Physiotherapy for Urinary Incontinence and Urogenital Prolapse

Kegel¹ was the first to report pelvic floor muscle training (PFMT) to be effective in the treatment of female urinary incontinence (UI) and pelvic organ prolapse (POP). According to his clinical experience and pre-post questions about the condition, he reported cure rates of over 84% for UI. These figures were not based on pad testing or more robust testing of the condition, randomized controlled trials with random allocation to either training or control or blinding of the investigator, factors that all may overestimate the results.

Today there is a huge body of evidence from randomized controlled trials and systematic reviews concluding that PFMT is effective in the treatment of stress (SUI) and mixed urinary incontinence (MUI) in women. Several systematic reviews have recommended conservative treatment and especially PFMT as first-line treatment for UI^{2, 3, 4, 5, 6, 7, 8} and for POP.⁸ However, many surgeons seem to regard minimally invasive surgery a better first-line option than PFMT. The skepticism against PFMT may be based on inappropriate knowledge of exercise science and physiotherapy, beliefs that there is still insufficient evidence for the effect of PFMT, that evidence for long-term effect is lacking or poor, and that women are not motivated to regularly perform PFMT.

To date, there are two main proven mechanisms on how PFMT may be effective in the prevention and treatment of SUI and POP:⁹

1.         Women learn to consciously contract before and during an increase in abdominal pressure ('the Knack'), and continue to perform such contractions as a behavioral modification to prevent descent of the pelvic floor.

2.         Women are taught to perform regular strength training over time to build up ‘stiffness’ and structural support of the pelvic floor.

There is basic research, case–control studies and randomized controlled trials to support both hypotheses.

EVIDENCE FOR CONSCIOUS PRE-CONTRACTION OF PELVIC FLOOR MUSCLES 'THE KNACK'

By intentional contraction of the PFM before and during an increase in intra-abdominal pressure there is a lift of the pelvic floor in a cranial and forward direction and a squeeze around the urethra, vagina and rectum.^{1, 10, 11, 12, 13} Ultrasonography and MRI studies have verified a lift in a cranial direction and movement of the coccyx in a forward, anterior and cranial direction.^{14, 15} Miller et al.¹⁶ named this voluntary pre-contraction the ‘Knack’, and in a single-blind randomized controlled trial, showed that the ‘Knack’ performed during a medium and deep cough reduced urinary leakage by 98.2 and 73.3%, respectively. Basic and functional anatomy research supports the ‘Knack’ as an effective maneuver to stabilize the pelvic floor.^{17, 18} However, to date there are no studies on how much strength is necessary to prevent descent during physical exertion, and we do not know whether regular counterbracing during daily activities will cause an automatic contraction of the PFM. A randomized controlled trial on women with POP found no effect of asking women to perform the 'Knack' on vaginal resting pressure or PFM strength and endurance.¹⁹ However, there are no studies prescribing the 'Knack' that have measured adherence to the prescription. Hence, we do not know to what degree women have followed the recommendation and actually changed their behavior during daily activities.

EVIDENCE FOR STRENGTH TRAINING

Kegel¹ described PFMT as physiological training or ‘tightening up’ the pelvic floor. The theoretical rationale for intensive strength training of the PFM to treat SUI is that strength training may build up the structural support for the internal pelvic organs by elevating the levator plate to a permanent higher location inside the pelvis and by enhancing hypertrophy and stiffness of the PFM and its connective tissue. This may increase the urethral resting pressure and the maximum urethral closure pressure, but also facilitate a more effective automatic motor unit firing (neural adaptation), preventing descent during increase in intra-abdominal pressure. The pelvic openings and the levator hiatus may narrow and the pelvic organs be held in place during increases in intra-abdominal pressure. In addition, a pelvic floor located at a higher level inside the pelvis may yield a much quicker and more coordinated response to an increase in intra-abdominal pressure, closing the urethra by increasing the urethral pressure.^{20, 21}

Ultrasound studies have shown that parous women have a more caudal location of the pelvic floor than nulliparous women.²² Difference in anatomical placement has also been shown between continent and incontinent women.^{17, 23}

