Physiology&Behavior,Vol. 50, pp. 945-949. ©Pergamon Press plc, 1991. Printed in the U.S.A.
0031-9384/91 $3.00 + .00
Activity-Based Anorexia: Relationship to Gender and Activity-Stress Ulcers L E E E. D O E R R I E S , * ~ : E R I C Z. S T A N L E Y * ( : A N D P A U L F. A R A V I C H t ~ : 1
*Department of Psychology, Christopher Newport College, Newport News, VA 23606 y'Department of Anatomy and Neurobiology, Eastern Virginia Medical School, Norfolk, VA 23501 4:Medical Research Service, Veterans Affairs Medical Center, Hampton, VA 23667 R e c e i v e d 1 April 1991 DOERRIES, L. E., E. Z. STANLEY AND P. F. ARAVICH. Activity-basedanorexia: Relationship to genderand activity-stress ulcers. PHYSIOL BEHAV 50(5) 945-949, 1991.--The phenomena of activity-based anorexia (ABA) and the activity-stress ulcer (ASU) both involve paradigms in which the food intake of rats is restricted and the animals are allowed voluntary exercise in a running wheel. This study determined the susceptibility of male and female rats to ABA and ASU following 25- and 30-percent losses of their original body weights. Male rats reached both weight loss criteria in a fewer number of days than did female rats. None of the male and female rats sacrificed at the 25% weight loss criterion evidenced gastric lesions; 52% of the animals sacrificed at the 30% weight loss criterion had one or more lesions. No gender differences with respect to gastric lesions were observed at the 30% weight loss criterion; however, at both weight loss criteria, females ate and ran more than males. It is concluded that ASUs are a consequence rather than a cause of ABA, and that there is a sexually "dimorphic" susceptibility to ABA but not ASUs. The utility of using a 25% weight loss criterion for defining ABA is also discussed. Activity-based anorexia Activity-stress ulcer Rats Male Sex differences Exercise Starvation Anorexia nervosa
Female
Sexual dimorphism
pendent studies suggest that males are more susceptible than are females to ABA. The contribution of gender factors to ABA, however, has not been systematically investigated. The purpose of this investigation was to determine the relationship between ASUs and ABA, and to determine if there is a sexually "dimorphic" sensitivity to either phenomenon. The susceptibility of male and female rats to ABA and ASUs following a 25-percent weight loss criterion and following a 30percent weight loss criterion was determined. If ASUs are a consequence rather than a cause of ABA, ASUs should be present in the 30-percent weight loss groups but not in the 25percent weight loss groups. If there is a sexually "dimorphic" sensitivity to either phenomenon, their manifestations should be associated with gender differences.
IT is well established that rats simultaneously exposed to a reslricted feeding schedule and allowed voluntary exercise in activity wheels are "anorexic" (i.e., eat less) compared to rats simply placed on restricted feeding schedules (3, 14, 16, 17). The phenomenon has been recently termed activity-based anorexia (ABA) (5) and embodies two characteristic features of anorexia nervosa: exercise and restricted feeding (2, 20, 21). It is also well established that simultaneous exposure to a reslricted feeding schedule and voluntary wheel-running exercise results in the development of severe gastric lesions in rats (6, 10-12, 18, 19). This phenomenon has been referred to as the activity-stress ulcer (ASU) (10). The relationship between ABA and ASUs has not been determined, although it has been proposed that ASUs are a consequence rather than a cause of the severe weight loss (12). Support for this contention is provided by data indicating that nearly all rats sustaining a greater than 30-percent weight loss develop extensive gastric pathology (6), and by preliminary data from our lab suggesting that no lesions are present in ABA animals following a 25-percent weight loss (4). Further research is clearly needed to confirrn these observations. The contribution of gender factors to ABA and ASUs is also unclear. For instance, two studies have reported that the number and severity of ASUs are not affected by gender (8,13). Two other studies report that male rats are more susceptible to ASUs and other forms of experimentally induced gastric ulcers than are female rats (7,9). Our own investigations of ABA have used either male rats (1) or female rats (3). The results of these inde-
METHOD
Subjects and Apparatus The subjects included 24 male and 24 female Holtzman Sprague-Dawley rats (Madison, WI). The rats were housed in a colony maintained at 22 (---2) degrees Centigrade with a 12/12hour light/dark cycle and fed Purina Rat Chow number 5001 (Ralston Purina Co., St. Louis, MO) as described below. Water was freely available throughout the experiment. The male rats were 41 days old, and the female rats were 42 days old at the start of the experiment, which followed a three-
tRequests for reprints should be addressed to Paul F. Aravich, Medical Research Service, Veterans Affairs Medical Center, Hampton, VA 23667.
