Secondary Logo

Journal Logo

Systematic Review

Exercise Interventions to Improve Pelvic Floor Muscle Functioning in Older Women With Urinary Incontinence: A Systematic Review

Fricke, Anja MSc1; Lark, Sally D. PhD1; Fink, Philip W. PhD2; Mundel, Toby PhD2; Shultz, Sarah P. PhD1,3

Author Information
Journal of Women's Health Physical Therapy: July/September 2021 - Volume 45 - Issue 3 - p 115-125
doi: 10.1097/JWH.0000000000000202
  • Free



Urinary incontinence, as defined by the International Continence Society (ICS), is a complaint of any involuntary urinary leakage.1–4 The World Health Organization (WHO) acknowledges incontinence as a set of diseases, while the International Classification of Functionality recognizes the associated severe disabilities.5 Global demographic trends indicate that the incidence of urinary incontinence will rise in the coming years, adding a significant health and social burden as well as increased economic costs.1 It is estimated that, worldwide, more than 200 million people live with urinary incontinence, and women are more affected than men.2,6 The incidence of incontinence increases from middle age onward and is associated with a reduced quality of life.2 In about 25% of people living with urinary incontinence, the health issue is considered to be severe; yet half the people living with incontinence do not discuss their problems with a health care professional or even family members or friends out of embarrassment.2,7

Chronic incontinence can be classified into 5 different types: overflow, functional, stress, urgency, and mixed incontinence.2,3,8 Stress urinary incontinence is considered to be the most common type of incontinence in older women.3,6,9–11 Stress urinary incontinence is defined as an involuntary loss of urine during activities that increase the intra-abdominal pressure such as sneezing or coughing.9,12 Urgency urinary incontinence is commonly a result of detrusor over activity. Stress and urgency urinary incontinence often coexist resulting in mixed urinary incontinence. Aging increases the prevalence of stress, urgency, and mixed incontinence and they most likely occur during the perimenopausal period (approximately 50 years of age).4,6,10,13 Menopause and the physiological effects of declining estrogen combined with aging increase the risk for postmenopausal women to experience urogynecology dysfunction, which can lead to urinary incontinence.14 It is thought that women undergo anatomical and neuromuscular changes and injury during childbirth, which decreases pelvic floor muscle strength. The resulting loss of muscular strength combined with the lack of tissue resilience resulting from menopause prohibits support of the bladder neck during physical stress and results in leakage.2–4,10 Some women might not experience incontinence until a small loss of pelvic floor muscle strength occurs within the normal aging process—particularly after menopause, through other injuries, or through some medications that might induce incontinence.2,4,15 Urinary incontinence, even if considered mild, can have a large impact on overall well-being and quality of life. Patients who live with urinary incontinence are more likely to have depression, reduced employment and work productivity, limited social and sexual function, and may be dependent on caregivers.5,6,16 Furthermore, urinary incontinence increases the likelihood for the elderly to be admitted to long-term care and the associated loss of independence and poorer quality of life.17,18 Thus, prevention of urinary incontinence in the early stages has become increasingly important.

Therapy options in the early stages of urinary incontinence focus on lifestyle and behavioral modifications, while options in the later stages can include surgery, medication, nerve stimulation, and containment products such as pads and catheters.2,5,17 Treatment of urinary incontinence with surgery and/or medications is associated with a range of side effects (including dry mouth and constipation) and high health care costs.2,19,20 Behavioral modifications have shown extensive benefits with very few risks in treating urinary incontinence. The WHO and the ICS recommend pelvic floor muscle training to treat stress, urgency, or mixed urinary incontinence in women.10 Women undergoing pelvic floor muscle training are believed to be 7 times more likely to be cured of urinary incontinence than those with no treatment.2 A previous systematic review by Dumoulin and Hay-Smith21 found that pelvic floor muscle training is better than no treatment or sham treatment, and the authors concluded that pelvic floor muscle training can cure or improve all forms of urinary incontinence. Although there were limited data comparing the efficacy of different types of pelvic floor muscle training, Hay-Smith et al22 concluded that pelvic floor muscle training with regular health professional contact was more effective than pelvic floor muscle training with little or no contact. Only one review examined the effects of multifaceted self-management interventions for older women with urinary incontinence.23 However, due to the quality of interventions included in the review, there was insufficient evidence to determine whether any of the interventions was superior to others in improving urinary incontinence symptoms. Therefore, the purpose of this review is to explore pelvic strengthening in a broader context of exercise and across multiple forms of chronic urinary incontinence. Specifically, we seek to examine which exercise interventions can show improvements in pelvic floor muscle strengthening and functioning and understand the impact of these on urinary incontinence in women older than 50 years.


Identification of Studies

This review protocol was not registered with PROSPERO; however, the PRISMA guidelines and checklist were used to guide this systematic literature review. A search of Medline, PsycINFO, Web of Science, CINAHL, and Scopus was conducted between April 2018 and September 2018. An additional updated search was performed in May 2020. Search criteria were chosen by the author (A.F.) with support from an independent professional librarian and verified again by the author (A.F.). The following search terms were defined: (woman* OR women* OR female*) AND (“therapeutic exercise*” OR “resistance training” OR “strength training” OR “exercise prescription*” OR “physical activity*” OR “physical training & conditioning” OR “physical training and conditioning” OR “physical exercise*” OR “exercise intervention*”) AND (aged OR elderly OR “older people” OR “very old” OR “65+ years” OR “old age” OR geriatric OR gerontology* OR “middle age*” OR “45-64 years” OR “late middle age*” OR “80 and over” OR “80+ years”) AND (incontinence* OR urinary OR urology*) AND pelvic.