None of the randomized controlled trials on strength training of the PFM to treat SUI in the general adult female population has evaluated the effect of PFMT on PFM tension or connective tissue stiffness, position of the muscles within the pelvic cavity, their cross-sectional area or neurophysiological function. However, in an uncontrolled trial of PFMT for SUI, Balmforth et al.²⁴ found that the position of the bladder neck was observed by ultrasound to be significantly elevated at rest, during Valsalva maneuver and squeeze after 14 weeks of supervised PFMT and behavioral modifications. In an assessor blinded randomized controlled trial of PFMT in women with POP, Brækken et al.¹⁹ found statistically significant changes in PFM strength, thickness, muscle length, levator hiatus area and position of the rectal ampulla and bladder neck in favor of the PFMT group compared to the control group. As the results were found in POP women there is a need for a similar study in women with SUI. However, since POP is considered a more complicated condition to treat, one may assume that even better results will be found in a group with SUI only.

In some studies the participants were tested both subjectively and objectively during physical activity, and showed no leakage during strenuous tests after the training period.^{25, 26, 27} Therefore, the effect, was most likely as a result of improved automatic muscle function and not only the ability to voluntarily contract before an increase in intra-abdominal pressure. Brækken et al.¹⁹ found that there was less widening in the levator hiatus area during Valsalva in the PFMT group compared to the control group, and suggested that this may be owing to increased automatic function of the PFM.

The rationale behind the use of PFMT to treat symptoms of overactive bladder (OAB) symptoms such as frequency, urgency and urgency urinary incontinence (UUI) is based on observations from electrical stimulation and urodynamic assessment during PFM contraction. Godec et al.²⁸ found that, during functional electrical stimulation (FES), hyperactivity of the bladder was diminished or completely abolished in 31 of 40 patients. One minute after stimulation cessation, the inhibition was still present.

Shafik and Shafik²⁹ investigated the effect of a voluntary PFM contraction on detrusor and urethral pressures in 28 women and men with OAB and 17 healthy volunteers. They found that during PFM contraction the urethral pressure significantly increased and the vesical pressure significantly decreased in both patients and healthy subjects. The change during PFM contraction was significantly larger in the healthy volunteers. The authors concluded that PFM contractions led to a decline of detrusor pressure, an increase of urethral pressures and suppression of the micturition reflex, and that their results encourage PFM contractions in the treatment of OAB.

Anecdotally, patients can successfully inhibit urgency, detrusor contraction and urinary leakage by walking, bending forwards, crossing their legs, using hip adductor muscles with or without conscious co-contraction of the PFM, or by conscious contraction of the PFM alone. After inhibition of the urgency to void and detrusor contraction, the patients may gain time to reach the toilet and thereby prevent leakage. There are two main hypotheses for the mechanism of PFMT to treat urgency incontinence:

●          Intentional contraction of the PFM during urgency, and holding of the contraction till the urge to void disappears;

●          Strength training of the PFM with long-lasting changes in muscle morphology may stabilize neurogenic activity.

None of the studies in this field (neither uncontrolled studies nor randomized controlled trials) have evaluated whether there are changes in the inhibitory mechanisms after PFMT. To date there is no consensus on the optimal exercise protocol to prevent or treat OAB.³⁰ The theoretical basis of how PFMT may work in the treatment of OAB therefore remains unclear.^{31, 32}

STRESS URINARY INCONTINENCE

Updated and comprehensive systematic reviews on PFMT in the treatment of SUI with detailed tables can be found in the Cochrane library^{4, 6, 7, 33} and the ICI consensus book.⁸

It is difficult to make meaningful comparisons between studies and groups of studies in this area because there is a great heterogeneity between studies. This heterogeneity involves inclusion criteria of the studies (several studies include women with SUI, UUI and MUI), use of different outcome measures, and different exercise regimens with a huge variety of training regimens. In addition, many researchers have used combined interventions (e.g. electrical stimulation and strength training, bladder training and strength training).

One important flaw in PFMT studies is a lack of assessment of the ability to contract the PFM correctly. Several research groups have shown that over 30% of women are unable to voluntarily contract the PFM at their first consultation even after thorough individual instruction.^{34, 35, 36, 37} Hay-Smith et al.³ reported that ability to contract PFM was checked before training in only 15 of 43 randomized controlled trials on the effect of PFMT for SUI, UUI and MUI. Common mistakes are to contract other muscles such as abdominals, gluteals and hip adductor muscles instead of the PFM.^{35, 38} In addition, Bump et al.³⁶ showed that as many as 25% of women may strain instead of squeeze and lift. If women are straining instead of performing a correct contraction, the training may harm and not improve PFM function. Proper assessment of the ability to contract the PFM is therefore mandatory.

