Search

Fos-like immunoreactivity (FLI) can indicate the location of neurons activated following expression of conditioned taste aversion (CTA). After one conditioning trial FLI has been identified in the intermediate nucleus of the solitary tract (iNTS) with little expression in other brain regions. The present study assessed the effect of increasing aversion strength on the magnitude and anatomical distribution of FLI during CTA expression. When animals received three rather than one conditioning trial, significant FLI was seen not only in the iNTS but also in the parabrachial nucleus (PBN), and the central nucleus of the amygdala (CNA), regions thought to be important in taste aversion learning.

The induction of c-Fos-like immunoreactivity (c-FLI) in the nucleus of the solitary tract (NTS) has been shown to be correlated with behavioral expression of a conditioned taste aversion (CTA). However, because this cellular response is also dependent on an intact amygdala, it may represent the activation of a stress-related autonomic response. The present experiments addressed this possibility by evaluating the correlation between c-FLI in the intermediate division of the NTS (iNTS) and 2 measures of conditioned fear: freezing and changes in mean arterial pressure (MAP) and heart rate (HR). Exposure to the taste conditioned stimulus (CS) resulted in a marked induction of c-FLI in the iNTS, whereas exposure to a fear CS did not. Further, exposure to a taste CS did not selectively lead to increases in MAP or HR. Results suggest that induction of c-FLI in the iNTS may reflect the activation of a cell population whose function is unique to the CTA paradigm.

Increases in c-Fos-like immunoreactivity (FLI) in the intermediate nucleus of the solitary tract (iNTS) have been seen consistently as a correlate of the expression of a conditioned taste aversion (CTA) when conditioning occurs using taste delivery through intraoral (I/O) infusions. The present study examined whether a similar FLI response would occur when conditioning was accomplished by presenting the taste solution in a bottle. I/O and bottle methods generated aversions that were comparable, when judged by the behavioral response of solution rejection. However, elevations in FLI were seen only in animals conditioned with the I/O method. This finding adds to evidence that the neural pathways underlying CTA learning differ as a function of the type of conditioning method used.

The specific role of insular cortex in acquisition and expression of a conditioned taste aversion was assessed using two different conditioning methods, which vary mode of taste delivery. Involvement of insular cortex in the induction of c-Fos-immunoreactivity in the nucleus of the solitary tract, a cellular correlate of the behavioral expression of a conditioned taste aversion, was also assessed. Electrolytic lesions of insular cortex blocked behavioral expression of a conditioned taste aversion and this was evident not only when lesions were placed prior to conditioning, but also when they were made after conditioning but before testing. In contrast to the effects on behavior, lesions did not completely block the c-Fos-immunoreactivity which accompanies re-exposure to the aversive taste. In addition, the blocking of behavioral evidence of aversion conditioning by cortical lesions was seen both in animals trained under an intraoral acquisition procedure and those trained with bottle-conditioning. This contrasts with previous work with amygdala lesions which showed that amygdala was absolutely necessary for taste aversions conditioned with the intraoral method but not for those conditioned using bottle presentation of the taste. Overall, these findings imply that the details of the neural circuitry involved in taste aversion learning, including its anatomical distribution, complexity and degree of redundancy, vary with the type of conditioning method employed.

Food aversions that are acquired as a result of unpleasant experiences with foods represent a potent defense mechanism against poisoning. However, this powerful and durable form of conditioning can also contribute to avoidance of foods that are not poisonous, and are, in fact, quite nutritious. This is because such foods may be coincidentally associated with unpleasant gastrointestinal symptoms, sometimes due to transient, unrelated illness, or unpleasant drug side effects. Most of the studies of naturally occurring learned food aversions in humans have been focused on subjects of college age, so we have limited information about the extent to which such food aversions occur in the elderly. Additionally, most studies have employed questionnaire or interview methods that may have some significant limitations in the accurate assessment of the incidence of food aversions. Thus, although food aversion learning has been thoroughly documented in the animal laboratory, its role in everyday food selection in humans, including the elderly, remains relatively unclear.

