- Email: [email protected]
The Effects of Food Deprivation, Nutritive and Non-Nutritive Feeding and Wheel Running on Gastric Stress Ulcers in Rats
Physiology & Behavior, Vol. 63, No. 2, pp. 219 –225, 1998 © 1998 Elsevier Science Inc. All rights reserved. Printed in the U.S.A. 0031-9384/98 $19.00 1 .00
PII S0031-9384(97)00427-7
The Effects of Food Deprivation, Nutritive and Non-Nutritive Feeding and Wheel Running on Gastric Stress Ulcers in Rats IMGAP YI AND FRIEDRICH K. STEPHAN1 Neuroscience Program, Department of Psychology, Florida State University, Tallahassee, FL 32306-1270, USA Received 21 November 1996; Accepted 22 July 1997 YI, I. AND F. K. STEPHAN.The effects of food deprivation, nutritive and non-nutritive feeding and wheel running on gastric ulcers in rats.PHYSIOL BEHAV 63(2) 219 –225, 1998.—Feeding and housing conditions that induce gastric lesions were investigated. Rats were housed in activity wheels or in hanging cages and exposed to food deprivation, ad lib cellulose or 6 g of cellulose per day for 5 days. Food-deprived rats in both housing conditions had ulcers in the rumen but many rats also had mucosal ulcers. Cellulose prevented rumenal ulcers but produced a tendency toward more severe mucosal ulcers. Ulcers in wheel-housed rats were somewhat larger but the difference was not significant. In a second experiment, rats were fed 6 g/day laboratory chow or 6 g/day chow 1 ad lib cellulose until b.wt. reached a preset criterion. On the average, about 10 days on the feeding regimen were required to induce ulcers in these groups. None of the rats had rumenal ulcers. Mucosal ulcers were reliably larger in rats that received cellulose in addition to 6 g of chow. There was no difference in ulcer area between wheel-housed and cage-housed rats. The results indicate that solid bulk, regardless of its caloric value or amount, protects the nonglandular stomach whereas noncaloric bulk tends to aggravate ulcers in the glandular stomach. A small amount of chow delays the rate of b.wt. loss and consequently ulcer formation. Furthermore, wheel running is not necessary to produce mucosal ulcers when food intake is insufficient to maintain b.wt. and b.wt. at sacrifice seems to be a good predictor of ulcer formation. © 1998 Elsevier Science Inc. Activity-stress ulcer Food deprivation
Stomach
Gastric mucosa
Corpus
IN rats, lesions induced by physical restraint (12) or by an activitystress procedure (11) develop in the lower, glandular portion of the stomach (corpus) and penetrate the muscularis mucosa of the gastric lining (8,16), characteristic of human peptic ulcers (4). On the other hand, earlier studies indicate that pylorus ligation or total food deprivation induces gastric lesions in the upper, nonglandular portion (rumen) of the stomach (13,14). It has also been reported that when electric shock is used as a stressor, rats given a liquid diet had more ulcers in the rumen than those given a solid diet, suggesting that ulcers in the corpus and rumen do not have a similar etiology (6). Solid food and even solid non-nutritive bulk are believed to decrease the incidence of nonglandular ulcers induced by food deprivation, presumably by reducing acidity in the rumen of the stomach (10). Interestingly, food-deprived rats housed in cages containing sawdust bedding developed ulcers confined to the glandular portion of the stomach (which contained a considerable amount of sawdust at sacrifice), whereas those housed in cages with grid floors had ulcers in both the rumenal and the glandular portion (5). At this time, it is not certain whether restricted feeding of solids is protective against ulceration in both nonglandular and glandular 1
Rumen
Cellulose
Restricted feeding
parts of the stomach or whether solids are protective for the rumen on the one hand, but offensive to the glandular stomach on the other hand. Recently, we found that 5 days of restricted feeding of laboratory chow (6 g per day) resulted in ulcers exclusively in the glandular stomach not only in rats housed in activity wheels but also in those housed in hanging cages without wheel access (18,20). This finding indicates that a deficient energy input leads to glandular ulceration, but that 6 g of powdered chow seems to provide sufficient solidity to protect the rumen from ulceration. However, in contrast to earlier reports (13,14), these results suggest that total food deprivation should produce ulcers in the glandular stomach in addition to rumenal ulcers. One purpose of the present study was to separately investigate the role of total food deprivation, of the solidity of diets and of caloric content in the generation of gastric ulcerations. The amount of food intake was manipulated by exposing rats to total deprivation, 6 g/day feeding or ad lib feeding and two different diets were used to manipulate caloric value: cellulose (non-nutritive) and laboratory chow. It has been suggested that activity-stress ulcers are the result of excessive running and that activity is the critical factor in this pathology. This view is supported by the observation that highly active rats in activity wheels develop the most severe ulcers and
To whom requests for reprints should be addressed.
219
220
YI AND STEPHAN
that food-yoked control rats in hanging cages that receive the same amount of food as paired wheel-housed rats rarely have ulcers (for review, see (7)). However, we recently reported that restricted feeding consistently results in glandular ulcers in rats even without access to activity wheels and that terminal body weight (indicative of the decrease of body fat reserves) is highly correlated with ulcer severity regardless of wheel access, suggesting that depletion of body fat reserves may be a common factor underlying the development of glandular ulcers in our restricted feeding procedure and in the activity-stress procedure (18 –20). Nevertheless, activity wheels seem to provide an ulcer-aggravating factor since wheelhoused rats had more severe ulcers than cage-housed rats even when their terminal body weights were matched by keeping the latter on restricted feeding for 3 additional days (20). Therefore, another purpose of this study was to conduct a comparison between the activity-stress procedure and the restricted feeding procedure in hanging cages. However, rather than restricting the feeding time, the amount of food was fixed by providing 6 g/day. As yet, no activity-stress studies have investigated the effect of total food deprivation or non-nutritive diets on resulting ulcers. Under each housing condition (activity wheel or hanging cage), rats were exposed to feeding conditions that differed in volume and in caloric value. METHODS
Animals and Housing Sixty-five male Sprague Dawley rats, weighing about 270 g, were housed in a colony maintained at a constant ambient temperature (22 6 2°C) with a standard light-dark 12:12 cycle (light on 0700 h). Food (Purina Rat Chow, No. 5001, 3.6 kcal/g) and water were available ad lib. After 10 days of habituation, 33 rats were housed individually in standard hanging cages. Access to food was controlled manually by securing a glass food container in the cage. Contact by the animal with a metal ring around the neck of the food cup completed a circuit and was recorded by computer. For another 32 rats, activity wheels with a small adjacent cage were used. Wheel revolutions were counted and stored in 10 min bins. Food access time (3 h/day beginning at 0900 h) was computer controlled by means of a motorized sliding door.
