Frequency‐Dependent Effects of Transcranial and Translumbosacral Magnetic Stimulation on Anorectal Neurophysiology: A Dose–Response Study in Healthy Women

ABSTRACT

Introduction

We recently showed that prolonged anorectal motor‐evoked potential (MEP) latencies are associated with impaired anorectal function in women with fecal incontinence. Repetitive transcranial magnetic stimulation (rTMS) and translumbosacral magnetic stimulation (rTLSMS) are emerging neuromodulatory treatments, yet optimal frequency and site parameters remain undefined. This study evaluated the acute neurophysiological effects of rTMS and rTLSMS in healthy volunteers.

Methods

In a randomized crossover study, ten healthy women (18–35 years) underwent TMS and TLSMS assessments measuring MEP latencies/amplitudes in anal and rectal muscles across 10 neural segments. Stimulation was delivered in randomized sessions: rTMS at 1 and 5 Hz (250 pulses) and rTLSMS at 1, 5, and 10 Hz (2,400 pulses).

Results

Five‐Hz rTMS significantly reduced latencies in five segments and increased amplitudes in five. In contrast, 1 Hz rTMS had no effect. For rTLSMS, 5 Hz reduced latency in two segments and increased amplitude in two segments, while no significant changes were observed in the remaining segments. Similarly, 10 Hz reduced latency in one sacro‐anal pathway, with no significant effects in other segments. Effect sizes supported 5 Hz stimulation (latency: Cohen's d = −0.24; amplitude: 5 Hz rTLSMS, d = 0.34).

Conclusion

Among the tested protocols, 5 Hz stimulation, particularly cortical rTMS, showed the most consistent effects, with significant latency reductions observed in multiple anorectal segments. This frequency‐dependent, exploratory study provides a mechanistic basis for further investigation in clinical populations.

Key Points

• Central and peripheral efferent pathway dysfunction is implicated in fecal incontinence, with disrupted brain–gut neural circuits contributing to impaired continence.

• In healthy subjects, we found that 5 Hz rTMS, and to a lesser extent 5 Hz rTLSMS, induced frequency‐dependent neurophysiological changes in anorectal motor pathways.

• These findings suggest that 5 Hz magnetic stimulation may have potential relevance for neuromodulatory approaches targeting anorectal dysfunction, warranting further investigation in clinical populations.

Plain Language Summary

Fecal incontinence (FI) affects many women and significantly reduces quality of life. In addition to damage to the anal sphincter, altered communication between the brain and the gut may also play a role. Previous research showed delayed nerve responses in the anal region of women with FI, indicating impaired nerve function. This study tested the effects of two types of magnetic stimulation—repetitive transcranial magnetic stimulation (rTMS) and translumbosacral magnetic stimulation (rTLSMS)—in healthy women. We measured nerve responses in the anal and rectal muscles at different frequencies (1 Hz, 5 Hz, 10 Hz). Results showed that rTMS at 5 Hz reduced response times and increased response strength in 50% of segments studied, whereas 1 Hz rTMS had no significant effects. rTLSMS at 5 Hz and 10 Hz showed more limited and less consistent changes. Overall, 5 Hz rTMS produced the most consistent neurophysiological effects. These findings provide a basis for further studies to evaluate the potential clinical relevance of magnetic stimulation in patients with anorectal disorders.

Five Hz rTMS and, to a lesser extent, rTLSMS, induce beneficial neurophysiological changes in anorectal function, suggesting their potential as effective treatments for fecal incontinence by improving brain‐gut neural circuits and pelvic floor biomechanics.

1. Introduction

Fecal incontinence (FI) is defined as the involuntary loss of feces, and it is a highly prevalent condition among individuals living in the community. Estimated prevalence rates range from 7% to 15% [1] and are notably higher in women [2]. In a recent study conducted by our group, we reported prevalence rates of 11.3%, 12%, and 18% among postpartum women, postmenopausal women aged 45–74 years, and women aged 75 years and older, respectively [3]. FI has a profound impact on QoL [4], leading to social isolation in the most severe cases.

FI has commonly been attributed to pelvic floor dysfunction, especially impairments of the anal sphincters resulting from muscular or neural injury. Other contributing factors include alterations in rectal sensitivity, localized in the pelvic floor region, and bowel dysfunction and loose stools [5, 6]. However, recent perspectives have broadened the conceptual framework of FI, recognizing it as a disorder of the brain–gut axis [7, 8, 9]. Rather than being solely a peripheral pelvic floor disorder, growing evidence suggests that cortical, supraspinal, and spinal alterations may also contribute to impaired continence mechanisms [10, 11].