Eligibility Criteria

All titles were scanned and duplicates removed by the first author (A.F.). Articles were further screened on their abstracts and full-text articles were retrieved. Full-text articles were reviewed to ensure that the research design conformed to the inclusion/exclusion criteria outlined later. When a decision could not be reached by the investigating author, 2 other authors (S.S. and S.L.) were consulted to determine eligibility for inclusion.23

Results from peer-reviewed journals during the years 1988 to May 2020 were included. Inclusion/exclusion criteria based on study design, participants, intervention, and outcomes were established a priori. Only randomized controlled trials with healthy women 50 years and older were included in this review. Studies must have had at least 1 control group and 1 exercise-only group. The exercise intervention group had the objective of training pelvic floor musculature. Outcome measures included the following: quality of life, urinary incontinence, and pelvic floor muscle strength—all assessed via validated measures. Quality-of-life data were compared to minimal clinically important differences (MCIDs) where appropriate MCIDs had been established and data were available.24

Critical Appraisal

Two authors (A.F. and S.L.) assessed the methodological quality of each included study using the previously published and validated PEDro scale.25 The PEDro scale was developed by the Physiotherapy Evidence Database to determine the quality of clinical trials; the scale is used extensively in systematic reviews across physiotherapy interventions.26 The scale consists of 11 yes-or-no questions relating to the internal validity and statistical information provided. Emphasis was placed on randomization, group comparison, measured outcomes, and outcome analysis. Critical appraisals were calculated out of 10 total points, as the first question in PEDro is not scored. Study quality was classified as excellent (7-10), good (5-6), fair (3-4), and poor (≤2) using the total score.27 If there was a lack of consensus between the 2 authors (A.F. and S.L.) concerning the scores of each criterion, a third author (S.S.) acted as arbitrator to reach agreement.


Study Selection

The initial data search showed 649 results from which only 8 studies met all the inclusion criteria.8,12,17,28–32 Second- and third-order reference lists were checked manually for relevant articles and did not reveal any additional eligible studies. The full article selection progress is shown in the Figure.

Selection of articles.

Assessment of Methodological Quality

Table 1 presents PEDro scale scores for each study. Three studies17,28,30 were classified as excellent research quality, while the remaining 5 studies8,12,29,31,32 were classified as good research quality. There was homogeneity in the PEDro scores across the studies (all studies scored ≥6); the studies scored similarly to the adherence hierarchy of PEDro items as outlined by de Morton.33 The most adhered to items on the PEDro scale include random allocation, between-group statistical comparison, point and variability measures, similarity at baseline, and >85% follow up. The studies in this systematic review followed that hierarchy and adhered to all of the aforementioned criteria, as well as the concealed allocation criterion. The least adhered to items on the hierarchy are therapist and subject blinding and intention-to-treat analysis.33 Because of the nature of these exercise interventions, very few studies complied with those final criteria. Kim et al12 and Tak Erwin et al30 included blinding of assessors on at least 1 key outcome, while only Martinho et al28 and Tak Erwin et al30 analyzed data by intention-to-treat. Tak Erwin et al30 research scored the highest out of all studies, with a score of 8 out of a possible 10.

Table 1. - PEDro Scores Across All Criteria
Criterion Overall PEDro Score
Random Allocation Concealed Allocation Similar at Baseline Blinding of Subjects Blinding of Therapists Blinding of all Assessors Measures Obtained From Minimum 85% Intention to Treat Between-Group Statistical Comparisons Point Measures Included
Aslan et al32 Y Y Y N N N Y N Y Y 6
Ferreira et al8 Y Y Y N N N Y N Y Y 6
Kim et al17 Y Y Y N N N Y N Y Y 6
Kim et al12 Y Y Y N N Y Y N Y Y 7
Martinho et al28 Y Y Y N N N Y Y Y Y 7
Ong et al29 Y Y Y N N N Y N Y Y 6
Tak Erwin et al30 Y Y Y N N Y Y Y Y Y 8
Tsai and Liu31 Y Y Y N N N Y N Y Y 6
Abbreviations: N, no; Y, yes.

Participant characteristics

A total of 629 women (319 in intervention and 310 in control) across all studies had a mean age of 61.9 years. The mean number of participants in the intervention groups was 39 and the mean number of participants in the control group was 38. Four of the studies only included participants who were diagnosed with stress urinary incontinence.8,17,29,31 Tak Erwin et al,30 Kim et al,12 Aslan et al,32 and Martinho et al28 included participants who were diagnosed with stress urinary incontinence, urgency incontinence, and mixed incontinence as well as some participants who were continent. From studies of mixed diagnoses where data were available,12,28,32 11% were continent, 19% were diagnosed with stress urinary incontinence, 34% were diagnosed with urgency urinary incontinence, and 36% were diagnosed with mixed urinary incontinence.

