Tobacco Documents Online

Reference How No. with Rates: age- diagnosed ulcers adjusted Current Sex cigarette ~ smokers Ratio smokers ~ Edwards, F. (1959) (22) Doctor 143 no M 10.1 6.0 1.7, yes Higgins, M.W. (1966) (28) Doctor 140 yes , M 7.1 5.2 1.4 - 47 yes F 2.8 1.4 2.0 - Friedman, G.D. (1974) (23) History 1520" yes M 12.2 5.8 2.1, yes 1092b yes F 6.3 3.9 1.6 yes Jedrychowski, W. (1974) (31) Doctor 106" no M 6.4 1.9 3.4 yes E6b no F .8 1.3 .6 yes Paffenbarger, R.S. (1974) (41) History 389 yes M 2.2" 1.5d 1.5" yes Goldbourt, U. (1975) (25) X-ray 895 no M 10.2 6.2 1.6 no s TABLE 1.-Peptic ulcer prevalence in smokers and nonsmokers (no. per 100) `Also, ratio > 1 within age and social class. "Not given-eatimated, using total population and reported ratee. 'Also, ratio > I within occupational groups. dSmoking categories in college, ulcers developed in 16 to 50 year follow-up. 9ao9s9c0 0 m ®

nonsmokers, with a trend of increased risk with increased number of prevalence of duodenal ulcer in this ou (3?1 first-generation Israelis of European descent increased, so did the al. made the interesting observation that as the smoking habits of nonsmokers (25). These differences were highly significant. Medalie, et of PUD (primarily duodenal) of 10.2 percent compared to 6.2 percent of,* InIsrael, the lifetime prevalence of PUD is 89/1000 men (37), similar to that in the United States. Smokers or ex-smokers had a prevalence cigarettes smoked. increases in the amount smoked studies provided evidence of increased frequency of peptic ulcer with are also consistently greater than 1.0. In addition, the majority of the ; ratios for women are similar with the exception of the Polish study, in ' which very few women smoked'. The ratios for ex-smokers (not shown) for men are very similar, the median being 1.7 and the mean 1.9. The ti smokers compared to nonsmokers. Despite the fact that these studies were done at different times and in four different countries, the ratios ' of the studies there was an increased prevalence of PUD in cigarette :a; 28; 31, 41) with a summary of their charaeteristics and results. In each. ulcers than nonsmokers?" is asked, results are st'rikingly, consistent. Table 1 lists~t'he six studies which investigated this problem (22, 23, 25, Thus, when the question, "Do cigarette smokers have more peptic Since cigarette smoking appears to be related to the prevalence of PUD, several other issues mustt be addressed. First, if a smoker does develop PUD, will cigarette smoking influence its healing and should the patient therefore be advised to stop smoking? Second, what, if any, role will smoking play in the chances of the patient dying from PUD? Course of Peptic Ulcer Disease who did, 86 percent (19/22) healed' as opposed to 61 percent of those who only decreased their smoking. The healing rate of the 24 nonsmokers was 58 percent, similar to that of smokers. Study design to continue smoking. Treatment for the ulcer disease was otherwise equivalent (although, not the same for all patients). The investigators then compared the two groups in regard to percent showing marked healing of the ulcer at 4 weeks (marked healing is defined as 2/3 or greater reductiom in ulcer size). Of those who. were advised to discontinue smoking, 75 percent showed marked healing, compared to only 58'percent of those who continued to smoke. In fact, 45 percent of the patients advised to stop smoking did not do so completely. Of those study, half were advised to stop smoking, the other half were allowed In a classic study, Doll; et al. (18) examined the effect of continued smoking on~ the healing rate of gastric uleers. Of the 80 smokers in the Effect on Healing and Recurrence 9-8

, ~~~Q town)'~wr y •d hould fany;~ PUD? : .~ inu n the -Wed "Wim ators rked '3 or' i_ to .3to it of hose hose 24 3ign 0 and technical aspects were offered as explanation for this latter observation. .. ~ . . Herrmann, and Piper (27) retrospectively looked at 101 patients with benign gastric ulcer, all radiologically diagnosed. At 3 weeks, 67 percent of nonsmokers had healed compared to 43 percent of smokers who continued smoking. Differences were less marked at 6 weeks (85 percent vs. 75 percent). Although the numbers were smaller, those smokers who stopped did not do as well' as either of the other two grouPs. The mean ulcer size in smokers was larger than in nonsmokers (120 mmz vs. 40 mm?): Those who smoked cigarettes and ingested~ salicylates had the largest ulcers, but mean ulcer