Influence of acute mental stress on the frequency and duration of gastroesophageal reflux in normal volunteers

Six fasted healthy male volunteers (mean age 21.7 years) were studied. A pH probe was inserted with the tip 5cm proximal to the gastroesophageal junction and recording begun 15 mins after completion of a standard meal. pH was recorded by computerized data logger which sampled every 6s. Subjects were stressed intermittently by a dichotic Listening task in which they attempted to track recorded word Usts heard simultaneously in different ears. Mean frequency and duration of reflux episodes was determined from two 30-min stress and control periods. Alterations in heart rate, systolic and diastolic blood pressure confirmed the effectiveness of the stressor. Mean ( ± SEM) duration of reflux per 30-min stress period ( l . l ±0.53 mins) was not significantly different to mean duration per 30-min control period (l.5 ±0.53 mins). The mean numberof reflux episodes per 30-min stress period (2.3 ± 1.1) did not differ significantly from the mean number of episodes per 30-min control period (2.8± 1.1). T herefore, in normal volunteers acute predictable emotio nal stress does not influence the frequency or duration of postprandial gastroesophageal reflux.

T HE GASTROINTESTINAL TRACT HAS long been recogn ized as an o rgan sensitive to emotio nal expression.The esophagus has been shown to take part in the reaction of the individual to physical as well as psychic stimul i. Emotional stress has long been recognized to precipitate spontaneous, non propulsive activity in the body o f the esophagus ( H) and increase the peristaltic amplitude of the distal esophagus (5).
Anecdotally, patient reports of dyspeptic symptoms and heartbu rn in response to stress are not infrequent.Up to 58% of referrals for esophageal manomctric stud ies have esophageal motility d isturbances (6).Fu rthermore, in patients with the irritable bowel synd rome OBS) (a condition with a high prevalence of stress related gastrointestinal symptoms), studies COOK AND COLLINS have reported between 37% and 56% to suffer from symptoms of gastrocsophageal reflux (7)(8)(9).As a group, !BS patients have significantly lower basal lower esophageal sphincter (LES) pressure (9,10) and abnormally frequent gastrocsophageal reflux judged by ambulatory pH monitoring (9).
The aim of this study was to determine in a~ympcomatic individuals, whether acute emotional stress influenced the frequency or duration of postprandial gastrocsophagcal reflux .
M ETHODS Six healthy male volunteers with mean age of21.7 years(SE = 0.56) and no h istory of gastrointestinal symptoms were ~tudied.Following an overnight fast, a pH probe (Spectrex, Massachusetts) was passed cransnasally without topical anesthesia and the rip positioned 5 cm proximal to the gastroesophageal junction.S ubjects were seated in the erect position and consumed a standard 1000 kcal meal in l 5 mins.The meal consisted of 114 g roast beef, two slices whole wheat bread, one pat butter, lettuce, o ne cup coffee with one creamer, salt and pepper, one milkshake (ice cream and l 42 g of 2% milk).Continuous mo nito ring of the esophageal pH was achieved with a computerized data logger (Conestogo Medical Electronics Ltd , Kitchener, Ontario).Esophageal pH sampling lasted 6 s.A reflux event was defined as a fall in pH of greater than one unit to a pH below 4, or a further fall from a level below 4 of at least one more pH unit.
The stressor used was a dichotic or dual listening task and has been described in detail elsewhere ( 1 1).Briefly, subjects attempted to track o ne of two different word lists heard simultaneously in different ears through stereo headpho nes.T he task was designed to simulate a stressful work environment in which the individ- ual is called upon to perform multiple tasks within a limited time while being distracted by noise.During control periods, subjects listened to relaxing music The effectiveness of the stressor was confirmed by monitoring extraintestinal mea sures of the stress response including heart rate, systolic and d iastolic blood pressure, each of which was measured every 5 mins.
Heart rate was measured by means of a 1-min ECG strip taken every 5 mins, and blood pressure was reco rded manually using a mercury sphygmomanometer.Expe rime ntal pro tocol: Each study compromised two 60-min sequences each consisting of a 30-min stress and a 30-min control period.The order of intervention (ie, stress-control or control-stress) was randomized.Subjects were located in a quiet room and no intrusions were permitted during the srudy.The first stress or con• trol period commenced 15 mins after com pletion of the meal.

