Introduction
Hand walking is a mainstay of equine postoperative exercise protocols and is also used preoperatively as a mechanism to reduce fill by ingesta within the digestive tract. In a survey of board-certified veterinary specialists, 84% to 86% of clinicians reported use of early postoperative mobility, specifically hand walking exercise, soon after colic surgery in an effort to reduce the risk of postoperative ileus.1,2 Hand walking is also commonly recommended for horses with gastrointestinal disease to encourage the movement of ingesta and gas through the digestive tract. In humans, early mobility after surgery has demonstrated a reduction in the risk of postoperative ileus, resulting in shortened hospital stays and decreased morbidity after gastrointestinal surgery.3–7 However, to our knowledge, the effect of hand walking exercise on gastrointestinal motility has not been evaluated in the horse.
Slow-wave electrical activity produced by the interstitial cells of Cajal provides the electrical stimulation necessary to initiate muscular contractions in the intestines.8 The rate of contractions is correlated to the frequency of the slow waves, whereas the amplitude, or power, of the slow waves reflects the strength of the intestinal contractions.9 Electrointestinography (EIG) is a noninvasive method to record and evaluate neuromuscular activity and represents the summation of electrical signals from different populations of pacemaker cells at multiple points throughout the intestines. Therefore, the EIG waveform represents the average myoelectrical activity over time for a targeted section of the gastrointestinal tract.10–13
In humans, gastric myoelectrical activity has been quantified with the use of skin surface electrodes to record electrogastrograms as a method to objectively measure motility abnormalities by EIG.14–17 In horses, this technology has been applied in research settings to identify the presence of postoperative ileus and the effects of drugs on intestinal motility patterns.18–21 The amplitude of the waveform has been used to describe the effects of an intervention on the strength and rate of contractions, whereas the distributions of the amplitude across the frequency range can be used to describe arrhythmias in slow-wave activity.18–22 As changes in slow-wave activity are not assessed on the basis of visual examination of the waveform, additional analytic tools, including power spectral density analysis techniques that measure the strength of a signal related to its frequency, are required for interpretation of EIG.10,23
The objective of this study was to evaluate the effects of hand walking exercise on myoelectrical activity in the cecum and left ventral colon (LVC) with the use of multichannel EIG in healthy, fasted horses placed on box stall rest. The goal was to determine the effects of hand walking on EIG dominant frequency (DF), dominant power (DP), and frequency distribution of total power (TP) to assess changes in intestinal myoelectrical rate, intensity, and rhythm. An additional goal of this study was to compare the findings of the EIG with contractions identified by transabdominal ultrasonography and auscultation of borborygmi. The hypothesis was that hand walking exercise would increase intestinal myoelectrical activity, ultrasonographic contractions, and borborygmi compared to stall rest alone.
Methods
Horses
Horses eligible for enrollment in the study were those of any full-sized breed, sex, and age > 2 years. A minimum sample size of 8 was calculated to provide a power of 80% with an α of 0.05, based on a comparable study19 that evaluated multichannel EIG in horses, with an expected 30% difference after treatment. Participants were selected as a convenience sample from the Michigan State University Teaching Herd at the College of Veterinary Medicine, with additional privately owned horses enrolled after obtaining informed consent from their owners. Horses with a history of colic or gastrointestinal disorders in the previous 6 months were excluded. Prior to admission to the study, a physical examination and ultrasonographic examination of the abdomen were performed to identify any evidence of systemic illness, preexisting gastrointestinal abnormalities, or evidence of previous abdominal surgery that would exclude a horse from the study. In addition, the horses did not receive administration of α-2 adrenergic receptor agonist or antagonist drugs, anticholinergic drugs, or NSAIDs during the 14 days prior to the study. This protocol was approved by the Michigan State University IACUC (protocol No. 202200211).
Study design
The study design was a 2-period, repeated-measures, crossover trial, allowing each horse to function as its own control both within and across the study periods to account for individual variation. The sequence in which the horses were assigned to the control (sedentary) or treatment (hand walking) group was provided by use of a random number generator (www.random.org; Randomness and Integrity Services Ltd). Eight horses were enrolled in this study. To randomize the study for temporal effects or acclimation to the experimental protocol, half (4 horses) were assigned to the exercise protocol and half (4 horses) were assigned to the sedentary protocol in the first period. A minimum of a 1-week washout was provided between study periods, in which the horses repeated the study in the alternate treatment group.
Fasting protocol
Horses selected for this study were on 24-hour turnout on grass pastures before enrollment. Prior to instrumentation and data collection, the horses were stalled in a 3 X 3-m box stall and fasted for 12 hours, with free access to water. Stall rest was used, in part, to alter gastrointestinal motility on the basis of a previous study24 that observed reduced intestinal contractions in horses placed on stall rest. The duration of feed restriction was extrapolated from published studies that utilized fasting as a model to induce mild ileus, with the reduction in motility noted in 1 study to be comparable to that produced by administration of α-2 agonists.25–27 Horses remained fasted during instrumentation and data collection.
Instrumentation
The horses were restrained in stocks or cross ties for instrumentation. The hair over the right flank and left ventral abdomen was removed with electric clippers and a No. 40 blade. Transabdominal ultrasonography was performed with a 3-MHz transducer, isopropyl alcohol, and ultrasonographic coupling gel to identify the location of the cecum and LVC in the right flank and left ventral abdomen, respectively.28 The skin was washed for 5 minutes with 4% chlorhexidine scrub, rinsed with tap water, and then towel dried.
