Dr SN Stafne, Department of Public Health and General Practice, Faculty of Medicine, Norwegian University of Science and Technology, PO Box 8905, 7491 Trondheim, Norway. Email firstname.lastname@example.org
Please cite this paper as: Stafne S, Salvesen K, Romundstad P, Torjusen I, Mørkved S. Does regular exercise including pelvic floor muscle training prevent urinary and anal incontinence during pregnancy? A randomised controlled trial. BJOG 2012;119:1270–1280.
Objective To assess whether pregnant women following a general exercise course, including pelvic floor muscle training (PFMT), were less likely to report urinary and anal incontinence in late pregnancy than a group of women receiving standard care.
Design A two-armed, two-centred randomised controlled trial.
Setting Trondheim University Hospital (St. Olavs Hospital) and Stavanger University Hospital, in Norway.
Population A total of 855 women were included in this trial.
Methods The intervention was a 12-week exercise programme, including PFMT, conducted between 20 and 36 weeks of gestation. One weekly group session was led by physiotherapists, and home exercises were encouraged at least twice a week. Controls received regular antenatal care.
Main outcome measures Self-reported urinary and anal incontinence after the intervention period (at 32–36 weeks of gestation).
Results Fewer women in the intervention group reported any weekly urinary incontinence (11 versus 19%, P = 0.004). Fewer women in the intervention group reported faecal incontinence (3 versus 5%), but this difference was not statistically significant (P = 0.18).
Conclusions The present trial indicates that pregnant women should exercise, and in particular do PFMT, to prevent and treat urinary incontinence in late pregnancy. Thorough instruction is important, and specific pelvic floor muscle exercises should be included in exercise classes for pregnant women. The preventive effect of PFMT on anal incontinence should be explored in future trials.
Pregnancy and delivery are risk factors for urinary and anal incontinence in later life.1,2 Urinary incontinence is defined as a ‘complaint of involuntary loss of urine’,3 whereas the term anal incontinence is defined as a ‘complaint of involuntary loss of feces or flatus’.3
The pelvic floor plays an important part in maintaining continence.4 Contraction of the pelvic floor muscles (PFMs) causes an inward lift and squeeze around the urethra, vagina and rectum,5 resulting in closure, stabilisation and resistance to downward movement.6 Randomised controlled trials have documented the effects of intensive pelvic floor muscle training (PFMT) and close follow-up with physiotherapists in the prevention and treatment of pregnancy-related urinary incontinence.7–11 Nevertheless, one Cochrane review and two other systematic reviews conclude that the efficacy of PFMT during pregnancy in preventing urinary and anal incontinence is still open to question.12–14 The Cochrane review strongly recommends that all future trials of PFMT during pregnancy should collect data on faecal incontinence, and highlights the need for large, pragmatic trials of population-based approaches, using intensive PFMT and recruiting antenatal women, regardless of continence status or parity.12 There has been a lack of trials investigating the effect of implementing PFMT in a more general training programme for pregnant women.
We conducted the present study as one part of a randomised controlled trial aimed to investigate the effects of regular exercise during pregnancy in the prevention of pregnancy-related diseases and complications during labour. The primary outcome measures were gestational diabetes and glucose metabolism.15 Another important research question was whether it was possible to prevent urinary and anal incontinence by including specific PFM exercises in the general exercise course, to improve the continence mechanisms before the problem arises (primary prevention), and also to detect incontinence at an early stage and thereby stop the development of the condition (secondary prevention).
The aim of this study was to assess whether pregnant women following a general exercise course including PFMT were less likely to report urinary and anal incontinence in late pregnancy than a group of women receiving standard care.
We conducted a two-armed, two-centre randomised controlled trial of a 12-week regular exercise programme versus standard antenatal care. The procedures followed were in accordance with ethical standards of research and the Declaration of Helsinki. The women received written information about the trial and signed informed consent forms. The participants were not compensated financially. The study was approved by the Regional Committee for Medical and Health Research Ethics (REK 4.2007.81), and was registered in Clinical.Trials.gov (NCT 00476567).
