Human pregnancy involves a wide spectrum of metabolic and cardiopulmonary adaptations that may alter the capacity to exercise. Specific pulmonary adaptations to pregnancy in the resting state include altered lung volumes and capacities (6,16,24,33), augmented respiratory sensitivity to CO2(3,18,25), and both increased tidal volume(VT) and minute ventilation (˙VE) at rest and during exercise(1,11,16,21-23,35). The oxygen cost of breathing is also increased because of augmented diaphragmatic work of breathing (2). This effect may be partly compensated by reduced airway resistance caused by a progesterone-induced reduction in bronchial smooth muscle tone(7). The result of these adaptations is that arterial oxygen tension (PaO2) is increased, arterial carbon dioxide tension (PaCO2) is reduced to approximately 30-34 Torr(16,18), and arterial pH is increased to approximately 7.46 (21-24) in the resting state.
An important theoretical consequence of the changes described above is a reduction in maternal ventilatory reserve. Previous studies have documented increases in the ventilatory equivalent for oxygen(˙VE/˙VO2) during submaximal exercise(11,16,21,22,23,34) and higher peak values for ˙VE at maximal exercise(17,20,27) in late gestation. Data further suggest that maximum voluntary ventilation is unchanged or moderately decreased (4). Thus, the capacity to increase˙VE in the transition from the resting state to maximal exercise is reduced. This decrement in ventilatory reserve could also help to explain why dyspnea is a common complaint in late gestation(2,8,24).
The central purpose of this study was to test the hypothesis that physical conditioning attenuates both ventilatory demand and respiratory perception of effort during standard submaximal exercise. Controlled longitudinal studies of aerobic conditioning in pregnancy are few in number(29,34), and basic information is needed to guide future research. In healthy nonpregnant subjects, such conditioning enhances fat utilization (carbohydrate sparing effect) and reduces both carbon dioxide output and the drive to ventilate during standard submaximal exercise(5,12,13,30). Other factors which, in theory, could cause a reduction in ventilatory demand (as reflected by˙VE/˙VO2) include reduced respiratory sensitivity to carbon dioxide, changes in respiratory mechanics to reduce the rate of dead space ventilation, and improved ventilation:perfusion relationships (as reflected by VD/VT). If any of these effects are observed after physical conditioning in pregnant women, this would help to compensate for pregnancy-induced reductions in ventilatory demand during exercise and may reduce respiratory perception of effort and symptoms of breathlessness in late gestation.
Subjects. Pregnant subjects were recruited via newspaper ads, posted announcements, and contact with local obstetricians in Kingston, Ontario. Prospective subjects were screened medically by the obstetricians monitoring their pregnancies using a written questionnaire that included a checklist to exclude contraindications to exercise in pregnancy and other questions to verify the subjects' health history and current pregnancy status. Specific exclusion criteria included the following: smoking before or during pregnancy; regular participation in recreational or occupational physical activity before or during pregnancy; presence of metabolic or cardiopulmonary diseases or conditions; maternal obesity or eating disorders; presence of either absolute or relative contraindications to exercise in pregnancy; and taking medications other than prenatal vitamins. A sedentary lifestyle before and during early pregnancy was also confirmed by completion of a standard physical activity questionnaire (Canadian Standardized Test of Fitness, 3rd Ed.). Qualifying subjects then chose voluntarily to participate as members of a physical conditioning group (exercised group, N = 34) or a control group (N = 25) who remained sedentary. Written informed consent was obtained before participation in data collection. The study design and informed consent form were approved by an institutional human ethics committee in conformance with guidelines for human experimentation published by Medicine and Science in Sports and Exercise.
Study design. The experimental group participated in a supervised physical conditioning program during the second and third trimesters (TM) and detrained during the immediate 3 months postpartum period. The control group remained sedentary during the same time periods. Data collection was conducted at the start of the second trimester (ENTRY/preconditioning), after both the second and third trimesters (end of TM2, end of TM3/postconditioning), and 3 months postpartum (nonpregnant control). Gestational age was calculated as the time since the first day of the last menstrual period as identified by the subject and confirmed by her obstetrician as part of the process of medical clearance. Any errors in this process were detected by comparison with values for gestational age at delivery recorded by qualified medical personnel at delivery.
Subjects in both groups entered the study on a staggered time schedule and each was studied over a 9-month time span. This design provided two types of control data: (i) longitudinal measurements from the sedentary control group; and, (ii) data from the experimental group obtained in the nonpregnant deconditioned state at three months postpartum.