Study Interventions

A summary of the study characteristics can be found in Table 2. Exercise interventions varied by frequency, duration, and mode of exercise. The most common duration was 3 to 4 months,12,17,29,31 with 2 studies completing shorter durations of 5 to 8 weeks28,32 and 2 studies completing longer durations of 6 months.8,30 Exercise frequency was not described in all studies29,32; in those studies that provided details, frequency ranged from 3 times daily,8 once daily,28,31 twice weekly,12,17 and once weekly.30 The majority8,12,17,29,31,32 of exercise interventions focused on specific pelvic floor muscle training, although Martinho et al28 and Tak Erwin et al30 did include prescribed exercises that did not directly activate the pelvic floor muscles.

Table 2. - Intervention Type and Outcome Measures of Selected Articles
Baseline Sample Size; Mean Age, y Type of Exercise Intervention Initial Training Model Ongoing Monitoring Duration and Frequency of Intervention Outcome Measures (Assessment)
Aslan et al32 I = 25 and C = 25
78.8 ± 4.8
Bladder training and Kegel exercise program Home-based (unsupervised) Weekly interviews 6-8 wk (1) Quality of life (King Health questionnaire, mini-mental test, Rankin scale)
(2) Severity of urinary incontinence (ICS 1-h pad test)
(3) Pelvic floor muscle strength (digital palpation)
Ferreira et al8 I = 17 and C = 17
50.7 ± 9.3
Pelvic floor muscle-strengthening program including a second intervention group with additional supervised exercise sessions Home-based (unsupervised) for control group; home-based (unsupervised) plus supervised weekly sessions for intervention group Monthly phone calls 6 mo; 3 times daily (1) Pelvic floor muscle strength (perineometer)
(2) Urinary incontinence (1-h pad test and voiding diary)
(3) Subjective perception of improvement (self-formulated question)
Kim et al17 I = 35 and C = 35
76.6 ± 5.0
Pelvic floor muscle-strengthening exercise sessions Group exercises (supervised); home-based (unsupervised) during follow-up Twice a week during group exercises; monthly during follow-up 12 wk; 60-min sessions twice weekly (1) Severity of urinary incontinence and cure rate (6-point scale based on the International Consultation on Incontinence Questionnaire)
(2) Function fitness test (including grip strength, walking speed, hip adductor muscle strength)
Kim et al12 I = 63 and C = 64
76.1 ± 4.3
Pelvic floor muscle-strengthening exercise sessions Group exercises (supervised); home-based (unsupervised) during follow-up Twice a week during group exercises; monthly during follow-up 3 mo; 60-min session twice weekly (1) Severity of urinary incontinence (urinary diary)
(2) Functional fitness (including grip strength, walking speed, hip adductor muscle strength)
Martinho et al28 I = 30 and C = 30
61.9 ± 8.6 and 61 ± 8.5
Two exercise interventions:
(1) Abdominopelvic training with a virtual reality game (intervention group)
(2) Strength exercises focusing on abdominopelvic cavity (control group)
Individual exercise sessions (supervised) Twice a week during exercises 5 wk; twice weekly (1) Pelvic floor muscle strength (digital palpation and vaginal dynamometry)
Ong et al29 I = 21 and C = 16
50.7 ± 11.0
Standardized pelvic floor muscle-strengthening program including biofeedback device for intervention group Home-based (unsupervised) with additional individual monthly exercise sessions (supervised) Once a month during exercise 16 wk (1) Severity of urinary incontinence (Australian pelvic floor questionnaire)
(2) Pelvic floor muscle strength (palpation using modified Oxford scale)
(3) Total urinary score, social life score, bothersome score, and subjective cure (Australian pelvic floor questionnaire)
Tak Erwin et al30 I = 85 and C = 70
84.6 ± 6.5
Behavioural instructions to improve control of micturition and physical exercise program to increase functional ability Group exercises (supervised) Once a week during exercise 6 mo; 60-min sessions once weekly (1) Severity of urinary incontinence (3-d bladder diary)
(2) Physical performance (physical performance test)
(3) Quality of life (SF-12 questionnaire and Incontinence Quality of Life Instrument)
Tsai and Liu31 I = 54 and C = 54
55.9 ± 9.1
Pelvic floor muscle-strengthening program including previous digital vaginal palpation to aid recognition of pelvic floor muscle Home-based (unsupervised) Telephone contact once a week 12 wk; daily (1) Severity of urinary incontinence (1-h pad test)
Abbreviations: C, control; I, intervention; ICS, International Continence Society.