size was significantly larger in smokers than in nonsmokers, even when those ingesting salicylates were excluded. Piper, et al. (.44); while investigating gastric ulcer, noted increased rates of recurrence for those discharged unhealed, for those with~ larger ulcers, and for smokers. In a 4-year follbw-up study of these patients, Piper, et al. (46) recently confirmed their previous report. They found that, of the 33 patients who were discharged with unhealed ulcers, 47 percent (8/17) of nonsmokers had recurrence, whereas 75 percent (12/16) of smokers ha& recurrence. Only one study has been made on the effect of smoking on the healing of duodenat ulcers. Peterson, et al. (42) recently showed for the first time the efficacy of antacids over placebo in the healing of duodenal ulcer (Table 2). In this study, 78 percent of the antacid- treated group healed at 4 weeks as compared to 45 percent of thee placebo group. When these groups were broken down~into smokers andd nonsmokers, 69 percent of the ulcers of nonsmokers who took placebo healed versus 32 percent of ulcers of smokers who took placebo (p < .05). In the antacid group{ 87 percent of nonsmokers healed versus 75 percent of smokers (p > .05). Nonsmokers showed good healing even on placebo; antacids appeared to make the most difference in treating the duodenal ulcers of smokers. Although there have been many recent clinical trials concerning the treatment of both gastric and duodenal' ulcers using the new histamine H2 receptor antagonist, cimetidine, none of these has carefully addressed the questiom of the influence of smoking on healing rates (67). Certainly, with all the international trials being undertaken to evaluate the plethora of new ulcer treatments, such as cimetidine, prostaglandins, bismuth, etc., the smoking habits of t,he patients should be examined. Such studies would provide information on the effect of smoking on the healing of untreated ulcers and~ on whether any of thee treatments can overcome the presumed adverse effect of smoking on healing. In summary, cigarette smoking, in males probably retards the healing rates of both gastric and duod'enal ulcers. 9-9 1 i4? a It R 0

.; 3;r - M:!ti TABLE 2.-Percentage of patients whose duodenal ulcers :wece .. healed by endoscopic examination at 4 weeks, ~`" .'~: classified according to treatment with placebo or' antacid and according to whether patients were `; smokers or nonsmokers of cigarettes. Numbers in parentheses are the number healed over the total number observed in each category. •'-..4~:a:r~ Percent healhd atA weeks 1 i* Y- t~~ : -....-. r.`. ~ ~ Smokera Nonsmokers Total '•: Placebo 3296 (8/25) 69% (9/13) 45% (17138) 3 Antacid 75% (21/28) 88% (7/8) 78% (28/36) ~ F.~$- . Total 55% (29/53) 76% (16/21) . TABLE 3.-Ulcer mortality of male cigarette smokers and .~ nonsmokers + ~~?C SOURCE: Peterson, W. L (42). ;S, R f No. of Rates: age- Ulcer Mortality • •{Doee" ' e erence deaths adjusted type ratio " response Hammond, E.C. (1958) 62 yes DU 2.2. (26) 46 yes GU >1.0- Ja Darn, H.F. (1959)' (W) 51 yes PUi 2.8 Weir, J:N. (1970) (64) 24 yes DU 5~ > >m i ~ 20 yes GU 74+ >1 Ye7 egimt . i :'. c~1 Doll, R. (1976) (I9) 79 yes PUi _ 2.5 yea . ,3 am •Smokeraincludeeegularcigaretteemuker,,many.ofwhomalsoamokedcigarsand PiPee- " ~~bRatio is 46i0. •Smoken include ex•emokers; nonsmokers include pipe and/or cigar. dRatiolor smokere of 1'pack/day to those smoking leas. DU - deodenal ulrcr;,GU - gaatric ulcer; PU- peptic.uleer. Effect on Mortality . . " . .yrl.r'~ . Mortality, as welli as morbidity„in PUD is related to cigarette smoking. = The four studies discussed below are summarized in Table 3. In one of the earliest and largest studies on smoking and death rates, Hammond ` and Horn (26)! pointed out smoking's harmful influence on PUD. ; Deaths from duodenal ulcer for smokers of more than a half pack per .: day of cigarettes were 2.5 times the rate for nonsmokers; for those smoking one-half pack per day or less, the ratp was 1.5 times the rate s# for nonsmokers. There were no gastric ulcer d'eaths among nonsmok- ,~ ers; but there were 46 among smokers; the death rate also increased ~ with smoking more than a half pack per day of cigarettes. Thus, _~ smoking was, clearly associated with a higher occurrence of death in ~ both types of ulcer disease. Dorn (20), in another large study, had similar results. The ratio of ~ observed deaths from both duod'enal ulcer and gastric ulcer in smokers '~ 9-10 50

