Efficacy of needleless transcutaneous electroacupuncture in synchronization with breathing for symptomatic idiopathic gastroparesis: A blinded and controlled acute treatment trial.

BACKGROUND: Needleless transcutaneous electroacupuncture (TEA) improves nausea and myoelectrical activity in diabetic gastroparesis (GP). Synchronized TEA (STEA), which combines synchronized breathing with TEA, is more potent than TEA in enhancing vagal activity in healthy subjects. AIMS: To investigate whether STEA improves symptoms, electrogastrogram (EGG) and vagal activity in idiopathic gastroparesis (IGP). METHODS: Eighteen IGP subjects underwent 2 randomized visits (sham at non-acupoints or real STEA at acupoints) consisted of a 30-minute baseline, an Ensure challenge to provoke nausea, followed by 60-minute treatment with sham or real STEA, and 15-minute observation period. Severity of nausea, EGG, and vagal activity (based on electrocardiogram and serum Pancreatic Polypeptide, PP) were recorded. RESULTS: In sham or STEA, the nausea scores of 2.7 ± 0.5 and 1.9 ± 0.5 at fasting baseline, respectively, increased to 5.9 ± 0.4 and 5.8 ± 0.3 during Ensure test (P < .05, vs baseline), subsequently reduced to 3.4 ± 0.6 with sham or 3.6 ± 0.6 with STEA, respectively (P < .05, vs Ensure period). Experiments with sham and STEA started with similar % of normal waves on EGG (66.4 ± 3.9 and 61.8 ± 3.0, respectively); decreased to 63. 5 ± 4.1 and 58.2 ± 2.8 during the Ensure test. After STEA, there was ~24% increase in % of normal waves, significantly different from the sham (6.0%) (P < .01). In sham or STEA, vagal activity was identical at baseline and after the Ensure. STEA induced a 3-fold increase in vagal activity compared with sham (P < .01). Ensure increased serum PP levels, and both treatments decreased the PP CONCLUSIONS: In IGP, STEA is not superior to Sham in decreasing nausea, but is more effective in improving gastric dysrhythmia.

Short pulse gastric electrical stimulation for cisplatin-induced emesis in dogs.

BACKGROUND: In a previous study, we investigated the ameliorating effect of gastric electrical stimulation (GES) with a single set of parameters on emesis and behaviors suggestive of nausea induced by cisplatin in dogs. The aim of this study was to investigate the effects of GES with different parameters on cisplatin-induced emesis in dogs. METHODS: Seven dogs implanted with gastric serosal electrodes were studied in six randomized sessions: one control session with cisplatin (2 mg kg(-1)) and five sessions with cisplatin plus GES of different parameters: GES-A: 14 Hz, 5 mA, 0.3 ms, 0.1 s on and 5 s off; GES-B: increased frequency and on-time; GES-C: increased frequency; GES-D: increased frequency and pulse width; and GES-E: increased frequency and amplitude. Gastric slow waves and emetic responses were recorded in each session. KEY RESULTS: (i) Cisplatin induced emetic responses and gastric dysrhythmia. The peak time of the emetic response was during the fourth hour after cisplatin. (ii) GES with appropriate parameters reduced cisplatin-induced emesis. The number of vomiting times during the 6 h after cisplatin was 7.0 ± 1.4 in the control, 4.7 ± 1.2 with GES-A (P = 0.179), 4.2 ± 1.2 with GES-B (P = 0.109), 7.0 ± 0.8 with GES-C (P = 0.928), 2.1 ± 0.3 with GES-D (P = 0.005) and 4.7 ± 1.5 with GES-E (P = 0.129). However, none of the GES parameters could improve gastric dysrhythmia. CONCLUSIONS & INFERENCES: Gastric electrical stimulation with appropriate parameters reduces cisplatin-induced emetic responses and behaviors suggestive of nausea in dogs. Among the tested parameters, GES with increased pulse width seems to produce better relief of cisplatin-induced emesis.

Hepatic electrical stimulation reduces blood glucose in diabetic rats.

