- Email: [email protected]
Sensory-specific satiety and its importance in meal termination
Neuroseienceand BiobehavioralReviews,Vol. 20, No. l, pp. 113--117,1996 Copyright© 1995 ElsevierScienceLtd Printed in Great Britain. All fights reserved 0149-7634/96$9.50 + .00
Pergamon
0149-7634(95)00048-8
Sensory-Specific Satiety and its Importance in Meal Termination M. M. H E T H E R I N G T O N
Department of Psychology, University of Dundee, DD1 4HN, Scotland, UK
HETHERINGTON, M.M. Sensory-specificsatiety and its importance in meal termination. NEUROSCI BIOBEHAV REV 20(1) 113-117, 1996. - - Pleasantness is important in influencing food choice, and may play a role in determining the amount of food consumed. Judgements of pleasantness decrease as the food is eaten. It has been proposed that this reflects the development of satiety to a specific food. However, consumers may not rate these changes as important in meal termination. Fifty-seven subjec,~swere given ad lib access to a test meal of cheese on crackers and at the end of this meal recorded the main reason for stopping from a possible seven statements. They then rank ordered the importance of each reason. One hour later, subjects were offered a choice of the same food, a different food, or no second course. Again reasons for stopping were recorded by those who selected a second course. The most common reason given for meal termination in the first course was "I got tired of eating that food" (40%) and for the second course "I felt full" (48%). Subjects were divided into those who rated fatigue and changes in pleasantness as important and those who rated fullness as more important. Significant differences in intake between these groups indicated that those who rated fatigue/hedonics as important consumed significantly fewer calories (275±23 kcal) than those who rated fullness as more important (424±65 kcal). It is argued that fatigue experienced by subjects may reflect sensory fatigue and that this is an important part of the development of sensory-specific satiety. Since subjects who rated gastric fullness as the most important reason for terminating the meat consumed more calories, it is suggested that this index of satiety may be relatively more crude than sensory or hedonic variables. Satiety
Food intake
Appetite
Sensory-specificsatiety
INTRODUCTION
again from an array of different foods is reduced (4). This dynamic aspect of pleasure as reflected in subjective judgements, rate of eating and food choice has b e e n n a m e d "sensory-specific satiety" (4). That satiety can be specific to a particular food as it is eaten appears to be a reliable p h e n o m e n o n and is observed across different species (7). However, the precise mechanisms which underly the expression of sensoryspecific satiety remain to be elucidated, although work on oral habituation (8) and neuronal models of food reward mechanisms (9) shed light on potentially important theoretical models (10). Berridge (11) has identified two distinguishable c o m p o n e n t s of the pleasure of eating. First of all liking, which is measured by pleasure responses in animals or subjective ratings of pleasantness in humans, and secondly, wanting, which is m e a s u r e d by assessing incentive motivation in animals or desire to eat (appetite for) a specific food in humans. Findings from neurophysiological studies in animals indicate that different regions of the brain control these two separable components of pleasure (11). Similarly, in humans pharmacological manipulations have revealed agents which will reduce hunger while sparing the pleasantness of foods (dfenfluramine; 12) and opioid blockade, which reduces the pleasantness of foods while not
E A T I N G IS highly pleasurable and of central importance in providing and maintaining energy stores. Le Magnen (1) suggested tha~L the stimulation to eat arises from both internal or metabolic signals, which underly the state of hunger, and the orosensory features of the food. The appraisal of the orosensory features of the food as positive (i.e., pleasant) is not an absolute property of the food, but rather depends upon energy status, previous experience of eating that food, and the hedonic response to the orosensory features of the food as it is perceived by the organism. If previous encounters with this food have b e e n positive, then the food is likely to be a p p r o a c h e d again (2). T h e hedonic response to a particular food begins during the preliminary stages of sensory processing (3). If the food does not look, smell or taste good it is unlikely to be eaten, although this evaluation :is d e p e n d e n t on deprivation state and food availability. As a food is eaten, there is a decrease in the pleasantness of the a p p e a r e a n c e , smell, texture and taste of that food, whereas other foods remain pleasant (4). This p h e n o m e n o n appear:~ to last for at least 1 h after eating (5). In addition, the rate of eating this food slows (6) and the likelihood of the same food being chosen 113
114
HETHERINGTON
Reasons for Stopping: First Course
Reasons for Stopping: Second Course reason
reason
(total sample)
m
had no secotxlcoupe" food tasted less pkasanr 1
food all gone"1
5
felt full"
felt full"
got tim5 cg food"
9
¢~hcr" 0
10
20
30
40
0
50
10
20
percentage
30
40
50
60
percentage
FIG. 1. The most important reason given for terminating the first course.