In our recently published study [11], we investigated corticoanorectal neurophysiology in women with FI using transcranial and translumbosacral magnetic stimulation. We demonstrated that patients exhibited significantly prolonged motor‐evoked potential (MEP) latencies in both cortico‐anal and lumbo‐anal pathways compared to age‐matched healthy controls. Furthermore, we identified a robust negative correlation between MEP latency and external anal sphincter squeeze pressure, suggesting that delayed neural conduction has functional consequences. These findings provide comprehensive evidence of both central and peripheral efferent pathway dysfunction in women with FI, supporting the hypothesis that disrupted brain–gut neural circuits contribute to impaired continence [10]. It also introduced the clinical potential of modulating these pathologically delayed neural pathways through targeted neuromodulation interventions. Magnetic neuromodulation has recently emerged as a promising therapeutic strategy for FI. A key clinical study was carried out by Rao and colleagues [12] in a randomized controlled trial using low‐frequency (1 Hz) translumbosacral magnetic stimulation (TLSMS) at 1, 5, and 15 Hz over six weeks, demonstrating significant improvements in FI symptoms, quality of life, and selected anorectal physiological measures. Their study included translumbosacral magnetic stimulation–evoked anal and rectal motor responses and showed that 1 Hz stimulation was associated with greater clinical and physiological improvement compared with higher frequencies. Stimulation and neurophysiological assessment in that study were limited to lumbosacral pathways, and frequency‐dependent effects on cortico‐anorectal conduction and acute segment‐specific latency modulation were not examined. Consequently, the optimal frequency for modulating central and segmental anorectal neural conduction remains to be fully characterized.

Other studies in neuromodulation, particularly in neurogenic dysphagia [13], chronic constipation [14], and pelvic floor disorders [15], have shown that frequency plays a critical role in shaping neural plasticity. Generally, low‐frequency stimulation (e.g., 1 Hz) has been associated with inhibitory effects, while higher frequencies (e.g., 5 Hz or more) can produce excitatory responses in corticospinal and autonomic pathways [16, 17, 18]. These findings suggest that specific frequencies may differentially influence targeted segments of the anorectal motor system. Understanding these frequency‐dependent effects is essential to rationally guide clinical application.

Building on this foundation, the present study was designed as a mechanistic follow‐up to our AJG 2025 paper. We have now evaluated whether repetitive transcranial magnetic stimulation (rTMS) and repetitive translumbosacral magnetic stimulation (rTLSMS) at varying frequencies acutely modulate corticoanorectal conduction in healthy volunteers (HV). This dose–response model intends to define the physiological responsiveness of anorectal motor pathways to magnetic stimulation at 1 Hz, 5 Hz, and 10 Hz, using MEP latency and amplitude as objective endpoints. Importantly, this design enables the characterization of frequency‐ and site‐specific effects on neural excitability as a necessary precursor to future interventional studies in patients with FI.

2. Materials & Methods

This was a randomized crossover dose–response study conducted in the GI Motility Unit at Hospital de Mataró (Barcelona), a secondary care hospital, on young female healthy volunteers (HV) (ClinicalTrials.gov code: NCT07192328). Participants had to be between 18 and 35 years, with no history of FI, and capable of understanding and providing informed consent. The exclusion criteria included pregnancy or breastfeeding, presence of a pacemaker, history of epilepsy or neurosurgery, withdrawal from the study, and poor treatment tolerance. Participants were informed of potential risks and instructed not to participate if pregnancy was suspected during the study period.

The study protocol was approved by the Hospital de Mataró Ethics Committee, under approval code 28/21. All participants provided written informed consent prior to their involvement in the study. All authors had access to the study data, reviewed, and approved the final manuscript.

2.1. Neurophysiological Evaluation

Transcranial magnetic stimulation (TMS) and TLSMS were used to obtain MEPs in the anal canal, rectum, and tibialis muscle (control). MEP latency (primary outcome) and amplitude (secondary outcome) were measured across different segments of the efferent pathways, including cortico‐rectal and cortico‐anal, and the peripheral segments lumbo‐rectal, lumbo‐anal, sacro‐rectal, and sacro‐anal, on both right and left sides. Participants remained in the prone position throughout the procedures.

