The effects of naloxone on flavor-calorie preference learning indicate involvement of opioid reward systems.
Because some aspect of metabolism is necessary for the flavor-calorie preference to develop, it is reasonable to seek out the postingestive events that make the learning possible. There is evidence that oxidation of glucose by the liver is necessary (Tordoff & Friedman, 1986; Tordoff, Tepper, & Friedman, 1987; Tordoff, Ulrich, & Sandler, 1990). Reflexive release of duodenal cholecystokinin (CCK) might serve as the signal for the learned preference (Mehiel, 1991, Mehiel & Bolles, 1988b), but data in support of that idea are equivocal. For example, Perez and Sclafani (1991) found evidence for flavor-CCK preferences, as did Mehiel (1991), but Mehiel (unpublished observations) has been unable to reliably replicate that finding. In contrast, Booth (1972; 1985) and his coworkers have published a number of reports showing that flavors paired with satiety are preferred.
Whatever the nature of the physiological processes that relate metabolic events and .psychological motive strength, the fact is that when the rat is offered a choice between a calorie-paired flavor and a noncalorie-paired flavor in a two-bottle test, it chooses to partake more of the calorie-paired flavor. Thus another question. relates to the form of the motive state when the preference is displayed. That is, does the rat have an expectancy about its intake of one flavor that is different from its expectancy about the intake of the other flavor? Fedorchak and Bolles (1988) argued that the rat learns a "nutritive expectancy" during conditioning trials. Simply stated, the rat learns that one flavor means food and the other flavor means no food.
Another way to explain flavor preference learning is to argue that the rat learns to like the flavor. Young (1961) made the point that while stimuli have informational value (e.g., nutritive expectancy), they also have affective value (e.g., liking). This idea, though elegant, has turned out to be very difficult to support with evidence. The taste reactivity test (Grill & Norgren, 1978) is one method that has been used successfully to argue for affective responses to tastants.
An area of investigation relating affective responses to taste has been the investigation of the brain opioid reward system. Central systems utilizing endogenous opioids as transmitters are argued to mediate the psychological experience of reward. In support of this idea, naloxone (an opioid antagonist) has been shown to reduce incentive value of sweetness in humans, without affecting levels of hunger (Fantino, Hosotte, & Apelbaum, 1986). By antagonizing the central "reward system" that utilizes brain opioids, naioxone presumably makes food taste less good. Similarly, it has been found that naloxone has no effect on the latency for mice-killing rats to kill their prey, but does greatly increase the latency to eat the mice (Walsh, White, & Albert, 1984). Naloxone also reliably shifts preference ratings for sweets such as saccharin and sucrose up the concentration gradient (Cooper, 1983; Le Magnen, Marfaing-Jallat, Miceli, & Devos, 1980; Lynch & Libby, 1983; Siviy & Reid, 1983). "Put plainly, the pleasure or palatability of consumed solid foods or liquids is thought to depend upon endogenous opioid activity." (Cooper & Kirkham, 1990, p. 240).
Because flavor preferences must on some level be a function of pleasure or palatability, it is possible that they are also related to endogenous opioid activity. That is, it ought to be possible to condition that activity to a flavor conditional stimulus (CS). It is possible that in flavor-calorie learning, a flavor CS that predicts the digestion of calories (US) results in a conditioned response (CR) that involves activation of the opioid reward system. The purpose of the present investigation was to gather evidence in support of that notion. Because naloxone reduces palatability of tastants, then if flavors paired with calories are increased in their palatability by conditioning, naloxone should reduce their palatability and thus reduce the preference. Furthermore, if endogenous opioid activity is involved in the learning of flavor-calorie preferences, then other effects related to endogenous opioid activity (such as reduced freezing to shock) should be present when the calorie-paired flavor is tasted.