Although an important role for the amygdala in taste aversion learning has been suggested by work in a number of laboratories, results have been inconsistent and interpretations varied. The present series of studies reevaluated the role of the amygdala in taste aversion learning by examining the extent to which conditioning methods, testing methods and lesioning methods, influence whether amygdala lesions dramatically affect conditioned taste aversion (CTA) learning. Results indicated that when animals are conditioned with an intraoral (I/O) taste presentation, lesions of amygdala eliminate evidence of conditioning whether animals are tested intraorally or with a two-bottle solution presentation. Dramatic effects of amygdala lesions on CTA learning were seen whether lesions were made electrolytically or using an excitotoxin. In contrast, when animals were conditioned using bottle presentation of the taste, electrolytic lesions attenuated CTAs but did not eliminate them, and excitotoxic lesions had no effect. These results are consistent with the hypothesis that neural structures critical for CTA learning may differ depending on the extent to which the method of conditioned stimulus delivery incorporates a response component.

The induction of c-Fos-like immunoreactivity (c-FLI) in the intermediate division of the nucleus of the solitary tract (iNTS) has been shown to be a cellular correlate of the behavioral expression of a conditioned taste aversion (CTA). To further define neuroanatomical structures and pathways that contribute to this cellular response and to CTA learning in general, electrolytic lesions of insular (gustatory) cortex (IC) were combined with immunostaining for c-FLI. Rats were given either unilateral or bilateral electrolytic lesions of insular cortex or 'sham' operations. Following surgery, 'paired' animals were given a single conditioning trial consisting of intraoral infusion of 5-ml 0.15% sodium-saccharin followed by injection with LiCl (0.15 M, 20 ml/kg, i.p.) while 'unpaired' controls received a non-contingent saccharin-LiCl presentation. Rats with bilateral lesions showed no behavioral evidence of having acquired a CTA. Increases in c-FLI in iNTS were evident, but reduced, relative to 'sham' animals. Rats with unilateral-lesions displayed a CTA by rejecting the saccharin, although increases in c-FLI on the side of the iNTS ipsilateral to the lesion were reduced relative to that seen in 'sham' animals. A comparison of these results with those obtained after amygdala lesions supports the conclusion that amygdala and insular cortex are necessary, but not sufficient, for the behavioral expression of a CTA., (Copyright 1998 Elsevier Science B.V. All rights reserved.)

Food deprivation increases the rewarding effects of self-administered drugs such as psychomotor stimulants and benzodiazepines. These drugs also possess aversive properties and can produce conditioned taste aversions (CTA). Because drug-seeking behavior is most likely affected by both the rewarding and aversive properties of drugs, we hypothesize that food deprivation might also attenuate a drug's aversive consequences. The CTAs induced by three different drugs (amphetamine, chlordiazepoxide, and LiCl) were assessed separately. Male Long-Evans rats were assigned to one of two feeding conditions: restricted (maintained at 80% of free-feeding body weight), or nonrestricted (with ad lib food). Both groups received CTA training, consisting of an intraoral infusion of a novel saccharin solution (10 min) followed immediately by one of two i.p. injections: paired rats received drug, and unpaired rats received a similar volume of saline. After 10 days of ad lib food access, saccharin was presented to all rats again, and the latency to reject the tastant was used as an index of CTA learning. The rats that had been food restricted at the time of conditioning exhibited attenuated CTAs relative to those that had not been deprived. These differences were seen only when a rewarding drug (amphetamine or chlordiazepoxide) and not when a nonrewarding drug (LiCl) was used as the unconditioned stimulus. In a separate experiment, we established that this effect is apparent only when the deprivation period precedes conditioning rather than precedes testing. The present results indicate that food deprivation modulates the acquisition of a CTA induced by amphetamine or chlordiazepoxide, but not LiCl.