Cortical TMS was performed using a Magstim BiStim stimulator with a 110 mm double‐cone coil (Magstim, UK). Stimulation was applied over the primary motor cortex at the vertex (BA4) (see Figure 1). The stimulation hotspot was identified by moving the coil in 1‐cm increments in the anteroposterior direction until optimal anal or rectal MEP responses were obtained. MEPs were recorded at an intensity 10% above the motor threshold, defined as the lowest stimulus intensity eliciting reproducible MEPs greater than 50 μV in five trials. Tibialis MEPs were recorded with surface electrodes [19].

FIGURE 1.

Methodology used to use repetitive cortical and lumbosacral magnetic stimulation. RTMS: Repetitive transcranial magnetic stimulation; rTLSMS: repetitive translumbosacral magnetic stimulation.

For TLSMS, a 90‐mm circular coil (Magstim, UK) was used to stimulate bilaterally the lumbar (L2–L3) and sacral (S2–S3) regions. Four stimulation sites were identified and marked approximately 4 cm lateral to the midline. Supramaximal stimulation was applied to assess the peripheral nerve conduction time along the lumbo‐rectal, lumbo‐anal, sacro‐rectal, and sacro‐anal pathways (see Figure 1).

Stimulation intensity for both TMS and TLSMS was kept constant within each session, and identical stimulation parameters were used for pre‐ and post‐stimulation assessments to ensure that any observed changes in MEP amplitude were not attributable to variations in stimulation intensity.

The MEPs were recorded using the BrainSight Neuronavigation system (Rogue Research Inc., Canada), with a custom anorectal probe for rectal and anal MEP recordings (Gaidhealtec, Scotland). Representative MEPs recorded after TMS and TLSMS are shown in Figure S1. MEP latency was measured from stimulus onset to the first waveform deflection, and the amplitude from peak‐to‐peak, with the average of five recordings calculated for each segment. Detailed methodological aspects have been described previously [11] (Figure 1).

2.2. Interventions

rTMS and rTLSMS were administered using a Magstim Rapid2 repetitive stimulator at pre‐identified hotspots cortical, lumbar, and sacral regions bilaterally [11, 20]. Each experimental session consisted of three phases: A pre‐treatment neurophysiological assessment, a stimulation treatment phase, and an immediate post‐treatment neurophysiological reassessment. During the treatment phase, participants received magnetic stimulation at a single site and frequency per session. Cortical rTMS was delivered using the same double‐cone coil employed for neurophysiological assessment, with 250 pulses applied at either 1 Hz or 5 Hz, at an intensity set at 10% below the cortical motor threshold. For peripheral stimulation, rTLSMS was delivered to the lumbar and sacral regions using a 70‐mm refrigerated air film coil (Magstim, UK), with a total of 2400 pulses (600 per hotspot) applied at 1 Hz, 5 Hz, or 10 Hz. No more than one stimulation modality or frequency was administered within the same session. Pre‐ and post‐treatment MEPs were recorded using identical stimulation parameters and intensities.

Each healthy participant underwent all five sessions in randomized order, with one‐week washout periods between each to prevent carry‐over effects. No control or placebo group was included. Randomization was performed using the GraphPad Randomizer.

2.3. Statistical Analyses

Continuous variables were described with mean ± SD and percentages. Given the predefined directional hypotheses based on prior neurophysiological evidence, one‐tailed t‐tests or their nonparametric equivalents (Wilcoxon signed‐rank tests) were used to compare pre‐ and post‐stimulation measures of continuous variables. A p‐value of < 0.05 was considered statistically significant. The effect size was assessed using Cohen's d index [21].

To account for the crossover design, carry‐over and period effects were formally evaluated using a Linear Mixed Model (LMM), with subject included as a random effect. Fixed effects included treatment, previous treatment received (carry‐over proxy), study period, anatomical segment, and outcome measure type. In addition, pre‐treatment baseline values were compared across study periods using Friedman tests to assess the adequacy of the washout period.

3. Results

A total of 10 healthy women (mean age 25.9 ± 5.28 years) were consecutively recruited. All participants completed the study procedures; therefore, no participants were excluded from the final analysis.

3.1. Repetitive Transcranial Magnetic Stimulation (rTMS)

Stimulation of the motor cortex with rTMS at 5 Hz resulted in a significant shortening of the latency in the following segments: Cortico‐anal, cortico‐rectal, lumbo‐rectal right, and bilateral sacro‐anal. In contrast, rTMS at 1 Hz did not produce significant differences in any of the studied segments (Table 1).

