Scandinavian Journal of Gastroenterology
ISSN: 0036-5521 (Print) 1502-7708 (Online) Journal homepage: http://www.tandfonline.com/loi/igas20
Influence of Stress on the Healing and Relapse of Duodenal Ulcers: A Prospective, Multicenter Trial of 2109 Patients with Recurrent Duodenal Ulceration Treated with Ranitidine G. Holtmann, D. Armstrong, E. Pöppel, A. Bauerfeind, H. Goebell, R. Arnold, M. Classen, L. Witzel, M. Fischer, M. Heinisch & A.L. Blum To cite this article: G. Holtmann, D. Armstrong, E. Pöppel, A. Bauerfeind, H. Goebell, R. Arnold, M. Classen, L. Witzel, M. Fischer, M. Heinisch & A.L. Blum (1992) Influence of Stress on the Healing and Relapse of Duodenal Ulcers: A Prospective, Multicenter Trial of 2109 Patients with Recurrent Duodenal Ulceration Treated with Ranitidine, Scandinavian Journal of Gastroenterology, 27:11, 917-923, DOI: 10.3109/00365529209000163 To link to this article: http://dx.doi.org/10.3109/00365529209000163
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 21
View related articles
Citing articles: 3 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=igas20 Download by: [New York University]
Date: 10 April 2016, At: 18:56
Influence of Stress on the Healing and Relapse of Duodenal Ulcers A Prospective, Multicenter Trial of 2109 Patients with Recurrent Duodenal Ulceration Treated with Ranitidine G . HOLTMANN, D. ARMSTRONG, E. POPPEL, A. BAUERFEINDt, H. GOEBELL, R. ARNOLD, M. CLASSEN, L. WITZEL, M. FISCHER, M. HEINISCH, A. L. BLUM & MEMBERS O F THE RUDER STUDY GROUP Division of Gastroenterology, University of Essen, Essen; Institute of Medical Psychology, University of Munich, and Dept. of Gastroenterology, Technical University, Munich; Division of Internal Medicine, University of Marburg, Marburg; Division of Internal Medicine, DRK Hospital, Berlin; and Institute of Numerical Statistics, Cologne; Germany; and Division of Gastroenterology, CHUV, Lausanne, Switzerland
Downloaded by [New York University] at 18:56 10 April 2016
Holtmann G , Armstrong D , Poppel E , Bauerfeind A, Goebell H, Arnold R, Classen M, Witzel L, Fischcr M. Heinisch M, Blum AL, Members of the RUDER Study Group. Influence of stress on the healing and relapse of duodenal ulcers. A prospective, multicenter trial in 2109 patients with recurrent duodenal ulceration treated with ranitidine. Scand J Gastroenterol 1992;27:917-923. The influence of psychologic factors on the healing and relapse of duodenal ulcers under treatment with ranitidine was studied in a prospective, multicenter trial in 2109 patients with an endoscopically proven duodenal ulcer (DU) and a history of recurrent duodenal ulceration. All patients received ranitidine (300 mg daily), and, after healing, 1899 patients continued maintenance treatment (ranitidine, 150 mg daily) for 2 years. A physician's assessment of stress (stress or no stress) was made at every consultation. In the healing phase an overall classification of stress as absent, intermittent, or continuous was made, and in the maintenance phase patients were classified dichotomously as having stress (stress on at least half of the follow-up consultations) or no stress. In addition, at the start of the healing phase stress was measured by means of a standardized questionnaire. Continuous stress, as assessed by the physicians, was associated with a lower 14-day healing rate (35.7%) than intermittent or absent stress (42.4%; relative risk (RR) for delayed healing in patients with continuous stress, 1.19; 95% confidence interval (Ci), 1.06-1.33; P < 0.02). Differences in the 14-day healing rate for patients with low and moderate stress scores (43.1%) compared with those with high and very high stress scores (37.9%) just failed to reach statistical significance (RR for patients with stress, 1.14; 95% CI, 0.99g1.29; P = 0.051). During the 1st year of maintenance treatment 18.3% of patients with stress, but 10.9% of patients without stress, had a DU relapse (RR of stress for DU relapse during the first year, 1.73; 95C1, 1.44-2.09; P < 0.001). In contrast, during the 2nd year a significant difference in the relapse rates for patients with (9.7%) and without stress (11.6%) was not found (RR, 0.84; 95% C1, 0.61-1.15; P>0.4). Thus, on the basis of the physicians' assessment, continuous stress delays duodenal ulcer healing, and stress increases relapse rates during the 1st but not during the 2nd year of maintenance therapy with an HZreceptor antagonist.
Key words: Peptic ulcer; psychologic factors; psychologic stress; ranitidine; relapse Gerald Holtmann, M . D . , University of Essen, Dept. of Internal Medicine, Dioision of Gastroenterology, Hufelandstr. 55, 0-4.300 Essen 1 , Germany
The close relationship between the brain and gastrointestinal functions was described more than 150 years ago by W. Beaumont (l),and since then several uncontrolled (2-7) and controlled (8-12) studies have been published indicating that acute mental stress affects the secretion and motility of the upper gastrointestinal tract. Furthermore, there is evidence of an association between the occurrence of duodenal ulcers and psychologic stress. Some anecdotal reports describe a coincidence of stressful life events and peptic ulcer disease (1S1.5). In contrast, controlled studies have yielded contradictory results: Sapira & Cross (16) and Magni et al.
t Deceased.
(17) found more stressful life events and more psychologic distress in patients with duodenal ulcers than in controls, whereas Piper et al. (18, 19) found little or no relationship between stressful life events and peptic ulcer disease. More recently, Feldman et al. (20) and Walker et al. (21) have shown that patients with peptic ulcer do not differ from nonulcer controls with regard to the number of life events, but peptic ulcer patients perceived life events more negatively than did controls. However, little is known about the influence of life stress on the healing and relapse of duodenal ulcers in patients treated with H2-receptor antagonists. Although there is some evidence that stress may delay duodenal ulcer healing (22, 23), there have been no large con-
Downloaded by [New York University] at 18:56 10 April 2016
918
G. Holtmann et al.
trolled, prospective studies. In particular, there is little information about the effect of stress on duodenal ulcer relapse. This study examines the influence of psychologic factors on the healing and relapse of duodenal ulcers in patients treated with the widely prescribed H2-receptor antagonist ranitidine. A study population of at least 2000 patients was necessary to ensure that any possible effect of stress and the concomitant effect of other risk factors on healing and relapse would not be overlooked. This necessitated a large multicenter study with close and regular monitoring to ensure that patients were not lost to follow-up during the maintenance phase. The following hypotheses were tested: 1) psychologic stress delays the healing of duodenal ulcers treated with 300 mg ranitidine daily; 2) stress increases relapse rates over a 2-year follow-up period during maintenance treatment with 150 mg ranitidine daily. MATERIALS AND METHODS The criteria for inclusion in the study were an endoscopically proven duodenal ulcer (DU) with a diameter greater than 5mm, a history of chronic DU disease, and at least one duodenal ulcer manifestation within the previous year before the patient had a relapse that was treated in accordance with the study protocol, age over 18 years, and no ulcer-related complication requiring surgical treatment. In addition, patients fulfilling the following criteria were excluded from the study: renal insufficiency, gastric ulcer, pregnancy or lactation period, alcohol abuse (>60 g pure alcohol per day), or evidence of drug abuse.
