CONTEMPO 1999 UPDATES LINKING EVIDENCE AND EXPERIENCE Stress and Peptic Ulcer Disease Susan Levenstein, MD Sigurd Ackerman, MD Janice K. Kiecolt-Glaser, PhD Andre ´ Dubois, MD A LEXANDER THE GREAT DIED AT THE age of 32 years, with acute ab- dominal pain that began after sev- eral days of binge drinking. Might it have been from a perforated peptic ulcer? Founding a great empire may be an ex- traordinary example of life stress, but stress is currently out of fashion as a cause of ulcer. 1 For many years the dominant etiologic model was exquisitely psycho- somatic: a vulnerable person—on grounds of personality and pepsinogen—encoun- ters a major life stress, and a duodenal ul- cer is born. 2 But after Helicobacter pylori proved to be a key and curable element in the ulcer diathesis, many concluded that the “real” cause had been found and had nothing to do with psychology. Re- search into stress effects on ulcer fell off precipitously, and the earlier literature was dismissed as misguided and naive, given the new, respectable status of peptic ul- cer as an infectious disease. In a recent telephone survey of ordinary Ameri- cans’ views of what causes an ulcer, the authors seemed to consider the wide- spread belief in a psychological compo- nent tantamount to a superstition deserving eradication. 3 But attempts to explain ulcer using H pylori and nonsteroidal anti- inflammatory drugs (NSAIDs) as sole etiologic factors are destined to fail. More than 80% of H pylori–infected people (and the vast majority of NSAID users) never develop an ulcer, while at least 10% of patients with non–NSAID-related pep- tic ulcers have no H pylori infection. 4 The field is therefore open for other factors working in conjunction with H pylori or causing ulcers through alternative pathways. The evidence that psychological stress is 1 of those factors is not invalidated by the discovery of H pylori. The German blitz in London, 5 the Kobe earthquake, 6 economic crisis in Sophia, 7 and sover- eignty negotiations in Hong Kong 8 have all been followed by an increase in pep- tic ulcers in both the stomach and the duodenum, as has being a prisoner of war. 9 In defined epidemiologic cohorts, subjects with psychological distress, 10 self-described “stress or strain,” 11 or con- crete life stressors 12 at baseline have in- creased incidence of ulcer over 9 to 15 years, an association that holds up to ad- justment for a variety of nonpsychoso- cial risk factors 10,11 and is similar whether the outcome is assessed by medical rec- ords or by self-report. 11 A large body of research has exam- ined the effect of stress on the upper gastrointestinal tract in animal models. Though most such models produce su- perficial gastric lesions (better models, using transgenic animals, may emerge in the future), enough is shared with chronic human peptic lesions to make consider- ation of evidence from these models worthwhile. In rats, susceptibility to gas- tric lesions is increased by such social stressors as premature separation of the rat pup from its mother. 13 Also, in rats taught to avoid ulcer-producing electric shocks by operant responses, the occur- rence of lesions can be drastically re- duced if the animal receives more infor- mation about the efficacy of its response, demonstrating that perceptions and their integration into retrievable information af- fect the course of ulcer formation. 14 Among potential mediators, several known behavioral risk factors for ul- cers—smoking, alcohol abuse, and lack of sleep 10 —have clear associations with real-life stress and are known to impair wound healing through their effects on immune function 15 ; sleep loss can also elevate cortisol levels. Individuals un- der stress may also be likely to increase NSAID use. 10 On the physiological side, stress is known to modify gastric blood flow, which plays an important role in the gastric mucosal barrier, and to af- fect possible mediators such as thyro- tropin-releasing hormone, cytokines, and corticotropin-releasing hormone. Stimulation of gastric acid secretion has historically been considered an- other mechanism by which stress in- creases susceptibility to duodenal ulcer- ation, and researchers have reported increases in acid secretion in associa- tion with psychosocial stressors, 16,17 es- pecially among patients with duodenal ulcer. 18 Some studies of responses to acute mental stressors in humans 19 and in nonhuman primates 20 have, how- ever, cast doubt on this mechanism. Per- haps these conflicting reports may be rec- onciled by postulating anomalous reaction patterns among ulcer-prone in- dividuals. 18 Stress seems to have vari- able effects on gastric motility: delayed gastric emptying could increase the risk of gastric ulcer, while accelerated emp- tying could increase the net acid load de- livered to the duodenum at any given level of gastric secretion, enhancing the risk of duodenal ulcer 21 ; skipped meals and poor sleep might increase duode- nal acid load still further. In most ulcer cases where stress is in- volved, H pylori is likely to be present as well. The impact of the 2 factors may be additive. Individuals infected with large Edited by Thomas C. Jefferson, MD, JAMA Fishbein Fellow. Author Affiliations: Department of Gastroenter- ology, Nuovo Regina Margherita Hospital, Rome, Italy (Dr Levenstein); Department of Psychiatry, St Luke’s-Roosevelt Hospital Center, New York, NY (Dr Ackerman); Department of Psychiatry, Ohio State Uni- versity College of Medicine, Columbus (Dr Kiecolt- Glaser); and Digestive Diseases Division, Uniformed Services University of the Health Sciences, Bethesda, Md (Dr Dubois). Corresponding Author and Reprints: Susan Leven- stein, MD, Via del Tempio 1A, 00186 Rome, Italy (e-mail: slevenstein@compuserve.com). 10 JAMA, January 6, 1999—Vol 281, No. 1 ©1999 American Medical Association. All rights reserved.