Postural effects when cycling in late pregnancy
Article Outline
Summary
Aim
This study assessed if upright cycling is preferable to semi-recumbent cycling during pregnancy.
Method
Healthy women with low risk singleton pregnancies were tested at 34–38 weeks gestation. They cycled for 12
min, either semi-recumbent (45°, n=27) or upright (n=23), at 135–145beatsmin−1.
Results
When semi-recumbent, minute ventilation was greater (p<0.03) at rest and systolic blood pressure and pulse pressure were greater during exercise (p<0.05). Exercise maternal heart rate, oxygen consumption, oxygen consumption per kilogram, minute ventilation, cardiac output, stroke volume, mean and diastolic blood pressures and arterio-venous oxygen difference were posture-independent. All increased with exercise (p<0.01), except stroke volume when semi-recumbent (p>0.05). Small post-exercise fetal heart rate increases (by 8beatsmin−1, p<0.05) were similar in both postures (n=11 in each sub-group), with no adverse changes. Fetal heart rate accelerations and uterine activity (n=11 in each sub-group) were not influenced by posture or exercise.
Conclusions
(1) Neither posture had a distinct advantage. (2) Both postures were safe for short duration cycling. (3) The same target maternal heart rates are suitable for both postures because they resulted in similar oxygen consumptions and fetal heart rates.
Keywords: Exercise, Pregnancy, Heart rate, fetal, Semi-recumbent, Cycling, Blood pressure
Introduction
Supine position exercise is contraindicated after the first trimester.1 Partial inferior vena caval obstruction due to the gravid uterus can decrease venous return, thereby limiting preload, stroke volume and cardiac output,2 blood pressure (BP) and fetal blood supply. After being advised to avoid supine exercise, pregnant women often question whether semi-supine exercise has similar problems. In particular, is it safe to continue using a semi-supine bicycle or should they change to an upright bicycle?
The first issue to consider is whether semi-supine exercise has any disadvantages compared with upright cycling. Compared with postpartum values, some studies found that cardiac output during pregnancy increases with upright cycling,3, 4 but not with semi-supine cycling.5
Secondly, does semi-supine exercise have any potential advantages? Uterine activity is lower with recumbent than upright cycling.6 It has not been reported if the semi-recumbent posture has a similar benefit over upright cycling.
Finally, does the fetal response to cycling, as indicated by changes in baseline fetal heart rate (FHR), differ between postures? This is the most important arbiter of safety.
This study is the first to compare maternal cardiorespiratory, uterine activity and FHR responses to moderate intensity semi-supine and upright cycling. Maternal heart rate (MHR) was used as the index of exercise intensity because target MHRs are often used when prescribing exercise for pregnant women. Tests were conducted in late pregnancy, when the greatest postural effects due to the gravid uterus would be expected.
Methods
Subjects
Healthy non-smokers with low-risk singleton pregnancies were recruited from the general community by advertising in local newspapers. They had no absolute or relative contraindications to exercise.1 Apart from haematinics or vitamin supplements, they were not on any medications. All gave written informed consent to the protocol, which had been approved by the Royal North Shore Hospital Ethical Review Committee. Volunteers were randomly assigned (1:1) to two groups.
Procedure and measurements
Subjects exercised at 34–38 weeks of gestation. To minimize confounding factors, dry laboratory temperature was maintained between 21.5 and 23°C. Also, the women fasted for 2h beforehand so that recently ingested food would not cause varying amounts of blood diversion away from exercising muscles to the gastrointestinal tract.
All measurements were made when semi-supine for group 1 and when seated upright for group 2. In the 30min period prior to baseline measurements, each woman was familiarized with the equipment.
Group 1 exercised on a semi-supine (45°) bicycle for 12min. The cycling rate was maintained by a metronome. Resistance was increased until MHR was 135–145beatsmin−1. Group 2 exercised for the same duration on an upright cycle ergometer, with workload increased until MHR was within the same range. All subjects exercised for at least 10min at the target level.