A small amount of conducting gel (Spectra 360 electrode gel; Parker Laboratories Inc) was applied to foam adhesive unipolar electrodes (EL501; Biopac Systems Inc), and the electrodes were applied to the skin with the self-adhesive pads supplemented with cyanoacrylate glue. Sixteen unipolar electrodes were placed in 2 staggered vertical rows in the right flank along the long axis of the cecum, with a seventeenth electrode placed just cranial to serve as a ground. Similarly, 16 unipolar recording electrodes were placed on the left ventral abdomen in 2 staggered horizontal rows along the long axis of the LVC in a cranial-to-caudal direction. The distance between each electrode pair was 15 cm (Figure 1). Skin resistance between each electrode was measured to confirm impedance was < 3 milliohms (EL-Check; Biopac Systems Inc). The electrodes were allowed to equilibrate for a minimum of 15 minutes prior to data collection.
Placement of electrodes and unipolar leads on the skin over the cecum (A) and left ventral colon (B) after identification of the organs with ultrasonography. The leads were placed in 2 staggered rows along the long axis of the organs, with the ground (black lead) placed cranial to the active leads (red and white) in the right flank.
Citation: Journal of the American Veterinary Medical Association 2024; 10.2460/javma.24.07.0486
The pairs of leads (Lead 1108; Biopac Systems Inc) were then connected to each electrode and attached to the signal amplifier (EGG100C; Biopac Systems Inc) and data acquisition system (MP160; Biopac Systems Inc). The system was monitored by a computer running proprietary software for data collection and analysis (AcqKnowledge Software, version 5.0; Biopac Systems Inc). The hardware was set with a gain of 2,000, low-pass filter set of 1.0 Hz, and high-pass filter of 0.005 Hz. Data acquisition rate was 2,000 samples/s with an acquisition length of 30 minutes for baseline and post-treatment periods.
Data collection
A baseline 30-minute EIG (baseline) was recorded for both the control and hand walking treatment groups simultaneously from the cecum and LVC. Physical examinations, including heart rate, respiratory rate, rectal temperature, mucous membrane color, capillary refill time, gastrointestinal borborygmi, and digital pulses were recorded at 0, 15, and 30 minutes of each EIG recording. The borborygmi were auscultated with a stethoscope placed lateral to the electrodes over the left ventral flank to obtain the contraction rate for the LVC. The cecal borborygmi were auscultated in the right dorsal flank, cranial to the most proximal electrodes over the base of the cecum. Only short sounds (< 3-second duration) consistent with mixing contractions were counted and averaged per minute to obtain a rate for the cecal base and LVC.29 Borborygmi were auscultated by a single observer to reduce bias (JMRT). In addition, transabdominal ultrasonography was used to evaluate intestinal motility as a secondary determinant of changes in intestinal motility.19 Cecal contractions were identified as movement of the cecal wall away from the body wall, and LVC motility was determined by counting changes in the appearance of the sacculations.19 The number of contractions of the cecum and LVC were counted over a 3-minute period at 0, 15, and 30 minutes for each EIG recording by a single observer (ASM), and the rate in contractions per minute for each time point was determined.
Immediately following the baseline EIG recording, the horses in the hand walking treatment group were provided a brisk 15-minute walk (approx 3 to 4 miles/h; Strava app; Strava Inc) on a lead in hand. The speed was determined by the horse’s stride length but was kept at a pace just under that which prevented the horse from breaking into a trot. Horses in the sedentary control group were stall rested or confined in stocks for 15 minutes. The EIG leads were reattached, and recordings of the EIG and physical and ultrasonographic examinations were repeated immediately after the hand walk or sedentary period and again at 2 hours after the initial EIG. Between EIG sessions, the horses were confined to a box stall and fasted, with access to water. After completion of the final data collection, the electrodes were removed. The horses were returned to pasture and monitored for signs of abdominal discomfort for 24 hours.
Electrointestinography analysis
The waveform was attenuated by use of the high-pass (0.2-Hz) and low-pass (0.03-Hz) filters to remove frequency variations outside of the selected bandwidth. The raw EIG signals were resampled at 1 sample/s with the proprietary software. An infinite impulse response digital filter was used, with a variable Q setting of 0.707. After converting to a format for processing by the algorithm (MATLAB; The MathWorks Inc), data were cleaned with the codes ‘filloutliers’ and ‘nearest,’ which uses the nearest nonoutlier value. The code ‘mean’ removed outliers > 3 SDs from the mean. A fast Fourier transformation calculated the frequency components, with computation of a single-sided spectrum. After defining the total frequency domain (0.03 to 0.2 Hz; 1.8 to 12 cycles/min [cpm]), the DF, DP, and TP for each channel during the baseline and 2 post-treatment sessions were obtained. The DF represents the slow-wave frequency of the organ of interest, DP is the amplitude of the slow wave at the DF, and TP is the intensity of the waveform over the frequency range of interest.23,30 The DP was compared to baseline as a ratio to determine an increase (> 1) or decrease (< 1) in contractility.31 Ratios are used to evaluate changes in motility, as DP cannot be directly compared within or between horses due to differences in electrode position, body wall thickness, and intestinal motion.30,32,33
Additional power spectral density analysis determined the TP for specific frequency ranges (0.030 to 0.033 Hz [1.8 to 2 cpm]; 0.033 to 0.067 Hz [2 to 4 cpm]; 0.067 to 0.13 Hz [4 to 8 cpm]; and 0.13 to 0.2 Hz [8 to 12 cpm]). Percentage power distribution was then calculated with the use of the TP in the specific frequency ranges divided by the TP of the entire waveform.34 The expected frequency range for each organ was defined as a narrow range (2 to 4 cpm), and arrhythmias were defined as an increased percentage of TP distribution in bands outside of that range.