without risking mortality. Two of the wheel-housed rats receiving chow 1 cellulose were sacrificed at higher b.wt. (177 and 179 g; about 2 days before they would have reached the b.wt. criterion) because of a severe reduction in wheel running and lethargic appearance. These rats had the largest ulcers among rats in Experiment 2. For both experiments, cellulose fiber (Dyet™; No. 401850) was mixed with 0.2% saccharine solution (500 mL/kg of cellulose) to increase its palatability and several drops of mineral oil (Crystal™) as a binder to reduce spillage. Preliminary observations showed that rats consumed 20 to 30 g of this mixture per day. Fresh cellulose and laboratory chow were supplied daily at 0900 h. For rats given laboratory chow with cellulose ad lib, 6 g of chow were given at 0900 h, and then free access to cellulose was allowed from 1100 h to the next day at 0900 h. Both experiments were conducted simultaneously with rats placed pseudorandomly in cages and activity wheels. Measurement of ulcerations. Rats were euthanized with an overdose of pentobarbital. After the esophagus and duodenum were ligated, the stomach was inflated by injection of 8 mL of 10% formalin solution. The stomach was then removed, cut along the greater curvature, rinsed with 10% formalin, pinned flat on a paraffin bed and preserved in 10% formalin. All stomachs were inspected under a low power light microscope, and loose blood clots were washed away with a 10% formalin solution. Ulcers with or without blood clots were easily identified as depressions and erosions in the mucosa. An image processing system (Macintosh II Quick Capture Board with S/W Image 1.22) was then used to outline the ulcerated portions of the stomach and to quantify the total area of ulceration. This measure reflects a combination of the number and the size of ulcers and provides a good index of the severity of the lesions. Data Analysis Data were analyzed by ANOVA with Bonferroni post hoc comparisons when appropriate. For most comparisons involving ulcer areas, p values for unequal variance (Levene’s test) are reported. Rats without ulcers were included in the data analysis. RESULTS
Procedures
Experiment 1
Experiment 1 (food deprivation/noncaloric feeding). Twenty rats were housed in hanging cages (H-C)and 19 rats were housed in activity wheels (A-W). Each housing condition had three subgroups matched for b.wt.: Group 1, total food deprivation [TFD: H-C (n 5 6) and A-W (n 5 6) rats]; Group 2, cellulose feeding of 6 g/day [CEL6: H-C (n 5 7) and A-W (n 5 7) rats]; Group 3, cellulose ad lib feeding [CELAL: H-C (n 5 7) and A-W (n 5 6) rats]. Rats given cellulose (6 g or ad lib) were in these feeding conditions for 4 days and were sacrificed 24 h after the last meal. Food-deprived rats were sacrificed at 0900 h after 5 days. Experiment 2 (caloric feeding). Thirteen rats were housed in hanging cages and 13 rats were housed in activity wheels. Each housing condition had two subgroups matched for body weight: Group 1, restricted feeding of 6 g/day powdered laboratory chow [RF6: H-C (n 5 7) and A-W (n 5 7) rats]; Group 2, restricted feeding of 6 g/day powdered laboratory chow plus cellulose ad lib feeding [R 1 C: H-C (n 5 6) and A-W (n 5 6) rats]. For all four groups, the decision to sacrifice was based on results from our previous work (18 –20). The main criterion was b.wt. loss to a predetermined level, i.e., about 165 g. In addition, rats were carefully monitored for any behavioral changes of distress. This ensured that all or most rats in these groups had gastric ulcers
Ulcerations. Only TFD rats in both H–C and A–W groups developed ulcers in the rumen. In the H–C group one of six rats had only nonglandular ulcers, another one had only glandular ulcers and four rats had both. In the A–W group three of six TFD rats had only nonglandular ulcers, one had only glandular ulcers and two rats had both. The other four subgroups which received cellulose developed solely glandular ulcers, regardless of the presence or absence of wheel access. Figure 1 shows representative animals from these groups. The ulcerated areas for individual rats are shown in Fig. 3 (top) and mean values are presented in Table 1. When the area of glandular ulcers was compared, the effect of feeding fell just short of significance [F(2,33) 5 2.42, p , 0.10], while housing was clearly not significant. Despite the small number of subjects and considerable variability, a meaningful trend toward increased mucosal ulcerations in the wheel-housed ad lib cellulose group seems to be present. A post hoc comparison between wheel-housed groups indicated that the ad lib cellulose group had a trend toward larger ulcers than the 6 g of cellulose group (p , 0.08). None of the other comparisons approached significance. Activity. The number of wheel revolutions increased sharply over the first 3 days and then declined for most rats on the fourth
FEEDING CONDITIONS AND STRESS ULCERS
221
FIG. 1. Representative stomachs of rats exposed to total food deprivation (TFD, top), 6 g of cellulose per day (CEL6, middle) or ad lib cellulose (CELAL). H and A designate hanging cage or activity wheel housing, respectively (some lesions in the nonglandular stomach are not visible in these photographs). Most lesions in the glandular stomach have tightly adhering blood clots.
and fifth day. A comparison of peak wheel revolutions (Day 3 or 4) revealed no significant differences among the three feeding conditions (Table 1). The number of contacts with the food jar in the hanging cage-housed groups also was not significantly different among groups. Both measures showed a shift toward increased activity into the light phase of the light-dark cycle on the last 2–3 days of the experiment. Terminal body weight. There was a significant main effect of feeding conditions on body weight [F (2,33) 5 3.9, p , 0.03]. The ad lib cellulose groups had higher terminal body weights than the other groups (p , 0.05). However, at sacrifice, the stomach and
intestines still contained undigested material, which most likely accounted for this difference. Experiment 2 Ulcerations. Figure 2 shows representative ulcers from the four groups. Ulcers for individual rats are shown in Fig. 3 (bottom) and means are presented in Table 2. There was a main effect of feeding on ulcerations [F(1,22) 5 6.99, p , 0.02] but the effect of housing was not significant. Post hoc comparisons indicated that ulcers in the hanging cage-housed rats receiving 6 g of chow and cellulose ad lib were larger than those fed 6 g of chow alone (p , 0.05).
