| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Naval Aerospace Medical Research Laboratory, Pensacola, Florida (E.R.M.); Departments of Medicine (K.L.K.) and Psychology (R.M.S.), Pennsylvania State University; and Department of Psychology, University of Missouri, Columbia, Missouri (J.F.T.).
Address reprint requests to: Eric R. Muth, PhD, Naval Aerospace Medical Research Laboratory, 51 Hovey Road, Pensacola, FL 32508-1046. Email: emuth{at}namrl.navy.mil
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: Electrogastrograms (EGGs) were recorded, and interbeat intervals (IBIs) were obtained from electrocardiographic recordings from 20 subjects during baseline and in response to a shock avoidance task (shock stimulus) and forehead cooling (dive stimulus). After each experimental period, subjects reported their emotional experience by rating descriptors ranging from serenity to excitement.
RESULTS: During the shock stimulus, IBIs decreased significantly (p < .05), gastric tachyarrhythmias increased (p < .05), and emotional arousal increased, as indexed by reports of increased interest, excitement, and activation. In contrast, during the dive stimulus, IBIs increased (p < .05), but there were no associated changes in gastric myoelectrical activity or emotional arousal.
CONCLUSIONS: Acute stress can evoke arousal and dysrhythmic gastric myoelectrical activity, and these acute changes, which occur in healthy individuals, may provide insight into functional gastrointestinal disorders.
Key Words: stress • autonomicnervous system • gastric myoelectrical activity • emotional responses • gastrointestinal system • electrogastrogram
Abbreviations: SNS = sympathetic nervous system; PNS = parasympatheticnervous system; GI = gastrointestinal; cpm = cycles perminute; ANS = autonomic nervous system; EGG =electrogastrogram; LF = low frequency; MF = mid frequency; HF = high frequency; ECG = electrocardiogram; IBI =interbeat interval.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As early as 1947, researchers and clinicians were curious about the relationship between stress and gastric motility. Wolf and Wolff (1) observed that Tom, a fistulated patient, showed increased gastric motility when he was angry and decreased gastric motility when he was fearful. Since Wolf and Wolff’s work, the role of stress in gastric motility has not been closely examined. The lack of research in this area can be explained by several factors. First, most tests of GI function (eg, endoscopy, colonoscopy, and pH/ manometry studies) are invasive and stressful. Hence, the tests themselves confound study results. Second, it is difficult to create a salient behavioral stress task in the laboratory.
Studies that have investigated the role of stress in GI function have investigated both sustained and acute stress. Motion sickness and long duration of exposure to cold can both be considered sustained stresses. Motion sickness has been shown to cause delayed oral-cecal transit times (2), decreased normal 3-cpm activity (3), and increased 4- to 9-cpm gastric tachyarrhythmia (4). Sustained cold stress has been shown to reduce gastric motor activity (5) and to decrease normal gastric myoelectrical activity (6). It has been shown that acute stress inhibits postcibal antral contractile activity and delays gastric emptying (7, 8). These responses to acute stress have been reproduced for sustained stress (9, 10). The effects of acute stress on gastric myoelectrical activity have not been examined.
Interest in the role daily life stress plays in gastric motility is growing as the mind-body interaction moves to the forefront in the struggle to diagnose and treat difficult functional GI disorders (11–13). It has been suggested that the ANS may provide a pathway for stress to affect the GI system (9). The development of noninvasive studies of GI and ANS function, specifically the EGG and spectral decomposition of cardiac variability, have opened a new window to study the role of stress in GI function. Furthermore, salient laboratory tasks have been shown to evoke changes in ANS activity using noninvasive measures (14, 15). These tasks may provide a link between the ANS, GI activity, and stress.
