Can an oscilloscope detect gastroparesis?

The autonomic neuropathy of the stomach in terminal. The significance of electrogastrography as a diagnostic tool


1 From the Department of Clinical Medicine, Department of Internal Medicine of the Medical Faculty of the University of Saarland, Homburg / Saar Medical Clinic III of the West Palatinate Clinic in Kaiserslautern (Head Physician: Prof. Dr. med. FW Albert) The autonomic neuropathy of the stomach in terminal kidney failure The importance of electrogastrography as diagnostic dissertation to obtain the degree of Doctor of Medicine of the Medical Faculty of the UNIVERSITY OF SAARLAND 2006 Submitted by: Martina Gradinger born on: in Mainz

2nd day of the doctorate: Dean: 1st reporter: Prof. Dr. med. M. Montanarh Prof. Dr. med. F.W. Albert 2nd Rapporteur:

3 TABLE OF CONTENTS Summary III - IV 1. Introduction 1 2. Basics 2.1 Anatomy and physiology of the stomach and gastric emptying Electrogastrography 7 3. Material and methods 3.1 Subject material Examination procedure Statistical evaluation results 4.1 Evaluation of the dominant frequency Evaluation of the frequency distribution pattern Evaluation of the dominant power evaluation of the power spectrum Evaluation of the slow wave coupling 49 5th discussion 5.1 Significance of the dominant frequency Significance of the frequency distribution pattern Significance of the power Significance of the slow wave coupling Assessment of the methodology and the current investigation 63 I.

4 CONTENTS 6. Bibliography Appendix Definition of terms List of figures Descriptive statistics V VI VIII 8. Acknowledgments XIII 9. Curriculum Vitae XIV II

5 SUMMARY Summary: Autonomic neuropathy is common in diabetics and dialysis patients and has been well studied with its effects on the cardiovascular system. As a possible cause, it can be partly responsible for the gastrointestinal complaints with the consequent malnutrition that are often found in these patients. A precise understanding of the underlying disorder of the neural impulses - influencing gastric motility - could provide assistance with regard to the existing therapy options. The aim of the present clinical work was therefore to use non-invasive electrogastrography to analyze the peculiarities of the autonomic neuropathy of the stomach and to check the value of the method as a diagnostic tool for this disease. For this purpose, measurements of the myoelectric impulses of the stomach via superficial skin electrodes were carried out in the fasting state and after taking a defined test meal on various groups of test subjects with assumed neuropathy (diabetics, patients with terminal renal insufficiency with diabetes mellitus and without diabetic metabolism) and compared with healthy test persons. After computer-aided evaluation by means of spectral analysis, the evaluation of the parameters dominant frequency, frequency distribution pattern, dominant power, power distribution pattern and slow wave coupling followed. In the patient groups, especially in the preprandial phase, a significant arrhythmia of the stomach in the frequency distribution pattern with rhythmization after ingestion as a feature of the neuropathy independent of the underlying disease was found. The other characteristics examined did not differ. Electrogastrography can thus represent a component in the diagnosis of autonomic neuropathy. The main feature of this neuropathy is the arrhythmia of the gastric pacemaker when the dominant frequency of the stomach is maintained as normal and the signal is properly propagated along the antral axis. III

6 SUMMARY Summary: The autonomic neuropathy of the stomach in endstage renal disease value of electrogastrography as a diagnostic tool. Autonomic neuropathy is highly prevalent in patients with diabetes or endstage renal disease. It is well studied in its effect on the cardiovascular system. Gastrointestinal disorders and malnutrition are common in uremic patients and the autonomic neuropathy can be responsible for these problems. Gastric motility is influenced by neuronal impulses. Exact understanding of the underlying disturbance of neuronal impulses could give further information regarding the existing therapy options. Aim of this study was to analyze the gastric autonomic neuropathy with the help of non invasive electrogastography and to investigate the clinical value of this method. We compared gastric myoelectrical impulses in four different groups of subjects. All groups were measured before and after a standardized test meal. The study was performed in healthy subjects, diabetic patients and patients with endstage renal disease with and without diabetes. A computer based running spectrum analysis was performed. Variables were assessed as distribution patterns of frequency and power, dominant frequency, dominant power and slow wave coupling. In all groups of patients the spectrum analysis disclosed preprandial gastric dysrhythmia, with normalization. These findings were independent of the underlying disease. The remaining examined parameters did not differ between all tested groups. The electrogastrography can thereby represent a helpful component in the diagnostics of autonomic neuropathy. The characteristic observation is the dysrhythmia of the gastric pacemaker with normal gastric dominant frequency and undisturbed slow wave coupling. IV