222
YI AND STEPHAN TABLE 1 EFFECTS OF NONCALORIC FEEDING AND HOUSING
Hanging cage groups TFD CEL6 CELAL Activity wheel groups TFD CEL6 CELAL
n
Rats with Ulcers
Area of Ulcers
Initial b.wt. (g)
Final b.wt. (g)
6
5 5 (R) 6 5
2.43 (1.39)* 5.51 (2.25) 8.42 (5.71) 7.92 (4.28)
229.8 (1.38)
177.8 (2.68)
NA
227.6 (1.87) 232.2 (1.29)
173.7 (2.62) 184.3 (2.43)
NA NA
229.2 (5.70)
171.0 (6.77)
9304 (2436)
238.6 (2.46) 235.5 (3.10)
179.2 (4.92) 187.1 (4.22)
10487 (2840) 9659 (2893)
7 7 6 7 6
3 5 (R) 5 5
6.58 (5.14) 3.46 (1.05) 5.20 (2.10) 21.25 (7.41)
Peak Wheel Revolutions
* Values are means with SEM in parentheses. CELAL 5 ad lib cellulose; (R) 5 ulcers in rumen (all other ulcers were in the glandular stomach). Initial b.wt. on Day 1 of the feeding condition.
However, a similar trend in the wheel-housed groups was not significant. The number of days to reach terminal body weights at sacrifice was not significantly different among groups. Activity. Peak levels of wheel running (Table 2) were significantly higher for rats receiving cellulose in addition to 6g chow than for rats receiving chow alone (p , 0.02). As in Experiment 1, wheel running and food jar contacts became more diurnal near the end of the experiment. Terminal body weights. Since rats were sacrificed near prede-
termined body weights, there were no significant differences among the groups. DISCUSSION
Total food deprivation led to ulcerations in both nonglandular and glandular portions of the stomach in one-half of the rats, but one rat each had ulcers exclusively in one portion or another. Although it has been reported that total food deprivation causes
FIG. 2. Representative stomachs of rats exposed to restricted feeding of 6 g/day chow (RF6, top), or 6 g/day chow and ad lib cellulose (R 1 C, bottom). H and A designate hanging cage and activity wheel housing, respectively.
FEEDING CONDITIONS AND STRESS ULCERS
223
FIG. 3. Total ulcerated area of the glandular stomach for individual rats in different feeding conditions. TFD 5 total food deprivation; CEL-6 5 6 g of cellulose per day; CEL-AL 5 ad lib cellulose; RF-6 5 restricted feeding of 6 g of chow per day; R 1 C 5 6 g of chow and ad lib cellulose.
only nonglandular ulcers, a careful reading of earlier studies reveals that both types of ulcers were observed, but with a dominance of nonglandular ulcers (13,14). Another study reports that prefeeding attenuates cold stress lesions in the glandular stomach and that food deprivation of nonstressed rats produced extensive rumenal lesions but also some glandular lesions after 2 to 6 days (17). Both the duration of food deprivation and the terminal body weight of the rats could influence whether or not glandular lesions occur. The present results confirm the prediction that food deprivation, like feeding of small amounts of chow, can lead to ulcers
in the mucosa of the rat stomach even in the absence of activity wheels (18 –20). In addition, the results show that small amounts of cellulose or ad lib cellulose protect the rumen from ulceration. On the other hand, ad lib cellulose produced a trend toward larger mucosal ulcers in rats housed in activity wheels. Restricted feeding of 6 g/day laboratory chow led to glandular (but not rumenal) ulcerations in both cage- and wheel-housed rats, consistent with our previous findings that a reduction in chow intake to 6 g/day is sufficient to cause glandular ulcers (18 –20). However, nearly twice as many days were required to induce
224
YI AND STEPHAN TABLE 2 EFFECTS OF RESTRICTED CALORIC FEEDING AND HOUSING
Hanging cage groups RF6 R 1 CEL Activity wheel groups RF6 R 1 CEL
n
Days to Sacrifice
Rats with Ulcers
Area of Ulcers
Initial Bodyweight (g)
Final Bodyweight (g)
7 6
9.6 (0.61)* 11.3 (1.25)
5 5
1.10 (0.50) 9.24 (3.06)
234.9 (2.56) 235.5 (1.87)
164.7 (2.09) 165.2 (2.62)
7 6
9.4 (1.11) 9.7 (1.20)
5 6
3.02 (1.96) 13.38 (6.55)
230.9 (2.54) 236.5 (3.73)
164.7 (2.09) 168.7 (2.82)
Peak Wheel Revolutions
NA NA 5586 (1350) 11440 (1833)
* Values are means with SEM in parentheses. R 1 CEL 5 6 g of chow 1 ad lib cellulose. Initial b.wt. is on Day 1 of the feeding condition.
ulcers compared to the noncaloric feeding groups in Experiment 1. This shows that a small amount of food significantly delays ulcer formation. As in Experiment 1, the addition of cellulose to 6 g of chow produced larger ulcerated areas. A protective effect against cold-restraint stress ulcers can be obtained by the administration of milk (3) or fish oil (1). Taken together, these findings indicate that ulcer severity may depend on the degree of caloric intake and that the location of ulcers depends mostly on the solidity of ingested food; i.e., solid food seems to protect the nonglandular stomach from ulceration in agreement with previous observations (9,10). However, the solidity of diets per se does not automatically protect the glandular stomach from damage; rather it can make the stomach more susceptible to ulceration, perhaps by stimulating gastric contractions and acid secretion. Observations in our laboratory show that glandular ulcers develop mostly along ridges (Figs. 1 and 2), which form a linear pattern in the glandular portion of the stomach and cross the arteries perpendicularly. It is possible that arteries or capillaries are susceptible to local ischemia due to hypermotility. Ischemia has been implicated in ulcerogenesis by a number of investigators (2,15). Our results, considered together with previous findings by others (5,6,9,10), suggest that ingestion of solid diets protects the rumen from ulceration. However, the ingestion of non-nutritive bulk seems to make the lower, glandular stomach more susceptible to ulceration. The mechanisms of these two different ulcerations have yet to be discovered.