Pregnant women booking for routine ultrasound at Trondheim University Hospital (St. Olavs Hospital) and Stavanger University Hospital were invited to participate in the trial. Women in Trondheim were recruited from April 2007 to June 2009, and women in Stavanger were recruited from October 2007 to January 2009. More than 97% of Norwegian pregnant women attend a routine scan at around 18 weeks of gestation, and these examinations are free of charge. During the inclusion period approximately 12 000 pregnant women had routine scans at the two hospitals. Inclusion criteria were age ≥18 years and a singleton live fetus. Exclusion criteria were high-risk pregnancies and/or diseases that could interfere with participation. For practical reasons we also excluded women who lived too far from the hospitals to attend weekly training groups (judged as more than 30-minutes drive).
Concealed randomisation in blocks of 30 was performed at the Unit for Applied Clinical Research, Norwegian University of Technology and Science, by a web-based computerised procedure. The staff involved with training or outcome assessments had no influence on the randomisation procedure. Because of the nature of the study it was not blinded.
Women in the intervention group received a standardised exercise programme, including aerobic activity, strength training (including specific PFM exercises) and balance exercises. The training protocol followed recommendations from The American College of Obstetricians and Gynecologists and the Norwegian National Report on Physical Activity and Health.16,17 Training sessions of 60 minutes, in groups of between eight and 15 women, instructed by a physiotherapist were offered once a week over a period of 12 weeks (between 20 and 36 weeks of gestation). Each group session consisted of three parts:
1 30–35 minutes of low-impact aerobics (no running or jumping). Step length and body rotations were reduced to a minimum, and crossing legs and sharp and sudden changes of position were avoided. The aerobic dance programme was performed at moderate intensity, defined as 13 and 14 on Borg’s rating scale of perceived exertion.18
2 20–25 minutes of strength exercises, using body weight as resistance, including exercises for the upper and lower limbs, back extensors, deep abdominal muscles and PFMs. Three sets of ten repetitions of each exercise were performed.
3 5–10 minutes of light stretching, body awareness, breathing and relaxation exercises.
In addition, women were encouraged to follow a written 45-minute home exercise programme, including PFMT, at least twice a week (30 minutes of endurance training and 15 minutes of strength and balance exercises).
Women in the intervention group were individually instructed in pelvic floor anatomy and how to contract the PFMs correctly by a physical therapist. Correct contraction was controlled by vaginal palpation. The PFMT followed principles for increasing the strength of skeletal muscles.19 Women were encouraged to perform three sets of eight to twelve close to maximum contractions of the PFMs, and were encouraged to hold the contraction for 6–8 seconds and, if possible, to add three fast contractions at the end of the contraction.20 PFMT was performed in different positions, with legs apart, to emphasise specific strength training of the PFMs and the relaxation of other muscles.
Adherence to the protocol was defined as exercising 3 days per week or more at moderate to high intensity. Performing the exercise programme was strongly emphasised, and recorded in the women’s personal training diary and through reports from the physiotherapists leading the training groups.
Women in the control group received standard antenatal care and the customary information given by their midwife or general practitioner. They were not discouraged from exercising on their own. Women in both groups received written information and recommendations on PFMT, diet and pregnancy-related lumbopelvic pain. The PFMT brochure includes detailed information about the pelvic floor and an evidence-based PFMT programme.9
Pre-tests were performed between 18 and 22 weeks of gestation, and post-tests took place between 32 and 36 weeks of gestation. The main outcomes were urinary and anal incontinence, as measured by self-report. All participants answered a questionnaire, including questions related to urinary incontinence (Sandvik’s severity index),21,22 and anal incontinence (St. Marks score),23 at both the start (18–22 weeks of gestation) and the end (32–36 weeks of gestation) of the intervention period. Women were asked to report their urinary leakage for the following situations with a ‘yes’ or ‘no’: (A) when coughing, sneezing or laughing; (B) while being physically active (running or jumping); (C) when making sudden changes in position or lifting; or (D) any leakage accompanied by a strong urgency to void. Urinary leakage was subclassified according to the definitions given in the standardised terminology of lower urinary tract symptoms.3 Women confirming any type of urinary leakage (A, B, C or D) were referred to as having urinary incontinence (UI). Women confirming loss of urine in association with A, B or C were defined as having stress urinary incontinence (SUI), whereas women confirming loss of urine in association with D were defined as having urge urinary incontinence (UUI). The outcome measures of UI were further divided into two severity categories with respect to frequency: ‘urinary leakage < once per week’ or ‘urinary leakage ≥ once per week (severe UI)’. Anal incontinence was registered as faecal and flatal incontinence during the previous 4 weeks, based on one question from the St. Marks score.23 Women reporting leakage of solid and/or liquid stool during the last 4 weeks were categorised as faecally incontinent, and women reporting flatus during the last 4 weeks were categorised as flatally incontinent.
