Consistent relationships between sensory properties of savory snack ...

International Journal of Obesity (2006) 30, 1685–1692 & 2006 Nature Publishing Group All rights reserved 0307-0565/06 $30.00 www.nature.com/ijo

ORIGINAL ARTICLE Consistent relationships between sensory properties of savory snack foods and calories influence food intake in rats SE Swithers, A Doerflinger and TL Davidson Department of Psychological Sciences and Ingestive Behavior Research Center, Purdue University, West Lafayette, IN, USA Objective: Determine the influence of experience with consistent or inconsistent relationships between the sensory properties of snack foods and their caloric consequences on the control of food intake or body weight in rats. Design: Rats received plain and BBQ flavored potato chips as a dietary supplement, along with ad lib rat chow. For some rats the potato chips were a consistent source of high fat and high calories (regular potato chips). For other rats, the chips provided high fat and high calories on some occasions (regular potato chips) and provided no digestible fat and fewer calories at other times (light potato chips manufactured with a fat substitute). Thus, animals in the first group were given experiences that the sensory properties of potato chips were strong predictors of high calories, while animals in the second group were given experiences that the sensory properties of potato chips were not predictors of high calories. Subjects: Juvenile and adult male Sprague–Dawley rats. Measurements: Following exposure to varying potato chip–calorie contingencies, intake of a novel, high-fat snack food and subsequent chow intake were assessed. Body weight gain and body composition as measured by DEXA were also measured. Results: In juvenile animals, exposure to a consistent relationship between potato chips and calories resulted in reduced chow intake, both when no chips were provided and following consumption of a novel high-fat, high-calorie snack chip. Long-term experience with these contingencies did not affect body weight gain or body composition in juveniles. In adult rats, exposure to an inconsistent relationship between potato chips and calories resulted in increased consumption of a novel high-fat, highcalorie snack chip premeal along with impaired compensation for the calories contained in the premeal. Conclusion: Consumption of foods in which the sensory properties are poor predictors of caloric consequences may alter subsequent food intake. International Journal of Obesity (2006) 30, 1685–1692. doi:10.1038/sj.ijo.0803329; published online 21 March 2006 Keywords: body weight; Pavlovian conditioning; energy regulation

Introduction Foods containing nondigestible fat substitutes (e.g., noncaloric sucrose polyesters) have been available in the US for about 10 years and are often used in the manufacture of reduced fat snack foods. Studies on the effects of consuming these foods on intake and body weight gain in humans have yielded inconsistent results. For example, in short-term (less than 1 day) tests, replacement of fat by nonabsorbable sucrose polyester has been demonstrated to produce no effects on energy intake (e.g., Hulshof et al.1 and Westerterp-

Correspondence: Dr SE Swithers, Department of Psychological Sciences, Purdue University, 703 Third Street, West Lafayette, IN 47907, USA. E-mail: [email protected] Received 4 August 2005; revised 16 February 2006; accepted 19 February 2006; published online 21 March 2006

Plantenga et al.2), while longer-term (12 days) studies suggest replacement results in lower energy consumption.3 Further, long-term (9 months) studies in obese individuals indicate that use of the fat substitute Olestra results in greater losses of body weight and body fat loss compared to fat reduction alone.4,5 In contrast, a long-term (1 year) study in free living humans demonstrated that a higher proportion of obese individuals consumed snacks containing Olestra (compared to overweight and normal-weight individuals), and that obese individuals that consumed the highest quantities of Olestra-containing snacks had significantly greater caloric intake during the first year after introduction of Olestra to the marketplace.6 Thus, the role that fat replacement or substitution may play in control of food intake and body weight remains uncertain. The purpose of the present research is to assess the possibility that consumption of reduced calorie, fat sub-

