Jan Tack, Translational Research Center for Gastrointestinal Disorders (TARGID), Department of Gastroenterology, University of Leuven, University Hospital Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium.
Menthol reduces intestinal motility in animal studies, an effect that is probably mediated by transient receptor potential channels. Peppermint oil (PO), with menthol as a major constituent, is widely used as a spasmolytic agent in irritable bowel syndrome. In the current study, we investigated the effect of acute PO administration on intragastric pressure (IGP) profiles and gastric sensorimotor functions in health.
Healthy volunteers underwent IGP measurement before and during continuous intragastric infusion of a nutrient drink (n = 13), and gastric barostat studies (n = 13). A single capsule of PO (182 mg) or placebo was administered during the studies in a randomized controlled crossover design. Throughout the studies, healthy volunteers scored 11 epigastric symptoms on a visual analogue scale (VAS); satiation was scored on a 6-point Likert scale during intragastric infusion.
During fasting, IGP and motility index (MI) of the proximal stomach decreased significantly after PO administration compared with placebo (P < 0.0001 and <0.05, respectively). In contrast, during intragastric infusion of the nutrient drink, no significant differences were detected between PO and placebo in IGP profiles, MI, satiation scores, and epigastric symptoms. The maximum infused volume, gastric compliance or sensitivity to balloon distention did not differ between both treatment arms. However, reduced appetite scores were seen during fasting after PO treatment, as compared with placebo (P = 0.01). Postprandial VAS scores were similar between PO and placebo.
Conclusions & Inferences
Peppermint oil reduces IGP, proximal phasic contractility, and appetite, with negligible effects on gastric sensitivity, tone, accommodation, and nutrient tolerance in health.
Peppermint (Mentha piperita L.) is a perennial herb endemic in Europe which has been used for centuries as a digestive aid in traditional medicine. Peppermint oil (PO) is extracted through distillation of various plant parts, and its main active constituent is menthol, among other volatile oils. In vitro animal studies have documented the inhibitory effects of menthol on gastrointestinal (GI) segmentation and motility.[2, 3] Based on recent clinical evidence, PO is widely being used as a spasmolytic agent in the treatment of the irritable bowel syndrome.[4-7]
Although the sites of action and pathways remain controversial,[3, 8, 9] the GI effects of menthol appear to be at least partly mediated by the melastatin8 (M8) member of the transient receptor potential (TRP) cation channel superfamily, which is also activated by cold temperatures and the cooling agent icilin. Transient receptor potential cation channels are increasingly being implicated in multiple cellular functions, including several regulatory aspects of the alimentary tract. Animal studies have detected TRPM8 channels in the dorsal root ganglia, vagal afferent neurons, the gastric fundus, colon, and small intestine.[2, 10-15]
Despite centuries of peppermint use and the recent insights in the TRPM8 channels, only few studies have examined the effects of TRPM8 ligands on GI sensorimotor functions. The effects of PO and menthol demonstrate significant species- and region-related variability.[15, 16] Furthermore, mechanistic studies in humans used combinations of several herbal extracts, rendering inferences about specific PO effects on gastric motor functions impossible.[17, 18] Finally, there is as yet no published data on the influence of PO on meal-induced accommodation, nutrient tolerance, and sensitivity to gastric distension.
Our group has recently implicated proximal gastric tone, through its effects on intragastric pressure (IGP), in the modulation of satiation during meal ingestion. Indeed, through a minimally invasive approach that involves the manometric measurement of IGP during intragastric infusion of a nutrient drink, we were able to demonstrate a close correlation between IGP and satiation changes. Nevertheless, it remains yet unclear how this modality compares to the gastric barostat, the current gold standard for the assessment of gastric accommodation.
Our aim was therefore to investigate the acute effect of an enteric-coated PO formulation on IGP profiles, gastric sensorimotor functions, and nutrient tolerance in health.
This study used a randomized, placebo-controlled, double-blind, crossover design in which healthy volunteers (HVs) participated in a gastric barostat study, and/or a protocol of IGP measurement during fasting and during intragastric nutrient infusion. Between consecutive visits, a minimum period of 1 week was respected. All procedures were approved by the Ethics Committee of the Leuven University Hospital, Belgium. Written, informed consent was obtained from all subjects.