Leptin (ob protein) and glucagon-like peptide-1-(7-36) amide (GLP-1) are peptides recently proposed to be involved in the regulation of food intake. Although the ability of exogenous leptin and GLP-1 to modulate consummatory behavior is consistent with the suggestion that these peptides are endogenous regulatory agents, central administration of these peptides may have aversive side effects, which could explain the anorexia. In the present experiment, exposure to a saccharine taste was immediately followed by central administration of leptin or GLP-1 to determine if these drugs could produce a conditioned taste aversion (CTA) in rats. At doses equated for producing comparable reductions in short-term food intake, GLP-1, but not leptin, generated a robust CTA. Although leptin caused no aversion, this peptide was the only drug to cause relatively long-term reductions in food consumption (16 h) and body weight (24 h). Hence, the results indicate that central GLP-1 produces aversive side effects, and it is argued that these nonspecific effects may explain the anorectic actions of GLP-1.

The induction of c-Fos-like immunoreactivity (c-FLI) in the intermediate division of the nucleus of the solitary tract (iNTS) has been shown to be a reliable cellular correlate of the acquisition and/or behavioral expression of a conditioned taste aversion (CTA). To begin to define neuroanatomical structures and pathways that contribute to this cellular response and to CTA learning in general, electrolytic lesions of the amygdala were combined with immunostaining for c-FLI. Rats were given either unilateral or bilateral electrolytic lesions of the amygdala or "sham' operations. Following surgery "paired' animals were given a single conditioning trial consisting of intraoral infusion of 5 ml 0.15% sodium-saccharin followed by injection with LiCl (0.15 M, 20 ml/kg, i.p.) while "unpaired' controls received a non-contingent saccharin-LiCl presentation. When tested, unilateral-lesioned rats displayed a CTA by rejecting the saccharin, but increases in c-FLI were evident only on the side of the iNTS contralateral to the lesion. Rats with bilateral lesions showed no evidence of having acquired a CTA and no increase in c-FLI in iNTS relative to unpaired controls. These findings support involvement of amygdala in CTA learning and suggest that a lateralized connection between amygdala and iNTS is necessary for the conditioned c-FLI which is induced by exposure to a conditioned aversive taste.

Conditioned taste aversion (CTA) learning is a robust form of classical conditioning in which animals rapidly associate a flavor with aversive internal symptoms. The present study assessed CTA learning in transgenic mice deficient in a specific nonreceptor tyrosine kinase (the fyn mutant). Fyn mutants show impaired long-term potentiation and marked deficits in acquisition of spatial learning tasks. To assess whether they are also impaired in CTA learning, fyn mutant and wild-type mice received 2 conditioning trials consisting of access to a flavored solution followed by administration of LiCl. Fyn mutant mice acquired significant CTAs following a single conditioning trial and these aversions were comparable to those seen in wild-type mice. These results indicate that the fyn mutation does not interfere with the acquisition of CTAs and hence that this mutation is not associated with a global learning deficit.

Fischer 344 (F344) rats fail to prefer sodium chloride (NaCl) solution to water, and this behavior is dramatically altered by bilateral transection of the chorda tympani nerve (CTX). Tests of retention and generalization of a conditioned taste aversion (CTA) to 0.15 M NaCl were used to assess alterations in salt taste perception after CTX. A CTA was established to 0.15 M NaCl in groups of F344 and Wistar rats after two pairings with LiCl (IP, 2% body weight; 0.15 M). After conditioning, animals received bilateral CTX or sham operations. Approximately 2 weeks after surgery animals were tested for retention of the CTA. Aversion to 0.15 M NaCl was evident in CTX and SHAM rats that had been conditioned prior to surgery, with no apparent difference in magnitude as a function of surgical condition. Thus, although CTX profoundly alters F344 rats' hedonic response to NaCl, it does not alter perceptual characteristics so markedly that NaCl is no longer recognized as the stimulus presented during aversion conditioning. Rather these studies suggest that in both the F344 and Wistar strains the chorda tympani nerve is not necessary for retention of a presurgical CTA to NaCl. These studies, therefore, do not provide evidence of changes in NaCl perceptual "quality" as a consequence of CTX either in F344 rats, where such changes were indicated by the preference data, or in the Wistar rat, where they were not.