Snack chip–calorie consistency and food intake SE Swithers et al

1686 stitutes might have unwanted consequences with respect to controlling caloric intake and body weight gain. It is welldocumented that animals, including humans, can learn to use the taste and other orosensory properties of foods to predict the nutritive or caloric consequences of eating (e.g., Beauchamp and Cowart,7 Booth,8 Davis and Smith,9 Sclafani10 and Sclafani11). Previously, we proposed that sweet tastes and food viscosity provide orosensory cues that are normally good predictors of caloric consequences.12–14 Furthermore, we hypothesized that experience with substances that taste sweet but contain no calories, or that are low viscosity but relatively high in calories, reduces the ability of animals to use sweetness and viscosity to anticipate the calories that are contained in the foods that they eat. Our results showed that rats that received this type of inconsistent experience ate more and gained more weight than rats that received only sweet, high calorie, food or foods that were both highly viscous and high calorie.13,15 The caloric density of highly palatable, high-fat foods is typically greater than that of foods that are less palatable and lower in fat. Thus, preabsorptive properties of palatability and fat taste would seem to be good predictors of the postingestive caloric consequences of eating. Based on simple principles of Pavlovian conditioning, experience with orosensory stimuli that are produced by consuming fat without the normal postingestive caloric consequences that usually accompany those orosensory cues would be expected to reduce the ability of fat-associated oral cues to predict calories in foods that are high in dietary fat. In the present studies, we examine whether providing animals with dietary experiences in which the sensory properties of a savory snack food (potato chips) are either consistently or inconsistently associated with calories affects control of food intake or body weight in short- and long-term tests in juvenile and adult rats. In the first experiments, juvenile rats were given two flavors of potato chips in addition to their normal maintenance chow. For one group, group Consistent, both flavored potato chips were high in fat and high in calories; thus, the sensory properties of chips were strong predictors of calories for this group. For a second group, group Inconsistent, one flavored chip was consistently high in fat and calories, while the alternative flavored chip was lower in calories and contained a fat substitute. Thus, for these animals, the sensory properties associated with potato chips were poor predictors of calories since chips sometimes provided high calories and sometimes provided lower calories. The short-term consequences of these experiences on consumption of a novel, high-fat snack chip, and subsequent caloric compensation were examined.

Experiments 1a and b Methods Subjects. Juvenile Sprague–Dawley (Harlan, IN, USA) male rats (Experiment 1a, n ¼ 10–12 per group; Experiment 1b International Journal of Obesity

n ¼ 10–12 per group), approximately 30 days of age at the start of training, were subjects in the current experiment. In Experiment 2, adult male, Sprague–Dawley (Harlan, IN) male rats weighing 330–370 g (approximately 90 days of age) at the start of the experiment were subjects. Animals were housed individually in hanging wire cages with laboratory chow (Lab Diets 5001) and water available ad lib except during testing as described below. The colony room was maintained on a 14:10 lightHdark cycle, with temperature maintained at 21–231C. Training. In Experiments 1a and b, animals were randomly assigned to one of three groups (Experiment 1a – Consistent, Inconsistent, Chow; Experiment 1b – Consistent, Inconsistent, Consistent Control). Animals in the Consistent condition received daily alternating exposure to 10 g of two high-fat, high-calorie snack foods (Lay’s potato chips; BBQ flavor Lay’ss potato chips; 5.5 kcal/g). The second group, group Inconsistent, received daily alternating exposure to 10 g of one high-fat, high-calorie diet (Lay’s potato chipss or BBQ Lay’ss potato chips; 5.5 kcal/g) and 10 g of a fatsubstituted, lower-calorie diet (Plain or BBQ Light! spotato chips; 2.75 kcal/g). The lower-calorie diet was manufactured with a noncaloric fat substitute (Oleans), which is designed to provide the sensory attributes of fat, without fat calories. All animals received one plain flavored chip and one BBQ flavored chip; the order of presentation and chip–calorie association were counterbalanced across animals. In Experiment 1a, the third group (Chow) received exposure to cups filled with 10 g of their maintenance chow diet. In Experiment 1b, the third group (Consistent Control) was designed to control for the amount of experience with high-fat chips. These animals received exposures to the high-fat, highcalorie chips on the same days as the Inconsistent animals, but received chow alone on the alternate days. The order of presentation was counterbalanced across animals. In Experiment 2, animals (n ¼ 12 per group) were randomly assigned to one of three groups, Consistent, Random Inconsistent and Consistent Control. Animals in the Consistent condition received daily exposure to 7 g of one of the two high-fat, high-calorie potato chips used in Experiments 1a and b. Animals received eight exposures to each of the two flavored chips presented in random order over the course of training. The second group, Random Inconsistent, received daily exposure to 7 g of one of the four potato chip diets. Two of the potato chips (plain and BBQ) were high fat and high calorie, whereas the remaining chips (plain and BBQ) were the fat-substituted, lower-calorie chips. All animals received four exposures to each of the two flavored high-fat, high-calorie chips and four exposures to each of the two flavored, low-calorie chips. The order of chip presentation was randomized over the course of training. The third group, Consistent Control, was designed to control for the amount of experience with high-fat chips. These animals received four exposures to each of two flavored high-fat, high-calorie chips on the same days as the