Physical conditioning program. The physical conditioning sessions were conducted 3 d·wk-1 by a qualified instructor and included both aerobic and muscular conditioning components. Aerobic conditioning consisted of upright stationary cycling at 75% of age-predicted maximal heart rate (140-150 beats·min-1). Exercise duration was increased during TM2 from 14 to 25 min per session and was held constant during TM3. The usual rate of progression was an increase of 2 min·wk-1 for the first 4 wk of conditioning, then 1 min·wk-1 to reach 25 min per session. Exercise records were kept by the class instructor for each subject and included attendance and cycling power output, steady-state heart rate(HR), and exercise duration for each conditioning session. At ENTRY, subjects from both groups were given the same standard advice (Health Canada Guidelines) on good nutrition during pregnancy and given strong encouragement to comply with this advice.
Measurements. Variables measured at each data collection time included anthropometric measurements, resting pulmonary function, and responses to graded submaximal exercise. Anthropometric characteristics included body height, body mass, and skinfold thicknesses (sites were biceps, triceps, subscapula, suprailium). Forced vital capacity (FVC, FEV1.0) was measured in the upright posture using a computerized spirometer (Cavitron, Model SC-20A).
Metabolic and cardiopulmonary functions were evaluated at rest in the sitting position and during three graded steady-state levels of upright cycle ergometer exercise using the Stage II protocol described in detail by Jones(14). Tests were conducted using a friction-type cycle ergometer (Monark, Model 868, Varberg, Sweden) using a pedaling cadence of 50 rpm. Each exercise level was 6 min in duration, with a brief (<5 min) rest between levels. Exercise power outputs for each subject were chosen to elicit steady-state heart rates of 110, 130, and 150 beats·min-1 in baseline tests conducted upon ENTRY to the study. The same exercise power outputs, representing light, moderate, and heavy steady-state work intensities, respectively, were used in subsequent assessments during pregnancy and postpartum. On ENTRY to the study subjects performed two exercise tests. The first was used to orient the subject to the testing procedures and to identify appropriate work rates to achieve the specific heart rate targets. The second was used to collect preconditioning baseline data.
At each 6-min exercise level, the subject breathed continuously through an Ellis-Lampman valve (dead space, 70 mL). Heart rate was measured electrocardiographically. After achievement of a cardiovascular steady-state(<5 beats·min-1 change in HR), oxygen uptake(˙VO2), carbon dioxide production (˙VCO2), and pulmonary ventilation (˙VE, f, VT) were determined by conventional open circuit spirometry using a Beckman Metabolic Measurement Cart calibrated with test gases of known fractional concentration. End-tidal carbon dioxide tension (PET CO2) was measured using a rapid-response infrared analyzer (Beckman LB-2, Anaheim, CA). Arterial carbon dioxide tension (PaCO2) was calculated from PETCO2 and VT using the equation of Jones et al.(15,26). Finally, respiratory perception of effort(RPEr) was determined using Borg's 15 point(6-20) scale (19,28).
Alveolar ventilation (˙VA) was calculated from the obtained data according to the following equation (14): ˙VA(L·min-1, BTPS) = ˙VCO2 (L·min-1, STPD× 863/PaCO2. Dead space ventilation per breath (VD) was calculated as: [(˙VE - ˙VA)/f] - valve dead space (70 mL). The rate of dead space ventilation (˙VD, L·min-1, BTPS) was then calculated as f × VD. The physiological dead space/tidal volume ratio(VD/VT) was employed as an estimate of the degree of matching of ventilation to perfusion during exercise (14). The ventilation equivalents for oxygen (˙VE/˙VO2) and carbon dioxide (˙VE/˙VCO2) were also determined, and the respiratory exchange ratio (RER) was calculated as an index of relative utilization of fat and carbohydrates.
Statistical analyses. Descriptive variables including mean age, parity, body height, and mass of the EG and CG on ENTRY to the study were compared using unpaired Student t statistics. Exercise records for the EG in the second and third trimesters were analyzed using paired Studentt-statistics. Changes in physical characteristics and responses to graded exercise within each group were evaluated by a one-way ANOVA for repeated measures with post-hoc analyses of significantF-ratios using paired Student t-statistics. To minimize Type I error, post-hoc analysis within groups were restricted to the following planned comparisons: ENTRY to TM2; ENTRY to the end of TM3; end of TM2 to the end of TM3; end of TM3 to PP. Where indicated, between-group effects were analyzed by calculating changes during the same time intervals. Changes in the EG versus CG were then compared for each time interval using unpaired Student t-statistics. Results of all statistical tests were considered significant if P < 0.05.