Reasons for Stopping: Second Course (those who had 2nd course) reason
influencing hunger (13). Under normal circumstances, without any pharmacological intervention, liking and wanting operate together and there may be minimal subjective awareness of the two different dimensions. It has been proposed that, although changes in pleasantness occur during a meal, consumers may be relatively unaware of this, or that these shifts are unimportant in terminating food intake (14). T o test this proposition a series of studies was conducted by Mook and Votaw (14) in which consumers were asked to give the main reason for terminating meals. Since the majority of participants selected fullness as the most common reason to stop eating and very few participants selected the hedonic option (i.e., the food stopped tasting as good, or the food tastes less good), the authors concluded that hedonic shifts are of little salience or importance in terminating a meal. However, a serious weakness of the three studies was that a meal was not offered to the consumers and so answers were based on m e m o r y rather than on the experience of just having eaten. The present study was devised to asssess reasons for stopping in participants who were eating a meal and who did not need to rely on memory. Also, since the previous work by Mook and Votaw did not record hedonic shifts, the present study also monitored food intake, and changes in pleasantness as well as the rank order of importance of various alternatives for terminating the meal. METHODS
Fifty-seven normal weight volunteers (45 women, 12 men) came to the laboratory for lunch, having eaten nothing since breakfast (minimum of 3 h deprivation). U p o n arrival, participants recorded their level of hunger, fullness, desire to eat (appetite) and the pleasantness of the taste of two foods on 100 m m visual analogue scales. The foods which were tasted and rated were cheese on cracker and chocolate. Participants were given ad lib access to a test meal of cheese on
food tasted less
food
got tinxl~ food
18.5
!
0
10
20
30
40
50
60
percentage
FIG. 2. Upper panel: reasons given for terminating the second course (total sample) and (lower panel) reasons given for terminating the second course (only those who consumed a second course).
crackers, and as much water as they wanted. When they had finished eating, participants were instructed to record the main reason for stopping from a possible seven statements. These statements were modified from Mook and Votaw (14) and were as follows: everyone else had finished eating, I had had as much as I am allowed; the food began to taste less pleasant; the food was all gone; I felt full; I got tired of eating the food; some other reason which the subject was asked to explain. They then rank ordered the importance of each reason where 1 was most important and 7 was least important. Participants rerated hunger, appetite, fullness and the pleasantness of the foods at 2, 20, 40 and 60 min after the meal. After the 60-min rating, the participants were offered ad lib access to a choice of the same food (cheese on crackers), a different food (chocolate), or no second course. Again when particpants had stopped eating, reasons for terminating the meal were recorded by those who selected a second course. Participants then rank ordered the importance of each reason.
SENSORY-SPECIFIC SATIETY
115
Energy Intake in the First Course
Changes in Pleasantness of Taste
(women only; n = 45)
Intake
:t
lO
[] eaten •
0"
uneaten
mean change
(mm)
-10"
(kcaO -20"
-30" -4O fatigue/hexlonics
fullness
fatigue/hedonics
fullness
Group
Group FIG. 3. Mean (±SEM) energy intake by group: those who ranked fatigue/hedonics as the most important factor and those who ranked fullness as the most important reason for terminating eating.
FIG. 4. Mean (±SEM) changes in pleasantness of the taste of the eaten (cheese on cracker) and uneaten (chocolate) foods by group.
RESULTS
Significantly more participants selected the different food (78%) for their second course compared to eating the same food (21%) again. This suggests that satiety was specific to the food eaten, and the pleasantnes and desire for the uneaten food was unchanged by eating the first course. Hedonic ratings did not differ across groups either before the meal began or in terms of the magnitude of changes in pleasantess as a function of eating (see Fig. 4). Similarly, fullness ratings prior to intake were not different. However, the group who rated fullness as the most important reason for terminating eating rated hunger and desire to eat as significantly higher (see Fig. 5).
