International Urogynecology Journal

, Volume 22, Issue 1, pp 17–22

A randomized controlled trial of antenatal pelvic floor exercises to prevent and treat urinary incontinence

Authors

  • Po-Chun Ko
    • Department of Obstetrics and GynecologyChang Gung Memorial Hospital at Linkou
    • Department of Obstetrics and GynecologyChang Gung Memorial Hospital at Linkou
    • College of MedicineChang Gung University
  • Shuenn-Dhy Chang
    • Department of Obstetrics and GynecologyChang Gung Memorial Hospital at Linkou
    • College of MedicineChang Gung University
  • Jian-Tao Lee
    • School of NursingChang Gung University
  • An-Shine Chao
    • Department of Obstetrics and GynecologyChang Gung Memorial Hospital at Linkou
    • College of MedicineChang Gung University
  • Po-Jen Cheng
    • Department of Obstetrics and GynecologyChang Gung Memorial Hospital at Linkou
    • College of MedicineChang Gung University
Original Article

DOI: 10.1007/s00192-010-1248-4

Cite this article as:
Ko, P., Liang, C., Chang, S. et al. Int Urogynecol J (2011) 22: 17. doi:10.1007/s00192-010-1248-4

Abstract

Introduction and hypothesis

The aim of the study was to evaluate the effect of antenatal pelvic floor muscle exercise (PFME) in the prevention and treatment of urinary incontinence during pregnancy and postpartum period.

Methods

Three hundred women were randomly assigned to the PFME group and control group. Urinary symptoms were measured by Urogenital Distress Inventory-6 (UDI-6), Incontinence Impact Questionnaire-7 (IIQ-7), and question of self-reported urinary incontinence. Questionnaire scores of the PFME and the control groups were compared and analyzed.

Results

During late pregnancy and the postpartum period, the PFME group had significantly lower total UDI-6 and IIQ-7 scores; their self-report rate of urinary incontinence was also less than the control group. Additionally, we found whether in PFME or control, women who delivered vaginally were more likely to develop postpartum urinary leakage than women who delivered by cesarean section.

Conclusions

PFME applied in pregnancy is effective in the treatment and prevention of urinary incontinence during pregnancy, and this effect may persist to postpartum period.

Keywords

Pelvic floor muscle exercisePregnancyPostpartumUrinary incontinence

Abbreviations

SUI

Stress urinary incontinence

PFME

Pelvic floor muscle exercise

IIQ-7

Incontinence Impact Questionnaire

UDI-6

Urogenital Distress Inventory

Introduction

Pregnancy and birth trauma are risk factors for urinary incontinence, which commonly develops during pregnancy or following delivery. Incidences of stress urinary incontinence (SUI) during pregnancy have been reported to range from 19.9% to 70% in nulliparous women [13], while in the postpartum period, the prevalence in primipara is between 0.7% and 35% [4, 5]. Pelvic floor muscle exercise (PFME) has been documented in the literature to be one of the treatment methods for urinary incontinence during pregnancy [6, 7]. Previous epidemiological data showed that the first delivery contributes the most to the development of SUI [8]. Therefore, the National Institute of Clinical Excellence [9] recommends PFME for all women in a first pregnancy for prevention of SUI. Antenatal PFME has been shown to reduce the incidence of postpartum SUI in the short term [6, 7, 10]; however, studies on the efficacy of PFME done in the antenatal period are rather limited. So, we conducted a randomized controlled trial to evaluate the effect of antenatal PFME in the prevention and treatment of urinary incontinence during pregnancy and after birth.

Methods

Nulliparous women, at 16 to 24 gestational weeks, receiving regular prenatal care in the obstetrics clinic at a university hospital, were invited to participate in the study. Women who agreed to join the study provided written informed consent. The study protocol (No. 96-1224B) was approved by our institutional review board. Exclusion criteria included multiparity, multiple gestations, severe pregnancy complications, high risk for preterm labor, pain during PFME, diseases that could interfere with participation, or would be unavailable for follow-up.