7 INTRODUCTION 1. Introduction: Malnutrition or malnutrition is a common problem in patients with end-stage renal disease. With a GFR of less than 10 ml / min, many patients consume less than 20 kcal per day and kg kg (IKITZLER, 1996). Those affected complain of constant nausea, a rapid feeling of fullness or pain in the upper abdomen. This is associated with weight loss and, as a result, a general weakness in the immune system. The cause of these symptoms is unclear (DUMITRASCU, 1995). Analogous to diabetic autonomic neuropathy with the same clinical picture, uremic autonomic neuropathy could be the basis of the symptoms in patients requiring dialysis. As a consequence of chronic renal insufficiency, neuropathies of various manifestations can appear. They become symptomatic in 50-90% of all people on dialysis. Changes in the peripheral nervous system are common. Peripheral sensory neuropathy should be mentioned here, which can lead to severe neuropathic pain and is associated with symmetrical, distally emphasized, mixed motor-sensitive deficits. Another clinical manifestation is the so-called restless legs syndrome, which is found in 20 40% of uraemics. The clinical picture of the autonomic dysfunction can still precede an impairment of the sensorimotor functions. The parasympathetic nervous system is impaired in 30-70% and the sympathetic nervous system in 10-40% of dialysis patients (CONVERSE, 1992; JASSAL, 1998). Damage to blood pressure control leads to orthostatic problems and a drop in blood pressure during dialysis. This manifestation of the autonomic neuropathy can be objectified by measuring the heart rate variability during breathing, Valsalva maneuvers or in the long-term ECG. In the area of ​​the gastrointestinal tract, the autonomic dysfunction manifests itself in the 1st

8 INTRODUCTION Extreme case of gastroparesis with the symptoms described at the beginning. The gold standard for the objective measurement of delayed gastric emptying is the gastric emptying scintigraphy introduced by GRIFFITH and co-workers in 1966. However, this is only the indirect evidence of a presumably underlying regulatory disorder in the sense of an autonomic neuropathy. A direct change in the neural impulses influencing gastric motility cannot be recorded as a result. This should be possible through non-invasive electrogastrography. The aim of the present work is to describe the possible value of this examination method in the diagnosis of the autonomic neuropathy of the stomach in dialysis patients. 2

9 BASICS 2. Basics: 2.1. Anatomy and physiology of the stomach and gastric emptying: The stomach forms the largest horn-shaped expansion of the digestive tract. It divides into the pars cardiaca, the gastric fundus, the gastric corpus, the antrum and the pylorus. The shape of the stomach is extremely different, and the position of the stomach is also not constant; it depends on the stomach contents and the position of the human body (WALDEYER and MAYET, 1987). The stomach wall consists of several layers (Fig. 1) -Mucosa -Muscularis mucosae -Submucosa -Muscularis propria -Serosa Fig. 1: Wall structure of the gastrointestinal tract in cross section 3

10 PRINCIPLES The active motor part of the stomach is formed by the muscularis propria. This is the only one in the digestive tract that has an additional third muscle layer. The task of the stomach is, on the one hand, to create a reservoir and, on the other hand, to chop it up and deliver it to the small intestine in portions: For this purpose, the stomach can be functionally divided into 2 regions. The gastric fundus forms a reservoir and the antrum represents the part in which the chyme is physically and chemically broken down by contractions. These activities are controlled and controlled on 3 levels: - Parasympathetic and sympathetic nervous system - Neuron plexus in the wall of the gastrointestinal tract (enteric brain) - Smooth muscle cells of the stomach wall The sympathetic nervous system reaches the stomach via the celiac plexus with fibers from the thoracic one Spinal cord segment. These supply the myenteric ganglia and above all deliver a dense network of fibers to the pylorus. The parasympathetic control of the stomach is also transported via the vagus nerve via the myenteric nerve plexus. The enteric brain forms a network of nerve plexuses that work as circuits between the sympathetic / parasympathetic and sensory stimuli in the stomach wall. With around million nerve cells, these nerve plexuses represent the largest number of neurons outside the central nervous system (FURNESS and COSTA, 1980). They form a neural network that spreads out in several layers in the wall of the entire gastrointestinal tract. The best characterized switching points are the myenteric plexus (Auerbach) and the submucosal plexus (Meissner). 4th

11 FUNDAMENTALS Smooth muscle forms the third level of control over gastrointestinal motility. Special pacemaker cells, which are characterized by a spontaneous depolarization of the resting potential of the membrane, give impulses for contraction. Through electrical coupling with the neighboring muscle cells in the form of a muscle syncytium, the contraction spreads over the circumference and in the longitudinal axis of the stomach. This pacemaker cell area lies in the middle of the great curvature (WEBER and KOHATSU, 1970). The resting membrane potential shows slow fluctuations, so-called slow waves, the frequency is about 3 / min. Not every depolarization of the pacemaker cells leads to a muscular contraction (SZURSZEWSKI, 1981). Only when a volley of rapid membrane potential fluctuations (so-called spike bursts) is grafted onto the slow wave does a contraction occur (MÜLLER-LISSNER, 1986). A distinction is therefore made between two types of electrical gastric activity. The electrical control activity (ECA = electrical control activity), represented by the gastric slow waves, and the electrical response activity (ERA = electrical response activity) in the form of the spike bursts, which are answered by muscle contractions. The interaction of these three levels of control is hormonally regulated depending on the composition and amount of the food consumed. 5