A conditioned taste aversion (CTA) is a form of classical conditioning in which animals avoid a taste (conditioned stimulus; CS) which has been previously paired with a treatment (unconditioned stimulus; US) that produces transient illness. Recently, a reliable cellular correlate of the behavioral expression of a CTA was identified using c-Fos immunostaining as a marker of neuronal activation. Exposure to a saccharin solution (CS) which had previously been paired with lithium chloride (LiCl; US) induced significant c-Fos-like immunoreactivity (c-FLI) in the intermediate zone of the nucleus of the solitary tract (NTS), a response that was quite similar to that displayed following administration of LiCl alone. The present studies employed a variant of the chronic decerebrate rat preparation to explore whether circuitry intrinsic to the brainstem is sufficient for the induction of c-FLI in both as an unconditioned response to the LiCl and as a conditioned response to the saccharin. Using chronic hemidecerebrate rats, which have a unilateral brain transection at the level of the superior colliculus, we found that the unconditioned c-FLI to LiCI was unaltered by the transection, while the conditioned expression of c-FLI to the CS taste was evident only on the side of the NTS which retained neural connections with the forebrain. These findings strongly implicate forebrain input in this cellular correlate of CTA learning and also indicate that the pathways mediating the response to the US (LiCl) and the CS (saccharin) differ.

Temporal parameters of taste aversion learning are known to differ markedly from other learning paradigms in that acquisition occurs despite lengthy delays between exposure to conditioned (CS) and unconditioned stimulus (US). Far less consideration has been paid to very brief CS-US intervals and the possibility that this learning may also be distinguished by an ineffectiveness of close temporal contiguity between CS and US. The effectiveness of a very brief CS-US interval (10 s) was compared with that of 2 lengthier intervals (15 and 30 min). Temporal control of CS delivery (0.15% saccharin solution) into the oral cavity and US delivery (7.5 mg/kg apomorphine hydrochloride) into circulation involved infusion pumps and indwelling catheters. Using a 1-trial learning paradigm, CS-US delays of 15 and 30 min led to significant aversions whereas the 10-s CS-US interval did not, suggesting that close temporal contiguity between CS and US is neither necessary nor sufficient for conditioned taste aversion acquisition.

Taste aversion conditioning is characterized by its prompt acquisition despite the introduction of long delays between CS (taste) and UCS (toxic drug). Although the dramatic changes in behavioral response to a taste after this conditioning are well documented, relatively little is known about the changes in neural activity that accompany this learning. c-Fos immunohistochemical staining was employed to define brain regions activated during the expression of a conditioned taste aversion (CTA). The present studies examined c-Fos immunoreactivity in the brains of rats after i.p. injection of LiCl or NaCl or after intraoral infusion of a saccharin CS. LiCl administration, a common unconditioned stimulus (UCS) in CTA experiments, was found to induce c-Fos protein in a number of brainstem regions, including the nucleus of the solitary tract (NTS), medial and lateral pontine parabrachial nucleus (PBN), and hypoglossal nucleus. Conditioned animals received a single pairing of the CS saccharin with the UCS LiCl, while controls were exposed to the CS saccharin but received non-contingent LiCl 24 h after saccharin exposure. Following saccharin re-exposure, conditioned animals showed patterns of neuronal activation to a taste which were similar to those activated by the UCS drug. Specifically, the pattern of c-Fos expression in conditioned animals was confined to the same region of the NTS which showed the most activation following the UCS LiCl. This pattern of activation was not evident in controls re-exposed to saccharin.(ABSTRACT TRUNCATED AT 250 WORDS)