The EGG is a reliable, noninvasive method of recording gastric myoelectrical activity. It is recorded by placing electrodes on the surface of the skin over the antrum of the stomach. The frequencies recorded by the EGG are identical to the frequencies of electrical activity when recorded in or directly on the stomach and the frequency of stomach contractions when they occur. As contractile activity in the stomach increases, the amplitude of the normal 3-cpm EGG rhythm increases (16). Normal 3-cpm activity can be replaced by erratic electrical waves, indicating dysrhythmias, which occur at frequencies of 0.25 to 2.25 cpm (bradygastria) and 4.0 to 9.75 cpm (tachyarrhythmia). These dysrhythmias disrupt 3-cpm activity and consequently reduce or abolish the contractile activity of the stomach (17). The significance of bradygastric activity is not well understood. Tachyarrhythmic activity (4.0–9.75 cpm), on the other hand, has been associated with the stress of motion sickness (18).
When cardiac variability, the majority of which occurs between 0.0 and 1.0 Hz (19), is plotted as a continuous function against time, three periodic fluctuations can be observed. These include a LF peak between 0.04 and 0.08 Hz, a MF peak centered at 0.1 Hz, and a HF peak between 0.15 and 0.5 Hz (20). The physiological basis for the HF component, or respiratory sinus arrhythmia, is well known, and many studies have validated the use of various respiratory sinus arrhythmia measures as indices of PNS activity (eg, 21). However, the physiological basis of the other periodicities in cardiac variability is controversial. It is thought that the LF wave is associated with thermoregulatory and peripheral vascular sympathetic influences. The degree to which variations in the LF wave are associated with either of these mechanisms has yet to be established. It seems that the LF wave may be mediated by both branches of the ANS (20). It has been proposed that the MF component reflects either an autonomous central nervous system rhythmicity or characteristics of the blood pressure (baroreceptor) control system and is most likely influenced by various efferent and afferent factors from both branches of the ANS (20). Thus, in terms of ANS activity, the spectrum can be divided into two ranges where everything from 0 to 0.15 Hz is influenced by a mix of PNS and SNS activity and where everything above 0.15 Hz is mainly influenced by PNS activity. Shifts in ANS activity away from PNS activity and toward SNS activity have been associated with increases in tachyarrhythmia during the stress of motion sickness (3, 4).
Threat of shock reliably elicits sympathetic ß-adrenergic cardiac responses (14, 15). Sympathetic activity has been associated with increased 4- to 9-cpm tachyarrhythmia in the EGG (4). Hence, it was hypothesized that a shock-avoidance task would cause increased emotional arousal and tachyarrhythmia. On the other hand, immersion of the face in cold water has been associated with parasympathetic activity. Eating has been associated with increases in PNS activity and normal 3-cpm activity in the EGG (3). Hence, it was hypothesized that the cold-face task would cause emotional relaxation and increased 3-cpm activity.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Apparatus
Electrogastrography.
EGGs were recorded using three disposable Silvon cutaneous electrodes (catalog no. 01–3530, NDM, Dayton, OH). The two active electrodes were attached to Fetrodes (UFI Corp., Morro Bay, CA), and the third was attached to a standard reference lead. Fetrodes (field effect transistors) provide an active, amplified circuit at the recording site that increases the signal-to-noise ratio and helps minimize recording artifacts such as subject movement. The EGG signal went to a Fetrode bioamplifier (model 2121FT) and then was sent through an EGG filter (model FEGGF) with a 3-dB roll-off at 0.16 Hz (9.6 cpm). The EGG filter was connected to a Gould recorder (model 3000, Cleveland, OH) equipped with a modified Universal coupler with a bandwidth of 0.008 to 0.3 Hz. Thus, the overall bandwidth of the system was 0.008 to 0.16 Hz. EGGs were digitized using an IBM-compatible PC and a MetraByte Dash-16 analogue to digital board at a rate of 4.267 Hz.
Electrocardiography.
ECG signals were recorded using Ag-AgCl electrodes (UFI). ECG signals were amplified using Grass 7P3 preamplifiers and 7DA amplifiers (Grass Instruments, Inc., Quincy, MA). ECGs were monitored online, and cardiac IBIs were stored using a Swan 386/20D PC and specialized software (22).
Emotion Reports.