TABLE 1.

Differences in latencies after cortical stimulation with rTMS 1 hz and rTMS 5 Hz.

Latencies	rTMS 1 Hz			rTMS 5 Hz
	Pretreatmente	Posttreatment		Pretreatmente	Posttreatment
Segment Studied	Mean ± SD	Mean ± SD	p‐value	Mean ± SD	Mean ± SD	p‐value
Tibial Latency	31.73 ± 2.23	33.19 ± 2.47	0.988	33.08 ± 2.169	33.44 ± 3.12	0.69
Cortico‐Anal Latency	23.8 ± 1.07	24.11 ± 0.78	0.893	24.26 ± 0.78	23.66 ± 0.97	0.007
Cortico‐Rectal Latency	19.86 ± 2.26	20.82 ± 2.19	0.986	20.3 ± 2.11	19.28 ± 1.63	0.034
Lumbo‐Anal right Latency	4.59 ± 0.76	4.51 ± 0.67	0.545	4.74 ± 0.91	5.02 ± 1.38	0.476
Lumbo‐Anal left Latency	4.54 ± 0.54	4.64 ± 0.45	0.813	4.90 ± 1.05	4.75 ± 0.53	0.41
Lumbo‐Rectal right Latency	3.99 ± 1.60	4.23 ± 1.52	0.736	4.24 ± 0.40	3.97 ± 0.56	0.038
Lumbo‐Rectal left Latency	4.04 ± 1.61	4.46 ± 1.51	0.946	4.15 ± 1.0	4.11 ± 0.60	0.337
Sacro‐Anal right Latency	4.34 ± 0.83	4.9 ± 1.71	0.762	4.36 ± 0.99	4.14 ± 0.80	0.033
Sacro‐Anal left Latency	4.39 ± 0.79	4.7 ± 1.84	0.406	4.90 ± 1.25	4.26 ± 0.58	0.01
Sacro‐Rectal right Latency	4.11 ± 1.88	4.74 ± 1.88	0.92	3.51 ± 0.63	3.55 ± 0.61	0.576
Sacro‐Rectal left Latency	4.59 ± 1.67	4.61 ± 1.66	0.723	3.61 ± 0.65	3.67 ± 0.43	0.65

Note: Bold values indicate statistically significant differences (p < 0.05).

Abbreviations: rTMS, repetitive transcranial magnetic stimulation; SD, (standard deviation).

Regarding the MEP amplitudes, a similar pattern was observed. The segments cortico‐anal, lumbo‐anal, and lumbo‐rectal left, and bilateral sacro‐anal demonstrated a significant increase in MEP amplitudes when stimulated at 5 Hz. No statistically significant changes were found when stimulating at 1 Hz (Table 2).

TABLE 2.

Differences in amplitudes of MEPs after cortical stimulation with rTMS 1 hz and rTMS 5 Hz.

Amplitudes	rTMS 1 Hz			rTMS 5 Hz
	Pretreatmente	Posttreatment		Pretreatmente	Posttreatment
Segment Studied	Mean ± SD	M Mean ± SD	p‐value	Mean ± SD	Mean ± SD	p‐value
Tibial Amplitude	126.8 ± 107.2	127.9 ± 121.7	0.476	382.8 ± 587.2	396.3 ± 586.6	0.344
Cortico‐Anal Amplitude	127.4 ± 48.9	110 ± 51.6	0.862	84.3 ± 41.9	135 ± 70.7	0.007
Cortico‐Rectal Amplitude	92.6 ± 78.5	85.7 ± 61.1	0.758	96.7 ± 46.5	77.7 ± 43.1	0.966
Lumbo‐Anal right Amplitude	253.6 ± 147.3	320 ± 267.2	0.216	211.7 ± 135.5	204.7 ± 71.9	0.594
Lumbo‐Anal left Amplitude	206.9 ± 132	195.6 ± 136.7	0.71	182.3 ± 97.7	226.3 ± 137.5	0.014
Lumbo‐Rectal right Amplitude	476 ± 344.2	360 ± 316.2	0.96	516 ± 418.8	466.7 ± 292.4	0.734
Lumbo‐Rectal left Amplitude	297.8 ± 165.4	284.7 ± 123.7	0.606	306.7 ± 204	401.3 ± 186.9	0.03
Sacro‐Anal right Amplitude	287.8 ± 183.7	262 ± 269.5	0.827	245.6 ± 128.8	286.7 ± 148.5	0.032
Sacro‐Anal left Amplitude	266.3 ± 149	255.1 ± 154.4	0.634	189.9 ± 108.8	238.4 ± 117.7	0.025
Sacro‐Rectal right Amplitude	203.9 ± 78.5	198.1 ± 95.5	0.59	302 ± 243.1	291.9 ± 231.3	0.627
Sacro‐Rectal left Amplitude	336.8 ± 276.7	295.1 ± 176.5	0.68	374.6 ± 130.1	332.8 ± 197.5	0.7