The most common reason given for meal termination in the first course was "I got tired of eating that food" (40%), followed by "I felt full", then "food all gone," and then "food tasted less pleasant" (see Fig. 1). Over half of the participants did not have a second course, however, of those who did, the most important reason for stopping was "I felt full" (48%), followed by "got tired of the food", and "food tasted less pleasant" (see Fig. 2). In this experiment, tiring of the food being eaten was rated as important in determining meal termination in both courses. Therefore, this phenomenon was salient to the participants, just as fullness was salient and :noted as important, particularly following the second course. Since participants had to rank order all reasons, a mean importance rating was determined for each reason. The three highest mean ranks were: got tired of the food (2.2±0.19), food tasted less pleasant (2.8±0.18), and felt full (2.85±0.2). Following the second course, the three highest mean ranks were: felt full (2.17_+0.3), got tired of the food (2.3+0.2) and food tasted less pleasant (3.1+0.3). Participants were-divided into those who rated fatigue and changes in pleasantness as important and those who rated fullness as more important. Significant differences in intake between these groups indicated that those who rated fatigue/hedonics as important consumed significantly fewer calories (275+23 kcal) than those who rated :fullness as more important (424±65 kcal; see Fig. 2). Frequency of selecting a second course was equivalent in both groups (about half in each), thus of those who rated fullness as important in terminating the first course, about half were still able to consume a second course. Energy intake in the second course was not different between the fatigue/hedonics group (238.2±19.2 kcal) and the fullness group (182.0±33.9 kcal). Overall, there was no difference in energy intake between the two groups.
DISCUSSION
When participants were given a test meal to consume and asked to rate the reasons for terminating eating, getting tired of the food was rated as important by
Subjective Ratings 80 ¸ [] fatigue/hedonics • fullness
6O
*p < 0.05
rating
(mm) 4O 20
hunger
fullness
desireto eat
FIG. 5. Mean (±SEM) ratings of hunger, fullness and desire to eat by group.
116
HETHERINGTON
more consumers than fullness after a first course. However, of those who selected a second course, fullness was ranked as the most important reason for stopping. It may be argued that fatigue experienced by participants reflects sensory fatigue and that this may contribute to the development of sensory-specific satiety. Since participants who rated fullness as the most important reason for terminating the meal consumed more calories in the first course, this index of satiety may be relatively more crude than sensory or hedonic variables. However, since the energy intake overall was not different between groups, perhaps differences in assigning reasons for meal termination are a matter of eating style. Participants who rated fullness as most important were hungrier and it could be that, since they started off more hungry and ate more, they were paying more attention to the development of fullness. Participants who rated getting tired of the food as important were less hungry and ate less, perhaps they were attending more to the food itself rather than hunger and appetite. Both groups demonstrated strong sensory-specific satiety, and so, although they selected different reasons for meal termination, both experienced a decrease in the pleasantness of the taste of the eaten food relative to the uneaten food. Both groups tended to select a second course in equal numbers and both were more likely to choose a different food in the second course rather than the same food. This selection of a different food could be due to the stimulation presented by a food which is different from the one eaten before, or to the decline in pleasantness of the eaten food. Indeed, it is likely that the decline in the pleasantness of the eaten food contributes to the salience and the desire to eat the uneaten food. However, it is puzzling that half of those who ranked fullness as the most important reason for stopping eating nevertheless went on to eat a second course. Thus the participants got tired of the food, they reported this alongside changes in the hedonic assessment of the eaten food, and then after the second course fullness was selected as the most important reason for termination. This sheds some light on the findings by Mook and Votaw (14) that, if asked at the
end of a large meal (i.e., one containing two courses), then the most likely reason for meal termination is indeed fullness. However, if asked at the end of a first course, as satiety is developing, the main reason for stopping is not fullness, but rather that the subject tires of eating that food. Clearly, asking participants to rate their reasons for stopping eating during a meal will yield different answers to asking participants to rely upon their memory. Changes in pleasantness (hedonic reasons) and getting tired of that food may be relatively transient and not remembered, compared to the longer-lasting feeling of fullness. Moreover, lack of awareness of the contribution of hedonic changes to meal termination does not imply that these changes are not important in the process of satiation. In the present study, the majority of participants who selected a second course chose a different food and not the same food again. This demonstrates a tendency to select a food which differs in sensory characteristics to that just eaten (15). In behavioral terms, their actions suggested a preference for the food which was not eaten, and, in terms of pleasantness ratings, their judgements demonstrated a decline with time. It is likely that the two phenomena are connected. As satiety to the eaten food develops, so the pleasantness and desire for that food declines and when a second course is selected, a different food is eaten. In conclusion, conscious awareness of hedonic shifts may not be necessary for such shifts to influence intake of the food which has been eaten to satiety. Rather, participants were aware that they were tiring of the food and that following a second course they felt full. Thus, the development of satiety may be in two discrete stages for some consumers--satiety to a specific food, and then general satiety for the meal. This proposal has yet to be examined systematically. However, since changes in pleasantness can occur before absorption takes place, this suggests that hedonic shifts occur early on in the meal, and this may explain why the participants in the study by Mook and Votaw (14) remembered fullness rather than changes in pleasantness as important in meal termination. Future studies should examine these issues more closely, paying particular attention to individual eating styles and stages of satiety as it develops.