Subjective reports of emotional experience were obtained using 13 descriptors: serenity, interest, relaxation, excitement, happiness, agitation, anger, sadness, tiredness, like, pleasantness, activation, and pain. Subjects rated their emotional experience during each task (described below) on a scale of 1 to 5 for each subjective descriptor, with 1 being "I did not feel any" and 5 being "I felt a lot." These descriptors were chosen to represent a two-dimensional valence/arousal model of emotion. In this model, descriptors represent four emotional quadrants: positive valence/high arousal; positive valence/low arousal, negative valence/high arousal; and negative valence/low arousal (23). Of the 13 descriptors measured, happiness, agitation, anger, sadness, like, and pleasantness are used to index valence. Serenity, interest, relaxation, excitement, tiredness, and activation are used to index arousal. Pain loads onto both the valence and arousal dimensions.
Respiratory Recordings.
Respiration was recorded using a mercury strain gauge placed around the subject’s chest. The strain gauge was connected to a bridge circuit (model 217, Parks Medical Electronics, Inc., Aloha, OR) and then to the Gould 3000 recorder equipped with a Universal coupler with a direct current setting. Respiration was recorded so that any portions of the EGG recordings that were contaminated with respiratory artifacts could be identified and deleted and to ensure that subjects were pacing their breathing as required.
Psychophysiological Stimuli
Reaction-Time/Shock-Avoidance Protocol (Shock Stimulus).
Threat of shock reliably elicits sympathetic ß-adrenergic cardiac responses (24, 25). Therefore, a reaction-time protocol was linked to a shock-avoidance stimulus to evoke SNS activity. The reaction-time protocol involved pressing a button as quickly as possible in response to a brief visual presentation of the word "go," which appeared in the center of the monitor screen with a concurrent audio tone. This sequence was followed by a random time interval during which the word "ready" was displayed on the monitor. The reaction-time stimulus was generated on an Apple IIE computer and displayed on an Apple monitor. Subjects were threatened with shock if their reaction time exceeded a certain level. Hence, the reaction-time/shock-avoidance (shock) stimulus is a psychophysiological task that yields both psychological and physiological changes due to the task and the threat of shock.
Shocks were administered using a Mark I Behavior Modifier (Farrall Instruments, Grand Island, NE), which produced voltages that can be varied in 10 steps of intensity from 0 to 0.66 V (alternating current). Before the experiment began, subjects set their own shock level by indicating to the experimenter the highest tolerable degree of shock as the level was raised from zero in quarter-step increments. Reaction-time feedback was given in milliseconds after each trial. Subjects were told that the reaction time required to prevent a shock would vary randomly from presentation to presentation. Thus, a subject could not necessarily gauge how fast reactions needed to be to avoid a shock based on reaction-time feedback. Subjects were therefore encouraged to react as quickly as possible on each trial to ensure avoidance of a shock. In actuality, only one shock was administered during the shock stimulus, after the first response in the third minute of the 4-minute shock stimulus. To maintain salience of the threat of shock, one shock was administered regardless of the reaction time. This stimulus has been shown to evoke SNS activity, as reflected in decreased IBIs and changing spectral decomposition of cardiac variability (14, 15).
Forehead Cooling Protocol (Dive Stimulus).
Immersion of the face in cold water while holding the breath produces a dive reflex. The dive reflex is characterized by increased PNS, bradycardia, and peripheral vasoconstriction (26, 27). The dive reflex can be simulated by holding a bag filled with cool water on the forehead (14, 15). In the present protocol, the dive stimulus was evoked by placing a plastic bag filled with cool water (6–9°C) on the subject’s forehead. This stimulus has been shown to evoke PNS activity, as reflected in increased IBIs and changing spectral decomposition of cardiac variability (14, 15).
