Ethanol reinforcement and its relationship to saccharin preference in Wistar rats

Alcohol,Vol. I1, No. 2, pp. 141-145,1994 Copyright©1994ElsevierScienceLtd Printedin the USA.All fightsreserved 0741-8329/94$6.00 + .00

Pergamon 0741-8329(93)E0001-F

Ethanol Reinforcement and Its Relationship to Saccharin Preference in Wistar Rats S. M I C H A E L B E L L , * B L A K E A. G O S N E L L , .1 D E A N D. K R A H N * A N D R I C H A R D A. M E I S C H t

*Department o f Psychiatry, University o f Wisconsin-Madison, Parkway Hospital, 6001 Research Park Blvd., Madison, WI 53719 "fDepartment o f Psychiatry and Behavioral Sciences, University o f Texas Health Sciences Center-Houston, MSI, SARC, 1300 Moursund Street, Houston, T X 77030 Received 23 M a r c h 1993; Accepted 9 September 1993 BELL, S. M., B. A. GOSNELL, D. D. KRAHN AND R. A. MEISCH. Ethanolreinforcement and its relationship to saccharinpreference in Wistarrats. ALCOHOL 11(2) 141-145, 1994.- Forty rats were given a choice between 0.1070sodium saccharin and water. Based on their intakes, three groups of six rats representing high, intermediate, and low saccharin preferences were selected. These rats were reduced to 80070 of their free-feeding weights. Ethanol was established as a reinforcer by use of a food-induced drinking procedure. Between-group differences were assessed based on response rates across acquisition sessions (0, 1, 2, 4, 5.7, 8070,w/v), a fluted-ratioseries (1, 2, 4, 8, 1), and a concentration series (8, 5.7, 4, 2, 2, 4, 5.7, 8, 11.3, 16, 22.6, 32, 8°70,w/v). In 29 of 32 conditions which were analyzed, the mean number of responses for ethanol was higher for the high saccharin preference group than for the low, and in 25 of 32 conditions, the intermediate group fell between the high and the low. However, there was considerable variability within groups across all conditions, such that mean between-group differences were not significant. This variability may be reduced by considering diet preferences in addition to saccharin preference. Nonetheless, these results offer limited support for the increasing body of evidence indicating a relationship between the factors mediating ethanol self-administration and those involving ingestion of palatable foods and fluids. Ethanol Alcohol Concentration-response

Self-administration Taste preference

Wistar rats

IN nonhuman subjects, a relationship between factors involved in eating behaviors and the self-administration of drugs has been identified. For example, dietary tryptophan supplements attenuate amphetamine (29) and cocaine (5) selfadministration in the rat. Marks-Kaufman and Lipeles (23) showed that rats with high preferences for fat during predrug baselines ultimately became morphine drinkers, while those animals who were low fat eaters during the baseline period did not consume the drug. In a separate study, fat-preferring rats consumed more ethanol than did carbohydrate-preferring rats (20). Finally, it is well documented that food-deprivation increases the self-administration of numerous drugs [see (6)]. Sweet tasting substances in particular tend to play an important role in the drug self-administration/ingesta relationship. Carroll and Boe (4) found that when rats were deprived of a glucose-saccharin solution they increased self-adminis-

Saccharin preference

Acquisition

Fixed ratio

tration of etonitazene. Lester and Greenberg (22) reported that when sucrose was provided to rats that drank ethanol, ethanol drinking markedly decreased. However, when the sucrose was subsequently removed, the ethanol drinking returned to even higher levels. Mice (C57BL/6J) characterized as morphine preferrers drank more sweetened solutions than mice (DBAJ2J) characterized as morphine avoiders (10). Also, there appears to be a close relationship between the factors influencing alcohol and saccharin intake in A A / A N A and P / N P rats (28). With a 1-h free-access paradigm, Gosnell and Krahn (13) demonstrated that rats with high saccharin preferences tended to drink more ethanol than rats with low saccharin preferences. In this study, a large group of male rats showing either high, intermediate, or low saccharin taste preferences was allowed to self-administer ethanol in operant chambers. Data