A general method was utilized to condition flavor-calorie preferences throughout the experiments that are described below. Typically rats are given experience with a caloric US such as dextrose with a flavor CS such as orange. A second flavor is paired with a noncaloric US like water or saccharin. After 8 to 10 serial discrimination trials, preference for the flavor paired with the caloric solution is assessed by a two-bottle preference test or by several one-bottle tests. Intakes during tests are fairly low as we are usually testing flavored water without the sugar or saccharin, so spillage is collected and factored into the intake score for each rat. We weigh the bottle, stopper, and spill dish before and after a test and take the difference score as the dependent measure.
Because naloxone reduces the hedonic value of tastants, rats conditioned to prefer a flavor paired with dextrose over a flavor paired with saccharin should have the preference disrupted compared to the preference displayed when injected with saline. Experiment 1 conditioned a preference as above, then tested each rat after injections of saline and injections of naloxone, and compared the intakes within rats across tests. Preferences should be lowered by naloxone.
Participants. Eight male, 60-day-old Long Evans rats purchased from Charles River served as subjects. The rats were individually housed in hanging stainless steel cages in the laboratory vivarium with lights on at 0600 and off at 1800 hours daily. Throughout the experiment the rats were restricted to 15 g of Purina Rat Chow daily except as noted below. Water was available ad libitum except during conditioning and testing.
Flavors, solutions, and drugs. Calories were derived from 10% w/v dextrose (Fisher Scientific) in tap water that was mixed daily. Flavors were cherry and grape flavored Unsweetened Kool Aid (General Foods) at a concentration of .25% w/v as in previous studies from this laboratory. Naloxone hydrochloride (Sigma) was injected ip. at a dose of 4 mg/ml/kg dissolved in saline (Fisher), and the saline vehicle was injected one ml/kg ip.
Procedure. The 8 rats were conditioned to prefer one of the flavors over the other during 4 days of ad libitum access to 100 ml of the flavored dextrose solution and 4 days ad libitum access to flavored water. Solutions were alternated daily and order of solution presentation and flavor paired with dextrose was counterbalanced across rats. Thus each rat had 4 days experience with flavored dextrose and 4 days experience with flavored water. Each conditioning day the rats were fed their 15-g ration prior to placing the refreshed conditioning bottles on the cage fronts, and intakes from the previous day's trial were recorded. Plain water was not available during the conditioning phase. On the first day following conditioning, half the rats were injected with naloxone and half were injected with saline vehicle. Fifteen min after the injection, each rat was offered two bottles of flavored water, one bottle contained the dextrose-paired flavor, the other had the water-paired flavor, but no dextrose was in the test. The rats were hungry, they had last been fed 24 hr before. The test took place at the normal conditioning time, 1100, and lasted 2 hr. Following the test, the rats were fed 15 g of Purina Chow, and plain water was returned. The next day, the test was replicated except that the drug condition was reversed for each rat. This concluded the experiment.
Results and Discussion
Figure 1 shows the intake during conditioning and the two-bottle tests. Data from the flavor preference tests were subjected to repeated measures ANOVA with flavor and drug as within rat factors and revealed a reliable effect for flavor, F(1, 7) = 7.68, p [less than] .03, because the rats drank more of the calorie-paired flavor than of the water-paired flavor. The main effect for drug was not significant, F(1, 7) = 2.66, p [less than] .25, because it did not reduce intake of the water-paired flavor, t(7) = 1.0, n.s. However, the drug by flavor interaction was significant, F(1, 7 = 10.02, p [less than] .02, because naloxone only reduced intake of the dextrose-paired flavor, t(7) = -2.44, p [less than] .05.
Naloxone significantly reduced intake of a dextrose-paired flavor, but did not reduce intake of a water-paired flavor. However, intake of the water-paired flavor was so low, it is possible that a floor effect prevented any decreased intake. Additionally, dextrose presumably tastes better than water, so naloxone may have reduced intake purely on the basis of taste. It would be more powerful to show that the effect is replicable when the noncalorie conditioning solution is also highly palatable.