BACKGROUND: The aim of this study was to investigate the feasibility and mechanisms of controlling blood glucose using hepatic electrical stimulation (HES). METHODS: The study was performed in regular Sprague-Dawley (SD) rats, streptozotocin-induced type 1 diabetic rats and Zucker diabetic fatty (ZDF) rats chronically implanted with one pair of stimulation electrodes on two lobes of the liver tissues. KEY RESULTS: (i) Hepatic electrical stimulation was effective in reducing blood glucose by 27%-31% at time points 60, 75 and 90 min after oral glucose in normal rats; (ii) HES reduced blood glucose in both fasting and fed states in both type 1 and type 2 diabetic rats; (iii) Chronic HES decreased the blood glucose level, and, delayed gastric empty and increased plasma glucagon-like peptide-1 (GLP-1) level; and (iv) No adverse events were noted in any rats during HES. Histopathological analyses and liver function tests revealed no electrode dislodgement, tissue damages or liver enzyme changes with HES. CONCLUSIONS & INFERENCES: Hepatic electrical stimulation is capable of reducing both fasting and fed blood glucose in normal, and type 1 and type 2 diabetic rats and the effect may be partially mediated via an increase in GLP-1 release.

Gastric slow waves, gastrointestinal symptoms and peptides in systemic sclerosis patients.

Impaired gastric slow waves, frequent gastrointestinal (GI) symptoms and altered GI peptides have been reported in Scleroderma (SSc) patients. The aim of this study was to investigate the associations among these three important components in GI dysmotility. Seventeen fasted SSc patients underwent four channel surface electrogastrography, measuring % of normal gastric slow waves or dysrhythmia. Patients completed a questionnaire designed by us to assess demographics, upper and lower GI symptoms (symptom presence, frequency and impact on quality of life, QOL), by YES/NO, Likert Scales and Visual Analogue Scales 1-100 mm (called GI Dysmotility Questionnaire, GIDQ) and health-related QOL by SF-36. Fasting plasma vasoactive intestinal peptide (VIP) and motilin levels were measured by peptide immunoassays. There were significant correlations between percentages of gastric dysrhythmias (bradygastria or arrhythmia) and a number of major GI symptoms such as nausea, abdominal bloating and pain. The plasma level of VIP was correlated positively with % dysrhythmia but negatively with % normal slow waves. Motilin was positively correlated with slow wave coupling (coordination). No major differences were noted in the measured peptides or gastric slow waves between limited SSc and diffuse SSc. Correlations were noted between SF-36 domain scores and our GIDQ scores. In SSc patients, gastric dysrhythmias are correlated with certain GI symptoms. Correlations are also noted between plasma VIP/Motilin levels and gastric slow waves. Thus in SSc, gastric dysrhythmias may be predictive of development of certain dyspeptic symptoms. Plasma VIP may be involved in the development of dysrhythmias.

Effect of intestinal pacing on small bowel transit and nutrient absorption in healthy volunteers.

BACKGROUND: Intestinal pacing (IP) has been previously shown to delay gastric emptying and reduce food intake in animals. The aims of this study were to investigate the effect and mechanism of IP on nutrient absorption in healthy volunteers. METHODS: Twelve healthy volunteers (six men, six women) were involved in a two-session (one session without IP and one with IP) study. At the beginning of each session, a nasal-duodenal feeding tube, with two ring electrodes (used for IP) on the tip of the tube, was incubated into the duodenum under endoscopy. After a complete recovery from the incubation, the duodenum was infused via the feeding tube with 150 ml 30% intralipid + 25 g D-xylose within 30 min, and the stool was collected for 24 h for the analysis of fecal lipid during which a controlled meal was taken. Then 100 ml 1mCi(99)Tc-labeled non-absorbable solution was infused within 3 min. The subject was asked to lie under a gamma camera for at least 1 h for the measurement of small bowel transit. The movement of isotopes was monitored by gamma camera at an interval of 10 s. The first appearance of isotopes in the cecum was considered as small intestinal transit time. The order of the two sessions was randomized and 1 week apart. In the IP session, intestinal pacing was performed via the pair of the ring electrodes for 2 h initiated at the beginning of infusion with a pacing frequency of 13 pulses/min, pulse width of 300 ms and amplitude of 5 mA. RESULTS: (1) IP significantly reduced lipid and D-xylose absorption. The fecal lipid was 6.6 +/- 4.6 g without IP and almost doubled with IP (11.1 +/- 6.5 g, P = 0.047). Similarly, the D-xylose in urine was 3.46 +/- 2.22 g with IP, which was significantly lower than that without IP (6.63 +/- 5.06 g, p = 0.049). (2) IP accelerated intestinal transit. The transit time was 39 +/- 17 min in the control session and reduced to 28 +/- 10 min in the IP session (p < 0.03). (3) Diarrhea was reported in one subject without IP but in six subjects with IP (p < 0.05). CONCLUSIONS: The increased fecal lipid and induction of diarrhea with intestinal pacing suggest that intestinal pacing is capable of inducing malabsorption. This effect maybe contributed to the acceleration of intestinal transit.