A total of 18 HVs (nine men, age: 32 ± 2.1 years) with normal body mass index (23.1 ± 0.6 kg m−2) were recruited: six of them participated only in the barostat protocol, five underwent IGP measurement, while the remaining seven HVs participated in both protocols. In each protocol, volunteers always participated in the placebo and PO treatment arms, thus enabling pair-wise comparisons of all parameters. None of the HVs had symptoms or a history of gastrointestinal disease, significant medical conditions, psychological disorders or allergies; none were taking medications or had any drug history. Healthy volunteers were asked to present for the studies after an overnight fast, and refrain from alcohol, tea, coffee, and cigarette smoking 12 h before the experiments.
Intragastric pressure measurement during fasting and nutrient drink infusion
A high-resolution solid-state manometer with 36 sensors interspaced every 1 cm (Manoscan 360°; Sierra Scientific Instruments, Los Angeles, USA) was positioned through the nose, so that at least two proximal sensors were spanning the lower esophageal sphincter (LES). Intragastric pressure was measured as the average pressure of the first five channels distal to the LES (approximately 3–8 cm distal to the LES border). A second infusion catheter (Flocare; Nutricia, Bornem, Belgium) was passed through the mouth, and its tip was advanced 5 cm under the LES. The catheters were fixed to the subjects' chin and the HVs assumed a comfortable semirecumbent position with the knees slightly bent (80°) in a custom-designed bed. After a baseline recording of 20 min, a single capsule of PO (182 mg, Tempocol®; Will-Pharma, Wavre, Belgium) or a placebo capsule was administered with 50 mL water. Enteric-coated PO capsules are completely tasteless and odorless. This dose of PO has demonstrated clinical efficacy in IBS trials, and it is the commercially available dose of PO in Belgium. Sixty minutes thereafter a nutrient drink (13% proteins, 48% carbohydrates, 39% lipids; Nutridrink®, Nutricia) was infused in the stomach by a peristaltic pump at a constant rate of 60 mL min−1. Healthy volunteers scored their satiation at 1-min intervals, using a 0 to 5 rating scale combining verbal descriptors (1, threshold; 5, maximum satiation). Furthermore, the HVs scored every 5 min on a 10-cm horizontal visual analogue scale (VAS) their sensations of hunger, appetite, fullness, bloating, nausea, belching, epigastric-substernal burning and cramps, and pain. Intragastric infusion was stopped as soon as the HVs scored maximally on one of the symptoms or when maximum satiation was reached.
Barostat study protocol
Each HV underwent two isobaric distension sessions and two gastric accommodation sessions on separate visits, on which they were randomly assigned to PO or placebo treatment (in total four visits per volunteer). The protocol of these studies has been previously detailed. A double lumen polyvinyl tube (Salem sump tube 14 Fr.; Sherwood Medical, Petit Rechain, Belgium) with an adherent 1200-mL plastic bag was first introduced through the mouth into the gastric fundus, and connected to a computer-driven volume-displacement barostat device (Synectics Visceral Stimulator, Stockholm, Sweden). The gastric balloon was initially unfolded to an air volume of 300 mL at a rate of 1 mL min−1, while the HVs lay supine on a bed. Following balloon deflation, the HVs assumed the semirecumbent position. Thereafter, the minimal distending pressure (MDP) was determined as the lowest pressure level that provided an intraballoon volume of 30 mL during 1-mmHg increments.
Isobaric distension session
Stepwise distensions were performed in 2-min increments of 2 mmHg starting from MDP. Healthy volunteers scored their upper abdominal sensations at the end of each distending step, using a 0 to 6 graphic rating scale that combined verbal descriptors. Healthy volunteers simultaneously filled out the previously described VAS. The end point of each distension was established at an intraballoon volume of 1000 mL, or when discomfort (score 5) or pain (score 6) was reported. The intraballoon pressure level was then set at MDP +2 mmHg. Twenty minutes hereafter, a single capsule of PO or placebo was administered with 50 mL water. An identical distension series was repeated 60 min post medication.