When rats are offered flavored sucrose, ethanol, starch, or vegetable oil; and the noncalorie solution is sweetened with saccharin, they still learn to prefer the calorie-paired flavor over the saccharin-paired flavor. Mehiel and Bolles (1988b) argued that unconditioned hedonic value of the calorie source had little to do with the preference learning, rather, the preference is accountable for by the caloric benefit of the solution. If that is true, then we should be able to find the same specific reduction of intake of the calorie-paired flavor after treatment with naloxone as was found in Experiment 1 even if the noncalorie solution is highly palatable. Alternatively, opioid antagonists do reduce intake of saccharin solutions (Cooper, 1983) and because saccharin carries no nutritive benefit, the effect is clearly a reduction in hedonic value. However, some dissociation between the solute and the flavor occurs in flavor-calorie conditioning. For example, rats prefer an ethanol-paired flavor over a saccharin-paired flavor, even though after conditioning they still prefer unflavored saccharin to unflavored ethanol (Mehiel, 1991; Mehiel & Bolles, 1984).
In Experiment 1 rats' intake of a preferred dextrose-paired flavor was suppressed by treatment with naloxone, but it was not clear whether or not floor effects constrained naloxone's effect on the noncalorie-paired flavor. Experiment 2 sought evidence that the reduction was specific to a flavor paired with calories, rather than a reduction of a flavor-flavor preference, and used longer one-bottle tests as a way of negating the floor effects. Experiment 2 was essentially a replication of Experiment 1 but ethanol was used as the calorie source, and the noncalorie solution was saccharin instead of water. In this way flavor-flavor conditioning worked against the hypothesis that a conditioned response (CR) of endogenous reward system activity stimulated by a flavor CS that had been paired with calories was being antagonized by the naloxone. Evidence that naloxone specifically reduces intake of an ethanol-paired flavor over a saccharin-paired flavor was sought.
Participants. Eight male Long Evans rats were 120 days old at the start of the experiment. Living conditions were as in Experiment 1.
Solutions and flavors. Flavors were Unsweetened Kool Aid (General Foods) .25% w/v in tap water. Ethanol was added to flavored water 5.5% v/v and sodium saccharin was .25% w/v in flavored water.
Procedure. Rats were conditioned to prefer either cherry or grape flavor by alternating exposures to flavored ethanol and flavored saccharin. Conditioning proceeded as in Experiment 1 with the following differences. Conditioning lasted 16 days with eight exposures to each solution. Additionally, all rats received their flavored ethanol on odd days, and on even days the amount of flavored saccharin available was yoked to each rats' intake of flavored ethanol. Thus the intakes of the two solutions was equal during conditioning. Experiment 1 had floor effect problems in the two-bottle tests, so in the present experiment tests were longer (3 hr) and were all one-bottle tests. Each rat was tested four times: two tests of the saccharin-paired flavor, two tests of the ethanol-paired flavor. One of each flavor test was preceded by an injection of naloxone and the other by an injection of saline. Test order was counterbalanced across flavor and drug. Test intakes were subjected to repeated measures ANOVA with drug and paired solution as within rat factors.
Figure 2 shows the intakes as a function of days in conditioning and during the tests. The rats increased their intake over the course of conditioning. The simple main effect of solution in the saline condition was the test of preference for the ethanol-paired flavor in the absence of drug, and it was significant, F(1, 7) = 27.7, p [less than] .001. The main effect for drug did not reach significance, F(1, 7) = 2.65, p [greater than] .05, because intake of the saccharin-paired flavor was not suppressed by naloxone. However, intake of the ethanol-paired flavor was reduced by naloxone as was revealed by the significant drug x flavor interaction, F(1, 7) = 8.55, p [less than] .025.
In the present experiment naloxone specifically reduced intake of an ethanol-paired flavor and not a saccharin-paired flavor when rats had sufficient experience to learn a preference for the flavor paired with ethanol. By using 3-hr, one-bottle tests, the floor effects found in Experiment 1 were overcome.
If it is true that naloxone reduces intake of a calorie-paired flavor by antagonizing conditioned activation of central reward systems stimulated by the taste of a flavor paired with calories, then it is possible that naloxone during the conditioning trials themselves might prevent learning of the preference. Experiment 3 sought evidence for that hypothesis.
