Effects of dietary and pharmacological manipulations on appetitive and consummatory aspects of feeding in non-human primates

Appetite 45 (2005) 110–120 www.elsevier.com/locate/appet

Research Report

Effects of dietary and pharmacological manipulations on appetitive and consummatory aspects of feeding in non-human primates Richard W. Foltin* Division on Substance Abuse, Department of Psychiatry, New York State Psychiatric Institute, College of Physicians and Surgeons of Columbia University, 1051 Riverside Drive, Unit 120, New York, NY 10032, USA Received 13 July 2004; revised 1 January 2005; accepted 11 March 2005

Abstract This study examined how pharmacological and behavioral manipulations affect appetitive and consummatory aspects of feeding of baboons. Baboons have access to food 24 h each day, but they must complete a two-phase operant procedure in order to eat. Responding on one lever during a 30-min appetitive phase was required before animals could start a consumption phase, i.e. a meal, where responding on another lever led to food delivery. Responding during the appetitive phase resulted in presentations of food-related stimuli only. Decreasing session length, increased appetitive behavior and increased meal size. Limiting the number of meals to a single 90 min meal each day but increasing the number of food pellets the animals received increased the size of meal, but did not increase appetitive behavior. These findings suggest that time since the previous meal has a greater effect on appetitive behavior than the size of the previous meal. Amphetamine (AMPH), which increases dopamine, decreased food intake at doses that did not affect appetitive behavior, indicating that appetitive and consummatory aspects of eating can be pharmacologically differentiated. Increasing how frequently animals could earn food-related stimuli in the appetitive phase and food in the consummatory phase increased both appetitive and consumatory behavior. Under these conditions, AMPH nearly doubled appetitive behavior at doses that decreased food intake by nearly 50 percent. When animals had one meal, of selfdetermined duration, meal size increased without affecting appetitive behavior, further demonstrating that appetitive behavior can be independent of the size of the previous meal and not predictive of the size of the subsequent meal. Under these conditions, AMPH decreased food intake at doses that did not affect appetitive behavior. In contrast, dexfenfluramine (DFEN), which increases serotonin, decreased both appetitive and consumatory behavior. Thus, it is possible to independently manipulate the appetitive and consummatory aspects of eating using both pharmacological and behavioral interventions indicating that it may be possible to develop medications that selectively affect appetitive or consummatory aspects of eating. q 2005 Elsevier Ltd. All rights reserved. Keywords: Motivation; Food intake; Deprivation; Amphetamine; Dexfenfluramine; Anorectic drugs; Model; Non-human primate

Introduction Sophisticated procedures have been developed for analyzing in both human and non-human animals the microstructure of a meal. Much is known about the physiological and behavioral variables affecting meal initiation, the pattern of eating within a meal, and meal termination (Barkeling, Rossner, & Sjoberg, 1995; DeCastro, 1990; Geliebter et al., 1992; Kissileff et al., 1996;

* Corresponding author. E-mail address: [email protected]

0195-6663/$ - see front matter q 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.appet.2005.03.011

Small, Jones-Gotman, & Dagher, 2003; Tolkamp, Allcroft, Austin, Nielsen, & Kyriazakis, 1998). These studies have emphasized the interaction between an organism consuming food and the consequences of that food consumption within a behavioral or physiological milieu. Similarly, the search for effective behavioral and pharmacological treatments for over-eating has focused on food consumption most often within the context of homeostasis. The possibility that it is more difficult to stop a behavior after it has started than to prevent a behavior from occurring, may play a role in high recidivism rates characteristic of appetitive disorders. A greater emphasis on the factors controlling the antecedents or appetitive aspects of eating behavior rather than on the consummatory aspects of eating may provide an useful avenue for improving treatment of eating disorders.

R.W. Foltin / Appetite 45 (2005) 110–120

As observed with drugs of abuse (see Wise, 1996 for a review), when rodents work for food there is increased dopamine (DA) release in the Nucleus Accumbens, as measured using microdialysis (e.g. Hernandez & Hoebel, 1988; Salamone, Cousins, McCullough, Carriero, & Berkowitz, 1994); increases that are larger in the presence of food deprivation (Radhakishun, van Ree, & Westerink, 1988). Thus, it appears that some central mechanisms overlap between eating behavior and drug taking. Because of the behavioral similarities between eating behavior and drug taking, and possible common central mechanisms mediating these two behaviors, the study of eating behavior can gain insight from the study of drug abuse. There is burgeoning evidence that antecedent stimulus conditions can determine behavioral output related to drug taking. For example, in laboratory animals, stimuli that have been paired with a drug of abuse can elicit respond for that drug and acquire conditioned reinforcing effects (e.g. Highfield, Mead, Grimm, Rocha, & Shaham, 2002; Woods & Winger, 2002). Thus, previously neutral stimuli that have been paired with primary reinforcement acquire motivational and reinforcing effects. As operationally defined by Bindra (1968), when such stimuli are presented prior to the initiation of instrumental responding and the probability of responding increases, then these stimuli have “incentive-motivational” properties. When such stimuli are presented upon completion of operant responding and the probability of responding increases then these stimuli have secondary or conditioned reinforcing effects. It is important to distinguish a stimulus with incentive-motivational value from one that is a Pavlovian classically conditioned stimulus. In the Pavlovian sense, the stimulus elicits, nearly without fail, a conditioned response (Mackintosh, 1974). By contrast, the behavioral response to a paired stimulus is modulated by current internal state as well as behavioral history. Thus, a hungry animal will respond for food when presented with a stimulus signaling that food’s availability, while a sated animal will not respond for more food when presented with a stimulus signaling that food’s availability. It is this flexibility in behavioral output following the presentation of an incentive-motivational stimulus that makes it a possible target for behavioral and pharmacological alteration (Bindra, 1978; Toates, 1981). In fact, it is possible to alter the behavioral effects of stimuli with neuropharmacological manipulations in laboratory rodents (e.g. Robinson & Berridge, 1993). Pharmacological agents that increase brain DA, such as AMPH, increase responding reinforced by conditioned reinforcers in rodents (e.g. Fletcher, 1995, 1996; Wyvell & Berridge, 2000), while pharmacological agents that increase brain serotonin (5-HT), such as DFEN decrease responding reinforced by conditioned reinforcers in rodents (Fletcher, 1995, 1996; Wilson, Costall, & Neill, 2000). Behavior that is reinforced by food-related stimuli, but not food itself, can be viewed as appetitive behavior, while behavior that is reinforced by food and food-related stimuli are consummatory behaviors. Thus, AMPH increases and DFEN decreases