Subjects. Seven of the 34 original members of the EG did not complete the experimental protocol because of moving from the area(N = 4), medical complications clearly unrelated to exercise(N = 1), and lack of interest (N = 2). Of the 25 women in the CG, five did not complete the study because of mild hypertension in late gestation (N = 3), an obstetric complication (N = 1), and delivery before the third trimester test session (N = 1), respectively. This resulted in final sample sizes of 27 and 20 in the EG and CG, respectively.
On ENTRY to the study, final EG and CG members were similar with respect to mean age (30.2 ± 0.7 vs 28.0 ± 1.0 yr), parity (0.7 ± 0.2 in both groups), body height (164.4 ± 1.2 vs 165.3 ± 1.1 cm), the sum of four skinfold thicknesses (65 ± 4 vs 69 ± 5 mm), FVC(3.30 ± 0.07 vs 3.27 ± 0.10 L·min-1), FEV1.0 (2.85 ± 0.08 vs 2.88 ± 0.08 L·min-1), and FEV1.0/FVC × 100 (87 ± 1 vs 88± 2%). Mean weight gain between ENTRY and the end of TM3 was 10 kg in both groups. No significant changes were observed for either FVC or FEV1.0 either within or between groups during the study. Mean gestational age at the time of exercise tests were also similar between groups. Values in the EG versus CG were: ENTRY, 17.0 ± 0.4 vs 18± 0.5 wk; end of TM2, 26.1 ± 0.2 vs 27.0 ± 0.2 wk; end of TM3, 37.3 ± 0.2 vs 37.2 ± 0.1 wk; PP, 13.1 ± 0.3 vs 14.0± 0.5 wk.
Mean work rates required to elicit steady-state heart rates of approximately 110, 130, and 150 beats·min-1 during exercise tests conducted at ENTRY were also similar between groups. Those for the CG were 17 ± 2, 42 ± 3, and 67 ± 3 W, and those for the EG were 21 ± 2, 46 ± 2 and 68 ± 3 W, respectively. These low work rates provided further confirmation of the sedentary state and low physical fitness of subjects in both groups at ENTRY(15).
Physical conditioning records. Analysis of physical conditioning records confirmed that EG members participated in physical conditioning for a mean duration of 20 wk (8.3 ± 0.3 wk during TM2; 11.7 ± 0.5 wk during TM3). HR during steady-state cycling was approximately 75% of agepredicted HR max. Mean exercise HRs were similar in TM2 and TM3 (146± 1 vs 147 ± 1 beats·min-1). Mean cycling duration(min/session) was significantly higher during TM2 vs TM3 (17.9 ± 0.3 vs 24.3 ± 0.2 min). All subjects reached 25 min per session by the end of the TM2 or early in TM3.
Heart rate responses. Within the CG (Fig. 1A), HR was significantly higher during pregnancy versus PP at rest and at the two lowest exercise levels. This effect was not observed at the highest exercise level. As expected, HR at rest and during exercise tended to increase with advancing gestational age in the CG, but the changes did not reach statistical significance. Within the EG (Fig. 1B), achievement of an aerobic conditioning response was confirmed by significant reductions in HR between ENTRY and both TM2 and TM3 (the period of physical conditioning) at all three exercise levels. This effect was not observed in the resting state and the magnitude of training-induced HR reduction increased as a function of the exercise intensity. Changes in HR between ENTRY and both TM2 and TM3 were significantly different in the EG versus CG at all 3 exercise levels.
Ventilatory demand. As expected, ventilatory demand (as reflected by ˙VE/˙VO2) was significantly greater in late gestation(end of TM3) compared with the nonpregnant state (PP) both at rest and at all three exercise levels in both groups (Figs. 2A and 2B). However, different patterns of change were observed in the control group and exercised group with advancing gestational age. Within the CG(Fig. 2A), values for ˙VE/˙VO2 at all three exercise levels were higher in late gestation (end of TM3) compared with those observed at the beginning and the end of the second trimester(ENTRY, end of TM2, respectively). Conversely, in the EG(Fig. 2B), significant increases in˙VE/˙VO2 were not observed between ENTRY and the end of TM3 (the period of physical conditioning) and the lowest values during pregnancy were observed at the end of TM2 (change from ENTRY to the end of TM2P < 0.05 at the highest exercise level). Absolute values at the end of both TM2 and TM3 were also systematically lower than those of the CG at each exercise level and changes from ENTRY to the end of both TM2 and TM3 (the period of physical conditioning) were significantly lower in the EG versus CG at the highest exercise level.