Statistical analysis
Analyses were conducted with the R Statistical language (version 4.3.2; R Core Team). Continuous variables were assessed for normality of the distribution with the Shapiro-Wilk test and visual examination of Q-Q plots and residuals (R packages ggplot and DHARMa). Data were described as median (IQR). The DF, DP ratios, and percentage TP distributions were determined from the power spectral density tracing, and the median was calculated for all leads for each organ at each time point. Comparisons were made to baseline for DF and DP ratios over time (baseline vs after intervention and at 2 hours after initial EIG) for both the control and hand walking groups for each organ. The DF and DP ratios for the cecum and colon were also compared between organs at each time point to identify differences between organs in the 2 treatment groups. A linear mixed-effect model was used for the EIG data (R packages lme4 and emmeans), with organ nested within horse as a random intercept to assess the effect of organ and time. A Friedman test was used for comparison between groups with a Nemenyi-Wilcoxon-Wilcox all-pairs test (R package PMCMRplus) with continuity correction applied for post hoc evaluation of significant differences. Correlation between borborygmi, ultrasonographic contractions, and DP were determined with a Spearman rank correlation (R package cor.test). Results were considered significant at P < .05.
Results
All horses completed the study without complications, and no horse demonstrated any sign of lameness at a walk. Data for 2 horses at 2 treatment time points (1 in the hand walking period and 1 in the control period) were lost due to file corruption, and these 2 sessions were not included in the results. Breeds represented included 3 American Quarter Horses, 2 draft crosses, and 1 each of Thoroughbred, Paint, and Standardbred. There were 7 mares and 1 gelding. The median age of the horses studied was 14 years (range, 13 to 21 years) with a median weight of 540 kg (range, 438 to 597 kg). Median ± IQR for heart rate (28.0 ± 5.5 beats/min), respiratory rate (12.0 ± 7.5 breaths/min) and temperature (37.6 ± 0.3 °C [99.6 ± 0.6 °F]) were within clinical reference ranges throughout the study. There was no difference in vital parameters between time points during the hand walking and sedentary periods (P > .06).
Electrointestinography analysis
Baseline DF in cycles per minute of the slow wave for the cecum (P = .486) and LVC (P = .395) in the sedentary control group showed no change compared to the EIG recording after intervention and at 2 hours (Table 1). The DF for the cecum (P = .883) and LVC (P = .884) in the hand walking group at baseline was also unchanged at both time points. Further comparisons with a linear mixed-effect model using horse as a random effect showed no effect of the treatments in either group, with or without hand walking (P > .128). The DF of the cecum was higher than the DF of the LVC at all time points in the hand walking group (P = .006) and in the sedentary controls (P = .019).
Median (IQR) dominant frequency (DF) in cycles per minute at baseline, after intervention (“post”; sedentary or hand walking), and at 2 hours after intervention (“2-hour”) obtained from the electrointestinography leads over the cecum and left ventral colon (LVC). The DF of the slow wave was unchanged over time for the cecum and LVC for both the sedentary and hand walking interventions. The DF of the cecum was higher than the DF of the LVC at all time points in both groups (P < .0191).
| Organ | Intervention | Time point | P value | ||
|---|---|---|---|---|---|
| Baseline | Post | 2-hour | |||
| Cecum | Sedentary | 2.53 (2.22–2.59) | 2.27 (2.12–2.49) | 2.30 (2.15–2.48) | .486 |
| Hand walking | 2.37 (2.22–3.14) | 2.32 (2.13–2.54) | 2.38 (2.18–2.48) | .833 | |
| LVC | Sedentary | 2.23 (2.00–2.40) | 2.26 (2.08–2.42) | 2.10 (2.01–2.14) | .395 |
| Hand walking | 2.15 (1.98–2.33) | 2.07 (2.01–2.18) | 2.12 (2.07–2.19) | .884 | |
The DP was determined from the power spectral density tracing and the median DP was calculated for the cecum and LVC for each horse and organ in each treatment group. The DP was assessed as a ratio for comparison of the treatment to baseline to account for the inherent variability for each horse, lead placement, and organ position.34 For the sedentary controls, baseline DP ratios for the cecum (P = .913) and LVC (P = .601) did not show a difference when compared to the EIG recorded immediately after treatment and at the 2-hour time point (P = .913; Table 2). In horses that were hand walked, there was no difference in DP ratios for the cecum (P = .808) and LVC (P = .957) comparing baseline, after hand walking, and at the 2-hour time point. After log transformation, comparisons with a linear mixed-effect model with horse as a random effect showed no effect of the treatments in either group, with or without hand walking (P > .466). The DP of the LVC was lower than the DP of the cecum at all time points in the hand walking group (P = .003), but no difference was observed between organs in the sedentary controls (P = .716).