Frequency, intensity and type of physical activity, including PFMT, were recorded for both groups at inclusion and at follow-up by self-report in a questionnaire. In addition, women in the intervention group registered PFMT in a training diary.
Based on a prevalence estimate of gestational diabetes of 9%, with a reduction to 4% (primary outcome of the trial), a study population of 381 patients in each group was needed in a two-sample comparison test with a 5% significance level and a power of 0.80. This sample size was able to detect a 0.2 SD difference on continuous variables, with a power of 0.80. Given the study population, we had a power of 0.79 to detect a risk of reduction in urinary incontinence of 10% from 50 to 40% in the two groups.
The statistical analyses were performed with spss 19. The data were analysed according to the ‘intention-to-treat’ principle using a chi-square test and binary logistic regression. The results were presented as crude and baseline-adjusted estimates, with 95% confidence intervals.
Women in the intervention and control groups were primarily analysed according to type of incontinence (UI, SUI, UUI, and faecal and flatal incontinence). Subgroup analyses were based on urinary, faecal and flatal continence status at the time of inclusion, for estimating the primary and secondary preventive effects of the intervention.
In all, 875 women consented to participate in the trial. Twenty women were excluded or withdrew before the first examination: thirteen did not meet the inclusion criteria, five miscarried and two had twin pregnancies. A total of 855 women were randomly allocated to an intervention group or a control group (Figure 1). However, 32 women in the intervention group and 61 in the control group were lost to follow-up. Data from 397 women in the intervention group and 365 women in the control group were included in a complete case analysis.
The groups were similar in baseline characteristics except for severe UI and SUI, which were more frequent in the control group (P = 0.05 and 0.01, respectively; Table 1). After the intervention period, significantly less women in the intervention group reported UI and SUI, irrespective of severity (Table 2). The findings were consistent when adjusting for baseline values. A lower proportion of women in the intervention group (3%) than in the control group (5%) reported faecal incontinence; however, the difference was not statistically significant (Table 2). There were no differences in weight or body mass index (BMI) between groups at follow-up.
Table 1. Baseline maternal characteristics of the study population
Data are means ± SDs.
*Data missing in 0.5% of cases.
**Data missing in 3.6% of cases.
***Data missing in 2.5% of cases.
****Data missing in 2.0% of cases.
(n = 429)
(n = 426)
Mean age – years
30.5 ± 4.4
30.4 ± 4.3
Parity – no. (%)
Previous vaginal delivery – no. (%)
Gestational week at booking
20.3 ± 1.5
20.3 ± 1.7
Booking BMI – kg/m2
24.7 ± 3.0
25.0 ± 3.4
Booking weight – kg
70.4 ± 9.8
70.8 ± 10.3
Exercise regularly – no. (%)
Exercise regularly ≥ 3 times per week at moderate to high intensity – no. (%)
PFMT – no. (%)
PFMT ≥ 3 times per week – no. (%)
UI ≥ 1 times per week – no. (%)*
SUI ≥ 1 times per week – no. (%)**
UUI ≥ 1 times per week – no. (%)
Faecal incontinence – no. (%)***
Flatal incontinence – no. (%)****
Table 2. Urinary incontinence (UI) at 32–36 weeks of gestation in the intervention group and control group, including women who did or did not report urinary leakage at inclusion
Unadjusted for baseline
Adjusted for baseline
n = 397
n = 365
*Data missing in 0.7% of cases.