REFERENCES 1. Le Magnen, J. Palatability: concept, terminology, and mechanisms. In: Boakes, R. A.; Popplewell, D. A.; Burton, M. J., eds. Eating habits: food, physiology and learned behaviour. Chichester: Wiley; 131-154. 2. Rozin, P.; Kalat, J. W. Specific hungers and poison avoidance as adaptive specializations of learning. Psychol. Rev., 78:459--486; 1971. 3. Rolls, E. T. The neural control of feeding in primates. In: Booth, D. A., ed. Neurophysiology of ingestion. Oxford: Pergamon; 1993:137-169. 4. Roils, B. J. Sensory-specific satiety. Nutr. Rev. 44:93-101. 5. Hetherington, M.; Burley, V. J.; Rolls, B. J. The time course of sensory-specific satiety. Appetite 12:57-68; 1989. 6. Bellisle, F.; Lucas, F.; Amrani, R.; Le Magnen, J. Deprivation, palatability and the microstructure of meals in human subjects. Appetite 5:85-94; 1984. 7. Berridge, K. C. Modulation of taste affect by hunger, caloric
8. 9. 10.
11. 12.
satiety, and sensoryspecific satiety in the rat. Appetite 16:103-120; 1991. Swithers, S. E.; Hall, W. G. Does oral experience terminate ingestion? Appetite 23:113-138; 1994. Rolls, E. T.; Treves, A. The relative advantages of sparse versus distributed encoding for associative neuronal networks in the brain. Network 1:407--421; 1990. Hetherington, M. M.; Rolls, B. J. Sensory-specific satiety: theoretical frameworks and central characteristics. In: Capaldi, E. D.; Powley, T. L., eds. The psychology of eating. Washington, DC: American Psychological Society, in press. Berridge, K. C. Puzzles of reward: palatability, appetite, brain and consciousness. Paper presented to Society for the Study of Ingestive Behavior, McMaster University, 16 August; 1994. Blundell, J. E.; Hill, A. J. On the mechanism of action of dexfenfluramine: effect on alliesthesia and appetite motivation in lean and obese subjects. Clin. Neuropharmacol. 11:S121-S134; 1988.
SENSORY-SPECIFIC SATIETY 13. Fantino, M.; Hosotte, J.; Apfelbaum, M. An opioid antagonist, naltrexone, reduces preference for sucrose in humans. Am. J. Physiol. 251:R91-R96, 1986. 14. Mook, D. G.; Votaw, IV[. C. How important is hedonism? Reasons given by college students for ending a meal. Appetite
117 18:69-75; 1992. 15. Rolls, B. J.; Hetherington, M. M. The role of variety in eating and body weight regulation. In: Shepherd, R. ed. Handbook of the psychophysiology of human eating. Chichester: Wiley; 1989:57-84.