Four8,29,31,32 of the studies were home-based exercise programs; however, 3 studies8,29,31 included regular communication with participants. Aslan et al32 provided participants with a scheduled bladder training and Kegel exercise instructions, which included explanations of the anatomy of the pelvic floor muscle system as well as digital palpation to aid awareness of the pelvic floor muscle. Weekly meetings with participants provided instructions for next week's program. Ferreira et al8 provided participants with a home-based exercise program detailing 8 to 10 pelvic floor muscle contractions in different positions 3 times a day for 6 months. A second group also received weekly supervised sessions. Participants were also educated on the anatomy and physiology of the pelvic floor muscles and lower urinary tract. Tsai and Liu31 provided participants with a home-based pelvic floor muscle exercise program and used deep vaginal palpation to aid participants with contracting the correct muscle. Ong et al29 had 2 different exercise groups performing home-based pelvic floor muscle training. Both groups underwent a pelvic floor muscle training protocol involving endurance and speed training. One exercise group was also provided with a biofeedback device to aid participants with correct muscle contraction29

Aslan et al32 interviewed participants at the end of each week for encouragement and to ensure correct progression. Ferreira et al8 conducted monthly phone calls with participants for encouragement. Participants in Tsai and Liu31 study attended the outpatient clinic every month, whereby deep vaginal palpation was performed to ensure that pelvic floor muscle exercises were performed correctly. Participants also used a diary to record daily pelvic floor muscle exercises and had weekly telephone contact with the researcher. Ong et al29 provided participants with individual monthly sessions with a physiotherapist for encouragement and correct progression.

Four studies12,17,28 conducted supervised exercise sessions. Kim et al12,17 educated participants on the anatomy of the pelvic floor muscle before performing supervised specific pelvic floor muscle exercises in different positions as well as general fitness exercises. Martinho et al28 individually supervised training but used 2 different exercise protocols. One group used a virtual reality protocol whereby participants used a gaming console to perform different exercises using pelvic movements that did not cause direct activation of the pelvic floor muscle. Four different exercises lasted 5 minutes each followed by abdominopelvic and lower limb stretching. The other exercise group performed abdominopelvic exercises on a gym ball. Participants performed pelvic mobility, stretching, strengthening, and relaxation exercises involving active pelvic floor muscle contractions. Tak Erwin et al30 included an exercise program, which consisted of weekly 1-hour training sessions for 22 weeks and additional homework exercises. Exercises included bladder training, pelvic floor muscle-strengthening, and physical exercises to improve functional ability.

Outcome Measures

Impact on Severity of Urinary Incontinence

Out of the 7 studies examining urinary incontinence severity, only 1 study30 showed no significant improvement of urinary incontinence symptoms in the intervention or control group. The remaining 6 studies8,12,17,29,31,32 reported significant improvements of urinary incontinence symptoms in their intervention groups. Urinary frequency complaints were significantly decreased in 3 studies12,17,29 and pad test weights were significantly reduced in the remaining 3 studies.8,31,32 Aslan et al32 using a home-based nonsupervised exercise program reported that 32% of participants had a decrease of pad weight. Tsai and Liu31 also involved a home-based exercise program but included digital palpation and reported a 21% decrease in pad weight. Exercise programs that included supervision however reported slightly higher improvements. Kim et al12,17 showed a 55% and 44% improvement of urine incontinence complains, respectively. Tak Erwin et al30 also reported a decrease in frequency of incontinence by 51%. Ferreira et al8 reported a decrease of pad test weight of 32% and 35% for the supervised intervention group and the home-based intervention group, respectively. Interestingly, Ong et al29 showed an 81% improvement of stress urinary incontinence complaints for the group using the biofeedback device as well as for the group without the biofeedback device. Across all studies home-based exercise programs8,29,32 showed an average improvement in urinary incontinence of 48%, compared with an average improvement of 46% in supervised programs,8,12,17,30 and 51% in home-based exercise programs that included a biofeedback device.29,31 Outcome measures can be found in Table 3.

Table 3. - Results of Selected Articles
Severity of Urinary Incontinence Pelvic Floor Muscle Strength Quality of Life
Aslan et al32 Statistically significant decrease in urgency and frequency complaints for treatment group compared with control group between preintervention and postintervention (P = .024 and P = .000 respectively).
Statistically significant decrease in involuntary urination in treatment group compared with control group after intervention (P = .000).
Statistically significant increase in strength in treatment group compared with control group after intervention (P = .000). Statistically significant improvement of general health perception, role limitations, physical limitations, and emotions in treatment group after intervention (P < .05).
Statistically significant improvement of severity measures in treatment group after intervention (P = .000).
Statistically significant improvement in “incontinence impact scores” in treatment group compared with control group (P < .05).
Ferreira et al8 Statistically significant improvements with pad tests (P < .05) in both intervention and control groups.
Less frequent episodes of urinary leakage in both intervention and control groups.
Statistically significantly higher increases in strength for intervention group that included supervised sessions (P = .006). Women in intervention group that included supervised sessions felt “improved, almost cured, or cured” compared with other intervention group with a significant difference between groups (P = .018).
The number needed to treat was 2.8 patients.
Kim et al17 Statistically significant decrease in frequency score of urine leakage for intervention group (P = .02). No statistically significant changes for control group. N/A 54.5% in intervention group reported being cured; 9.4% in control group reported being cured (P < .001).
The number needed to treat was 2.2 patients.
MCID for frequency score of urine leakage not met34
Kim et al12 Statistically significant decrease in urine leakage score for intervention group (P = .007). No statistically significant changes for control group.
44.1% in intervention group reported being cured of urine leakage.
Martinho et al28 N/A Digital palpation: Statistically significant improvements in pelvic floor muscle strength for both intervention groups (P < .05).
Vaginal dynamometry: Statistically significant improvement in average strength and endurance in virtual reality intervention group (P = .02 and P = .05, respectively). Statistically significant improvements in maximum strength and endurance for other intervention group (P = .02 and P = .04, respectively)
Ong et al29 Statistically significant improvement in stress urinary incontinence score for intervention and control groups (P < .001 and P = .001, respectively). No statistically significant differences between groups Statistically significant improvement of pelvic floor muscle strength in intervention and control groups (P = .001 and P = .059, respectively). No statistically significant differences between groups. No statistically significant differences of social life score, bothersome score, total urinary score, and subjective cure for either groups.
Tak Erwin et al30 No statistically significant improvements in urinary incontinence symptoms for either group. Reduction of urinary incontinence symptoms only reported in compliant study participants irrespective of group allocation. N/A No statistically significant improvements in quality of life for either group.
MCID for Incontinence Quality of Life Instrument met35
Tsai and Liu31 Statistically significant decrease in weight of test pads in intervention group (P < .001). No statistically significant differences for control group. N/A N/A
Abbreviations: MCID, minimal clinically important difference; N/A, not available.