Women were recruited into the trial from April 2008 to October 2008, and followed up until October 2009. Three hundred fifteen women gave their signed consent to participate in this trial. Fifteen women were excluded or withdrew before the first examination: Five had pregnancy complications, four had twin pregnancies, and six withdrew for unknown reasons. Three hundred women were randomly allocated to a PFME group or to a control group. Randomization was achieved by selection of sealed envelopes, which were opened at entry. Women of the PFME group were individually instructed by a physical therapist about pelvic floor anatomy and how to contract the pelvic floor muscles correctly before exercise. Correct contraction was assessed by observation of inward movement of perineum during contraction. The exercise group followed a specially designed pelvic floor muscle training course, as published previously by Reilly et al. [10]. However, we did not ask pregnant women to increase the number of contractions per repetition from eight to 12 at the third trimester as Reilly et al. did. The exercises comprised three repetitions of eight contractions each held for 6 s, with 2-min rest between repetitions. These were repeated twice daily at home with additional training in groups once a week for 45 min by a physical therapist, and lasted throughout a 12-weeks’ period. Each training group included around 10 women. Group training was performed in sitting and standing positions with legs apart to emphasize specific strength training of the pelvic floor muscles and relaxation of other muscles. Motivation was strongly emphasized by the physiotherapist. Compliance in the PFME group was monitored using diaries to record the number of daily exercises performed from the first examination to 36 gestational weeks. We considered women in the exercise group who reported practice PFME at least 75% as adherent. Women of the control group also received regular prenatal care and received the customary written postpartum instructions that did not include PFME from the hospital. If women performed PFME before entry, they were excluded from the study; however, out of ethic consideration, pregnant women were not discouraged from performing PFME on their own.

At entry, all participants were asked to complete two validated questionnaires, including Incontinence Impact Questionnaire (IIQ-7) (a seven-item measure of urogenital distress) and Urogenital Distress Inventory (UDI-6) (a six-item measure of incontinence effect) [11]. Total scores of IIQ-7 and UDI-6 as well as individual scores of six questions of the UDI-6 were calculated as previously reported [12]. In addition to the IIQ-7 and UDI-6, we asked all women two questions: “Did you go urinate more than 7 times during the day” and “Do you leak urine at any time: never, seldom, weekly or daily?” Both urinary frequency and urinary incontinence were registered in accordance with the International Continence Society’s definitions [13]. Women reporting urinary incontinence once per week or more during the last month were categorized as incontinent. All women were interviewed by a research nurse and administered the questionnaires via face-to-face method at 36 gestational weeks and postpartum day 3, and via telephone 6 weeks and 6 months postpartum. Participant characteristics abstracted from patient charts included age, body mass index, delivery method, and baby birth weight.

Previous reports demonstrated that the incidence of urinary incontinence during pregnancy was 42% [5], so we assumed that the effect of PFME in treatment of urinary incontinence was 50%. In order to achieve 85% power to detect the difference in self-reported urinary incontinence between groups of PFME and control, a sample size of at least 290 was required. Characteristics of the PFME group and the control group were assessed by paired Student t test for continuous data and Chi-square test for proportions. Comparison of questionnaire scores of UDI-6 and IIQ-7 between the PFME group and the control group was assessed by the Mann–Whitney U test. The association of self-reported urinary incontinence and PFME were assessed by the Chi-square test. We used the generalized estimating equations to evaluate the odds ratios of self-reported urinary incontinence between the PFME group and the control group. We also used the generalized estimating equations to evaluate the odds ratios of self-reported urinary incontinence between groups of vaginal delivery and cesarean section. A P value of <0.05 was considered statistically significant. Statistical analyses were performed using SPSS 15.0 for Windows (SPSS, Inc., Chicago, IL, USA).

Results

Three hundred women were randomly assigned to the PFME group (n = 150) or the control group (n = 150) (Fig. 1). Over 80% women of exercise group attended entire group training program, and none were absent from the training more than twice. All women were followed up for 6 months after delivery. At 36 gestational weeks, 87% of the treatment group reported practice of PFME at least 75% of the time. Participant characteristics are shown in Table 1. Patient’s age, body mass index, delivery method, baby birth weight, labor duration, episiotomy, instrument assisted delivery, and severe perineal laceration of degree 3 or 4 were not significantly different between the two groups. During late pregnancy and the postpartum period, the PFME group had significantly lower total UDI-6 and IIQ-7 scores as well as lower scores for questions 2, 3, and 4 of the UDI-6 Questionnaire (Table 2). Question 2 concerns urine leakage related to the feeling of urgency; question 3 looks into urine leakage related to activity, coughing, or sneezing; question 4 investigates small amounts of urine leakage. Total IIQ-7 scores show that quality of life related to incontinence is significantly better in the PFME group compared with the control group at 36 pregnancy weeks, and 6 weeks and 6 months postpartum.
https://static-content.springer.com/image/art%3A10.1007%2Fs00192-010-1248-4/MediaObjects/192_2010_1248_Fig1_HTML.gif
Fig. 1