12 FUNDAMENTALS Fig. 2: Control of gastric emptying by inhibiting and promoting influences When fasting, episodes of motor rest alternate with episodes of motor activity in a regular sequence. These so-called interdigestive motor complexes run in 3 phases with a cycle duration of around minutes. In the first phase there is motor rest, in the second phase irregular contractile activity is detectable, the third phase shows regular propulsive contractions. This fasting cycle is used to empty indigestible solids. This phenomenon was described for the first time in the small intestine by SZURSZEWSKI. This complex pattern is interrupted by the ingestion of food. (KONTUREK and DOMSCHKE, 1998) 6

13 BASICS 2.2. Basics of Electrogastrography Electrogastrography is the name given to the non-invasive recording of electrical activity of the stomach using superficial abdominal skin electrodes. For the first time in 1922, ALVAREZ recorded an electrical signal from the stomach via skin electrodes. From the correlation between the pulse fluctuations and the peristaltic contractions of the stomach visible through the abdominal wall in his slim patient, he concluded that the signal obtained was of gastric origin. Subsequent studies with simultaneous registration of signals via superficial skin electrodes and internal serosal electrodes in dogs (SMOUT, 1980; ATANASSOVA, 1995) and also in humans (ABELL and MALAGELADA, 1985; COLESKI, 2004) revealed that electrogastrography was indeed in is able to record the gastric slow waves. The optimal position of the electrodes was achieved by placing them along the axis of the antrum (MIRIZZI, 1983). Due to the considerable individual anatomical differences, this should be determined sonographically (PFAFFENBACH, 1995) Since the electro-gastrographic signal is only weak overall and can be heavily overlaid by artifacts, the sole visual interpretation of the original recording has ultimately proven to be unsuitable for evaluation. The introduction of the computer-aided evaluation with the help of the fast Fourier transformation simplified the interpretation of the EGG signal. A mathematical function is used to map the prevailing frequencies in the spectrum from the vibrations obtained during data recording. The frequency with the greatest power in the analyzed signal is called the dominant frequency. The frequency range defined as normal is between 2 and 4 wave cycles per minute (cpm), tachygastric frequencies above 4 cpm, and bradygastric frequencies below 2 cpm. 7

14 BASICS A further development towards a more differentiated evaluation is the so-called running spectrum analysis (VAN DER SCHEE, 1987, CHEN and McCALLUM 1991). The examination period is divided into intervals and the dominant frequencies are calculated by analyzing the recorded waves over these defined periods of time. Since this is an averaging, the selected period should be as short as possible. With the slow oscillation of around three cycles per minute, a period of 2 minutes is useful. Shorter collection periods would no longer be able to capture the slow fluctuations. This approach enables changes in the spectrum over time - e.g. after stimulation by a test meal - to be recognized. In addition, the period recording allows a quantification of arrhythmias in the sense of bradygastria, tachgastria or arrhythmias. Above all, the purity of the waveform is reflected in the power of the transformed curve and is given as the dominant power. The unit is decibel (db), but it is also influenced by the amplitude of the original curve. By measuring over several bipolar electrodes, the propagation speed of the signal over the antral axis - so-called slow wave coupling - can be determined. 8th

15 BASICS a) Example of an inconspicuous EGG signal from a healthy subject. The associated running spectrum curves each have a clear peak at the same frequency. b) Example of a signal recording of an EGG in a dialysis patient. The running spectrum curves sometimes show no clear peak at all. The analysis shows a high proportion of dysrhythmias. Fig.3: Representation of the EGG recording as original curves with the associated running spectrum 9

16 BASICS Fig. 4: Spectrum analysis pre and in the form of the running spectrum to determine the frequency distribution pattern and representation of the total spectrum pre and ... The frequency with the greatest power is defined as the dominant frequency. 10

17 MATERIAL AND METHODS 3. Material and methods: 3.1. Subject groups A total of 52 subjects were examined. They were divided into the following study groups: 1.) healthy subjects (n = 10) 2.) patients with diabetes mellitus, not kidney failure (n = 14) 3.) dialysis patients without diabetes mellitus (n = 13) 4.) dialysis patients Diabetes mellitus (n = 15) The patient groups were recruited from the general inpatient population of the Medical Clinic III of the West Palatinate Clinic in Kaiserslautern, as well as from the dialysis patients treated as inpatients at the clinic. All test persons had to have completed a fasting phase of at least 8 hours and be able to fully eat the breakfast given during the measurement. The examinations were therefore always carried out in the morning. All examined persons had a body mass index (BMI) of at least 18 kg / m 2 and at most 30 kg / m 2; cachectic or very obese patients were not included in the study. A prerequisite for all study participants was the negative of acute gastrointestinal complaints. Previous operations in the gastrointestinal tract (with the exception of appendectomy) and an existing pregnancy were also considered exclusion criteria. 11