Pergamon
0149-7634(95)00048-8
Sensory-Specific Satiety and its Importance in Meal Termination M. M. H E T H E R I N G T O N
Department of Psychology, University of Dundee, DD1 4HN, Scotland, UK
HETHERINGTON, M.M. Sensory-specificsatiety and its importance in meal termination. NEUROSCI BIOBEHAV REV 20(1) 113-117, 1996. - - Pleasantness is important in influencing food choice, and may play a role in determining the amount of food consumed. Judgements of pleasantness decrease as the food is eaten. It has been proposed that this reflects the development of satiety to a specific food. However, consumers may not rate these changes as important in meal termination. Fifty-seven subjec,~swere given ad lib access to a test meal of cheese on crackers and at the end of this meal recorded the main reason for stopping from a possible seven statements. They then rank ordered the importance of each reason. One hour later, subjects were offered a choice of the same food, a different food, or no second course. Again reasons for stopping were recorded by those who selected a second course. The most common reason given for meal termination in the first course was "I got tired of eating that food" (40%) and for the second course "I felt full" (48%). Subjects were divided into those who rated fatigue and changes in pleasantness as important and those who rated fullness as more important. Significant differences in intake between these groups indicated that those who rated fatigue/hedonics as important consumed significantly fewer calories (275±23 kcal) than those who rated fullness as more important (424±65 kcal). It is argued that fatigue experienced by subjects may reflect sensory fatigue and that this is an important part of the development of sensory-specific satiety. Since subjects who rated gastric fullness as the most important reason for terminating the meat consumed more calories, it is suggested that this index of satiety may be relatively more crude than sensory or hedonic variables. Satiety
Food intake
Appetite
Sensory-specificsatiety
INTRODUCTION
again from an array of different foods is reduced (4). This dynamic aspect of pleasure as reflected in subjective judgements, rate of eating and food choice has b e e n n a m e d "sensory-specific satiety" (4). That satiety can be specific to a particular food as it is eaten appears to be a reliable p h e n o m e n o n and is observed across different species (7). However, the precise mechanisms which underly the expression of sensoryspecific satiety remain to be elucidated, although work on oral habituation (8) and neuronal models of food reward mechanisms (9) shed light on potentially important theoretical models (10). Berridge (11) has identified two distinguishable c o m p o n e n t s of the pleasure of eating. First of all liking, which is measured by pleasure responses in animals or subjective ratings of pleasantness in humans, and secondly, wanting, which is m e a s u r e d by assessing incentive motivation in animals or desire to eat (appetite for) a specific food in humans. Findings from neurophysiological studies in animals indicate that different regions of the brain control these two separable components of pleasure (11). Similarly, in humans pharmacological manipulations have revealed agents which will reduce hunger while sparing the pleasantness of foods (dfenfluramine; 12) and opioid blockade, which reduces the pleasantness of foods while not
E A T I N G IS highly pleasurable and of central importance in providing and maintaining energy stores. Le Magnen (1) suggested tha~L the stimulation to eat arises from both internal or metabolic signals, which underly the state of hunger, and the orosensory features of the food. The appraisal of the orosensory features of the food as positive (i.e., pleasant) is not an absolute property of the food, but rather depends upon energy status, previous experience of eating that food, and the hedonic response to the orosensory features of the food as it is perceived by the organism. If previous encounters with this food have b e e n positive, then the food is likely to be a p p r o a c h e d again (2). T h e hedonic response to a particular food begins during the preliminary stages of sensory processing (3). If the food does not look, smell or taste good it is unlikely to be eaten, although this evaluation :is d e p e n d e n t on deprivation state and food availability. As a food is eaten, there is a decrease in the pleasantness of the a p p e a r e a n c e , smell, texture and taste of that food, whereas other foods remain pleasant (4). This p h e n o m e n o n appear:~ to last for at least 1 h after eating (5). In addition, the rate of eating this food slows (6) and the likelihood of the same food being chosen 113
114
HETHERINGTON
Reasons for Stopping: First Course
Reasons for Stopping: Second Course reason
reason
(total sample)
m
had no secotxlcoupe" food tasted less pkasanr 1
food all gone"1
5
felt full"
felt full"
got tim5 cg food"
9
¢~hcr" 0
10
20
30
40
0
50
10
20
percentage
30
40
50
60
percentage
FIG. 1. The most important reason given for terminating the first course.