Impact on Pelvic Floor Muscle Strength

All 4 studies8,28,29,32 that examined pelvic floor muscle strength saw significant improvements in strength after the intervention. Aslan et al32 showed that 56% of participants had increased pelvic floor muscle strength following the intervention. Interestingly, Ong et al29 also saw greater strength improvements in the group that used a biofeedback (87%) device compared with the group without the biofeedback device (33%). Both Aslan et al32 and Ong et al29 used the subjective measures to infer muscle strength of digital palpation with a rating scale. Ferreira et al8 (using a vaginal perineometer) found significant increases in strength in both intervention groups, but noted that percentage of strength improvements from baseline to posttest was significantly higher in the group that additionally included supervised exercise sessions (78%) compared with the group that only had home-based exercise programs (32%). Martinho et al28 also found significant improvements in pelvic floor muscle strength when using both a subjective and objective measure to examine pelvic floor muscle strength, although the objective measure using vaginal dynamometry was also able to further specify improvements between groups for average, endurance, and maximum strength. The exercise group using a virtual reality program showed an improvement of 29% in maximal strength, 36% in average strength, and 57% in endurance. Greater improvements however were seen in the exercise group that included pelvic floor muscle exercise, whereby maximal strength improved by 74%, average strength by 52%, and endurance by 75%.28 As only 4 studies measured pelvic floor muscle strength, results of improvement in pelvic floor muscle strength should be taken with caution.

Impact on Quality of Life Measures

Four studies measured quality of life via validated questionnaires and rating scales, while Ferreira et al8 assessed perception of improvement with a single question previously used by other studies but not confirmed to be valid. To assess quality of life, Aslan et al32 used the King Health questionnaire, Ong et al29 used the social life, bothersome, and subjective cure score from the Australian pelvic floor questionnaire, and Tak Erwin et al30 used the 12-Item Short Form Health Survey (SF-12) questionnaire and Incontinence Quality of Life Instrument. Two studies29,30 showed no significant improvements in quality of life after their intervention. Significant improvements of general health perception, role limitations, physical limitations, emotions, severity measures, and incontinence impact scores were found in one study using the quality-of-life assessment.31 Furthermore, a statistically significantly larger amount of women in the intervention groups reported being cured compared with the control groups.8,17 The number needed to treat for Ferreira et al8 was 2.8 patients, while the number needed to treat for Kim et al17 was 2.2 patients. A lack of data prevented MCID criteria analysis for Aslan et al.32 Ferreira et al8 and Ong et al29 used questionnaires and exercise protocols for which no MCIDs have been established yet. MCID analysis is presented for Kim et al17 and Tak Erwin et al30 in Table 3.

Follow-up Results

Only 312,17,32 out of the 8 studies analyzed in this review included follow-up assessments of at least 6 months. Aslan et al32 reported significant improvements in urinary frequency and urgency complaints, pad test results, and pelvic floor muscle strength in the treatment group compared with the control group at the 6-month evaluation (P < .05, P < .05, and P < .001, respectively).32 Kim et al17 reported that the cure rate for urine leakage was at 30.8% 1 year after the intervention, thus showing a significant improvement (P < .001).17 Kim et al12 also reported significant improvements at the 7-month follow-up evaluation for urine leakage (P < .001), and cure rates for all types of urinary incontinence (stress: P < .001) in the treatment group compared with the control group (2.7 patients needed to treat).


Summarizing current pelvic floor muscle training interventions and their impact on urinary incontinence can help to inform further research in an effort to find an optimal training solution. The studies included in this review comprised a variety of exercise interventions and were all able to elicit significant improvements for women older than 50 years in one or more of the following outcomes: severity of urinary incontinence, pelvic floor muscle strength, and quality of life. These statistically significant improvements also have meaningful clinical implications—all outcome measures improved by over 20%. These findings are supported by Cacciari et al,36 which also found that pelvic floor muscle training was effective in improving urinary incontinence compared with no treatment. Exercise parameters (eg, frequency, duration, and mode) varied greatly across interventions, with little impact to the magnitude of improvement. However, a significant determining factor was prevalence of supervision and feedback. Exercise interventions that were supervised or included some type of biofeedback device showed greater and faster improvements than interventions without additional assistance. Similarly, a review by Price et al37 found that studies showed greater self-reported improvements with urinary incontinence if pelvic floor muscle training was performed in a supervised program.