RESULTS
Effectiveness of the stressor: Under the influence of the dichotic lis tening task, heart rate increased significantly (P< 0.001) by a mean of 6.12 ± 0.85 beats/ min.Significant elevation of systolic blood pressure (P <0.02) and diastolic blood pressure (P <O.O l ) were also observed in response to the stressor.Systolic blood pressure rose by a mean of 4. 7 ± l. 4 mmHg and diastolic blood pressure rose by 3.2±0.6mmHg (Table L, Figure 1).Neither sequence effect nor stress sequence interaction effects were significant, indicating no apparent adaptation to the srressor on second exposure.The mean number of reflux events per 30 mins during stress (2.33 ± l.12) did not differ significantly from the mean number of reflux events during control conditions (2.83± 1.13).Mean duration of reflux events per 30 mins under srress conditions ( l.08±0.53)did not differ significantly from that of control conditions (1.48 ± I.53)(Table 1, Figure 2).

DISCUSSION
On the basis of alterations in several different measures of the autonomic response it was demonstrated that the study population was stressed by the dichotic listening task.However, a significant influence of stress on the frequency or duration of gastroesophageal reflux events could not be demonstrated .
The lack of correlation between stress and reflux in asymptomatic volunteers may be explained by one or more theories.Symptoms attributed to the esophagus in response to srress are not related to gastroesophageal reflux.The potency of the stressor was insufficient to provoke reflux or the asymptomatic subjects studied were insufficiently susceptible to the degree of srress that might be expected to produce a resµonse in predisposed individuals.Multiple factors are operational in susceptible individuals before a particular stimulus can "trigger" reflux events.
In addressing these issues.o ne must consider that, while esophageal motility disturbances (rather than reflux events) might account for stress related "heartburn", there has been no clear and consistent demonsrration of manometric abnormalities concurrent with esophageal symptoms attributed to those abnormalities (12).With regard to the adeq uacy of the dichotic listening task in chis situation, comparable autonomic cardiovascu lar changes have previously been demonstrated together with pharyngoesophageal sphincter hyperconicity in normal subjects ( 13), and others have induced esophageal peristaltic abnormalities using a comparable stressor (5).However, the cask used is an acute and predictable srress and the present study may not be as readily extrapolated to the more chronic or unpredictable stress events of daily li fe.
The most feasible explanation for these findings is based on the heterogeneity of factors implicated in gastroesophagcal reflux.Gascroesophageal reflux in normal subjects occurs nearly exclusively by the mechanism of transient complete relaxation of the LES ( 14).In reflux patients.rransicnt LES relaxation or a persistently low basal LES pressure arc determinants of reflux ( 15,16).However, a low or absent LES pressure does not guarantee the occu rrence of gastroesophageal reflux .Other factors such as intragastric volume ( 17).intra-abdominal pressure ( 18).relaxation of the diaphragmatic hiatus ( 19) and possibly longitudinal esophageal shortening (20) appear to be important.It is possible that stress contributes to gastrocsophageal reflux but that several of the above criteria need to be operative before s tress can trigger a refl ux event.
Transient complete relaxations of the LES do not occur in asymptomatic su bjects during sleep but arc observed during arousa l from slee p or during full wakefullncss (14).Cervical vagotomy in dogs eliminates transient LES relaxation in response to gastric distension (21 ).Labyrinthine stimulation in human volunteers significantly lowers basa l LES pressure ( 11 ).These data strongly suggest that central nervous system arousal.through brain stem and vagal pathways, may be an important determinant of LES relaxation and hence gastroesophageal reflux.Furthermore, the repeated demonsrration of a high prevalence of heartburn (4. 7, 8 ), abnormal 24 hesophageal pH profilcs (9) and low basal LES pressure (9.10) in !BS patients suggests that susceptibility to stress might be a factor in the expression of their esophageal as well as intestinal symptoms.
Our inabil ity to demonstrate stress related reflux in asymptomatic volunteers docs not rrcclu<le the prospect of stress playing a rorential role in causing reflux in rredisposed individuals.As outlined, the interplay of factors resulting in gasrroesophage;il reflux is complex.For examrle. it is not known whether stress will rrecipimte reflux in individuals with low hasal LES rressurc (such as IBS or reflux ratientsl.Furthermore, it is suspected bur not known whether certain psychological Gastm-e,ophag<.!al reflux 1n the irritable hmvd ,yndmn1l' Gui 1986.271127 > I profiles might render an indiv1<lual more susceptible ro the effect:, of stress. Factors orerative in the perception of esophageal pam are equally poorly under-stoo<l and almost certainly depend on local as well as psychic facto rs (22.23) For example.stress and psychological profile may influence esophageal sensitivity or threshold to stimuli such as low pH or distension ( 21).The stress prone individ-ual might not reflux more frequently than normal.but might perceive each event more readily.
In spite of the apparent lack of correlation between stress and reflux in asymptomatic volunteers, the available evidence would suggest the investigation of the hypothesis in patient ropulations intuitively rredisposcd to stress and/or reflux is warranted.

TABLE 1
Influence of emotional stress on extraintestinal stress measures and postprandial gastroesophageal reflux events srress).All values are expressed as mean ±SEM.