18 MATERIAL AND METHODS Master data of the groups examined: Healthy subjects: initials Age (years) Gender BMI K.M. 46 W 24 G.E. 84 W 19 S.A. 31 M 27 T.N. 50 W 20 B.W. 40 M 24 H.M. 30 W 18 L.R. 61 W 21 T.O. 32 M 30 K.C. 34 W 20 H.O. 36 M 30 MW 44.40 23.40 Std Dev 17.05 4.62 All healthy test subjects denied taking medication regularly. In addition, all of them were non-smokers. 12th

19 MATERIAL AND METHODS Diabetic: Initials Age Gender BMI Diabetes Treatment F.W. 75 W 22 Insulin Therapy B.A. 43 M 20 insulin therapy M.L. 67 W 24 Insulin Therapy W.K. 66 W 25 insulin therapy L.M. 79 W 21 oral antidiabetic agents W.B. 67 M 23 oral antidiabetic agents R.O. 73 M 22 oral antidiabetic agents R.B. 67 M 23 oral antidiabetic agents J.H. 78 W 20 oral antidiabetic agents B.B. 53 W 25 insulin therapy H.R. 70 W 22 Insulin Therapy L.E. 54 W 23 oral antidiabetic agents P.P. 48 M 19 oral antidiabetic agents S.W. 77 M 20 oral antidiabetic agents MW 65.50 22.07 Std dev 11.59 1.90 13

20 MATERIAL AND METHODS Dialysis patients without diabetes mellitus: Initials Age Gender BMI Dialysis duration (months) K.W. 55 M 20 6 T. P. 71 M M.E. 56 W K.R. 63 W J.G. 75 M H.W. 65 M C.A. 70 M K.K. 67 M G.G. 70 W E.H. 75 M D.G. 66 M P. K. 83 M M.K. 64 M MW 67.69 23.0 hrs 7.67 2.24 The dialysis patients had to be treated with a stable dialysis regimen for at least 3 months. 14th

21 MATERIAL AND METHODS Dialysis patients with diabetes mellitus: Initials Age Gender BMI Dialysis duration (months) G.M. 55 W K.G. 64 W K.R. 53 W L. K. 56 M K.G. 71 M B.E. 70 M A.H. 63 M S.A. 38 M 24 9 W.A. 79 M G.H. 65 W 24 4 J.K. 72 M S.H. 58 M 28 7 L.F. 83 W B.M. 39 W S.R. 40 M MW 60.40 24.07 Std Dev 13.94 1.87 The examined diabetics requiring dialysis were all adjusted to insulin therapy with regard to their diabetes mellitus. 15th

22 MATERIAL AND METHODS 3.2. Device and software used: The examinations were carried out with the Polygraf ID data acquisition device from Medtronic. The computer-aided evaluation of the recorded data was carried out with the POLYGRAM-NET TM ElectroGastroGrapy application software, which is also a product of Medtronic Functional Diagnostics A / S. Fig. 5: Polygraph Fig. 6: Examination and evaluation unit 16

23 MATERIAL AND METHODS 3.3 Examination procedure: The examination was carried out with the patient lying down. The skin was roughened and degreased with a special peeling paste in order to reduce the skin resistance and thus the risk of artifacts. After applying and then drawing in an electrode gel, the electrodes were glued to the skin areas prepared in this way. The positioning of the electrodes followed a fixed scheme. The reference electrode is at the lower limit of the sternum, the 1st electrode is glued in the midline between the xiphoid process of the sternum and the navel, the 2nd electrode at an angle of 45 degrees approx. 4 6 cm next to it on the patient's left side 3. Shifted another 4 6 cm further to the left at an angle of 45 degrees upwards. A 4th ground electrode is at the same height as the first. For better recognition of artifacts, the breathing excursions and other movement influences are also derived via a movement sensor. Fig. 7: Schematic representation of the electrode position. 17th

24 MATERIAL AND METHODS After an impedance check in which the EGG device tests the skin contact resistance in the area of ​​each individual electrode, the actual measurement began. The potentials in the fasting phase were derived over a period of about minutes. Then the test person received a breakfast consisting of a roll with jam, a yogurt and a cup of coffee. Breakfast had to be eaten in a maximum of 15 minutes. To do this, the test person was allowed to sit up. The electrodes were not removed during mealtimes to prevent changes in position between preprandial and er recording. This food intake period was not taken into account in the later evaluation. The EGG was then registered for 45 minutes. After completion of the investigation, the original curve was first visually checked with the marking of possible artifacts. Periods marked as artefacts were excluded from the final analysis. Then the special software calculated the dominant frequency and the dominant power of the frequency spectrum as well as the frequency distribution pattern both pre and. 18th

25 MATERIAL AND METHODS Fig. 8: The marked artifact, here easily identified by a deflection of the motion sensor in the same direction (bottom curve), is removed from the calculation in the computer-aided analysis. 19th