The frequency and duration of the interventions varied greatly in the studies included in this review; however, both exercise parameters seemed to have little decisive impact on the results. These results indicate that exercise interventions do not necessarily need to be of high frequency or long duration to elicit positive effects on urinary incontinence.

All exercise interventions showed significant improvement in their outcome measures; however, interventions that were supervised produced substantially greater improvements in urinary leakage (20%) and pelvic floor muscle strength (50%) compared with exercise interventions that were not supervised. Previous studies have shown that over 30% of women are unable to contract the pelvic floor muscle in isolation even after an initial consultation.38,39 Women commonly contracted hip adductors, abdominal and gluteal muscles instead of the pelvic floor muscle. These muscles do not act as structural support to the pelvic organs, and are subsequently unable to help prevent or manage urinary incontinence.40 However, contracting the extrinsic musculature does cause a co-contraction of the pelvic floor muscles in healthy women, which explains why improvements (although smaller in magnitude) were still seen even with unsupervised home-based exercise interventions that required isolated pelvic floor muscle contractions.40 The success of the co-contraction may not be as evident in older women, as a review by Fu et al23 showed that pelvic floor muscle exercises improved women's urine leakage but did not necessarily significantly improve pelvic floor muscle strength. The findings of both this systematic review and that of Fu et al23 further highlight the importance of supervised and detailed education of correct pelvic floor muscle activation particularly for older women. It is therefore recommended to provide detailed education and frequent appointments during a training program.23

Studies showed a more definitive impact on pelvic floor muscle strength when biofeedback devices29 or digital palpation31 was used to help participants contract pelvic floor muscles in isolation. Using a biofeedback device for a pelvic floor muscle exercise intervention does, however, have its drawbacks. Biofeedback devices use surface electromyography, which is unable to determine speed or strength of muscle contractions.41 Some studies have also reported high dropout rates, as a biofeedback device can have a poor device tolerance, in some cases even causing pain, vaginitis, and bleeding.29 Digital palpation was commonly performed on a weekly or monthly basis and might therefore see higher compliance rates than using a biofeedback device daily.31Although biofeedback devices resulted in greater improvements of urinary incontinence complaints compared to digital palpation, it is recommended that digital palpation is considered when prescribing pelvic floor exercise for urinary incontinence, as intolerance, injury, and dropout rates are lower in studies using digital palpation.

Unfortunately, less is known about the long-term effects of the exercise interventions, as only 3 studies included a long-term follow-up assessment. While all 3 studies showed significant improvements in their follow-up assessments, 2 of these studies encouraged participants to keep exercising during the follow-up period and to attend monthly group exercise sessions.12,17 Only 1 study included a follow-up period of 6 months, whereby participants followed the exercise program on their own without check-ups during that period. This particular study showed significant improvements in the follow-up assessment compared with baseline, which could indicate that a 6- to 8-week-long intervention might be sufficient to improve urinary incontinence complaints.32 Importantly, research by Cammu et al42 showed that 66% of successful initial pelvic floor muscle training continued to demonstrate favorable results even after a 10-year follow-up.42 As also suggested by Dumoulin and Hay-Smith,21 further studies are needed to examine the lasting effects of an intervention program on improved urinary incontinence particularly in older women.

The evidence presented in this review does align with previous review articles21,22,35,43 examining the effects of pelvic floor muscle exercise interventions and highlights the importance of exercise interventions for postmenopausal women to improve pelvic floor muscle functioning. Guidelines from major US organizations are in general agreement with other international groups (ie, Best Practice Advocacy Centre of New Zealand and UK's National Institute for Health and Care Excellence [NICE]), which recommend supervised pelvic floor muscle training for at least 3 months' duration as the first line of treatment for women with stress or urinary incontinence.44 Studies in this review however showed that shorter durations could also show improvement in the outcome measures. While the use of perineometry or pelvic floor electromyography as biofeedback devices for routine training is not recommended, digital palpation could be a viable option for maintaining the maximum benefit during pelvic floor muscle training. It is important to note that all of these recommendations were last updated in 2006. The European Association of Urology (EAU) guidelines on incontinence were last updated in 2009 and agree with the NICE guidelines. Interestingly, the EAU guidelines note that the addition of a biofeedback device for pelvic floor muscle training does not appear to be of benefit.45 However, the study by Ong et al29 has been published since that update and showed that pelvic floor muscle training with a biofeedback device produced greater and faster improvements. Further studies are therefore needed to examine the benefits and risks of the latest biofeedback devices in pelvic floor muscle training.

This systematic review is not without limitations. First, articles were scanned and identified by the lead author. Only where the lead author could not make a decision were 2 other authors consulted to determine whether articles met inclusion or exclusion criteria. Although the more appropriate methodology would have been to use 2 authors throughout the entire selection process, another systematic review using 1 author for the selection process has previously been published.46 Studies that were included in this review were classified as good or excellent research quality by the PEDro scale. However, the uniform adherence to the hierarchy of PEDro criteria resulted in overall similar scores. It is important to note that, while there was homogeneity in the research quality, there was significant heterogeneity in the exercise prescription parameters.