**Data missing in 6.0% of cases.
***Data missing in 1.0% of cases.
****Data missing in 2.8% of cases.
*****Data missing in 2.2% of cases.
UI ≥ 1 time per week*
SUI ≥ 1 time per week**
UUI ≥ 1 time per week***
Subgroup analyses were performed for nulliparous and multiparous women. At baseline, the prevalences of severe UI and SUI were lower in the intervention group among nulliparous women. After the intervention (at 32–36 weeks of gestation) the prevalence of UI among nulliparous women was 35 versus 47% in the intervention and control groups, respectively (OR 0.6, adjusted for baseline values; 95% CI 0.4–0.9; P = 0.03), and the prevalence of severe UI was 6 versus 14% (OR 0.4, adjusted for baseline values; 95% CI 0.2–0.9; P = 0.02). In multiparous women, the prevalence of faecal incontinence at 32–36 weeks of gestation was lower in the intervention group: 3 versus 8% (OR 0.2, adjusted for baseline values; 95% CI 0.1–0.8; P = 0.03). No other significant differences were seen in a subgroup analysis according to parity.
Baseline characteristics of subgroups based on groups stratified according to urinary continence status at inclusion are presented in Table 3. Sixty-nine percent of the women who were continent and 40% of the women who were incontinent were nulliparous. The subgroup analyses showed that in women who were continent for urine at inclusion, the proportions of women reporting SUI ≥ once per week and UUI were significantly lower (P = 0.03 and 0.006, respectively) in the intervention group after the intervention period (Table 4). Furthermore, in women in either group who were incontinent for urine at inclusion, the proportion reporting UI, UI ≥ once per week and SUI after the intervention period was lowest in the intervention group (P = 0.002, 0.03 and 0.007, respectively). These findings were consistent when adjusting for baseline values (Table 4).
Table 3. Baseline characteristics of subgroups stratified according to continence status at inclusion
n = 256
n = 172
n = 244
n = 180
Data are means ± SDs.
Mean age – years
30.1 ± 4.0
31.2 ± 4.9
30.0 ± 4.4
31.0 ± 4.1
Parity – no. (%)
Booking weight – kg
70.0 ± 10.0
71.0 ± 9.5
70.3 ± 10.1
71.6 ± 10.7
Booking BMI – kg/m2
24.5 ± 3.0
24.9 ± 3.1
24.9 ± 3.4
25.2 ± 3.5
UI ≥ 1 times per week – no. (%)
SUI ≥ 1 times per week – no. (%)
UUI ≥ 1 times per week – no. (%)
Table 4. Subgroup analysis of urinary incontinence (UI) at 32–36 weeks of gestation in the intervention group (IG) and control group (CG). Subgroups are based on groups stratified according to continence status at inclusion
Unadjusted for baseline
Adjusted for baseline
Continent at inclusion
n = 235
n = 208
UI ≥ 1 time per week
SUI ≥ 1 time per week
UUI ≥ 1 time per week
Incontinent at inclusion
n = 157
n = 156
UI ≥ 1 time per week
SUI ≥ 1 time per week
UUI ≥ 1 time per week
There were no differences in age, parity, weight and BMI based on groups stratified according to faecal continence status at inclusion. Among women who were faecally continent at inclusion, 2% (8/359) in the intervention group and 4% (13/332) in the control group reported faecal incontinence at follow-up (OR 0.6; 95% CI 0.2–1.4; P = 0.20). Among women reporting faecal incontinence at inclusion, 19% (4/21) in the intervention group and 20% (3/15) in the control group reported faecal incontinence at follow-up (OR 0.9; 95% CI 0.2–5.0; P = 0.94).
Women reporting flatal incontinence at inclusion were younger than flatally continent women (30.0 ± 4.1 versus 30.7 ± 4.4 years), and a larger proportion of them was nulliparous (65 versus 54%). Among women who were flatally continent at inclusion the proportion of flatal incontinence in both groups after the intervention period was 21% (58/275 versus 53/258; OR 1.0; 95% CI 0.7–1.6; P = 0.88). Among women reporting flatal incontinence at inclusion, 69% (75/109) in the intervention group and 74% (69/93) in the control group reported flatal incontinence at follow-up (OR 0.8; 95% CI 0.4–1.4; P = 0.40).
In the intervention group, 95% reported weekly PFMT at follow-up, compared with 79% in the control group (P < 0.001). The proportion of women in the intervention group performing PFMT three times per week or more was 67%, compared to 40% in the control group (P < 0.01).