Experiments 1 and 2 have shown that naloxone disrupts a preference for a calorie-paired flavor when rats are tested for preference following conditioning. These data suggest that blockade of endogenous opioid activity that may occur due to ingestion of nutrients might result in blunted or no preference being learned for a flavor that is paired with the nutrients. In the present experiment, we gave rats experience with flavored dextrose and flavored saccharin, but half the rats had naloxone paired with the dextrose and the other half had naloxone paired with the saccharin. We predicted that naloxone-dextrose rats would not prefer the dextrose-paired flavor and naloxone-saccharin rats would show a dextrose-paired flavor preference.
Participants. Sixteen male Long Evans rats 90 days old served as subjects. The rats were assigned to either the NAL-DEX or the NAL-SAC groups by matching their weights. The two groups had mean weights of 375.6 and 374 g. Rats were housed as in the previous experiments.
Solutions, flavors, and drugs. Cherry and grape Kool Aid were used as flavors in the same concentration as before. Saccharin was again .25% w/v and dextrose was 10% w/v. Naloxone and saline were as before.
Procedure. In the first phase of the experiment, rats were trained to drink at the same time each day by restricting their access to water. All rats were placed in individual cages and given 15 g of Purina Rat Chow at 1100 on the first day of the experiment. No water was available until the next day, when water bottles were retuned for 1 hr. At the end of the hour, water intakes were recorded and the rats were fed 15 g of chow. This procedure was followed for 5 days.
In the second phase, injections of naloxone or saline preceded access to flavored conditioning solutions. Rats in Group NAL-DEX were injected with saline, then 15 min later were given 35 ml of flavored saccharin that was available for 1 hr. Following the saccharin, the rats were fed their 15-g ration. The next day the rats were injected with naloxone and 15 min later the flavored dextrose was presented for 1 hr. Conditioning proceeded in that manner for 8 days with each odd day being a replication of Day 1 and each even day a replication of Day 2. Rats in group NAL-SAC received the same treatment, but the drug/solute relationship was reversed. Thus naloxone was paired with saccharin instead of dextrose.
On the 9th day of the conditioning phase, a two-bottle test between cherry and grape flavored water ensued. The test lasted 1 hr and occurred at the regular conditioning time. Intake of the dextrose-paired flavor was converted to a preference ratio (dextrose-paired flavor divided by total intake of both flavors) and subjected to ANOVA.
During Days 10 - 16 the rats had ad libitum chow and water rations. On Day 17, rats were restricted to 15 g of chow overnight, and the next day a second preference test occurred. In this test, saccharin and dextrose were combined (sac/dex) and flavored with cherry or grape. Each rat received two bottles, one with cherry flavored sac/dex and one with grape flavored sac/dex. Intakes were subjected to repeated measures ANOVA with flavor as a within rat factor and group as a between rat factor. This ended the experiment.
Intakes during the conditioning phase were analyzed by ANOVA with groups as a between rat factor and days and solutions as within rat factors. The groups did not differ in overall intake during conditioning, F(1, 14) = 1.04, p [greater than] .05. Consumption did significantly increase over the four exposures to the solutions from 13.9 ml on the first day to 21.3 ml on the last day, F(3, 42) = 14.62, p [less than] .001. There was no group by day interaction, F(3, 42) = 2.8, p [greater than] .05.
The rats drank reliably more of the dextrose solution than of the saccharin solution collapsed over drugs and days, F(1, 14) = 8.2, p [less than] .02. Of particular interest was the significant drug by solution interaction. Rats drank more dextrose than saccharin if they had saline paired with dextrose. Rats that had naloxone paired with dextrose drank more saccharin than dextrose, F(1, 14) = 48.3, p [less than] .0001. Finally, the group by day by solution interaction was significant, F(3, 42) = 8.7, p [less than] .001. This was because rats in Group NAL-DEX drank more flavored saccharin than flavored dextrose on every trial, while rats in Group NAL-SAC drank more flavored dextrose than flavored saccharin on every trial. Figure 3 displays the intakes by the two groups as a function of days and drug condition.