26 MATERIAL AND METHODS 3.4. Statistical evaluation The statistical evaluation of the data was carried out with the Wilcoxon Sign Rank test for connected samples within the individual groups for the parameters surveyed. The analysis to distinguish between the different study populations was carried out with the Mann-Whitney U-Test. The SPSS software was used for the calculation

27 RESULTS 4. RESULTS 4.1. Evaluation of the dominant frequency The dominant frequency represents the frequency in periods per minute (cpm) that is associated with the dominant power in the overall spectrum. In the evaluation, this parameter was initially considered for each examination section, i.e. preprandially and separately, and checked for differences between the different examination groups. We then calculated the change in the dominant frequency that resulted after taking the test meal. For this purpose, this was evaluated individually within each group of test persons. 21

28 RESULTS The dominant frequency preprandial in all groups was in the normogastric frequency range between 2-4 cpm. The mean of all subjects was 3.02 (+ /) cpm. There were no measurably significant differences between the individual study groups. Comparison of the dominant frequency preprandial cpm Healthy diabetics Dialysis Diabetics requiring dialysis Frequency Fig. 9: Representation of the preprandial dominant frequencies of the individual study groups (mean values) 22

29 RESULTS The e dominant frequency was also in the normogastric frequency range in all groups. The mean value over all subjects is calculated with (+ /) cpm. There were no significant differences between the different groups of test subjects. Comparison of dominant frequency cpm healthy diabetics dialysis patients dialysis dependent diabetics frequency Fig.10: Representation of the mean values ​​of the dominant frequencies from the individual test subject groups. 23

30 RESULTS After taking the standardized test meal, the dominant frequency in the healthy subjects increased significantly within the normogastric range. A mean value of 2.96 (+ /) cpm was calculated preprandially. Postprandial it was (+ /) cpm. Frequency change / healthy P

31 RESULTS Preprandial and postprandial normogastric (3.06 (+ /) and (+ /)) fluctuations in potential were measured in all examined diabetics. Here, too, there is a significant increase in the dominant frequency after taking the test meal. Frequency change / diabetic P

32 RESULTS A significant increase in the dominant frequency could also be seen in the dialysis patients. A mean value of 3.0 (+ /) cpm was measured preprandially. Postprandial the mean value was (+ /) cpm change in frequency / dialysis patients P

33 RESULTS In the group of diabetics requiring dialysis, there was a significant increase in the dominant frequency from preprandial () cpm to () cpm. Here, too, the dominant frequencies are always within the ranges defined as normogastric. Frequency change / diabetics requiring dialysis P

34 RESULTS Summary: In all study groups there was a significant increase in the dominant frequency after taking the test meal. This moved both preprandially and within the so-called normogastric frequency spectrum between 2 cpm and 4 cpm. In the comparative evaluation of the examined subjects, no differences in the dominant output frequencies could be found. There were also no measurable differences. Ultimately, the extent of the frequency increase is to be regarded as the same in all groups. 28

35 RESULTS 4.2. Evaluation of the frequency distribution pattern In addition to the normogastric actions, potential fluctuations in other frequency ranges must always be measured. At the same time, there are also phases in which no clear frequency pattern can be recognized. These are called arrhythmic. In the evaluation, the proportion of the individual frequencies in the total time of the examination per examination phase was determined. First of all, the difference in the measured spectrum was analyzed preprandially and also between the different groups of test persons. Then the evaluation of the change in the frequency distribution by taking the test meal within the respective group was carried out. 29

36 RESULTS In the group of healthy volunteers, the percentage of normogastric myoelectric actions preprandially is significantly higher than in the examined patients. There is no discernible difference between diabetics, dialysis patients and diabetics requiring dialysis. Comparison of the proportion of normogastric actions preprandial P <% Healthy diabetics Dialysis patients Diabetics requiring dialysis Percentage Fig.15: Representation of the proportion of actions in the normogastric frequency range of the study groups preprandial. 30th

37 RESULTS Postprandially, the percentage of potential fluctuations in the range of 3-4 cpm is identical in all study groups; there are no significant differences. Comparison of the proportion of normogastric actions% healthy diabetics dialysis patients diabetics requiring dialysis percentage 50.75 50.136 49.169 58.36 Fig. 16: Representation of the percentage of normogastric frequency components of the study groups 31

38 RESULTS In healthy individuals, the majority (/%) of the measured actions are in the normogastric range. Postprandial does not change anything in the distribution pattern. Frequency distribution pre-u. / Healthy 100% 80% 60% 40% 20% 0% preprandial arrhythmic tachygastric bradygastric normogastric Fig. 17: Comparison of the frequency distribution pattern pre- and in healthy people percentage of normogastric actions% normal frequency preprandial normal frequency percentage 66.86 50.75 Fig. 18: The decrease in normogastric frequencies is not significant. 32