The type and pattern of ulcers were very similar in all groups in agreement with our previous findings (18 –20). However, in the present study, access to activity wheels did not induce significantly larger ulcers. In part, this may be due to the small number of subjects and the large variability in daily wheel revolutions. In Experiment 2, peak activity was significantly higher for rats fed chow and cellulose, but compared to chow-fed rats the area of ulcers fell short of statistical significance. A second factor was that the terminal b.wt. of cage- and wheel-housed rats were nearly identical; this measure is a good predictor of ulcer area. The variability in ulcer areas within groups suggests that these ulcers develop very quickly, rapidly enlarge over 1 or 2 days and would quickly lead to mortality. Overall, these results strongly suggest that activity-stress ulcers share a common cause with ulcers induced by restricted feeding in hanging cage-housed rats; i.e., decreased food intake and subsequent depletion of body fat reserves may trigger a common ulcerogenic process. The increase in activity observed in wheelhoused rats may be a manifestation of metabolic stress (impending starvation) but is not necessary for ulcer formation. The origin and nature of the hypothesized stress signal remain to be determined but are highly correlated with b.wt. and fat reserves (19). ACKNOWLEDGEMENTS
Work was funded by the FSU Foundation. The authors thank Dr. Karen Glendenning for the use of an image processing system for ulcer assessment and Mr. Charles Badland for photography.
REFERENCES 1. Al-Harbi, M. M.; Islam, M. W.; Al-Shabanah, O. A.; Al-Gharably, N. M. Effect of acute administration of fish oil (omega-3 triglyceride) on gastric ulceration and secretion induced by various ulcerogenic and necrotizing agents in rats. Food Chem. Toxicol. 33:553–558; 1995. 2. Caldwell, M. T. P.; Suart, R. C.; Byrne, P. J.; Gorey, P. F.; Hennessy, T. P. J. Microvascular changes in experimental gastric stress ulceration: The influence of allopurinol, cimetidine, and misoprostol. J. Surg. Res. 55:135–139; 1993. 3. Dial, E. J.; Romero, J. J.; Lichtenberger, L. M. Gastroprotection by dairy foods against stress-induced ulcerogenesis in rats. Dig. Dis. Sci. 40:2295–2299; 1995. 4. Doughty, D. B.; Jackson, D. B. Gastrointestinal disorders. St. Louis: Mosby; 1993:86 – 88. 5. Driscoll, P.; Martin, J. R.; Kugler, P.; Baettig, K. Environmental and genetic effects on food-deprivation induced stomach lesions in male rats. Physiol. Behav. 31:225–228; 1983. 6. Glavin, G. B.; Mikhail, A. A. Stress and ulcer etiology in the rat. Physiol. Behav. 16:135–139; 1976.
7. Lambert, K. G. The activity-stress paradigm: Possible mechanisms and applications. J. Gen. Psychol. 120:21–32; 1993. 8. Manning, F. J.; Wall, H. G.; Montgomery, C. A. Jr.; Simmons, C. J.; Sessons, G. R. Microscopic examination of the activity-stress ulcer in the rat. Physiol. Behav. 21:269 –274; 1978. 9. Mikhail, A. A.; Gabriel, R. M.; Glavin, G. B. Ulcer reduction by non-nutritive bulk in pylorus ligated rats. Physiol. Behav. 19:437– 439; 1977 10. Mikhail, R. R.; Hirchberg, J. Ulceration in the rat’s forestomach: Its reduction by non-nutritive bulky substances. Physiol. Behav. 8:769 – 770; 1972. 11. Pare, W. P. The influence of food consumption and running activity on the activity-stress ulcer in the rat. Dig. Dis. 20:262–273; 1975. 12. Pare, W. P.; Glavin, G. B. Restraint stress in biomedical research: A review. Physiol. Behav. 10:339 –370; 1986. 13. Pare, W. P.; Temple, L. J. Food deprivation, shock stress and stomach lesions in the rat. Physiol. Behav. 11:371–375; 1973. 14. Pfeiffer, C. J.; Debro, F. R.; Muller, P. J. Gastric pathologic and
FEEDING CONDITIONS AND STRESS ULCERS biochemical changes induced by starvation of weanling rats. Life Sci. 5:509 –519; 1966. 15. Piasecki, C. K.; Thrasivoulou, C. Spasm of gastric muscularis mucosae might play a key role in causing focal mucosal ischemia and ulceration. Dig. Dis. Sci. 38:1183–1189; 1993. 16. Pintar, K.; Schulte, W. J.; Barboriak, J. J. Histologic study of gastric lesions in food-restricted rats. Proc. Soc. Exp. Biol. Med. 145:1353– 1358; 1974. 17. Wozniak, D. F.; Goldstein, R. Effects of deprivation duration and
225 prefeeding on gastric stress erosions in the rat. Physiol. Behav. 24: 231–236; 1980. 18. Yi, I.; Bays, M. E.; Stephan, F. K. Stress ulcers in rats: The role of food intake, body weight, and time of day. Physiol. Behav. 54:375–381; 1993. 19. Yi, I.; Stephan, F. K. Body fat reserves attenuate gastric ulcers induced by restricted feeding in rats. Physiol. Behav. 59:931–936; 1996. 20. Yi, I.; Stephan, F. K.; Bays, M. E. Activity-stress ulcers in rats: The role of pre-entrainment to meal time. Physiol. Behav. 58:67–73; 1995.