1687 Inconsistent animals, but received chow alone on the alternate days. The order of presentation of the chips was randomized across training. Consistent and Random Inconsistent animals had one cup of chips presented each day 4 days per week for 4 weeks; Consistent Control animals had one cup of chips presented 2 days per week for 4 weeks. Potato chips were provided in small enamel camping cups attached to the inside of the cage. Potato chips were available for approximately 23 h; cups were then removed and contents were weighed to determine intake. Laboratory chow and water were available ad lib throughout the experiment except during testing as noted below. Chip intake, chow intake and body weight were recorded daily.

Testing. Following eight consecutive days (Experiment 1a and b) or 16 total days (Experiment 2) of exposure to the appropriate chip diets, animals were given lab chow for 1 day, and then food was removed overnight. Half of the animals in each group were presented with a premeal on the following morning following approximately 18 h food deprivation; the other half of the animals in each of the groups received no premeal. The premeal consisted of a novel, high fat, snack food with a caloric density similar to the high-calorie potato chip training diet (Bugless Original chip). The premeal (3 g) was offered for 1 h in Experiments 1a and b and for 30 min in Experiment 2. Standard lab chow was then presented to the animals, and food intake and body weight measures were made at 1, 2, 4 and 24 h. Animals were then were maintained for 3 days on their maintenance chow alone, food deprived overnight, and animals that had previously received the premeal were tested with no premeal and vice versa; the order of preload presentations was counterbalanced across animals. In Experiment 1b, following the intake tests, the animals were returned to their scheduled daily, alternating delivery of 10 g potato chips for an additional 30 exposures (5 days per week for 6 weeks). Then, body composition of the animals was determined using DEXA. Animals were briefly anesthetized with ketamine and xylazine (70 and 7 mg/kg, respectively, i.p.), then scanned using a pDEXA Sabre machine. To determine whether exposure to the potato chips diets during the juvenile phase had any persistent effects on body weight or body composition, animals were maintained on chow diets (with no potato chips) for an

Table 1

additional 3 months following the long-term exposure, then body composition was reassessed by DEXA. All experimental procedures were approved by the Purdue University Animal Care and Use Committee. Statistical analyses. The effects of short-term training on body weight gain were examined using a one-way ANCOVA with starting body weight as the covariate. The effects of the training on food intake on days when chow alone was available (days 9 and 14) were examined separately for Experiments 1a and b using one-way (diet) repeated measures (day of testing) ANOVAs. The effects of the training on intake following the premeal were examined using a three-way (time premeal training condition), repeated measures ANOVA, with time of testing and premeal condition as within factors and training condition as a between factor. Long-term body weight gain in Experiment 1b was examined using a one-way (diet), repeated measures ANOVA. Body composition was analyzed using a one-way (diet), repeated measures (age) ANOVA. Follow-up ANOVAS of significant effects were conducted where indicated. A P-value of o0.05 was taken as significant for all effects.