The area postrema (AP), a chemoreceptor trigger zone for nausea and vomiting, has been implicated in taste aversion conditioning with LiCl. In addition to taste aversion acquisition, the present studies indicate that a number of other responses to LiCl administration are eliminated by lesions of the AP. These include a behavioral response, 'lying-on-belly' as well as two physiological responses, delayed stomach emptying and hypothermia. These findings suggest that the area postrema is critically involved in the detection of LiCl and in a wide range of responses to this toxin. They also provide strong evidence that the failure to acquire conditioned taste aversions to LiCl-paired flavors after AP lesions can be attributed to the absence of a significant 'illness' response in lesioned animals.

Fischer 344 (F344) rats are atypical in their lack of preference for any concentration of NaCl solution over water. It was hypothesized that abnormal signals mediated by the chorda tympani nerve (CT) could be causally involved in NaCl avoidance by F344 rats. This study assessed whether CT transection would normalize the salt preference of F344 rats. Preference for NaCl solutions (0.6%, 0.8%, and 1.0%) versus water was assessed using two-bottle preference tests. At all concentrations tested, CT-transected animals preferred NaCl solutions to water. This preference differed dramatically from the avoidance of these solutions by controls. These findings are striking, particularly because CT transections have generally failed to significantly affect NaCl preference in other rat strains. The results are consistent with the hypothesis that in F344 rats the avoidance of the taste of NaCl stems from input mediated by the CT.

Tumor necrosis factor (TNF) or cachectin has been proposed as an important mediator of cancer anorexia and cachexia. The present studies examined the extent to which TNF administration generates symptoms similar to those produced by tumor growth. Like the growth of certain tumors, TNF administration was found to be associated with the development of strong aversions to a novel diet. Area postrema lesions were found to significantly attenuate the effects of TNF on intake of a novel diet, a finding previously reported for tumor anorexia. In addition, the anorexic effects of TNF differed considerably as a function of the novelty of the diet. When the available diet was novel, effects of TNF in lowering food intake were substantial, whereas more modest effects were seen when the diet was familiar. These findings provide evidence for parallels between TNF- and tumor-induced anorexias. Nonetheless, these studies also confirm previous observations of the rapid development of tolerance to the anorexic effects of TNF, a finding that is not consistent with a role for TNF as a critical mediator of tumor anorexia.

The present experiments examined whether behavioral conditioned responses (CRs) develop to lithium chloride (LiCl)-paired tastes and whether these CRs are similar to the behaviors that follow administration of the drug. Rats were exposed to a saccharin solution via intraoral infusions before being injected with either LiCl or saline. CRs were assessed after conditioning when the saccharin conditioned stimulus was delivered alone. The unconditioned response to LiCl delivery is a very distinctive posture that has been termed "lying-on-belly." The present study indicates that this behavior pattern also occurs after the delivery of a LiCl-paired taste solution. The similarity between these unconditioned and conditioned behaviors is consistent with the hypothesis that responses are conditioned during taste aversion acquisition and that CRs are similar to those that are generated by the drugs used in conditioning.

Growth of a Leydig cell [LTW (m)] tumor in rats is associated with the development of significant hypophagia and severe aversions to the available diet. Lesions of the area postrema and nearby caudal medial area of the solitary tract were found to block or greatly attenuate the anorexia and food aversions that typically accompany the growth of this tumor. Thus these tumor-induced symptoms may be associated with the detection of blood-borne chemicals by cells located in or near the area postrema.

The potency of food stimuli as targets in aversion learning was examined using an interference paradigm. In the first study, foods interfered with liquid aversions but liquids did not interfere with food aversions. In the second study, food aversions were found to be resistant to interference by other foods. These findings suggest that foods are relatively potent targets in aversion conditioning in that they are resistant to interference by both foods and drinks. The final study examined the contribution of flavor intensity and nutrient density to the potency of aversion conditioning. The relative potency of foods over drinks may reflect differences in their intensity as well as their different roles in toxin avoidance and nutrient selection.