111

appetitive behavior. Because drugs that increase DA are also effective in decreasing food intake (Leibowitz, 1978), it should be possible to see an increase in appetitive behavior, but a decrease in consummatory behavior in the same animals following AMPH administration. We have recently confirmed this possibility in non-human primates (Foltin, 2001). This dissociation supports the argument that the appetitive effects of food-related stimuli are mediated by different mechanisms than food consumption. We have developed a laboratory model that simulates appetitive and consummatory aspects of eating behavior based on the procedures developed by Collier and colleagues (e.g. Collier, 1983; Collier, Hirsch, & Kanarek, 1977). In our laboratory, baboons work for food by pulling on response levers, and food delivery is always accompanied by a visual signal, usually flashing lights. The food-related stimuli are conditioned reinforcers and baboons will work for the presentation of the stimuli only. Baboons have access to food 24 h each day, but they must complete a two-phase operant procedure in order to eat. Responding during the initial 30-min appetitive phase is reinforced by only the food-related stimuli, while responding during the latter consumption phase is reinforced with food and food-related stimuli. Every 10 responses completed during the appetitive phase is reinforced by the foodrelated stimuli (Kelleher, 1966). While the first response on one lever starts the appetitive phase, and the baboon needs to only make a minimum of 10 responses, with the last response occurring after 30 min in order to switch to the consumption phase, baboons respond more often than the minimum amount. It is possible to see a wide range of responding during the appetitive phase (Berridge & Robinson, 1998). All the responding during a single consumption phase comprises a single meal. While the appetitive phase must last about 30 min, the duration of the consumption phase is variable, as the phase continues until a baboon pauses eating for 10 min, i.e. meal duration and size can vary. In order to have another meal, the baboon must start another 30-min appetitive phase. Because this procedure allows the amount of responding during the appetitive and consumption phases of a meal to vary, it should be possible to determine the behavioral variables that might affect the two phases of a meal. Furthermore, as data indicate that it is possible to pharmacologically differentiate the two phases of a meal, it was also of interest to see if differences in baseline appetitive and consummatory behavior affect how pharmacological manipulations alter the two phases of a meal.

Methods Animals Eight adult male baboons (Papio cynocephalus anubis), weighing 26.3–31.8 (MeanZ28.9 kg) kg at the start of

112

R.W. Foltin / Appetite 45 (2005) 110–120

the study, were individually housed in standard non-human primate cages (0.94!1.21!1.52 m high) at The New York State Psychiatric Institute. Over approximately 1 year of these studies, the baboons’ weights increased by 2 kg to a mean weight of 30.9 kg (26.3–36.2 kg). The baboons had 7–16 years experience responding under FR schedules, and had participated in other studies on the effects of some of the same manipulations on response maintained under a similar operant schedule (Foltin, 2001, 2004). The room was illuminated with fluorescent lighting from 7:00 AM to 7:00 PM daily. The only food that baboons received daily, other than pellets delivered during the session, were two chewable vitamins (‘Kiddy Chews,’ Schein Pharmaceutical, Inc., Port Washington, NY), two pieces of fresh fruit (80–100 kcal, each), and a dog biscuit (150 kcal, Old Mother Hubbard, Inc., Lowell, MA). Water was available ad libitum from a spout located at the back of each cage. All aspects of animal maintenance and experimental procedures complied with the US National Institutes of Health Guide for Care and Use of Laboratory Animals, and were approved by the New York State Psychiatric Institute Animal Care and Use Committee. Apparatus A response panel holding, from bottom to top, a food hopper, two Lindsley levers spaced 0.30 m apart (Gerbrands, Arlington, MA), four stimulus lights (two above each lever), and a pellet dispenser (BRS-LVE model PDC-005, Beltsville, MD) was attached to the front of each cage. All schedule contingencies were programmed using Pascal on Macintosh (Cupertino, CA) computers located, along with the interface, in an adjacent room. Schedule of reinforcement Responding under each phase of a two-phase chain schedule of reinforcement was on a separate response manipulandum. The session began with the illumination of a single light above the appetitive lever. The first response on the appetitive lever began a 30-min timer and illuminated a second light over the appetitive lever, i.e. the 30-min appetitive phase was indicated by the illumination of two lights above the appetitive lever. The appetitive phase was a FI 30 min schedule, with a FR 10 second-order phase [FI 30 0 (FR 10:S)]. Thus, after every 10th response during the FI phase, the stimuli associated with reinforcer delivery during the second phase were presented. There was a 10 min limited hold for the appetitive phase, such that after the expiry of the 30 min FI, the next FR 10 had to be completed within 10 min. Failure to complete a FR 10 within 10 min canceled that appetitive phase, and extinguished one light over the appetitive lever such that only a single light was illuminated over the appetitive lever. The baboon received no indication that the 30-min interval had elapsed. The first FR 10 completed after 30 min resulted in the two lights

above the left lever being extinguished and a single light above the right lever being illuminated, signaling the availability of food under the FR consumption phase of the chain schedule. The consumption phase of the chain schedule was maintained under a FR 10 schedule of food reinforcement (1 grain-based “dustless” banana-flavored 1g food pellet; 3.34 kcal/g: 20.1% protein, 3.3% fat, 55.3% carbohydrate, 3.3% ash,!5% moisture and 4.0% fiber; Bio-serv, Frenchtown, NJ). After a 10-min interval in which no responses occurred, the consumption phase terminated, i.e. meal-size was determined by each baboon. The single light above the right consumption lever was then extinguished, and the single light above the left appetitive lever was again illuminated. In order to initiate another meal, the baboon was required to start another 30-min appetitive phase by pulling on the left lever. Responding was studied using two types of paired stimuli: flashing lights and 1-s time-outs. Under the flashing-light condition, food pellets were accompanied by the flashing over a 12 s interval (1 s on: 1 s off), of all four stimulus lights above both levers, and an additional 18 s of darkness, when all stimulus lights were extinguished. Under the 1-s time-out condition, food pellets were accompanied by only 1 s of darkness, when all stimulus lights were extinguished. These stimuli were also presented after every 10th response during FI-appetitive phases. Daily sessions began each day, seven days a week, at 0900. Procedure Session length Six baboons participated in a study on the effects of session length on appetitive and consumatory behavior. Flashing lights were paired with food delivery and presented following every 10 responses during appetitive phases. Baboons had been stabilized on this feeding regime with 24 h/day access to food for several years prior to the start of this study. The length of the daily session was decreased to 18, 12, 8, 6, 4, 2 h and 90 min consecutively. Each duration was in effect for a minimum of 6 days or until responding was stable (no upward or downward trends in pellet consumption for 3 consecutive days). The total number of light flashes during appetitive phases and total number of food pellets (with light flashes), number of meals (consumption phases), the latency to the first pellet delivery of the first meal (including the time required to complete the first meal appetitive phase), and the running rate (responses/s timed from the first response after reinforcer delivery to the next reinforcer delivery) during the first appetitive and consumption phases of each session were calculated. Data were summarized using 2-factor repeated-measures analyses of variance (ANOVA): the first factor was session length, and the second factor was day (last 3 days under each length condition). There were 7 planned comparisons: data obtained under each length were compared to data obtained under the 24-h access