Several randomized controlled trials have demonstrated that PFM exercise is more effective than no treatment.^{16, 26, 39, 40, 41, 42} In addition, a number of randomized controlled trials have compared PFMT alone with the use of vaginal resistance devices, biofeedback or vaginal cones.^{6, 8, 33} Interestingly, Ramsey and Thou⁴³ is one of very few research groups that did not find any significant effect of PFMT. In this study there was no check of the women’s ability to contract, adherence to the training protocol was poor, and the placebo group contracted gluteal muscles and external rotators of the hips; activities that may give co-contractions of the PFM.^{38, 44}

As for surgery⁴⁵ and pharmacology studies,⁴⁶ a combination of cure and improvement measures is often reported. To date there is no consensus on what outcome measure to choose as the gold standard for cure.^{47, 48} Subjective cure and improvement rates of PFMT reported in randomized controlled trials in studies including groups with SUI and MUI vary between 56 and 70%.⁸ Short-term cure rates of 44–70%, defined as ≤2 g of leakage on different pad tests, have been found after PFMT.^{26, 27, 40, 42, 49, 50} The highest cure rates were shown in two single-blind randomized controlled trials of high methodological quality. The participants had thorough individual instruction by a trained physiotherapist, combined training with biofeedback or electrical stimulation, and close follow-up once or every second week. Adherence was high, and drop-out was low.^{27, 49} Because biofeedback and electrical stimulation have not shown any additional effect to PFMT in randomized controlled trials and systematic reviews (biofeedback randomized controlled trials being flawed because of difference in training regimens in favor of the biofeedback intervention),^{6, 8} one could hypothesize that the key factors for success are most likely close follow-up and more intensive training.

What is the best pelvic floor muscle training regimen?

Because of use of different outcome measures and instruments to measure PFM function and strength, it is impossible to combine results between studies, and it is difficult to conclude which training regimen is the more effective. Also the exercise regimen (type of exercise, frequency, duration and intensity) varies significantly between studies.⁸

Bø et al.²⁵ have shown that instructor-followed up training is significantly more effective than home exercise. In this study individual assessment and teaching of correct contraction was combined with strength training in groups in a 6-month training program. The women were randomized to either an intensive training program consisting of seven individual sessions with a physiotherapist, combined with 45 minutes weekly PFMT classes, and three sets of 8–12 contractions per day at home or the same program without the weekly intensive exercise classes. The results showed a much better improvement in both muscle strength and leakage in the intensive exercise group: 60% reported to be continent/almost continent in the intensive exercise group compared to 17% in the less intensive group. A significant reduction of urinary leakage, measured by pad test with standardized bladder volume, was only demonstrated in the intensive exercise group.

This study demonstrated that very little effect can be expected after training without close follow-up. It is worth noting that the significantly less effective group in this study had seven visits with a skilled physiotherapist and that adherence to the home training program was high. More intensive training has also been shown to be more effective in other studies and this is now the conclusion of several systematic reviews.^{5, 7, 8, 51} There is a dose–response issue in all training regimens.^{52, 53} Therefore, one reason for disappointing effects shown in some clinical practices or studies may be as a result of insufficient training stimulus and low dosage. Dosage in exercise training includes frequency (number of exercises sessions per week), duration (number of weeks or months of exercising) and intensity (percentage of maximum e.g. percentage of one repetition maximum in strength training) of the exercises. If low dosage programs are chosen as one arm in a randomized controlled trial comparing PFMT with other methods, PFMT is bound to be less effective.⁵⁴

Pelvic floor muscle training with biofeedback

Biofeedback has been defined as ‘a group of experimental procedures where an external sensor is used to give an indication on bodily processes, usually in the purpose of changing the measured quality’.⁵⁵ Kegel¹ based his training protocol on thorough instruction of correct contraction using vaginal palpation and clinical observation. He combined PFMT with the use of vaginal squeeze pressure measurement as biofeedback during exercise. Today, a variety of biofeedback apparatus is commonly used in clinical practice to assist with PFMT.