39 RESULTS Preprandially, arrhythmic fluctuations in potential are predominantly observed in the diabetic group. After taking the test meal, a rhythmization in favor of the normogastric frequency components can be observed. There are no noticeable changes in the area of ​​bradygastric and tachygastric actions. Frequency distribution pre-u. / Diabetics 100% 80% 60% 40% 20% 0% Preprandial Arrhythmic 60.221 40.757 Tachygastric 2.16 3.2 Bradygastric 6.13 5.91 Normogastric 31.521 50.136 Fig. 19: Comparison of the frequency distribution pattern pre- and percentage in diabetics normogastric actions P <0.05% normogastric preprandial normogastric percentage 31,521 50,136 Fig. 21: The increase in the percentage of normogastric actions is significant. 33

40 RESULTS Preprandial arrhythmic actions in the frequency spectrum predominate in dialysis patients (/%). Postprandially, normogastric actions in particular can be measured. Bradygastric and tachygastric proportions are not affected by mealtime. Frequency distribution pre- and dialysis patients 100% 80% 60% 40% 20% 0% preprandial arrhythmic 64.715 38.654 tachygastric 2.18 5.37 bradygastric 9.72 6.83 normogastric 23.385 49.169 . in dialysis patients the percentage of normogastric actions 100% P

41 RESULTS A clear increase in the percentage of normogastric actions can also be seen in the group of dialysis patients with concomitant diabetes mellitus. Frequency distribution pre-u. / Diabetics requiring dialysis 100% 80% 60% 40% 20% 0% preprandial arrhythmic tachygastric bradygastric normogastric Fig. 23: Comparison of the frequency distribution pattern pre- and in diabetics requiring dialysis percentage of normogastric actions% normogastric percentage 34.1 percentage percentage of normogastric 73.36 percentage of normogastric percentage 58.1 percent normalogastric Fig. 24: The increase in the normogastric percentage is significant. A significant change can also be seen in the bradygastric actions. 35

42 RESULTS Summary: While no significant differences in the distribution pattern between pre-u potential fluctuations were found in the healthy test persons, a significant change in the sense of a rhythmization with a higher proportion of normogastric actions was observed in all other groups examined after taking the test meal. Bradygastria and tachygastria are not influenced; a decrease in bradygastria was only observed in the group of diabetics requiring dialysis. 36

43 RESULTS 4.3. Evaluation of the dominant power The dominant power represents the dominant power of the analyzed signal. It depends on the amplitude of the output signal and the purity of the waveform of the original curve. For this parameter, too, a comparison was first made between the study groups in the individual study phases, then the power behavior within each group was analyzed. 37

44 RESULTS The mean value of the preprandial dominant power of all study groups was / db. The individual groups do not differ significantly. Preprandial power 60 db Healthy diabetics Dialysis patients Diabetics requiring dialysis Power (db) Fig. 25: The dominant power does not differ between the individual study groups preprandially. 38

45 RESULTS Postprandially, a value of / db was determined for all test subjects examined. When looking at the individual groups in this investigation phase, no measurable differences were found. e Power db Healthy diabetics Dialysis patients Diabetics requiring dialysis Power (db) 48.15 47.65 50.94 54.26 Fig. 26: There are also no significant differences between the different groups. 39

46 RESULTS Power behavior after taking the test meal. The preprandial dominant performance in the healthy subjects was calculated with / db. Postprandial the value was / db. The increase was not significant (p> 0.5). Power / Healthy db Preprandial Postprandial power (db) 46.56 48.15 Fig. 27: Change in dominant power in healthy subjects. 40

47 RESULTS In diabetics, the dominant power increased from / db preprandial to / db. The demonstrable increase in dominant power in diabetics is significant. Power / Diabetic P

48 RESULTS Dialysis patients also show a significant increase in dominant power from preprandial to. At a baseline of / db before taking the test meal, a performance of / was measured. Power / dialysis patients P

49 RESULTS The electro-gastrographic signals of diabetics requiring dialysis show a significant change in the dominant power in the sense of an increase in performance. There is an increase of / - 5.8 db to / db. Power / diabetics requiring dialysis 60 P

50 RESULTS 4.4. Evaluation of the preprandial power spectrum in the healthy subjects, the performance moves mainly in the normogastric range in contrast to the patient groups. Here the power predominates in the bradygastric frequency component. Power share in the normogastric range% p = P

51 RESULTS The percentage of power in the tachygastric area does not differ with /% in healthy people, /% in diabetics, / or / in diabetics requiring dialysis. Postprandially, there are no significant differences in the power distribution between the different study groups. 45

52 RESULTS After taking the test meal, the proportion of performance in the bradygastric frequency range increases. However, this increase is not significant. Power distribution / healthy 100% 80% 60% 40% 20% 0% preprandial tachgastric bradygastric normogastric Fig.33: Power distribution pattern preprandial and in healthy postprandial diabetics there are no significant differences in the power distribution pattern compared to the preprandial phase. Power distribution pre-u. / Diabetics 100% 80% 60% 40% 20% 0% Preprandial Tachygastric Bradygastric Normogastric Figure 34: Power distribution pattern preprandially and in diabetics 46