PII S0031-9384(97)00427-7
The Effects of Food Deprivation, Nutritive and Non-Nutritive Feeding and Wheel Running on Gastric Stress Ulcers in Rats IMGAP YI AND FRIEDRICH K. STEPHAN1 Neuroscience Program, Department of Psychology, Florida State University, Tallahassee, FL 32306-1270, USA Received 21 November 1996; Accepted 22 July 1997 YI, I. AND F. K. STEPHAN.The effects of food deprivation, nutritive and non-nutritive feeding and wheel running on gastric ulcers in rats.PHYSIOL BEHAV 63(2) 219 –225, 1998.—Feeding and housing conditions that induce gastric lesions were investigated. Rats were housed in activity wheels or in hanging cages and exposed to food deprivation, ad lib cellulose or 6 g of cellulose per day for 5 days. Food-deprived rats in both housing conditions had ulcers in the rumen but many rats also had mucosal ulcers. Cellulose prevented rumenal ulcers but produced a tendency toward more severe mucosal ulcers. Ulcers in wheel-housed rats were somewhat larger but the difference was not significant. In a second experiment, rats were fed 6 g/day laboratory chow or 6 g/day chow 1 ad lib cellulose until b.wt. reached a preset criterion. On the average, about 10 days on the feeding regimen were required to induce ulcers in these groups. None of the rats had rumenal ulcers. Mucosal ulcers were reliably larger in rats that received cellulose in addition to 6 g of chow. There was no difference in ulcer area between wheel-housed and cage-housed rats. The results indicate that solid bulk, regardless of its caloric value or amount, protects the nonglandular stomach whereas noncaloric bulk tends to aggravate ulcers in the glandular stomach. A small amount of chow delays the rate of b.wt. loss and consequently ulcer formation. Furthermore, wheel running is not necessary to produce mucosal ulcers when food intake is insufficient to maintain b.wt. and b.wt. at sacrifice seems to be a good predictor of ulcer formation. © 1998 Elsevier Science Inc. Activity-stress ulcer Food deprivation
Stomach
Gastric mucosa
Corpus
IN rats, lesions induced by physical restraint (12) or by an activitystress procedure (11) develop in the lower, glandular portion of the stomach (corpus) and penetrate the muscularis mucosa of the gastric lining (8,16), characteristic of human peptic ulcers (4). On the other hand, earlier studies indicate that pylorus ligation or total food deprivation induces gastric lesions in the upper, nonglandular portion (rumen) of the stomach (13,14). It has also been reported that when electric shock is used as a stressor, rats given a liquid diet had more ulcers in the rumen than those given a solid diet, suggesting that ulcers in the corpus and rumen do not have a similar etiology (6). Solid food and even solid non-nutritive bulk are believed to decrease the incidence of nonglandular ulcers induced by food deprivation, presumably by reducing acidity in the rumen of the stomach (10). Interestingly, food-deprived rats housed in cages containing sawdust bedding developed ulcers confined to the glandular portion of the stomach (which contained a considerable amount of sawdust at sacrifice), whereas those housed in cages with grid floors had ulcers in both the rumenal and the glandular portion (5). At this time, it is not certain whether restricted feeding of solids is protective against ulceration in both nonglandular and glandular 1
Rumen
Cellulose
Restricted feeding
parts of the stomach or whether solids are protective for the rumen on the one hand, but offensive to the glandular stomach on the other hand. Recently, we found that 5 days of restricted feeding of laboratory chow (6 g per day) resulted in ulcers exclusively in the glandular stomach not only in rats housed in activity wheels but also in those housed in hanging cages without wheel access (18,20). This finding indicates that a deficient energy input leads to glandular ulceration, but that 6 g of powdered chow seems to provide sufficient solidity to protect the rumen from ulceration. However, in contrast to earlier reports (13,14), these results suggest that total food deprivation should produce ulcers in the glandular stomach in addition to rumenal ulcers. One purpose of the present study was to separately investigate the role of total food deprivation, of the solidity of diets and of caloric content in the generation of gastric ulcerations. The amount of food intake was manipulated by exposing rats to total deprivation, 6 g/day feeding or ad lib feeding and two different diets were used to manipulate caloric value: cellulose (non-nutritive) and laboratory chow. It has been suggested that activity-stress ulcers are the result of excessive running and that activity is the critical factor in this pathology. This view is supported by the observation that highly active rats in activity wheels develop the most severe ulcers and
To whom requests for reprints should be addressed.
219
220
YI AND STEPHAN
that food-yoked control rats in hanging cages that receive the same amount of food as paired wheel-housed rats rarely have ulcers (for review, see (7)). However, we recently reported that restricted feeding consistently results in glandular ulcers in rats even without access to activity wheels and that terminal body weight (indicative of the decrease of body fat reserves) is highly correlated with ulcer severity regardless of wheel access, suggesting that depletion of body fat reserves may be a common factor underlying the development of glandular ulcers in our restricted feeding procedure and in the activity-stress procedure (18 –20). Nevertheless, activity wheels seem to provide an ulcer-aggravating factor since wheelhoused rats had more severe ulcers than cage-housed rats even when their terminal body weights were matched by keeping the latter on restricted feeding for 3 additional days (20). Therefore, another purpose of this study was to conduct a comparison between the activity-stress procedure and the restricted feeding procedure in hanging cages. However, rather than restricting the feeding time, the amount of food was fixed by providing 6 g/day. As yet, no activity-stress studies have investigated the effect of total food deprivation or non-nutritive diets on resulting ulcers. Under each housing condition (activity wheel or hanging cage), rats were exposed to feeding conditions that differed in volume and in caloric value. METHODS
Animals and Housing Sixty-five male Sprague Dawley rats, weighing about 270 g, were housed in a colony maintained at a constant ambient temperature (22 6 2°C) with a standard light-dark 12:12 cycle (light on 0700 h). Food (Purina Rat Chow, No. 5001, 3.6 kcal/g) and water were available ad lib. After 10 days of habituation, 33 rats were housed individually in standard hanging cages. Access to food was controlled manually by securing a glass food container in the cage. Contact by the animal with a metal ring around the neck of the food cup completed a circuit and was recorded by computer. For another 32 rats, activity wheels with a small adjacent cage were used. Wheel revolutions were counted and stored in 10 min bins. Food access time (3 h/day beginning at 0900 h) was computer controlled by means of a motorized sliding door.