Patient flow chart

Table 1

Characteristics of patients

Group

PFME (n = 150)

Control (n = 150)

P value

Age

31.66 ± 3.42

31.29 ± 3.78

0.38

BMI before pregnancy

21.78 ± 4.10

22.18 ± 3.38

0.35

BMI after pregnancy

27.24 ± 5.14

27.71 ± 3.41

0.34

Vaginal delivery

102 (68)

107 (71)

0.53

Baby birth weight

3,121 ± 532.58

3,145.91 ± 415.26

0.65

Labor duration (min)a

319 ± 176

326 ± 149

0.52

Episiotomya

99 (97)

104 (97)

0.87

Instrumental deliverya

6 (6)

7 (7)

0.44

Severe perineal lacerationa

10 (10)

10 (9)

0.63

Data are presented as mean ± standard deviation or n (%), and calculated with the paired Student t test or Chi-square test

BMI body mass index

aData of women who delivered vaginally are shown in 102 of PFME group and 107 of control group

Table 2

Comparisons of UDI-6 and IIQ-7 between PFME group and control group

Questionnaire

PFME

Control

P value

UDI-6a

1.27 ± 1.54

1.36 ± 1.55

0.49

UDI-6b

3.44 ± 3.26

4.66 ± 3.32

<0.01

UDI-6c

1.42 ± 2.04

2.31 ± 2.16

<0.01

UDI-6d

0.81 ± 1.36

1.54 ± 1.59

<0.01

UDI-6e

0.35 ± 0.84

0.86 ± 1.14

<0.01

IIQ-7a

1.11 ± 2.47

1.21 ± 2.44

0.18

IIQ-7b

3.77 ± 6.01

5.28 ± 5.61

<0.01

IIQ-7d

1.73 ± 3.57

2.86 ± 3.52

<0.01

IIQ-7e

0.77 ± 2.07

1.56 ± 2.20

<0.01

Data are presented as mean ± standard deviation and calculated with Mann–Whitney U test

aPregnancy at 16 to 24 weeks

bPregnancy at 36 weeks

cThree days after delivery

dSix weeks after delivery

eSix months after delivery

The incidence of self-reported urinary incontinence was analyzed by the generalized estimating equations with results produced as follows: There is a significantly lower incidence of self-reported urinary incontinence in the PFME group than the control group (odds ratio = 1.63, 95% confidence interval between 1.52 and 1.74, P < 0.05) during late pregnancy (36 pregnancy weeks) and 6 months postpartum (Table 3); with regard to the incontinence–delivery method relationship, we found that in either the PFME group (Table 4) or the control group (Table 5), women with vaginal delivery had a higher self-reported urinary incontinence rate than women with cesarean delivery; overall (putting both groups together), in the postpartum period, women experienced urinary incontinence more frequently if they delivered through the vagina route than by cesarean section (odds ratio = 4.89, 95% confidence interval between 4.37 and 5.46, P < 0.05).
Table 3

Self-reported urinary incontinence during pregnancy and after birth

Time

PFME

Control

P value

16–24 weeks

41 (27)

45 (30)

0.61

36 weeks

52 (34)

76 (51)

<0.01

3 days after delivery

46 (30)

62 (41)

0.07

6 weeks after delivery

38 (25)

53 (35)

0.06

6 months after delivery

25 (16)

42 (27)

0.04

Data are presented as n (%) and calculated with Chi-square test

Table 4

Self-reported urinary incontinence in PFME group delivered by vaginal or cesarean delivery

Time

VD (n = 102)

CS (n = 48)

P value

Odds ratio (95% CI)

16–24 weeks

28 (27)

13 (27)

0.96

1.02 [0.47, 2.21]

36 weeks

34 (33)

18 (38)

0.62

0.83 [0.41, 1.71]