This review has shown that exercise interventions targeting the pelvic floor muscles may be effective in improving pelvic floor muscle strength, urinary incontinence symptoms, and even quality-of-life measures in women older than 50 years. Additionally, follow-up assessments indicated that exercise interventions could be effective for long-term improvements. Exercise interventions that were supervised, or used digital palpation, showed greater and faster improvements than interventions without additional assistance, particularly for pelvic floor muscle strength. Pelvic strengthening was effective in the broader context of exercise, but interventions that ensured correct pelvic floor muscle activation (via supervision, biofeedback, and digital palpation) resulted in greater benefits.


1. Abrams P, Cardozo L, Fall M, et al. The standardisation of terminology in lower urinary tract function: report from the standardisation sub-committee of the International Continence Society. Urology. 2003;61(1):37–49.
2. Norton P, Brubaker L. Urinary incontinence in women. Lancet North Am Ed. 2006;367(9504):57–67.
3. Chmielewska D, Stania M, Kucab-Klich K, et al. Electromyographic characteristics of pelvic floor muscles in women with stress urinary incontinence following sEMG-assisted biofeedback training and Pilates exercises. PLoS One. 2019;14(12):e0225647.
4. Ptak M, Ciećwież S, Brodowska A, et al. The effect of pelvic floor muscles exercise on quality of life in women with stress urinary incontinence and its relationship with vaginal deliveries: a randomized trial. Biomed Res Int. 2019;2019:5321864.
5. Riemsma R, Hagen S, Kirschner-Hermanns R, et al. Can incontinence be cured? A systematic review of cure rates. BMC Med. 2017;15(1):63.
6. Khandelwal C, Kistler C. Diagnosis of urinary incontinence. Am Fam Physician. 2013;87(8):543–550.
7. Dugan SA, Lavender MD, Hebert-Beirne J, Brubaker L. A pelvic floor fitness program for older women with urinary symptoms: a feasibility study. PM R. 2013;5(8):672–676.
8. Ferreira M, Santos PC, Duarte JA, Rodrigues R. Exercise programmes for women with stress urinary incontinence. Primary Health Care. 2012;22(3):24–27.
9. Donahoe-Fillmore B, Brahler CJ, Chorny W, Ingley A, Kennedy J, Osterfeld V. A comparison of two pelvic floor muscle training programs in females with stress incontinence: a pilot study. J Womens Health Phys Ther. 2011;35(1):29–29.
10. Abrams P, Andersson KE, Apostolidis A, et al. 6th International Consultation on Incontinence. Recommendations of the International Scientific Committee: evaluation and treatment of urinary incontinence, pelvic organ prolapse and faecal incontinence. Neurourol Urodyn. 2018;37(7):2271–2272.
11. Singh U, Agarwal P, Verma ML, Dalela D, Singh N, Shankhwar P. Prevalence and risk factors of urinary incontinence in Indian women: a hospital-based survey. Indian J Urol. 2013;29(1):31–36.
12. Kim H, Yoshida H, Suzuki T. The effects of multidimensional exercise treatment on community-dwelling elderly Japanese women with stress, urge, and mixed urinary incontinence: a randomized controlled trial. Int J Nurs Stud. 2011;48(10):1165–1172.
13. Singh N, Rashid M, Bayliss L, Graham P. Pelvic floor muscle training for female urinary incontinence: does it work? Arch Gynecol Obstet. 2016;293(6):1263–1269.
14. Bardsley A. Drug therapies for post-menopause urinary incontinence. Nurse Prescribing. 2012;10(8):384–391.
15. Tsakiris P, Oelke M, Michel MC. Drug-induced urinary incontinence. Drugs Aging. 2008;25(7):541–549.
16. Nygaard I, Turvey C, Burns TL, Crischilles E, Wallace R. Urinary incontinence and depression in middle-aged United States women. Obstet Gynecol. 2003;101(1):149–156.
17. Kim H, Suzuki T, Yoshida Y, Yoshida H. Effectiveness of multidimensional exercises for the treatment of stress urinary incontinence in elderly community-dwelling Japanese women: a randomized, controlled, crossover trial. J Am Geriatr Soc. 2007;55(12):1932–1939.
18. Gavira Iglesias FJ, Caridad y Ocerin JM, Perez del Molino Martin J, et al. Prevalence and psychosocial impact of urinary incontinence in older people of a Spanish rural population. J Gerontol A Biol Sci Med Sci. 2000;55(4):207–214.
19. Landis JR, Kaplan S, Swift S, Versi E. Efficacy of antimuscarinic therapy for overactive bladder with varying degrees of incontinence severity. J Urol. 2004;171(2):752–756.
20. Andersson KE, Appell R, Cardozo LD, et al. The pharmacological treatment of urinary incontinence. BJU Int. 1999;84(9):923–947.
21. Dumoulin C, Hay-Smith J. Pelvic floor muscle training versus no treatment, or inactive control treatments, for urinary incontinence in women. Cochrane Database Syst Rev. 2010;(1):CD005654.
22. Hay-Smith EJC, Herderschee R, Dumoulin C, Herbison GP. Comparisons of approaches to pelvic floor muscle training for urinary incontinence in women. Cochrane Database Syst Rev. 2011(12):CD009508.