945
946
DOERRIES, STANLEY AND ARAVICH
day habituation period to the colony for the males and a fourday habituation period for the females. The male rats were slightly younger than the female rats to better match the starting weights of the two genders. Of the 24 female rats that began the experiment, four animals, two in the 25-percent weight loss group and two in the 30-percent weight loss group, failed to lose the 25 and 30-+ 1.5 percent of original body weight (criteria used to define the ABA and ASU paradigms, respectively). Of the 24 male rats, one male failed to reach the criterion in each of the respective weight loss groups. Data analyses are based on 10 female and 11 male rats in each of the weight loss conditions. The male rats in the 25- and 30-percent weight loss groups weighed 187.9-+ 3.2 and 187.5 -+ 3.2 grams, respectively. Female rats in the same weight loss groups weighed 181.1 -+ 1.5 and 182.1 -+ 1.4 grams, respectively. The male rats weighed slightly but significantly (p0.05). Similarly, an analysis of the median pathology ratings for male and female rats at this weight loss criterion (Mann-Whitney U-test) failed to indicate a significant sex difference (U = 49.5, p>0.05). Finally, a t-test for independent samples performed on the mean cumulative lesion lengths for male and female rats in the 30-percent groups also revealed no significant gender difference, t(9j = 1.4, p>0.05. In order to determine whether activity or food intake levels immediately preceding sacrifice influenced the development of gastric pathology, separate 2-way ANOVAs were performed on mean total wheel turns and mean total food intake during the 24-hour period prior to sacrifice for male and female rats that developed lesions versus those that did not. Neither activity level nor food intake was indicative of which animals developed lesions. In contrast to ASUs, a sexually "dimorphic" susceptibility to ABA was observed. Figure 2 shows the mean number of days required by male and female rats to reach the 25- and 30-percent weight loss criteria. To lose 25 percent of their original body weight (ABA), female rats required 2.0--0.6 days longer than males. To lose 30 percent of their original body weight (ASU), female rats took 2.6-+0.6 days longer than male rats. A 2-way ANOVA performed on these data indicated a significant gender effect, F(1,3g)= 15.5, p0.05). Hence, the sexually "dimorphic" susceptibility to ABA was not dependent upon the degree of sustained weight loss. Figure 3 summarizes the mean total wheel turns for male and female rats in both the 25- and 30-percent weight toss conditions on their final day in the experiment. At the 25-percent weight loss criteria, female rats ran twice as much as male rats during the 24-hour period prior to sacrifice. At the 30-percent weight loss criteria, female rats ran one third more wheel revolutions than male rats. A 2-way ANOVA performed on these data indicated the gender difference to be reliable, F(1,38)= 11.21, pO.05). Figure 4 summarizes the mean total food intake for male and female rats in both the 25- and 30-percent weight loss conditions 24 hours prior to sacrifice. At both weight loss criteria, female rats ate more than the male rats. A 2-way ANOVA revealed this gender difference to be reliable, F(1,38)=11.86, pO.05) interaction effect. DISCUSSION This investigation demonstrates that activity-stress ulcers are a consequence rather than a cause of activity-based anorexia, and
ACTIVITY-BASED ANOREXIA AND GENDER
947
FIG. 1. Representative stomachs taken from male (A) and female (B) rats in the 25-percent weight loss groups. Notice the absence of gross gastric lesions. By contrast, there was a marked incidence of gastric pathology in the 30-percent weight loss groups. The gross pathology seen in all lesioned females and 3 of 5 male rats was moderate (C), as indicated by the arrows, or severe (D). Scale bar: 1 cm.