This research was conducted while S.M.B. was at the University of Texas Health Sciences Center-Houston and B.A.G. and D.D.K. were at the University of Michigan. To whom requests for reprints should be addressed. 141

142

BELL ET AL.

were systematically recorded concerning differences in saccharin taste preference and ethanol self-administration, including acquisition, fixed-ratio performance, and concentration-response relationships. METHODS

Saccharin Taste Preferences and Group Selection Forty male Wistar rats (Charles River Laboratories, Wilmington, MA) were housed individually in stainless steel cages with lights on from 0700 to 1900 and food (Purina rat chow) and water ad lib. The rats (when 80 days old) were waterdeprived for 23 h prior to each of the first two test sessions, transferred to separate cages, and provided access to 0.1% (w/v) sodium saccharin solution from a 25-ml buret for 1 h. Water deprivation induced all rats to drink saccharin solution during these first two test sessions, thus preventing confounding by neophobia in subsequent sessions. During the next 12 sessions, rats were provided two burets during the 1-h test sessions: one containing 0.1% saccharin solution, the other containing water. The positions of the burets were reversed prior to each session. Following test sessions rats were returned to their home cages where food and water were freely available. Based on the saccharin intakes averaged over the last four test sessions, 18 rats (i.e., three groups of 6 rats each) were chosen that represented groups of low, intermediate, and high saccharin drinkers. Mean (+_ SEM) intakes of these groups were 3.3 _+ 0.2, 7.7 _+ 0.3, and 12.0 _+ 1.0 ml. Water intakes were extremely low; average intake for all three groups was less than 0.2 ml. Therefore, water intake was not considered when forming the groups. As an additional test of the differences in saccharin preference among the three groups, the selected rats were given an additional session in which water and saccharin were available, and then six sessions in which saccharin and water were alternated as the only fluid in the session. The mean of the final two water trials was subtracted from the mean of the final two saccharin trials. With this procedure, the group means were 4.9 + 1.0, 7.0 _ 1.9, and 12.9 _+ 1.4 ml. Thus, the groups maintained their ordinal rankings when tested with water available concurrently with saccharin and when saccharin was alternated with water. All further analyses were based only on the initial determination of saccharin preference. The testing of initial saccharin preference and the selection of subjects was performed at the University of Michigan by two of the investigators (B.A.G. and D.D.K., now at the University of Wisconsin). Selected rats were shipped to the laboratory in Texas where the ethanol self-administration experiments were conducted (no information was supplied to the Texas laboratory staff regarding saccharin preferences, and housing conditions were similar in the two labs). Upon arrival at the Texas laboratory, the 18 rats were food-deprived and gradually reduced to 800/o of the free-feeding weights recorded at the completion of the saccharin preference tests [see (6)]. Daily allotments of food were varied to maintain the rats at these weights throughout the self-administration experiments. The ranges of weights for low, intermediate, and high saccharin preference groups were 292-365, 329-388, and 332351 g, respectively, following weight reduction. Formal testing in Texas began approximately 30 days after the final test day in Michigan. Unless stated otherwise, water was always freely available in the home cages.

Ethanol Self-Administration Apparatus. Each experimental chamber was equipped with a pair of levers, each located below a metal spout. Below each