Biofeedback is not a treatment on its own. It is an adjunct to training, measuring the response from a single PFM contraction. In the area of PFMT both vaginal and anal surface EMG, and urethral and vaginal squeeze pressure measurements have been used to make patients more aware of muscle function, and to enhance and motivate patients’ effort during training.^{3, 6} However, one should be aware that erroneous attempts at PFM contractions (e.g. by straining) may be registered by manometers and dynamometers, and contractions of other muscles than the PFM may affect surface EMG activity. Therefore, biofeedback cannot be used to register a correct contraction.

Since Kegel first presented his results, several randomized controlled trials have shown that PFMT without biofeedback is more effective than no treatment for SUI.⁴ In a Cochrane review Herderschee et al.⁶ found 24 randomized controlled trials or quasi randomized trials comparing PFMT with and without biofeedback. They concluded that the use of biofeedback may provide benefit in addition to PFMT. However, in women with SUI or MUI, all but two randomized controlled trials have failed to show any additional effect of adding biofeedback to the training protocol for SUI.^{27, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67} Berghmans et al.⁵⁸ demonstrated quicker progress in the biofeedback group. In the study of Glavind et al.⁶¹ a positive effect was demonstrated. However, this study was confounded by a difference in training frequency, and the effect might be owing to a double training intensity, the use of biofeedback, or both. The results support the studies concluding that there is a dose–response issue in PFMT.

Very few of the studies comparing PFMT with and without biofeedback have used the exact same training regimen in the two groups.⁶ For example Pages et al.⁶³ compared 60 minutes of group training five days a week with 15 minutes of individual biofeedback training 5 days a week, and found that the individualized biofeedback training protocol was more effective assessed by the women’s report and measurement of PFM strength. When the two groups under comparison receive different  amounts  of training in addition to biofeedback, it is impossible to conclude what is causing a possible effect. In addition, other factors flaw the results of studies comparing PFMT with and without biofeedback. As PFMT is effective without biofeedback, a large sample size may be needed to show any beneficial effect of adding biofeedback to an effective training protocol. In most published studies comparing PFMT with PFMT combined with biofeedback, the sample sizes are small, and type II error may have been the reason for negative findings.⁶ However, in the two largest randomized controlled trials published, no additional effect was demonstrated from adding biofeedback.

Many women may not like to undress, go to a private room and insert a vaginal or rectal device to exercise.⁶⁸ On the other hand, some women find it motivating to use biofeedback to control and enhance the strength of the contractions when training. Therefore, when available, biofeedback should be given as an option for home training, and the physiotherapist should use any sensitive, reliable and valid tool to measure PFMT strength development at office follow-up.

Pelvic floor muscle training with vaginal weighted cones

Vaginal cones are weights that are put into the vagina above the levator plate.³³ The cones were developed by Plevnik in 1985.³ The theory behind their use is that the PFM are contracted reflexively or voluntarily when the cone is perceived to slip out. The weight of the cone is supposed to give a training stimulus and make women contract harder with progressive weight. In a Cochrane review, combining studies including 17 randomized controlled trial or quasi randomized studies with 1484 women with both SUI and mixed incontinence (six trials published as abstracts only) it was concluded that training with vaginal cones is more effective than no treatment.³³

Several randomized controlled trials have compared PFMT with and without vaginal cones for SUI.^{26, 62, 69, 70, 71} Bø et al.²⁶ found that PFMT was significantly more effective than training with cones both to improve muscle strength and reduce urinary leakage. In three other studies there were no differences between PFMT with and without cones.^{62, 70, 71} Cammu and van Nylen⁷⁰ reported very low compliance and therefore did not recommend use of cones. Also in the study of Bø et al.,²⁶ women in the cone group had great motivational problems. Laycock et al.⁶² had a total drop-out rate in their study of 33%.

The use of cones can be questioned from an exercise science perspective. Holding the cone for the recommended time of 15–20 minutes may result in decreased blood supply, decreased oxygen consumption, muscle fatigue and pain, and recruit contraction of other muscles instead of the PFM. Arvonen et al.⁶⁹ used ‘vaginal balls’ and followed general strength training principles. They found that training with the balls was significantly more effective in reducing urinary leakage than regular PFMT.

Pelvic floor muscle training or electrical stimulation for SUI?