Results Experiments 1a and b During training, two animals (one in group Consistent and one in group Inconsistent) in Experiment 1a and three animals (one each in groups Consistent, Inconsistent and Consistent Control) in Experiment 1b failed to consume at least 75% of the greatest amount of potato chip supplements consumed by any animal, and their data were excluded from further analysis. Mean grams of potato chip intake in the remaining animals did not differ across the Consistent and Inconsistent groups (Table 1). In both Experiments 1a and b, body weights of all groups were similar at the outset of training, and animals gained similar amounts of weight across the 8 days of training (Table 1). When chow intake was assessed following training on days that no potato chips were provided, intake was affected by the training experience in both Experiments 1a and b. In Experiment 1a, animals given consistent experience with the chips and calories consumed significantly fewer calories of chow on days when no chips were provided compared to animals given either

Body weight and chip consumption in Experiments 1a and b

Group

Mean daily potato chip intake (g) Start body weight (g) End body weight (g)

Experiment 1a

Experiment 1b

Consistent (n ¼ 11)

Inconsistent (n ¼ 11)

Chow control (n ¼ 10)

Consistent (n ¼ 10)

Inconsistent (n ¼ 10)

Consistent control (n ¼ 10)

6.5770.22 82.2771.23 135.9173.08

6.9570.24 84.4571.94 140.2773.52

NA 85.672.03 146.0074.13

6.8270.24 70.671.82 123.173.82

6.9770.23 69.671.23 123.572.86

7.3670.29 70.571.68 120.672.45

International Journal of Obesity


1688 inconsistent experience, or no experience; in addition, animals consumed significantly more chow on the second day of measurement, but there was no interaction with diet (main effect of diet training; F[2, 28] ¼ 6.64, Po0.05; main effect of test day; F[1, 28] ¼ 13.66, Po0.05; Figure 1a). In Experiment 1b, animals in both the Consistent and Consistent Control groups consumed significantly fewer calories of chow on days when no chips were provided compared to animals in the Inconsistent group; animals also consumed significantly more chow on the second day of measurement, but there was no interaction with diet condition (main effect of diet training; F[2, 25] ¼ 3.51, Po0.05; main effect of test day; F[1, 25] ¼ 60.9; Po0.05; Figure 1b). In Experiments 1a and b, there were no significant differences in the quantity of the novel, high-fat, highcalorie premeal consumed across groups (Figure 2). Following this premeal, intake of chow was significantly affected by

the premeal, the time during testing, and the training experience of the animal (Experiment 1a; main effect of diet; F[2, 29] ¼ 9.47, Po0.05; main effect of premeal; F[1, 29] ¼ 12.61, Po0.05; main effect of time; F[3, 87] ¼ 2525, Po0.05; time diet interaction; F[6, 87] ¼ 4.63, Po0.05; premeal time interaction; F[3, 87] ¼ 28.04, Po0.05; Experiment 1b; main effect of premeal; F[1, 28] ¼ 17.18, Po0.05; main effect of time; F[3, 84] ¼ 2389, Po0.05; time diet interaction; F[6, 84] ¼ 4.83, Po0.05; premeal time interaction; F[3, 84] ¼ 10.74, Po0.05). Post hoc tests revealed that animals in both experiments consumed significantly fewer calories following the premeal compared to following no premeal. In addition, in both experiments, animals in the Consistent group consumed significantly fewer grams of chow after 24 h (but not at other time points) compared to the Inconsistent and Control groups, which did not differ from one another (Figures 3 and 4).

60

80 *

*

24-hr intake (Calories)

24-hr intake (Calories)

80

40

20

0 Consistent Inconsistent

60

40

20

0

Chow Control

*

Diet

Consistent Inconsistent Consistent Control Diet

Figure 1 Chow intake in Experiment 1a (left) and Experiment 1b (right) on days when no potato chip supplements were provided (collapsed across chow day 1 and chow day 2). *Po0.05 compared to Inconsistent group.

b

3.0 2.5

Premeal Intake (grams)


a

2.0 1.5 1.0

3.0 2.5 2.0 1.5 1.0 0.5

0.5

0.0

0.0 Consistent Inconsistent Diet

Chow Control

Consistent Inconsistent Consistent Control Diet

Figure 2 Intake of the novel, high-calorie snack food in Experiment 1a (left) and Experiment 1b (right).