Meal-induced accommodation session
Intraballoon pressure was set at MDP +2 mmHg. Twenty minutes hereafter, a single capsule of PO or placebo was administered with 50 mL water. Sixty minutes post treatment, the volunteers consumed a nutrient drink (200 mL; 300 kcal, Nutridrink®, Nutricia). Intraballoon volume measurements continued for 60 min. The HVs were asked to fill out the VAS for 11 epigastric symptoms at 5-min intervals throughout the session.
Evaluation of gastric tone after intraduodenal administration of PO
In four HVs (two females), a feeding polyvinyl tube (Flocare, Nutricia) was advanced transnasally into the second duodenal portion under fluoroscopic guidance. The barostat catheter was subsequently inserted and intraballoon pressure level was set at MDP +2 mmHg. Balloon volume was registered for 20 min before and 60 min after intraduodenal bolus administration of the content of a PO capsule. The purpose of this uncontrolled preliminary experiment was to assess whether the potential effect of PO on gastric tone varies according to route and site of administration.
Intragastric pressure study
The original data were exported from the recording software (Manoview Analysis, Sierra Scientific Instruments, Los Angeles, CA, USA) to Excel (Microsoft, New York, NY, USA). To minimize influences on IGP arising from movement, coughing, sneezing, or swallowing artifacts, a moving median transformation was applied to the original data, over a moving 1-min frame. A baseline value was then calculated per channel from the moving median data as the average pressure in the last 10 min of the stabilization period. Subsequently, the moving median values were corrected for the baseline value of the respective channel. Eventually, data were summarized as the average baseline-corrected values of the five selected channels below the LES, as described above.
During nutrient drink infusion, IGP generally demonstrates a consistent pattern, including a prompt pressure decrease shortly after the initiation of infusion, followed by a progressive pressure increase toward baseline levels. This profile was quantified using the following calculated parameters: relaxation amplitude (difference between average IGP during the last 10 min before the infusion and pressure nadir during the infusion); time to nadir pressure in s; pressure recovery slope in mmHg s−1 (slope of regression line between nadir pressure and termination of the infusion). In addition, the maximum infused volume and area under the satiation-time curve were determined. These parameters have been rigorously validated in HVs.[19-21]
The phasic contractile activity of the fundus was compared between treatment arms during the baseline and postmedication periods: this was accomplished by a baseline IGP reconstruction, after filtering out respiratory artifacts by means of a computerized algorithm.[22, 23] A motility index (MI) was calculated per 5-min interval as the natural logarithm of the area between the IGP signal and the baseline.
During the isobaric distensions, the average balloon volume was determined for each distending step. Perception (discomfort) thresholds were, respectively, defined as the lowest levels of pressure and the corresponding volume that evoked a perception score of 1 (5) or more. Gastric compliance was calculated as the slope and y-intercept of the pressure–volume curve obtained from stepwise distensions, using linear regression analysis. Analysis was limited to the pressure range for which a value was obtained in over 75% of the subjects.
To evaluate gastric tone and accommodation, the mean intraballoon volume was calculated over consecutive 5-min intervals. The first four (20 min), the next 12 intervals (60 min), and the final 12 intervals constitute the baseline, preprandial, and postprandial periods, respectively. Treatment-induced changes in gastric tone were quantified as the difference between the average intragastric volume during the baseline and preprandial periods. Gastric accommodation was expressed as the difference between the mean postprandial and mean preprandial volumes. Consecutively, a MI was calculated per 5-min interval as the natural logarithm of the area between the volume signal and the baseline normalized over time.
Treatment effects on epigastric symptoms were evaluated by calculating separate cumulative VAS scores for each specific symptom during the isobaric distensions (as area under the pressure–perception curve), and during the baseline, preprandial, and postprandial periods, respectively (as area under the time–perception curve).
Data are presented as the mean ± standard error of the mean (SEM). Paired t-tests were used to compare mean parameter values between treatment arms. Treatment effects over time on gastric tone, accommodation, MI, and IGP were evaluated separately in the baseline, preprandial, and postprandial periods (where applicable) using a two-way analysis of variance (anova) for repeated measurements. Differences were considered significant at the 5% level.
Effect of PO on IGP, phasic gastric contractility, and nutrient tolerance
During fasting, IGP and proximal gastric MI significantly decreased after PO administration as compared with placebo (repeated measures anova P = 0.005 and P = 0.047, respectively, Figs 1 and 2).