lever was a row of stimulus lights, and above each spout was a second row of lights. The drinking device has been described in greater detail elsewhere (2). The experimental chambers were contained within a darkened room with masking noise provided from a single speaker. Experiments were controlled and data recorded by a PDP-11 computer utilizing SKED software. At the beginning of a session the lights below the active lever were illuminated. When a rat operated a lever, the lights below that lever were turned off and the lights above the spout were illuminated. While the upper lights were on, the rat was able to obtain liquid from the drinking system with licks of its tongue. Each tongue lick on the metal spout completed a drinkometer circuit and resulted in a brief activation (5 ms) of a solenoid-operated valve. Each opening of the valve delivered an average of 0.01 ml of liquid through the metal spout directly into the rat's mouth. The delivery component was programmed so that each lever press (fixed ratio 1 [FR 1]) resulted in access to 10 consecutive reinforced licks that cumulatively yielded an average of 0.I ml of liquid. When a series of 10 licks had been completed, the lights above the drinking spout were extinguished and the lights below the lever were again illuminated. When the spout lights were not illuminated, licks had no programmed consequences. When the spout lights were illuminated, lever presses also had no programmed consequences. Procedure. Prior to the beginning of the experiment, rats were water-deprived for 24 h and trained to press a lever to obtain deliveries of water. Initial training sessions were 3 h in length. Water was available only in the experimental chambers under an FR 1 reinforcement schedule. Daily food allotments were provided in the chambers during these training sessions. These conditions remained in effect for 18 consecutive sessions. Subsequently, continuous access to water in the home cages was restored and 60-min daily sessions were begun. An initial water baseline was established by providing water alternately from the left and right spouts until stability was obtained. A food-induced (postprandial) drinking procedure was then instituted in which 10 g of food were provided 30 min into a 90-min session (the remainder of the rats' daily allotments was given postsession in the home cages) under an FR 1 reinforcement schedule. Responses made during the first 30 min of the session were not food-induced and were attributed to the available liquid. An ethanol solution was presented alternately (left-right) from a single spout in increasing concentrations: 0, 1, 2, 4, 5.7, and 80/o w/v. Ethanol solutions were mixed approximately 20 h prior to the start of the session. After six sessions at each of the ethanol concentrations the within-session food amount was reduced at a rate of 1 g per session. Once food was completely faded from the sessions and six sessions were obtained in the absence of food, the session length was reduced to 60 min. A training period was then instituted in which the reinforcement schedule was manipulated: FR 1, 2, 4, 8, and 1. Following six sessions of training at each of these conditions testing sessions began. Water was made available concurrently with an 8 % ethanol solution and the FR manipulation series was repeated. Following six sessions at each of these FR conditions an ethanol concentration series was presented: 8, 5.7, 4, 2, 2, 4, 5.7, 8, 11.3, 16, 22.6, 32, and 80/o (w/v). Six consecutive sessions were conducted at each concentration with an FR 1 reinforcement schedule. The rats were then food-satiated and run in a final series of 24 sessions, with 8% ethanol and water available concurrently. The position of the concurrently available solu-

SACCHARIN PREFERENCE AND ETHANOL DRINKING

143

tions was alternated (left-right) on a daily basis throughout the study.

RESULTS Saccharin Intake After all self-administration procedures were completed, saccharin preferences were retested using methods identical to those used for initial group selection. The three groups maintained the same relative rankings as observed at the beginning of the experiment. However, intake levels in all three groups increased from the first test to the retest. Mean saccharin intakes for the retest were 7.2 ± 1.0, 12.1 +_ 2.7, and 14.9 ± 2.5 ml. Pre- and postethanol self-administration measures of saccharin preference were significantly correlated (Pearson r = 0.49; Spearman rho = 0.64, p < 0.05).