Intakes during the first two-bottle preference test were converted to preference ratios and subjected to ANOVA. Rats in Group NAL-SAC preferred the flavor paired with dextrose (M = .90 +/- .02) whereas rats in Group NAL-DEX did not (M = .55 +/- .07) and the difference between the groups was reliable, F(1, 14) = 17.1, p [less than] .001.
Intakes during the final two-bottle test between cherry and grape dex/sac solutions was subjected to ANOVA. The analysis revealed no significant main effects for groups, F(1, 14) = 2.86, or solution, F(1, 14) = .72, but the group by solution interaction was significant, F(1, 14) = 7.54, p [less than] .025. The interaction was due to a preference for the dextrose-paired flavor in Group NAL-SAC and a lack of preference for the dextrose-paired flavor in Group NAL-DEX. Figure 4 shows the interaction.
In the present experiment, rats that had naloxone during dextrose trials did not develop a preference for a dextrose-paired flavor, while rats that had naloxone during saccharin trials did develop a preference for the dextrose-paired flavor. It is unlikely that this was caused by an aversion for the flavor concomitant with naloxone, since inspection of Figure 3 shows no decrease over days of flavored solutions consumed under the influence of naloxone. However, these data do not rule out flavor-flavor mechanisms in the conditioning trials, because naloxone might be having its effect simply by reducing the palatability of whatever solution it is paired with.
In Experiment 1 rats' preference for a flavor paired with dextrose over a flavor paired with water was disrupted in a two-bottle test by administration of naloxone. In Experiment 2, expression of a preference for a flavor paired with ethanol over a flavor paired with saccharin was similarly blocked by naloxone. Experiment 3 showed that preference for a dextrose-paired flavor over a saccharin-paired flavor does not develop when dextrose trials are concomitant with naloxone administration.
Experiment 4 used another approach to find evidence that flavor preferences based on calories might depend on endogenous reward systems. Because increased levels of endogenous opioids decrease fear and thus freezing to shock (Fanselow, 1991), another way of testing for the contribution of "reward system" activity aroused by flavor-calorie conditioning would be to use the freezing response as a probe for conditioned endogenous opioid activity. Rats should freeze less when shocked after tasting a calorie-paired flavor than when shocked after tasting a noncalorie paired flavor if the flavor is a CS and opioid activity is a CR.
Participants. Fourteen male Sprague Dawley rats about 180 days old were used. The rats were housed in the same conditions as in the previous experiments.
Solutions and flavors. Orange and grape Unsweetened Kool Aid served as the flavor cues and were mixed in tap water at a concentration of .25% w/v as in Experiment 1. Once again, dextrose (10% w/v) was added to make the calorie solution. Sodium saccharin (Sigma) was .25% w/v. The conditioning procedure was shortened by giving the rats two bottles each day, one containing 50 ml of flavored dextrose, the other 50 ml of flavored saccharin. Flavors were counterbalanced between rats so that half had grape dextrose and orange saccharin while the other half had the reverse. Position of the bottles was alternated daily.
Shock. Shock was administered in a Lafayette Operant Conditioning Apparatus (Model 81320). The chamber measured 27 cm long by 21 cm wide by 21 cm high and was constructed of aluminum end walls and clear plastic sides and top. Shock was delivered through the grid floor made up of metal rods spaced 1.5 cm apart. One mA shocks were 1 second in duration and the chamber was cleaned with ammonium hydroxide between subjects.