1689 Long-term exposure to the potato chip diets did not affect body weight gain either when the chips were available, or after the exposure to the chips had been discontinued (Figure 5). Owing to technical difficulties, DEXA data were not obtained from all animals at both ages; only animals for which DEXA data were available at both time points were included (n ¼ 7–9 per group). Analysis of body composition at the end of long-term exposure to potato chips and after a subsequent 3-month exposure to chow alone indicated that there were no effects of the training diet on body composition at either time point (Figure 6).

Experiment 2 During training, two animals (one consistent and one random inconsistent) consumed less than 75% of the largest

25

Premeal No Premeal

350

*

300 Body weight gain (grams)

24-hr Chow Intake (grams)

30

quantity of potato chips consumed; their data were excluded from analysis. Potato chip consumption did not differ across the consistent and random inconsistent conditions in the remaining animals (Table 2). Body weights did not differ across the groups at the beginning or end of the 4 weeks of training (Table 2). When chow intake was assessed following training, there were no significant effects of diet training on chow intake. Premeal intake was significantly affected by diet training (main effect of diet; F[2, 31] ¼ 5.44, Po0.05; Figure 7). Animals in the Random Inconsistent group consumed significantly more of the premeal compared to animals in the Consistent and Consistent Control groups. When chow intake was assessed following the novel premeal, intake was affected by the training condition, premeal, and time during testing (main effect of diet; F[2, 31] ¼ 5.57, Po0.05; main effect of premeal; F[1, 31] ¼ 16.84, Po0.05; main effect of time; F[3, 93] ¼ 3436, Po0.05;

20 15 10 5

Consistent Inconsistent Consistent Control

250 200 150 100 50

0 Consistent

Inconsistent Group

Training 1

Figure 3 Intake of chow diets following a novel, high-calorie premeal in animals with Consistent, Inconsistent or no experience with potato chips and calories. *Po0.05 compared to Inconsistent and Chow Control groups.

25

2

3 4 5 6 Exposure time (weeks)

Premeal No Premeal

30

*

Consistent Inconsistent Consistent Control

25

20 15 10

Adult

Figure 5 Body weight gain during 8 days of training, weekly during longterm exposure, and following 3 months maintenance on chow alone (adult).

Body fat (percent)

24-hr Chow Intake (grams)

30

0

Chow Control

20 15 10

5 5 0 Consistent

Inconsistent Group

Consistent Control

Figure 4 Intake of chow diets following a novel, high-calorie premeal in animals with Consistent, Inconsistent or Consistent Control with potato chips and calories. *Po0.05 compared to Inconsistent and Consistent Control groups.

0 Juvenile

Adult Age

Figure 6 Body fat percentage assessed by DEXA immediately following long-term exposure to varying potato chip–calorie relationships (Juvenile) and after 3 subsequent months maintenance on chow alone (Adult).



1690 premeal diet interaction; F[2, 31] ¼ 3.61, Po0.05; premeal time interaction; F[3, 93] ¼ 4.10; Po0.05). Post hoc analyses indicated that animals in the Consistent and Consistent Control conditions consumed significantly less

Body weight and chip consumption in Experiment 2

Group

Experiment 2 Consistent (n ¼ 11)

Mean daily potato chip intake (g) 6.8970.08 Start body weight (g) 358.472.5 End body weight (g) 413.073.6

Discussion

Random inconsistent (n ¼ 11)

Consistent control (n ¼ 12)

6.6870.15 6.7870.14 357.475.1 357.474.1 407.877.8 410.075.4

3.0


*

*

2.5

2.0

1.5

1.0

0.5

0.0 Consistent

Random Inconsistent Diet

Consistent Control

Figure 7 Intake of novel, high-calorie premeal in adult rats given Consistent, Random Inconsistent or Consistent Control experience with potato chips and calories. *Po0.05 compared to Random Inconsistent.