Healing phase On the 1st day all patients with an endoscopically verified DU gave written informed consent to their inclusion in the study. Data concerning the history, socioeconomic status, and social situation were recorded, and treatment was started with 300 mg ranitidine taken at 1800 h. After 14 days of treatment all patients underwent a second upper GI endoscopy. If the ulcer had healed, the patients were enrolled in the maintenance phase of the study, and treatment was started with 150 mg ranitidine taken at 1800 h. If healing was not complete, patients continued to receive 300 mg ranitidine until complete ulcer healing was documented endoscopically; it was proposed that the repeat endoscopies should be performed 4 and 8 weeks after the start of the healing phase, but this was not obligatory, and, in general, they were performed at the discretion of the physician, when he/ she believed clinically that the ulcer had healed. Patients whose ulcers healed within 2 weeks were classified as fast healers, and all other patients as slow healers. In 19 patients there was a persistent ulcer (after 8 or 3 1 2 weeks of treatment). In accordance with the study protocol, these patients were classified as non-responders, and a new treatment
regimen (for example, gastric surgery), independent of the study protocol, was started. Data from 2109 patients under the care of 546 general practitioners or gastroenterologists performing upper GI endoscopy in outpatients were collected for the healing period. Study monitoring was performed by 48 physicians located at regional hospital centers: 5 study chairmen, 12 professional monitors in a data evaluation center, and 4 biostaticians who coordinated action in case of protocol violation. For statistical analysis data from 1923 patients (91.2%) were used; data from the remaining 186 patients could not be used for one or more of the following reasons: patients dropped out before healing was documented ( n = 66), no history of previous peptic ulcer disease (n = 61), ulcer diameter less than 5 mm at the time of diagnosis ( n = 39), treatment with ulcer-healing drugs other than ranitidine ( n = 23), alcohol abuse ( n = 6), and other reasons ( n = lo). The patient characteristics are summarized in Table I.
Maintenance phase During the 2-year follow-up period the patient attended every 3 months for follow-up appointments. Upper GI endoscopy was performed immediately in the case of duodenal ulcer symptoms and, in all cases, at the end of the 1st
Table I. Sociodemographic characteristics of the study population; data on 1923 patients were available (mean 2 SEM, range or number of patients (%) with the particular characteristic) Age (years) Gender* Men Women Weight (kg) Height (cm) Marital status Married Divorced Single Partnership Widowed Age at begin. of duodenal ulcer disease (years) Years since the first manifestation Patients with >3 recurrences within the past 2 years Complication noted in the records or recalled by the patient Perforation Bleeding Pyloric stenosis Other Smokers Ex-smokers Non-smokers
48 19-90 1318 601 73 34-120 172 144200 1444 131 174
89 85 36 12-86 12
f 13
(68.5%) (31.3%) f12 28
(75.1% (6.8% (9.0% (4.6% (4.4% 213 2 10
63% 38 296 78 23 1028 431 464
(2.0% (15.4% (4.1% (1.2% (53.5% (22.4% (24.1%
* Data available for 1919 patients; otherwise data available for all 1923 patients.
Stress and Healing and Relapse of DU
Downloaded by [New York University] at 18:56 10 April 2016
and 2nd year. In the case of a relapse the dose of ranitidine was increased to that used during the healing period (300 mg daily) until healing was confirmed, at which time the maintenance dosage was resumed. For the 1st year of follow-up, data from 1899 and for the 2nd year data from 1671 patients were available (87% of patients with a documented healing completed the 2-year follow-up). Patients with 2-year followup were not different from those patients who did not complete the study with regard to sociodemographic data, the presence of risk factors for ulcer relapse, and the course of the disease (for example, duration of the disease or the number of relapses before the study). Assessment of psychologic stress When the diagnosis of duodenal ulcer had been confirmed (day l), psychologic stress was measured by means of a standardized questionnaire consisting of 20 items (Table 11). The patient scored each item from 0 to 3 depending on the extent to which the statement described his (or her) feelings of stress (grade 0 = none; grade 1 = slight; grade 2 = moderate; grade 3 = high). Thus, the minimum possible score was 0, and the maximum was 60. The questionnaire score was then used to classify the patients as having no or low (score S l S ) , moderate (score 16-25), high (score 2 6 35), or very high (score >35) stress. This questionnaire has
Table 11. Items of the stress questionnaire Item no. 1. 2.
3. 4.
5. 6. 7. 8. 9. 10.
11. 12. 13.
14.
15. 16. 17. 18. 19. 20.
I experience loneliness In the past I have been successful in realizing my aims and wishes* I try to foresee unpleasant events, to avoid them My everyday life is monotonous I always have good experiences with my fellow human beings* My job is considered to be strenuous My family life is harmonious* For some time past there have been strokes of fate Sometimes I have a premonition that gets me in a state I calm down quickly after excitement* I always try to work out problems in new ways* I don’t believe that my environment and problems will result in sickness* In our society the individual is usually not allowed to show his feelings I am looking forward to events in the near future; I am really pleased* Whenever I overcome a problem, the next problem follows I expect to recover from my disease* Unknown situations or unexpected events create disturbance and tensions At this time I am in serious economic and financial difficulties My apartment fulfills all my needs’ There are opponents I have to protect myself from to prevent harm
* Low scores on these items indicate presence of stress; stress scores were calculated accordingly.
919
been specifically developed and validated for this study. Cross-validation with the Freiburg Personality Inventory (FPI) (24) indicated that high scores on this stress questionnaire were associated with low life satisfaction (scale 1 of the FPI, r = -0.60, P < 0.01) and intense life strain (scale 7 of the FPI, r = 0.48, P < 0.01). At each visit the practitioner had to answer the question as to whether the patient was stressed at the time of the follow-up visit. There were no criteria given regarding what should be considered ‘stressed’. Therefore, the individual classification of stressed or not stressed was based on the physicians’ concepts and perception of stress rather than somehow standardized criteria. The practitioners’ estimation of stress was repeated at every scheduled follow-up appointment during the healing and follow-up periods. Furthermore, the treating physician did not have a key for evaluation of the stress questionnaire. They gave their assessment of stress without knowledge of the stress value described above. In the healing phase, therefore, patients were classified as having no stress, intermittent stress (stress at some but not at all appointments), or continuous stress (stress at all appointments). During the maintenance phase patients were classified as having stress if the physician considered them to be stressed on at least half of their follow-up appointments; all other patients were classified as having no stress. Statistical analysis In the healing phase the effect of stress was evaluated separately for the physicians’ assessment of stress and for the results of the self-administered questionnaire. For the physicians’ assessment, the numbers of fast healers (patients who had healed completely within 2 weeks) in each of the three groups (continuous, intermittent, and no stress) were compared by means of Fisher’s exact test (two-tailed) (25, 26). For the self-administered questionnaire the numbers of fast healers in the four groups (low, moderate, high, and very high stress) were compared, using again Fisher’s exact test (two-tailed). In the maintenance phase, the numbers of patients who had relapsed in the two stress groups (stress or no stress) were compared using Fisher’s exact test (twotailed). A significant level of 5% was assumed when testing the hypotheses of the effects of stress on healing and relapse. Statistical analysis was conducted using the Statistical Package for Social Sciences (SPSS) and Biomedical Computer Programs (BMDP). RESULTS Healing phase Overall healing rates. After 2 weeks of treatment healing was found in 39.5% of the patients; these patients were classified as fast healers. Cumulative healing rates after 4 and 8 weeks were 70.9% and 93.2%, respectively. No significant differences in the healing rates were found for male and female patients.
920
G . Holtmann et al.