Procedure. Conditioning proceeded for 5 days. Each day at 1100 rats' intakes from their two bottles were recorded, they were fed their 15-g ration of chow, and new bottles were put on the cages. On the day following the last conditioning day rats had ad libitum water and 15 g of chow. Next day, each rat was allowed to drink either its dextrose-paired or its saccharin-paired flavor for 2 min. Intakes were not measured, the idea was simply to let each rat taste its preferred flavor. Then an assistant placed a rat into the shock chamber. The observer was not aware of the group from which the rat was taken. After a 1-rain exploration period, a 1-sec shock occurred. The rat was observed constantly and the number of seconds freezing during each 15-sec interval was recorded. At the end of the first min. a second shock occurred and the observations were repeated. The rat received 8 shocks, separated by 1-min intervals, in which freezing was recorded. The rats were tested individually, half after their dextrose flavor and half after their saccharin flavor. Following the test, they were fed 15 g of chow and ad libitum water was returned. The next day, the testing process was repeated, but each rat had the other flavor.
After a 2-day period of 15 g of chow and ad libitum water, a two-bottle preference test for flavor-calorie preference took place. The rats were offered 50 ml each of their dextrose-paired and their saccharin-paired flavors for 2 hr and intakes were recorded at the end of the test. This concluded the experiment.
During conditioning, the rats drank more of the flavored dextrose than of the flavored saccharin. The mean intake of flavored dextrose rose over days from 37.9 to 40 ml (which was all that could be obtained from the bottles, because they tilt slightly away from the cage fronts), while the mean intake of flavored saccharin decreased from 24.9 to 14.5 ml. The results of the shock-induced freezing tests are displayed in Figure 5. The total number of seconds freezing during each 8-min test by each rat was subjected to repeated measures ANOVA with dextrose-paired flavor versus saccharin-paired flavor as the within rat factor. Rats froze more after their saccharin flavor than after their dextrose flavor, F(1, 26) = 4.83, p [less than] .05. The results of the two-bottle preference test confirmed that the rats had learned a preference for their calorie-paired flavor, F(1, 26) = 34.8, p [less than] .001. Figure 6 shows the mean intakes during the test. A correlation between each rat's preference ratio (consumption of calorie flavor divided by total consumption) and proportion of time freezing after tasting the calorie-paired flavor was computed. The correlation was small and not significant, r = -.06, p [greater than] .05.
The finding that rats froze less after tasting their calorie-paired flavor brings converging evidence to the idea that flavor-calorie preferences might be mediated by an endogenous opioid reward system. Although beyond the scope of these investigations, it is likely that some neuroendocrine result of metabolism is associated with activity in this reward system (see Blundell & Lawton, 1993, for an excellent review).
The present set of experiments provides evidence that some form of endogenous opioid activity is involved in flavor-calorie preference learning. Antagonism by naloxone disrupted expression of a dextrose- or ethanol-based flavor preference (Experiments 1 & 2) and blocked the learning of a dextrose-paired flavor preference (Experiment 3). Additionally, tasting a flavor that had been previously paired with dextrose resulted in less freezing to shock, compared to freezing after tasting a flavor that had been previously paired with saccharin (Experiment 4). Although these investigations only looked at reward system activity in an indirect way, it is reasonable to assert that more direct investigations of the relationship between central opioid reward systems and flavor-calorie preference learning are warranted.
Opiate antagonists such as naltrexone lower the pleasantness of a sweet solution in humans but have no effect on hunger (Fantino, Hosotte, & Apelbaum, 1986). Naloxone also decreases food intake in humans (Atkinson, 1982; Morley & Levine, 1982; Trenchard & Silverstone, 1983) while again having no effect on hunger (Blundell & Hill, 1988). Chronic peripheral treatment with naloxone reduces food intake in free feeding rats (Brands, Thornkill, Hirst, & Gowdey, 1979), and peripheral injections of naloxone reduces eating by hungry rats (Holtzman, 1974; Rodgers, 1978). In sheep, infusion of opiates into the ventricles increases food intake, and peripheral injections of naloxone decreases food intake (Baile, Keim, Della-Ferra, & McLaughlin, 1980).
Pretreatment with sweet tastes can provide analgesia by activating opioid systems. Sucrose consumption results in reduced rat pup distress vocalizations (Blass, 1986) and reduced pain thresholds and increased analgesic efficacy of exogenous morphine (Schoenbaum, Martin, & Roane, 1989; Roane & Martin, 1990).