without risking mortality. Two of the wheel-housed rats receiving chow 1 cellulose were sacrificed at higher b.wt. (177 and 179 g; about 2 days before they would have reached the b.wt. criterion) because of a severe reduction in wheel running and lethargic appearance. These rats had the largest ulcers among rats in Experiment 2. For both experiments, cellulose fiber (Dyet™; No. 401850) was mixed with 0.2% saccharine solution (500 mL/kg of cellulose) to increase its palatability and several drops of mineral oil (Crystal™) as a binder to reduce spillage. Preliminary observations showed that rats consumed 20 to 30 g of this mixture per day. Fresh cellulose and laboratory chow were supplied daily at 0900 h. For rats given laboratory chow with cellulose ad lib, 6 g of chow were given at 0900 h, and then free access to cellulose was allowed from 1100 h to the next day at 0900 h. Both experiments were conducted simultaneously with rats placed pseudorandomly in cages and activity wheels. Measurement of ulcerations. Rats were euthanized with an overdose of pentobarbital. After the esophagus and duodenum were ligated, the stomach was inflated by injection of 8 mL of 10% formalin solution. The stomach was then removed, cut along the greater curvature, rinsed with 10% formalin, pinned flat on a paraffin bed and preserved in 10% formalin. All stomachs were inspected under a low power light microscope, and loose blood clots were washed away with a 10% formalin solution. Ulcers with or without blood clots were easily identified as depressions and erosions in the mucosa. An image processing system (Macintosh II Quick Capture Board with S/W Image 1.22) was then used to outline the ulcerated portions of the stomach and to quantify the total area of ulceration. This measure reflects a combination of the number and the size of ulcers and provides a good index of the severity of the lesions. Data Analysis Data were analyzed by ANOVA with Bonferroni post hoc comparisons when appropriate. For most comparisons involving ulcer areas, p values for unequal variance (Levene’s test) are reported. Rats without ulcers were included in the data analysis. RESULTS
Procedures
Experiment 1
Experiment 1 (food deprivation/noncaloric feeding). Twenty rats were housed in hanging cages (H-C)and 19 rats were housed in activity wheels (A-W). Each housing condition had three subgroups matched for b.wt.: Group 1, total food deprivation [TFD: H-C (n 5 6) and A-W (n 5 6) rats]; Group 2, cellulose feeding of 6 g/day [CEL6: H-C (n 5 7) and A-W (n 5 7) rats]; Group 3, cellulose ad lib feeding [CELAL: H-C (n 5 7) and A-W (n 5 6) rats]. Rats given cellulose (6 g or ad lib) were in these feeding conditions for 4 days and were sacrificed 24 h after the last meal. Food-deprived rats were sacrificed at 0900 h after 5 days. Experiment 2 (caloric feeding). Thirteen rats were housed in hanging cages and 13 rats were housed in activity wheels. Each housing condition had two subgroups matched for body weight: Group 1, restricted feeding of 6 g/day powdered laboratory chow [RF6: H-C (n 5 7) and A-W (n 5 7) rats]; Group 2, restricted feeding of 6 g/day powdered laboratory chow plus cellulose ad lib feeding [R 1 C: H-C (n 5 6) and A-W (n 5 6) rats]. For all four groups, the decision to sacrifice was based on results from our previous work (18 –20). The main criterion was b.wt. loss to a predetermined level, i.e., about 165 g. In addition, rats were carefully monitored for any behavioral changes of distress. This ensured that all or most rats in these groups had gastric ulcers
Ulcerations. Only TFD rats in both H–C and A–W groups developed ulcers in the rumen. In the H–C group one of six rats had only nonglandular ulcers, another one had only glandular ulcers and four rats had both. In the A–W group three of six TFD rats had only nonglandular ulcers, one had only glandular ulcers and two rats had both. The other four subgroups which received cellulose developed solely glandular ulcers, regardless of the presence or absence of wheel access. Figure 1 shows representative animals from these groups. The ulcerated areas for individual rats are shown in Fig. 3 (top) and mean values are presented in Table 1. When the area of glandular ulcers was compared, the effect of feeding fell just short of significance [F(2,33) 5 2.42, p , 0.10], while housing was clearly not significant. Despite the small number of subjects and considerable variability, a meaningful trend toward increased mucosal ulcerations in the wheel-housed ad lib cellulose group seems to be present. A post hoc comparison between wheel-housed groups indicated that the ad lib cellulose group had a trend toward larger ulcers than the 6 g of cellulose group (p , 0.08). None of the other comparisons approached significance. Activity. The number of wheel revolutions increased sharply over the first 3 days and then declined for most rats on the fourth
FEEDING CONDITIONS AND STRESS ULCERS
221
FIG. 1. Representative stomachs of rats exposed to total food deprivation (TFD, top), 6 g of cellulose per day (CEL6, middle) or ad lib cellulose (CELAL). H and A designate hanging cage or activity wheel housing, respectively (some lesions in the nonglandular stomach are not visible in these photographs). Most lesions in the glandular stomach have tightly adhering blood clots.
and fifth day. A comparison of peak wheel revolutions (Day 3 or 4) revealed no significant differences among the three feeding conditions (Table 1). The number of contacts with the food jar in the hanging cage-housed groups also was not significantly different among groups. Both measures showed a shift toward increased activity into the light phase of the light-dark cycle on the last 2–3 days of the experiment. Terminal body weight. There was a significant main effect of feeding conditions on body weight [F (2,33) 5 3.9, p , 0.03]. The ad lib cellulose groups had higher terminal body weights than the other groups (p , 0.05). However, at sacrifice, the stomach and
intestines still contained undigested material, which most likely accounted for this difference. Experiment 2 Ulcerations. Figure 2 shows representative ulcers from the four groups. Ulcers for individual rats are shown in Fig. 3 (bottom) and means are presented in Table 2. There was a main effect of feeding on ulcerations [F(1,22) 5 6.99, p , 0.02] but the effect of housing was not significant. Post hoc comparisons indicated that ulcers in the hanging cage-housed rats receiving 6 g of chow and cellulose ad lib were larger than those fed 6 g of chow alone (p , 0.05).