Note: Bold values indicate statistically significant differences (p < 0.05).

Abbreviations: rTMS, repetitive transcranial magnetic stimulation; SD, (standard deviation).

3.2. Repetitive Translumbosacral Magnetic Stimulation

With regard to MEP latencies, significant shortening of the latencies was observed in the cortico‐anal and sacro‐anal right segments when stimulated with rTLSMS at 5 Hz. Also, sacro‐anal left latency was significantly shortened when stimulating at 10 Hz. No other segments showed statistically significant changes at 1 Hz, 5 Hz, or 10 Hz (See Table 3).

TABLE 3.

Differences in latencies after peripheral stimulation with rTLSMS 1 hz, rTLSMS 5 Hz, and rTLSMS 10 Hz.

Latencies	rTLSMS 1 Hz			rTLSMS 5 Hz			rTLSMS 10 Hz
	Pretreatmente	Posttreatment		Pretreatmente	Posttreatment		Pretreatmente	Posttreatment
Segment Studied	Mean ± SD	Mean ± SD	p‐value	Mean ± SD	Mean ± SD	p‐value	Mean ± SD	Mean ± SD	p‐value
Tibial Latency	31.52 ± 3.41	31.67 ± 4.26	0.787	33.60 ± 4.98	33.20 ± 4.7	0.375	32.56 ± 2.11	32.53 ± 2.87	0.469
Cortico‐Anal Latency	23.87 ± 0.49	24.09 ± 3.07	0.577	24.14 ± 0.92	23.58 ± 0.91	0.028	23.79 ± 0.96	23.82 ± 0.95	0.629
Cortico‐Rectal Latency	20.10 ± 1.85	19.54 ± 2.04	0.167	20.18 ± 1.13	19.46 ± 1.65	0.138	19.2 ± 1.66	19.22 ± 1.01	0.451
Lumbo‐Anal right Latency	5.51 ± 1.76	5.26 ± 1.80	0.367	4.80 ± 1.14	4.61 ± 0.54	0.583	5.04 ± 1.62	5.26 ± 1.83	0.815
Lumbo‐Anal left Latency	5.05 ± 0.77	4.87 ± 0.68	0.4	4.49 ± 0.55	4.59 ± 0.46	0.75	5.32 ± 1.87	4.81 ± 0.66	0.723
Lumbo‐Rectal right Latency	3.61 ± 0.96	3.48 ± 1.06	0.534	4.18 ± 0.35	4.20 ± 1.80	0.513	3.87 ± 0.69	3.85 ± 0.59	0.479
Lumbo‐Rectal left Latency	3.93 ± 1.28	4.03 ± 0.123	0.722	4.80 ± 1.06	4.31 ± 0.89	0.344	3.96 ± 0.78	3.83 ± 0.65	0.292
Sacro‐Anal right Latency	4.95 ± 1.29	4.42 ± 1.15	0.202	4.27 ± 0.63	4.04 ± 0.85	0.045	4.52 ± 0.81	4.60 ± 0.71	0.371
Sacro‐Anal left Latency	4.67 ± 1.14	4.76 ± 0.9	0.496	4.47 ± 0.84	4.02 ± 0.59	0.072	4.92 ± 1.28	4.45 ± 0.69	0.038
Sacro‐Rectal right Latency	3.50 ± 1.09	3.28 ± 1.36	0.086	3.41 ± 0.91	3.22 ± 1.06	0.106	3.44 ± 0.49	3.24 ± 0.82	0.142
Sacro‐Rectal left Latency	3.48 ± 1.32	3.38 ± 1.01	0.398	3.99 ± 0.67	4.40 ± 1.4	0.688	3.95 ± 0.66	3.66 ± 0.66	0.195

Note: Bold values indicate statistically significant differences (p < 0.05).