Systolic and diastolic BP (Korotkoff phase IV) [mercury sphygmomanometer] were measured at rest and after 10min of exercise. MHR [electrocardiogram], oxygen consumption (VO2), minute ventilation, tidal volume and respiratory rate [in-house automated system] and cardiac output were measured at rest and after 6min of exercise when steady state was attained. Subjects breathed through a Hans Rudolph three-way valve [model 8600] connected in tandem with a two-way valve [model 2600]. Expired air passed through a pneumotachograph [Morgan Number 080] and mixing box. Mixed expired air was continuously sampled for carbon dioxide (CO2) [Datex CD 30] and oxygen (O2) [Servomex 580A] until steady state was obtained, i.e. a change of <0.1% in both over 1min.7
Cardiac output was measured by the CO2 rebreathing technique. This indirect Fick technique relies on accurate indirect estimation of the mixed venous and arterial partial pressure of CO2 (
). Mixed venous 
was estimated by the Collier method,8 without correction for the “downstream” effect. Arterial 
was estimated from end-tidal 
and during measurements made at rest, a forced expiration was used to collect end-tidal air.7, 9 The mixed venous and arterial 
values were converted to a mixed venous-arterial content CO2 difference (Cv-a CO2) using the McHardy–Jones formula, then corrected for each subject's resting haemoglobin concentration [American Optical Haemoglobinometer].7
The precision of this technique was established in our laboratory by comparison with measurements made by the established technique of thermodilution, in non-pregnant subjects. The coefficient of variation was 13.4% (±standard deviation [S.D.] 8.5%) at rest and 7.0% (±S.D. 3.8%) during moderate intensity exercise.
Stroke volume, arterio-venous O2 difference, mean BP, pulse pressure and VO2 per kilogram values were derived by standard formulae.
In the sub-groups used to measure FHR and uterine activity, a Hewlett-Packard 8040A cardiotocogram with autocorrelation was employed. The incidence of FHR accelerations (i.e. an increase of >15beatsmin−1 for >15s)10 was assessed during the 12min periods immediately before and after exercise. FHR was not measured during exercise because abdominal wall movement can interfere with cardiotocogram results.11 To enable FHR to be measured immediately post-exercise, when adverse FHR changes are most likely to occur,12 there was no formal cool-down period. However, subjects were advised to gently move their legs to prevent venous pooling. The incidence of uterine contractions causing an increase in uterine tone of 5mmHg or more for at least 15s was noted for the 12min period immediately before exercise and during exercise.
Statistical analyses
Intra-group exercise-related changes were assessed with paired Student t-statistics. The exception was FHR, where repeated measures of analysis of variance was used. When p was <0.05, Newman–Keul's multiple comparison test was applied post hoc to determine when FHR was different in recovery compared with baseline. Inter-group comparisons were made using independent two-sample Student t-statistics. Regression lines of individual values of cardiac output plotted against VO2 were also calculated for both groups. The level of statistical significance was set at p<0.05.
Results
Demographic and delivery data (Table 1)
The two groups were well-matched (p>0.05). All babies were healthy. They were born between 37 and 42 weeks of gestation, with birthweights above the tenth percentile for gestational age and 5min Apgar scores of eight or more.
Table 1. Demographic and delivery data (mean±standard deviation)
| Group 1 (n=27) | Group 2 (n=23) | |
|---|---|---|
| Primiparous:multiparous | 15:12 | 15:8 |
| Age (years) | 30 (5) | 30 (4) |
| Weight (kg) | 71 (9) | 69 (5) |
| Height (cm) | 165 (6) | 165 (6) |
| Gestation at test (weeks) | 36 (2) | 35 (1) |
| Trained:neither:sedentary | 5 (19%):6 (22%):16 (59%) | 5 (22%):3 (13%):15 (65%) |
| Birthweight (kg) | 3.4 (0.3) | 3.5 (0.5) |
| Gestation at delivery (days) | 279 (7) | 282 (6) |
Maternal cardiorespiratory parameters (Table 2)
At rest, the lower tidal volume (p<0.01) when semi-supine was more than offset by a faster respiratory rate (p<0.01), resulting in a greater minute ventilation (p<0.03).
