Electromyography (EMG) is used to record action potentials (APs) generated by the depolarization of muscle fibers [1]. Surface electromyography (sEMG) is commonly used to record electrical activity of the pelvic floor muscles (PFMs) with active electrodes placed on the skin of the perineum or on the mucosa within the anal, vaginal canal or urethra. sEMG recordings are typically interpreted as representing the activity originating from striated muscles closest to the electrodes [2]. Results from animal studies raise questions about this simplified interpretation. Recordings with sEMG made via the rectal wall in dogs [3], and other similar methods with other species, include APs that are generated by smooth muscle, referred to as spike potentials. In humans, it is likely that sEMG recordings via these electrodes will include spike potentials from smooth muscle along with APs from striated muscles. Presence of smooth muscle spike potentials would lead to over-estimation of activity attributed to striated muscles and challenge the traditional interpretation of EMG recordings. Based on animal data smooth muscle spike potentials would be identifiable based on their duration (>20 ms vs. 5-20 ms for striated muscle APs), discharge rate (1-5 Hz vs. >10 Hz for striated muscle APs) and inability to be controlled voluntarily. Whether smooth muscle APs contaminate anal or vaginal sEMG recordings of the PFMs and impact on the analysis of striated muscle sEMG amplitude remains unclear. Although sEMG recordings of gastrointestinal smooth muscle have been made in humans with specialised electrodes, we are unaware of studies that have reported smooth muscle activity using sEMG for PFM applications in humans. This study aimed to investigate whether potentials are identifiable in PFM sEMG recordings from the anal (male) and vaginal (female) canal that would be attributed to smooth muscle, and if present, to evaluate their properties.

Data were acquired from two male and one female participants. Methods were approved by the Institutional Medical Research Ethics Committee. PFM sEMG recordings were made with custom-made electrodes that consisted of 2 parallel recording surfaces (1 cm long, sterling silver wire) soldered to stranded tinned copper conductor. Wires were passed into a 25-cm medical grade polyvinyl chloride tube, and a small hole was cut between the electrode to apply gentle suction. For male participants (30 and 53 years old), the probe involved 2 pairs of recording surfaces separated by 4 cm, such that when the electrodes were placed in the anal canal recordings could be made simultaneously from the puborectalis and external anal sphincter muscles. Male participants were assessed using 2 probes fixed either side of the balloon of a Foley catheter that could be inflated to increase the separation between the left and right electrode at the level of the puborectalis muscle. For the female participant (32 years old), an electrode with a single pair of recording surfaces was placed into the vaginal canal to record the activity of the right levator ani muscle. Electrodes were single-use, sterilized, and a water-based lubricant was applied to facilitate insertion. The ground electrode was placed over the right iliac crest. Electrodes were connected to a preamplifier (NL844 Pre-Amplifier, Neurolog, Digitimer Ltd). EMG signals were amplified 5000 times for male participants and 2000 times for the female participant (NL820 Isolation Amplifier, Neurolog, Digitimer Ltd), filtered (20-1000 Hz) with a notch filter at 50 Hz (NL125, Neurolog, Digitimer Ltd), and sampled at 2000 Hz using a Power1401 data acquisition system and Spike2 software (v7, Cambridge Electronic Design Ltd).

Male participants were positioned in a reclined long sitting and the female participant was positioned in a supine lying on an examination table. EMG recordings were monitored in real-time. Clear periods of spontaneous smooth muscle APs were identifiable based on the AP duration (>20 ms) and slow discharge rate. When present, a physiotherapist instructed the participant, who was skilled in relaxing or contracting their PFMs, to either relax or contract their PFMs at various intensities (10%, 20%, 30%, 50%, or 100%). In some cases, recordings were made with filter settings at 0-1000 Hz to enable analysis of the slow frequency components of the smooth muscle spike potentials.

EMG recordings were visually inspected and analyzed using Spike2 software. When identified, the properties of smooth muscle potentials were extracted, including discharge rate, duration and amplitude. Data are presented as a range of values. We observed whether the occurrence of smooth muscle APs changed according to voluntary contraction and relaxation of the PFMs. We compared EMG amplitude at rest and during contractions between 3-s periods of data with and without presence of smooth muscle potentials.

Potentials with properties consistent with smooth muscle spike potentials were identified at rest and during PFM contractions in male and female participants using the EMG probes designed to assess the electrical activity of the PFMs. Discharge rate of smooth muscle spike potentials were not influenced by PFM contractions, which confirms that they were not influenced by voluntary drive. Smooth muscle spike potentials were more easily detected at rest when there was minimal summation with striated muscle APs. Smooth muscle spike potentials had a discharge rate ranging between 1.2-4.4 Hz, a duration ranging between 50-152 ms, and an amplitude ranging between 15-63 µV. When the high-pass filter was changed to 0 Hz, a slower frequency component of smooth muscle APs was observed – APs were longer (91-178 ms) and had a greater amplitude (78-97 µV). When sEMG amplitude was compared between periods with the striated PFMs at rest, with and without smooth muscle APs, the amplitude was increased by 51-523%. Smooth muscle APs were less distinct as the intensity of PFM contractions increased, and the over-estimation of activity as the intensity of PFM contractions increases was ~53% during 10% PFM contraction and ~23% during 50% PFM contraction.

We made several other observations that provide additional insight into the smooth muscle recordings. In males, smooth muscle APs were identified on both sides, but they were rarely present on both sides at the same time. This implies that recordings arise from longitudinal rather than circular smooth muscle fibers. Synchronous smooth muscle APs on both sides were observed in a very small number of instances. This might reflect the activity of circular fibers or a chance observation of simultaneous activity of longitudinal fibers bilaterally. Smooth muscle APs were present regardless of the inflation of the balloon between the electrodes in the male participants, suggesting that APs were not influenced by this minor distension at the anorectal junction.

This report provides evidence that smooth muscle spike potentials can be recorded with sEMG electrodes in the anal and vaginal canal in humans. When present, smooth muscle spike potentials influence EMG amplitude and could interfere with the estimation of activity of striated PFMs. The impact of this contamination was greatest at rest and reduced during PFM contractions. Accurate evaluation of EMG data of striated muscles should involve selection of periods of recording that are not contaminated by smooth muscle APs. We provide measures of smooth muscle AP properties that could assist with their identification.

This study is the first to report presence of smooth muscle spike potentials in sEMG recordings designed to assess striated PFM activity. Findings demonstrate the importance of considering the potential contribution of smooth muscle spike potentials in PFM sEMG data.

Bo K, Frawley HC, Haylen BT, et al. An International Urogynecological Association (IUGA)/International Continence Society (ICS) joint report on the terminology for the conservative and nonpharmacological management of female pelvic floor dysfunction. Neurourol Urodyn. 2017;36(2):221-244. doi:10.1002/nau.23107
Frawley H, Shelly B, Morin M, et al. An International Continence Society (ICS) report on the terminology for pelvic floor muscle assessment. Neurourol Urodyn. 2021;40(5):1217-1260. doi:10.1002/nau.24658
Shafik A, El-Sibai O. Transmission of rectal electric waves: is it through circular or longitudinal smooth muscle layers or both? Arch Physiol Biochem. 2001;109(2):127-134. doi:10.1076/apab.109.2.127.4274