23. Fu Y, Nelson EA, McGowan L. Multifaceted self-management interventions for older women with urinary incontinence: a systematic review and narrative synthesis. BMJ Open. 2019;9(8):e028626.
24. Hays RD, Woolley JM. The concept of clinically meaningful difference in health-related quality-of-life research—how meaningful is it?. Pharmacoeconomics. 2000;18(5):419–423.
25. Verhagen AP, de Vet HC, de Bie RA, et al. The Delphi list: a criteria list for quality assessment of randomized clinical trials for conducting systematic reviews developed by Delphi consensus. J Clin Epidemiol. 1998;51(12):1235–1241.
26. Elkins MR, Moseley AM, Sherrington C, Herbert RD, Maher CG. Growth in the Physiotherapy Evidence Database (PEDro) and use of the PEDro scale. Br J Sports Med. 2013;47(4):188–189.
27. Collado-Garrido L, Parás-Bravo P, Calvo-Martín P, Santibáñez-Margüello M. Impact of resistance therapy on motor function in children with cerebral palsy: a systematic review and meta-analysis. Int J Environ Res Public Health. 2019;16(22):4513.
28. Martinho NM, Silva VR, Marques J, Carvalho LC, Iunes DH, Botelho S. The effects of training by virtual reality or gym ball on pelvic floor muscle strength in postmenopausal women: a randomized controlled trial. Braz J Phys Ther. 2016;20(3):248–257.
29. Ong TA, Khong SY, Ng KL, et al. Using the Vibrance Kegel device with pelvic floor muscle exercise for stress urinary incontinence: a randomized controlled pilot study. Urology. 2015;86(3):487–491.
30. Tak EC, van Hespen A, van Dommelen P, Hopman-Rock M. Does improved functional performance help to reduce urinary incontinence in institutionalized older women? A multicenter randomized clinical trial. BMC Geriatr. 2012;12:51.
31. Tsai Y-C, Liu C-H. The effectiveness of pelvic floor exercises, digital vaginal palpation and interpersonal support on stress urinary incontinence: an experimental study. Int J Nurs Stud. 2009;46(9):1181–1186.
32. Aslan E, Komurcu N, Beji NK, Yalcin O. Bladder training and Kegel exercises for women with urinary complaints living in a rest home. Gerontology. 2008;54(4):224–231.
33. de Morton NA. The PEDro scale is a valid measure of the methodological quality of clinical trials: a demographic study. Aust J Physiother. 2009;55(2):129–133.
34. Lim R, Liong ML, Lim KK, Leong WS, Yuen KH. The minimum clinically important difference of the international consultation on incontinence questionnaires (ICIQ-UI SF and ICIQ-LUTSqol). Urology. 2019;133:91–95.
35. Yalcin I, Patrick DL, Summers K, Kinchen K, Bump RC. Minimal clinically important differences in Incontinence Quality-of-Life scores in stress urinary incontinence. Urology. 2006;67(6):1304–1308.
36. Cacciari LP, Dumoulin C, Hay-Smith EJ. Pelvic floor muscle training versus no treatment, or inactive control treatments, for urinary incontinence in women: a Cochrane systematic review abridged republication. Braz J Phys Ther. 2019;23(2):93–107.
37. Price N, Dawood R, Jackson SR. Pelvic floor exercise for urinary incontinence: a systematic literature review. Maturitas. 2010;67(4):309–315.
38. Bø K, Larsen S, Oseid S, Kvarstein B, Hagen R, Jørgensen J. Knowledge about and ability to correct pelvic floor muscle exercises in women with urinary stress incontinence. Neurourol Urodyn. 1988;7(3):261–262.
39. Hesse U, Schussler B, Frimberger J, Obernitz N, Senn E. Effectiveness of a three step pelvic floor reeducation in the treatment of stress urinary incontinence: a clinical assessment. Neurourol Urodyn. 1990;9(4):397–398.
40. Bø K. Urinary incontinence, pelvic floor dysfunction, exercise and sport. Sports Med. 2004;34(7):451–464.
41. Vigotsky AD, Halperin I, Lehman GJ, Trajano GS, Vieira TM. Interpreting signal amplitudes in surface electromyography studies in sport and rehabilitation sciences. Front Physiol. 2018;8:985.
42. Cammu H, Van Nylen M, Amy J. A 10-year follow-up after Kegel pelvic floor muscle exercises for genuine stress incontinence. BJU Int. 2000;85(6):655–658.
43. Ferreira M, Santos P. Pelvic floor muscle training programmes: a systematic review. Acta Med Port. 2011;24(2):309–318.
44. Viktrup L, Summers KH, Dennett SL. Clinical practice guidelines on the initial assessment and treatment of urinary incontinence in women: a US focused review. Int J Gynaecol Obstet. 2004;86(suppl 1):S25–S37.
45. Thüroff JW, Abrams P, Andersson KE, et al. EAU guidelines on urinary incontinence. Actas Urol Esp. 2011;35(7):373–388.
46. Barrow DR, Abbate LM, Paquette MR, et al. Exercise prescription for weight management in obese adults at risk for osteoarthritis: synthesis from a systematic review. BMC Musculoskelet Disord. 2019;20(1):610.

exercise; pelvic floor muscle; urinary incontinence

© 2021 Academy of Pelvic Health Physical Therapy, APTA.