R.W. Foltin / Appetite 45 (2005) 110–120

condition. The error terms used in these contrasts were derived from the Length by Day interaction. Results were considered significantly different at P!0.05, using HuynhFeldt corrections Effect of number of pellets per food delivery under 2-h access conditions Eight baboons participated in a study that increased the number of pellets earned per each food delivery: 1, 5 or 10 pellets per delivery when baboons had access to food 2 h each day. Flashing lights accompanied food delivery and were presented following every 10 responses during appetitive phases. Each reinforcer magnitude was in effect for a minimum of 8 days or until responding was stable. Data were summarized using 2-factor repeated-measures ANOVA: the first factor was reinforcer magnitude, and the second factor was day (last 5 days under each reinforcer condition). There were two planned comparisons: data obtained under the 5-pellet and 10-pellet conditions were each compared to data obtained under the 1-pellet condition. The error term used in these contrasts was derived from the Magnitude by Day interaction. D-Amphetamine under 2-h access conditions with different stimuli paired with food Seven baboons participated in a study assessing the effects of AMPH on appetitive and consummatory behavior under two food-stimulus conditions when baboons had access to food 2 h each day. Initially, responding during the appetitive phase was reinforced by flashing lights that also accompanied food delivery (1 pellet), as used in the previous studies. After collecting data on the effects of AMPH, responding during the appetitive phase was reinforced by 1 s of darkness that also accompanied food delivery. After responding again stabilized, a dose–response function was then determine for AMPH. D -Amphetamine sulfate (0.06-0.50 mg/kg, Sigma Chemical Corp., St Louis, MO), was given intramuscularly (i.m.) in a thigh muscle (location varying among sessions) on Tuesday and Friday of each week at 0900, with placebo injections given occasionally on other days of the week. Drug doses are expressed as total weight of the salt or base. A complete dose–response function was determined in 2–3 weeks, and the dose–response function was determined twice under each paired-stimulus condition. Doses were administered only when response of the two previous days were stable. Dose order systematically varied within and between baboons such that all possible dosing orders were tested. Data were summarized using 3-factor repeated-measures ANOVA: the first factor was test condition (first vs. second), the second factor was drug condition (placebo vs. active; there was one placebo session for each active dose session), and the third factor was dose (four doses). There were 4 planned comparisons: each of the active drug doses was compared to the placebo doses. The error term used in these

113

contrasts was derived from the Drug by Dose interaction. Because of the large difference in the number of appetitive and consumption reinforcers earned under placebo conditions for both stimulus conditions, the data were not statistically compared between stimulus conditions. D-Amphetamine and dexfenfluramine with 1-s time-out under single-meal conditions Eight baboons participated in a study assessing the effects of AMPH and DFEN (dexfenfluramine hydrochloride; 0.12-1.0 mg/kg, Sigma Chemical Corp.) on appetitive and consummatory behavior when animals had a single meal of self-determined size each day with the 1-s time-out accompanying food delivery. Upon completion of the previous study, the restriction on session length was replaced by a restriction on the number of meals such that only one meal was available each session: the first meal continued until 10 min had elapsed without a response. Once responding stabilized under these conditions, a single dose–response function for AMPH and then a single dose– response function for DFEN was determined as described above. Data analyses were also as described above except there was no test factor in the ANOVAs.

Results Session length Under the 24-h access condition, baboons had about 3 meals (consumption phases), earned about 45 light flashes, i.e. baboons responded about 450 times during appetitive phases, and ate about 300 food pellets, i.e. baboons responded about 3000 times. As shown in the top left panel of Fig. 1, the total daily number of light flashes was only decreased under the 4-h session condition [F(1, 14)Z 7.6, P!0.03] compared to the 24-h access condition. Because baboons consumed only one meal per day when the session length was 4 h or less, responding during the first appetitive phase provided the best estimate of appetitive behavior. Baboons earned about 10 stimulus presentations during the first appetitive phase under the 24-h access condition. Decreasing the session length significantly increased the number of light flashes during the first appetitive phase [F(1,14)OZ9.1, P!0.003, for all contrasts]. Baboons earned about 35 light flashes before the first meal during both the 2-h and 90-min sessions, i.e. appetitive responding increased by about 350 percent. As shown in the top right panel of Fig. 1, decreasing the session length to 18 h, 12 h or 8 h significantly increased up to 130 percent the total number of food pellets consumed [F(1,14)OZ37.5, P!0.0001, for all contrasts]. By contrast, decreasing the session length to 6 h or less significantly decreased the total number of food pellets consumed to less than one third of that observed under the 24-h access condition [F(1,14)OZ6.6, P!0.02, for all contrasts].

114

R.W. Foltin / Appetite 45 (2005) 110–120

Appetitive Reinforcers

Consumption Reinforcers

60

450

50

375

§

Number

40 § 30

* * *

20

*

10

*

*

*

Total Session

Number of Food Pellets

§

§ §

300

§ 225 150

*

*

*

*

*

§ §

Total Session

75

First Appetitive Phase

First Appetitive Phase 0

0 24 hr

12 hr 18 hr

6 hr 8 hr

24 hr

2 hr 4 hr

90 min

12 hr 18 hr

Meals

2 hr 4 hr

90 min

Response Rate

3.0

4

6 hr 8 hr

§

First Appetitive Phase 2.5

Number

§ §

2

§ §

§

§

responses/s

§

3

§

First Consumption Phase

2.0 1.5 1.0

1

*

0.5

Total Session

0.0

0 24 hr

12 hr 18 hr

6 hr 8 hr

24 hr

2 hr 4 hr

90 min

12 hr 18 hr

6 hr 8 hr

2 hr 4 hr

90 min

Session Duration Fig. 1. Top Left Panel: Mean total daily number of stimulus presentations delivered during appetitive phases of the entire session and delivered during the first appetitive phase of the session as a function of session duration. An § indicates a data point that differs from the 24 h total session data, and an * indicates a data point that differs from the 24 h first appetitive phase data (P!0.05). Top Right Panel: Mean total daily number of food pellets delivered during the entire session and delivered during the first consumption phase (meal) of the session as a function of session duration. An § indicates a data point that differs from the 24 h total session data, and an * indicates a data point that differs from the 24 h first meal data (P!0.05). Bottom Left Panel: Mean total daily number of meals as a function of session duration. An § indicates a data point that differs from the 24 h total session data (P!0.05). Bottom Right Panel: Mean response rate during the first appetitive phase and meal of the session as a function of session duration. An § indicates a data point that differs from the 24 h first meal data, and an § indicates a data point that differs from the 24 h first appetitive phase data (P!0.05). Error bars represent 1 SEM.