53 RESULTS Postprandially, the proportion of performance in the bradycardial frequency range in dialysis patients is significantly lower than; the rest of the performance spectrum does not change directionally. Power distribution pre-u./ Dialysis patients 100% 80% 60% 40% P <% 0% preprandial tachgastric bradygastric normogastric Fig.35: Power distribution pattern preprandial and in dialysis patients Diabetics requiring dialysis show a significant increase in power in the normogastric area. Power distribution in pre- and / or diabetics requiring dialysis 100% 80% 60% 40% 20% P <% prepandial tachygastry bradygastry normogastry Fig. 36: Power distribution pattern preprandially and in diabetics requiring dialysis 47

54 RESULTS Summary: There was a clearly noticeable increase in dominant power in all study groups. However, this is not statistically significant in the healthy test subjects. The level of power did not differ between the groups in either phase of the investigation. When examining the distribution of power over the defined frequency ranges of normogastric, bradygastric and tachygastric preprandially, a preponderance in the bradygastric range was found in all patient groups, while the healthy subjects showed the greatest proportion of power in the normogastric range. There are no significant differences in postprandial terms. 48

55 RESULTS 4.5. Evaluation of the slow wave coupling With the calculation of the slow wave coupling percentage, in a given period the preprandial or the percentage of time at which the frequency with which the gastric slow waves propagate between two adjacent electrodes is less than 0 , Is 2 cpm. This is a measure that records the transmission of the myoelectric signal within the cell syncytium. 49

56 RESULTS There is no significant difference in the percentage of potential fluctuations that are registered as temporally coupled in healthy volunteers from preprandial. slow wave coupling / healthy people% preprandial% 60.67 55.32 Figure 37: Percentage of slow wave coupling prepandial and in healthy people In diabetics, the percentage of slow wave coupling increases significantly when the test meal is taken. slow wave coupling / diabetics 100% P

57 RESULTS There are no significant differences in slow wave coupling between preprandial and dialysis patients. slow wave coupling / dialysis preprandial% 51.3 51.492 Figure 39: Percentage of slow wave coupling preprandial and in dialysis patients Even diabetics requiring dialysis show no statistically clear differences in slow wave coupling after taking the test meal. slow wave coupling / pre-prandial diabetics subject to dialysis% 54.057 57.743 Fig. 40: Percentage of pre-prandial and post-prandial slow wave coupling in diabetics subject to dialysis. 51

58 RESULTS Summary: There are no significant differences in the percentage of slow wave coupling between the individual study groups and also within the respective groups preprandially or after taking the test meal. A significantly higher percentage of slow wave coupling can only be observed in diabetics. 52

59 DISCUSSION 5. Discussion The assessment of the myoelectric activity of the stomach by means of electrogastrography has been investigated for a wide variety of clinical pictures since it was first described by ALVAREZ in 1922. After a phase of oblivion, the method is now experiencing growing interest, especially through the development of computer-controlled evaluations (JONDERKO, 2005). The reasons for this are certainly the easier interpretation options for the signal, combined with the advantage of having a non-invasive, easy-to-perform examination option available. What is still unclear, however, is the exact statement of the individual parameters recorded during an examination with regard to their clinical significance. Meaning of the dominant frequency: The most easily understood parameter is the so-called dominant frequency. The dominant frequencies in all of the groups we examined were in the normogastric range, i.e. between 2 4 cpm. There was also a significant increase in the dominant frequency after taking the test meal. This behavior has been well documented in numerous studies for healthy volunteers. So describe e.g. LEVANON et al in a study of 24 healthy people aged 22 to 91 years showed a clear increase in the dominant frequency after taking the test meal.Similarly, SIMONIAN et al in a multicenter study on a total of 61 healthy test persons saw a significant increase in the dominant frequency of 2.98 ( + /) to 3.08 (+/- 0.04) cpm in the first hour. With regard to possible pathological changes in gastric emptying, different results can be read in the literature with regard to the increase in the dominant frequency after taking a test meal. 53

60 DISCUSSION JEBBINK (1994) e.g. saw no differences in myoelectric activity in well-controlled diabetics, regardless of the severity of symptoms of existing neuropathy. In contrast to this, and also to our results in 1997, LIN did not describe any increase in the dominant frequency from preprandial to with chronic kidney disease. But here, too, the dominant frequencies measured were in the normogastric range, he also saw no significant increase in the dominant frequency in a study with patients with functional dyspepsia. In a study of the myoelectric activity of the stomach in healthy children and those with functional dyspepsia by CHEN (1998), normogastric dominant frequencies were found in both healthy and sick children. The frequency increase seen in both study groups within normogastrics, however, turned out to be insignificant in the patients. Ultimately, the dominant frequency and its increase after taking a test meal does not seem to be a sufficiently meaningful parameter to differentiate an autonomic neuropathy of the stomach. 54