Premeal No Premeal

Random Inconsistent *

30

20

10

0

*

1

*

2 4 Time (hr)

24

Cumulative Chow Intake (grams)


Consistent 40

Taken together, the results of the present experiments suggest that experiences with foods and their caloric consequences may contribute to modulation of food intake over the short-term in juvenile and adult rats. In the Experiments 1a and b, juvenile rats given a consistent relationship between potato chips and high calories decreased chow intake relative to naı¨ve animals and animals given an inconsistent relationship, suggesting that Consistent animals may have learned to anticipate the availability of high-calorie foods. In Experiment 2, adult animals with an inconsistent relationship between potato chips and calories overconsumed a novel, high-fat chip premeal and then failed to compensate for the calories contained in that premeal, suggesting an impaired ability to modulate intake compared to animals with a history of consistent chip–calorie relationships. Thus, these data suggest that one factor that might contribute to the increasing prevalence of overweight and obesity is increasing exposure to dietary experiences in which sensory properties of foods do not provide strong predictors of caloric consequences. The prediction of the present set of experiments was that animals given experience with inconsistent relationships between the sensory properties of a food and its caloric consequences would have an impaired ability to compensate for calories provided in a novel food with sensory properties similar to that experience during training. While the results provided some support for this hypothesis, the outcomes of the experiments are not definitive. For example, in the first set of experiments, juvenile animals were given experiences Consistent Control

40


Table 2

chow following the premeal indicating compensation for calories delivered in the premeal. In contrast, animals in the Random Inconsistent condition consumed similar quantities of chow following the premeal and following no premeal indicating that they did not compensate for the calories contained in the premeal (Figure 8).

Premeal No Premeal 30

20

10

0

1

2 4 Time (hr)

24

40

* Premeal No Premeal

30

20

10

0

*

1

*

2 4 Time (hr)

24

Figure 8 Chow intake following a novel, high-calorie premeal in adult rats given Consistent, Random Inconsistent or Consistent Control experience with potato chips and calories. *Po0.05 compared to no premeal condition.



1691 in which the sensory properties of potato chips were either good or poor predictors of calories. Under those circumstances, animals for which the potato chips consistently predicted high calories showed decreased chow intake relative to animals that were naı¨ve to potato chips, as well as relative to animals that had experiences in which the sensory properties of potato chips were inconsistently related to calories. These results are consistent with the idea that providing animals with the opportunity to learn that the sensory properties of a food can provide consistent cues for calories results in changes in ingestive behavior. Animals that do not have the opportunity to learn this relationship, either because they have not been provided with the experience (chow control animals) or because they have explicitly received experience that sensory properties are not good cues for calories (Inconsistent animals), consume increased quantities of chow. However, in contrast to initial predictions, Consistent animals demonstrated decreased chow intake both following a novel premeal with sensory properties similar to the training diet and when no premeal was provided. That is, animals did not appear to use prior learning about the sensory properties of the premeal to modulate subsequent food intake. The results could be due to Consistent animals having learned to anticipate the availability of a high-calorie diet supplement (potato chips). Decreased chow intake might then reflect a compensatory response that persisted even when no chips were provided. As the other groups did not have this high-calorie food regularly available, they could not anticipate its presence, and did not adjust their chow intake. Alternatively, a failure to demonstrate effects of the training on caloric compensation following the premeal may suggest that previous experience with sensory property – calorie consistency does not contribute to the control of food intake. Animals in all groups consumed the entire premeal provided, and adjusted chow intake in a similar fashion relative to when no premeal was provided. These results suggest that for these juvenile rats, the previous history of food–calorie consistency did not contribute in determining food intake. The results from our study of the effects of long-term experience with consistent versus inconsistent potato chip– calorie relationships also provided no evidence that body weight and adiposity were affected by these experiences. Several possible explanations for this lack of difference in body weight are possible. First, because animals were juveniles at the start of testing, all animals were showing rapid increases in body weight across the period under study. These rapid increases in body weight may have obscured potential differences in body weight gain across groups. For example, previous work in animals at similar ages has demonstrated that diets of differing viscosity do not produce differences in body weight across the same time frame, but do result in increased adiposity and adult body weight gain (e.g, Swithers and Davidson15). However, in the present study, neither adiposity nor adult body weight gain were affected by the consistency of the experience.

