Table 1

Studies assessing the effect of pelvic floor muscle exercises during pregnancy to prevent urinary incontinence including both women with and without urinary incontinence at inclusion

AuthorDesignSubjectsTraining protocolLosses to follow-up/AdherenceOutcomes (Numbers and percentage (%))
Sampselle et al302-Arm RCT
1. Control (n=38): Routine care
2. Intervention (n=34): A tailored PFMT programme
N=72 primigravid women recruited at 20-week pregnancy. Some women had existing UI. Groups comparable at baseline. Single centre, USA1. Control: routine care
2. A tailored PFMT programme beginning with muscle identification progressing to strengthening. 30 contractions/ at max or near-max intensity from 20-week pregnancy. Correct VPFMC checked
Losses to follow-up: 36
adherence PFMT:
  • ▸ 35-week pregnancy: 85%

  • ▸ 1-year postpartum: 62–90%

Adverse events not stated.
Self-reported adherence.
Partial ITT analysis
Change in mean UI symptom score:
          Control  Intervention p
35-week pregnancy :  0.20  −0.02   0.07
6-week postpartum: 0.25  −0.06   0.03
6-month post partum: 0.15  −0.11   0.05
12-month post partum: 0.06  0.00   0.74
PFM strength: NS difference (low numbers)
Hughes et al(abstract) 272-arm RCT
1. Control (n=583): routine care
2. Intervention (n=586): a tailored PFMTprogramme
N=1169 pregnant nulliparous women recruited at 20-week pregnancy.
Some women had existing UI.
Single centre, UK
1. Control: routine care that may have included advice on PFMT.
2. Intervention: one individual session with physiotherapist, and one group PFMT session between 22-week and 25-weekpregnancy. Home training daily for up to 11 months.VPFMC checked
Losses to follow-up:
40% at 6-week postpartum
27% at 3-month postpartum
34% at 6-month postpartum
461/586 women in the intervention group attended the PFMT session
SUI Bristol Female Urinary Tract Symptoms Questionnaire:
          Control  intervention
36-week pregnancy:   66%   61%
OR (95% CI) 0.78 (0.59  to 1.04)
6-month postpartum:  38%   36%
OR (95% CI) 0.90 (0.64 to 1.28)
Reilly et al29
Agur et al31




8-year follow-up
2-Arm RCT
1. Control (n=129): routine care
2. Intervention (n=139): 20-week intensive PFMT
1. Control (n=85):
2. Intervention(n=79)
N=268 primigravid, continent women with increased bladder neck mobility recruited at 20-week pregnancy. Single centre, UK164/268 (61%) of the original group1. Control: routine antenatal care (verbal advice)
2. Intervention: individual PFMT with physiotherapist at monthly intervals from 20-week until delivery, with additional home exercises 3 sets of 8 contractions (each held for 6 s) repeated twice daily. Instructed to contract the PFM when coughing or sneezing
Losses to follow-up at 12 months: 14%
Adherence PFMT:
  • ▸ 11% completed less than 28 days of PFMT

  • ▸ 46% completed 28 days or more of PFMT

Adverse events not stated ITT analysis
  • ▸ 38% in the intervention group were doing PFMT twice or more per week

Self-reported UI at 3-month postpartum:
1. Control: 36/110 (32.7%)
2. Intervention: 23/120 (19.2%)
RR (95% CI) 0.59 (0.37 to 0.92) p=0.023
Quality of life: higher score in the exercise group
p=0.004
Pad test: NS difference
Bladder neck mobility: NS difference
PFM strength: NS difference
Self-reported UI at 8 years follow-up:
1. Control: 38.8%
2. Intervention: 35.4%
p=0.75
Mørkved et al28







Mørkved et al(abstract) 32
6-year follow-up
2-Arm RCT
1. Control (n=153): customary information from general practitioner/midwife
2. Intervention (n=148 m): 12-weeks of intensive PFMT

1. Control (n=94)
2. Intervention (n=94)
N=301 primigravid women recruited at 20-weekpregnancy. Some women had existing UI. Three outpatient physiotherapy clinics in Norway