As shown in Table 1, f decreased moderately at all three work rates (P < 0.05 from ENTRY to the end of TM3 at the highest work rate) during the period of physical conditioning, whereas values tended to increase slightly with advancing gestational age in the CG(between group effects P < 0.05 for ENTRY to the end of TM3 at the lowest and highest work rates). In both groups, VT decreased significantly from the end of TM3 to PP. VT increased significantly from ENTRY to the end of TM3 within the EG at the lowest work rate.
Respiratory exchange ratio. As illustrated inFigure 3A, values for RER in the CG were significantly higher at PP compared with those at the end of TM3 (as well as compared with those at ENTRY and the end of TM2) at all three exercise levels. Nonsignificant trends existed for small increases in RER with advancing gestational age, but this effect reached statistical significance only between the end of TM2 and the end of TM3 in the resting state. In the EG, RER values during exercise at PP tended to be higher than those in late gestation, but this effect reached statistical significance only at the second exercise level. Values for RER during exercise decreased from ENTRY to the end of TM2(P < 0.05 at the second exercise level). Also, absolute values at the end of both TM2 and TM3 and PP also tended to be lower than those of the CG. Between group statistics approached but did not reach statistical significance for this effect.
Respiratory sensitivity to carbon dioxide. It is well established that evaluation of the relationship of ˙VE to ˙VCO2 is a valid model for the study of respiratory sensitivity during exercise(13,14). In both groups at all four measurement times, a positive linear relationship was observed for ˙VE expressed as a function of increasing ˙VCO2.˙VE/˙VCO2 was significantly higher in late gestation(end of TM3) versus PP at rest and at all three exercise levels in both groups(Fig. 4A and B). ˙VE/˙VCO2 also tended to increase with advancing gestational age in both groups. This effect was significant between ENTRY and TM3 at all three exercise levels and between the end of TM2 and the end of TM3 at the two lowest exercise levels within the CG. Within the EG, this effect was significant between ENTRY and the end of TM3 at the second exercise level, and from the end of TM2 and the end of TM3 at the highest exercise level. The increases in ˙VE during pregnancy and with advancing gestational age within both groups resulted from increases in VT with no significant change in f at all three exercise levels.
As shown in Figure 5A and B, values for PaCO2 were significantly reduced during pregnancy in both groups at all three exercise levels. Also in both groups, values tended to decrease slightly with advancing gestational age. In this regard, values decreased significantly between ENTRY and the end of TM3 in the CG at the highest exercise level. Within the EG, values decreased significantly between ENTRY and the end of TM3 at the lowest exercise level and between the end of TM2 and the end of TM3 at the second exercise level.
Alveolar/dead space ventilation. The relationship between˙VA and ˙VCO2 appears in Figure 6A and B. At all three exercise levels ˙VA/˙VCO2 values were significantly higher in both groups in late gestation (end of TM3) compared with those in the nonpregnant state (PP). Evidence also existed in both groups for an increase in ˙VA/˙VCO2 with advancing gestational age. Within the CG (Fig. 6A), significant increases in˙VA/˙VCO2 were observed from the end of TM2 to the end of TM3 at the second exercise level and from ENTRY to the end of TM3 at the highest exercise level. Within the EG (Fig. 6B), significant increases in ˙VA/˙VCO2 were observed from ENTRY to the end of TM3, and from the end of TM2 to the end of TM3 at the two lowest exercise levels.
Within the EG a significant reduction in ˙VA/˙VCO2 was observed from ENTRY to the end of TM2 (Fig. 6B). The only significant between group result was a significantly greater increase in˙VA/˙VCO2 between ENTRY and TM2 in the CG versus EG at the highest exercise level. Both of these findings were attributed to hyperventilation of some EG subjects at the highest exercise level during exercise tests conducted at ENTRY.
Within the CG, physiological dead space per breath (VD, L) was significantly greater in late gestation (end of TM3) compared with that in the nonpregnant state (PP) at all three work rates: 0.20 ± 0.01 versus 0.15± 0.01 L at Level 1; 0.25 ± 0.02 versus 0.17 ± 0.01 L at Level II; 0.27 ± 0.02 versus 0.19 ± 0.01 L at Level III. This effect was less evident in the EG since changes reached statistical significance only at the highest work rate (0.26 ± 0.02 vs 0.18± 0.01 L). However, between group effects did not reach statistical significance.