Adherence to the general exercise protocol (i.e. exercising 3 days per week or more at moderate to high intensity) in the intervention group was 55% (n = 217). In comparison, 10% of women in the control group exercised 3 days per week or more at moderate to high intensity at follow-up (P < 0.001). In the intervention group, 72% exercised at least once per week. UI and anal incontinence were not more frequent among women in the intervention group who adhered to the protocol (n = 217), compared with the total control group at the end of the intervention.
No serious adverse events related to physical exercise were seen, and the outcomes of pregnancy were similar in the two groups.15
This trial demonstrates that pregnant women who followed a 12-week course of regular exercise including PFMT were less likely to report UI and SUI in late pregnancy than women given standard care, including written instructions for PFMT. Among pregnant women who were continent for urine at the time of inclusion, significantly fewer women in the intervention group reported SUI once per week or more and UUI at follow-up, indicating a primary preventive effect of the intervention. In the subgroup of pregnant women categorised as incontinent for urine at inclusion, significantly fewer women in the intervention group reported UI, SUI once per week or more, and SUI in late pregnancy, compared with women in the control group (secondary prevention). A lower proportion of women in the intervention group reported faecal incontinence compared with the control group; however, the difference was not statistically significant. A subgroup analysis of multiparous women showed a significantly lower prevalence of faecal incontinence in the intervention group after the intervention.
The strengths of the present trial are the large number of participants, the high follow-up rate and the use of a computerised randomisation procedure. Despite the randomised design, some of the baseline characteristics were significantly different between the groups. This was probably the result of chance, and was accounted for in the statistical analysis. The investigators were aware of group allocation; however, as the prevalence of incontinence was based on self-reporting, the lack of blinding of investigators should not have had any impact on the results.
The prevalence of urinary leakage in pregnancy has been reported to vary between 9 and 74%.24,25 This variation probably arises from different study populations (various proportions of nulliparous/parous women), different definitions of incontinence (self-reports or standardised pad tests) and different pregnancy lengths. In the Norwegian Mother and Child Cohort Study including 43 279 women from 1999 to 2006, 58% of the women reported any UI, and 35% of women reported any UI weekly or more in gestational week 30.26 Anal incontinence is less studied in pregnancy, but a prevalence of 6% in late pregnancy has been reported.27
The prevalence of urinary leakage in mid and late pregnancy in the present trial was lower than in a previous trial conducted in the same geographic area 10 years ago, using the same outcome measures to classify women as continent or incontinent as used in the present trial.9 Mørkved et al.9 found that 32% of nulliparous women in the intervention group and 48% in the control group reported leaking urine once per week or more in gestational week 36. In the present trial, leaking urine once per week or more in gestational weeks 32–36 was reported by 6% versus 14% of nulliparous women in the two groups, respectively. As PFMT is known to have a high cure rate,28 the difference between these studies may be the result of an increased focus on PFMT in pregnant women and health care professionals, following the results of the previous trial. In the current trial 60% of the study participants reported doing any PFMT at the time of inclusion, and 25% reported doing PFMT three times per week or more. In the previous trial 30% performed any PFMT at inclusion.9
The success of training depends on the ability to effectively contract the PFMs. It is estimated that 30% of women are unable to contract the PFMs on a first attempt.29,30 Bump et al.29 found that only 49% performed an ideal voluntary PFM contraction after brief standardised verbal instruction. In the present trial all women in the intervention group were individually instructed in pelvic floor anatomy, and the correct contraction was controlled with vaginal palpation. Women in both groups received written information on PFMT, including an evidence-based exercise programme, and at follow-up the number of women performing PFMT had increased in both groups. At follow-up, a large proportion of women in the control group reported doing regular PFMT during the intervention period. Still, the intervention group demonstrated significantly less UI in late pregnancy. This finding suggests that thorough instructions in correct PFM contraction, intensive PFMT and close follow-up are important to increase the success rate.
