary and sponses to endurance cycle exercise in pregnant and postpartum ventilatory responses to pregnancy, immersion and exercise. women. Int. J. Sports Med. 19: 193–198. J. Appl. Physiol. 66: 857–862. McAuliffe, F., Kametas, N., Costello, J., Rafferty, G.F., Greenough, Borg, G. 1982. Psychophysiological bases of perceived exertion. A., and Nicolaides, K. 2002. Respiratory function in singleton Med. Sci. Sports Exerc. 14: 377–381. and twin pregnancy. Br. J. Obstet. Gynaecol. 109: 765–769. Brooks, G.A., Fahey, T.D., and White, T.P. 1996. Exercise physiol- Ohtake, P.J., and Wolfe, L.A. 1998. Physical conditioning attenu- ogy: human bioenergetics and its applications. 2nd ed. Mayfield, ates respiratory responses to steady-state exercise in late gesta- Mountain View, Calif. tion. Med. Sci. Sports Exerc. 30: 17–27. Casaburi, R., Starer, T.W., and Wasserman, K. 1987. Mediation of Physical Activity Readiness Medical Examination for Pregnancy reduced ventilatory response to exercise after endurance train- (PARmed-X for pregnancy). 2002. Canadian Society for Exer- ing. J. Appl. Physiol. 63:1533–1538. cise Physiology. Available from http://www.csep.ca/forms.asp Contreras, G., Gutierrez, M., Beroize, T., Fantin, A., Oddo, H., [cited 14 September 2004]. Villarroel, L., et al. 1991. Ventilatory drive and respiratory mus- Pivarnik, J.M., Lee, W., and Miller, J.F. 1991. Physiological and cle function in pregnancy. Am. Rev. Respir. Dis. 144: 837–841. perceptual responses to cycle and treadmill exercise during Crapo, R.O. 1996. Normal cardiopulmonary physiology during pregnancy. Med. Sci. Sports Exerc. 23: 470–475. pregnancy. Clin. Obstet. Gynecol. 39: 3–16. Pivarnik, J.M., Ayres, N.A., Mauer, M.B., Cotton, D.B., Kirshon, Davies, G.A.L., Wolfe, L.A., Mottola, M.F., and MacKinnon C. B., and Dildy, G.A. 1993. Effects of maternal aerobic fitness on 2003. Joint SOGC/CSEP clinical practice guideline: exercise in cardiorespiratory responses to exercise. Med. Sci. Sports Exerc. pregnancy and the postpartum period. Can. J. Appl. Physiol. 28: 25: 993–998. 329–341. Popcock, S.J. 1984. Clinical trials: a practical approach. John Field, S.K., Bell, S.G., Cenaiko, S.F., and Whitelaw, W.A. 1991. Wiley and Sons, N. Y. Relationship between inspiratory effort and breathlessness in Press, W.H., Flannery, B.P., Teukolsky, S.A., and Vetterling W.T. pregnancy. J. Appl. Physiol. 71: 1897–1902. 1992. Numerical recipes: the art of scientific computing. 2nd ed. Hoogeveen, A.R. 2000. The effect of endurance training on the Cambridge University Press, London, U. K. ventilatory response to exercise in elite cyclists. Eur. J. Appl. Taggart, N.R., Holliday, R.M., Billewicz, W.Z., Hytten, F.E., and Physiol. 82: 45–51. Thomson, A.M. 1967. Changes in skinfolds during pregnancy. Hughson, R.L., Northey, D.R., Xing, H.C., Dietrich, B.H., and Br. J. Nutr. 21: 439–451. Cochrane, J.E. 1991. Alignment of ventilation and gas fraction Taylor, R., and Jones, N.L. 1979. The reduction by training of CO for breath-by-breath respiratory gas exchange calculations in ex- 2 output during exercise. Eur. J. Cardiol. 9: 53–62. ercise. Comp. Biomed. Res. 24: 118–128. Jennings, D.B. 1994. Respiratory control during exercise: hor- Webb, K.A., Wolfe, L.A., and McGrath, M.J. 1994. Effects of acute and chronic maternal exercise on fetal heart rate. J. Appl. mones, osmolality, strong ions and PaCO2. Can. J. Appl. Physiol. 19: 334–339. Physiol. 77: 2207–2213. Jensen, D., Wolfe, L.A., Slatkovska, L., Webb, K.A., Davies, Wolfe, L.A., and Davies, G.A.L. 2003. Canadian guidelines for ex- G.A.L., and O’Donnell, D.E. 2005. Effects of human pregnancy ercise in pregnancy. Clin. Obstet. Gynecol. 46: 488–495.

For personal use only. on the ventilatory chemoreflex response to carbon dioxide. Am. Wolfe, L.A., Ohtake, P.J., and Mottola, M.F. 1989. Physiological J. Physiol: Regul. Integr. Comp. Physiol. 288: R1369–R1375. interactions between pregnancy and chronic exercise. Exerc. Jones, N.L., Robertson, D.G., and Kane, J.W. 1979. Difference be- Sport Sci. Rev. 17: 295–301. Wolfe, L.A., Walker, R.M.C., Bonen, A., and McGrath, M.J. 1994. tween end-tidal and arterial PCO2 in exercise. J. Appl. Physiol. 47: 954–960. Effects of pregnancy and chronic exercise on respiratory re- Keppel, G., and Wickens, TD. 2004. Design and analysis: a responses to graded exercise. J. Appl. Physiol. 76: 1928–1934. searchers handbook. 4th ed. Prentice Hall, Upper Saddle River, Wolfe, L.A., Kemp, J.G., Heenan, A.P., Preston, R., and Ohtake, N.J. P.J. 1998. Acid-base regulation and control of ventilation in hu- Kochan-Vintinner, A. 1999. Active living during pregnancy: physi- man pregnancy. Can. J. Physiol. Pharmacol. 76: 815–827. cal activity guidelines for mother and baby. Edited by L.A. Wolfe, L.A., Heenan, A.P., and Bonen, A. 2003. Aerobic condi- Wolfe, and M.F. Mottola. Canadian Society for Exercise Physi- tioning effects on substrate responses during graded cycling in ology, Ottawa, ON, Canada. Available from http://www.csep.ca/ pregnancy. Can. J. Physiol. Pharmacol. 81: 696–703. publicationsmain.html [cited 14 September 2004]. Wolfe, L.A., Charlesworth, S.A., Glenn, N.M., Heenan, A.P., and Lotgering, F.K., Spinnewijn, W.E.M., Struijk, P.C., Boomsma, F., Davies, G.A.L. 2005. Effects of pregnancy on maternal work and Wallenberg, H.C.S. 1998. Respiratory and metabolic re- tolerance. Can. J. Appl. Physiol. 30(2): 212–232. Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by MCGILL UNIVERSITY on 12/08/11

1. Brooke Kalisiak, Theresa Spitznagle. 2009. What Effect Does an Exercise Program for Healthy Pregnant Women Have on the Mother, Fetus, and Child?. PM&R 1:3, 261. [CrossRef] 2. DennisJensenD. Jensena(email: [email protected])Katherine A.WebbK.A. WebbbDenis E.O’DonnellD.E. O’Donnellb. 2007. Chemical and mechanical adaptations of the respiratory system at rest and during exercise in human pregnancy. Applied Physiology, Nutrition, and Metabolism 32:6, 1239-1250. [Abstract] [Full Text] [PDF] [PDF Plus] 3. Michael S Kramer, Sheila W McDonald, Michael S KramerAerobic exercise for women during pregnancy . [CrossRef] For personal use only. Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by MCGILL UNIVERSITY on 12/08/11