As shown in Figure 7A and B, the rate of dead space ventilation (˙VD, L·min-1) at any given˙VCO2 tended to be higher in late gestation (end of TM3) compared with that in the nonpregnant state (PP) in both groups. However, some evidence existed to suggest that this effect was greater in CG versus EG. In this regard, changes in ˙VD/˙VCO2 between the end of TM3 and PP were significant at all three exercise levels in the CG, whereas, in spite of greater statistical power, this effect was significant only at the highest exercise level in the EG (Fig. 7B). In addition, the reduction in ˙VD/˙VCO2 from the end of TM3 to PP was significantly greater in the CG versus EG at the lowest exercise level. Finally, the change from ENTRY to the end of TM3 reached statistical significance at the lowest exercise level in the CG (Fig. 7A).
No significant changes were observed either within or between groups for VD/VT (Fig. 8A and B). However, data trends that approached statistical significance were observed at all three exercise levels within the CG to increase in VD/VT with advancing gestational age (i.e., ENTRY to end of TM3, end of TM2 to end of TM3), and for higher values in late gestation (end of TM3) compared with those in the nonpregnant state (PP).
Respiratory Perception of Effort. Within the CG(Fig. 9A), no significant changes were observed for RPEr with advancing gestational age or in late gestation (end of TM3) compared with those in the nonpregnant state (PP). Conversely, significant reductions were observed at the two highest exercise levels within the EG(Fig. 9B) from ENTRY to the end of both TM2 and TM3 (the period of physical conditioning). Between group changes were significantly greater in the EG versus CG between ENTRY and the end of both TM2 and TM3 at the highest exercise level. No subjects in either group complained of severe shortness of breath or dyspnea either at rest or during exertion in association with any of the exercise tests.
Plotting of RPEr data as a function of ˙VE also produced significant results. Within the CG (Fig. 10A), values for RPEr/˙VE were significantly lower in late gestation (end of TM3) compared with those in the nonpregnant state (PP). Within the EG(Fig. 10B) significant reductions in RPEr/˙VE were observed from ENTRY to the end of TM3 at the two highest exercise levels and from ENTRY to the end of TM2 at the second exercise level, suggesting that respiratory effort at any given level of ˙VE was lower after physical conditioning. Changes in RPEr/˙VE between ENTRY and TM3 were also significantly greater in the EG versus CG at the two highest exercise levels.
The central focus of this study was to test the hypothesis that maternal aerobic conditioning attenuates pregnancy-induced increases in ventilatory demand (as reflected by ˙VE/˙VO2 during standard submaximal weight-supported exercise). Strengths of the study include its controlled longitudinal design, accurate quantification of the physical conditioning program, and verification of a substantial aerobic conditioning effect in the experimental (exercised) group. A nonrandomized study design was employed to promote compliance to the exercise program in the experimental group and to prevent cross-over in the control group during the 20-wk physical conditioning program. However, subjects were recruited from a homogenous middle socioeconomic population and were allowed to choose their group assignments only after meeting specific entry criteria. This resulted in experimental and control groups that were similar at study entry with respect to mean age, parity, body height, adiposity (as reflected by skinfold thicknesses), and aerobic fitness (as reflected by heart rate/power output relationships during exercise testing). Values for FVC and FEV1.0 were also similar between groups, and values above 80% for FEV1.0/FVC× 100 confirmed that no subjects in either group exhibited evidence of obstructive pulmonary disease. While we chose our experimental design based on practicality and historical precedent, we cannot rule out the possibility that some of our results might have been affected by the fact that random assignment of subjects to the two groups was not performed.
As expected, substantial increases in ˙VE/˙VO2 were observed within the CG during the third trimester at all three exercise levels. In contrast, values in the EG tended to decrease from ENTRY to the end of TM2 (P < 0.05 at the highest exercise level), and changes between ENTRY and the end of TM3 were not statistically significant. Between group statistics also confirmed that responses in the EG versus CG were significantly different from one another at the highest exercise level. Thus, our experimental hypothesis was strongly supported.
The observed attenuation of ventilatory demand in the EG versus CG (as reflected by ˙VE/˙VO2) must be explained by one or more of the following factors: attenuation of pregnancy-induced increases in respiratory sensitivity to CO2; reduced CO2 production and RER, resulting in reduced ventilatory drive; attenuation of pregnancy-induced increases in dead space ventilation (14). The study results did not support the hypothesis that pregnancy-induced increases in respiratory sensitivity are attenuated by physical conditioning. In both groups, positive linear relationships were observed between both˙VE and ˙VA and ˙VCO2 at each testing time. The effects of pregnancy and advancing gestational age on these relationships were also similar in direction and magnitude. Finally, values achieved for calculated PaCO2 were similar between groups at all three work rates at each testing time. These findings are consistent with those of a parallel study that examined the effects of a similar aerobic conditioning program on˙VE/˙VCO2 relationships during graded exercise in pregnancy (35).
