Pelvic floor muscle (PFM) training, the most commonly used physical therapy treatment for women suffering from stress urinary incontinence (SUI), can improve their symptoms (1). Up to date PFM training is described as based on voluntary PFM contractions (1) even though fast involuntary reflexive PFM contractions seem crucial during SUI provoking situations. Training procedures for involuntary reflexive muscle contractions are widely implemented in rehabilitation and sports but not yet in PFM rehabilitation. Therefore, the research group developed two PFM training protocols, one including standard physiotherapy (PT) and one additionally focusing on involuntary reflexive PFM contractions. Both protocols involved an intervention of 16 weeks including nine personal PT consultations and 78 short home training sessions (weeks 1–5: 3x/week & 3x/day; weeks 6–16: 3x/week & 1x/day). The study aim was to compare the two PT protocols regarding their effect on SUI.

The present study was designed as a prospective, triple-blinded (participant, investigator, statistician) randomized controlled trial (intention to treat) with two PT intervention groups: CON = control group (standard PT), EXP = experimental group (standard PT + involuntary reflexive PFM training). The primary outcome was the International Consultation on Incontinence Modular Questionnaire Urinary Incontinence short form (ICIQ-UIsf). Secondary and tertiary outcomes were the International Consultation on Incontinence Modular Questionnaire Urinary Incontinence Symptoms Quality of Life (ICIQ-LUTSqol), a modified 20-min pad-test, PFM strength digitally assessed according to the Oxford Grading Scale and adherence to the home exercise protocol.

Sample size was estimated theoretically based on the primary outcome ICIQ-UIsf and the following assumptions: Effect size = 0.1, α = 0.05, power (1–β error probability) = 0.8, number of groups = 2, number of measurements = 10; correlations among repeated measures were estimated conservatively low with 0.5. Based on these assumptions, a total sample size of N = 80 was estimated. In anticipation of dropouts (+10 %: N = 8) or a violation of normality assumption (+10 %: N = 8), a final sample size of N= 96 (48 participants per group) was chosen.

This study adhered strictly to the CONSORT flow diagram. 

To analyze within and between group differences (CON vs. EXP) and the development over time (pre-intervention, PT1, PT2, … post-intervention) concerning the primary outcome, mixed effect regression models were used, which allow to account for within-participant correlations by random effects. The significance level was P≤0.05.

One hundred fourteen participants were assessed, and 18 participants were excluded based on exclusion criteria. Therefore, 96 participants could be assigned to randomization and group allocation (EXP: N = 48; CON: N = 48). Four participants were lost after randomization. Subsequently, baseline assessment of 45 (EXP) and 47 (CON) participants was completed.
Eight participants had incomplete primary outcome ICIQ-UIsf data. The missing data were replaced by the last valid data based on the intention-to-treat and last observation carried forward approach.

There were no significant differences in baseline characteristics between both groups in terms of body measurements, number of births and adherence to therapy schedule.

The analysis of intervention effects revealed that the total score of the primary outcome decreased significantly over time (pre/post) by about 3 points for both groups (EXP: 10.3/7.4; CON: 10.0/7.0), however, did not differ between groups.

The overall significance of group differences was tested by using Type III tests of fixed effects. Since it was not clear whether group differences vary over time, both the main effect of the group and the interaction between the group and time was tested. Overall, there were no statistically significant differences between EXP and CON at any time.

Except for EXP LUTSqol Part A, all pre/post comparisons of muscle strength test, 20-min pad test, and LUTSqol Part B showed a significant improvement in both groups, however no difference between the groups.

This study showed clinically relevant improvements in SUI in the CON as well as in the EXP group (2). However, participants of both groups were not completely cured from SUI after the intervention. These results are comparable to pre/post differences in former physiotherapy intervention studies (1). Therefore, the present as well as former findings show that PFM training generally can improve but rarely perfectly restore SUI. A possible reason could be that relevant training methods (hypertrophy training, intramuscular coordination, power and power endurance training) are widely disregarded in PFM training protocols or cannot be applied on the PFM as easily as on other skeletal muscles. For the current study a standardized therapy plan based on training methods used for e.g. leg muscle training (3) was developed for both groups (CON and EXP) with a specific focus on involuntary reflexive PFM training for the EXP group as this seems to be a main issue of SUI and, so far, was not included in common PFM training (1). 
The following aspects could have impaired the outcome of the present study: Compared to a common phase duration of skeletal muscle training methods (3), the training methods were applied during rather short time phases because of feasibility reasons regarding a pragmatic clinical trial approach (e.g. 9 PT sessions, totally 16 weeks for all training method phases together) and recruitment and participation reasons (e.g. reasonable therapy and study time span for participants); furthermore, therapy method phases were terminated without assessing obtainment of the related therapy methods’ effects (e.g. specific sensorimotor abilities, muscle mass or strength, rate of force development, fatigue resistance). Due to scientific requirements of a strict standardization of a training protocol, the used standardization - as necessary for an RCT - did prohibit adaptation of training parameters to individual factors and progress status. And, compared to common skeletal muscle training (3), the applied training methods might not have been high enough regarding intensity, exhaustion and strain due to feasibility reasons (e.g. no external weights). This seems to concern especially hypertrophy training, intramuscular coordination, power and power endurance training.

This RCT showed a clinically relevant effect on SUI of two PT protocols - one including standard PT and one additionally focusing on involuntary reflexive PFM contractions. However, there was no difference regarding this effect between those two PT protocols. Future studies should, on the one hand, use criteria oriented and not time-oriented PFM training protocols, i.e. individualized PFM training protocols and progress. Therefore, criteria for goal obtainment of each PFM training method phase would have to be developed and defined. 
And, on the other hand, training methods for PFM hypertrophy, intramuscular coordination, power and power endurance training, which are comparable to “common” skeletal muscle training i.e. performed with higher intensities and workout, should be tested.

Dumoulin C, Cacciari LP, Hay-Smith EJC. Pelvic floor muscle training versus no treatment, or inactive control treatments, for urinary incontinence in women (Review). Cochrane Database of Systematic Reviews. 2018(10). Art. No.: CD005654
Nyström E, Sjöström M, Stenlund H, Samuelsson E. ICIQ symptom and quality of life instruments measure clinically relevant improvements in women with stress urinary incontinence. Neurourol Urodyn. 2015;34(8):747-51
American College of Sports Medicine. American College of Sports Medicine position stand. Progression models in resistance training for healthy adults. Med Sci Sports Exerc. 2009;41(3):687–708