® ~ ~ . ~ ~ :~~:: wur ~~. krl~i.:' to expected deaths from~ these diseases was 2.8. Those who smoked more than two packs per day had more deaths than those who smoked~ one to two packs per day, who in turn fared worse than those who smoked less than one pack per day. In a prospective study of smoking and mortality in 68,153 middle- aged men, Weir and Dunn (64), just as Hammond and Horn (26), found no deaths from gastric ulcer in nonsmokers but a significant number of smokers dying from gastric uleer disease. Their results, however, for duodenal ulcer were completely opposite, in that the relative risk of death from duodenal ulcer in smokers was half that in nonsmokers: Why this discrepancy should exist is not clear. Doll and Peto (19), in a study of more than 10,000 British physicians, found a significant increase in death from peptic ul+eer disease (specific location of uleer not st'ated) in smokers as compared to nonsmokers, with a higher rate in moderate or heavy smokers than in light smokers. Finally, Din and Small (15) proposed that the long-term survival of patients after gastrectomy was decreased by smoking. They felt the increased mortality rate was due to cigarette smoking (and perhaps alcohol, too)~ and not to the operation. The evidence for this is unclear. A summary of the important data from the four studies (19; 20, 26, 64), which bear on the epidemiological question, "Does smoking influence a person's chance of dying from his ulcer disease?" can be found in Table 3. These data show that mortality from gast'ric ulcer is greater in male cigarette smokers than in nonsmokers and, except in one study (64), also is greater in male cigarette smokers with duodenal ulcer disease. In the study that was the exception, the results are clbud'ed by inclusion of ex-smokers in the smoking group. So, in general, it can be concluded that male cigarette smokers have more than a twofold greater chance of dying from ulcer disease than nonsmokers. It is not clear how much of this excess risk is due to the increased prevalence of ulcer disease in smokers an& how much is due to the reduced ability of the smoker to survive an ulcer due to a greater prevalence of chronic heart and lung, disease. The Question of the Etiological Role of Smoking In Peptic Ulcer Disease The studies reviewed have consistently shown an increased; frequency of PUD in smokers as opposed to nonsmokers. In additionj the frequency of PUD rises with increases in the amount smoked'y and smoking appears to retard peptic ulcer healing. All this, of course, does not provide a definitive answer to the question: "Is cigarette smoking a cause of peptic ulcer disease, or is it just associated with a cause such, as genetic predisposition, personality type, and so on?" Epidemiologi- cal, case-control, and genetic studies cannot exclude the possibility that cigarette smoking is only associated with the cause(s) of PUD. An ~ 0 0 ~ 0 yt t I@ 0 0 ~ 8 @ y ® g ® ® ~ 0 9-11 hy ® i- - ~

essential' link in establishing whether cigarette smoking is a causative different from duodenal ulcer (49), what other factors may smoking protect the duodenum; does smoking interfere with this defense ~:~ mechanism? Finally, since the pathogenesis of gastric ulcer may be '"' secretion is of mterest. Pancreatic buffering of acid may serve to ulcer hypersecrete acid (68), so the effect of smoking on gastric acid ..',• marked overlap with: normals, on the average, patients with duodenal that (with rare exceptions) acid must be present (30): Although there is' -1~ why certain patients develop PUD under any condition. We do know effect on physiological mechanisms that might allow an ulcer yt;o develop. This question is difficult to deal with since it is still not known factor in PUD is a convincing demonstration that smoking has aii influence that might alter the stomach s,defenses. Tco~. , pentagastrin and carbachof (50): Nicotine alone did not produce any ulcers in~ the animals. Radecki, et al. (47) studie& the response of cats to nicotine in both the basal an& pentagastrin-stimulated states. Doses of nicotine up to 200 µg/kg did not alter acid secretion in either state. A dose of 400 µg/kg hydrochloric acid perfusion (51) or by subcutaneous infusion of pepsin output. Robert and his colleagues have shown that nicotine can increase the number and severity of duodenal ulcers formed in rats by of nicotine to the chronically treated rats inhibited gastric acid and vagotomy or anterior hypothalamic lesions (57). Acute administration their