Moore et al.⁸ concluded that PFMT is better than electrical stimulation as first-line conservative therapy for SUI, especially if PFMT is supervised. Hennalla et al.,³⁹ Hofbauer et al.⁴¹ and Bø et al.²⁶ found that PFMT was significantly better than electrical stimulation to treat SUI. Laycock and Jerwood⁷² and Hahn et al.⁷³ found no difference, and Smith⁷⁴ found that electrical stimulation was significantly better. Bidmead et al.,⁷⁵ Goode et al.,⁷⁶ Hofbauer et al.⁴¹ and Knight et al.⁷⁷ found no effect of adding electrical stimulation to PFMT.

Many of the electrical stimulation studies are flawed with small numbers, and future  randomized controlled trials with better methodological quality should be repeated.⁸ However, electrical stimulation has shown side-effects⁷⁸ and is less well tolerated by women than PFMT.²⁶ In addition, Bø and Talseth⁷⁹ found that voluntary PFM contraction increases urethral pressure significantly more than electrical stimulation, and several consensus statements have concluded that strength training is more effective than electrical stimulation in humans.^{80, 81}

Is surgery more effective than pelvic floor muscle training for SUI?

Only two published studies have compared surgery with PFMT as first-line treatment for SUI. In the randomized controlled trial of Klarskov et al.⁸² the patients had different surgeries according to their problems. The PFMT program was described as group training with five or more sessions with a physiotherapist, and it is not clear whether the participants had vaginal palpation to make sure they were able to contract the PFM correctly. At 4 months the PFMT group was less likely to report cure compared to women who had surgery. However, there was no statistical difference in the proportions reporting cure/improvement. At 12 months 10/24 women in the PFMT group reported satisfaction with the initial therapy versus 19/26 in the surgery group. Adverse effects were reported only in the surgery group, including new UUI, retropubic or pelvic pain or dyspareunia. In a recently published 12 months follow-up of a randomized controlled trial comparing PFMT with mid-urethral sling (230 patients in each group) subjective improvement rate was 64.4% versus 90.8%, subjective cure rate 58.8% versus 85.2% and objective cure rate (cough stress test with a bladder volume of at least 300 ml) 58.8% versus 76.5% in the PFMT and surgery groups, respectively.⁸³ However, the study may be flawed by a selection bias as the women had to agree to surgery to volunteer, and the usual complications after surgery were reported.

Adverse effects of pelvic floor muscle training

Few, if any, adverse effects have been found after PFMT.⁸ Lagro-Jansson et al.⁸⁴ found that one woman reported pain with exercise and three had an uncomfortable feeling during the exercises. Aukee et al.⁵⁷ reported no side-effects in the training group, but found that two women interrupted the use of home biofeedback apparatus because they found the vaginal probe uncomfortable. These women were both postmenopausal. In other studies no side-effects have been found.²⁶

Long-term effect of pelvic floor muscle training for SUI

Several studies have reported long-term effect of PFMT.^{8, 85} However, usually women in the non-treatment or less effective intervention groups have gone on to further treatment after cessation of the study period. Therefore, follow-up data are usually reported for either all women or for only the group with the best effect. In the systematic review by Bø and Hilde⁸⁵ 19 studies were included (1141 women followed between 1 and 15 years). Statistical meta-analysis was not performed owing to high heterogeneity. Only two of the long-term studies provided follow-up interventions. Losses to follow-up during the long-term period ranged between 0 and 39%. Long-term adherence to PFMT varied between 10% and 70%. The initial success rate for SUI and MUI was maintained in the long-term in five studies. Long-term success based on being a responder to the original trial varied between 41% and 85%. Surgery rates in the long term varied between 4.9% and 58%. It was concluded that short-term outcome of PFMT can be maintained at long-term follow-up without incentives for continued training. However, the high heterogeneity in both interventional and methodological quality in short-and long-term pelvic floor muscle training studies is a challenge when analysing long-term effect.

As for surgery,^{45, 86} there are only a few long-term studies including clinical examination.⁸⁵ Klarskov et al.⁸⁷ assessed only some of the women originally participating in the study. Lagro-Janssen et al.⁸⁸ evaluated 88 of 110 women with SUI and UUI or MUI 5 years after cessation of training and found that 67% remained satisfied with the condition. Only seven of 110 had been treated with surgery. Moreover, satisfaction was closely related to compliance to training and type of incontinence, with women with mixed incontinence being more likely to lose the effect. Women with SUI had the best long-term effect, but only 39% of them were exercising daily or ‘when needed’.