Under 24-h access conditions, baboons ate about 125 food pellets during the first meal of the session. Decreasing the session length to 18, 12, 8, 6 or 4 h significantly increased up to 170 percent the number of food pellets consumed during the first meal [F(1,14)OZ6.6, P!0.02, for all contrasts]. Decreasing the session length significantly decreased the number of meals [bottom left panel of Fig. 1; F(1,14)OZ 10.1, P!0.006, for all contrasts] with baboons having one meal per day when the session length was 4 h or less. The fact that baboons earned 10 light flashes before the first meal when they had 24 h access to food and about 35 light flashes before the first meal when they had 2 h access to food, indicates that decreasing session length increased appetitive behavior.

Mean response rate during the first appetitive phase was about 0.5 responses/s, while the mean response rate during the first meal was about 1.3 responses/s. As shown in the bottom right panel of Fig. 1, decreasing the session length to 2 h significantly increased the rate of responding during the first meal [F(1,14)OZ25.7, P!0.0001, for both contrasts], while decreasing the session length to 90 min significantly increased the rate of responding during the first appetitive phase [F(1,14)OZ7.8, P!0.03]. Baboons began the first meal of the session about 150 min after the start of the session under the 24 h access condition (data not shown). Decreasing the session length significantly decreased the latency to the first meal to 30–40 min [F(1,14)OZ93.9, P!0.0001, for all contrasts].

R.W. Foltin / Appetite 45 (2005) 110–120

115

Effect of number of pellets per food delivery

D-Amphetamine under 2-h access conditions with different stimuli paired with food Fig. 2 compares the effects of AMPH on feeding behavior under both stimulus conditions when baboons had 2-h access to food each day. Under baseline conditions when flashing lights accompanied food, baboons earned about 25 reinforcers during the appetitive phase (top panel of Fig. 2), while under baseline conditions when a 1-s timeout accompanied food, baboons earned about 75 reinforcers during the appetitive phase (there was one appetitive and consumption phase per session). This increase was most likely due to the fact that the time-out, when responding had no consequences, after stimulus presentations was 30 s under the flashing light condition, compared to 1 s under the other condition, which provided the baboons more time to respond. 0.50 mg/kg AMPH significantly decreased appetitive behavior under the flashing-light condition [F(1,18)Z 13.8, P!0.001]. By contrast, the only significant effect of AMPH on appetitive behavior under the 1-s time-out condition was to increase the number of reinforcers following the 0.12 and 0.25 mg/kg doses [F(1,18)OZ4.8, P!0.04, for both contrasts]. Under baseline conditions when flashing lights accompanied food, baboons earned about 150 food pellets (bottom panel of Fig. 2), while under baseline conditions when a 1-s time-out accompanied food, baboons earned about 220 food pellets. Under both stimulus conditions,

200

Flashing Lights

§ §

1-s Time out

Number

150 100 50

*

0 0

0.06

0.12

0.25

0.50

Consumption Reinforcers 250 Number of Food Pellets

When baboons received 1 food pellet per delivery and had only 2-h access to food each day they earned an average of 29.9G2.1 light flashes and 148.3G2.1 food deliveries in a 2-h session, i.e. during the first and only meal of that session. Increasing the number of food pellets per delivery to 5 did not affect appetitive behavior, but decreased the number of food deliveries to 48.0G1.9 [F(1,42)Z1017.5, P!0.001]. Because 5 pellets were delivered, the total daily pellet intake increased to 239.8G9.3 [F(1,42)Z9.8, P!0.02]. Similarly, increasing the number of food pellets per delivery to 10 did not affect appetitive behavior, but decreased the number of food deliveries to 51.1G1.9 [F(1,42)Z956.2, P!0.001], such that total daily pellet intake increased to 510.6G25.8 [F(1,42)Z153.2, P!0.001]. Increasing the number of food pellets per delivery also increased the latency to the daily meal from 37.7G1.9 min with 1 pellet to 42.2G2.3 min with 5 pellets (n.s.) to 45.5G2.1 min with 10 pellets [F(1,42)Z5.6, P!0.03]. Finally, increasing the number of food pellets per delivery also decreased the rate of responding during the daily meal from 2.60G0.16 responses/s with 1 pellet to 1.56G0.09 responses/s with 5 pellets [F(1,42)Z20.4, P!0.0005] to 0.56G0.08 responses/s with 10 pellets [F(1,42)Z82.3, P!0.0001]. Thus, increasing the number of pellets earned per delivery during a 2-h session, increased food consumption without affecting appetitive behavior.

Appetitive Reinforcers

200

§

150 100

*

* *

50 §

0 0

0.06

0.12

0.25

§ 0.50

Amphetamine Dose (mg/kg) Fig. 2. Mean total daily number of stimulus presentations delivered during the appetitive phase (Top Panel) and food pellets delivered during the consumption phase (Bottom Panel) of the session as a function of type of paired stimulus and AMPH dose. An § indicates a data point that differs from placebo data under the 1-s time-out condition, and an * indicates a data point that differs from placebo data under the flashing-light condition (P!0.05). Error bars represent 1 SEM.