Table 111. Comparison of the stress score and treating physicians’ estimation of stiess during the healing phase (includes i609.patients with data on stress and physicians’ assessment of stress)
Downloaded by [New York University] at 18:56 10 April 2016
Practitioners’ estimation Stress score
Without stress
With stress
Score < 15 (low), n = 298 15 < Score < 25, (moderate), n = 796 25 < Score s 35 (high), n = 392 Score < 35 (very high), n = 123
194 (65.1%)
104 (34.9%)
338 (42.4%)
458 (57.6%)
90 (23.0%)
302 (77.0%)
7 (5.7%)
116 (94.3%)
Assessment of psychologic stress: practitioners’ assessment. On the basis of the physicians’ assessment, 740 (38%) of the 1923 patients were classified as having no stress. The remaining 1183 patients were classified as being stressed: in 367 (19% of total) the stress was intermittent, and in 816 (42%) it was continuous. Stress questionnaire (patient self-assessment). Stress scores were available for 1609 patients; these showed that , patients 298 patients had low stress (test scores ~ 1 5 )796 had moderate stress (score 1&25), 392 patients had high stress (score 26-34), and 123 patients had very high stress (Table 111). Comparison of the two assessment procedures showed that 94.3% of patients with very high stress (>35) were also classified by their practitioner as being stressed (intermittent or continuous). However, only 65.1 % of
50
r
I
without
patients having low stress on the questionnaire (G1.5) were as having no stress. gls0 classified b y their Overall, patients with no stress according to their practitioners had a mean questionnaire score of 19.6 6 (LSD); patients with intermittent stress had a mean score of 22 2 7, and those with continuous stress had a mean score of 26 8. Psychologic stress and healing rates: practitioners’ assessment. The 14-day healing rates in patients with continuous stress (35.7%) was significantly lower than in patients with intermittent stress (42.8%; P < 0.05) or without stress (42.2%; P < 0.02). Therefore, the relative risk (RR) of patients with continuous stress as compared with patients without and with intermittent stress (14-day healing rate, 42.4%) for delayed healing is 1.12 (95% CI, 1.04-1.20; P < 0.02). There was no significant difference in the healing rates between those with intermittent stress and those without stress (Fig. 1). Furthermore, there was no statistically significant difference in the presence of risk factors for slow ulcer healing (for example, smoking, non-steroidal antiinflammatory drug intake) between subjects with and without stress. Ulcer healing occurred within 14 days in 37.7% of smokers and ex-smokers, in contrast to 45.3% of nonsmokers ( P < 0.001). Stressed smokers had a 14-day healing rate of 34.9% as compared with 38.4% in smokers without stress ( P < 0.05). There was no significant difference in the 14-day healing in non-smokers with and without stress (41.8 versus 45.6; P 2 0.4). N o association between stress and smoking habits or changes in the smoking habits was found. Stress questionnaire. In patients with low or moderate stress scores, the 14-day healing rate was 43.1%, compared with a 14-day healing rate of 37.9% in patients with high
*
**
intermittent
continuous
Stress Fig. 1 . Percentage of each stress group (practitioners’ assessment) who were healed within 14 days ( * P < 0.05, * * P < 0.01).
Downloaded by [New York University] at 18:56 10 April 2016
Srress and Healing and Relapse of DU
921
"
low
moderate
high
very high
Intensity of stress (stress score) Fig. 2. 14-day healing rates and stress (results of the standardized questionnaire, P = 0.051 for the comparison of subjects with low stress and moderate versus high and very high stress). and very high stress scores (RR for slow healing in patients with high and very high stress scores as compared with patients with low and moderate stress scores, 1.14; 95% CI, 0.998-1.29; P = 0.051). In Fig. 2, 14-day healing rates for patients with low, moderate, high, and very high stress scores are shown.
Maintenance phase (2-year follow-up study) Relapse rates During the 1st year of the follow-up, duodenal ulcer relapses were documented in 247 patients (13.0%; 95% CI, 11S 1 4 . 4 ) . Forty-nine patients with relapse and 179 patients without relapse during the 1st year did not participate in the 2nd year of the follow-up. Thus, 198 patients with a relapse during the first year and 1473 patients without a relapse were studied during the 2nd year of maintenance treatment (= 1671 patients). During the 2nd year 11.7% (95% CI, 10.313.3) of patients without relapse during the 1st year of maintenance treatment had a relapse. In addition, relapses occurred in 78 of 198 patients with a relapse during the 1st year. Differences in the presence of risk factors for DU relapse and sociodemographic data for patients with and without participation in the long-term treatment phase were not found. Stress and relapse rates During the 1st year of the follow-up the relapse rate in patients with stress (stress on at least half of the follow-up consultations) was 18.9%, compared with 10.9% in patients without stress (RR of stress for D U relapse during the first year, 1.73; 95% CI, 1.44-2.09; P < 0.001); during the 2nd year 44 of 454 (9.7%) patients with stress relapsed, compared with 141 of 1217 (11.6%) patients without stress (RR of
stress for DU relapse during the 2nd year of maintenance treatment, 0.84; 95% CI, 0.61-1.15; P > 0.4). Cumulated relapse rates for patients who completed 2 years of maintenance treatment were 120 of 454 (26.4%) with stress and 263 of 1217 (21.6%) patients without stress (RR of stress for DU relapse during 2-year maintenance treatment, 1.22; 95% CI, 1.01-1.47; P < 0.05). The significantly higher relapse rate of patients with stress during the 1st year of maintenance treatment was independent of the presence or absence of stress during the healing period (Fig. 3). Differences in the presence of risk factors for ulcer relapse (for example, smoking behavior) were not found for subjects with and without stress. Relapses occurred in 20.4% of non-stressed but in 30.5% of stressed smokers ( P < 0.001). DISCUSSION There is a widespread belief that psychologic factors are involved in the pathogenesis of peptic ulceration. Thus, psychologic stress is accused of delaying the healing and of promoting relapses of duodenal ulcers. This assumption is consistent with controlled and uncontrolled studies showing that stress can alter both gastric acid secretion and gastroduodenal motility (1-8, 12, 27). However, there are very few pertinent studies, and previous studies have yielded conflicting results (13, 1&21, 28). The present prospective, multicenter study was conducted to examine, in a large and inhomogeneous population, the influence of stress on the healing of DU. Identical treatment for all patients was chosen and consisted of curative treatment with ranitidine (300 mg daily). Subsequently, maintenance treatment with 150 mg ranitidine daily was given for 2 years. The results of
Downloaded by [New York University] at 18:56 10 April 2016
922
G . Holtmann et al.