61 DISCUSSION 5.2. Significance of the frequency distribution pattern What is striking in our investigations is an overall low percentage of normal-frequency actions in favor of a high proportion of arrhythmias in the overall spectrum, especially preprandial. This percentage is somewhat below the values ​​given in the literature. One of the reasons for this could be the schematic layout of the electrodes, which are not, e.g. recommended by MIRIZZI (1983) and placed over the antral axis in a sonographically controlled manner. This makes the signal weaker and more prone to artifacts, and the risk of measurement inaccuracies increases. This was accepted in favor of the desire to have a method available that can be carried out at the patient's bedside during dialysis treatment without great effort for the patient. The small number of healthy subjects examined also plays a role, since few of the examined with strongly arrhythmic gastric potentials had a comparatively strong influence on the statistical result. Postprandially, we saw a clear rhythmization in the sense of an increase in the proportion of normogastric actions in the overall spectrum. In its 2004 multicenter study, SIMONIAN found a proportion of normogastric potential fluctuations in the frequency spectrum of 77 +/- 3% in healthy people during the fasting period. Postprandially, this was 81 +/- 2%. (for comparison / or / in our test group). In a comparative study between healthy volunteers and patients with chronic renal insufficiency, LIN (1997) sees a percentage of / - 2.5% preprandially and / - 1.8% in its healthy subjects. As in ours, the group with renal failure had a significantly lower proportion of normal-frequency actions preprandially, regardless of whether the patients were under a co-existing 55

62 DISCUSSION Suffering from diabetes mellitus or not. In the study mentioned, however, the clear rhythmic increase in the consumption of a test meal was not observed, which we were able to demonstrate in all 3 patient groups (diabetics, dialysis patients and diabetics requiring dialysis). KO (1998) also sees a low percentage of normogastric actions preprandially, but he measures an increase after taking the test meal. What is remarkable in this study is the influence of the time of measurement. KO found significant differences when taking the measurement 1 hour after completing dialysis treatment or before dialysis. After dialysis, the electrogastrogram obtained was significantly more arrhythmic, and the extent of bradygastria increased in particular. KO blames complex pathophysiological reversible changes caused by dialysis for this. He mentions factors such as Disorders in the hormone or electrolyte balance. SIRINEK (1984) describes changes in gastrin, insulin, potassium, calcium, urea and creatinine after 4 hours of hemodialysis treatment. In 1994, JEBBINK showed that hyperglycaemia can induce gastric myoelectric arrhythmias. Our examinations were carried out exclusively while hemodialysis was ongoing. This may also partly explain the high level of arrhythmias, as the changes in the electrolyte balance mentioned by KO are more intense during dialysis than after a compensation period of 1 hour post-dialytic. Simultaneous measurements of the above-mentioned laboratory parameters in the course of dialysis could provide information on this. KAWAGASHI (1997) examined diabetics with and without proven autonomic neuropathy. He found a significantly lower rate of frequencies in the normogastric range (37 +/- 4%) in patients with autonomic neuropathy than in patients without neuropathy (/ - 5.2%). MAYAUDON described something similar in 1999 in asymptomatic diabetics with proven cardiac autonomic neuropathy. He examined 51 type 1 diabetics with no gastrointestinal 56

63 DISCUSSION complaints. The percentage of regular normogastric actions was significantly lower than that of a healthy comparison group. In the study mentioned, tachygastric actions were particularly noticeable. In contrast to Jebbink, Mayaudon could not demonstrate any influence from hyperglycaemia. In a measurement of a total of 10 type 1 diabetics by MANTIDES (1997), significantly more dysrhythmias were found in asymptomatic patients compared to an equally large healthy control group. PFAFFENBACH (1998) found no disorders of the myoelectric gastric activity in diabetics despite dyspeptic symptoms. In contrast, he found more tachygastria in hyperthyroid patients. A work by BARCZYNSKI (2001) is also interesting in this context. He examined electro-gastrographically patients with hyperthyroidism and after correction for the euthyroid metabolic situation as a model of a reversible autonomic neuropathy. In the hyperthyroid metabolic situation, an increased rate of dysrhythmias, especially bradygastria, was noticed. After normalization of the metabolic situation, the arrhythmias disappeared in favor of a normal EGG course. The presence of the autonomic neuropathy was checked at the same time using cardiac parameters (heart rate variability). We have not performed any tests in our patient groups for extra-gastric manifestations of a possible existing autonomic neuropathy. With preprandial significantly reduced normogastric frequency proportions between and% in favor of dysrhythmic actions, the presence of a manifest neuropathy appears at least likely in the patients examined by us, taking into account the studies by KAWAGASHI, MAYUAUDON and BARCZYNSKI Diseases described. In 1998, CHEN showed a significant decrease in arrhythmias by taking a test meal in children with functional dyspepsia. PFAFFENBACH (1997) also shows a preprandial change in the frequency spectrum at the expense of regular actions in adult patients aged 57