Table 2. Comparison of maternal cardiorespiratory responses to semi-supine and upright cycling in late pregnancy (mean±standard deviation)
| Rest | Exercise | |||
|---|---|---|---|---|
| Semi-supine (n=27) | Upright (n=23) | Semi-supine (n=27) | Upright (n=23) | |
| Maternal heart rate (beatsmin−1) | 85 (15) | 84 (12) | 137 (15) | 142 (17) |
| O2 consumption (mLmin−1) | 207 (48) | 260 (50) | 1163 (251) | 1247 (279) |
| O2 consumptionkg−1 (mLmin−1kg−1) | 3.9 (0.7) | 3.9 (0.7) | 16.8 (3.9) | 18.2 (3.9) |
| Minute ventilation (Lmin−1) | 9.4 (1.7) | 8.3 (1.6)* | 38.1 (8.7) | 37.3 (10.3) |
| Tidal volume (mL) | 616 (124) | 715 (127)* | 1601 (460) | 1892 (514) |
| Respiratory rate (breathsmin−1) | 16 (4) | 12 (3)* | 25 (7) | 21 (7) |
| Cardiac output (Lmin−1) | 6.3 (1.3) | 5.9 (1.3) | 11.3 (1.7) | 12.3 (1.9) |
| Stroke volume (mL) | 76 (18) | 71 (15) | 82 (16) | 89 (18) |
| Systolic BP (mmHg) | 110 (9) | 107 (8) | 157 (13) | 146 (12)* |
| Diastolic BP (mmHg) | 68 (8) | 70 (6) | 75 (8) | 78 (7) |
| Mean arterial BP (mmHg) | 82 (7) | 82 (6) | 102 (8) | 101 (8) |
| Pulse pressure (mmHg) | 42 (9) | 37 (6) | 82 (13) | 68 (11)* |
| Arterio-venous O2 difference (mL100mL−1) | 44 (10) | 46 (11) | 105 (16) | 102 (17) |
All women completed the exercise test without any problems. Similar target MHRs were achieved in both groups. The increase in stroke volume was significant with upright cycling (18mL, ±S.D. 19, p<0.01) but not with semi-supine cycling (5mL, ±S.D. 18, p>0.05). All other parameters increased with exercise in both groups (p<0.01). Exercise values for systolic BP and pulse pressure were higher with semi-supine cycling (p<0.05). Maternal resting and exercise values were otherwise similar in both postures.
When regression lines of individual values of cardiac output plotted against VO2 were calculated for the two groups, the intercept (p=0.87) and slope (0.64) of the two lines were identical. Thus, when exercise cardiac output was corrected for the workload achieved, it was similar in both postures.
FHR response (Table 3)
The same intensity was achieved in sub-groups 1 and 2 (MHR 135, ±S.D. 11beatsmin−1 versus 136, ±S.D. 16beatsmin−1, respectively, p>0.05). Resting FHR and post-exercise increases in FHR were similar in both postures (p>0.05). FHR increases were not observed until 5min post-exercise and had not returned to resting values by 10min. The mean frequency of FHR accelerations and uterine activity were the same in both postures and did not change with exercise.