188/301 (62%) returned the questionnaire
1. Control: customary information from general practitioner/midwife. Not discouraged from PFMT. Correct PFM contraction checked at enrolment
2. Intervention: 12 weeks of intensive PFMT (in a group) led by physiotherapist, with additional home exercises 10 max contractions (each held for 6 s) and to the last 4 were 3–4 fast contractions added, repeated twice daily, between 20-week and 36-week pregnancy. Correct VPFMC checked at enrolment
Control group received information about the results of the trial and the training programme, about 1 year  after delivery
Losses to follow-up 12/301(5 intervention and 7 controls).
Adherence to PFMT:
  • ▸ 81% adherence to PFMT in the intervention group

Adverse events not stated
ITT analysis
45% adherence to PFMT in both groups
Self-reported UI at 36-week pregnancy:
1. Control: 74/153 (48%)
2. Intervention: 48/148 (32%)
RR (95% CI) 0.67 (0.50 to 0.89) p=0.007
UI at 3 months postpartum:
1. Control: 49/153 (32%)
2. Intervention: 29/148 (19.6%)
RR (95% CI): 0.61 (0.40 to 0.90) p=0.018
PFM strength: sign difference in favour of the intervention group
UI at 6 years follow-up:
1. Control: 17%
2. Intervention: 23%
p=0.276
Gorbea Chávez et al(abstract)262-arm RCT
1. Control (n=34 after dropouts) No PFMT
2. Intervention (n=38 after dropouts) PFMT
75 pregnant nulliparous continent women recruited at 20-week pregnancy.Single setting, Mexico1. Control: requested not to perform PFMT during pregnancy or postpartum
2. Intervention: individual PFMT with physiotherapist. 10 VPFMC each held for 8 s each followed by 3 fast 1 s contraction; 6 s rest. Clinic appointments weekly for 8 weeks, then weekly phone calls up to 20 weeks
Biofeedback and training diary. Correct VPFMC checked
Losses to follow-up 3/75 (4%)
Adherence to PFMT:
84% attended 7 or 8 physiotherapy appointments.
ITT analyses
Urinary incontinence:
         Control Intervention  p 28-week pregnancy:  17%  0     <0.05 35-week pregnancy: 47%  0   <0.05 6 weeks postpartum:  47%  15%    <0.05 
Mason et al222-arm RCT
1. Control (n=148)
2. Intervention (n=141):PFMT
N=311 nulliparous pregnant women with no symptoms of SUI at 11–14-week pregnancy
Two hospitals in England
1. Control
2. Intervention: 45 min physiotherapy class once per month for 4 months. Additional home exercises 8–12 max contractions (each held for 6 s) and to the last 4 were 3–4 fast contractions added, repeated twice daily, between 20-week and 36-week pregnancy. Correct VPFMC checked at enrolment in most women
Losses to follow-up: 8%
Some significant differences between responders and non-responders
90 women (31.4%) completed all sets of questionnaires
91/141 (49.1%) in the intervention group attended a PFMT class
Significantly more PFMT in the intervention group compared to the control group
Self-reported UI at 36-week pregnancy:
1. Control: 51/96 (53%)
2. Intervention: 24/60 (40%)
OR (95%CI) 1.7 (0.884 to 3.269) p=0.138
UI at 3 months postpartum:
1. Control: 33/80 (41.3%)
2. Intervention: 23/68 (33.8%)
OR (95%CI) 1.374 (0.702 to 2.688) p=0.397
No significant difference in symptoms and episodes of UI, between groups
Ko et al212-arm RCT
1. Control (n=150): Routine care
2. Intervention (n=150): 20 weeks of intensive PFMT
N=300 nulliparous women recruited at 16–24-week pregnancy. Some women had existing UI. Single centre, Taiwan1. Control: routine antenatal care
2. Intervention: Individual PFMT with physiotherapist once per week between 20–36-weekpregnancy, with additional home exercises three sets of eight contractions (each held for 6 s) repeated twice daily. Instructed to contract the PFM when coughing or sneezing
Losses to follow-up: no.
Adherence PFMT:
  • ▸ 87% practised PFMT at least 75% of the time