Involvement of the hippocampus and neuronal nitric oxide synthase [correction of synapse] in the gastric electrical stimulation therapy for obesity.

BACKGROUND: Gastric electrical stimulation (GES) has been introduced for treating obesity. However, possible central mechanisms remain to be revealed. Hippocampus has been shown to be involved in the regulation of gastrointestinal functions. Changes in hypothalamic neuronal nitric oxide synthase (nNOS) have been observed in genetically obese rodents. The aim of this study was to investigate the involvement of nNOS with GES in the rodent hippocampus. METHODS: The effect of GES on gastric distension (GD) neurons was investigated using four different sets of parameters (GES-A, pulse train of standard parameters; GES-B, reduced on time; GES-C, increased pulse width, and GES-D: reduced pulse frequency), and the expression of nNOS in hippocampus was observed by fluoimmunohistochemistry staining. RESULTS: CA1 region neurons (90.8%) responded to GD, 50.6% of which showed excitation (GD-E neurons) and 49.4% showed inhibition (GD-I neurons). Most of GD-responsive neurons (63.3%) were excited with GES. The response to GES was associated with stimulation strength, pulse width and frequency. GD-E neurons (62.5%, 76.9%, 100%, and 62.3%) and GD-I (63.6%, 47.1%, 85.7% and 50.0%) showed excitatory responses to GES-A, GES-B, GES-C, and GES-D, respectively (P < 0.05, GES-C vs. others). nNOS immunoreactive (nNOS-IR) positive neurons were observed in hippocampus CA1, CA2-3 regions and the dentate gyrus. The expression of nNOS-IR positive neurons was significantly decreased in CA1 and CA2-3 region (P < 0.05) after GES (para-C) for 2 h. CONCLUSIONS: Excitation of GD-responsive neurons and reduced expression of nNOS in the hippocampus are indicative of the central effect of GES.

Cisplatin-induced gastric dysrhythmia and emesis in dogs and possible role of gastric electrical stimulation.

The aim of this study was to investigate the effect of cisplatin on gastric myoelectrical activity and the role of gastric electrical stimulation in the treatment of cisplatin-induced emesis in dogs. Seven dogs implanted with electrodes on the gastric serosa were used in a two-session study. Cisplatin was infused in both the control session and the gastric electrical stimulation session, and gastric electrical stimulation was applied in the gastric electrical stimulation session. Gastric slow waves and emesis, as well as behaviors suggestive of nausea, were recorded during each session. The results were as follows: (1) cisplatin induced vomiting and other symptoms and induced gastric dysrhythmia. The percentage of normal slow waves decreased significantly during the 2.5 h before vomiting (P=0.01) and the period of vomiting (P<0.001). (2) Gastric electrical stimulation reduced emesis and the symptoms score. The total score in the control session was higher than that in the gastric electrical stimulation session (P=0.02). However, gastric electrical stimulation had no effects on gastric dysrhythmia. It is concluded that cisplatin induces emesis and gastric dysrhythmia. Gastric electrical stimulation may play a role in relieving chemotherapy-induced emetic responses and deserves further investigation.

A novel method of 2-channel dual-pulse gastric electrical stimulation improves solid gastric emptying in dogs.

BACKGROUND: Gastric electrical stimulation (GES) is known to improve vomiting with short pulses, normalize dysrhythmia with long pulses, and accelerate gastric emptying with 2 channels. The aim of this study was to assess the effects of a new method GES, namely, 2-channel GES with dual pulses on gastric emptying of solids as well as gastric dysrhythmia and emetic responses. METHODS: Seven beagle dogs implanted with 4 pairs of electrodes were studied. A novel method of GES was proposed: 2-channel dual-pulse GES in which each stimulus was composed of a short pulse followed with a long pulse, and stimulation was delivered at 2 different locations. The study was performed to test the effects of this new method of GES on vasopressin-induced delayed gastric emptying of solids, gastric dysrhythmia, and emetic responses. RESULTS: (1) Vasopressin-induced gastric dysrhythmia and emetic responses, as well as delayed gastric emptying of solids (P < .01). (2) Two-channel, but not 1-channel, dual-pulse GES was able to accelerate vasopressin-induced delayed gastric emptying of solids. (3) Both 1- and 2-channel dual-pulse GES was capable of improving dysrhythmia and emetic responses (P < .01). CONCLUSIONS: The novel method of 2-channel dual-pulse GES is capable of accelerating gastric emptying of solids and improving dysrhythmia and emetic responses induced by vasopressin. This new method of GES may have a potential for gastroparesis.