In order to determine whether some nutrients are more salient targets than others for the development of learned taste aversions a large group of normal human subjects were surveyed about their experiences with such aversions. The targets of these spontaneously occurring food aversions were identified and classified into general food categories. A prominent category for human aversions proved to be foods which were protein sources. This finding was similar to our observations in the clinical setting as well as those we have obtained using an animal model. It suggests that taste and/or postingestive properties of protein foods may make them particularly salient targets for food aversion learning.

A modified learned food aversion paradigm simulated the prolonged food and illness exposure likely to characterize nutrient deficiencies or tumor growth. In this paradigm, a continuously available food was associated with a slow, continuous infusion of LiCl provided by osmotic minipump. Significant aversions were acquired when the available diet was novel but not when it was familiar. Effects of drug infusions on daily food intake were found to parallel those on aversion formation. Marked, persistent suppression of food intake was seen in drug-treated animals consuming a novel diet, but only transient declines were seen in those with a familiar diet. By separating the direct from the conditioned effects on food intake of chronic drug infusions, these studies provided strong evidence that learned food aversions can lead to anorexia.

The effectiveness of a method for reducing the incidence of chemotherapy-induced learned food aversions was examined. Candy (coconut or rootbeer Lifesavers) was used as a scapegoat and given between the consumption of a meal and the administration of chemotherapy to determine whether this would lead to a greater willingness to consume items in that meal at a future test. This procedure produced evidence that the scapegoat had a significant protective effect: children were twice as likely to eat some portion of their test meal at the time of assessment if they had received the scapegoat at conditioning than when there was no intervention. Thus, the consumption of strongly flavored candies before chemotherapy appears to be a simple and effective way to reduce the impact of chemotherapy on preference for normal menu items.

Adult Fischer-344 (F-344) rats fail to prefer NaCl solutions of any concentration to water and, in fact, avoid NaCl at solution concentrations which other strains of rats reliably prefer. In contrast, immature rats of a number of strains have been reported to express an elevated salt preference relative to adults. The present studies examine the development of the F-344 rat's salt aversion by testing salt preference in the preweaning (Day 10) and weanling (Days 20-24) F-344 rat. At Day 10, F-344 rats ingested significantly more 1% and 2% NaCl solutions than water during "towel testing" and their preference was comparable to that displayed by other strains at this age. By the time of normal weaning (Days 20-24) significant avoidance of NaCl solutions in the isotonic range was displayed by F-344 rats, and their preference was strikingly different from other rat strains. This was evident in a 24-hr 2-bottle preference test as well as a 25 min "towel test." Since NaCl preference of F-344 rats is comparable to other rat strains at Day 10 but dramatically different by Day 20, a change must occur during this time which reverses the hedonic value of NaCl stimulation. Identification of factors responsible for this reversal may provide insight into the causes both of the F-344 rat's dislike of salt and the immature rats elevated salt preference.

Children with neoplastic diseases were offered an unusual ice cream before their drug treatments. Patients experiencing gastrointestinal toxicity due to the drugs were subsequently less likely to choose that ice cream again than controls. This suggests that taste aversions induced by drug-associated symptoms may contribute to the appetite loss experienced by cancer patients.