At least some of these effects may be conditioned to flavors paired with dextrose as well, because rats show (a) reduced preference for flavors paired with dextrose or ethanol when pretreated with naloxone, (b) are prevented from forming flavor preferences for dextrose-paired flavors consumed after naloxone injections, and (c) reduced freezing to shock after sampling a dextrose-paired flavor. Conditioned activation of opioid reward systems may be what we mean when we say the rat learns to like a flavor.
ATKINSON, R. L. (1982). Naloxone decreases food intake in obese humans. Journal of Clinical Endocrinology & Metabolism, 55, 196-198.
BAILE, C. A., KEIM, D. A., DELLA-FERRA, M. A., & MCLAUGHLIN, C. L. (1980). Opiate antagonists and agonists and feeding in sheep. Physiology & Behavior, 26, 1019-1023.
BLASS, E. M. (1986). Functional interaction between the positive affect of sweet and the negative affect of pain and distress. Appetite, 7, 243.
BLUNDELL, J. E., & HILL, A. J. (1988). On the mechanism of action of dexfenfluramine: effect on alliesthesia and appetite motivation in lean and obese subjects. Clinical Neuropharmacology, 11 (Suppl. 1), S121 - S134.
BLUNDELL, J. E., & LAWTON, C. L. (1993). Pharmacological aspects of appetite. In A. J. Stunkard & T. A. Wadden (Eds.), Obesity: Theory & therapy (2nd ed.). New York: Raven Press, Ltd.
BOOTH, D. A. (1972). Conditioned satiety in the rat. Journal of Comparative and Physiological Psychology, 81, 457-471.
BOOTH, D. A. (1985). Food-conditioning eating preferences and aversions with interoceptive elements: conditioned appetites and satieties. Annals of the New York Academy of Sciences, 443, 22-41.
BRANDS, B., THORNKILL, J. A., HIRST, M., & GOWDEY, C. W. (1979). Suppression of food intake and body weight gain by naloxone in rats. Life Sciences, 24, 1773-1778.
COOPER, S. J. (1983). Effects of opiate agonists and antagonists on fluid intake and saccharin choice in the rat. Neuropharmacology, 22, 323-328.
COOPER, S. J., & KIRKHAM, T. C. (1990). Basic mechanisms of opioids' effects on eating and drinking. In L. D. Reid (Ed.), Opioids, bulimia, & alcohol abuse and alcoholism (pp. 91-110). New York: Springer Verlag.
FANSELOW, M. J. (1991). Analgesia as a response to aversive Pavlovian conditional stimuli: Cognitive and emotional mediators. In M. R. Denny (Ed.), Fear, avoidance, & phobias (pp. 61-86). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
FANSELOW, M., & BIRK, J. (1982). Flavor-flavor associations induce hedonic shifts in taste preference. Animal Learning & Behavior, 10, 223-228.
FANTINO, M., HOSOTTE, J., & APELBAUM, M. (1986). An opioid antagonist, naltrexone, reduces preference for sucrose in humans. American Journal of Physiology, 251, R91-R96.
FEDORCHAK, P. M., & BOLLES, R. C. (1987). Hunger enhances the expression of calorie- but not taste-mediated conditioned flavor preferences. Journal of Experimental Psychology: Animal Behavioral Processes, 13, 73-79.
FEDORCHAK, P. M., & BOLLES, R. C. (1988). Nutritive expectancies mediate cholecystokinin's suppression-of-intake effect. Behavioral Neuroscience, 102, 451-455.
GRILL, H. J., & NORGREN, R. (1978). The taste-reactivity test. I. Mimetic responses to gustatory stimuli in neurologically normal rats. Brain Research, 143, 263-269.
HOLTZMAN, S. G. (1974). Behavioral effects of separate and combined administration of naloxone and d-amphetamine. Journal Pharmacology & Experimental Therapy, 189, 51-60.
LE MAGNEN, J., MARFAING-JALLAT, P., MICELI, D., & DEVOS, M. (1980). Pain modulating and reward systems: A single brain mechanism? Pharmacology, Biochemistry & Behavior, 12, 729-733.