Induction of heme oxygenase-1 improves impaired intestinal transit after burn injury.

BACKGROUND: Burn injury has been shown to impair intestinal transit. The induction of heme oxygenase (HO)-1, the rate-limiting enzyme in heme degradation, has been demonstrated to provide protection against various injuries. The aim of this study was to investigate whether the induction of HO-1 by hemin would improve impaired intestinal transit after burn injury. METHODS: Burn/sham rats were divided into 3 groups: saline solution, hemin (HO-1 inducer), and hemin plus tin protoporphyrin IX. Intestinal transit was measured with the use of phenol red and assessed with the geometric center. The gene and/or protein expression of inducible nitric oxide synthase (iNOS), cyclooxygenase (COX)-2, interleukin (IL)-1beta, HO-1, and p38 mitogen-activated protein kinase (p38 MAPK) was measured by real-time polymerase chain reaction and/or by Western blot analysis. RESULTS: Intestinal transit was delayed with burn injury and improved significantly with the induction of HO-1; burn injury significantly activated p38 MAPK and myeloperoxidase and increased gene and/or protein expression of iNOS, COX-2, IL-1beta, and HO-1. The administration of hemin led to a significant decrease in the activation of p38 MAPK and myeloperoxidase and the gene and/or protein expression of iNOS, COX-2, and IL-1beta. CONCLUSION: The induction of HO-1 improves burn-induced delayed intestinal transit. The beneficial effect of hemin treatment could be linked, at least in part, to the down-regulation of iNOS, COX-2, and IL-1beta expression, which suggests that the induction of HO-1 may provide an effective therapeutic measure for gut dysmotility after burn injury.

Ghrelin improves burn-induced delayed gastrointestinal transit in rats.

Delayed gastrointestinal transit is common in patients with severe burn. Ghrelin is a potent prokinetic peptide. We aimed at testing the effect of ghrelin on burn-induced delayed gastrointestinal transit in rats. Gastric emptying (GE), intestinal transit (IT), and colonic transit (CT) studies were performed in male Sprague-Dawley rats. Rats were randomized into two main groups as follows: sham injury and ghrelin-treated burn injury with doses of 0, 2, 5, and 10 nmol/rat ip 6 h after burn. Sham/burn injury was induced under anesthesia. Rats received a phenol red meal 20 min following ghrelin injection. Based on the most effective ghrelin dose, 1 mg/kg sc atropine was given 30 min before the ghrelin in one group of rats for each study. The rats in each group were killed 30-90 min later; their stomachs, intestines, and colons were harvested immediately, and the amount of phenol red recovered was measured. Percentage of gastric emptying (GE%) and geometric center for IT and CT were calculated. We found 1) severe cutaneous burn injury significantly delayed GE, IT, and CT compared with sham injury (P < 0.05); 2) ghrelin normalized both GE and IT, but not the CT; 3) the most effective dose of ghrelin was 2 nmol/rat; and 4) atropine blocked the prokinetic effects of ghrelin on GE% and IT. In conclusion, ghrelin normalizes burn-induced delayed GE and IT but has no effect on CT in rats. The prokinetic effects of ghrelin are exerted via the cholinergic pathway. Ghrelin may have a therapeutic potential for burn patients with delayed upper gastrointestinal transit.

Implantable gastric stimulation alters expression of oxytocin- and orexin-containing neurons in the hypothalamus of rats.