Median (IQR) dominant power (DP) in millivolts at baseline, after intervention (“post”; sedentary or hand walking), and at 2 hours after intervention (“2-hour”) obtained from the electrointestinography leads over the cecum and LVC. The amplitude of the slow wave was unchanged over time for the cecum and LVC for both the sedentary and hand walking interventions (P > .466). The DP of the LVC was lower than the DP of the cecum at all time points in the hand walking group (P = .003), but no difference was observed between organs in the sedentary controls (P = .716).
| Organ | Intervention | Time point | P value | ||
|---|---|---|---|---|---|
| Baseline | Post | 2-hour | |||
| Cecum | Sedentary | 0.011 (0.007–0.014) | 0.010 (0.007–0.012) | 0.010 (0.009–0.010) | P = .913 |
| Hand walking | 0.009 (0.008–0.014)a | 0.011 (0.009–0.017)a | 0.011 (0.011–0.012)a | P = .601 | |
| LVC | Sedentary | 0.006 (0.005–0.008) | 0.008 (0.007–0.012) | 0.011 (0.006–0.017) | P = .808 |
| Hand walking | 0.007 (0.006–0.010)b | 0.006 (0.005–0.012)b | 0.008 (0.005–0.010)b | P = .957 | |
Different superscript letters between rows indicate significant differences in DP between organs in the hand walking treatment group (P < .05).
The power distribution in the EIG power spectral density tracing was calculated as a percentage by dividing the TP in each frequency range by the TP of the waveform. The frequency range from 2 to 4 cpm was defined for this study as the normal rhythm based on the observed dominant frequency. For both cecum and LVC in both treatment groups, this frequency range had the highest percentage of TP compared to frequency ranges above and below (P < .001; Figure 2). For horses in the sedentary treatment group, there was no change in power distribution for the cecum (P > .368) and LVC (P > .975) in all frequency ranges after intervention and at 2 hours after the baseline EIG. Horses in the hand walking group also showed no change in TP distribution for the cecum (P > .787) and LVC (P > .898) for each frequency range compared to baseline immediately after and at 2 hours after hand walking.
Box plots of the total power (TP) distributions in percentages. The TP distributions for the cecum in the sedentary (A) and hand walking (B) groups were unchanged at all time points (P > .368). The TP distributions for the left ventral colon were also unchanged compared to baseline for all time points in the sedentary (C) and hand walking (D) groups (P > .898). The normal frequency range was defined as 2 to 4 cycles/min (cpm).
Citation: Journal of the American Veterinary Medical Association 2024; 10.2460/javma.24.07.0486
Gastrointestinal borborygmi
Gastrointestinal borborygmi were recorded at all time points, and the median obtained for each time point (baseline, after, and at 2 hours). The number of contractions per minute auscultated for the cecum overall was 1.66 (IQR, 2.0), and the rate was not different across time and between treatments (P > .088; Table 3). The median number of contractions per minute auscultated for the LVC during the study was 1.33 (IQR, 0.67) and was also not affected by time or treatment (P > .29). Median borborygmi for each time point as borborygmi per minute were compared to ultrasonographic contraction rate and EIG DP.
Median (IQR) borborygmi in contractions per minute at baseline, after intervention (“post”; sedentary or hand walking), and at 2 hours after intervention (“2-hour”) auscultated over the cecum and LVC. No significant difference was observed for each time point compared to baseline (P > .088).
| Organ | Intervention | Time point | ||
|---|---|---|---|---|
| Baseline | Post | 2-hour | ||
| Cecum | Sedentary | 2.33 (1.67–2.83) | 2.67 (1.83–3.00) | 2.00 (1.33–2.33) |
| Hand walking | 1.33 (1.33–2.17) | 1.33 (1.17–2.00) | 1.33 (0.83–2.00) | |
| LVC | Sedentary | 1.33 (1.33–1.67) | 1.33 (1.00–1.83) | 1.33 (1.17–1.67) |
| Hand walking | 1.33 (1.00–1.50) | 1.00 (0.83–1.50) | 1.33 (1.00–1.50) | |
Ultrasonographic contraction rate
Ultrasonographic observations of cecal (2.33; IQR, 1.58) and LVC (1.00; IQR, 0.33) motility in contractions per minute were unchanged for both sedentary and hand walked horses at all time points (P > .119; Table 4). There was a significant relationship between auscultated borborygmi per minute and the number of cecal and LVC contractions measured with ultrasonography for both treatment and control groups (ρ = 0.629; P < .001). The EIG DP was not correlated with either borborygmi or ultrasonographic contractions (P > .2081).
Median (IQR) ultrasonographic contractions per minute at baseline, after intervention (“post”; sedentary or hand walking) and at 2 hours after intervention (“2-hour”) obtained over the cecum and LVC. The rate of contractions was unchanged for both sedentary and hand walked horses at all time points (P > .119).
| Organ | Intervention | Time point | ||
|---|---|---|---|---|
| Baseline | Post | 2-hour | ||
| Cecum | Sedentary | 2.33 (1.83–2.67) | 3.00 (2.17–3.17) | 2.67 (1.67–3.33) |
| Hand walking | 2.00 (1.67–2.67) | 1.67 (1.33–2.17) | 2.00 (1.33–2.67) | |
| LVC | Sedentary | 1.33 (1.00–1.33) | 1.00 (0.83–1.00) | 1.00 (0.67–1.00) |
| Hand walking | 0.67 (0.67–1.17) | 0.67 (0.67–1.00) | 1.00 (0.83–1.00) | |
Discussion
This study used multichannel EIG, ultrasonographic examinations, and auscultation of borborygmi to describe the effects of hand walking on the myoelectrical rhythms and motility of the cecum and LVC in healthy, fasted, stall-rested horses. The results of this study do not support the hypothesis that a brief episode of hand walking would increase gastrointestinal motility, as no differences were noted between horses that were hand walked and those that were sedentary on the basis of the measures of intestinal motility assessed in this study. The EIG DP, which has been noted to correlate with contractions of the intestine observed by ultrasonography in normal horses, did not correlate with contractions in the current investigation.19,35 However, it was observed that ultrasonographic evidence of contractility did correlate with auscultation in this study, indicating that these noninvasive methods performed similarly regarding identification of intestinal contractions.