3 days after delivery

39 (38)

7 (15)

<0.01

3.63 [1.48, 8.88]

6 weeks after delivery

32 (31)

6 (13)

0.01

3.2 [1.24, 8.29]

6 months after delivery

21 (21)

4 (8)

0.06

2.85 [0.92, 8.83]

Data are presented as n (%) and calculated with generalized estimating equations

VD vaginal delivery, CS cesarean section, CI confidence interval

Table 5

Self-reported urinary incontinence in the control group who delivered by vagina or cesarean section

Time

VD (n = 107)

CS (n = 43)

P value

Odds ratio (95% CI)

16–24 weeks

31 (29)

14 (33)

0.67

0.84 [0.39, 1.81]

36 weeks

52 (49)

24 (56)

0.42

0.75 [0.37, 1.52]

3 days after delivery

57 (52)

5 (12)

<0.01

8.34 [3.05, 22.8]

6 weeks after delivery

48 (45)

5 (12)

<0.01

6.18 [2.26, 16.9]

6 months after delivery

39 (35)

3 (7)

<0.01

7.05 [2.04, 24.3]

Data are presented as n (%) and calculated with generalized estimating equations

VD vaginal delivery, CS cesarean section, CI confidence interval

Discussion

Using the validated UDI-6 and IIQ-7 questionnaires, we found that PFME led to improvement of urinary incontinence and quality of life in late pregnancy and up to 6 months postpartum. Reilly et al. [10] found no difference between the study groups on any scale of the Kings Health Questionnaire, but a higher score for the general health measure in the Short Form-36 in the PFME group. After factor analysis, which resulted in four domains of Incontinence Impact Questionnaire-30 [14], Woldringh et al. [15] reported that the impact on emotional health and recreational activities was stronger than on social relationships and physical activities, but no difference existed between the PFME and control groups. The reasons for different results among the three studies might be due to different questionnaires, inclusion criteria for participants, and the frequency of the pelvic floor muscle training.

In addition to total scores of UDI-6 and IIQ-7, individual scores for question 2 (urge incontinence), question 3 (stress incontinence), and question 4 (drops) in UDI-6 were also found to have showed significant difference between the PFME group and the control group at 36 gestational weeks and up to 6 months postpartum. Our results are similar to those of several previous randomized controlled trials [6, 7, 10, 16]. Sampselle et al. [6] reported significantly less urinary incontinence in the training group at 35 weeks’ pregnancy, 6 weeks postpartum, and 6 months postpartum when compared with the control group. In the study by Morkved et al. [7], following attendance in an intensive 12-week PFME program, the training group had significantly less urinary incontinence than control group at 36 weeks’ pregnancy (32% versus 48%) and 3 months after delivery (20% versus 32%). Reilly et al. [10] studied only women with bladder neck mobility and found that fewer women in the training group reported urinary incontinence at 3 months postpartum, 19.2% compared with 32.7% in the control group. In their study of 80 pregnant women with urinary incontinence in their 20th to 34th weeks of pregnancy who were randomly allocated to PFME group or to control group, Dinc et al. [16] reported that 43.2% of the women in the PFME group continued to have urinary incontinence at 36 to 38 gestational weeks, which decreased to 17.1% at 6 to 8 weeks postpartum; but that in the control group, as high as 71.4% of the women had persistent urinary incontinence at 36 to 38 gestational weeks; and even 6–8 weeks postpartum, 39.4% still had it.

Results from a randomized controlled trial [15] addressing antenatal PFME in women with urinary incontinence during pregnancy are inconsistent with those of the aforementioned trials. Using urinary incontinence severity scale and a 7-day bladder diary for outcome measurement, Woldringh et al. [15] reported that PFME did not lower the incidence of urinary incontinence in pregnant women any more than the control group did. However, in their series, women in the control group received some instruction on PFME as well. In our series, the PFME group had significantly lower urinary incontinence than the control group in late pregnancy (34% versus 51%) and at 6 months postpartum (16% versus 27%), but did not demonstrate the same result at early postpartum. We speculated that vaginal delivery might contribute to pelvic floor damage through muscular, fascial, and nervous injuries, in which tissue damage took time to restore through the early postpartum period. In a previous study, there was electromyographic evidence of re-innervation in the pelvic floor muscles after first vaginal delivery in 80% of those studied during early postpartum period [17]. It is possible that episiotomy interferes with PFME at immediate postpartum period; however, there were similar episiotomy rates in both groups of exercise and control, thereby rendering the influence on the outcome insignificant. Other potential obstetric risk factors of postpartum urinary incontinence consistently reported in the medical literature included birth weight, duration of labor, and forceps delivery [18, 19].