Another possible explanation of a low prevalence in the current trial may be differences in study populations. In the trial by Mørkved et al.9 pregnant women were invited to participate with the explicit goal to study the effects of PFMT on UI. The aims of the current trial were to assess the effects of a more general exercise programme on several pregnancy-related conditions. Women experiencing urinary leakage may have been more likely to sign up in the previous trial aiming to improve continence, and women with severe urinary leakage may have been less likely to sign up for a general exercise trial including activities or situations that often cause incontinence. In the previous trial 32% of the study population reported weekly UI at inclusion,9 compared with 13% in the current trial. The study population in the present trial was found to be representative for pregnant women in Norway regarding age, BMI, parity and level of exercise.15
Three previous trials have addressed primary prevention of UI during pregnancy.8,31,32 Two trials reported highly significant preventive effects of PFMT.8,32 The third trial had conflicting results,31 but the response rate of participants was low (about 50%). The Cochrane review addressing prevention of urinary and faecal incontinence included mixed prevention and treatment trials.12 Data from subgroups of previously continent women were added to the analyses of a primary preventive effect of PFMT on UI,7,9 and the results indicated that for women having their first baby, antenatal PFMT appears to reduce the prevalence of UI in late pregnancy.12 Thus, the review suggested new, large, pragmatic trials using intensive PFMT and recruiting antenatal women regardless of continence status or parity, and collecting data on both urinary and faecal incontinence.12 Subgroup analyses in the present trial support the Cochrane review conclusions that the prevention and treatment effects on UI in late pregnancy were predominantly seen in women bearing their first baby, with a 40–60% reduction in risk.
Anal incontinence is less prevalent, but this condition is clearly a social and psychological burden that reduces the quality of life.33 The preventive effect of PFMT on the development of anal incontinence has been sparsely studied. One previous trial included PFMT in a general fitness programme. This trial demonstrated no difference in urinary or anal incontinence between the intervention and control groups. However, this trial was underpowered and included no control of the women’s ability to contract the PFMs correctly.34
In the present trial, the proportion of women reporting faecal incontinence in the intervention group was reduced from 5 to 3% after the intervention period. In the control group the proportion reporting faecal incontinence increased from 4 to 5% at follow-up. Although this difference was not statistically significant, it may suggest a possible clinically significant effect of the intervention. Another interesting finding was the preventive effect on faecal incontinence in multiparous women. This finding indicates that in women with possible previous birth-related injury to the PFMs, PFMT might prevent faecal incontinence in a later pregnancy.
In the present trial, women in the intervention group were encouraged to follow a general exercise protocol three times per week or more during the 12-week intervention period. Fifty-five percent of the women in the intervention group adhered to the protocol, and 72% of the women in the intervention group exercised at least once per week. For comparison, 10% of women in the control group reported exercising 3 days per week or more, and 30% exercised at least once per week. Women who are physically active are more frequently exposed to higher and more repetitive increases in abdominal pressure compared with more sedentary women. The exercise protocol in the present trial included physical fitness exercises designed for pregnant women, and did not include running or jumping. However, the aerobic part of the programme consisted of weight-bearing exercises causing increased abdominal pressure and increased risk of urinary leakage. Nevertheless, fewer women in the intervention group reported UI, and women adhering to the exercise protocol reported less UI than women who did not adhere in the intervention group at follow-up.
In addition to better structural support for the bladder neck with PFMT, the intervention group may have developed increased awareness and skill of timing the contraction with the event that elicits leakage. Miller et al.35 found that 80% of women with de novo stress incontinence in gestational week 35 were able to reduce leakage during coughing by using the Knack maneuver (i.e. tighten the PFMs in preparation for a known leakage-provoking event), with 55% eliminating leakage completely.
An epidemiological study including 27 936 women reported a prevalence of any urinary incontinence of 25% in the adult population, with the prevalence of incontinence increasing with age.36 Among the nulliparous women in that study (n = 3339), 10% reported UI.37 Data suggest that pregnancy contributes to pelvic floor dysfunction later in life,38 and that the first pregnancy and delivery is associated with a high prevalence of objective and subjective bowel dysfunction.39 In a cohort study it was found that most women experiencing postpartum UI had symptoms of UI during the third trimester.25 Another study found that 5 years after the first delivery, 33% of women reported UI.40 Although UI is common, it reduces the quality of life and affects social, physical, occupational and leisure activities.41 The present trial demonstrates that PFMT during the second half of pregnancy reduced the prevalence of UI in late pregnancy. However, it is important to assess any possible long-term effects postpartum. When comparing the prevalence of UI in the present trial with a previous trial from the same area,9 we found a significant reduction over the last 10 years. This may be linked to different study populations, but may also be the result of an increased awareness and knowledge about prevention and treatment of UI among healthcare professionals and the general population, indicating a preventive effect of information, instructions and encouragement to exercise the PFMs.