Investigation of the effects of hand walking on intestinal motility patterns was motivated by the widespread belief that early ambulation can improve gastrointestinal function, increase intestinal motility, and reduce ileus. A recent survey of diplomates of the American College of Veterinary Surgeons, European College of Veterinary Surgeons, American College of Veterinary Emergency Critical Care, European College of Equine Internal Medicine, and American College of Veterinary Internal Medicine noted that while 53% of respondents believed that early mobility did not play a significant role in postoperative ileus, 47% of clinicians implemented early postoperative exercise as part of a multimodal approach for prevention of ileus.1,2 Interestingly, hand walking exercise was implemented by 84% to 86% of veterinary specialists to prevent and manage postoperative ileus, with more than half starting exercise within 24 hours of surgery.1,2 We have observed in our practice that hand walking is commonly prescribed for any horse with signs of abdominal discomfort, but the effects of exercise on intestinal motility and contractions had not been evaluated objectively. In humans, early ambulation has been called into question, as 1 study36 noted patients implanted with seromuscular electrodes after laparotomy for gastrointestinal disease demonstrated no change in slow-wave frequency or intestinal spike potentials in those provided early ambulation compared to a bed-rested control. In addition, a recent meta-analysis37 showed no evidence to support early ambulation after gastrointestinal surgeries in humans as a means to reduce morbidity or length of stay. The results of this research are consistent with the human literature and were unable to provide any evidence that motility patterns were altered by hand walking in normal fasted horses.
The lack of effect of exercise on intestinal motility in this study was an interesting finding, as to our knowledge the evaluation of the effects of hand walking exercise has not been assessed as an isolated factor affecting motility. A previous study26 assessing stall rest and a short period of fasting on ultrasonographic contractions reported a reduction in contractility of the cecum and ventral colon compared to baseline prior to the fast. An additional study38 that evaluated horses transitioned from pasture turnout management to stall rest noted ultrasonographic contractions of the cecum and ventral colon were also decreased; it was assumed that both alterations in feeding patterns and reduced activity were the cause of the motility changes. A separate report24 evaluated normal horses over 14 days after transition from pasture to stall rest with light exercise and found that motility was reduced on ultrasound on day 2 for the cecum and colon but only in the left ventral colon on day 4. Finally, in a report39 investigating healthy stall-rested ponies receiving once-daily hand walking exercise, gastrointestinal function was improved, noted by increased digestibility of dry matter and fiber when exercised at a walk compared to faster speeds. While it is unlikely that the single episode of hand walking used in this study would alter digestibility or transit time, a longer evaluation of horses with repeated, intermittent hand walking may show additional effects on overall transit time and gastrointestinal motility, which may not be assessed by slow-wave activity measured from a single treatment.
The frequency band for normal EIG slow-wave activity was defined as 2 to 4 cpm for both the cecum and colon based on the observed DF at baseline, with approximately 50% of the slow waves falling within this frequency range in the current study. Similar results have been reported in other studies evaluating EIG of the equine cecum and LVC.19,22,35 Hand walking had no effect on the percentage of time that slow waves were within this normal range, and arrhythmias were not noted in either the control or treatment group at any time point. It was interesting to note that fasting showed a reduction in the TP in the 2- to 4-cpm frequency bands compared to previous work in fed horses.35 Fasting was expected to reduce EIG amplitude, due to its relationship with contractions. The similar distributions over the EIG spectrum in both sedentary and hand walked horses indicate that the frequency distribution and rhythmicity were comparable to normal fed horses.35,40
The DP of the EIG slow wave was not affected by hand walking in this study in either the cecum or the LVC. The DP is the intensity and amplitude of the waveform at the DF defined by the Fourier analysis30 and has shown a correlation to contractions measured with implanted strain gauges.23 Ratios of DP have been noted to decrease in horses treated with medications that can reduce intestinal motility and should be expected to increase with appropriate stimulation.19,22,31,35 The observed lack of correlation of DP with intestinal contractions may be due to the fact that the EIG is a measure of slow waves over a large area of the organ, whereas the ultrasonographic contraction rate and borborygmi are localized to a more specific, focal region of the intestine. While other studies have identified a relationship between both TP or DP and observed ultrasonographic contractions, the correlations have been weak, and the lack of correlation in this study may relate to differences in EIG technique or individual horse variability.19,35
Auscultation correlated well across both treatments and time but did not show any effect of the interventions in either the treatment or control groups. Borborygmi were expected to decrease after fasting, and while other studies have assessed borborygmi by use of a grading scale, we elected to record a rate per minute to allow for direct comparison with ultrasonographic contractions.25,41 Agreement between observers and objective measurement of abdominal borborygmi intensity have shown poor correlation, and the use of additional observers for auscultation would likely not have improved the accuracy of the results, due to the low interobserver agreement noted in other studies.25,27 Although borborygmi were not altered by hand walking, auscultation may not be sensitive enough to identify changes in motility in normal horses. This is supported by other studies in the literature that have found no effect of lidocaine or bit-chewing on borborygmi compared to placebo controls.29,42,43
Ultrasound is commonly used as a noninvasive method to measure intestinal motility in experimental studies and clinical trials.24–26,29,38,42,44 The rate of intestinal contractions in the current investigation was similar to that of previous studies comparing fasted horses and stall rest, with the colon specifically noted to have a lower contraction rate than that of the cecum.24–26,38 The lower rate of contractility in the colon is regarded as a normal finding, proposed to be related to the digestive patterns of the horse that favor longer retention times for bacterial fermentation and water reabsorption.45 While hand walking was not shown to have an effect on ultrasonographic contractions in the current study, it is possible that ultrasound was not sensitive to changes in motility patterns, as the acoustic window can only view a small portion of the equine gastrointestinal tract. Alternatively, it is possible that hand walking may affect other segments of the intestines, or the experimental model was not effective in reducing motility patterns to a level that would demonstrate a measurable effect.