LYNCH, W. C., & LIBBY, L. (1983). Naloxone suppresses intake of highly preferred saccharin in food deprived and sated rats. Life Sciences, 33, 1909-1914.
MEHIEL, R. (1991). Hedonic shift conditioning with calories. In R. C. Bolles (Ed.), The hedonics of taste (pp.107-126). Hillsdale, NJ: Lawrence Erlbaum.
MEHIEL, R., & BOLLES, R. C. (1984). Learned flavor preferences based on caloric outcome. Animal Learning & Behavior, 12, 421-427.
MEHIEL, R., & BOLLES, R. C. (1988a). Learned flavor preferences based on calories are independent of initial hedonic value. Animal Learning & Behavior, 16, 383-387.
MEHIEL, R., & BOLLES, R. C. (1988b). Hedonic shift learning based on calories. Bulletin of the Psychonomic Society, 26, 459-462.
MORLEY, J. E., & LEVINE, A. S. (1982). The role of the endogenous opiates as regulators of appetite. American Journal of Clinical Nutrition, 35, 757-761.
PEREZ, E., & SCLAFANI, A. (1991). CCK conditions flavor preference in rats. American Journal of Physiology, 260, R179-R185.
RAMIREZ, L. (1994). Flavor preferences conditioned with starch in rats. Animal Learning and Behavior, 22, 181-187.
ROANE, D. S., & MARTIN, R. J. (1990). Continuous sucrose feeding decreases pain threshold and increased morphine potency. Pharmacology, Biochemistry & Behavior, 35, 225-229.
RODGERS, R. J. (1978). Opioid peptides, brain and behavior: a brief review. Irish Journal of Medical Science, 147, suppl. 1, 57-60.
SCHOENBAUM, G. M., MARTIN, R. J., & ROANE, D. S. (1989). Relationships between sustained sucrose-feeding and opioid tolerance and withdrawal. Pharmacology, Biochemistry & Behavior, 34, 911-914.
SCLAFANI, A. (1991). The hedonics of sugar and starch. In R. C. Bolles (Ed.), The hedonics of taste, (p. 59-87). Hillsdale, NJ: Lawrence Erlbaum.
SCLAFANI, A., & NISSENBAUM, J. W. (1988). Robust conditioned flavor preference produced by intragastric starch infusions in rats. American Journal of Physiology, 255, R672-R675.
SIVIY, S. M., & REID, L. D. (1983). Endorphinergic modulation of acceptability of putative reinforcers. Appetite, 4, 249-257.
TORDOFF, M. G., & FRIEDMAN, M. I. (1986). Hepatic portal glucose infusions decrease food intake and increase food preference. American Journal of Physiology, 251, R192-R196.
TORDOFF, M. G., TEPPER, B. J., & FRIEDMAN, M. I. (1987). Food flavor preferences produced by drinking glucose and oil in normal and diabetic rats: Evidence for conditioning based on fuel oxidation. Physiology & Behavior, 41, 481-487.
TORDOFF, M. G., ULRICH, P. M., & SANDLER, F. (1990). Flavor preferences and fructose: Evidence that the liver detects the unconditioned stimulus for calorie-based learning. Appetite, 14, 29-44.
TRENCHARD, E., & SILVERSTONE, T. (1983). Naloxone reduces the food intake of normal human volunteers. Appetite, 4, 43-50.
YOUNG, P. T. (1961). Motivation and emotion. New York: Wiley.
WALSH, M. L., WHITE, L. R., & ALBERT, D. J. (1984). Dissociation of prey killing and prey eating by naloxone in the rat. Pharmacology, Biochemistry & Behavior, 21, 5-7.
|Printer friendly Cite/link Email Feedback|
|Publication:||The Psychological Record|
|Date:||Jun 22, 1996|
|Previous Article:||The effect of a change in body weight on running and responding reinforced by the opportunity to run.|
|Next Article:||Arbitrarily applicable relational responding and sexual catogorization: a critical test of the derived difference relation.|