10080-

During the food-induced drinking acquisition procedure total session responses for all three groups were an inverted U-shaped function of ethanol concentration. However, responding during the initial 30-min period increased as concentration was increased from 0%0 to 4%0, but remained relatively constant as concentration was increased further to 8°70. Mean total session ethanol intake (g drug/kg body weight) was compared: Intake for the low saccharin group peaked at the 40/0 ethanol condition (1.37 g/kg), the intermediate group peaked at 5.7% (1.32 g/kg), and the high group peaked at 8070 (1.37 g/kg). During the first 30 rain of the acquisition sessions mean ethanol intake increased as concentration increased for all three groups. For the low, intermediate, and high groups, intakes when 8°70 ethanol was available were 0.50, 0.52, and 0.70 g/kg. When food was removed from the session, total session responses for all three groups decreased markedly, but responses during the initial 30-min period remained about the same. For low, intermediate, and high saccharin preference groups, one-tailed, paired t tests (n = 6, p < 0.05) revealed a significant increase in mean initial 30-min responding from when water was available (2.9 ± 0.8, 6.8 ± 2.0, and 7.6 ± 2.0 responses) to when 8% ethanol was available-this was true both when food was given in the session (20.3 ± 7.7, 22.6 ± 5.9, and 30.2 +_ 8.1) and after food was removed from the session (16.3 ± 4.4, 23.2 ± 6.3, and 25.2 + 5.7). For low, intermediate, and high groups, mean ethanol intakes following the removal of food from the sessions were 0.58, 0.56, and 0.65 g/kg. Ethanol served as a reinforcer for all rats, since responding for 8070 ethanol was higher than responding for water when the two were offered concurrently and when each was offered alone. When the response requirement for concurrently available 8% ethanol and water was increased, ethanol- but not waterreinforced responding increased for all groups (Fig. 1). However, the increased responding was not proportional to the increased FR size, such that fewer ethanol deliveries were obtained as the fixed ratio was increased. Water responses were approximately equal for all groups and were far less than ethanol responses at all tested fixed ratios. At FR 1, ethanol concentration was gradually decreased from 8% to 2%, then gradually increased from 2% to 32%, and finally retested at 807o. As indicated in Fig. 2, all groups varied their intake as a function of concentration, with maximumresponding occurring around 5.7-807o ethanol. At three different points during the ethanol concentration manipulation, 807o ethanol was tested. Response values at the first 8070

•

_/

I~n~ghrm ediate Low

j,~

¢n == 60"6 40200 40

>~ 30"6

Ethanol Self-Administration

:

20-

10-

0 1

2

4

8

1

Fixed-ratio size

FIG. 1. Mean number of responses or deliveries as a function of fixed-ratio size. Each point represents the group mean (n = 6) over the six sessionsat each condition; groups represented high, intermediate, and low saccharin taste preference. Closed symbols represent responses for 8070 (w/v) ethanol, and open symbols represent responses for concurrently available water.

retest are approximately the same as those for the first test, for all three groups. Following exposure to the high ethanol concentrations, however, the final retest showed much smaller between-group differences. This pattern was also seen in the response curve as a w h o l e - t h a t is, at the higher concentrations, the differences among the group means diminished. During acquisition, the fixed-ratio series of tests, and the ethanol concentration series, the high saccharin preference group consistently responded more than the low saccharin preference group (on 29 of 32 comparisons) and the intermediate saccharin preference group fell between these groups (on 25 of 32 comparisons). However, there was great within-group variability across conditions, and the groups did not significantly differ from one another when analyzed with analysis of variance. As a final test, all 18 rats were food-satiated and allowed to self-administer for a series of 24 dally sessions. For all rats, levels of responding for ethanol became negligible. Also, the low, intermediate, and high saccharin preference groups did not differ from each other in dally intakes; over the last six sessions of the series, mean intakes of 8070 ethanol were 0.7 + 0.2, 1.0 + 0.3, and 0.7 + 0.2 ml, and mean intakes of concurrently available water were 1.0 ± 0.2, 0.9 ± 0.3, and 0.9 ± 0.2 ml.

DS ICUSSO IN The present study investigated ethanol self-administration by rats that were selected on the basis of saccharin taste preference. Responding for ethanol was greater by the high saccharin preference group than by the low on 29 of 32 comparisons, and the intermediate fell between the two groups 25 of 32

144

BELL ET AL, 40~ • ¢

]"

High Intejmediate

i

o

S

20-

E=

Y

10-

8

5.7

4

2

2

4

5.7

8

11.3

16

22.6

32

Ethanol concentration (% w/v)