Procedures
Subjects were seated in a comfortable lounge chair in a small, sound-attenuated room. The experimental procedure was explained, and informed consent forms were signed. Self-adhesive electrodes for recording the EGG were attached to the subject’s abdomen: one electrode between the umbilicus and the xiphoid process, and one in the upper quadrant, left side, just below the costal margin, approximately 8 cm left of the midline. The EGG reference electrode was placed in the right upper quadrant approximately 6 cm to the right of the subject’s midline. One ECG lead was placed directly beneath the center of the left collarbone, and the other electrode was positioned on the right electrode on the right side beneath the last floating rib, directly above the hip. An ECG reference electrode was located beneath the electrode on the right side. Two electrodes were placed on the back of the left calf approximately 5 cm apart for shock delivery.
Paced breathing was maintained throughout the experiment because vagally mediated rhythmic variations in heart rate are ideally assessed under conditions of controlled breathing (28, 29). The pace was set at 15 breaths per minute by a metronome to avoid the influence of respiration on cardiac sympathetic activity that can occur at slower respiration rates (21) and to keep respiration frequency at a higher rate than EGG frequencies of interest. Each subject practiced the paced breathing technique until comfortable with the procedure.
During the baseline period, subjects sat quietly while pacing their breathing. After the baseline period, subjects experienced the shock or dive stimulus; the order of these two stimuli was counterbalanced between subjects. Each experimental period was 4 minutes long and was followed by a recovery period of simple relaxation lasting 6 to 8 minutes. Immediately after each of the two experimental periods, subjects reported their emotional status using the 13 descriptors listed above.
Data Reduction
ANS Activity.
Frequency domain measures of heart rate variability were derived from the autoregressive spectra and calculated for each 4-minute test period. The autoregressive program sequentially recognized peaks for individual R spikes and calculated the autoregressive coefficients that defined the power spectrum for that series of IBIs. Power values were calculated in power spectral density units (ms2/Hz) for each spectral component in the model that gave the best statistical fit (22). Peaks in the LF (0.039–0.15 Hz) and HF (0.18–0.35 Hz) ranges were extracted. The power in the LF component is thought to be due to baroreceptor-mediated blood pressure variations and is a function of relatively slow modulation by both sympathetic and vagal mechanisms (21, 30). The power in the HF component reflects relatively fast respiratory modulation of cardiac parasympathetic activity (21). Hence, the LF component is used as a rough estimate of SNS activity, and the HF component is used as a rough estimate of PNS activity.
For each subject, a LF peak, HF peak, and mean IBI were calculated for each 4-minute test period. This resulted in a LF peak, HF peak, and mean IBI for the baseline, shock, and dive periods. Differences were calculated for each subject as the difference between shock and baseline (shock - baseline) and the difference between dive and baseline (dive - baseline).
Electrogastrogram.
EGG signals from each test period were spectrally decomposed using the fast Fourier transform with a Hamming window (1024 samples). Power estimates were obtained for every 0.25 cpm from 0.25 to 15.0 cpm (see Ref. 16 for more details). Spectral estimates were summed for the following bands: 3 cpm (2.5–3.75), tachyarrhythmia (4.0–9.75 cpm), bradygastria (0.25–3.5 cpm), and total power (0.25–15 cpm). The percentage of total power found in the 3-cpm and tachyarrhythmia bands was also calculated.
Emotion Score.
Subjects rated each of the 13 emotional descriptors listed above after each experimental period. For each descriptor, difference scores were calculated for each subject as the difference between mean shock and baseline scores (shock - baseline) and the difference between emotion scores after dive and baseline (dive - baseline).
Statistical Analyses
Results from within the shock and dive stimuli and between the shock and dive stimuli were compared using within-subject t tests. p values were adjusted within the IBI, ANS, EGG, and emotion score comparisons using a modified Bonferroni correction (31) to control the family-wise error rate at 
= 0.05. This procedure involved three steps. First, the p value results from a group of comparisons being corrected (eg, the ANS comparisons) were arranged from most to least significant. Second, critical values for rejecting the null hypothesis were calculated using the formula j/N, where j = the number of the test performed (1st test, 2nd test, etc.), = family-wise error rate (0.05 in this case), and N = total number of tests being performed for the variable (eg, EGG). Third, each p value was then compared with the corresponding critical value. If the p value was less than the critical value, the comparison was considered statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Effects of Shock and Dive Stimuli on Gastric Myoelectrical Activity
During the shock stimulus, 3-cpm EGG activity decreased slightly compared with activity at baseline (t(19) = 1.82, p < .10). During the dive stimulus, 3-cpm EGG activity showed little change relative to baseline. However, the decrease in 3-cpm activity during the shock stimulus was significantly different from the increase during the dive stimulus (t(19) = 2.69, p < .05).