pepsin~ output (p C 0':01): This effect could be blocked by either cigarettes per day) doubled~ their gastric acid output and increased,; ~ nicotine 3 times daily for 15 days (the equivalent of smoking 10 to 15 also studied' by the same investigators (58). Rats receiving 100 µg/kg percent ett:anol had no effect on acid secretion but reduced pepsin output (56). The effects of chronic nicotine administration in~ rats was Nicotine, 100 µg/kg, injected~ into rats, depresse& histamine-stimu- lated secretion of acid and pepsim It also depressed basal secretion and submaximal pentagastrin-stimulated secretion. Tobacco smoke in 10 topical nicotine. barrier LO back diffusion of hydrogen ions by either intravenous or, effect on mucosal blood flow, and no interruption of the mucosal r; I in either basal aci& output or half-maximal gastric acid secretion stimulated~ by histamine or pentagastrin. In addition, they found no acid secretioni in the fasting state. Konturek, et al. (36) studied the i ~ effect of intravenous nicotine (100 µg/kg) in d'ogs and foun& no change studies (53) in dogs showed that neither cigarette smoking zior subcutaneous injections of 0.2, 0.4, or 1 mg of nicotine increased gastric have been performed in rats, cats, dogs, and man-many with contradictory results even in the same species: One of the earliest Studies of the effects of smoking or nicotine on gastric acid secretion' Gastnc Secretion ' : 9-12

.r ;~. 9= :._ ~ depressed stimulated acid secretion by 30 percent; it also produced restlessness, vomiting, and diarrhea. Nicotine (200 µg/kg) did, however, potentiate the development of pentagastrin-induce& experi- mental duodenal ulcers in these cats (35). Studies of the effects of smoking on acid secretion in human subjects have given contradictory results. Schnedorf and Ivy (53) studied' the effect of acute smoking on acid secretion in 40 normals (smokers and nonsmokers) and in 20 patients with duodenal ulcer. Mean acid output fell: smoking in both the normals and the ulcer patients, but no statisticali analysis was done, so the significance of the decrease eannot be evaluated. Steigmann„ et al. (55) reported that 26 of 44 controls and 40 of 45 ulcer patients increased acid production while smoking an unfiltered cigarette; a control study without smoking was not done. Cooper and Knight (12) recorded no difference in basal acid secretion between 60 patients with duodenal ulcer who smoked during the testt and 60 patients who did' not. Fung and Tye (24) investigated the effect'ss of smoking 3 cigarettes per hour on 16 smokers and 16 nonsmokers, 23 of whom had duodenal ulcer and 7, gastric ulcer. There was no significant difference between basal acid' output and acid output during smoking in either group. Another study showed that smoking four cigarettes an hour did not alter acid, pepsin, or mucus production in either normal subjects or ulcer patients who were smokers (65). This is particularly interesting in that the same laboratory reported different findings 15 years earlier when they found that smoking increased gastric secretion in man (45). Murthy, et al. (40) studied secretory response to smoking one cigarette per 15 minutes for 1 hour in smokers with~ duodenal ulcer and in normal smokers and nonsmok- ers. In the first 15 minutes, there was a significant increase in acid secretion in the ulcer patients. No significant effect was seen in either group of normals. Debas, et al. (14) studied'~ 12 subjects„6 smokers and 6 nonsmokers, of both sexes. The subjects smoked three cigarettes per hour while gastric secretion was maintained at half maximalirate with pentagastrin. Smoking caused no significant change in, mean rate of acid secretion or pepsin secretion in either group. In a: separate study (10); the same investigators found that while cigarettes alone had no effect on acid output, nausea indticed by smoking in nonsmokers did inhibit acU production, Debas and~ Cohen (13) noted that smoking produced substantial inhibition of acid secretion in the majority of subjects during the first test but this could not be reproduced on repeated testing. They suspected that the inhibition was due to nausea, not smoking, per se. They also reported (13) that intravenous infusion of 2 mg of nicotine produced essentially no change in~ pentagastrin- stimulated acid and pepsin secretion in eight subjects. Wilkinson and Johnston (86) also studied the effects of smoking on pentagastrin-stimulated acid secretion and found depression of acid output in response to smoking one or two cigarettes in three groups (38 9-13 P a