The most logical explanation for reduced ventilatory demand after aerobic conditioning would be a reduced CO2 output at any given work rate or˙VO2 (i.e., lower RER) as a result of greater utilization of fat versus carbohydrate as a metabolic fuel(5,12,30). In this regard, values for RER at the two highest exercise levels decreased moderately between ENTRY and the end of TM3 in the EG, whereas corresponding values increased slightly in the CG. These data trends approached but did not reach statistical significance within or between groups. However, it is noteworthy that a parallel study of aerobic conditioning during the second and third TMs of pregnancy(35) resulted in significant reductions in RER at higher work rates (90, 120 W) during graded exercise testing. Thus, it seems probable that slightly greater fat versus carbohydrate utilization did, in fact, contribute to attenuation of pregnancy-induced increases in ventilatory demand following aerobic conditioning.
Evidence also existed to support the hypothesis that aerobic conditioning attenuates pregnancy-induced increases in the rate of dead space ventilation(˙VD, L·min-1). From a mathematical viewpoint, this appeared to be the result of lesser increases in the EG versus CG in dead space ventilation per breath (VD, L) at the two lowest work rates and a significant reduction in breathing frequency in the EG between ENTRY and the end of TM3 at the highest exercise level. Thus, the study results support the view that training-induced attenuation of increases in˙VE/˙VO2 during strenuous work in late gestation are the result of the combined effects of enhanced fat utilization (causing reduced CO2 output), attenuation of pregnancy-induced increases in VD per breath, and more efficient ventilatory mechanics. A training-induced reduction in respiratory sensitivity was not supported by our findings.
In agreement with previous reports (22,31), VD/VT was not increased significantly at any exercise level in either group, suggesting that ventilation:perfusion is well preserved in exercising pregnant women. However, values in late gestation in the CG were higher than those observed either in early gestation or in the nonpregnant state. This effect approached statistical significance and was not observed in the EG. Thus, a beneficial effect of physical conditioning to preserve ventilation:perfusion is not disproved by our results and should be examined in greater detail in future studies.
The present study is the first to report the interactive effects of advancing gestational age and aerobic conditioning on RPEr. Physiological factors that may increase respiratory work of breathing and RPEr at any given work rate during pregnancy include a small increase in ˙VO2, significant increases in both inspired and expired ˙VE, augmented VT/FVC, and increased involvement of the diaphragm and accessory respiratory muscles (8,19,34). A reduction in airway resistance that is attributed to a progesterone-induced reduction in bronchial smooth muscle tone may be helpful to offset these effects(10). Indeed, data from our control group indicated that RPEr is not significantly increased during submaximal weight-supported exercise at any gestational age. Furthermore, when RPEr data were expressed as a function of ˙VE, the values were significantly lower in late gestation compared with those in the nonpregnant state.
The present study results also demonstrated significant training-induced reductions in RPEr at any given absolute work rate and also when RPEr data were expressed as a function of ˙VE. These data support the hypothesis that aerobic conditioning may be useful to prevent or treat pregnancy-induced symptoms of breathlessness. Further study is therefore recommended to identify the mechanisms of reduced RPEr following physical conditioning in pregnant women. These may include an increase in maximum breathing capacity (so that a given level of ˙VE is perceived as less stressful), increased aerobic endurance of respiratory muscles, and more efficient respiratory mechanics. As discussed above, the present study provided some evidence for improvement of ventilatory mechanics as a factor that could contribute to this effect.
The present study results differ significantly from the earlier findings of South-Paul et al. (29). The authors reported the responses of 10 healthy women to a physical conditioning regimen conducted 3 d·wk-1 between the 20th and 30th week of gestation. A nonexercising control group (N = 7) was also studied. Physical conditioning sessions included both aerobic and muscular conditioning components. The aerobic conditioning component included stationary cycling with progressive increases in intensity (60-80% of “maximal” heart rate) and duration (not specified). A graded cycle ergometer test to volitional fatigue was conducted pre- and postconditioning and included evaluation of HR, ˙VO2, RER, ˙VE, f, VT, and ˙VE/˙VO2 of a standard power output of 75 W and at peak exercise. The only significant change after physical conditioning was a moderate increase (177 mL) in VT at peak exercise in association with a nonsignificant increase (225 mL·min-1) in O2 at peak exercise. No significant changes were observed for any exercise variable measured at 75 W in either group. Failure to achieve training-induced changes comparable with those of the present study may be a result of the use of a shorter (10 vs 20 wk) less intensive conditioning regimen and/or employment of smaller sample sizes.