doses of 0.12 mg/kg or larger of AMPH decreased the number of food pellets to similar levels [flashing-light condition: F(1,18)OZ7.6, P!0.01, for all contrasts; 1-s time-out condition: F(1,18)OZ7.8, P!0.01, for all contrasts]. 0.50 mg/kg AMPH significantly increased the latency to the only meal under the flashing-light condition [F(1,18)Z62.7, P!0.0001], while doses of 0.12 mg/kg or larger of AMPH increased the latency to the only meal under the 1-s time-out condition [F(1,18)OZ21.1, P!0.0002, for all contrasts]. Because some animals did not respond following the larger AMPH doses, it was not possible to analyze response rate. D-Amphetamine and dexfenfluramine with 1-s time-out under single-meal conditions Fig. 3 compares the effects of AMPH and DFEN on feeding behavior when baboons had a single appetitive phase and meal each day and a 1-s time-out accompanied food. Under baseline conditions, baboons earned 75–110 reinforcers during the appetitive phase. Only the largest AMPH dose significantly decreased appetitive behavior [F(1,21)Z5.8, P!0.02], while the 3 largest DFEN doses all decreased appetitive behavior [F(1,21)OZ5.1, P!0.04,

116

R.W. Foltin / Appetite 45 (2005) 110–120

Appetitive Reinforcers

Consumption Reinforcers

160

600 Number of Food Pellets

Amphetamine Dexfenfluramine

Number

120

§

80

40

§ §

500 §

400

§ 300 200

* 100

* *

*

0

*

0

0

0.06

0.12

0.25

0.50

AMPH

0

0.06

0.12

0.25

0.50

AMPH

0

0.12

0.25

0.50

1.00

DFEN

0

0.12

0.25

0.50

1.00

DFEN

Latency to Only Meal

Response Rate

375

2.5

* 300

2.0

min

responses/s

*

225 150

§

*

75

§

1.5 1.0 0.5

* 0

0.0 0

0.06

0.12

0.25

0.50

AMPH

0

0.06

0.12

0.25

0.50

AMPH

0

0.12

0.25

0.50

1.00

DFEN

0

0.12

0.25

0.50

1.00

DFEN

Dose (mg/kg) Fig. 3. Mean total daily number of stimulus presentations delivered during the appetitive phase (Top Left Panel) and food pellets delivered during the consumption phase (Top Right Panel) of the session, latency to the first meal of the session (Bottom Left Panel), and response rate during the first meal and appetitive phase (Bottom Right Panel) as a function of drug and drug dose. An § indicates a data point that differs from DFEN placebo data, and an * indicates a data point that differs from AMPH placebo data (P!0.05). Error bars represent 1 SEM.

for all contrasts]. Under baseline conditions, baboons earned about 425 food pellets during the only meal. All AMPH doses significantly decreased the number of food pellets [F(1,21)Z20.6, P!0.0002, for all contrasts], while only the 2 largest DFEN doses decreased the number of food pellets [F(1,21)OZ17.3, P!0.0009, for both contrasts]. Thus, most AMPH doses decreased food intake without affecting appetitive behavior, while most DFEN doses decreased food intake and appetitive behavior. Baboons began the only meal of the session about 35 min after the start of the session. The three largest AMPH doses increased the latency to the first meal [F(1,21)OZ7.1, P!0.02, for all contrasts], while only the largest DFEN dose increased the latency to the first meal [F(1,21)Z6.9, P!0.05]. The mean response rate during the only meal of the day was about 1.5 responses/s. Only the largest AMPH [F(1,21)Z7.3, P!0.02] and DFEN [F(1,21)Z4.5, P!0.05] doses decreased response rate during the only meal [the increase in response rate following 0.12 mg/kg AMPH was most

likely spurious]. Neither drug affected response rate during the appetitive phase.

Discussion The results of the present series of experiments clearly indicate that the appetitive and consummatory phases of eating can be behaviorally and pharmacologically differentiated. Three behavioral variables, one paired-stimulus variable and two pharmacological variables were all shown to affect the appetitive and consummatory phases of eating resulting in a three-way interaction. The first behavioral variable to be manipulated was the length of the daily session. After animals were accustomed to 24-h/day access to food, decreasing the session length to between 18 h and 8 h/day increased both food appetitive and consummatory phases of eating. Limiting access to food decreased the number of meals and increased meal size.

R.W. Foltin / Appetite 45 (2005) 110–120

This finding is similar to the well known phenomena that increasing the effort to obtain a meal decreases the daily number of meals and increases meal size in laboratory animals (Collier, 1983, 1985; Foltin & Fischman, 1988). It is interesting to note that food intake increased by nearly 50 percent when the session was shortened by 6 h, i.e. 18 h/day condition. The 6 h of feeding that were lost occurred from 0300 to 0900. This precluded the animals from having their usual first meal of the day, which occurred shortly after the room lights were illuminated at 0700 each day. Finally, reducing the session length to 4 h increased appetitive behavior, but decreased food intake: food intake decreased because the brief session length precluded the animals from responding for as many pellets as they ate under the 24-h access condition. Anthropomorphically speaking, depriving animals of breakfast increased food consumption during the rest of the 8-h day. While breakfast has often been viewed as an important meal because performance is improved after eating in the morning (see Kanarek, 1997 for review), other data in humans indicate that breakfast plays a role in caloric regulation (Lluch, Hubert, King, & Blundell, 2000). Consumption of a low-calorie breakfast or no breakfast at all usually leads to much greater food consumption at lunch or later in the day (Hubert, King, & Blundell, 1998). The present results in non-human primates support a key role for the first meal of the day in caloric homeostasis. This anthropomorphic conclusion should be tempered by that fact that we did not prevent the animals from having a different meal, such as the first meal after the session started at 0900, as a comparison condition. The clearest indicator of a change in appetitive behavior is the finding that has the session length decreased baboons pressed the lever more, earning more light flashes, during the first appetitive phase of the day: increases in appetitive behavior were linearly related to decreases in session length. A second measure of appetitive behavior is the latency to the first meal of the session. Decreasing the session length decreased the latency to the first meal. Decreasing access to food and amount eaten is well-known to decrease the latency to initiate responding for food or eating during a session in laboratory animals (e.g. Collier & Levitsky, 1968; Foltin, 2001) and humans (Johnstone et al., 2002). The largest increase in appetitive behavior was observed when the session length precluded the baboons from eating as much food as they ate when food was continuously available. Thus, decreasing the session length increased the motivation to eat as shown by an increase in appetitive behavior. The second behavioral variable to be manipulated was the number of food pellets per delivery. Increasing the number of food pellets from 1 to 5 or 10 increased the total number of pellets consumed in the single meal of the day, but decreased the number of responses made for food during the 2 h session. While this appears parodoxical, animals responded less because less work was required to earn