that there is a sexually "dimorphic" susceptibility to ABA but not to ASUs. The failure to find evidence of gastric lesions in rats reduced by 25 percent of their original body weight clearly indicates that gastric pathology does not cause the weight loss associated with activity-based anorexia. Our data showing that approximately 50 percent of the rats in the 30-percent weight loss condition had one or more gastric lesions suggests that the gastric pathology is more likely a consequence rather than a cause of the severe weight loss. This view is supported by other data (12) and the fact that nearly all rats reduced by 32 to 35 percent of their original body weight evidence extensive gastric pathology (6,10). When our findings are compared with data from previous studies of ASU, it appears that the weight loss threshold required for the development of gastric pathology is
TABLE 1 PERCENTAGEWITH LESIONS,MEDIANLESIONRATING,AND MEAN ( -+SE) CUMULATIVELESIONLENGTH(ram) OF MALEAND FEMALE SUBJECTSIN THE 30-PERCENTWEIGHTLOSS GROUPS
1o .c "z . -~ u° g L ~ • o ~
~
Male
8
\\\
\\\
6 \\\
,~
\\\ \\\ \\\ \\\ \\\ \\\
2
25Z Gender
N
Perccntage With Lesions
Median Lesion Rating
Mean Cumulative Lesion Length (ram)
Male
II
45
3.0
1.8 ---0.9
Female
16"
60
3.0
0.6 -0.1
~ Female
Weight-Loss
30Z Criterion
FIG. 2. Mean ( _ SE) days to reach the 25- or 30-percent weight loss criteria as a function of gender. (*p
0031-9384/91 $3.00 + .00
Activity-Based Anorexia: Relationship to Gender and Activity-Stress Ulcers L E E E. D O E R R I E S , * ~ : E R I C Z. S T A N L E Y * ( : A N D P A U L F. A R A V I C H t ~ : 1
*Department of Psychology, Christopher Newport College, Newport News, VA 23606 y'Department of Anatomy and Neurobiology, Eastern Virginia Medical School, Norfolk, VA 23501 4:Medical Research Service, Veterans Affairs Medical Center, Hampton, VA 23667 R e c e i v e d 1 April 1991 DOERRIES, L. E., E. Z. STANLEY AND P. F. ARAVICH. Activity-basedanorexia: Relationship to genderand activity-stress ulcers. PHYSIOL BEHAV 50(5) 945-949, 1991.--The phenomena of activity-based anorexia (ABA) and the activity-stress ulcer (ASU) both involve paradigms in which the food intake of rats is restricted and the animals are allowed voluntary exercise in a running wheel. This study determined the susceptibility of male and female rats to ABA and ASU following 25- and 30-percent losses of their original body weights. Male rats reached both weight loss criteria in a fewer number of days than did female rats. None of the male and female rats sacrificed at the 25% weight loss criterion evidenced gastric lesions; 52% of the animals sacrificed at the 30% weight loss criterion had one or more lesions. No gender differences with respect to gastric lesions were observed at the 30% weight loss criterion; however, at both weight loss criteria, females ate and ran more than males. It is concluded that ASUs are a consequence rather than a cause of ABA, and that there is a sexually "dimorphic" susceptibility to ABA but not ASUs. The utility of using a 25% weight loss criterion for defining ABA is also discussed. Activity-based anorexia Activity-stress ulcer Rats Male Sex differences Exercise Starvation Anorexia nervosa
Female
Sexual dimorphism
pendent studies suggest that males are more susceptible than are females to ABA. The contribution of gender factors to ABA, however, has not been systematically investigated. The purpose of this investigation was to determine the relationship between ASUs and ABA, and to determine if there is a sexually "dimorphic" sensitivity to either phenomenon. The susceptibility of male and female rats to ABA and ASUs following a 25-percent weight loss criterion and following a 30percent weight loss criterion was determined. If ASUs are a consequence rather than a cause of ABA, ASUs should be present in the 30-percent weight loss groups but not in the 25percent weight loss groups. If there is a sexually "dimorphic" sensitivity to either phenomenon, their manifestations should be associated with gender differences.
