Although intravaginal dynamometry (IVD) is recommended as the best approach to assess levator ani muscle (LAM) function in research settings, it may not be accessible nor acceptable in some populations, including older females, children/adolescents, or transgender men. Transperineal ultrasound imaging (USI) may be a more acceptable alternative for the assessment of LAM function. Indeed, transient changes in pelvic morphology, including shortening of the levator plate and bladder neck excursion are strongly correlated with LAM electromyography when measured in supine [1]. However, we don’t know whether it is valid to use transient changes in pelvic morphology measured using USI as an indirect measurement of LAM force-generating capacity. Further, the relationship between transient changes in pelvic morphology and LAM force-generating capacity may differ in supine vs upright positions. 

The aims of this study were to 1) evaluate whether body position (supine vs. standing) impacts measures of pelvic floor morphology measured using ultrasound imaging (USI) and/or LAM force measured using IVD and 2) to investigate the relationship between changes in levator plate length and bladder neck height seen on dynamic transperineal USI and maximal vaginal closure force measured using IVD, in supine and standing positions.

This was a secondary analysis of data acquired through a previous cross-sectional, observational study, and received approval from the local institutional research ethics board and consent from research participants [2]. Thirty women (assigned female at birth) who had received previous instruction by a pelvic health physiotherapist on how to perform LAM contractions, and had performed home LAM exercise training for at least three weeks were recruited from local physiotherapy clinics. Potential participants were excluded if they were pregnant, <6 months post-partum, experienced dyspareunia, had a history of pelvic trauma, had pelvic organ prolapse (POP) beyond stage 2, or had previous gynaecologic surgery for incontinence or POP. 

Participants attended two visits in the research laboratory. At the first visit, three rest and three LAM maximum voluntary contraction (MVC) efforts were performed while force data were collected using a custom mechatronic IVD with the arms opened to a 35mm diameter. These tasks were performed in a supine then in a standing position. At the second visit, two-dimensional transperineal USI was performed while participants repeated the same rest and MVC tasks as the first visit. The USI video was captured while keeping the pubic symphysis, the bladder neck, and the anorectal angle clearly visible within the image frame throughout the three repetitions of the LAM MVC in both the supine and the standing positions. 

IVD outcomes included: force at rest, forces acquired before (baseline force) and during (peak force, relative peak force, and rate of force development) the three MVC tasks performed in each position. USI data were processed to identify the baseline morphology (from rest trials) and transient changes in levator plate length (LPL) and bladder neck height (BNH) relative to baseline observed during the three MVCs performed in each position. For all measures, the median value from the three trials was retained for analysis.

The effect of testing position (supine vs. standing) was analysed for each outcome using paired-samples t-tests. Parametric tests were conducted with bootstrapping based on 1000 samples with replacement and bias corrected accelerated 95% confidence intervals. A Bonferroni correction was applied (α =.005). The associations between relative peak force measured using IVD and transient changes in pelvic morphology observed on USI in  both supine and standing were assessed using multiple linear regression models with an uncorrected significance level (α =.05).

Twenty-six females (mean +/- SD, age =42 +/- 2 years, height =1.66 +/- 0.01m, weight =70.40 +/- 2.38 kg, n =22 parous) participated. 

A summary of IVD and USI outcomes is presented in Figure 1. IVD baseline force was higher in standing compared to supine, while relative peak force and rate of force development were lower in standing compared to supine. Absolute peak force was not different between standing and supine. USI outcomes showed that baseline LPL and LPL at the peak of the LAM contraction were longer in standing than in supine. BNH was lower at baseline in standing and was lower at the peak of the LAM contraction. There was no difference in the extent of LPL shortening or BN elevation observed during the LAM MVC between supine and standing. 

The regression models showed a significant negative relationship between relative peak force measured through IVD and change in LPL observed during LAM MVCs in both supine (adj. R-squared =.19, p =.05) and standing (adjusted R-squared =.49, p =.00) (Fig 2).

The findings that baseline force is higher and relative peak force measured using IVD is lower in standing compared to supine are consistent with previous reports [3]. The increased baseline force is likely the result of increased loading of the dynamometer arms by the weight of the abdominal/pelvic organs. These results highlight the importance of assessing LAM function in standing to reflect LAM force generating capacity in a functional position.

While the results show that MVC force measured through dynamometry is linearly related to the extent of shortening of the LPL seen on USI in both supine and standing positions, the relationships are not strong, explaining less than 20% of the variance in the data for the supine position and approximately half of the variance in the model for the standing position. 

The lack of a significant relationship between change in the BNH and LAM force measured through IVD is not surprising. The position of the BNH within the pelvic cavity is defined in relation to a plane drawn between the pubic symphysis and the anorectal junction. As the levator ani muscles contract, the anorectal junction moves, changing the position and angle of the reference plane and impacting the measured BNH. Changes in BNH are not recommended as a reflection of LAM contractile force. Higher order measures, such as the velocity or acceleration of landmarks may show stronger relationships with IVD measures of force or power and should be considered in future work.

While baseline and peak forces achieved during MVC are higher when measured using IVD in standing compared to supine, the relative peak force is lower. LPL is longer and BNH is more caudal when measured using USI in standing compared to supine, both at rest and at the peak of the MVC. Separate normative data sets are therefore needed for recordings made in standing and supine.

The relationship between LAM MVC force and the change in LPL observed during MVC was significant in both standing and supine positions, but, at best, only half of the variance in the data was explained by the model. It is not recommended to use the extent of change in the BNH or LPL observed on USI during LAM MVC as a reflection of force generated by the LAMs, either in research or in clinical practice.

Thibault-Gagnon S, Auchincloss C, Graham R, McLean L. The temporal relationship between activity of the pelvic floor muscles and motion of selected urogenital landmarks in healthy nulliparous women. J Electromyo. Kinesiol. 38 (2018) 126-135.
Czyrnyj CS, Berube M-E, Brooks K, Varette K, McLean L. Reliability and validity of a mobile home pelvic floor muscle trainer: The Elvie Trainer. Neurourol. Urodyn. 39 (2020) 1717-1731.
Bø K, Finckenhagen HB. Is there any difference in measurement of pelvic floor muscle strength in supine and standing position? Acta Obstet. Gynecol. Scand.  82 (2003) 1120-1124.