117

multiple pellets. In spite of the increase in meal size, there was no effect of reinforcer magnitude on appetitive behavior. This provides a clear demonstration of a disconnection between appetitive and consummatory behavior. When session length was decreased, the increase in appetitive behavior was predictive of an increase in meal size. But when the meal-size was increased by providing more pellets per delivery, there was no corresponding increase in appetitive behavior. This difference was most likely due to the fact that although animals ate more food during the 2-h session when more pellets were delivered, they were still 22-h of food deprivation between sessions. The disconnect between size of the previous meal and appetitive behavior before the next meal suggests that appetitive behavior is related more to the time since the previous meal, than the size of the previous meal. Furthermore, in contrast to the finding when session length was decreased, the amount of appetitive behavior did not predict the size of the following meal. Assuming that ratings of appetite are a measure of appetitive behavior in humans, the existing data indicates that appetitive behavior measures also do not predict subsequent food intake (e.g. Doucet, St-Pierre, Almeras, & Tremblay, 2003). As observed here, in laboratory rodents, increasing the number of food pellets per delivery reliably decreases total responding, but increases food intake (Allison, Miller, & Wozny, 1979). Increasing reinforcer magnitude by increasing the number of pellets per delivery decreased the rate of response during consumption phases replicating previous findings in rodents (Collier, Johnson, Hill, & Kaufman, 1986) and baboons (Foltin, 1994) who earned all of their daily food intake during laboratory sessions. This indicates that response rate during a meal does not correlate with meal size. The third behavioral variable to be manipulated was the size of the single meal each day. Because the failure to see changes in appetitive behavior when the number of food pellets per delivery was varied may have been due to the fact the baboons were always 22 h food deprived, we then allowed the baboons to eat a single meal, with no limit on size, each day, rather than being limited to a 2 h session. When animals were permitted to eat an unlimited amount of food in one meal, food intake nearly doubled. In spite of the doubling in responding during the single consumption phase, there was no change in responding during the single appetitive phase. These findings again demonstrate that size of the previous meal is not a critical variable in determining appetitive behavior before the next meal. Unfortunately, the baboons ate quickly so that the only meal of the day ended 2 1/2–3 h after the start of the session, such that baboons were still 21–22 h food deprived each day. The finding that eating a larger meal without changing duration of deprivation failed to affect appetitive behavior, further suggest that appetitive behavior is related more to the time since the previous meal, than the size of the previous meal.

118

R.W. Foltin / Appetite 45 (2005) 110–120

Again, the amount of appetitive behavior did not predict subsequent meal size. The first pharmacological manipulation examined the effects of AMPH when the sessions were 2 h long and either flashing lights or a 1-s time-out accompanied food delivery. Decreasing the duration of the stimulus from 30 s (flashing lights) to 1 s (1 s of darkness) increased both appetitive behavior by 200 percent and food intake by 50 percent. Under the flashing-light condition, only the largest dose of AMPH decreased appetitive behavior, while the three largest doses decreased food intake, indicating that consummatory behavior was more sensitive to the effects of AMPH than appetitive behavior. By contrast, under the 1-s condition the intermediate doses of AMPH increased appetitive behavior by 100 percent and decreased food intake by up 50 percent; the largest AMPH dose decreased food intake without affecting appetitive behavior. Peripherally administered AMPH can produce a wide range of behavioral effects in humans and laboratory animals. Studies using administration of AMPH into discrete brain regions have localized some of the behavioral effects relevant to eating behavior. AMPH infused into Nucleus Accumbens has been shown to increase responding reinforced by stimuli that have been paired with a primary reinforcer, as observed here (Whishaw & Kornelsen, 1993), while AMPH infused into the lateral hypothalamus has been shown to decrease food intake (Leibowitz, 1978), as observed here. Thus, under certain conditions, peripherally-administered AMPH produces both increases and decreases in behavior in the same animal in the same session (Cohen & Branch, 1991; Foltin & Evans, 1999; Kornblith & Hoebel, 1976). For example, we have shown that AMPH decreased food intake of rhesus monkeys, but simultaneously increased the amount of time that monkeys spent in a location that was paired with food (Evans & Foltin, 1997). One procedure commonly used with laboratory rodents to demonstrate the conditioned effects of stimuli paired with primary reinforcement is to train the animals to associate the stimulus cues with primary reinforcement, then allow the animals to respond in an operant chamber to receive only the paired cues. AMPH, when given under extinction conditions (no primary reinforcer present), increases responding that results in the delivery of the paired cues only (Fletcher, 1995). Observations of this type have led to the general conclusion that drugs that increase DA enhance the conditioned reinforcing effects of stimuli paired with primary reinforcement (e.g. Fletcher, 1996; Wyvell & Berridge, 2000), and enhance the motivational value of stimuli associated with a primary reinforcer (Robinson & Berridge, 1993). Within this framework, the dramatic increase in appetitive behavior under the 1-s time-out condition would be viewed as an enhancement of the motivational value of the stimuli due to increased DA following AMPH administration. One difficulty with this interpretation is the fact that similar increases were not

observed under the flashing-light condition. It is possible that there was a ceiling effect under that condition. When given placebo, baboons earned about 25 stimulus presentations during the appetitive phase under the flashing-light condition. As shown in Fig. 1, baboons can earn up to 40 stimulus presentations during a single appetitive phase under those conditions, indicating that responding was not at ceiling levels. We have recently completed a study that examined the effects of AMPH under these two stimulus conditions, but in the absence of food restrictions, i.e. 24-h session (Foltin, 2004). Under free-feeding conditions, AMPH increased food appetitive under the flashing-light condition, but not the 1-s time-out condition. Comparing the present results with those of the previous study indicates that food restriction interacts with AMPH and the pairedstimulus conditions to determine the effects of AMPH on appetitive behavior. The results of the second pharmacological manipulation support this assumption. Because some animals did not respond at all after receiving the larger AMPH doses, it was decided to examine the effects of AMPH when baboons had access to a single large meal each day. In this way, druginduced pauses in initiating food appetitive would not limit food consumption. AMPH (1) produced long dose-dependent increases in the latency to the first meal; (2) did not increase appetitive behavior; and (3) once again decreased food intake at doses that did not affect appetitive behavior. As previously reported, the AMPH dose–response function for food intake was shifted to the left of the dose–response function for appetitive behavior, indicating that intake was decreased while appetitive behavior was either increased or unaffected (Foltin, 2001, 2004). Several previous studies indicate that the behavioral effects of AMPH interact with food deprivation, and that this interaction varies across measures (e.g. Cole, 1979; Cole & Gay, 1971). In humans, 24 h food deprivation did not alter the subjective effects of AMPH, e.g. drug liking, but participants were more likely to identify the drug as a stimulant in the fasted state (Zacny & de Wit, 1989). In laboratory rats, increasing food deprivation increased the effect of AMPH on rearing, but not on locomotor activity (Cole, 1980). Bassareo and Di Chiara (1999) demonstrated that food deprivation modulated the release of DA in the nucleus accumbens in response to appetitive stimuli, suggesting that changes in DA may also play a role in the increase in food appetitive following AMPH administration under deprivation conditions. Finally, because food deprivation is a stressor and stress can increase motivation (Shaham, Shalev, Lu, De Wit, & Stewart, 2003), it is possible that stress may be involved in the interaction between AMPH, deprivation and type of paired stimuli. In contrast to AMPH, DFEN decreased both appetitive behavior and food intake and produced only modest increases in the latency to the meal, replicating data obtained in baboons with 24 h access to food (Foltin, 2001). Although several studies have reported that drugs