SUI in young women is usually the result of pelvic floor damage during vaginal delivery. Whether cesarean delivery may prevent such damage and consequent development of postpartum urinary incontinence is uncertain. Rortveit et al. [20] investigated the association between urinary incontinence and childbirth in a community-based cohort of 15,307 women. Their study demonstrated an increased risk of urinary incontinence among women who delivered by cesarean section as compared with nulliparous women who never delivered. On this result, Rortveit et al. [20] commented that the mechanical strain accumulated antenatally may add to the risk associated with pregnancy itself. In one study, cesarean section seemed to protect against the development of postpartum urinary incontinence, but 3 months after delivery, the statistically significant influence of the mode of delivery had disappeared [19]. In our study, the mode of delivery showed a significant difference in the prevalence of self-reported urinary incontinence between women who had vaginal delivery and those who delivered by cesarean section in the PFME group (Table 4) and the control group (Table 5) at postpartum period. However, at 6 months postpartum, the difference had disappeared in the PFME group but not in the control group, indicating that pelvic floor muscle training seems to be beneficial to the restoration of continence in women who delivered vaginally. Using the generalized estimating equations analysis to assess urinary incontinence rate, we found the PFME group had a significantly lower self-reported urinary incontinence rate than the control group. Likewise, by the generalized estimating equations, vaginal delivery was associated with a significantly higher postpartum urinary incontinence rate than cesarean delivery.

The limitations of our study included lack of data from long-term follow-up and no objective assessments such as the measurement of pelvic floor muscle strength, pad test, perineal ultrasound, or urodynamic study. In addition, for ethic consideration, we cannot forbid women of control group to perform exercise by themselves that possibly have an impact of antenatal PFME on urinary incontinence. Long-term follow-up is important to assess continence following further pregnancies; hence, we believe in using diagnostic tests and outcome measurements that would cause just minimal discomfort to the participants if any because they are all healthy pregnant women. In our study, only subjective questionnaires were used as outcome measurement. Several studies investigating the effect of antenatal PFME on pregnancy-related urinary incontinence showed poor compliance to receive invasive tools for survey [5, 6, 10]. Regarding the measurement of pelvic floor muscle strength, previous studies were inconsistent: Some investigators found that pelvic floor muscle strength was stronger in the study group who did the PFME in the antenatal period [7, 16]. Other studies showed no difference in pelvic floor muscle strength between the groups PFME and control [6, 10]. However, the strength of pelvic floor muscle is not necessarily related to the rate of postpartum urinary incontinence. Previous studies showed fewer women in the training group reporting significantly lower postpartum urinary incontinence but no difference in pelvic floor muscle strength between the groups after exercise [6, 10]. The monitoring of the adherence to the training program between sessions is more important than the total number of session followed [14]. In this study, all women of exercise group were individually instructed about correct pelvic floor muscle contractions. The exercise group followed a 12-week designed PFME course, including group training once per week and daily training at home. Skilled physiotherapists led the training groups, giving instructions on PFME. They especially emphasized the importance of adherence to the training protocol and urged the pregnant women to follow the program strictly. Over 80% of our participants were adherent in that they reported practice of PFME at least 75% of the time.

Conclusion

PFME applied in pregnancy is effective in the treatment and prevention of urinary incontinence during pregnancy, and this effect may persist to the postpartum period. The delivery method also makes an impact on urinary incontinence: In both PFME and control groups, more women who delivered vaginally experienced postpartum urinary leakage than those delivered by cesarean section.

Acknowledgment

This work was supported by the Medical Research Project, Chang Gung Memorial Hospital (CMRPG370321 and BMRP412). The authors would like to thank Drs. Yao-Lung Chang, Chao-Nin Wang, and Ho-Yen Chueh, and physiotherapists of our urogynecological division for their assistance to accomplish this study.

Conflicts of interest

None.

Copyright information

© The International Urogynecological Association 2010