BACKGROUND: Implantable gastric stimulation (IGS) has been proposed as a therapeutic option for treating obese patients. We studied the underlying central mechanism behind the reduction of food intake and body weight by gastric electrical stimulation (GES), by studying the expression of anorexigenic and orexigenic-peptide containing neurons in the hypothalamus. METHODS: Oxytocin antiserum, orexin antiserum and c-Fos protein-antiserum were used for immunostaining technique. Brain sections were obtained from the control rats and those with 2 hours of GES with parameters typically used in the treatment of obesity in humans. RESULTS: A) 2-hr IGS increased the expression of oxytocin-immunoreactive (IR) neurons in the paraventricular nucleus (PVN, 23.7+/-1.8 vs 31.1+/-2.2, P<0.05) and the supraoptic nucleus (SON, 29.0+/-2.2 vs 39.7+/-2.5, P<0.01). However, the expression of orexinIR neurons in the lateral hypothalamic area (LHA) was decreased (27.8+/-2.6 vs 20.6+/-1.7, P<0.01). B) The expression of c-Fos positive neurons was increased in the PVN and SON with IGS. A coexistence of oxytocin-IR positive neurons and c-Fos-IR ones in the PVN and SON from the adjacent brain sections was observed, confirming the activation of OT-containing neurons in the PVN and SON following IGS. CONCLUSION: IGS increases the expression of hypothalamic neurons containing the anorexigenic neuropeptide, oxytocin, and decreases the expression of neurons containing the orexigenic neuropeptide, orexin. This central anorexigenic effect of GES may contribute to the reduced appetite and increased satiety in obese patients with IGS therapy.

Central mechanisms of gastric electrical stimulation involving neurons in the paraventricular nucleus of the hypothalamus in rats.

BACKGROUND: Gastric electrical stimulation (GES) has been shown to be effective in inducing satiety in patients with obesity, as well as in treating nausea and vomiting in patients with gastroparesis. It is conceivable that these effects are mediated via the central nervous system, although there is a lack of studies or evidences. The aims of this study were to investigate the effects of GES with different parameters on neurons in the paraventricular nucleus (PVN) of the hypothalamus and to study whether the neuronal response was related to stimulation parameters. METHODS: Extracellular potentials of single neurons in the PVN were recorded in anesthetized rats at baseline, during gastric distention (GD) and during GES with various sets of parameters. RESULTS: Of 115 PVN neurons tested by GD, 104 were GD-responsive neurons. Most of these GD-responsive neurons were responsive to GES of various parameters. 1) GES with parameters used in treating obesity, activated neurons both excited (GD-E) and inhibited (GD-I) by gastric distention; it excited 48.4% (16/33) of the GD-E neurons and 60.4% (29/48) of GD-I neurons. 2) GES with increased pulse width was of the highest efficacy in exciting the GD-responsive neurons, whereas a reduction in stimulation frequency resulted in a lower activation. 3) GES with parameters used in treating gastroparesis solicited neuronal responses opposite to those with parameters used for treating obesity. CONCLUSION: GES activates gastric-related neurons in the PVN, and the excitatory effect of GES seems related to stimulation strength. The GES used for treating gastroparesis and the GES used for treating obesity seem to solicit different neuronal responses in the PVN.

Systematic review: pathophysiology and management of gastrointestinal dysmotility in systemic sclerosis (scleroderma).

Gastrointestinal dysmotility in systemic sclerosis (scleroderma) is prevalent in 90% of patients, increasing morbidity and in some cases mortality. The resultant gastrointestinal complications are usually extensive, involving many regions of the gut from the oesophagus to the anus. Collagen replacement of vascular and enteric smooth muscle results in hypomotility, lumen dilatation, tensile rigidity and eventual loss of organ functions. The aim of this paper is to provide an overview of systemic sclerosis-related gastrointestinal dysmotility and available/potential therapeutic options. We evaluated published data on the pathophysiology and management of gastrointestinal dysmotility in systemic sclerosis patients using the MEDLINE database for English and non-English articles from 1966 to July 2005. Based on this systematic review, lifestyle and medical therapy approaches are preferred as they often improve and/or ameliorate symptoms. Surgery is only recommended with serious, rare complications such as bowel perforation or ischaemia. Alternative therapies such as acupuncture-based therapies are well tolerated, with clinical improvement and may be of potential therapeutic benefit for systemic sclerosis gastrointestinal dysmotility. Further elucidation of initiating and persistent mechanisms of systemic sclerosis-related gastrointestinal dysmotility will optimize the development of a multidisciplinary and more directed treatment regimen.

Gastric electrical stimulation (GES) with parameters for morbid obesity elevates lower esophageal sphincter (LES) pressure in conscious dogs.