This study had several limitations. Observers were not blinded to the treatment when assessing borborygmi and ultrasonographic contractions, but the correlation of the observations between the 2 observers lends strength to the data obtained. The sample size, while adequate on the basis of power calculations, was small and may not reflect the population as a whole, especially those with naturally occurring gastrointestinal disease. The use of each horse as its own control for the EIG, combined with the crossover design, should have improved the robustness of the results. A larger study to monitor horses with gastrointestinal disease, as well as those fasted for surgical procedures, may provide a better design to demonstrate the true effect of hand walking on gastrointestinal contractions in a clinical setting. It would also be useful to evaluate motility prior to fasting and stall rest, to determine the effect of the model on intestinal motility at baseline. It has been reported that a 12-hour fast decreased gastrointestinal motility similar to the effects of xylazine administration; therefore, comparisons prior to baseline measurements would help to identify the magnitude of change after treatment.25 While the exercise provided to these horses simulated commonly used protocols for treatment of colic, the effects of longer periods of hand walking may provide different results. The addition of biomarkers of transit time, such as nondigestible beads, video endoscopic capsules, or radiopaque markers, could also provide additional evidence for or against the use of hand walking to improve intestinal motility.29,42,46
Additional limitations related to the EIG used in this study were characteristic of EIG recordings in general. The high signal-to-noise ratio in standing horses along with motion artifacts continue to be significant limitations to the interpretation and assessment of both equine and human EIG.47,48 Gastrointestinal organs with inherent rhythmicity, including the heart and diaphragm, can also interfere with EIG recording.49 The use of band pass filters as well as postprocessing algorithms improve the ability to isolate the EIG waveform from the organ of interest, but recording of physiologic myoelectrical activity in live animals is complex and makes EIG analysis less than straightforward.
In conclusion, this study does not provide evidence to support the mechanistic role of hand walking to improve motility patterns of the cecum and LVC in normal horses undergoing fasting and box stall rest. The results demonstrate correlation between auscultated borborygmi and intestinal contractions noted on ultrasound but no correlation with the EIG DP as a measure of global contractility. Further research is needed in clinical cases to confirm any benefits of hand walking, and the addition of other methods for monitoring motility and transit time may provide alternative methods to assess the effect of hand walking on equine gastrointestinal motility.
Acknowledgments
The authors would like to acknowledge Ms. Stesha Payne for her assistance with data collection and organization and Dr. Selin Aviyente for her role in the development of the algorithm for electrointestinography analysis.
Disclosures
The authors have nothing to disclose. No AI-assisted technologies were used in the generation of this manuscript.
Funding
Support for this investigation was provided by the Michigan Animal Health Foundation.
ORCID
A. S. Munsterman https://orcid.org/0000-0001-9922-2129
References
- 1.↑
Lefebvre D, Hudson NP, Elce YA, et al. Clinical features and management of equine post operative ileus (POI): survey of diplomates of the American Colleges of Veterinary Internal Medicine (ACVIM), Veterinary Surgeons (ACVS) and Veterinary Emergency and Critical Care (ACVECC). Equine Vet J. 2016;48(6):714-719. doi:10.1111/evj.12520
- 2.↑
Lefebvre D, Pirie RS, Handel IG, Tremaine WH, Hudson NP. Clinical features and management of equine post operative ileus: survey of diplomates of the European Colleges of Equine Internal Medicine (ECEIM) and Veterinary Surgeons (ECVS). Equine Vet J. 2016;48(2):182-187. doi:10.1111/evj.12355
- 3.↑
Svensson-Raskh A, Schandl AR, Ståhle A, Nygren-Bonnier M, Fagevik Olsén M. Mobilization started within 2 hours after abdominal surgery improves peripheral and arterial oxygenation: a single-center randomized controlled trial. Phys Ther. 2021;101(5):pzab094. doi:10.1093/ptj/pzab094
- 4.
Delaney CP, Zutshi M, Senagore AJ, Remzi FH, Hammel J, Fazio VW. Prospective, randomized, controlled trial between a pathway of controlled rehabilitation with early ambulation and diet and traditional postoperative care after laparotomy and intestinal resection. Dis Colon Rectum. 2003;46(7):851-859. doi:10.1007/s10350-004-6672-4
- 5.
Kibler VA, Hayes RM, Johnson DE, Anderson LW, Just SL, Wells NL. Cultivating quality: early postoperative ambulation: back to basics. Am J Nurs. 2012;112(4):63-69. doi:10.1097/01.NAJ.0000413460.45487.ea
- 6.