Abbreviations: rTLSMS, repetitive translumbosacral magnetic stimulation; SD, (standard deviatiYon).

Regarding the amplitudes when stimulating peripherally at 1 Hz, 5 Hz, or 10 Hz, we found a significant increase in the MEP amplitude only in the sacro‐anal right and left segments with rTLSMS at 5 Hz. No other changes were statistically significant when stimulating peripherally (See Table 4).

TABLE 4.

Differences in amplitudes after peripheral stimulation with rTLSMS 1 hz, rTLSMS 5 Hz, and rTLSMS 10 Hz.

Amplitudes	rTLSMS 1 Hz			rTLSMS 5 Hz			rTLSMS 10 Hz
	Pretreatmente	Posttreatment		Pretreatmente	Posttreatment		Pretreatmente	Posttreatment
Segment Studied	Mean ± SD	Mean ± SD	p‐value	Mean ± SD	Mean ± SD	p‐value	Mean ± SD	Mean ± SD	p‐value
Tibial Amplitude	140 ± 147.5	249 ± 247.6	0.049	214 ± 202.8	379 ± 442.9	0.075	134 ± 113.9	148 ± 107.4	0.266
Cortico‐Anal Amplitude	107 ± 54.6	104 ± 58.5	0.584	155 ± 141.1	152 ± 100.9	0.537	119 ± 52.5	109 ± 41.1	0.694
Cortico‐Rectal Amplitude	100 ± 46.5	115 ± 58.9	0.22	137 ± 144.8	117 ± 94.8	0.785	103 ± 83.5	152 ± 177.7	0.417
Lumbo‐Anal right Amplitude	348 ± 223.4	310 ± 262.5	0.729	152 ± 84.9	192 ± 113.4	0.277	234 ± 226.9	186 ± 134.2	0.819
Lumbo‐Anal left Amplitude	487 ± 179.6	317 ± 151.4	0.945	178 ± 150.4	215 ± 161.7	0.121	210 ± 194.5	218 ± 210.6	0.348
Lumbo‐Rectal right Amplitude	463 ± 373.5	396 ± 276.8	0.888	309 ± 166.2	433 ± 272.3	0.5	386 ± 296.5	382 ± 265.7	0.527
Lumbo‐Rectal left Amplitude	487 ± 179.6	317 ± 151.4	1	229 ± 139.8	230 ± 168.5	0.486	298 ± 120.3	316 ± 192.9	0.389
Sacro‐Anal right Amplitude	303 ± 243.6	424 ± 244.6	0.32	228 ± 102.9	379 ± 137.1	0.007	248 ± 196.3	287 ± 243.2	0.258
Sacro‐Anal left Amplitude	303 ± 256.2	282 ± 231.7	0.687	196 ± 111.6	245 ± 92.6	0.039	202 ± 140.4	206 ± 146.7	0.469
Sacro‐Rectal right Amplitude	401 ± 227.9	398 ± 156	0.513	292 ± 208.2	300 ± 153.9	0.438	237 ± 166.2	348 ± 228.6	0.077
Sacro‐Rectal left Amplitude	285 ± 187.1	455 ± 333.9	0.139	404 ± 413.4	492 ± 467.2	0.264	342 ± 257.9	336 ± 372	0.844

Note: Bold values indicate statistically significant differences (p < 0.05).

Abbreviations: rTLSMS, repetitive translumbosacral magnetic stimulation; SD, (standard deviation).

We also conducted an effect size analysis using Cohen's d, a standardized measure to quantify the magnitude of treatment effects. This analysis was applied to assess the impact of rTMS (1 Hz and 5 Hz) and rTLSMS (1 Hz, 5 Hz, and 10 Hz) on both MEP latency and amplitude across all neurophysiological segments. To evaluate the overall efficacy of each stimulation protocol, we calculated the mean Cohen's d across segments, providing a summary estimate of the average treatment effect on both latency and amplitude.

The greatest overall effect on latency was observed with both rTMS and rTLSMS at 5 Hz (mean Cohen's d = −0.24). However, statistically significant pre–post differences in latency were found in five neural segments following rTMS at 5 Hz (Table 1), compared to only two segments following rTLSMS at 5 Hz (cortico‐anal and right sacro‐anal; Table 3). In contrast, after 1 Hz rTMS, opposite effects to those observed at 5 Hz were noted. A comparison of the results is shown in Figure 2.

FIGURE 2.