percent in normals, 21 percent in duodenal! ulcer patients, and 18 3.- ~ ~ 7a. chronic smoking on acid secretion. . ,- ,-- secretiom A few studies found inhibition of acid secretion by smoking; but these involved first attempts at smoking with, a gastric tube in place. Such procedures often produce nausea which by itself can inhibit acid secretion. There has been no systematic study of the effect of smoking one or a few cigarettes exerts an inconsistent effect on acid - r ~+c , 3s~ In summary most of the studies in human subjects have shown that-x and elevation of blood pressure whilp smokin•"AZRE g. percent in gastric ulcer patients). All subjects experienced tachycardi effect in normal secretors hypersecretors (BAO 5 to 16.5 mEq hr); but produced only a smalt . smoking lowered duodenal pR from a range of 6.2-74 to 1.7-2.5 in five ',~~ that smoking may alter the d'uodenal environment. They found that A, affects pancreatic buffering mechanisms. Murthy, et al. (39) showed .'9D - ., has not been clearly shown to inerease gastric secretion, so perhaps it duodenal ulcer formation by an effect on duodenal acidity. Stnoking development of duodenali ulcers; thus,, smoking might influence ' It is generally accepted that an acid milieu is required for the ~ Pancreatic Secretion tion of pancreatic secretion tion and noted graded inhibition of bicarbonate secretion~ (23 to 62 percent). All values returned to control levels after cessation of the nicotine. Similarly, nicotine (100~µg kglh-1) reduced hepatic bile volume and bicarbonate by 50 percent. In~ a~ subsequent study (34), they reconfirmed that intravenous nicotine reduced the pancreatic response to intravenous secretin, Topical nicotine, however, did not alter the response to secret'in: In addition, as the dose of secretin was increased from .37 to 3 U kg-1 h-1, the inhibition of bicarbonate secretion by intravenous nicotine decreased from 75 to 15 percent. To examine the effect of' nicotine on pancreatic secretion induced by endogenous secretin, pancreatic secretion~ was, stimulated by intraduodenal admin- istration of HCI with a response equivalent to .75 U kg-1 h-1' of intravenous secretin. Both intravenous nicotine and topical nicotine reduced the response to the acid' by about 25 Cpercent. However, nicotine had, no significant effect on cholecystokinin-induced' stimula- intravenously) to dogs on a background of maximal secretin stimula- ~ pancreatic or bilhary secretion in dogs during smoking. Konturek and his colleagues (36) gave graded doses of nicotine (12.5 to 100 µg kg-1 h-1' Schnedorf and Ivy (53) found no significant change in either . Boden and his associates (7) found, in their dog experiments that ~w basal and HCI (9.6 mEq/30 min) stimulated bicarbonate outputs were insignificantly decreased by intravenous infusion of nicotine (100 µg ~ kgl h-1), and nicotine did not decrease bicarbonate output in response to intravenous secretin (1.0' U kg-1 h-I). In addition, nicotine had no 9-14 ' .`:°~,

9 r, : _. ~ Ma ® z.: PJ : significant effect on the serum secretin level (measured by radioimmu- noassay) except to delay the appearance of the peak value. It shouU be noted that Boden used 2.4 times as much acid to stimulate pancreatic secretion as did Konturek, et al. (31,). Solomon, et al. (54) studied the effect of nicotine on the rabbit pancreas. Nicotine infused at rates of 100 to 400 µg kgl' h-1' decreased pancreatic secretion in a dose-dependent fashion. Since nicotine is a stimulant of autonomic ganglia (62), the effect of norepinephrine and epinephrine was studied. Norepinephrine at 2 or 4 µg kgl min-1 and epinephrine at 2 µg kgl' inhibited secretory flow and bicarbonate output. Phenoxybenzamine; an~ a-adrenergic blocker, increased water and bicarbonate secretion and; blocked the inhibitory action of nicotine and norepinephrine on pancreatic secretion. On the basis of these results, they concluded that nicotine indirectly inhibits pancreatic secretion by stimulating catecholamine release, an effect that is negated by alpha adrenergic blockade. The evidence for smoking's effect in man parallels that in animals. Bynum and his colleagues (9) studied the acute effects in light and heavy chronic smokers of smoking four cigarettes an hour on bicarbonate output in response to secretin, The light smokers responded normally to secretin during the controli period but had' decreased pancreatic