IT is well established that rats simultaneously exposed to a reslricted feeding schedule and allowed voluntary exercise in activity wheels are "anorexic" (i.e., eat less) compared to rats simply placed on restricted feeding schedules (3, 14, 16, 17). The phenomenon has been recently termed activity-based anorexia (ABA) (5) and embodies two characteristic features of anorexia nervosa: exercise and restricted feeding (2, 20, 21). It is also well established that simultaneous exposure to a reslricted feeding schedule and voluntary wheel-running exercise results in the development of severe gastric lesions in rats (6, 10-12, 18, 19). This phenomenon has been referred to as the activity-stress ulcer (ASU) (10). The relationship between ABA and ASUs has not been determined, although it has been proposed that ASUs are a consequence rather than a cause of the severe weight loss (12). Support for this contention is provided by data indicating that nearly all rats sustaining a greater than 30-percent weight loss develop extensive gastric pathology (6), and by preliminary data from our lab suggesting that no lesions are present in ABA animals following a 25-percent weight loss (4). Further research is clearly needed to confirrn these observations. The contribution of gender factors to ABA and ASUs is also unclear. For instance, two studies have reported that the number and severity of ASUs are not affected by gender (8,13). Two other studies report that male rats are more susceptible to ASUs and other forms of experimentally induced gastric ulcers than are female rats (7,9). Our own investigations of ABA have used either male rats (1) or female rats (3). The results of these inde-
METHOD
Subjects and Apparatus The subjects included 24 male and 24 female Holtzman Sprague-Dawley rats (Madison, WI). The rats were housed in a colony maintained at 22 (---2) degrees Centigrade with a 12/12hour light/dark cycle and fed Purina Rat Chow number 5001 (Ralston Purina Co., St. Louis, MO) as described below. Water was freely available throughout the experiment. The male rats were 41 days old, and the female rats were 42 days old at the start of the experiment, which followed a three-
tRequests for reprints should be addressed to Paul F. Aravich, Medical Research Service, Veterans Affairs Medical Center, Hampton, VA 23667.
945
946
DOERRIES, STANLEY AND ARAVICH
day habituation period to the colony for the males and a fourday habituation period for the females. The male rats were slightly younger than the female rats to better match the starting weights of the two genders. Of the 24 female rats that began the experiment, four animals, two in the 25-percent weight loss group and two in the 30-percent weight loss group, failed to lose the 25 and 30-+ 1.5 percent of original body weight (criteria used to define the ABA and ASU paradigms, respectively). Of the 24 male rats, one male failed to reach the criterion in each of the respective weight loss groups. Data analyses are based on 10 female and 11 male rats in each of the weight loss conditions. The male rats in the 25- and 30-percent weight loss groups weighed 187.9-+ 3.2 and 187.5 -+ 3.2 grams, respectively. Female rats in the same weight loss groups weighed 181.1 -+ 1.5 and 182.1 -+ 1.4 grams, respectively. The male rats weighed slightly but significantly (p0.05). Similarly, an analysis of the median pathology ratings for male and female rats at this weight loss criterion (Mann-Whitney U-test) failed to indicate a significant sex difference (U = 49.5, p>0.05). Finally, a t-test for independent samples performed on the mean cumulative lesion lengths for male and female rats in the 30-percent groups also revealed no significant gender difference, t(9j = 1.4, p>0.05. In order to determine whether activity or food intake levels immediately preceding