In a 5-year follow-up, Bø and Talseth⁸⁹ examined only the intensive exercise group and found that urinary leakage was significantly increased after cessation of organized training. Three of 23 had been treated with surgery. Of the women, 56% had a positive closure pressure during cough and 70% had no visible leakage during cough at 5-year follow-up; 70% of the patients were still satisfied with the results and did not want other treatment options.

Cammu et al.⁹⁰ used a postal questionnaire and medical files to evaluate the long-term effect on 52 women who had participated in an individual course of PFMT for urodynamic SUI. Eighty-seven per cent were suitable for analysis – 33% had had surgery after 10 years. However, only 8% had undergone surgery in the group that had originally had success after training, whereas 62% had undergone surgery in the group initially dissatisfied with training. Successful results were maintained after 10 years in two-thirds of the patients originally classified as successful.

Bø et al.⁹¹ reported current status of lower urinary tract symptoms from questionnaire data 15 years after cessation of organized training. They found that the short-term significant effect of intensive training was no longer present: 50% from both groups had interval surgery for SUI, however, more women in the less intensive training group had surgery within the first 5 years after ending the training program. There were no differences in reported frequency or amount of leakage between women who had or had not had surgery, and women who had surgery reported significantly more severe leakage and to be more bothered by urinary incontinence during daily activities than those who had not.

The general recommendations for maintaining muscle strength are one session per week of moderate to hard intensity exercises.⁹² The intensity of the contraction seems to be more important than frequency of training. So far, no studies have evaluated how many contractions subjects have to perform to maintain PFM strength after cessation of organized training. In the study by Bø and Talseth⁸⁹ PFM strength was maintained 5 years after cessation of organized training with 70% exercising more than once a week. However, number and intensity of exercises varied considerably between successful women.⁹³

Which women benefit from pelvic floor muscle training for SUI?

Several researchers have looked into factors affecting outcome of PFMT on UI.⁸ No single factor has been shown to predict outcome, and it has been concluded that many factors traditionally supposed to affect outcomes such as age and severity of incontinence may be less crucial than previously thought. Factors that appear to be most associated with a positive outcome are thorough teaching of correct contraction, motivation, adherence with the intervention, and intensity of the program.

Some women may find the exercises hard to conduct on a regular basis.⁹⁴ However, when analysing results of randomized controlled trials, adherence to the exercise program is generally high and drop-out rate is low.⁸ In a few studies low adherence and high drop-out rates have been reported.^{43, 62} Knowledge about behavioral sciences such as pedagogy and health psychology, and ability to explain and motivate patients may be a crucial factor to enhance adherence and minimize drop-outs from training. In some studies such strategies have been followed, and high adherence has been achieved.^{94, 95}

In a study of Alewijnse⁹⁴ most women followed advice of training 4–6 times a week 1 year after cessation of the training program. The following factors predicted adherence with 50%: positive intention to adhere, high short-term adherence levels, positive self-efficacy expectations and frequent weekly episodes of leakage before and after initial therapy.

Patients did not comply with treatment for a wide variety of reasons: long-lasting and time-consuming treatments, requirement of lifestyle changes, poor client/patient interaction, cultural and health beliefs, poor social support, inconvenience, lack of time, motivational problems and travel time to clinics have been listed.⁹⁶

Sugaya et al.⁹⁷ used a computerized pocket-size device giving a sound three times a day to remind the person to perform PFMT. To stop the sound the person needed to push a button, and by pushing the button for each contraction, adherence was registered: 46 women were randomly assigned to either instruction to contract the PFM following a pamphlet or with the same pamphlet together with the sound device and instruction on how to use the device. The results showed a significant improvement in daily incontinence episodes and pad test only in the device group: 48% were satisfied in the device group compared to 15% in the control group. It was reported that patients in the device group felt obliged to perform PFMT when the chime sounded.

Are there alternative exercise regimens to pelvic floor muscle training for UI?

In a systematic review on transverse abdominal training to treat female SUI, Bø et al.⁹⁸ concluded that there was no evidence to recommend training of the abdominals instead or in addition to PFMT. On the contrary contraction of the transverse abdominal muscles has been shown to open the levator hiatus area and move the pelvic floor downwards.⁹⁹ In a systematic review Bø and Herbert¹⁰⁰ analysed the effect of alternative exercise regimens for the PFM and found three randomized controlled trials for abdominal training, two randomized controlled trials for Pilates and two randomized controlled trials for the Paula method. None of the alternative exercises proved to be better than or yielded additional effect to PFMT. No randomized controlled trials were found for the effect of yoga, Tai Chi, balance, posture or respiration exercises to prevent or treat SUI in women.