R.W. Foltin / Appetite 45 (2005) 110–120

that primarily increase 5HT decrease responding maintained by conditioned reinforcers (Fletcher, 1995, 1996), it is unclear if the decrease in appetitive behavior is due to that effect or a decrease in motivation to eat (Chaki & Nakazato, 2001; Hensrud, 2000). Regardless, DFEN does not produce the dissociation, as produced by AMPH, between appetitive behavior and food intake. Collier and his colleagues, in a long-term series of studies, convincingly argued that eating behavior is too complex to be accounted for solely by reference to homeostatic mechanisms (Collier, 1983, 1985). More recently, Berridge and Robinson (2003) suggested, also based on a long series of studies, that an understanding of reinforced behavior requires integration of three behavioral processes. An individual must learn the relationships between behavior and outcome. Outcomes must produce a change in affect. And, motivational variables must modulate these relationships. Consumption of a commodity is affected by these three processes and each of the processes is multiply determined. The present study clearly demonstrates that the appetitive and consummatory aspects of eating behavior can be behaviorally and pharmacologically differentiated. The complex interaction among the behavioral and pharmacological manipulations used here clearly indicate that eating behavior is far from understood. It would be important to know how experience and deprivation modulate these responses. A better understanding of how neuropharmacological and behavioral manipulations affect the appetitive and consummatory aspects of eating behavior will be an asset in the development of drugs and behavioral therapies that affect eating behavior.

Acknowledgements This research was supported by DA-04130 from The National Institute on Drug Abuse, and approved by the New York State Psychiatric Institute Animal Care and Use Committee. The assistance of Julian Perez, Angel Ramirez, and Drs Suzette Evans, Margaret Haney and Mohamed Osman is gratefully acknowledged.

References Allison, J., Miller, M., & Wozny, M. (1979). Conservation in behavior. Journal of Experimental Psychology: General, 108, 4–34. Barkeling, B., Rossner, S., & Sjoberg, A. (1995). Methodological studies on single meal food intake characteristics in normal weight and obese men and women. International Journal of Obesity, 19, 284–290. Bassareo, V., & Di Chiara, G. (1999). Differential responsiveness of dopamine transmission to food-stimuli in nucleus accumbens shell/core compartments. Neuroscience, 89(3), 637–641. Berridge, K. C., & Robinson, T. E. (1998). What is the role of dopamine in reward: hedonic impact, reward learning, or incentive salience? Brain Research. Brain Research Reviews, 28, 309–369. Berridge, K. C., & Robinson, T. E. (2003). Parsing reward. Trends in Neurosciences, 26, 507–513.

119

Bindra, D. (1968). Neuropsychological interpretation of the effects of drive and incentive-motivation on general activity and instrumental behavior. Psychological Review, 75, 1–11. Bindra, D. (1978). How adaptive behavior is produced: A perceptualmotivational alternative to response-reinforcement. The Behavioral and Brain Sciences, 1, 41–91. Chaki, S., & Nakazato, A. (2001). Recent advances in feeding suppressing agents: Potential therapeutic strategy for the treatment of obesity. Expert Opinions in Therapuetic Patents, 11(11), 1677–1692. Cohen, S. L., & Branch, M. N. (1991). Food-paired stimuli as conditioned reinforcers: Effects of D-amphetamine. Journal of the Experimental Analysis of Behavior, 56, 277–288. Cole, S. O. (1979). Interaction of food deprivation with different measures of amphetamine effects. Pharmacology Biochemistry and Behavior, 10, 235–238. Cole, S. O. (1980). Deprivation-dependent effects of amphetamine on concurrent measures of feeding and activity. Pharmacology Biochemistry and Behavior, 12, 723–727. Cole, S. O., & Gay, P. E. (1971). Interaction of amphetamine and food deprivation on a food-motivated operant. Communications in Behavioral Biology, 6, 345–347. Collier, G., Hirsch, E., & Kanarek, R. (1977). The operant revisited. In W. K. Honig, & J. E. R. Staddon (Eds.), Handbook of operant behavior. Englewood Cliffs, NJ: Prentice Hall. Collier, G., & Levitsky, D. A. (1968). Operant running as a function of deprivation and effort. Comparative Physiology and Psychology, 66(2), 522–523. Collier, G. H. (1983). Life in a closed economy: The ecology of learning and motivation. In M. D. Zeiler, & P. Harzem (Eds.), Advances in analysis of behaviour. New York: Wiley. Collier, G. H. (1985). Satiety: An ecological perspective. Brain Research Bulletin, 14, 693–700. Collier, G. H., Johnson, D. F., Hill, W. L., & Kaufman, L. W. (1986). The economics of the law of effect. Journal of the Experimental Analysis of Behavior, 46, 113–136. Doucet, E., St-Pierre, S., Almeras, N., & Tremblay, A. (2003). Relation between appetite ratings before and after a standard meal and estimates of daily energy intake in obese and reduced obese individuals. Appetite, 40, 137–143. Evans, S. M., & Foltin, R. W. (1997). The effects of D-amphetamine on the reinforcing effects of food and fluid using a novel procedure combining self-administration and location preference. Behavioural Pharmacology, 8, 429–441. Fletcher, P. J. (1995). Effects of d-fenfluramine and metergoline on responding for conditioned reward and the response potentiating effect of nucleus accumbens D-amphetamine. Psychopharmacology, 118, 155–163. Fletcher, P. J. (1996). Injection of 5-HT into the nucleus accumbens reduces the effects of D-amphetamine on responding for conditioned reward. Psychopharmacology, 126, 62–69. Foltin, R. W. (1994). Does package size matter? A unit-price analysis of ‘demand’ for food in baboons. Journal of the Experimental Analysis of Behavior, 62, 293–306. Foltin, R. W. (2001). Effects of amphetamine, dexfenfluramine, diazepam, and other pharmacological and dietary manipulations on food ‘seeking’ and ‘taking’ behavior in non-human primates. Psychopharmacology, 158(1), 28–38. Foltin, R. W. (2004). Effects of amphetamine, dexfenfluramine, diazepam, and dietary manipulations on responding reinforced by stimuli paired with food in nonhuman primates. Pharmacology, Biochemistry, and Behavior, 77(3), 471–479. Foltin, R. W., & Evans, S. M. (1999). The effects of D-amphetamine on intake of food and a sweet fluid containing cocaine. Pharmacology, Biochemistry, and Behavior, 62, 457–464. Foltin, R. W., & Fischman, M. W. (1988). The effects of varying procurement costs on food intake in baboons. Physiology and Behavior, 43, 493–499.