During intragastric infusion of the nutrient drink, no differences occurred between PO and placebo in relaxation amplitude (3.5 ± 0.7 vs 3.8 ± 0.6 mmHg, P = 0.66), time to nadir pressure (4.5 ± 0.7 vs 4.3 ± 0.6 min, P = 0.84), and IGP recovery slope (0.33 ± 0.07 vs 0.27 ± 0.07 mmHg min−1, P = 0.448).
The maximum tolerated nutrient volume was not significantly different between PO and placebo (704 ± 48 vs 761 ± 34 mL, P = 0.121, Fig. 3A). Accordingly, the area under the time-satiation curve did not differ between PO and placebo during the first 10 min of the infusion (117.7 ± 10.1 vs 99.1 ± 10.2, P = 0.124, Fig. 3B). In addition, VAS scores were comparable between placebo and PO for each of the 11 individual symptoms (data not shown).
Effect of PO on gastric sensitivity and compliance during isobaric balloon distensions
The mean MDP was comparable between the PO and placebo arms (7.9 ± 0.6 vs 7.4 ± 0.3 mmHg, P = 0.235). The pressures needed to induce first perception of distension or discomfort, and the corresponding intraballoon volumes were not significantly different (Table 1). Progressively higher distending pressures produced progressively larger intraballoon volumes (Fig. 4A). No significant differences were demonstrated between treatment arms in the slope or intercept of the pressure–volume curves, either pre- or post medication (Table 1).
Table 1. Comparison of gastric sensitivity and compliance between the placebo and peppermint oil treatment arms
Threshold pressure, mmHg
11.4 ± 0.5
11.2 ± 0.9
Discomfort pressure, mmHg
18.8 ± 0.5
21.8 ± 3.3
Threshold volume, mL
183.3 ± 25.1
145.9 ± 27.5
Discomfort volume, mL
558.6 ± 49.8
500.0 ± 52.5
Compliance, mL mmHg−1
49.6 ± 4.8
48.1 ± 7.7
8.1 ± 12.4
1.8 ± 15.1
Threshold pressure, mmHg
10.6 ± 0.4
11.5 ± 0.8
Discomfort pressure, mmHg
17.3 ± 0.8
16.8 ± 0.9
Threshold volume, mL
237.6 ± 32.8
311.8 ± 51.6
Discomfort volume, mL
564.6 ± 34.7
538.9 ± 46.4
Compliance, mL mmHg−1
51.6 ± 4.3
42.0 ± 5.6
78.5 ± 19.9
112.4 ± 39.4
The area under the pressure–perception curves did not differ between PO and placebo during the baseline (4.5 ± 0.6 vs 4.3 ± 0.4 min, P = 0.815) and postmedication distensions (4.0 ± 0.4 vs 4.5 ± 0.3 min, P = 0.385, Fig. 4B). Accordingly, VAS scores were comparable for all 11 epigastric symptoms (Fig. 5A).
Effect of PO on gastric tone, accommodation, and epigastric symptoms
Ingestion of the nutrient drink caused an immediate relaxation of the proximal stomach, reflected by intraballoon volume increase (Fig. 6). Volume change over time was not significantly different between the two study arms during the baseline, postmedication, and postprandial periods (repeated measures anova P = 0.755, 0.999 and 0.999, respectively, Fig. 6). Similarly, gastric accommodation was similar after PO and placebo (180.5 ± 38.6 mL vs 155.4 ± 40.2 mL; t-test P = 0.751).
Moreover, there was no significant difference in barostat MI after PO compared with placebo during the fasting period (mean MI 32.7 ± 4.2 vs 27.0 ± 3.2, repeated measures anova P = 0.998) and the postprandial period (mean MI 22.5 ± 2.22 vs 21.2 2.7, repeated measures anova P = 0.673).
Cumulative fasting appetite scores were significantly lower after PO as compared with placebo treatment (P = 0.017, Fig. 5B). In contrast, VAS scores for the remaining symptoms were comparable during the baseline and postprandial periods.