Our results also differ significantly from those of Pivarnik et al.(23). The authors used a cross-sectional approach to compare the responses of 10 aerobically-conditioned women with those of six sedentary controls. Subjects in the conditioned group were recruited from running clubs, fitness centers, and other athletic organizations, whereas sedentary subjects were recruited from the patient population receiving routine prenatal care at a major hospital. Measurements were obtained in the resting state and during steady-state semi-recumbent stationary cycling using conventional open circuit spirometry at a pulse rate target of 140 beats·min-1. Exercise tests were repeated at approximately 25- and 36-wk gestation and 12-wk postpartum. As in the present study, conditioned pregnant subjects exhibited lower values (P < 0.05) during exercise for ˙VE/˙VO2 and VD/VT than sedentary subjects. However, in contrast to the present findings and those from another major study from this laboratory (35), substantially lower values were also observed for˙VE/˙VCO2 in conditioned versus sedentary pregnant subjects, suggesting reduced respiratory sensitivity in the former. Since evidence for increased respiratory sensitivity has not been reported in controlled longitudinal studies of physical conditioning in pregnant women(29,35), it appears that the conditioned subjects of Pivarnik et al. (23) may have been a selected group with special characteristics not entirely attributable to high levels of aerobic fitness.
In conclusion, the present study results support our original hypothesis that moderate aerobic conditioning during the second and third pregnancy trimesters reduces both ventilatory demand and RPEr during standard submaximal exercise. If ventilatory demand is reduced, then ventilatory reserve during standard submaximal exercise must be increased and the mother's ability to provide oxygenated blood to the fetus under stressful physiological conditions may be enhanced. From a clinical perspective, aerobic conditioning may also be useful to prevent or treat exertional dyspnea during pregnancy and may help to preserve maternal exercise tolerance.
The authors wish to thank Michele C. Amey for time and expertise in typing, editing and production of artwork for the final manuscript.
Financial support from the Advisory Research Committee (Queen's University), Fitness Canada, the Canadian Fitness and Lifestyle Research Institute, the Ontario Ministry of Health, and Natural Sciences and Engineering Research Council of Canada is gratefully acknowledged. P. J. Ohtake was supported by a Franklin Bracken fellowship (Queen's University) and a traineeship from the Heart and Stroke Foundation of Canada.
Address for correspondence: Larry A. Wolfe, Ph.D., FACSM, School of Physical and Health Education, Queen's University, Kingston, Ontario, Canada K7L 3N6.
1. Artal, R., R. Wiswell, Y. Romem, and F. Dorey. Pulmonary responses to exercise in pregnancy. Am. J. Obstet. Gynecol.
2. Bader, R. A., M. E. Bader, and D. J. Rose. The oxygen cost of breathing in dyspnoeic subjects as studied in normal pregnant women.Clin. Sci.
3. Bayliss, D. A. and D. E. Millhorn. Central neural mechanisms of progesterone action: application to the respiratory system.J. Appl. Physiol.
4. Berry, M. J., R. G. McMurray, and V. L. Katz. Pulmonary and ventilatory responses to pregnancy, immersion, and exercise. J. Appl. Physiol.
5. Casaburi, R., T. W. Starer, and K. Wasserman. Mediation of reduced ventilatory response to exercise after endurance training. J. Appl. Physiol.
6. Contreras, G., Gutiérrez, T. Beroiza, et al. Ventilatory drive and respiratory muscle function in pregnancy. Am. Rev. Resp. Dis.
7. Cugell, D. W., N. R. Frank, E. A. Gaensler, and T. L. Badger. Pulmonary function in pregnancy I. Serial observations in normal women. Am. Rev. Tuberc. Pulm. Dis.
8. Field, S. K., S. G. Bell, D. F. Cenaiko, and W. A. Whitelaw. Relationship between inspiratory effort and breathlessness in pregnancy. J. Appl. Physiol.