PELVIC ORGAN PROLAPSE

No randomized controlled trials or studies using other designs have been found to evaluate the effect of PFMT on POP in primary prevention, i.e. to stop prolapse from developing.⁸ Search on PubMed revealed eight randomized controlled trials assessing PFMT to treat POP or POP symptoms. One study compared PFMT with no treatment,¹⁰¹ while Piya-Anant¹⁰² compared a combination of PFMT + advice to drink water and eat vegetables to reduce constipation and straining at stool, and compared this with no treatment. In most randomized controlled trials PFMT has been conducted in combination with lifestyle intervention and compared against lifestyle interventions alone. Lifestyle interventions have included use of pre-contraction of the PFM before and during increase in intra-abdominal pressure; 'the Knack' and advice to avoid pushing down during defecation^{19, 103} or general lifestyle advice.^{104, 105, 106}  None have compared the effect of these lifestyle interventions with untreated controls, and there is no report of adherence to these protocols. Hence, the effect of lifestyle interventions on POP is still unknown. Brækken et al.¹⁹ did not find any effect of advice to use the Knack on PFM strength, thickness, levator hiatus area or position of the bladder or rectal ampulla.

The randomized controlled trials are all in favor of the efficacy of PFMT to be effective in treating POP, demonstrating statistically significant improvement in symptoms^{101, 103, 104, 105, 106, 107, 108} and/or prolapse stage.^{102, 103, 104, 107, 108} Frawley et al. ¹⁰⁶ did not find a significant change in stage of POP, but significant changes were found in some of the individual POP quantification (POPQ) measurements.

Is there any difference in outcome based on the intervention?

All studies used PFMT and all research groups included vaginal palpation to assess the ability to perform a correct PFM contractions. The training period varied between 14 weeks and 2 years, and number of visits with the physiotherapist varied between four and 18 times. PFMT was taught individually in all trials and was combined with a home training program. Drop outs were low and adherence high in all studies. The highest number of visits with the physiotherapist was in the study of Brækken et al.^{19, 103} which included 18 visits over 6 months. This study had high adherence and only two drop outs, and showed the best overall results in change in POP stage and symptom reduction. interestingly a randomized controlled trial from India compared PFMT with a self-instruction manual, training the patients for 6 months as well, and showed the same overall beneficial results in symptoms and POPQ stage.¹⁰⁸ Owing to use of different outcome measures to assess POP symptoms it is difficult to compare results based on dose–response issues. No studies compared different training regimens.

PFMT as adjunct to prolapse surgery

Jarvis et al.¹⁰⁹ studied the effect of PFMT and bladder/bowel training on women undergoing surgery for POP/UI, with an randomized controlled trial of 60 women. The number of women undergoing POP-only surgery was not specified. Thirty women were randomized to each of the treatment and control groups. The intervention consisted of PFMT, contraction of the PFM prior to rises in intra-abdominal pressure, bladder/bowel training and advice to reduce straining during voiding and defecation. Significant improvements in bladder quality of life and urinary symptom specific scores were found in the treatment group. Subjects in the treatment group also demonstrated an increase in digital palpation score and maximum vaginal squeeze pressure compared with subjects in the control group, who showed a decrease in squeeze pressure.

In an assessor-blinded randomized controlled trial comparing the effect of POP surgery with and without a structured physiotherapy program, Frawley et al.¹¹⁰ did not find a significant effect of PFMT on UI at 1 year follow-up after surgery. The physiotherapy intervention comprised a PFM strength training protocol, supplemented by bladder and bowel advice. This was provided over eight sessions: one preoperative and seven postoperative sessions; day 3 postoperatively, week 6, 7, 8, 10 and 12, and a final appointment at 9 months postoperatively. Hence, none of the published randomized controlled trials on PFMT in conjunction with POP surgery have evaluated the effect on POP stage or symptoms and future studies are warranted in this area.