FIG. 2. Mean number of responses for ethanol as a function of concentration (% w/v). Each point represents the group mean (n = 6) over the six sessions at each condition; groups represented high, intermediate, and low saccharin taste preference. Closed symbols represent responses for ethanol, and open symbols represent responses for concurrently available water. Within-group variability was great for all groups at all conditions. SEMs are shown for high and low saccharin groups but, for visualization purposes, not shown for the intermediate group.

times. However, due to a high degree of within-group variability, statistical analysis of group means indicated no significant differences between the groups. These results are contrary to previous studies in which rats had free access to ethanol (13,25). Important differences may exist between tests under free-access conditions and those in operant chambers. It is possible that any differences in ethanol responding among the three groups were due solely to nonspecific 'differences in liquid intake. Sinclair et al. (28) showed that strains of rats selected for high ethanol consumption (AA) drank much more saccharin than the line developed for low ethanol intake (ANA). Furthermore, the A A strain drank more of bitter, salty, and sour solutions than did the ANA. Perhaps rats selected for high saccharin preference will drink any solution in higher quantities, or conversely, rats selected for low saccharin preference will drink any solution in lower quantities. However, this is not likely the sole explanation, since 1) in our study, concurrently available water values were equally low across groups in pre- and postethanol self-administration saccharin taste-preference trials; 2) during the conditions in which water was available concurrently with ethanol, the low saccharin preference group actually self-administered slightly more water than the high saccharin preference group on 12 of 17 conditions; 3) when the rats were food-satiated, there was no difference between ethanol and water responses, and responses for both solutions were equally low across groups; and 4) Sinclair et al. (28) found little difference between P and NP rats (another strain developed for high and low ethanol consumption) with regard to intake of salty, bitter, and sour solutions, but differences were found in preferences for saccharin. Individual differences in saeehaxin preference have previously been shown to be related to rat's initial selection of ethanol (18). In a two-bottle choice paradigm for measuring

ethanol preference, provision of a third choice of saccharin, sucrose, or fat solution resulted in a decrease in alcohol consumption. Conversely, when the third choice was subsequently removed, ethanol drinking returned to even higher levels (22). Intakes of ethanol have also been shown to be directly related to initial preferences for saccharin (13,25). The present results, though, offer at best only limited support to these findings. In addition to an apparent relationship between sweet taste and ethanol consumption, a relationship between macronutrient preference and ethanol consumption has been reported. Alcohol-preferring A A rats have higher fat intakes in diet self-selection situations than do alcohol-nonpreferring A N A rats (11). Furthermore, the provision of ethanol to A A rats caused a reduction in carbohydrate intake, and there was a significant correlation between the amount of ethanol consumed and the size of this reduction. In outbred SpragueDawley rats, we have found that rats with a high baseline preference for fat consumed more alcohol than those with a low fat preference (20). In a nonchoice paradigm, rats fed a high-fat diet drank more alcohol than rats fed a control diet (26). Thus, it seems that, as has been observed with rats with preferences for sweets, rats with preferences for high-fat diets prefer ethanol. Although to the best of our knowledge the direct relationship between sweet and fat preference has not been studied in rats, we suggest that perhaps the factors mediating taste and diet preferences are independent, though both related to ethanol intake. Since the rats in this study were selected only on the basis of saccharin preference, it is possible that some of the within-group variability could have been reduced if diet selection had been considered as well. Further studies will test this hypothesis. In clinical populations there is evidence of a relationship between diet and taste preference and intake of ethanol and other substances of abuse as well. Anorexic and bulimie pa-