Pooled Cohen's d values representing the effect size on the latencies of each treatment condition. Negative values (in green) reflect an improvement (shortening of the latencies). RTMS: Repetitive transcranial magnetic stimulation; rTLSMS: repetitive translumbosacral magnetic stimulation.

Regarding amplitude in the effect size analysis, the greatest effect was again observed with rTLSMS at 5 Hz, which showed a pooled Cohen's d of 0.34. However, only two segments demonstrated statistically significant pre–post differences. For rTMS at 5 Hz, the pooled Cohen's d was 0.18; nonetheless, this condition showed the highest number of segments with statistically significant changes (See Figure 3).

FIGURE 3.

Pooled Cohen's d values representing the effect size on the amplitudes of MEPs of each treatment condition. Positive values (in green) indicate improvement (increasing of MEPs amplitudes). RTMS: Repetitive transcranial magnetic stimulation; rTLSMS: repetitive translumbosacral magnetic stimulation.

The Linear Mixed Model showed no significant carry‐over effect (previous treatment: F(4356) = 0.97, p = 0.422) and no significant period effect (period: F(3,457) = 0.11, p = 0.953), indicating that neither treatment sequence nor temporal progression across sessions systematically influenced the neurophysiological outcomes. Baseline pre‐treatment values also did not differ significantly across study periods (Friedman test, all p > 0.05), further supporting the adequacy of the washout period.

4. Discussion

This study aimed to explore the relative effects of different stimulation sites and frequencies on neural conduction (as measured by latency) and MEP amplitude in pathways associated with anorectal function. Building on our previous research, which demonstrated prolonged latencies in patients with FI compared to age‐matched HV and revealed a negative correlation between latency and sphincter strength [11], we proposed latency reduction as a potential functionally relevant neurophysiological marker linked to anorectal motor function performance. Although anorectal MEP latency represents only one component of the multifactorial pathophysiology of fecal incontinence, the within‐subject latency reductions observed in healthy volunteers suggest acute modulation of neural conduction along anorectal motor pathways. We hypothesize that even small changes in latency may be physiologically meaningful, particularly in patients with more markedly delayed latencies associated with cortico‐anorectal neuropathy, as previously described also by our group in women with FI.

By focusing on healthy volunteers, the present study was able to characterize acute within‐subject neurophysiological modulation under controlled conditions, thereby minimizing confounding effects related to disease severity, structural damage, or aging. These findings serve as a mechanistic framework that will need to be further examined in patient populations, particularly in older women with FI, where age‐related factors such as neuromuscular degeneration and comorbid pelvic floor dysfunction may further impair anorectal neurophysiology.

One of the main findings of this study is that 5 Hz rTMS applied to the motor cortex produced a statistically significant shortening of conduction latencies in five distinct anorectal pathways, whereas 1 Hz rTMS elicited no significant changes in latency. These results are consistent with an excitatory influence of 5 Hz stimulation on corticospinal pathways involved in anorectal control, while 1 Hz may induce inhibitory effects—an interpretation consistent with existing literature on rTMS applied in other neurological conditions, such as swallowing function [16]. Given that shorter latencies have been associated with improved external anal sphincter function [11, 22], our findings imply that high‐frequency rTMS may enhance neural conduction along anorectal motor pathways. Moreover, these improvements occurred in the same neural segments that were found to be impaired in our earlier cohort of patients with FI, suggesting that these pathways are not only pathophysiologically relevant but also physiologically responsive to stimulation.

Regarding MEP amplitudes, although our previous studies did not demonstrate clear differences between patients and HV, increased post‐stimulation amplitudes have been reported in other neurophysiological contexts [23]. In the present study, 5 Hz rTMS was associated with increased MEP amplitudes in several segments, whereas 1 Hz stimulation showed no significant effects. These findings are consistent with an excitatory effect of corticospinal pathways induced by 5 Hz cortical stimulation; however, in the present framework, amplitude was considered a secondary outcome and interpreted as supportive of the primary latency findings rather than as an independent functional marker.

Among the peripheral stimulation conditions, 5 Hz rTLSMS showed the most consistent effects compared to 1 Hz and 10 Hz. It produced statistically significant latency reductions in two segments—the cortico‐anal and right sacro‐anal pathways. Although these effects were less consistent than those observed with cortical rTMS, they suggest a frequency‐dependent excitatory influence. In comparison, 10 Hz rTLSMS produced a significant change in only one segment (left sacro‐anal), and 1 Hz did not result in any significant latency change.