0 Healing phase: without stress phase: MainteMnce
without stress
with stress
with stress
without stress
without Stress
with stress
with Stress
Fig. 3. InRuence of stress (practitioners’assessment) on the 1-year relapse rate (***P< 0.001 for the comparison of subjects with and without stress during the 1st year of the maintenance phase) this study confirm the two hypotheses tested. First, patients considered by their physicians to be under continuous stress in the healing phase had a slower DU healing than patients who were not stressed or patients with intermittent stress. Second, during the 2-year maintenance treatment patients with stress had a significantly higher relapse rate than patients without stress. Even though stress, as assessed by the physician, was associated with delayed ulcer healing and DU relapses, effects of stress as measured by the questionnaire just failed statistical significance (RR, 1.14; 95% CI, 0.998-1.29; P = 0.051). This finding suggests that physicians, when they assess the patients’ stress, are taking more into consideration than the internal stress level that the standardized questionnaire was designed to measure. Even though patients with high stress scores in general were classified by the physician as stressed, the physician’s judgement of stress is actually a composite assessment of the patient’s current lifestyle and life situation, his recent experience of life events, and his response to stress; it may even be that the physician is subconsciously drawing conclusions on the basis of his clinical experience of patients who are likely to respond poorly to treatment and heal slowly or relapse rapidly. Thus, the outcomes of the physicians’ assessment and the questionnaire measurement should not be seen as mutually contradictory: rather, they suggest that the questionnaire measures a specific facet of the patient’s stress (in particular, as indicated by the results of the questionnaire cross-validation, both strain and life satisfaction in general), whereas the physician is able to provide an even more general assessment of the extent of the patient’s stress status. Nevertheless,
some between-physician variation in the assessment of stress will result from the present approach to assessment. The stress assessment of the present study, however, reflects what the participating physicians considered to be stress. The relapse rate was also related to the presence or absence of stress in the healing phase as measured by the questionnaire. The overall relapse rate during the 1st year of maintenance treatment was higher in patients with low stress scores during healing, particularly in those who were stressed (physicians’ assessment) during maintenance treatment (23.8%). A possible explanation of this finding is that low stress scores might detect not only those patients who are genuinely unstressed and, thus, have a low risk of relapse, but also those stressed patients whose repressive coping strategies lead them to deny their stress. However, this repressive coping strategy seems to be associated with an increased risk for a relapse. There are several possible reasons why this study has detected an effect that has not been seen in previous studies. The most important reason is that all previous studies have examined relatively small numbers of patients. Studies with relatively small numbers of subjects may fail to show an effect if this effect on healing or relapse is small. It is also possible that stress and psychologic factors are important in some but not all subjects and therefore might be overlooked in small samples. Another reason for not seeing an effect may be the fact that psychologic stress is not a dichotomous variable: some patients are more stressed than others, and intensity of stress may change with time. The ability to cope with psychologic strains differs from individual to individual and might influence the effect of stress. In addition, psy-
Stress and Healing and Relapse of DU
Downloaded by [New York University] at 18:56 10 April 2016
chologic stress is not just an accumulation of stressful events; it is a multifactorial influence in which the stress is d u e to a variable number of events whose effect might be modified by other factors such as experience, social circumstances, and personality (29-33). Thus, it is not only t h e life events themselves but also the various modifying factors which determine the individual response t o potentially stressful life events. Recently, Ellard et al. (36) emphasized t h e importance of t h e duration and quality of the stress. Chronic stressors (those of 6 or more months’ duration) were significantly and independently associated with the onset and
relapse of DU In conclusion, psychologic factors affect the healing and the relapse rate of DU. T h e effect of stress on healing and relapse rate is independent of other risk factors such as smoking. Furthermore, the results suggest that a simple assessment of stress by the physician may identify those patients who will heal more slowly and have an increased risk for a DU relapse. ACKNOWLEDGEMENTS This comprehensive study has been financed by Cascan & Co. KG Wiesbaden, Germany, and SNF grant 32-26369 89; the essential administrative management was handled by Cascan and IFNS (Institut fur Numerische Statistik) under t h e supervision of the chairman. REFERENCES 1 . Beaumont W. Experiments and observations on the gastric juice and the physiology of digestion. Plattsburgh: F. P. Allen, 1833. 2. Bennett TI, Venables J. The effects of the emotions on gastric secretion and motility in the human being. Br Med 3 1920;2.6623. 3. Wolf S, Wolff HG. An experimental study of a man and his stomach. 2nd ed. New York: Oxford University Press, 1947. 4. Eichhorn R, Tracktir J . The relationship between anxiety, hypnotically induced emotions and gastric secretion. Gastroenterology 1955;29:422-31. 5. Kehoe M, Ironside W. Studies on the experimental evocation of depressive responses using hypnosis. 11. The influence of depressive responses upon the secretion of gastric acid. Psychosom Med 1963;25:403-19. 6 . Wittkower E. Zur affektiven Beeinflussbarkeit der Magensekretion. Klin Wochenschr 1931;10:1811-3. 7. Heller MH, Levine J . Sohler TP. Gastric acid and normally produced anxiety. Psychosom Med 1953;15:509-12. 8. Sonnenberg A, Donga M, Erckenbrecht JF, Weinbeck M. The effect of mental stress induced by noise on gastric acid secretion and mucosal blood Row. Scand J Gastroenterol 1984;19 Suppl 89:45-8. 9. Thompson DG, Richelson E, Malagelada JR. Perturbation of gastric emptying and duodenal motility through the central nervous system. Gastroenterology 1982;83:12OC-6. 10. Badgley LE, Spiro HM, Senay E. Effect of mental arithmetic on gastric secretion. Psychophysiology 1969;5:633-7. 1 1 . Holtmann G, Kriebel R, Singer MV. Mental stress and gastric acid secretion: Do personality traits influence the response? Dig Dis Sci 1990;35:998-1007. 12. Holtmann G , Singer MV, Krievel R, Stacker KH, Goebell H. Differential effects of acute mental stress on interdigestive Received 2 March 1992 Accepted 3 June 1992
923
secretion of gastric acid, pancreatic enzymes, and gastroduodenal motility. Dig Dis Sci 1989;34:1701-7. 13. Davies DT, Wilson ATM. Observations on life-history of chronic peptic ulcer. Lancet 1937;2:1353-60. 14. Mittleman B, Wolff HG. Emotions and gastroduodenal function: experimental studies on patients with gastritis, duodenitis and peptic ulcer. Psychosom Med 1942;4:5-62. 15. Peters MN, Richardson CT. Stressful life events, acid hypersecretion, and ulcer disease. Gastroenterology 1983;84:114-9. 16. Sapira JD, Cross MR. Pre-hospitalization life changes in gastric ulcer versus duodenal ulcer. Psychosom Med 1982;44:121. 17. Magni G, Salmi A , Paterlini A, Merlo A. Psychological distress in duodenal ulcer and acute gastroduodenitis. Dig Dis Sci 1982;27:1081-4. 18. Piper DW, Ariotti D, Grieg M, Brown R. Chronic duodenal ulcer and depression. Scand J Gastroenterol 1980;15:201-3. 19. Piper DW, McIntosh JH, Ariotti DE, Calogiuri JV, Brown RW, Shy CM. Life events and chronic duodenal ulcer: a case control Study. Gut 1981;22:1011-7. 20. Feldman M, Walker P, Green JL, Weingarden K. Life events stress and psychosocial factors in man with peptic ulcer disease: a multidimentional case-controlled study. Gastroenterology 1986;91:1370-9. 21. Walker P, Luther J , Samloff IM, Feldman M. Life events stress and psychosocial factors in men with peptic ulcer disease. Gastroenterology 1988;94:323-30. 22. Whiten JT, Bright-Asare P. Stress may contribute to delayed duodenal ulcer healing by increased acid secretion [abstract]. Gastroenterology 1984;86:1298. 23. Mason JB, Moshal MG, Naidoo V , Schlemmer L. The effect of stressful life situations on the healing of duodenal ulceration. S Afr Med J 1981;60:734-7. 24. Fahrenberg J, Hampel R, Selg H. Das Freiburger Personlichkeits Inventar. Gottingen: Hogrefe, 1984. 25. Winer BJ. Statistical principles in experimental design. New York: McGraw-Hill, 1971. 26. Mehta CR, Patel NR. A network algorithm for performing Fisher’s exact test in r x c contingency tables. J Am Stat Assoc 1983;78:427-34. 27. Holtmann G, Kriebel R, Singer MV. Effects of acute mental stress on gastric acid, pancreatic output and gastroduodenal motility. A short review. In: Singer MV, Goebell H, editors. Nerves and gastrointestinal tract. London: MTP Press, 1989: 523-30. 28. Cobbs S, Rose RM. Hypertension, peptic ulcer and diabetes in air traffic controllers. JAMA 1973;224:48%92. 29. Depue RA, Monroe SM. Conceptionalisation and measurement of human disorder in life stress research: the problem of chronic disturbance. Psycho1 Bull 1986;99:3&51. 30. Alexander F. Psychosomatic medicine: Its principles and applications. New York: Norton, 1950. 31. Weiner F, Thaler M, Reiser MF, Mirsky IA. Aetiology of duodenal ulcer. I . Relation of specific psychological characteristics to rate of gastric secretion (serum pepsinogen). Psychosom Med 1957;19:1-10. 32. McIntosh JH, Nasiry RW, Frysman M, Waller SL, Piper DW. The personality pattern of patients with chronic peptic ulcer. A case control study. Scand J Gastroenterol 1983;18:945-50. 33. Magni G, DiMario F, Conlon P, Nacarato R. Personality patterns and duodenal ulcer relapse under antisecretory treatment. Can J Psychiatry 1987;32:77-8. 34. Sonnenberg A , Muller-Lissner SA, Vogel E, et al. Predictors of duodenal ulcer healinR - and relapse. Gastroenterology -. I981;81:1061-7. 35. Sontae S. Graham DY. Belsito A. et al. Cimetidine, cigarette smokGg,’ and recurrence of duodenal ulcer. N Engl Med 1984;311:68%93. 36. Ellard K, Beaurepaire J, Jones M, Piper D, Tennant C. Acute and chronic stress in duodenal ulcer disease. Gastroenterology 1990;99:1628-32.