OVERACTIVE BLADDER

In clinical practice, many patients with OAB symptoms are treated with PFMT with and without biofeedback, electrical stimulation, bladder training, or medication, and often many of the interventions are combined. When different methods are combined it is not possible to analyse which of the interventions  may cause the effect. In most systematic reviews on efficacy of PFMT to prevent and treat UUI, studies including patients with symptoms or urodynamic diagnosis of SUI, UUI and MUI are combined. This makes it impossible to understand the real effect of the different interventions on each condition.

Although there are new theories suggesting PFM dysfunction as a common cause of the two main diagnoses (SUI and UUI),^{111, 112} the pathophysiology of UUI is not yet thoroughly understood, and may be very different from patient to patient. Optimally, the physiotherapy intervention should relate to the underlying pathophysiological condition. PFMT may have different cure and improvement rates for SUI and UUI, and the combination of heterogeneous patient groups in systematic reviews and meta-analyses may disseminate the real cure rate for each of the diagnoses. In addition, and most importantly; an optimal PFMT protocol may be different for the two conditions owing to a different theoretical rationale.

There are several systematic reviews including PFMT to treat OAB.^{4, 6, 7, 8, 30, 31, 32, 113, 114} Four randomized controlled trials using PFMT alone to treat symptoms of OAB were found.^{32, 115, 116, 117} The studies had moderate to high methodological quality.

Nygaard et al.¹¹⁵ found significant improvement in many variables in the subgroup of women with detrusor instability. There was no differences in outcome between the two randomized groups, and no comparison with non-treated controls. Berghmans et al.³² did not demonstrate any significant effect of their exercise protocol compared to an untreated control group. Wang et al.¹¹⁷ found that the significant subjective improvement/cure rate of OAB was the same between the electrical stimulation group and in the biofeedback-assisted PFMT group, but lower in the PFMT home training group. Millard¹¹⁶ did not show any additional benefit for a simple PFMT protocol (two-page written instruction, no assessment of ability to contract, and no follow-up or supervised training). The effect of PFMT on OAB is therefore questionable. Berghmans et al.³² and Millard¹¹⁶ included intentional contraction of the PFM to inhibit detrusor contractions in addition to a strength training program. However, we have no information about how many conducted the exercises in Millard’s study, and Berghmans et al.³² also included bladder training in their protocol. The protocol from Berghmans et al. did not show any effect when compared with untreated controls, but if there had been an effect it would not be possible to tell whether this was as a result of the exercises or the bladder training. In Millard’s study¹¹⁶ a very weak exercise protocol was conducted. There was no control of ability to contract the PFM, patients were left alone to exercise, and there was no report on adherence to the exercise protocol. The exercise period varied between 9 and 12 weeks in duration in the four randomized controlled trials in this area. This may be too short to treat a complex condition such as OAB. Both the NICE and ICI guidelines^{51, 118} have concluded that PFMT is recommended for UUI. However, as have been cited above, there is sparse knowledge from  randomized controlled trials evaluating the effect of PFMT on UUI alone, and the recommendations are based on studies with MUI.

Randomized controlled trials with high methodological quality, systematic reviews and several Cochrane reviews have concluded that there is level 1, grade A evidence that PFMT is more effective than no treatment, sham or placebo treatment for SUI. PFMT is recommended as first-line treatment for SUI. There is no evidence to suggest that adding the use of biofeedback, electrical stimulation or vaginal cones brings any additional effect over PFMT alone.

There is now level 1, grade A evidence that PFMT is effective in treatment of POP. There are no studies in primary prevention or lifestyle intervention. However, morphological changes have been found after PFMT which may indicate a possible preventive effect. There is a need for further randomized controlled trials to investigate the effect of PFMT in combination with POP surgery on POP symptoms and stage of prolapse. There are no long-term studies on the effectiveness of PFMT for POP, but such studies would be difficult to perform as women are frequently offered other treatments, e.g. surgery, after cessation of the PFMT intervention. One can assume that PFMT must be continued to maintain short-term results.

There are few randomized controlled trials in the area of PFMT and OAB and the results are difficult to interpret. In general the studies have moderate to high methodological quality, but the exercise protocols may not have been optimal. Because the pathophysiological background for OAB is not clear, it is difficult to plan an optimal training protocol. There is a need for more basic research to understand the role of a voluntary PFM contraction in inhibition of the micturition reflex and future  randomized controlled trials with high interventional and methodological quality are recommended.