BACKGROUND: Treatment of obese patients with the implantable gastric stimulator (IGS) was reported to improve reflux symptoms, independent of weight loss. We evaluated the effect of gastric electrical stimulation on LES pressure in conscious dogs. METHODS: 8 dogs were studied. GES with three sets of parameters was randomly applied via a pair of electrodes implanted in the fundus on separate days. Manometry was performed with a Dent-Sleeve catheter passed through an esophageal canula. The involvement of the cholinergic pathway was also tested. RESULTS: 1) Stimulation with IGS parameters (40 Hz, 0.3 ms, 6 mA, 2 seconds on and 3 seconds off) induced a significant increase in LES pressure (29.9+/-4.8 mmHg), and remained significantly higher during the post-stimulation period (32.6+/-9.6 mmHg) compared to baseline (24.5+/-3.8 mmHg), P<0.01.2) Long pulse stimulation (10 cpm, 300 ms, 8 mA) tended to increase LES pressure from 29.6+/-4.4 mmHg of baseline to 31.8+/-4.9 mmHg with stimulation, to 32.6+/-4.5 mmHg after discontinuation (P=0.08); 3) Modified IGS parameters (40 Hz, 2 ms, 6 mA, 2 seconds on and 3 seconds off) did not induce a significant change in LES pressure during and after stimulation. 4) Effect of stimulation with IGS parameters on LES pressure was blocked by intravenous atropine. CONCLUSION: GES with IGS parameters significantly increases LES pressure in conscious dogs. This effect is mediated by the cholinergic pathway. These results suggest that GES may be able to benefit GERD patients and obese patients with GERD.

Effects and mechanisms of implantable gastric stimulation on gastric distention in conscious dogs.

BACKGROUND: While implantable gastric stimulation (IGS) on the lesser curvature has been reported to induce weight loss in obese patients, its mechanisms involving gastric mechanical activity remain largely unknown. The aim of this study was to investigate the effect and mechanism of IGS on gastric tone in canines. METHODS: 8 healthy dogs were implanted with a gastric cannula on anterior stomach and a pair of electrodes at lesser curvature. Gastric tone was assessed with an electronic barostat in 2 randomized sessions, a control and a session with L-NNA (nitric oxide synthase inhibitor). The control session included 30 min baseline, followed by 30 min with IGS; the L-NNA session included 30 min baseline, 30 min immediately after a bolus of L-NNA, and 30 min with IGS. RESULTS: 1) In the control session, IGS significantly increased the proximal gastric volume from baseline 91.3+/-7.1 ml to 186.3+/-27.1 ml (P<0.05); 2) L-NNA markedly reduced the proximal gastric volume from 110.3+/-11.9 ml to 56.9+/-10.8 ml (P<0.01); and subsequent IGS did not significantly increase the proximal gastric volume (P>0.05). CONCLUSION: IGS significantly inhibits proximal gastric tone or induces gastric distention, and this inhibitory effect is mediated via the nitrergic pathway.

Impaired gastric slow wave rhythmicity in patients after bone marrow or stem cell transplant.

Patients after bone marrow or stem cell transplant often develop gastrointestinal symptoms. The aim of this study was to investigate possible impairment of gastric myoelectrical activity in these patients. The study was performed in 15 patients who had had bone marrow or stem cell transplant and 13 healthy subjects. Gastric myoelectrical activity was assessed using electrogastrography. The electrogastrogram (EGG) was made for 30 min in the fasting state and 60 min after a test meal (475 kcal; turkey sandwich). Overall and minute-by-minute spectral analyses were performed to derive various parameters of the EGG. Compared with the healthy controls, the patients showed a significantly higher percentage of arrhythmia (no obvious rhythmicity observed in the EGG) in both fasting (17.6 +/- 3.8% vs 7.1 +/- 2.17%, P < 0.02) and fed (11.4 +/- 2.65% vs 4.19 +/- 1.04%, P < 0.02) state. The patients showed a significantly higher instability coefficient of the dominant frequency in the fasting state than in the controls (0.51 +/- 0.06 vs 0.29 +/- 0.18, P < 0.008). The total average symptom score was 3.93 +/- 0.84 in the patients and 0 in the controls, and a relatively weak but significant correlation was found between the symptom scores and the percentage of arrhythmia in the patients in fed state (r = 0.69, P < 0.02). It was concluded that patients with bone marrow or stem cell transplant have excessive arrhythmia that is correlated with their dyspeptic symptoms.