ISSN: 0036-5521 (Print) 1502-7708 (Online) Journal homepage: http://www.tandfonline.com/loi/igas20
Influence of Stress on the Healing and Relapse of Duodenal Ulcers: A Prospective, Multicenter Trial of 2109 Patients with Recurrent Duodenal Ulceration Treated with Ranitidine G. Holtmann, D. Armstrong, E. Pöppel, A. Bauerfeind, H. Goebell, R. Arnold, M. Classen, L. Witzel, M. Fischer, M. Heinisch & A.L. Blum To cite this article: G. Holtmann, D. Armstrong, E. Pöppel, A. Bauerfeind, H. Goebell, R. Arnold, M. Classen, L. Witzel, M. Fischer, M. Heinisch & A.L. Blum (1992) Influence of Stress on the Healing and Relapse of Duodenal Ulcers: A Prospective, Multicenter Trial of 2109 Patients with Recurrent Duodenal Ulceration Treated with Ranitidine, Scandinavian Journal of Gastroenterology, 27:11, 917-923, DOI: 10.3109/00365529209000163 To link to this article: http://dx.doi.org/10.3109/00365529209000163
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 21
View related articles
Citing articles: 3 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=igas20 Download by: [New York University]
Date: 10 April 2016, At: 18:56
Influence of Stress on the Healing and Relapse of Duodenal Ulcers A Prospective, Multicenter Trial of 2109 Patients with Recurrent Duodenal Ulceration Treated with Ranitidine G . HOLTMANN, D. ARMSTRONG, E. POPPEL, A. BAUERFEINDt, H. GOEBELL, R. ARNOLD, M. CLASSEN, L. WITZEL, M. FISCHER, M. HEINISCH, A. L. BLUM & MEMBERS O F THE RUDER STUDY GROUP Division of Gastroenterology, University of Essen, Essen; Institute of Medical Psychology, University of Munich, and Dept. of Gastroenterology, Technical University, Munich; Division of Internal Medicine, University of Marburg, Marburg; Division of Internal Medicine, DRK Hospital, Berlin; and Institute of Numerical Statistics, Cologne; Germany; and Division of Gastroenterology, CHUV, Lausanne, Switzerland
Downloaded by [New York University] at 18:56 10 April 2016
Holtmann G , Armstrong D , Poppel E , Bauerfeind A, Goebell H, Arnold R, Classen M, Witzel L, Fischcr M. Heinisch M, Blum AL, Members of the RUDER Study Group. Influence of stress on the healing and relapse of duodenal ulcers. A prospective, multicenter trial in 2109 patients with recurrent duodenal ulceration treated with ranitidine. Scand J Gastroenterol 1992;27:917-923. The influence of psychologic factors on the healing and relapse of duodenal ulcers under treatment with ranitidine was studied in a prospective, multicenter trial in 2109 patients with an endoscopically proven duodenal ulcer (DU) and a history of recurrent duodenal ulceration. All patients received ranitidine (300 mg daily), and, after healing, 1899 patients continued maintenance treatment (ranitidine, 150 mg daily) for 2 years. A physician's assessment of stress (stress or no stress) was made at every consultation. In the healing phase an overall classification of stress as absent, intermittent, or continuous was made, and in the maintenance phase patients were classified dichotomously as having stress (stress on at least half of the follow-up consultations) or no stress. In addition, at the start of the healing phase stress was measured by means of a standardized questionnaire. Continuous stress, as assessed by the physicians, was associated with a lower 14-day healing rate (35.7%) than intermittent or absent stress (42.4%; relative risk (RR) for delayed healing in patients with continuous stress, 1.19; 95% confidence interval (Ci), 1.06-1.33; P < 0.02). Differences in the 14-day healing rate for patients with low and moderate stress scores (43.1%) compared with those with high and very high stress scores (37.9%) just failed to reach statistical significance (RR for patients with stress, 1.14; 95% CI, 0.99g1.29; P = 0.051). During the 1st year of maintenance treatment 18.3% of patients with stress, but 10.9% of patients without stress, had a DU relapse (RR of stress for DU relapse during the first year, 1.73; 95C1, 1.44-2.09; P < 0.001). In contrast, during the 2nd year a significant difference in the relapse rates for patients with (9.7%) and without stress (11.6%) was not found (RR, 0.84; 95% C1, 0.61-1.15; P>0.4). Thus, on the basis of the physicians' assessment, continuous stress delays duodenal ulcer healing, and stress increases relapse rates during the 1st but not during the 2nd year of maintenance therapy with an HZreceptor antagonist.
Key words: Peptic ulcer; psychologic factors; psychologic stress; ranitidine; relapse Gerald Holtmann, M . D . , University of Essen, Dept. of Internal Medicine, Dioision of Gastroenterology, Hufelandstr. 55, 0-4.300 Essen 1 , Germany
The close relationship between the brain and gastrointestinal functions was described more than 150 years ago by W. Beaumont (l),and since then several uncontrolled (2-7) and controlled (8-12) studies have been published indicating that acute mental stress affects the secretion and motility of the upper gastrointestinal tract. Furthermore, there is evidence of an association between the occurrence of duodenal ulcers and psychologic stress. Some anecdotal reports describe a coincidence of stressful life events and peptic ulcer disease (1S1.5). In contrast, controlled studies have yielded contradictory results: Sapira & Cross (16) and Magni et al.
t Deceased.
(17) found more stressful life events and more psychologic distress in patients with duodenal ulcers than in controls, whereas Piper et al. (18, 19) found little or no relationship between stressful life events and peptic ulcer disease. More recently, Feldman et al. (20) and Walker et al. (21) have shown that patients with peptic ulcer do not differ from nonulcer controls with regard to the number of life events, but peptic ulcer patients perceived life events more negatively than did controls. However, little is known about the influence of life stress on the healing and relapse of duodenal ulcers in patients treated with H2-receptor antagonists. Although there is some evidence that stress may delay duodenal ulcer healing (22, 23), there have been no large con-
Downloaded by [New York University] at 18:56 10 April 2016
918
G. Holtmann et al.
trolled, prospective studies. In particular, there is little information about the effect of stress on duodenal ulcer relapse. This study examines the influence of psychologic factors on the healing and relapse of duodenal ulcers in patients treated with the widely prescribed H2-receptor antagonist ranitidine. A study population of at least 2000 patients was necessary to ensure that any possible effect of stress and the concomitant effect of other risk factors on healing and relapse would not be overlooked. This necessitated a large multicenter study with close and regular monitoring to ensure that patients were not lost to follow-up during the maintenance phase. The following hypotheses were tested: 1) psychologic stress delays the healing of duodenal ulcers treated with 300 mg ranitidine daily; 2) stress increases relapse rates over a 2-year follow-up period during maintenance treatment with 150 mg ranitidine daily. MATERIALS AND METHODS The criteria for inclusion in the study were an endoscopically proven duodenal ulcer (DU) with a diameter greater than 5mm, a history of chronic DU disease, and at least one duodenal ulcer manifestation within the previous year before the patient had a relapse that was treated in accordance with the study protocol, age over 18 years, and no ulcer-related complication requiring surgical treatment. In addition, patients fulfilling the following criteria were excluded from the study: renal insufficiency, gastric ulcer, pregnancy or lactation period, alcohol abuse (>60 g pure alcohol per day), or evidence of drug abuse.