Tachyarrhythmia increased significantly during the shock stimulus compared with baseline (t(19) = 3.08, p < .05). This increase was not significantly greater than the slight increase that occurred during the dive stimulus.
Figure 1 is an EGG recording from a subject during baseline and during the shock and dive stimuli. These EGG tracings are reconstructed analogue signals based on the raw digitized EGG signal. One minute of EGG signal from each condition is shown. Notice the shift in frequency from a 3-cpm rhythm during baseline to a 6-cpm tachyarrhythmia during the shock period. The EGG frequency during the dive period is 3 cpm and is similar to the baseline pattern.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Results of the present study show that SNS activation is accompanied by changes in gastric myoelectrical activity. During the shock stimulus, normal 3-cpm EGG activity decreased, whereas gastric tachyarrhythmia increased. Not surprisingly, emotional arousal also increased. It is not clear that PNS activity increased during the dive stimulus. However, normal 3-cpm gastric myoelectrical activity increased during the dive stimulus relative to the shock stimulus, and emotional arousal decreased, indicating that some relaxation did occur. This relaxation brought subjects to a physiological and emotional state that was similar to that at baseline. These findings demonstrate that arousal caused by a SNS stress task is accompanied by objective changes in GI function and subjective changes in emotional state.
Increased SNS activity seems to have a variety of effects on gastric myoelectrical activity. Stern et al. (6) showed that 3-cpm EGG activity was suppressed during the cold pressor test, a painful stimulus that increases SNS activity. In contrast to the painful stimulus of a cold pressor, the emotion of "disgust" was associated with a lack of 3-cpm EGG response in subjects who were sham fed a hot dog on a bun (32). In the present study, a reaction-time task coupled with the threat of an electrical shock decreased 3-cpm activity but also increased gastric tachyarrhythmia activity. In contrast to cold pressor tests or sham feeding disgusting foods, both of which produced only a decrease in 3-cpm activity, the shock task produced tachyarrhythmia as ANS activity shifted toward SNS dominance and away from PNS activity. This result is similar to results of motion sickness studies, in which the stress of motion sickness has been associated with both a decrease in 3-cpm activity and an increase in tachyarrhythmia, along with a decrease in PNS activity and an increase in SNS activity (3, 4).
The dive stimulus proved to be a less salient stimulus. Subjects were relatively relaxed at baseline in a laboratory setting that alone could have been considered stressful. Because of this floor effect, it was difficult to bring subjects to a further level of relaxation. In addition, the cold water proved to be a little too cold and was painful for some subjects. In future studies, this task needs to be revised to use slightly warmer water, or perhaps a better relaxation task should be found.
This study set out to examine the relationships between ANS activity, GI function, and emotional reports. It was shown that a brief (4-minute) SNS task that evoked subjective arousal also evoked dysrhythmic gastric myoelectrical activity. The GI changes that occur in healthy individuals during acute stress may provide insight into functional GI disorders. The availability of a noninvasive measure of GI function coupled with proven behavioral manipulations of ANS activity provides a paradigm for studying acute stress in both healthy individuals and patients with functional GI disorders.
Received for publication July 20, 1998.
Revision received January 20, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  T. Ritz and B. Dahme Implementation and Interpretation of Respiratory Sinus Arrhythmia Measures in Psychosomatic Medicine: Practice Against Better Evidence? Psychosom Med, July 1, 2006; 68(4): 617 - 627. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. P. M. Vianna, J. Weinstock, D. Elliott, R. Summers, and D. Tranel Increased feelings with increased body signals Soc Cogn Affect Neurosci, June 1, 2006; 1(1): 37 - 48. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