An alternative possibility is that animals were learning to use the flavor cues to predict the caloric consequences, and thus over the long-term, the Consistent and Inconsistent groups did not differ in predictability. Such a possibility can be examined in future studies, for example using the Random Inconsistent design employed in the experiment with adult rats. In addition, the high-caloric density of the diets employed may have obscured the effects of the taste–calorie consistency relationship. The high-fat, high-calorie chip contained more calories than the animal’s maintenance chow, while the lower-calorie chip contained fewer calories. As a result, animals consistently consuming the high-fat chip were receiving significantly more calories (equal to roughly 50% of their daily caloric intake) than animals sometimes consuming the lower-calorie chip. Perhaps the direct effects of consuming large quantities of high-calorie chips were sufficient to offset the indirect effects of predicting the delivery of high calories. Thus, one additional approach would be to assess the consequences of providing smaller quantities of the high-calorie diets. Finally, it is possible that long-term experience with the consistency of food–calorie relationships, such as those employed here, does not result in significant changes in body weight regulation in juvenile animals. While the long-term consequences of consistent and inconsistent potato chip–calorie relationships in adult animals remain untested, data from short-term intake tests did reveal that providing experiences in which neither flavor cues nor other sensory properties of chips were predictive of calories did result in alteration of food intake. Animals given Inconsistent experience with chips and calories consumed greater quantities of a novel, high-fat snack food premeal compared to animals with consistent experience between high fat, savory food and calories. Further, despite the increased intake of the premeal, Inconsistent animals failed to compensate for any of the calories in the premeal, consuming similar quantities of chow following the highcalorie premeal and following no premeal at all. In contrast, animals given a Consistent relationship between chips and calories showed an immediate caloric compensation for the premeal calories, reducing chow intake within the first hour to adjust for the calories consumed in the premeal. This caloric compensation was maintained for the duration of the 24 h test. In addition, unlike the results seen with juveniles, differences in chow intake were observed only following the novel savory premeal, and not when no premeal was provided. These data suggest that in these adult animals, the previous experience with sensory properties of the potato chips and calories modulated responding to the novel chip premeal. Although age-related differences cannot presently be excluded, the observed differences between the juvenile and adult results likely reflect differences in experimental design. For example, in adults, the flavor (plain or BBQ) was not predictive of calories whereas in juveniles it was. In addition, in juveniles, high- and low-calorie chips were provided on alternating days, while in adults the order of presentation was randomized. Thus, relative to the juveniles, International Journal of Obesity


1692 adult animals in the Inconsistent group had fewer cues available to predict caloric composition. The mechanisms by which providing consistent relationships between the sensory properties of foods and their calories influence control of food intake are unknown. However, studies of Pavlovian conditioning suggest a number of possibilities. For example, Pavlov first reported that the sensory properties of food and stimuli that are cognitively associated with food (e.g., the sound of metronome) can induce secretion of saliva, gastric acid and pepsin in dogs. Thus, one possible mediator of the observed effects is that orosensory cues that predict the delivery of nutrients can acquire the capacity to evoke what are termed ‘cephalic phase’ responses. Cephalic phase responses are physiological reflexes that are evoked preabsorptively by stimuli related to food 16,17 which are usually transient, fractional components of larger physiological changes that occur when food actually enters the gastrointestinal tract. A predictive relationship between the sensory properties of a food and its postingestive, caloric consequences may enable foods to elicit anticipatory conditioned cephalic phase reflexes, which may promote efficient regulation of body weight (e.g., Cannon18 and Woods and Ramsay19). However, the role of such conditioned cephalic phase responses in modulating food intake or body weight has yet to be demonstrated, and will require additional experiments.