Healing phase On the 1st day all patients with an endoscopically verified DU gave written informed consent to their inclusion in the study. Data concerning the history, socioeconomic status, and social situation were recorded, and treatment was started with 300 mg ranitidine taken at 1800 h. After 14 days of treatment all patients underwent a second upper GI endoscopy. If the ulcer had healed, the patients were enrolled in the maintenance phase of the study, and treatment was started with 150 mg ranitidine taken at 1800 h. If healing was not complete, patients continued to receive 300 mg ranitidine until complete ulcer healing was documented endoscopically; it was proposed that the repeat endoscopies should be performed 4 and 8 weeks after the start of the healing phase, but this was not obligatory, and, in general, they were performed at the discretion of the physician, when he/ she believed clinically that the ulcer had healed. Patients whose ulcers healed within 2 weeks were classified as fast healers, and all other patients as slow healers. In 19 patients there was a persistent ulcer (after 8 or 3 1 2 weeks of treatment). In accordance with the study protocol, these patients were classified as non-responders, and a new treatment
regimen (for example, gastric surgery), independent of the study protocol, was started. Data from 2109 patients under the care of 546 general practitioners or gastroenterologists performing upper GI endoscopy in outpatients were collected for the healing period. Study monitoring was performed by 48 physicians located at regional hospital centers: 5 study chairmen, 12 professional monitors in a data evaluation center, and 4 biostaticians who coordinated action in case of protocol violation. For statistical analysis data from 1923 patients (91.2%) were used; data from the remaining 186 patients could not be used for one or more of the following reasons: patients dropped out before healing was documented ( n = 66), no history of previous peptic ulcer disease (n = 61), ulcer diameter less than 5 mm at the time of diagnosis ( n = 39), treatment with ulcer-healing drugs other than ranitidine ( n = 23), alcohol abuse ( n = 6), and other reasons ( n = lo). The patient characteristics are summarized in Table I.
Maintenance phase During the 2-year follow-up period the patient attended every 3 months for follow-up appointments. Upper GI endoscopy was performed immediately in the case of duodenal ulcer symptoms and, in all cases, at the end of the 1st
Table I. Sociodemographic characteristics of the study population; data on 1923 patients were available (mean 2 SEM, range or number of patients (%) with the particular characteristic) Age (years) Gender* Men Women Weight (kg) Height (cm) Marital status Married Divorced Single Partnership Widowed Age at begin. of duodenal ulcer disease (years) Years since the first manifestation Patients with >3 recurrences within the past 2 years Complication noted in the records or recalled by the patient Perforation Bleeding Pyloric stenosis Other Smokers Ex-smokers Non-smokers
48 19-90 1318 601 73 34-120 172 144200 1444 131 174
89 85 36 12-86 12
f 13
(68.5%) (31.3%) f12 28
(75.1% (6.8% (9.0% (4.6% (4.4% 213 2 10
63% 38 296 78 23 1028 431 464
(2.0% (15.4% (4.1% (1.2% (53.5% (22.4% (24.1%
* Data available for 1919 patients; otherwise data available for all 1923 patients.
Stress and Healing and Relapse of DU
Downloaded by [New York University] at 18:56 10 April 2016
and 2nd year. In the case of a relapse the dose of ranitidine was increased to that used during the healing period (300 mg daily) until healing was confirmed, at which time the maintenance dosage was resumed. For the 1st year of follow-up, data from 1899 and for the 2nd year data from 1671 patients were available (87% of patients with a documented healing completed the 2-year follow-up). Patients with 2-year followup were not different from those patients who did not complete the study with regard to sociodemographic data, the presence of risk factors for ulcer relapse, and the course of the disease (for example, duration of the disease or the number of relapses before the study). Assessment of psychologic stress When the diagnosis of duodenal ulcer had been confirmed (day l), psychologic stress was measured by means of a standardized questionnaire consisting of 20 items (Table 11). The patient scored each item from 0 to 3 depending on the extent to which the statement described his (or her) feelings of stress (grade 0 = none; grade 1 = slight; grade 2 = moderate; grade 3 = high). Thus, the minimum possible score was 0, and the maximum was 60. The questionnaire score was then used to classify the patients as having no or low (score S l S ) , moderate (score 16-25), high (score 2 6 35), or very high (score >35) stress. This questionnaire has
Table 11. Items of the stress questionnaire Item no. 1. 2.
3. 4.
5. 6. 7. 8. 9. 10.
11. 12. 13.
14.
15. 16. 17. 18. 19. 20.
I experience loneliness In the past I have been successful in realizing my aims and wishes* I try to foresee unpleasant events, to avoid them My everyday life is monotonous I always have good experiences with my fellow human beings* My job is considered to be strenuous My family life is harmonious* For some time past there have been strokes of fate Sometimes I have a premonition that gets me in a state I calm down quickly after excitement* I always try to work out problems in new ways* I don’t believe that my environment and problems will result in sickness* In our society the individual is usually not allowed to show his feelings I am looking forward to events in the near future; I am really pleased* Whenever I overcome a problem, the next problem follows I expect to recover from my disease* Unknown situations or unexpected events create disturbance and tensions At this time I am in serious economic and financial difficulties My apartment fulfills all my needs’ There are opponents I have to protect myself from to prevent harm
* Low scores on these items indicate presence of stress; stress scores were calculated accordingly.
919
been specifically developed and validated for this study. Cross-validation with the Freiburg Personality Inventory (FPI) (24) indicated that high scores on this stress questionnaire were associated with low life satisfaction (scale 1 of the FPI, r = -0.60, P < 0.01) and intense life strain (scale 7 of the FPI, r = 0.48, P < 0.01). At each visit the practitioner had to answer the question as to whether the patient was stressed at the time of the follow-up visit. There were no criteria given regarding what should be considered ‘stressed’. Therefore, the individual classification of stressed or not stressed was based on the physicians’ concepts and perception of stress rather than somehow standardized criteria. The practitioners’ estimation of stress was repeated at every scheduled follow-up appointment during the healing and follow-up periods. Furthermore, the treating physician did not have a key for evaluation of the stress questionnaire. They gave their assessment of stress without knowledge of the stress value described above. In the healing phase, therefore, patients were classified as having no stress, intermittent stress (stress at some but not at all appointments), or continuous stress (stress at all appointments). During the maintenance phase patients were classified as having stress if the physician considered them to be stressed on at least half of their follow-up appointments; all other patients were classified as having no stress. Statistical analysis In the healing phase the effect of stress was evaluated separately for the physicians’ assessment of stress and for the results of the self-administered questionnaire. For the physicians’ assessment, the numbers of fast healers (patients who had healed completely within 2 weeks) in each of the three groups (continuous, intermittent, and no stress) were compared by means of Fisher’s exact test (two-tailed) (25, 26). For the self-administered questionnaire the numbers of fast healers in the four groups (low, moderate, high, and very high stress) were compared, using again Fisher’s exact test (two-tailed). In the maintenance phase, the numbers of patients who had relapsed in the two stress groups (stress or no stress) were compared using Fisher’s exact test (twotailed). A significant level of 5% was assumed when testing the hypotheses of the effects of stress on healing and relapse. Statistical analysis was conducted using the Statistical Package for Social Sciences (SPSS) and Biomedical Computer Programs (BMDP). RESULTS Healing phase Overall healing rates. After 2 weeks of treatment healing was found in 39.5% of the patients; these patients were classified as fast healers. Cumulative healing rates after 4 and 8 weeks were 70.9% and 93.2%, respectively. No significant differences in the healing rates were found for male and female patients.
920
G . Holtmann et al.