222
YI AND STEPHAN TABLE 1 EFFECTS OF NONCALORIC FEEDING AND HOUSING
Hanging cage groups TFD CEL6 CELAL Activity wheel groups TFD CEL6 CELAL
n
Rats with Ulcers
Area of Ulcers
Initial b.wt. (g)
Final b.wt. (g)
6
5 5 (R) 6 5
2.43 (1.39)* 5.51 (2.25) 8.42 (5.71) 7.92 (4.28)
229.8 (1.38)
177.8 (2.68)
NA
227.6 (1.87) 232.2 (1.29)
173.7 (2.62) 184.3 (2.43)
NA NA
229.2 (5.70)
171.0 (6.77)
9304 (2436)
238.6 (2.46) 235.5 (3.10)
179.2 (4.92) 187.1 (4.22)
10487 (2840) 9659 (2893)
7 7 6 7 6
3 5 (R) 5 5
6.58 (5.14) 3.46 (1.05) 5.20 (2.10) 21.25 (7.41)
Peak Wheel Revolutions
* Values are means with SEM in parentheses. CELAL 5 ad lib cellulose; (R) 5 ulcers in rumen (all other ulcers were in the glandular stomach). Initial b.wt. on Day 1 of the feeding condition.
However, a similar trend in the wheel-housed groups was not significant. The number of days to reach terminal body weights at sacrifice was not significantly different among groups. Activity. Peak levels of wheel running (Table 2) were significantly higher for rats receiving cellulose in addition to 6g chow than for rats receiving chow alone (p , 0.02). As in Experiment 1, wheel running and food jar contacts became more diurnal near the end of the experiment. Terminal body weights. Since rats were sacrificed near prede-
termined body weights, there were no significant differences among the groups. DISCUSSION
Total food deprivation led to ulcerations in both nonglandular and glandular portions of the stomach in one-half of the rats, but one rat each had ulcers exclusively in one portion or another. Although it has been reported that total food deprivation causes
FIG. 2. Representative stomachs of rats exposed to restricted feeding of 6 g/day chow (RF6, top), or 6 g/day chow and ad lib cellulose (R 1 C, bottom). H and A designate hanging cage and activity wheel housing, respectively.
FEEDING CONDITIONS AND STRESS ULCERS
223
FIG. 3. Total ulcerated area of the glandular stomach for individual rats in different feeding conditions. TFD 5 total food deprivation; CEL-6 5 6 g of cellulose per day; CEL-AL 5 ad lib cellulose; RF-6 5 restricted feeding of 6 g of chow per day; R 1 C 5 6 g of chow and ad lib cellulose.
only nonglandular ulcers, a careful reading of earlier studies reveals that both types of ulcers were observed, but with a dominance of nonglandular ulcers (13,14). Another study reports that prefeeding attenuates cold stress lesions in the glandular stomach and that food deprivation of nonstressed rats produced extensive rumenal lesions but also some glandular lesions after 2 to 6 days (17). Both the duration of food deprivation and the terminal body weight of the rats could influence whether or not glandular lesions occur. The present results confirm the prediction that food deprivation, like feeding of small amounts of chow, can lead to ulcers
in the mucosa of the rat stomach even in the absence of activity wheels (18 –20). In addition, the results show that small amounts of cellulose or ad lib cellulose protect the rumen from ulceration. On the other hand, ad lib cellulose produced a trend toward larger mucosal ulcers in rats housed in activity wheels. Restricted feeding of 6 g/day laboratory chow led to glandular (but not rumenal) ulcerations in both cage- and wheel-housed rats, consistent with our previous findings that a reduction in chow intake to 6 g/day is sufficient to cause glandular ulcers (18 –20). However, nearly twice as many days were required to induce
224
YI AND STEPHAN TABLE 2 EFFECTS OF RESTRICTED CALORIC FEEDING AND HOUSING
Hanging cage groups RF6 R 1 CEL Activity wheel groups RF6 R 1 CEL
n
Days to Sacrifice
Rats with Ulcers
Area of Ulcers
Initial Bodyweight (g)
Final Bodyweight (g)
7 6
9.6 (0.61)* 11.3 (1.25)
5 5
1.10 (0.50) 9.24 (3.06)
234.9 (2.56) 235.5 (1.87)
164.7 (2.09) 165.2 (2.62)
7 6
9.4 (1.11) 9.7 (1.20)
5 6
3.02 (1.96) 13.38 (6.55)
230.9 (2.54) 236.5 (3.73)
164.7 (2.09) 168.7 (2.82)
Peak Wheel Revolutions
NA NA 5586 (1350) 11440 (1833)
* Values are means with SEM in parentheses. R 1 CEL 5 6 g of chow 1 ad lib cellulose. Initial b.wt. is on Day 1 of the feeding condition.
ulcers compared to the noncaloric feeding groups in Experiment 1. This shows that a small amount of food significantly delays ulcer formation. As in Experiment 1, the addition of cellulose to 6 g of chow produced larger ulcerated areas. A protective effect against cold-restraint stress ulcers can be obtained by the administration of milk (3) or fish oil (1). Taken together, these findings indicate that ulcer severity may depend on the degree of caloric intake and that the location of ulcers depends mostly on the solidity of ingested food; i.e., solid food seems to protect the nonglandular stomach from ulceration in agreement with previous observations (9,10). However, the solidity of diets per se does not automatically protect the glandular stomach from damage; rather it can make the stomach more susceptible to ulceration, perhaps by stimulating gastric contractions and acid secretion. Observations in our laboratory show that glandular ulcers develop mostly along ridges (Figs. 1 and 2), which form a linear pattern in the glandular portion of the stomach and cross the arteries perpendicularly. It is possible that arteries or capillaries are susceptible to local ischemia due to hypermotility. Ischemia has been implicated in ulcerogenesis by a number of investigators (2,15). Our results, considered together with previous findings by others (5,6,9,10), suggest that ingestion of solid diets protects the rumen from ulceration. However, the ingestion of non-nutritive bulk seems to make the lower, glandular stomach more susceptible to ulceration. The mechanisms of these two different ulcerations have yet to be discovered.