9. Gazioglu, K., N. L. Katrieder, M. Rosen, and P. N. Yu. Pulmonary function during pregnancy in normal women and in patients with cardiopulmonary disease. Thorax
10. Gee, J. G. L., B. S. Packer, J. E. Miller, and E. D. Robin. Pulmonary mechanics during pregnancy. J. Clin. Invest.
11. Guzman, C. A. and R. Caplan. Cardiorespiratory response to exercise during pregnancy. Am. J. Obstet. Gynecol.
12. Hanson, J. S. and W. H. Nedde. Long-term physical training effect in sedentary females. J. Appl. Physiol.
13. Jones, N. L. Use of exercise in testing respiratory control mechanisms. Chest
14. Jones, N. L. Clinical Exercise Testing
, 4th Ed. Philadelphia: W.B. Saunders, 1997.
15. Jones, N. L., D. G. Robertson, and J. W. Kane. Difference between end-tidal and arterial PCO2
in exercise. J. Appl. Physiol.
16. Knuttgen, H. G. and K. Emerson. Physiological response to pregnancy at rest and during exercise. J. Appl. Physiol.
17. Lotgering, F. K., M. B. Van Doorne, P. C. Struijk, J. Pool, and H. C. S. Wallenberg. Maximal aerobic exercise in pregnant women: heart rate, O2
production and ventilation.J. Appl. Physiol.
18. Lyons, H. A. and R. Antonio. The sensitivity of the respiratory center in pregnancy and after administration of progesterone.Trans. Assoc. Am. Physicians
19. Mahler, D. A. and M. B. Horowitz. Perception of breathlessness during exercise in patients with respiratory disease.Med. Sci. Sports Exerc.
20. McMurray, R. G., A. C. Hackney, V. L. Katz, M. Gall, and W. J. Watson. Pregnancy-induced changes in the maximal physiological responses during swimming. J. Appl. Physiol.
21. Pernoll, M. L., J. Metcalfe, P. A. Kovach, R. Wachtel, and M. J. Durham. Ventilation at rest and exercise in pregnancy and postpartum. Respir. Physiol.
22. Pivarnik, J. M., W. Lee, T. Spillman, S. L. Clark, D. B. Cotton, and J. F. Miller. Maternal respiration and blood gases during aerobic exercise performed at moderate altitude. Med. Sci. Sports Exerc.
23. Pivarnik, J. M., N. A. Ayres, M. B. Mauer, D. B. Cotton, B. Kirshan, and G. A. Dildy. Effects of maternal aerobic fitness on cardiorespiratory responses to exercise. Med. Sci. Sports Exerc.
24. Prowse, C. M. and E. A. Gaensler. Respiratory and acid-base changes during pregnancy. Anesthesiology
25. Regensteiner, J. G., W. D. Woodard, D. D. Hagerman, et al. Combined effects of female hormones and metabolic rate on ventilatory drives in women. J. Appl. Physiol.
26. Robbins, P. A., J. Conway, D. A. Cunningham, S. Khamnei, and D. J. Paterson. A comparison of indirect methods for continuous estimation of arterial PCO2
in men. J. Appl. Physiol.
27. Sady, S. P., M. W. Carpenter, P. D. Thompson, M. A. Sady, B. Maydon, and D. R. Coustan. Cardiovascular response to cycle exercise during and after pregnancy. J. Appl. Physiol.
28. Silverman, M., J. Barry, H. Hellerstein, J. Janos, and S. Kelsen. Variability in the perceived sense of effort in breathing during exercise in patients with chronic obstructive pulmonary disease. Am. Rev. Resp. Dis.
29. South-Paul, J. E., K. R. Rajagopal, and M. S. Tenholder. The effect of participation in a regular exercise program upon aerobic capacity during pregnancy. Obstet. Gynecol.
30. Taylor, R. and N. L. Jones. The reduction by training of CO2
output during exercise. Eur. J. Cardiol.
31. Templeton, A. and G. R. Kelman. Maternal blood gases.(PA
), physiological shunt and VD
in normal pregnancy. Br. J. Anaesth.
32. Webb, K. A., L. A. Wolfe, and M. J. McGrath. Effects of acute and chronic maternal exercise on fetal heart rate. J. Appl. Physiol.
33. Weinberger, S. E., S. T. Weiss, W. R. Cohen, J. W. Weiss, and T. S. Johnson. Pregnancy and the lung. Am. Rev. Resp. Dis.
34. Wolfe, L. A., P. J. Ohtake, and M. F. Mottola. Physiological interactions between pregnancy and aerobic exercise.Exerc. Sport Sci. Rev.
35. Wolfe, L. A., R. M. C. Walker, A. Bonen, and M. J. McGrath. Effects of pregnancy and chronic exercise on respiratory responses to graded exercise. J. Appl. Physiol.