Table 111. Comparison of the stress score and treating physicians’ estimation of stiess during the healing phase (includes i609.patients with data on stress and physicians’ assessment of stress)
Downloaded by [New York University] at 18:56 10 April 2016
Practitioners’ estimation Stress score
Without stress
With stress
Score < 15 (low), n = 298 15 < Score < 25, (moderate), n = 796 25 < Score s 35 (high), n = 392 Score < 35 (very high), n = 123
194 (65.1%)
104 (34.9%)
338 (42.4%)
458 (57.6%)
90 (23.0%)
302 (77.0%)
7 (5.7%)
116 (94.3%)
Assessment of psychologic stress: practitioners’ assessment. On the basis of the physicians’ assessment, 740 (38%) of the 1923 patients were classified as having no stress. The remaining 1183 patients were classified as being stressed: in 367 (19% of total) the stress was intermittent, and in 816 (42%) it was continuous. Stress questionnaire (patient self-assessment). Stress scores were available for 1609 patients; these showed that , patients 298 patients had low stress (test scores ~ 1 5 )796 had moderate stress (score 1&25), 392 patients had high stress (score 26-34), and 123 patients had very high stress (Table 111). Comparison of the two assessment procedures showed that 94.3% of patients with very high stress (>35) were also classified by their practitioner as being stressed (intermittent or continuous). However, only 65.1 % of
50
r
I
without
patients having low stress on the questionnaire (G1.5) were as having no stress. gls0 classified b y their Overall, patients with no stress according to their practitioners had a mean questionnaire score of 19.6 6 (LSD); patients with intermittent stress had a mean score of 22 2 7, and those with continuous stress had a mean score of 26 8. Psychologic stress and healing rates: practitioners’ assessment. The 14-day healing rates in patients with continuous stress (35.7%) was significantly lower than in patients with intermittent stress (42.8%; P < 0.05) or without stress (42.2%; P < 0.02). Therefore, the relative risk (RR) of patients with continuous stress as compared with patients without and with intermittent stress (14-day healing rate, 42.4%) for delayed healing is 1.12 (95% CI, 1.04-1.20; P < 0.02). There was no significant difference in the healing rates between those with intermittent stress and those without stress (Fig. 1). Furthermore, there was no statistically significant difference in the presence of risk factors for slow ulcer healing (for example, smoking, non-steroidal antiinflammatory drug intake) between subjects with and without stress. Ulcer healing occurred within 14 days in 37.7% of smokers and ex-smokers, in contrast to 45.3% of nonsmokers ( P < 0.001). Stressed smokers had a 14-day healing rate of 34.9% as compared with 38.4% in smokers without stress ( P < 0.05). There was no significant difference in the 14-day healing in non-smokers with and without stress (41.8 versus 45.6; P 2 0.4). N o association between stress and smoking habits or changes in the smoking habits was found. Stress questionnaire. In patients with low or moderate stress scores, the 14-day healing rate was 43.1%, compared with a 14-day healing rate of 37.9% in patients with high
*
**
intermittent
continuous
Stress Fig. 1 . Percentage of each stress group (practitioners’ assessment) who were healed within 14 days ( * P < 0.05, * * P < 0.01).
Downloaded by [New York University] at 18:56 10 April 2016
Srress and Healing and Relapse of DU
921
"
low
moderate
high
very high
Intensity of stress (stress score) Fig. 2. 14-day healing rates and stress (results of the standardized questionnaire, P = 0.051 for the comparison of subjects with low stress and moderate versus high and very high stress). and very high stress scores (RR for slow healing in patients with high and very high stress scores as compared with patients with low and moderate stress scores, 1.14; 95% CI, 0.998-1.29; P = 0.051). In Fig. 2, 14-day healing rates for patients with low, moderate, high, and very high stress scores are shown.
Maintenance phase (2-year follow-up study) Relapse rates During the 1st year of the follow-up, duodenal ulcer relapses were documented in 247 patients (13.0%; 95% CI, 11S 1 4 . 4 ) . Forty-nine patients with relapse and 179 patients without relapse during the 1st year did not participate in the 2nd year of the follow-up. Thus, 198 patients with a relapse during the first year and 1473 patients without a relapse were studied during the 2nd year of maintenance treatment (= 1671 patients). During the 2nd year 11.7% (95% CI, 10.313.3) of patients without relapse during the 1st year of maintenance treatment had a relapse. In addition, relapses occurred in 78 of 198 patients with a relapse during the 1st year. Differences in the presence of risk factors for DU relapse and sociodemographic data for patients with and without participation in the long-term treatment phase were not found. Stress and relapse rates During the 1st year of the follow-up the relapse rate in patients with stress (stress on at least half of the follow-up consultations) was 18.9%, compared with 10.9% in patients without stress (RR of stress for D U relapse during the first year, 1.73; 95% CI, 1.44-2.09; P < 0.001); during the 2nd year 44 of 454 (9.7%) patients with stress relapsed, compared with 141 of 1217 (11.6%) patients without stress (RR of
stress for DU relapse during the 2nd year of maintenance treatment, 0.84; 95% CI, 0.61-1.15; P > 0.4). Cumulated relapse rates for patients who completed 2 years of maintenance treatment were 120 of 454 (26.4%) with stress and 263 of 1217 (21.6%) patients without stress (RR of stress for DU relapse during 2-year maintenance treatment, 1.22; 95% CI, 1.01-1.47; P < 0.05). The significantly higher relapse rate of patients with stress during the 1st year of maintenance treatment was independent of the presence or absence of stress during the healing period (Fig. 3). Differences in the presence of risk factors for ulcer relapse (for example, smoking behavior) were not found for subjects with and without stress. Relapses occurred in 20.4% of non-stressed but in 30.5% of stressed smokers ( P < 0.001). DISCUSSION There is a widespread belief that psychologic factors are involved in the pathogenesis of peptic ulceration. Thus, psychologic stress is accused of delaying the healing and of promoting relapses of duodenal ulcers. This assumption is consistent with controlled and uncontrolled studies showing that stress can alter both gastric acid secretion and gastroduodenal motility (1-8, 12, 27). However, there are very few pertinent studies, and previous studies have yielded conflicting results (13, 1&21, 28). The present prospective, multicenter study was conducted to examine, in a large and inhomogeneous population, the influence of stress on the healing of DU. Identical treatment for all patients was chosen and consisted of curative treatment with ranitidine (300 mg daily). Subsequently, maintenance treatment with 150 mg ranitidine daily was given for 2 years. The results of
Downloaded by [New York University] at 18:56 10 April 2016
922
G . Holtmann et al.