Kasalar I, Sarigol Ordin Y. Relation of early mobility with gastrointestinal symptoms and pain in patients undergoing abdominal surgery. J Perianesth Nursing. 2024;39(4):604-610. doi:10.1016/j.jopan.2023.11.001
- 7.↑
Kırtıl İ, Kanan N. Effect of early mobilization on gastrointestinal functions after abdominal surgery: systematic review. Arşiv Kaynak Tarama Dergisi. 2021;30:166-176. doi:10.17827/aktd.873681
- 8.↑
Fintl C, Hudson NP. The interstitial cells of Cajal of the equine gastrointestinal tract: what we know so far. Equine Vet J. 2010;42(4):372-377. doi:10.1111/j.2042-3306.2010.00073.x
- 9.↑
Farrugia G. Interstitial cells of Cajal in health and disease. Neurogastroenterol Motil. 2008;20(suppl 1):54-63. doi:10.1111/j.1365-2982.2008.01109.x
- 10.↑
DiBaise JK, Park FL, Lyden E, Brand RE, Brand RM. Effects of low doses of erythromycin on the 13C Spirulina platensis gastric emptying breath test and electrogastrogram: a controlled study in healthy volunteers. Am J Gastroenterol. 2001;96(7):2041-2050.
- 11.
Xu X, Wang Z, Hayes J, Chen JDZ. Is there a one-to-one correlation between gastric emptying of liquids and gastric myoelectrical or motor activity in dogs? Dig Dis Sci. 2002;47(2):365-372. doi:10.1023/A:1013726223531
- 12.
Camilleri M, Hasler WL, Parkman HP, Quigley EM, Soffer E. Measurement of gastrointestinal motility in the GI laboratory. Gastroenterology. 1998;115(3):747-762. doi:10.1016/S0016-5085(98)70155-6
- 13.↑
Bortoff A. Electrical transmission of slow waves from longitudinal to circular intestinal muscle. Am J Physiol. 1965;209(6):1254-1260. doi:10.1152/ajplegacy.1965.209.6.1254
- 14.↑
Hamilton JW, Bellahsene BE, Reichelderfer M, Webster JG, Bass P. Human electrogastrograms: comparison of surface and mucosal recordings. Dig Dis Sci. 1986;31(1):33-39. doi:10.1007/BF01347907
- 15.
Pezzolla F, Riezzo G, Maselli MA, Giorgio I. Electrical activity recorded from abdominal surface after gastrectomy or colectomy in humans. Gastroenterology. 1989;97(2):313-320. doi:10.1016/0016-5085(89)90066-8
- 16.
Familoni BO, Kingma YJ, Bowes KL. Study of transcutaneous and intraluminal measurement of gastric electrical activity in humans. Med Biol Eng Comput. 1987;25(4):397-402. doi:10.1007/BF02443360
- 17.↑
Chen J, McCallum RW. Response of the electric activity in the human stomach to water and a solid meal. Med Biol Eng Comput. 1991;29(4):351-357. doi:10.1007/BF02441654
- 18.↑
Ekstrand C, Michanek P, Gehring R, et al. Plasma atropine concentrations associated with decreased intestinal motility in horses. Front Vet Sci. 2022;9:951300. doi:10.3389/fvets.2022.951300
- 19.↑
Koenig JB, Martin CE, Nykamp SG, Mintchev MP. Use of multichannel electrointestinography for noninvasive assessment of myoelectrical activity in the cecum and large colon of horses. Am J Vet Res. 2008;69(6):709-715. doi:10.2460/ajvr.69.6.709
- 20.
Okamura K, Sasaki N, Kikuchi T, et al. Effects of mosapride on motility of the small intestine and caecum in normal horses after jejunocaecostomy. J Vet Sci. 2009;10(2):157-160. doi:10.4142/jvs.2009.10.2.157
- 21.↑
Okamura K, Sasaki N, Yamada M, Yamada H, Inokuma H. Effects of mosapride citrate, metoclopramide hydrochloride, lidocaine hydrochloride, and cisapride citrate on equine gastric emptying, small intestinal and caecal motility. Res Vet Sci. 2009;86(2):302-308. doi:10.1016/j.rvsc.2008.07.008
- 22.↑
Sasaki N, Murata A, Lee I, Yamada H. Evaluation of equine cecal motility by ausculation, ultrasonography and electrointestinography after jejunocecostomy. Res Vet Sci. 2008;84(2):305-310. doi:10.1016/j.rvsc.2007.04.009
- 23.↑
Smout AJ, van der Schee EJ, Grashuis JL. What is measured in electrogastrography? Dig Dis Sci. 1980;25(3):179-187. doi:10.1007/BF01308136
- 24.↑
Williams S, Horner J, Orton E, et al. Water intake, faecal output and intestinal motility in horses moved from pasture to a stabled management regime with controlled exercise. Equine Vet J. 2015;47(1):96-100. doi:10.1111/evj.12238
- 25.↑
VanderBroek AR, Reef VB, Aitken MR, Stefanovski D, Southwood LL. Assessing gastrointestinal motility in healthy horses comparing auscultation, ultrasonography and an acoustic gastrointestinal surveillance biosensor: a randomised, blinded, controlled crossover proof of principle study. Equine Vet J. 2019;51(2):246-251. doi:10.1111/evj.12990
- 26.↑
Mitchell CF, Malone ED, Sage AM, Niksich K. Evaluation of gastrointestinal activity patterns in healthy horses using B mode and Doppler ultrasonography. Can Vet J. 2005;46(2):134-140.
- 27.↑
Naylor JM, Poirier KL, Hamilton DL, Dowling PM. The effects of feeding and fasting on gastrointestinal sounds in adult horses. J Vet Intern Med. 2006;20(6):1408-1413. doi:10.1111/j.1939-1676.2006.tb00759.x
- 28.↑
Hendrickson EH, Malone ED, Sage AM. Identification of normal parameters for ultrasonographic examination of the equine large colon and cecum. Can Vet J. 2007;48(3):289-291.