Table 3. Comparison of fetal heart rate and uterine activity responses to semi-supine and upright cycling in late pregnancy (mean±standard deviation)
| Semi-supine | Upright | |
|---|---|---|
| Sub-group 1 (n=11) | Sub-group 2 (n=11) | |
| FHR (beatsmin−1) | ||
| Rest | 136 (10) | 134 (9) |
| Post-exercise change: at 0–1min | 4 (10) | 4 (8) |
| Post-exercise change: at 5min | 8 (10)* | 7 (5)* |
| Post-exercise change: at 10min | 7 (7)* | 8 (11)* |
| FHR accelerations (accelerationsmin−1) | ||
| Rest | 0.3 (0.2) | 0.4 (0.2) |
| Post-exercise change | −0.1 (0.3) | 0.0 (0.3) |
| Uterine activity (contractionsmin−1) | ||
| Rest | 0.02 (0.04) | 0.06 (0.11) |
| During exercise | 0.07 (0.11) | 0.07 (0.14) |
| Change during exercise | 0.05 (0.11) | 0.02 (0.10) |
Discussion
The present study is the first to show that semi-supine cycling can achieve similar stroke volume and cardiac output values to upright cycling, even in late pregnancy. There is a paucity of studies on non-pregnant adults for comparison. In boys, a parallel study13 showed that cardiac index was qualitatively higher with 45° semi-supine cycling than with upright cycling (11.95Lmin−1m−2 versus 9.51Lmin−1m−2, respectively) but this may have been due to differences in heart rates (198beatsmin−1 versus 182beatsmin−1, respectively). At 140–150beatsmin−1, boys were reported to have the same stroke volume and cardiac output in both postures but the semi-recumbent (70° tilt) position tested was not very different to the upright position.14
Before postulating that the failure for stroke volume to increase with semi-supine cycling is due to mild inferior vena caval obstruction, further research is required. It may be physiological. Even in non-pregnant subjects, increases in stroke volume with semi-supine cycling are not as marked as with upright cycling.13 The range of fitness levels seen was similar in both groups and so was unlikely to have contributed to this postural difference. Furthermore, a smaller parallel study of semi-supine cycling at 138beatsmin−1 at 36 weeks of gestation found that stroke volume increased qualitatively in both sedentary women (by 19mL, n=6) and physically active women (by 23mL, n=10).5 It is not clear why they found a greater increase in stroke volume than our study. The CO2 rebreathing technique that we used was described in detail to highlight that it was very similar to their method.
Reassuringly, cardiac output with semi-supine cycling was high enough to cause the same mean and diastolic BPs as upright cycling. Mean and diastolic BPs are important because they correlate with fetoplacental perfusion during upright cycling.15 The effects of minute ventilation (for O2 intake), mean BP (for blood supply and O2 delivery to tissues) and a-v O2 difference (i.e. O2 extraction by tissues) combine to meet O2 demands (i.e. VO2). If blood supply was compromised with semi-supine cycling, a compensatory greater increase in a-v O2 difference would be expected, but this was not seen. Exercise minute ventilation, mean BP and a-v O2 difference were the same in both postures. BP was not limited with semi-supine cycling: systolic BP was actually greater than with upright cycling. This study also confirmed earlier findings that short duration, moderate intensity cycling is well-tolerated by the mother in either posture.4, 5, 16
Most importantly, this is the first report demonstrating that FHR responses to cycling are posture-independent. There were no potentially adverse indicators, such as fetal bradycardia or FHR deceleration to suggest that the babies were at risk. Small FHR increases are commonly seen post-exercise and are not problematic.17 FHR changes were not due to changes in uterine activity.
Another original finding is that posture did not influence uterine activity. Women prone to uterine contractions are unlikely to benefit if they change from an upright bicycle to a semi-supine bicycle.
There are several limitations. This is a single centre, parallel study with a small number of subjects. The duration tested was the shortest that would allow steady state to be achieved and measurements to be made, based on preliminary work. If any adverse postural differences occurred, we wanted them to be as short-lived as possible. Our findings should not be extrapolated to longer durations, higher intensities or multiple gestation (which might potentially cause more pronounced partial inferior vena caval obstruction when semi-supine). Future research is required to assess these conditions.
The American College of Obstetricians and Gynecologists recommends that pregnant women, without absolute or relative contraindications, should exercise for 30min on most days.1 If a pregnant woman wants to continue using a semi-recumbent bicycle, this could be achieved by combining 10min of cycling with other activities, such as walking. Alternatively, she could cycle for 10min, thrice daily. If she wanted to cycle for 30min continuously, an upright bicycle remains the first choice. Upright cycling for 30min has been shown to be safe for the mother and baby16 but maternal and FHR responses to moderate duration semi-supine cycling have not been reported.