sacrifice influenced the development of gastric pathology, separate 2-way ANOVAs were performed on mean total wheel turns and mean total food intake during the 24-hour period prior to sacrifice for male and female rats that developed lesions versus those that did not. Neither activity level nor food intake was indicative of which animals developed lesions. In contrast to ASUs, a sexually "dimorphic" susceptibility to ABA was observed. Figure 2 shows the mean number of days required by male and female rats to reach the 25- and 30-percent weight loss criteria. To lose 25 percent of their original body weight (ABA), female rats required 2.0--0.6 days longer than males. To lose 30 percent of their original body weight (ASU), female rats took 2.6-+0.6 days longer than male rats. A 2-way ANOVA performed on these data indicated a significant gender effect, F(1,3g)= 15.5, p0.05). Hence, the sexually "dimorphic" susceptibility to ABA was not dependent upon the degree of sustained weight loss. Figure 3 summarizes the mean total wheel turns for male and female rats in both the 25- and 30-percent weight toss conditions on their final day in the experiment. At the 25-percent weight loss criteria, female rats ran twice as much as male rats during the 24-hour period prior to sacrifice. At the 30-percent weight loss criteria, female rats ran one third more wheel revolutions than male rats. A 2-way ANOVA performed on these data indicated the gender difference to be reliable, F(1,38)= 11.21, pO.05). Figure 4 summarizes the mean total food intake for male and female rats in both the 25- and 30-percent weight loss conditions 24 hours prior to sacrifice. At both weight loss criteria, female rats ate more than the male rats. A 2-way ANOVA revealed this gender difference to be reliable, F(1,38)=11.86, pO.05) interaction effect. DISCUSSION This investigation demonstrates that activity-stress ulcers are a consequence rather than a cause of activity-based anorexia, and
ACTIVITY-BASED ANOREXIA AND GENDER
947
FIG. 1. Representative stomachs taken from male (A) and female (B) rats in the 25-percent weight loss groups. Notice the absence of gross gastric lesions. By contrast, there was a marked incidence of gastric pathology in the 30-percent weight loss groups. The gross pathology seen in all lesioned females and 3 of 5 male rats was moderate (C), as indicated by the arrows, or severe (D). Scale bar: 1 cm.
that there is a sexually "dimorphic" susceptibility to ABA but not to ASUs. The failure to find evidence of gastric lesions in rats reduced by 25 percent of their original body weight clearly indicates that gastric pathology does not cause the weight loss associated with activity-based anorexia. Our data showing that approximately 50 percent of the rats in the 30-percent weight loss condition had one or more gastric lesions suggests that the gastric pathology is more likely a consequence rather than a cause of the severe weight loss. This view is supported by other data (12) and the fact that nearly all rats reduced by 32 to 35 percent of their original body weight evidence extensive gastric pathology (6,10). When our findings are compared with data from previous studies of ASU, it appears that the weight loss threshold required for the development of gastric pathology is
TABLE 1 PERCENTAGEWITH LESIONS,MEDIANLESIONRATING,AND MEAN ( -+SE) CUMULATIVELESIONLENGTH(ram) OF MALEAND FEMALE SUBJECTSIN THE 30-PERCENTWEIGHTLOSS GROUPS
1o .c "z . -~ u° g L ~ • o ~
~
Male
8
\\\
\\\
6 \\\
,~
\\\ \\\ \\\ \\\ \\\ \\\
2
25Z Gender
N
Perccntage With Lesions
Median Lesion Rating
Mean Cumulative Lesion Length (ram)
Male
II
45
3.0
1.8 ---0.9
Female
16"
60
3.0
0.6 -0.1
~ Female
Weight-Loss
30Z Criterion
FIG. 2. Mean ( _ SE) days to reach the 25- or 30-percent weight loss criteria as a function of gender. (*p