Effect of intraduodenal administration of PO on fasting gastric volume
Direct intraduodenal administration of PO did not alter balloon volume over time as compared with baseline preadministration values (mean pre- and postadministration volume: 154.8 ± 24.2 mL and 133.9 ± 15.5 mL, respectively, repeated measures anova P = 0.142).
This study explored the acute effects of an enteric-coated PO formulation on IGP, gastric sensorimotor functions and nutrient tolerance, in normal-weight healthy adults. Our results demonstrate significant effects of PO during fasting namely a reduction in IGP levels, phasic contractility, and appetite scores. Peppermint oil did not influence gastric sensitivity/compliance during isobaric distensions. We observed negligible effects of PO on gastric accommodation to a nutrient drink and on IGP profiles during intragastric infusion. Peppermint oil did not affect satiation and epigastric symptoms during the infusion or the liquid meal.
The observed clear-cut effect of PO on fasting IGP could be interpreted as a result of sustained reduction in gastric tone. This is in concordance with studies that demonstrated menthol-induced relaxation of intestinal and colonic muscle strips.[2, 3] Also, relaxation of guinea-pig fundic strips and of the proximal stomach in man were induced by a combination of multiple herbal extracts including PO, although the contribution of PO to these effects remains unclear. Transient receptor potential M8 superfamily channel activation on the other hand has been implicated as the underlying mechanism of cooling-induced contractions in the rat gastric fundus.[15, 16] Adding to the controversy, the barostat data did not indicate any effect of PO on proximal gastric tone. In support and further confirmation of this notion comes the fact that gastric volumes were not significantly altered after intraduodenal administration of PO in a subset of our HVs; thus, a potential interference of the balloon with capsule emptying in the duodenum and PO availability is not a plausible explanation for the lack of effect on proximal gastric tone as measured with the barostat.
Based on our data, the reduction in fasting IGP should thus be attributed to a simultaneous reduction in gastric phasic contractility after PO administration, reflected in decreased MI of the proximal stomach. This finding is in concordance with preliminary or indirect evidence from previous studies namely reduced contraction duration and amplitude in the gastric corpus after direct intraduodenal administration of PO, and following a single oral dose of a herbal combination containing PO. Spraying of L-menthol on the gastric mucosa yielded similar findings, in a controlled trial reporting suppression of endoscopically assessed gastric peristalsis. The temporal correlation between the reduction of IGP and gastric MI further supports our hypothesis. The fact that changes in phasic contractility were shown with IGP measurement, but not with barostat measurements, suggests that IGP monitoring may have superior sensitivity to assess phasic motor events in the proximal stomach. We assume that during fasting conditions the balloon may be influenced by antral activity due to the small intragastric volume, thus masking a potential effect of PO on fundic MI. This is in line with the intended design of the barostat which mainly serves to quantify changes in tone.
Our study was not designed to investigate the receptors and pathways mediating the effects of PO in gastric motility. There is some experimental evidence to support a direct action on gastric smooth muscle, involving reduction in calcium influx. On the other hand, although TRPM8 channel expression has been demonstrated in the fundic muscular layer of certain species, it is still controversial whether menthol affects motility through this receptor.[15, 26] Moreover, most of this evidence comes from studies on isolated muscle strips. Thus, given the expression of TRPM8 channels in vagal afferents projecting to the upper GI, and the well-known dependence of gastric phasic contractility on the vagus nerve, the observed effects may be initiated by TRPM8-receptor coupling in duodenal vagal afferents and mediated through a vagal circuit. Evidently, these hypotheses would require further study employing selective TRPM8-channel inhibition at different levels. These studies would ideally investigate any further effects of TRPM8 ligands on the intestinal peptides that influence gastric contractility and the gastric migratory motor complex (MMC) namely motilin and ghrelin.[28, 29]
Also, while most studies focus on TRPM8 as the primary molecular target for menthol, TRPM8-deficient mice still exhibit a considerable amount of menthol-sensing dorsal root and trigeminal neurons.[30, 31] However, Karashima et al. provided evidence that another member of the TRP superfamily, TRPA1, is also sensitive to menthol, at least at submicromolar to low micromolar concentrations. As TRPA1 is widely expressed throughout the gut and is shown to control visceral sensation and GI motility,[12, 33, 34] it is also possible that some of the effects observed in this study are mediated by TRPA1.