SACCHARIN PREFERENCE AND ETHANOL DRINKING tients, who binge on high fat foods (7) and who have increased preferences for sweet tastants (8,9), also tend to demonstrate increased rates of alcoholism and other substance abuse (3,16,19). Conversely, w o m e n treated for alcohol problems may exhibit greater incidences o f eating disorders (17,19,21, 27). Recovering alcoholics who have the longest periods of postdetoxification sobriety report markedly higher intakes of sugar in beverages compared to those with shorter periods of postdetoxification sobriety (32). Sweet cravings are frequently reported by opiate addicts, and large intakes of sweet foods have been reported in this population (24,30). It is also interesting to note in literature published by Alcoholics Anonymous that recovering alcoholics claim that "many of u s even many who said they had never liked s w e e t s - h a v e found that eating or drinking something sweet allays the urge to drink" (1). In a survey of eating and drinking habits among adolescents, increased alcohol intake was associated with increased intake o f sugars, sweets, fats, and oils, and with decreased intakes of complex carbohydrates (31). U.S. adults and Finn-

145 ish men who are classified as high alcohol users eat more fat than those classified as low alcohol users (14,15), and data from the first National Health and Nutrition Examination Survey ( H A N E S I) indicate that "the most salient difference in nutrient intake between drinkers and nondrinkers was the substantially lower carbohydrate intake o f drinkers" [(14), p. 289]. There is an increasing body of evidence linking the mechanisms involved in diet selection and in abuse o f ethanol and other substances. Although the present results offer only limited support for this link, we suggest that further studies are warranted to investigate the variability, interactions, and commonalities o f these behaviors. ACKNOWLEDGEMENTS The research was supported by grant DA 06827 from the National Institute on Drug Abuse. R.A.M. is the recipient of Research Scientist Award DA 00159 from the National Institute on Drug Abuse. The authors wish to thank David Averbach, Lisa Haymes, and Sara Rizvi for their expert technical assistance.

REFERENCES 1. Alcoholics Anonymous. Living sober. New York: Alcoholics Anonymous World Services; 1987:23. 2. Beardsley, P. M.; Meisch, R. A. A precision drinking device for rats tested with water, etonitazene, and ethanol. Pharmacol. Biochem. Behav. 14:871-876; 1981. 3. Bulik, C. M. Drug and alcohol abuse by bulimic women and their families. Am. J. Psychiatry 144:1604-1606; 1987. 4. Carroll, M. E.; Boe, I. N. Increased intravenous drug selfadministration during deprivation of other reinforcers. Pharmacol. Biochem. Behav. 17:563-567; 1982. 5. Carroll, M. E.; Lac, S. T.; Asencio, M.; Kragh, R. Intravenous cocaine self-administration in rats is reduced by dietary L-tryptophan. Psychopharmacology 100:293-300; 1990. 6. Carroll, M. E.; Meisch, R. A. Increased drug-reinforced behavior due to food-deprivation. In: Thompson, T.; Dews, P. B.; Barrett, J. E., eds. Advances in behavioral pharmacology, vol. 4. New York: Academic Press; 1984:47-88. 7. Drewnowski, A. Physiological basis of taste preference in eating disorders. In: Blinder, B. J.; Chaitin, B. F.; Golstein, R., eds. The eating disorders: Medical and psychological basis of diagnosis and treatment. New York: PMA; 1988:121-127. 8. Drewnowski, A.; Bellisle, F.; Aimez, P.; Remy, B. Taste and bulimia. Physiol. Behav. 41:621-626; 1987. 9. Drewnowski, A.; Halmi, K. A.; Pierce, B.; Gibbs, J. Taste and eating disorders. Am. J. Clin. Nutr. 46:442-450; 1987. 10. Forgie, M. L.; Beyerstein, B. L.; Alexander, B. K. Contribution of taste factors and gender to opioid preference in C57BL and DBA mice. Psychopharmacology 95:237-244; 1988. 11. Forsander, O. A. The interaction between voluntary alcohol consumption and dietary choice. Alcohol Alcohol. 23:143-149; 1988. 12. Forsander, O. A.; Sinclair, J. D. Protein, carbohydrate, and ethanol consumption: Interactions in AA and ANA rats. Alcohol 5: 233-238; 1988. 13. Gosnell, B. A.; Krahn, D. D. The relationship between saccharin and alcohol intake in rats. Alcohol 9:203-206; 1992. 14. Gruchow, H. W,; Sobocinski, K. A.; Barboriak, J. J.; Scheller, J. G. Alcohol consumption, nutrient intake and relative body weight among US adults. Am. J. Clin. Nutr. 42:289-295; 1985. 15. Hasunen, K.; Pekkarinen, M.; Nuutinen, O. Alcohol consumption and dietary intake of Finnish men. Second European Nutrition Conference, Munich 1976. Nutr. Metabol. 21:132-133; 1977. 16. Jonas, J. M.; Gold, M. S.; Sweeney, D,; Pottash, A. L. C. Eating disorders and cocaine abuse: A survey of 259 cocaine abusers. J. Clin, Psychiatry 48:47-50; 1987. 17. Jones, D. A.; Cheshire, N.; Moorhouse, H. Anorexia nervosa,