Similarly, among the peripheral stimulation, only 5 Hz rTLSMS induced increases in MEP amplitude, and these effects were limited to the sacro‐anal segments, whereas other frequencies showed no significant effects. Overall, peripheral stimulation at 5 Hz demonstrated less consistent and more segment‐specific effects compared with cortical rTMS. These results must also be viewed in light of prior clinical studies. Rao and colleagues demonstrated that low‐frequency (1 Hz) TLSMS delivered over six weeks resulted in clinical improvement in patients with FI, supporting its therapeutic value [12]. While our study did not observe significant physiological effects with 1 Hz stimulation in HV, this does not contradict those findings. Rather, it suggests that different frequencies may engage distinct neuromodulatory mechanisms—particularly when comparing acute neurophysiological responses in healthy individuals to cumulative therapeutic effects observed after repeated stimulation in patients.

In terms of effect size, the analysis using Cohen's d further supports the frequency‐dependent nature of the observed neurophysiological effects. Both 5 Hz rTMS and 5 Hz rTLSMS were associated with latency reductions, each showing small negative effect sizes (d = −0.24), consistent with an excitatory modulation on neural conduction. In contrast, 1 Hz rTMS was associated with latency prolongation (positive d values), suggesting a directionally opposite or inhibitory response. Peripheral 1 Hz and 10 Hz stimulation showed smaller effect sizes, with only one segment showing a significant change following 10 Hz rTLSMS.

For MEP amplitudes, 5 Hz rTLSMS was associated with the largest positive effect size (d = 0.34), while cortical 5 Hz rTMS showed a smaller yet still significant effect (d = 0.18). Stimulation at the other frequencies produced negative changes. Although amplitude effect sizes were small, 5 Hz rTMS produced statistically significant changes across a greater number of neural segments than peripheral stimulation, indicating a more consistent and spatially distributed modulatory effect.

Taken together, these findings suggest that 5 Hz stimulation, particularly via cortical rTMS, shows the most consistent pattern of modulation of neurophysiological parameters associated with anorectal function. The remaining question for clinical translation is whether cortical or translumbosacral stimulation offers superior therapeutic outcomes, an issue that warrants further investigation in patient populations with FI. Importantly, our findings also suggest that stimulation parameters may need to be tailored not only by site but also by desired neurophysiological outcome (e.g., excitatory vs. inhibitory), a principle supported by similar research in dysphagia, chronic constipation, and other pelvic floor dysfunctions [24].

The limitations of this study include the small sample size (n = 10) and the use of healthy young women, which may limit generalizability to clinical populations. Although this approach was chosen for mechanistic and translational reasons, hormonal fluctuations across the menstrual cycle were not controlled for and may have contributed to inter‐individual variability. While formal analyses did not identify significant carry‐over or period effects, residual influences between sessions cannot be completely excluded, given the small sample size and crossover design. However, prior neurophysiological evidence suggests that MEP changes induced by a single stimulation session are transient and unlikely to persist beyond hours to days [25, 26], supporting—though not fully guaranteeing—the adequacy of a one‐week washout period for mechanistic crossover studies. Finally, effect sizes were small, and not all changes reached statistical significance, and the use of one‐tailed statistical tests may increase the risk of type I error. Therefore, results should be interpreted with caution and confirmed in future studies. Further investigation with larger cohorts, parallel‐group designs, and repeated stimulation sessions will be required to assess the durability and clinical relevance of these findings in patients with FI.

This study suggests that 5 Hz rTMS, and to a lesser extent 5 Hz rTLSMS, induce acute neurophysiological changes in anorectal pathways consistent with excitatory modulation. These exploratory findings provide a mechanistic basis for future studies evaluating the clinical relevance of frequency‐specific magnetic stimulation in fecal incontinence.

Author Contributions

Lluís Mundet: study conceptualization, data acquisition, statistical analysis, manuscript writing. Tennekoon Buddhika Karunaratne: data acquisition, manuscript writing and review. Alba Raventós: data acquisition. Pere Clavé: discussion of results and writing review.

Funding

This work was supported by the Departament de Salut, Generalitat de Catalunya, SLT017/20/000125, SLT017/20/000236. Instituto de Salud Carlos III, ICI20/00117.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Examples of MEP after Transcranial Magnetic Stimulation and Translumbosacral Magnetic Stimulation.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.