120

R.W. Foltin / Appetite 45 (2005) 110–120

Geliebter, A., Melton, P. M., McCray, R. S., Gallagher, D. R., Gage, D., & Hashim, S. A. (1992). Gastric capacity, gastric emptying and test-meal intake in normal and bulimic women. American Journal of Clinical Nutrition, 56, 656–661. Hensrud, D. D. (2000). Pharmacotherapy for obesity. Medical Clinics of North America, 84, 463–476. Hernandez, L., & Hoebel, B. G. (1988). Food reward and cocaine increase extracellular dopamine in the nucleus accumbens as measured by microdialysis. Life Sciences, 42, 1705–1712. Highfield, D. A., Mead, A. N., Grimm, J. W., Rocha, B. A., & Shaham, Y. (2002). Reinstatement of cocaine seeking in 129X1/SvJ mice: Effects of cocaine priming, cocaine cues and food deprivation. Psychopharmacology, 161, 417–424. Hubert, P., King, N. A., & Blundell, J. E. (1998). Uncoupling the effects of energy expenditure and energy intake: Appetite response to short-term energy deficit induced by meal omission and physical activity. Appetite, 31, 9–19. Johnstone, A. M., Faber, P., Gibney, E. R., Elia, M., Horgan, G., Golden, B. E., et al. (2002). Effect of an acute fast on energy compensation and feeding behaviour in lean men and women. International Journal of Obesity and Related Metabolic Disorders, 26, 1623–1628. Kanarek, R. (1997). Psychological effects of snacks and altered meal frequency. British Journal of Nutrition, 77, S105–S120. Kelleher, R. (1966). Conditioned reinforcement in second-order schedules. Journal of the Experimental Analysis of Behavior, 9, 475–485. Kissileff, H. R., Wentzlaff, T. H., Guss, J. L., Walsh, B. T., Devlin, M. J., & Thornton, J. C. (1996). A direct measure of satiety disturbance in patients with bulimia nervosa. Physiology and Behavior, 60, 1077– 1085. Kornblith, C. L., & Hoebel, B. G. (1976). A dose–response study of anorectic drug effects on food intake, self-stimulation, and stimulationescape. Pharmacology, Biochemistry, and Behavior, 5, 215–218. Leibowitz, S. F. (1978). Identification of catecholamine receptor mechanisms in the perifornical lateral hypothalamus and their role in mediating amphetamine and L-DOPA anorexia. In S. Garattini, & R. Samanin (Eds.), Central mechanisms of anorectic drugs. New York: Raven Press. Lluch, A., Hubert, P., King, N. A., & Blundell, J. E. (2000). Selective effects of acute exercise and breakfast interventions on mood and motivation to eat. Physiology Behavior, 68, 515–520. Mackintosh, N. J. (1974). The psychology of animal learning. New York: Academic Press.

Radhakishun, F. S., van Ree, J. M., & Westerink, B. H. C. (1988). Scheduled eating increases dopamine release in the nucleus accumbens of food-deprived rats as assessed with on-line brain dialysis. Neuroscience Letters, 85, 351–356. Robinson, T. E., & Berridge, K. C. (1993). The neural basis of drug craving: an incentive-sensitization theory of addiction. Brain Research Reviews, 18, 247–291. Salamone, J. D., Cousins, M. S., McCullough, L. D., Carriero, D. L., & Berkowitz, R. J. (1994). Nucleus accumbens dopamine release increases during instrumantal lever pressing for food but not free food consumption. Pharmacology Biochemistry and Behavior, 49, 25–31. Shaham, Y., Shalev, U., Lu, L., De Wit, H., & Stewart, J. (2003). The reinstatement model of drug relapse: History, methodology and major findings. Psychopharmacology, 168, 3–20. Small, D. M., Jones-Gotman, M., & Dagher, A. (2003). Feeding-induced dopamine release in dorsal striatum correlates with meal pleasantness ratings in healthy human volunteers. Neuroimage, 19, 1709–1715. Toates, F. M. (1981). The control of ingestive behaviour by internal and external stimuli—a theoretical review. Appetite, 2, 35–50. Tolkamp, B. J., Allcroft, D. J., Austin, E. J., Nielsen, B. L., & Kyriazakis, I. (1998). Satiety splits feeding behaviour into bouts. Journal of Theoretical Biology, 194, 235–250. Whishaw, I. Q., & Kornelsen, R. A. (1993). Two types of motivation revealed by ibotenic acid nucleus accumbens lesions: Dissociation of food carrying and hoarding and the role of primary and incentive motivation. Behavioral Brain Research, 55, 283–295. Wilson, A. W., Costall, B., & Neill, J. C. (2000). Manipulation of operant responding for an ethanol-paired conditioned stimulus in the rat by pharmacological alteration of the serotonergic system. Journal of Psychopharmacology, 14, 340–346. Wise, R. A. (1996). Addictive drugs and brain stimulation reward. Annual Review of Neuroscience, 19, 319–340. Woods, J. H., & Winger, G. D. (2002). Observing responses maintained by stimuli associated with cocaine or remifentanil reinforcement in rhesus monkeys. Psychopharmacology, 163, 345–351. Wyvell, C. L., & Berridge, K. C. (2000). Intra-accumbens amphetamine increases the conditioned incentive salience of sucrose reward: enhancement of reward ‘wanting’ without enhanced ‘liking’ or response reinforcement. Journal of Neuroscience, 20, 8122–8130. Zacny, J. P., & de Wit, H. (1989). Effects of food deprivation on subjective responses to D-amphetamine in humans. Pharmacology, Biochemistry, and Behavior, 34, 791–795.