0 Healing phase: without stress phase: MainteMnce
without stress
with stress
with stress
without stress
without Stress
with stress
with Stress
Fig. 3. InRuence of stress (practitioners’assessment) on the 1-year relapse rate (***P< 0.001 for the comparison of subjects with and without stress during the 1st year of the maintenance phase) this study confirm the two hypotheses tested. First, patients considered by their physicians to be under continuous stress in the healing phase had a slower DU healing than patients who were not stressed or patients with intermittent stress. Second, during the 2-year maintenance treatment patients with stress had a significantly higher relapse rate than patients without stress. Even though stress, as assessed by the physician, was associated with delayed ulcer healing and DU relapses, effects of stress as measured by the questionnaire just failed statistical significance (RR, 1.14; 95% CI, 0.998-1.29; P = 0.051). This finding suggests that physicians, when they assess the patients’ stress, are taking more into consideration than the internal stress level that the standardized questionnaire was designed to measure. Even though patients with high stress scores in general were classified by the physician as stressed, the physician’s judgement of stress is actually a composite assessment of the patient’s current lifestyle and life situation, his recent experience of life events, and his response to stress; it may even be that the physician is subconsciously drawing conclusions on the basis of his clinical experience of patients who are likely to respond poorly to treatment and heal slowly or relapse rapidly. Thus, the outcomes of the physicians’ assessment and the questionnaire measurement should not be seen as mutually contradictory: rather, they suggest that the questionnaire measures a specific facet of the patient’s stress (in particular, as indicated by the results of the questionnaire cross-validation, both strain and life satisfaction in general), whereas the physician is able to provide an even more general assessment of the extent of the patient’s stress status. Nevertheless,
some between-physician variation in the assessment of stress will result from the present approach to assessment. The stress assessment of the present study, however, reflects what the participating physicians considered to be stress. The relapse rate was also related to the presence or absence of stress in the healing phase as measured by the questionnaire. The overall relapse rate during the 1st year of maintenance treatment was higher in patients with low stress scores during healing, particularly in those who were stressed (physicians’ assessment) during maintenance treatment (23.8%). A possible explanation of this finding is that low stress scores might detect not only those patients who are genuinely unstressed and, thus, have a low risk of relapse, but also those stressed patients whose repressive coping strategies lead them to deny their stress. However, this repressive coping strategy seems to be associated with an increased risk for a relapse. There are several possible reasons why this study has detected an effect that has not been seen in previous studies. The most important reason is that all previous studies have examined relatively small numbers of patients. Studies with relatively small numbers of subjects may fail to show an effect if this effect on healing or relapse is small. It is also possible that stress and psychologic factors are important in some but not all subjects and therefore might be overlooked in small samples. Another reason for not seeing an effect may be the fact that psychologic stress is not a dichotomous variable: some patients are more stressed than others, and intensity of stress may change with time. The ability to cope with psychologic strains differs from individual to individual and might influence the effect of stress. In addition, psy-
Stress and Healing and Relapse of DU
Downloaded by [New York University] at 18:56 10 April 2016
chologic stress is not just an accumulation of stressful events; it is a multifactorial influence in which the stress is d u e to a variable number of events whose effect might be modified by other factors such as experience, social circumstances, and personality (29-33). Thus, it is not only t h e life events themselves but also the various modifying factors which determine the individual response t o potentially stressful life events. Recently, Ellard et al. (36) emphasized t h e importance of t h e duration and quality of the stress. Chronic stressors (those of 6 or more months’ duration) were significantly and independently associated with the onset and
relapse of DU In conclusion, psychologic factors affect the healing and the relapse rate of DU. T h e effect of stress on healing and relapse rate is independent of other risk factors such as smoking. Furthermore, the results suggest that a simple assessment of stress by the physician may identify those patients who will heal more slowly and have an increased risk for a DU relapse. ACKNOWLEDGEMENTS This comprehensive study has been financed by Cascan & Co. KG Wiesbaden, Germany, and SNF grant 32-26369 89; the essential administrative management was handled by Cascan and IFNS (Institut fur Numerische Statistik) under t h e supervision of the chairman. REFERENCES 1 . Beaumont W. Experiments and observations on the gastric juice and the physiology of digestion. Plattsburgh: F. P. Allen, 1833. 2. Bennett TI, Venables J. The effects of the emotions on gastric secretion and motility in the human being. Br Med 3 1920;2.6623. 3. Wolf S, Wolff HG. An experimental study of a man and his stomach. 2nd ed. New York: Oxford University Press, 1947. 4. Eichhorn R, Tracktir J . The relationship between anxiety, hypnotically induced emotions and gastric secretion. Gastroenterology 1955;29:422-31. 5. Kehoe M, Ironside W. Studies on the experimental evocation of depressive responses using hypnosis. 11. The influence of depressive responses upon the secretion of gastric acid. Psychosom Med 1963;25:403-19. 6 . Wittkower E. Zur affektiven Beeinflussbarkeit der Magensekretion. Klin Wochenschr 1931;10:1811-3. 7. Heller MH, Levine J . Sohler TP. Gastric acid and normally produced anxiety. Psychosom Med 1953;15:509-12. 8. Sonnenberg A, Donga M, Erckenbrecht JF, Weinbeck M. The effect of mental stress induced by noise on gastric acid secretion and mucosal blood Row. Scand J Gastroenterol 1984;19 Suppl 89:45-8. 9. Thompson DG, Richelson E, Malagelada JR. Perturbation of gastric emptying and duodenal motility through the central nervous system. Gastroenterology 1982;83:12OC-6. 10. Badgley LE, Spiro HM, Senay E. Effect of mental arithmetic on gastric secretion. Psychophysiology 1969;5:633-7. 1 1 . Holtmann G, Kriebel R, Singer MV. Mental stress and gastric acid secretion: Do personality traits influence the response? Dig Dis Sci 1990;35:998-1007. 12. Holtmann G , Singer MV, Krievel R, Stacker KH, Goebell H. Differential effects of acute mental stress on interdigestive Received 2 March 1992 Accepted 3 June 1992
923
secretion of gastric acid, pancreatic enzymes, and gastroduodenal motility. Dig Dis Sci 1989;34:1701-7. 13. Davies DT, Wilson ATM. Observations on life-history of chronic peptic ulcer. Lancet 1937;2:1353-60. 14. Mittleman B, Wolff HG. Emotions and gastroduodenal function: experimental studies on patients with gastritis, duodenitis and peptic ulcer. Psychosom Med 1942;4:5-62. 15. Peters MN, Richardson CT. Stressful life events, acid hypersecretion, and ulcer disease. Gastroenterology 1983;84:114-9. 16. Sapira JD, Cross MR. Pre-hospitalization life changes in gastric ulcer versus duodenal ulcer. Psychosom Med 1982;44:121. 17. Magni G, Salmi A , Paterlini A, Merlo A. Psychological distress in duodenal ulcer and acute gastroduodenitis. Dig Dis Sci 1982;27:1081-4. 18. Piper DW, Ariotti D, Grieg M, Brown R. Chronic duodenal ulcer and depression. Scand J Gastroenterol 1980;15:201-3. 19. Piper DW, McIntosh JH, Ariotti DE, Calogiuri JV, Brown RW, Shy CM. Life events and chronic duodenal ulcer: a case control Study. Gut 1981;22:1011-7. 20. Feldman M, Walker P, Green JL, Weingarden K. Life events stress and psychosocial factors in man with peptic ulcer disease: a multidimentional case-controlled study. Gastroenterology 1986;91:1370-9. 21. Walker P, Luther J , Samloff IM, Feldman M. Life events stress and psychosocial factors in men with peptic ulcer disease. Gastroenterology 1988;94:323-30. 22. Whiten JT, Bright-Asare P. Stress may contribute to delayed duodenal ulcer healing by increased acid secretion [abstract]. Gastroenterology 1984;86:1298. 23. Mason JB, Moshal MG, Naidoo V , Schlemmer L. The effect of stressful life situations on the healing of duodenal ulceration. S Afr Med J 1981;60:734-7. 24. Fahrenberg J, Hampel R, Selg H. Das Freiburger Personlichkeits Inventar. Gottingen: Hogrefe, 1984. 25. Winer BJ. Statistical principles in experimental design. New York: McGraw-Hill, 1971. 26. Mehta CR, Patel NR. A network algorithm for performing Fisher’s exact test in r x c contingency tables. J Am Stat Assoc 1983;78:427-34. 27. Holtmann G, Kriebel R, Singer MV. Effects of acute mental stress on gastric acid, pancreatic output and gastroduodenal motility. A short review. In: Singer MV, Goebell H, editors. Nerves and gastrointestinal tract. London: MTP Press, 1989: 523-30. 28. Cobbs S, Rose RM. Hypertension, peptic ulcer and diabetes in air traffic controllers. JAMA 1973;224:48%92. 29. Depue RA, Monroe SM. Conceptionalisation and measurement of human disorder in life stress research: the problem of chronic disturbance. Psycho1 Bull 1986;99:3&51. 30. Alexander F. Psychosomatic medicine: Its principles and applications. New York: Norton, 1950. 31. Weiner F, Thaler M, Reiser MF, Mirsky IA. Aetiology of duodenal ulcer. I . Relation of specific psychological characteristics to rate of gastric secretion (serum pepsinogen). Psychosom Med 1957;19:1-10. 32. McIntosh JH, Nasiry RW, Frysman M, Waller SL, Piper DW. The personality pattern of patients with chronic peptic ulcer. A case control study. Scand J Gastroenterol 1983;18:945-50. 33. Magni G, DiMario F, Conlon P, Nacarato R. Personality patterns and duodenal ulcer relapse under antisecretory treatment. Can J Psychiatry 1987;32:77-8. 34. Sonnenberg A , Muller-Lissner SA, Vogel E, et al. Predictors of duodenal ulcer healinR - and relapse. Gastroenterology -. I981;81:1061-7. 35. Sontae S. Graham DY. Belsito A. et al. Cimetidine, cigarette smokGg,’ and recurrence of duodenal ulcer. N Engl Med 1984;311:68%93. 36. Ellard K, Beaurepaire J, Jones M, Piper D, Tennant C. Acute and chronic stress in duodenal ulcer disease. Gastroenterology 1990;99:1628-32.