The studies included in this chapter examine the learned food aversions that develop as a result of cancer and cancer treatment. Clinical studies have shown that cancer patients can develop learned aversions to a novel ice cream flavor when it is consumed before drug treatments that produce nausea and vomiting. They also provided evidence that patients can acquire aversions to food in their usual diets when these foods are eaten before similar drug treatments. Observations in the clinic, supported by complementary studies with animal models, suggest that learned aversions are more likely to arise to protein foods than to other nutrient sources and that the presentation of a novel food in association with drug treatments may act as a "scapegoat" in blocking the development of aversions to foods in the normal diet. Laboratory studies using transplantable tumors in rats have shown that tumor growth can be associated with the development of strong aversions to the available diet. These aversions are specific to the diet eaten during tumor growth and they appear to play a causal role in the development of tumor-induced anorexia. The food aversions apparent in animals with certain experimental tumors point to physiological consequences of tumor growth that act as unconditioned stimuli in taste aversion conditioning. The identification of these changes and development of methods for correcting them are the current goals of our research in this area.

Subdiaphragmatic vagotomy produces hypophagia and weight loss in normal rats and can reverse the hyperphagia and obesity of rats with ventromedial hypothalamic lesions. Vagotomy surgery can also produce symptoms of nausea and discomfort. Since such symptoms are highly effective as unconditioned stimuli in food aversion conditioning, the present studies examined whether some of the depression in food intake observed in rats with vagotomy could be due to the development of aversions to the foods eaten after their surgery. In the first study, significant aversions developed to the specific novel diet consumed after vagotomy, results indicating that the symptoms associated with vagotomy can serve as effective unconditioned stimuli in the acquisition of learned food aversions. The second study compared vagotomized animals consuming familiar laboratory chow with those consuming a novel diet. In contrast to the novel diets, learned aversions did not develop to the familiar chow, and hypophagia was less persistent and severe. It is concluded that learned food aversions can contribute to the appetite and weight loss exhibited by vagotomized animals. Consideration of the conditions under which these aversions arise after vagotomy surgery may allow for the design of studies so as to minimize the aversions and thereby separate these nonspecific effects from direct regulatory deficits produced by vagotomy.

A learned food aversion paradigm was developed to investigate characteristics of aversion learning likely to be involved in nutrient selection. The development of aversions to a relatively familiar, complete food was examined in rats with continuous access to that food. Learned food aversions were observed which were specific to the particular diet present during a sequence of toxic drug injections. In subsequent studies, novel liquids and solids were introduced in association with drug treatment to examine their effectiveness as interference stimuli. Flavored solutions, which have previously been reported to interfere with conditioned taste aversions, had little or no effect on learned aversions to the diet. This lack of interference by novel flavored liquids was coupled with significant conditioned aversions to the novel liquids themselves. Novel flavored solutions were contrasted with novel foods, which were extremely effective as interference stimuli in this paradigm. Possible bases for the observed differences in interference effects of foods and drinks are examined.

Previous work has suggested that aversions to foods are relatively resistant to interference effects. Although these findings have been interpreted as reflecting the potency of foods as targets in aversion conditioning, resistance to interference could more generally characterize intense conditioned aversions. In the present studies with Wistar rats we examined whether interference with food aversion conditioning could be demonstrated by reducing either the drug dose (cyclophosphamide; 20 vs 40 mg/kg) or the number of conditioning trials. In both cases interference was produced; that is, the presence of a second food did interfere with the conditioning of aversions to the first or "target" food. Thus, under conditions which would be expected to produce slower or weaker conditioning, interference with food aversions was facilitated.

LTW(m) Leydig cell tumors were removed from Wistar Furth rats after the animals had developed anorexia and strong aversions to a novel diet consumed during tumor growth. Previous studies had reported recovery from anorexia within 24-48 hr of tumor excision but had not measured diet preference to find out if food aversions persisted beyond recovery. In the present study, which allowed the rats to choose freely between the aversive diet and a nonaversive diet for several weeks, aversions persisted from four to 25+ days. No signs of anorexia were evident during this testing phase, which followed 10 days of post-operative recuperation in which the rats were restricted to a diet of familiar laboratory chew. These results are consistent with the hypothesis that certain anorexigenic tumors, such as the LTW(m), can substantially affect food selection long after their direct effects on appetite have dissipated.