On the other hand, our data argue against a significant influence of PO on proximal gastric sensitivity to distension, as quantified by perception and discomfort thresholds, by area under the pressure–perception curve, and by VAS scoring. It is important to note that these findings do not necessarily exclude efficacy of PO in dyspeptic patients who are hypersensitive to mechanical stimuli; this hypothesis, warrants testing in future clinical studies.
Before this study, the effect of PO on meal-induced accommodation had not been directly addressed. The accommodation response involves relaxation of the proximal gastric muscular layer upon initiation of a meal, and is mediated through a vago-vagal reflex which is integrated in the brainstem. Fundic relaxation is effected primarily through inhibitory nitrergic pathways, although inhibition of cholinergic pathways may also contribute to the reflex. In this study, we did not observe any effects on IGP profiles or in accommodation parameters during a nutrient drink challenge. Although our study was powered to detect a 30% change in postprandial balloon volumes, and clinically important differences in gastric accommodation were previously detected with similar samples,[37, 38] we cannot exclude the possibility that a type II error may have masked a modest effect. Moreover, the omission of a dose–response study of PO represents a potential limitation of our data. On the other hand, there are several lines of evidence that indirectly support our findings. First, animal studies have previously established that menthol-induced relaxation of gastrointestinal muscle strips is independent of cholinergic antagonism or nitrergic pathways.[3, 8, 9, 16] Second, a randomized controlled trial found no significant effect of a 90 mg PO capsule on the gastric emptying rate of liquids, suggesting lack of a robust effect on fundic tone in response to a liquid meal. Similarly, a PO-containing formula minimally increased the retention of liquids between 10 and 50 min postprandially compared with placebo, while no differences were found in the lag phase, T1/2 and AUC for liquid emptying. Third, unaffected satiation/symptom scores and IGP profiles during intragastric infusion in this study also argue against a significant effect of PO on the gastric accommodation reflex.
To the best of our knowledge, this is the first report of an effect of PO on the control of food intake in humans. In our hands PO reduced fasting appetite. Relevant animal data are inconclusive, as a single study failed to detect differences in feeding behavior or weight gain between rabbits fed with peppermint and other herbs. The mechanism of the appetite effect cannot be safely inferred based on our data. It is generally accepted that enteric peptide hormones released during nutrient exposure mediate satiety during the late postprandial period, via local vagal stimulation or through hypothalamic signaling. Given the long pretest fasting period of our protocol, enteric peptides are unlikely to mediate the appetite effect of PO, which does not refer to satiety, but rather to the return of hunger pangs during prolonged fasting. The concomitant reduction of appetite scores and MI, however, raise the question whether decreased appetite is mediated through a reduction in phasic gastric contractility. In this perspective, our group has recently presented preliminary data that generation of hunger pangs requires motilin receptor stimulation and/or a typical phase-III pattern of the MMC. In addition, we found significant correlations between motilin levels, MI during phase III and hunger scores. Whether the specific trigger of hunger pangs in humans is the rise of motilin levels or the emergence of gastric phase-III activity per se still remains to be determined. Therefore, future studies of PO should evaluate motilin levels in addition to gastric motility and hunger ratings.
In conclusion, in this study acute oral administration of PO did not affect gastric sensitivity, compliance, or accommodation reflex to a liquid drink in health. Satiation and meal-related symptoms were also unaffected. During fasting, PO reduced IGP, phasic contractility of the proximal stomach, and appetite ratings. Further studies are warranted to refine the role of TRPM8 channel activation and antagonism on gastric motility, control of food intake and determine the underlying mechanisms.
Dr Papathanasopoulos was supported by an international scholarship from the Hellenic Society of Gastroenterology. Werend Boesmans and Ricard Farré are postdoctoral fellows of the FWO, Belgium.
This work was supported by an FWO grant and a Methusalem grant to Jan Tack, M.D., Ph.D.
None of the authors has a competing interest.
AP & AR performed the study; AP analyzed the data and wrote the manuscript; AP, JT, and PJ designed the study and reviewed the manuscript; WB, RF, & PVB reviewed the manuscript and contributed to the manuscript writing.