- 29.↑
Patton ME, Leise BS, Baker RE, Andrews FM. The effects of bit chewing on borborygmi, duodenal motility, and gastrointestinal transit time in clinically normal horses. Vet Surg. 2022;51(1):88-96. doi:10.1111/vsu.13745
- 30.↑
Chen JD. Spectral analysis of electrogastrogram and its clinical significance. World J Gastroenterol. 1996;2(suppl 1):9-11. doi:10.3748/wjg.v2.iSuppl1.9
- 31.↑
Chen JD, McCallum RW. Electrogastrographic parameters and their clinical significance. In: Chen JD, McCallum RW, eds. Electrogastrography: Principles and Applications. Raven Press; 1994:45-73.
- 32.↑
Chang FY. Electrogastrography: basic knowledge, recording, processing and its clinical applications. J Gastroenterol Hepatol. 2005;20(4):502-516. doi:10.1111/j.1440-1746.2004.03751.x
- 33.↑
Levanon D, Zhang M, Orr WC, Chen JD. Effects of meal volume and composition on gastric myoelectrical activity. Am J Physiol. 1998;274(2):G430-G434. doi:10.1152/ajpgi.1998.274.2.G430
- 34.↑
Koch KL, Stewart WR, Stern RM. Effect of barium meals on gastric electromechanical activity in man: a fluoroscopic-electrogastrographic study. Dig Dis Sci. 1987;32(11):1217-1222. doi:10.1007/BF01296369
- 35.↑
Munsterman A, Moreira ASD, Kottwitz J. Evaluation of the effects of detomidine on equine myoelectrical activity using electrointestinography. J Vet Em Crit Care. Forthcoming.
- 36.↑
Waldhausen JH, Schirmer BD. The effect of ambulation on recovery from postoperative ileus. Ann Surg. 1990;212(6):671-677. doi:10.1097/00000658-199012000-00004
- 37.↑
Willner A, Teske C, Hackert T, Welsch T. Effects of early postoperative mobilization following gastrointestinal surgery: systematic review and meta-analysis. BJS Open. 2023;7(5):zrad102. doi:10.1093/bjsopen/zrad102
- 38.↑
Williams S, Tucker CA, Green MJ, Freeman SL. Investigation of the effect of pasture and stable management on large intestinal motility in the horse, measured using transcutaneous ultrasonography. Equine Vet J. 2011;43(39):93-97. doi:10.1111/j.2042-3306.2011.00399.x
- 39.↑
Schaafstra FJWC, van Doorn DA, Schonewille JT, et al. Effect of exercise on apparent total tract digestibility of nutrients and faecal recovery of ADL and TiO2 in ponies. Animal. 2018;12(12):2505-2510. doi:10.1017/S1751731118000356
- 40.↑
Sasaki N, Lee I, Ayukawa Y, Yamada H. Clinical applications of electrointestinography in the horse. J Equine Sci. 2004;15(4):85-92. doi:10.1294/jes.15.85
- 41.↑
Sasaki N, Mizuno Y, Yoshihara T. The application of electrocecography for evaluation of cecum motility in horses. J Vet Med Sci. 1998;60(11):1221-1226. doi:10.1292/jvms.60.1221
- 42.↑
Patton ME, Andrews FM, Bogers SH, et al. Effects of bit chewing on gastric emptying, small intestinal transit, and orocecal transit times in clinically normal horses. Animals (Basel). 2023;13(15):2518. doi:10.3390/ani13152518
- 43.↑
Rusiecki KE, Nieto JE, Puchalski SM, Snyder JR. Evaluation of continuous infusion of lidocaine on gastrointestinal tract function in normal horses. Vet Surg. 2008;37(6):564-570. doi:10.1111/j.1532-950X.2008.00421.x
- 44.↑
Gomaa N, Uhlig A, Schusser GF. Effect of Buscopan compositum on the motility of the duodenum, cecum and left ventral colon in healthy conscious horses. Berl Munch Tierarztl Wochenschr. 2011;124(3-4):168-174.
- 45.↑
Ross MW, Cullen KK, Rutkowski JA. Myoelectric activity of the ileum, cecum, and right ventral colon in ponies during interdigestive, nonfeeding, and digestive periods. Am J Vet Res. 1990;51(4):561-566. doi:10.2460/ajvr.1990.51.04.561
- 46.↑
Elfenbein JR, Robertson SA, MacKay RJ, KuKanich B, Sanchez L. Systemic and anti-nociceptive effects of prolonged lidocaine, ketamine, and butorphanol infusions alone and in combination in healthy horses. BMC Vet Res. 2014;10(suppl 1):S6. doi:10.1186/1746-6148-10-S1-S6
- 47.↑
Verhagen MA, Van Schelven LJ, Samsom M, Smout AJPM. Pitfalls in the analysis of electrogastrographic recordings. Gastroenterology. 1999;117(2):453-460. doi:10.1053/gast.1999.0029900453
- 48.↑
Parkman HP, Hasler WL, Barnett JL, Eaker EY; American Motility Society Clinical GI Motility Testing Task Force. Electrogastrography: a document prepared by the gastric section of the American Motility Society Clinical GI Motility Testing Task Force. Neurogastroenterol Motil. 2003;15(2):89-102. doi:10.1046/j.1365-2982.2003.00396.x
- 49.↑
Mintchev MP, Bowes KL. Impact of external factors on the stability of human electrogastrograms. Med Biol Eng Comput. 1996;34(3):270-272. doi:10.1007/BF02520087