The type and pattern of ulcers were very similar in all groups in agreement with our previous findings (18 –20). However, in the present study, access to activity wheels did not induce significantly larger ulcers. In part, this may be due to the small number of subjects and the large variability in daily wheel revolutions. In Experiment 2, peak activity was significantly higher for rats fed chow and cellulose, but compared to chow-fed rats the area of ulcers fell short of statistical significance. A second factor was that the terminal b.wt. of cage- and wheel-housed rats were nearly identical; this measure is a good predictor of ulcer area. The variability in ulcer areas within groups suggests that these ulcers develop very quickly, rapidly enlarge over 1 or 2 days and would quickly lead to mortality. Overall, these results strongly suggest that activity-stress ulcers share a common cause with ulcers induced by restricted feeding in hanging cage-housed rats; i.e., decreased food intake and subsequent depletion of body fat reserves may trigger a common ulcerogenic process. The increase in activity observed in wheelhoused rats may be a manifestation of metabolic stress (impending starvation) but is not necessary for ulcer formation. The origin and nature of the hypothesized stress signal remain to be determined but are highly correlated with b.wt. and fat reserves (19). ACKNOWLEDGEMENTS
Work was funded by the FSU Foundation. The authors thank Dr. Karen Glendenning for the use of an image processing system for ulcer assessment and Mr. Charles Badland for photography.
REFERENCES 1. Al-Harbi, M. M.; Islam, M. W.; Al-Shabanah, O. A.; Al-Gharably, N. M. Effect of acute administration of fish oil (omega-3 triglyceride) on gastric ulceration and secretion induced by various ulcerogenic and necrotizing agents in rats. Food Chem. Toxicol. 33:553–558; 1995. 2. Caldwell, M. T. P.; Suart, R. C.; Byrne, P. J.; Gorey, P. F.; Hennessy, T. P. J. Microvascular changes in experimental gastric stress ulceration: The influence of allopurinol, cimetidine, and misoprostol. J. Surg. Res. 55:135–139; 1993. 3. Dial, E. J.; Romero, J. J.; Lichtenberger, L. M. Gastroprotection by dairy foods against stress-induced ulcerogenesis in rats. Dig. Dis. Sci. 40:2295–2299; 1995. 4. Doughty, D. B.; Jackson, D. B. Gastrointestinal disorders. St. Louis: Mosby; 1993:86 – 88. 5. Driscoll, P.; Martin, J. R.; Kugler, P.; Baettig, K. Environmental and genetic effects on food-deprivation induced stomach lesions in male rats. Physiol. Behav. 31:225–228; 1983. 6. Glavin, G. B.; Mikhail, A. A. Stress and ulcer etiology in the rat. Physiol. Behav. 16:135–139; 1976.
7. Lambert, K. G. The activity-stress paradigm: Possible mechanisms and applications. J. Gen. Psychol. 120:21–32; 1993. 8. Manning, F. J.; Wall, H. G.; Montgomery, C. A. Jr.; Simmons, C. J.; Sessons, G. R. Microscopic examination of the activity-stress ulcer in the rat. Physiol. Behav. 21:269 –274; 1978. 9. Mikhail, A. A.; Gabriel, R. M.; Glavin, G. B. Ulcer reduction by non-nutritive bulk in pylorus ligated rats. Physiol. Behav. 19:437– 439; 1977 10. Mikhail, R. R.; Hirchberg, J. Ulceration in the rat’s forestomach: Its reduction by non-nutritive bulky substances. Physiol. Behav. 8:769 – 770; 1972. 11. Pare, W. P. The influence of food consumption and running activity on the activity-stress ulcer in the rat. Dig. Dis. 20:262–273; 1975. 12. Pare, W. P.; Glavin, G. B. Restraint stress in biomedical research: A review. Physiol. Behav. 10:339 –370; 1986. 13. Pare, W. P.; Temple, L. J. Food deprivation, shock stress and stomach lesions in the rat. Physiol. Behav. 11:371–375; 1973. 14. Pfeiffer, C. J.; Debro, F. R.; Muller, P. J. Gastric pathologic and
FEEDING CONDITIONS AND STRESS ULCERS biochemical changes induced by starvation of weanling rats. Life Sci. 5:509 –519; 1966. 15. Piasecki, C. K.; Thrasivoulou, C. Spasm of gastric muscularis mucosae might play a key role in causing focal mucosal ischemia and ulceration. Dig. Dis. Sci. 38:1183–1189; 1993. 16. Pintar, K.; Schulte, W. J.; Barboriak, J. J. Histologic study of gastric lesions in food-restricted rats. Proc. Soc. Exp. Biol. Med. 145:1353– 1358; 1974. 17. Wozniak, D. F.; Goldstein, R. Effects of deprivation duration and
225 prefeeding on gastric stress erosions in the rat. Physiol. Behav. 24: 231–236; 1980. 18. Yi, I.; Bays, M. E.; Stephan, F. K. Stress ulcers in rats: The role of food intake, body weight, and time of day. Physiol. Behav. 54:375–381; 1993. 19. Yi, I.; Stephan, F. K. Body fat reserves attenuate gastric ulcers induced by restricted feeding in rats. Physiol. Behav. 59:931–936; 1996. 20. Yi, I.; Stephan, F. K.; Bays, M. E. Activity-stress ulcers in rats: The role of pre-entrainment to meal time. Physiol. Behav. 58:67–73; 1995.