bicarbonate output while smoking. Heavy smokers had'a decreased response to secretin during the control' period and this was not further affected by smoking. In a: study of subjects who smoked regularly (5); smoking three cigarettes significantly decrease& basal bicarbonate output. Brown (8) investigated the effect of smoking on~ pancreatic secretion in 14 healthy smokers, 7 heavy and 7 light smokers. Heavy smokers had lower responses to secretin (2 U/kg) than light smokers. In addition, smoking cigarettes reduced even further the volume and bicarbonate content of the duodenal juice in~both groups. Murthy, et al. (40) studied the effects of smoking in smokers with and without duodenal ulher and in nonsmokers. They found that smoking depressed basaU bicarbonate and volume in both normalg and patients with duodenal ulcer and in both smokers and nonsmokers. Changes in plasma nicotine were inversely correlated with pancreatic secretion. In addition, smoking had no effect oni gastrin or secret'inn levels as measured! by radioimmunoassay. Bloom and Ward (4) reported depressed secretin release irt response to intraduodenal acid instillation in patients with~ dtzodenal ulcer in contrast to controls. Actually, the increase in secretin over basal values was approximately the same in the ulcer patients as in the normal controls. Those patients who smoked more had smaller peak secretin values than lighter smokers. There was no difference in secretin release between smoking and nonsmoking controls. A subsequent study by Isenberg, et al. (29), using the same radioimmunoassay for 9-15 I m I I ~ 0 U

secretin, did not demonstrate a difference in secretin release between - I Four studies in man (5, 8, 9, 40) all, show decreases in bicarbonate output in response to smoking. There is no evidence that this is due to duodenal ulcer patients and normals. In light of this, the purported' effect of smoking on secretin release must be questioned: inhibition of secretin release. Pvloric Reflux and Gastric Ulcer `' What is smoking's relationship to the pathogenesis of gastric ulcer? The possible causes of gastric ulcer have been reviewed (49); and several' hypotheses have been proposed. Various pharmacologic agents have been shown to disrupt the mucosal barrier to back diffusionl of hydrogen ions, whichl might contribute to the development of gastric ulcer. However, no such effect has been demonstrated with smoking 7S (36). Another hypothesis is that excessive reflux of duodenal contents, i.e. bile and pancreatic juice, through an incompetent pyloric sphincter, Other possible etiological relationships have been examined. Ed- wards andl Coghill (21) found that chronic atrophic gastritis was twice as common in persons who smoked more than 20 cigarettes a day as in nonsmokers. Since the majority of patients with gastric ulcer have chronic atrophic gastritis (1), smoking may predispose to gastric ulcer the degree of bile reflux in gastric ulcer patients (16); gastric aspirates during the hour of smoking as compared to the control hour. Dippy an& his colleagues found that smoking increased gastr~c secretion, a so notice more mar e i e s aimng o eir t ll , r u o g s pp p y r.~.. found that smoking increased radiologic evidence of duodenogastric reflux. Whitecross„ et al; (65), while studying the effect of smoking on 1 d' k d b'1 t'' f h ' ~~- rted evious work b ftead and Grech (J H(61) This 8) who ' cigarette decreased basal pressure significantly from 10.2 to 7.9 mm ~ may be implicated inl the pathogenesis of gastric ulcer (52). Recently, manometric studies of the human pylorus showed that smoking one by producing chronic atrophic gastritis, which in turn may be a ~; into the stomach, which could; be relevant in the light of the hypothesis ulcer. Smoking also appears to increase reflux of duodenal contents '~ ~ mechanism by which smoking could favor occurrence of duodenal' U*P capacity to neutralize gastric acid is a plausible but unproven --,A inhibit pancreatic secretion of bicarbonate; the consequent lowered 1 little or no effect of smoking on acid secretion. Smoking and nicotine I If smoking does indeed influence the development and course of pepticc ulcer disease, how does it do so? Experiments investigating the effect of smoking and~ nicotine on gastrointestinal function in animals and man have not established conclusively any mechanisms by which smoking might contribute to peptic ulcer forrnation: Most studies show Summary r. precursor of gastric ulcer. 9-16