Acknowledgements Supported by NIH grants R01HD44179 and R01HD29792. Portions of these data were presented at the 2005 meeting of the Society for the Study of Ingestive Behavior. Thanks to Melissa McCurley, Curt Studebaker, Daira Springer, Erica Hamilton and Carolina Campanella for their assistance with data collection.

References 1 Hulshof T, de Graaf C, Weststrate JA. Short-term effects of high-fat and low-fat/high-SPE croissants on appetite and energy intake at three deprivation periods. Physiol Behav 1995; 57: 377–383.


2 Westerterp-Plantenga MS, Wijckmansduijsens NEG, Tenhoor F, Weststrate JA. Effect of replacement of fat by nonabsorbable fat (sucrose polyester) in meals or snacks as a function of dietary restraint. Physiol Behav 1997; 61: 939–947. 3 Degraaf C, Hulshof T, Weststrate JA, Hautvast J. Nonabsorbable fat (sucrose polyester) and the regulation of energy intake and body weight. Am J Physiol Reg Int Comp Physiol 1996; 39: R1386–R1393. 4 Bray GA, Lovejoy JC, Most-Windhauser M, Smith SR, Volaufova J, Denkins Y et al. A 9-mo randomized clinical trial comparing fat-substituted and fat-reduced diets in healthy obese men: the Ole Study. Am J Clin Nutr 2002; 76: 928–934. 5 Lovejoy JC, Bray GA, Lefevre M, Smith SR, Most MM, Denkins YM et al. Consumption of a controlled low-fat diet containing olestra for 9 months improves health risk factors in conjunction with weight loss in obese men: the Ole’ Study. Int J Obes Relat Metab Disord 2003; 27: 1242–1249. 6 Satia-Abouta J, Kristal AR, Patterson RE, Neuhouser ML, Peters JC, Rock CL et al. Is olestra consumption associated with changes in dietary intake, serum lipids, and body weight? Nutrition 2003; 19: 754–759. 7 Beauchamp GK, Cowart BJ. Congenital and experiential factors in the development of human flavor preferences. Appetite 1985; 6: 357–372. 8 Booth D. Conditioned satiety in the rat. J Comp Physiol Psychol 1972; 81: 457–471. 9 Davis J, Smith G. Learning to sham feed: behavioral adjustments to loss of physiological postingestional stimuli. Am J Physiol 1990; 259: R1228–R1235. 10 Sclafani A. Learned controls of ingestive behaviour. Appetite 1997; 29: 153–158. 11 Sclafani A. Post-ingestive positive controls of ingestive behavior. Appetite 2001; 36: 79–83. 12 Davidson TL, Swithers SE. A Pavlovian approach to the problem of obesity. Int J Obes Relat Metab Disord 2004; 28: 933–935. 13 Davidson TL, Swithers SE. Food viscosity influences caloric intake compensation and body weight in rats. Obes Res 2005; 13: 537–644. 14 Swithers SE, Davidson TL. Obesity: outwitting the wisdom of the body? Curr Neurol Neurosci Rep 2005; 5: 159–162. 15 Swithers SE, Davidson TL. Influence of early dietary experience on energy regulation in rats. Physiol Behav 2005; 86: 669–680. 16 Giduck SA, Threatte RM, Kare MR. Cephalic reflexes: their role in digestion and possible roles in absorption and metabolism. J Nutr 1987; 117: 1191–1196. 17 Mattes RD. Nutritional implications of the cephalic-phase salivary response. Appetite 2000; 34: 177–183. 18 Cannon WB. The Wisdom of the Body. W.W. Norton & Co. Inc.: New York, 1932. 19 Woods SC, Ramsay DS. Pavlovian influences over food and drug intake. Behav Brain Res 2000; 110: 175–182.