Midwives and other exercise prescribers (such as doctors, physiotherapists, fitness leaders and personal trainers) often use target MHRs as the index of intensity because MHRs can be easily and repeatedly measured by pregnant women. The validity of using MHR is based on the linear relationship in pregnant subjects between MHR and VO2, which is the gold standard.18 A potential problem is that the relationship between MHR and VO2 can vary with different types of exercise.19 However, the same target MHRs apply to cycling in both postures because they cause similar VO2s.
Conclusions
Midwives can advise healthy pregnant women with low-risk singleton pregnancies that short duration, moderate intensity upright and semi-supine cycling are both safe for the mother and baby. After taking into account the mothers’ cardiorespiratory changes, uterine activity and FHR responses, neither posture was found to be preferable. Pregnant women who want to cycle for short durations do not need to change from stationary semi-recumbent bicycles to upright bicycles. They can use the type of bicycle that is most readily available and suits them best. A further useful finding for exercise prescribers is that the same target MHRs apply to both postures (because they cause similar VO2s, as well as similar FHR responses).
Acknowledgement
This research project was funded by Royal North Shore Hospital.
References
- American College of Obstetricians and Gynecologists. Exercise during pregnancy and the postpartum period (Committee opinion Number 267, January 2002). Int J Gynaecol Obstet. 2002;77(1):79–81
- . Hemodynamics at rest and during exercise in normal pregnancy as studied by cardiac catheterization. J Clin Invest. 1955;34:1524–1536
- . Cardiorespiratory responses to exercise during pregnancy. Am J Obstet Gynecol. 1970;108:600–605
- . Cardiovascular response to cycle exercise during and after pregnancy. J Appl Physiol. 1989;66:336–341
- . Effects of maternal aerobic fitness on cardiorespiratory responses to exercise. Med Sci Sports Exerc. 1993;25:993–998
- . Comparative evaluation of uterine response to exercise on five aerobic machines. Am J Obstet Gynecol. 1990;162:754–756
- . 2nd ed.. Clinical exercise testing. vol. 236. Sydney: W.B. Saunders; 1982;p. 238–9
- . Determination of mixed venous CO2 tensions by rebreathing. J Appl Physiol. 1956;9:25–29
- . Validation of the CO2 rebreathing method for measuring cardiac output in patients with hypertension or heart failure. J Lab Clin Med. 1976;88:672–682
- . Antepartum nonstress fetal heart rate monitoring: systematic analysis of baseline patterns and decelerations as an adjunct to reactivity in the prediction of fetal risks. Am J Obstet Gynecol. 1985;153:633–637
- . “Pseudo” fetal bradycardia during maternal exercise [Letter]. J Appl Physiol. 1987;62:849–850
- Fetal heart rate response to maternal exertion. J Am Med Assoc. 1988;259:3006–3009
- . Cardiac responses to progressive exercise in normal children: a synthesis. Med Sci Sports Exerc. 2000;32:253–259
- . Comparison of upright and semi-recumbent postures for exercise echocardiography in healthy children. Am J Cardiol. 2005;95:918–921
- . Effect of exercise on maternal hemodynamics and placental blood flow in healthy women. Acta Obstet Gynecol Scand. 1988;67:21–25
- . Maternal rectal temperature and fetal heart rate responses to upright cycling in late pregnancy. Br J Sports Med. 1996;30:32–35
- . Exercise and pregnancy: a review. Birth. 1994;21:85–95
- . Effect of pregnancy on heart rate/oxygen consumption calibration curves. Med Sci Sports Exerc. 2002;34:750–755
- . Cardiovascular responses of pregnant women during aerobic exercise in water: a longitudinal study. Int J Sports Med. 1988;9:443–447
PII: S1871-5192(06)00089-8
doi:10.1016/j.wombi.2006.09.002
© 2006 Australian College of Midwives. Published by Elsevier Inc. All rights reserved.