18. 19. 20. 21. 22.

23. 24.

25.

26. 27. 28. 29.

30. 31. 32.

bulimia and alcoholism: Association of eating disorder and alcohol. J. Psychiatr. Res. 19:377-380; 1985. Kampov-Polevoy, A. B.; Kasheffskaya, O. P.; Sinclair, J. D. Initial acceptance of ethanol: Gustatory factors and patterns of alcohol drinking. Alcohol 7:83-85; 1989. Krahn, D. D. The relationship of eating disorders and substance abuse. J. Subst. Abuse 3:239-253; 1991. Krahn, D. D.; Gosnell, B. A. Fat-preferring rats consume more alcohol than carbohydrate-preferring rats. Alcohol 8:313-316; 1991. Lacey, J. H.; Moureli, E. Bulimic alcoholics: Some features of a clinical subgroup. Br. J. Addict. 81:389-393; 1986. Lester, D.; Greenberg, L. A. Nutrition and the etiology of alcoholism: The effects of sucrose, saccharin and fat on the selfselection of ethyl alcohol by rats. Q. J. Stud. Alcohol 13:533650; 1952. Marks-Kaufman, R.; Lipeles, B. J. Patterns of nutrient selection in rats orally self-administering morphine. Nutr. Bchav. 1:33-46; 1982. Morabia, A.; Fabre, J.; Chee, E.; Zeger, S.; Orsat, E.; Robert, A. Diet and opiate addiction: A quantitive assessment of the diet of noninstitutionalized opiate addicts. Br. J. Addict. 84:173-180; 1989. Overstreet, D. H.; Kampov-Polevoy, A. B.; Rezvani, A. H.; Murrelle, L.; Halikas, J. A.; Janowsky, D. S. Saccharin intake predicts ethanol intake in genetically heterogeneous rats as well as different rat strains. Alcohol. Clin. Exp. Res. 17:366-369; 1993. Pekkanen, L.; Erikson, K. Dietary-induced changes in voluntary ethanol consumption and ethanol metabolism in the rat. Br. J. Nutr. 40:103-113; 1978. Peveler, R.; Falrburn, C. Eating disorders in women who abuse alcohol. Br. J. Addict. 85:1633-1638; 1990. Sinclair, J. D.; Kampov-Polevoy, A.; Stewart, R.; Li, T.-K. Taste preferences in rat lines selected for low and high alcohol consumption. Alcohol 9:155-160; 1992. Smith, F. L.; Yu, D. S. L.; Smith, D. G.; Leccese, A. P.; Lyness, W. H. Dietary tryptophan supplements attenuate amphetamine self-administration in the rat. Pharmacol. Biochem. Behav. 25: 849-855; 1986. Weiss, G. Food fantasies of incarcerated drug users. Int. J. Addict. 17:905-912; 1982. Yamamoto, M. E.; Block, G. D.; Ishii, E. Food patterns among adolescents: Influence of alcohol consumption. Alcohol. Clin. Exp. Res. 15:359; 1991. Yung, L.; Gordis, E.; Holt, J. Dietary choices and likelihood of abstinence among alcoholic patients in an outpatient clinic. Drug Alcohol Depend. 12:355-362; 1983.