Adverse events not stated.
ITT analysis
Self-reported UI at 36-week pregnancy: sjekk
1. Control: 76/150 (51%)
2. Intervention: 52/150 (34%)
p<0.01
Self-reported UI at 3days postpartum:
1. Control: 62/150 (41%)
2. Intervention: 46/150 (30%)
p=0.06
Self-reported UI at 6 weeks postpartum:
1. Control: 53/150 (35%)
2. Intervention: 38/150 (25%)
p=0.06
Self-reported UI at 6-month postpartum:
1. Control: 42/150 (27%)
2. Intervention: 25/150 (16%)
p=0.04
Significant improvement of in the intervention group in Scores on the Incontinence Impact Questionnaire and Urogenital Distress Inventory, in late pregnancy and up to 6-month postpartum
Bø and Haakstad172-arm RCT
1. Control (n=53):
2. Intervention (n=52): 12–16-week aerobic fitness class including PFMT
N=105 nulliparous women recruited within 24-week pregnancy. Some women had existing UI.
Single centre, Norway
1.Control:
2. Intervention: 12–16 weeks of aerobic exercise classes twice per week during pregnancy, including intensive PFMT (in a group) led by aerobic instructor. Additional home exercises 10 max contractions (each held for 6 s) and to the last 4 were 3–4 fast contractions added ×3/day. Correct VPFMC was not checked at enrolment
Losses to follow-up: 21/105 (10 intervention and 11 control).
Adherence to training sessions: 40%
Adverse events not stated
Not IIT analysis
Self-reported UI at 36–38-week pregnancy:
1. Control : 7/53
2. Intervention: 9/52
Self-reported UI at 3-month postpartum:
1. Control: 6/53
2. Intervention: 5/52
No significant difference
Stafne et al232-arm RCT
1. Control (n=426): customary information from general practitioner/midwife.
2. Intervention (n=429): 12-week intensive PFMT
N=855 pregnant women recruited
20-week pregnancy. Some women had existing UI. Two hospitals in in Norway
1.Control: customary information from general practitioner/midwife and written information. Not discouraged from PFMT
2. Intervention: 12 weeks of exercise class including led by physiotherapist, with additional home exercises 3×10 max contractions (each held for 6 s and to the last 4 were 3–4 fast contractions added) at least three times per week between 20-week and 36-week pregnancy. Correct VPFMC checked at enrolment
Losses to follow-up: 93/855 (32 intervention and 61 controls).
Adherence to PFMT:
  • ▸ 67% adherence to PFMT in the intervention group

  • ▸ 40% adherence to PFMT in the control group

No adverse events
ITT analysis
Self-reported UI at 34–38-week pregnancy:
Any UI
1. Control: 192/365 (53%)
2. Intervention: 166/397 (42%) p=0.004
UI once per week or more
3. Control: 68/365 (19%)
4. Intervention: 44/397 (11%) p=0.004
Dias A et al(abstract)183-arm RCT
1. Control group (n=29)
2. Supervised group (n=29)
3. Observational group (n=29)
N=87 primigravidas women recruited 18-week pregnancy Some women had existing UI. Single centre Brazil1. Control: no exercising
2. Supervised:exercising under supervision of a physiotherapist monthly + daily home exercises
3. Observational group: unsupervised daily home exercises
Losses to follow-up: ?Self-reported UI at 38-week pregnancy:
1. Control: 96%
2. Supervised: 6,9%
3. Observational: 6.9%
  • ITT, intention to treat analysis; max, maximum; NS, non-significant; PFM. pelvic floor muscles; PFMT, pelvic floor muscle training; RCT, randomised controlled trial; RR, relative risk; SUI; stress urinary incontinence; UI, urinary incontinence; VPFMC, voluntary pelvic floor muscle contraction.