Anal incontinence is less prevalent, but this condition is clearly a social and psychological burden that reduces the quality of life.23 No previous trial has assessed the preventive effect of PFMT on the development of faecal incontinence. In the intervention group, the proportion of women reporting faecal incontinence was reduced from 5 to 3% after the intervention period. In the control group the proportion reporting faecal incontinence increased from 4 to 5% at follow-up. Although this difference was not statistically significant, it may suggest a possible clinically significant effect of the intervention.
The prevalence of faecal incontinence during pregnancy is low, and the present trial was slightly underpowered to assess faecal incontinence properly, as shown by the wide confidence intervals. One important issue for future research is to assess whether PFMT in pregnancy can reduce faecal incontinence in pregnancy. Furthermore, it is important to assess the effects of regular exercise, including PFMT, during pregnancy on the duration of labour, the long-term effects on urinary and anal continence, and to investigate the associations between outcome, prior continence status and parity.
The results from the present trial indicate that pregnant women should do PFMT to prevent and treat UI in late pregnancy. Thorough instruction in correct PFM contraction and PFMT is important, and specific PFM exercises should be included in exercise classes for pregnant women. Any possible long-term effects on UI and the preventive effect of PFMT on anal incontinence should be explored further.
Disclosure of interests
None of the authors have a conflict of interest to declare.
Contribution to authorship
SNS, physiotherapist and PhD student, participated in planning the main study, coordinated the data collection, organised the training programme, analysed the data, wrote the first draft and finalised the article. KÅS, Professor of Obstetrics and Gynaecology, participated in the planning of the main study, and revising and finalising the article. PRR, Professor of Epidemiology, participated in the data analyses, and in revising and finalising the article. IHT, physiotherapist, participated in the interpretation of data, and in revising and finalising the article. SM, Professor of Physiotherapy and principal investigator, initiated and planned the main study, supervised the training programme and participated in the interpretation of the data, as well as drafting and finalising the article.
Details of ethics approval
The study was approved by the Regional Committee for Medical and Health Research Ethics (REK 4.2007.81). Clinical trial registration: http://www.clinicaltrials.gov, NCT00476567.
The trial was supported by the Norwegian Fund for Postgraduate Training in Physiotherapy, and the Liaison Committee for Central Norway Regional Health Authority (RHA) and the Norwegian University of Science and Technology. The sponsors of the study had no role in the study design, data collection, data analysis, data interpretation, writing the article or decision to submit for publication.
The authors thank all of the physiotherapists (Marit Lindland Ree, Wilma van de Veen, Karen Schei, Marte Sundby and Henriette Tokvam Larsen) and the medical secretaries (Elin Ørndahl Holthe and Heidi Larsen) at the two hospitals for their efforts in performing this study with exercise classes and testing, as well as the women who participated in this study.
1. Background: Describe what is currently known about the prevalence of urinary and anal incontinence in pregnancy, and women in general. What is the existing evidence for the efficacy of pelvic floor muscle training (PFMT) on the prevention and treatment of incontinence?
2. Methods: This trial used self-reported symptoms of incontinence as its primary outcome measure, in line with the most recent Cochrane review.1 Discuss the advantages and disadvantages of this approach. There was a difference between the two groups concerning the proportion of women with severe urinary incontinence at baseline. The authors addressed this imbalance by adjusting odds ratios for baseline differences. Another approach would have been to ensure the comparability of the two groups for key prognostic variables, using stratified randomisation or minimisation. Describe and compare these methods.
3. Results and implications: Almost 80% of the women in the control group performed PFMT at least once a week, a high proportion compared with other estimates of PFMT in pregnant women.2 Still, a significant reduction in urinary incontinence was seen in the intervention group compared with controls. Explore what features of the intervention may have increased its effectiveness. The loss to follow-up was much higher in the control group, particularly for ‘reasons unknown’. Discuss the implications in terms of bias. This study investigates the effects of PFMT during pregnancy – consider the need for longer-term follow-up. Based on the findings, should this intensive exercise programme be recommended routinely? (Data S1)