Reasons for Stopping: Second Course (those who had 2nd course) reason
influencing hunger (13). Under normal circumstances, without any pharmacological intervention, liking and wanting operate together and there may be minimal subjective awareness of the two different dimensions. It has been proposed that, although changes in pleasantness occur during a meal, consumers may be relatively unaware of this, or that these shifts are unimportant in terminating food intake (14). T o test this proposition a series of studies was conducted by Mook and Votaw (14) in which consumers were asked to give the main reason for terminating meals. Since the majority of participants selected fullness as the most common reason to stop eating and very few participants selected the hedonic option (i.e., the food stopped tasting as good, or the food tastes less good), the authors concluded that hedonic shifts are of little salience or importance in terminating a meal. However, a serious weakness of the three studies was that a meal was not offered to the consumers and so answers were based on m e m o r y rather than on the experience of just having eaten. The present study was devised to asssess reasons for stopping in participants who were eating a meal and who did not need to rely on memory. Also, since the previous work by Mook and Votaw did not record hedonic shifts, the present study also monitored food intake, and changes in pleasantness as well as the rank order of importance of various alternatives for terminating the meal. METHODS
Fifty-seven normal weight volunteers (45 women, 12 men) came to the laboratory for lunch, having eaten nothing since breakfast (minimum of 3 h deprivation). U p o n arrival, participants recorded their level of hunger, fullness, desire to eat (appetite) and the pleasantness of the taste of two foods on 100 m m visual analogue scales. The foods which were tasted and rated were cheese on cracker and chocolate. Participants were given ad lib access to a test meal of cheese on
food tasted less
food
got tinxl~ food
18.5
!
0
10
20
30
40
50
60
percentage
FIG. 2. Upper panel: reasons given for terminating the second course (total sample) and (lower panel) reasons given for terminating the second course (only those who consumed a second course).
crackers, and as much water as they wanted. When they had finished eating, participants were instructed to record the main reason for stopping from a possible seven statements. These statements were modified from Mook and Votaw (14) and were as follows: everyone else had finished eating, I had had as much as I am allowed; the food began to taste less pleasant; the food was all gone; I felt full; I got tired of eating the food; some other reason which the subject was asked to explain. They then rank ordered the importance of each reason where 1 was most important and 7 was least important. Participants rerated hunger, appetite, fullness and the pleasantness of the foods at 2, 20, 40 and 60 min after the meal. After the 60-min rating, the participants were offered ad lib access to a choice of the same food (cheese on crackers), a different food (chocolate), or no second course. Again when particpants had stopped eating, reasons for terminating the meal were recorded by those who selected a second course. Participants then rank ordered the importance of each reason.
SENSORY-SPECIFIC SATIETY
115
Energy Intake in the First Course
Changes in Pleasantness of Taste
(women only; n = 45)
Intake
:t
lO
[] eaten •
0"
uneaten
mean change
(mm)
-10"
(kcaO -20"
-30" -4O fatigue/hexlonics
fullness
fatigue/hedonics
fullness
Group
Group FIG. 3. Mean (±SEM) energy intake by group: those who ranked fatigue/hedonics as the most important factor and those who ranked fullness as the most important reason for terminating eating.
FIG. 4. Mean (±SEM) changes in pleasantness of the taste of the eaten (cheese on cracker) and uneaten (chocolate) foods by group.
RESULTS
Significantly more participants selected the different food (78%) for their second course compared to eating the same food (21%) again. This suggests that satiety was specific to the food eaten, and the pleasantnes and desire for the uneaten food was unchanged by eating the first course. Hedonic ratings did not differ across groups either before the meal began or in terms of the magnitude of changes in pleasantess as a function of eating (see Fig. 4). Similarly, fullness ratings prior to intake were not different. However, the group who rated fullness as the most important reason for terminating eating rated hunger and desire to eat as significantly higher (see Fig. 5).
The most common reason given for meal termination in the first course was "I got tired of eating that food" (40%), followed by "I felt full", then "food all gone," and then "food tasted less pleasant" (see Fig. 1). Over half of the participants did not have a second course, however, of those who did, the most important reason for stopping was "I felt full" (48%), followed by "got tired of the food", and "food tasted less pleasant" (see Fig. 2). In this experiment, tiring of the food being eaten was rated as important in determining meal termination in both courses. Therefore, this phenomenon was salient to the participants, just as fullness was salient and :noted as important, particularly following the second course. Since participants had to rank order all reasons, a mean importance rating was determined for each reason. The three highest mean ranks were: got tired of the food (2.2±0.19), food tasted less pleasant (2.8±0.18), and felt full (2.85±0.2). Following the second course, the three highest mean ranks were: felt full (2.17_+0.3), got tired of the food (2.3+0.2) and food tasted less pleasant (3.1+0.3). Participants were-divided into those who rated fatigue and changes in pleasantness as important and those who rated fullness as more important. Significant differences in intake between these groups indicated that those who rated fatigue/hedonics as important consumed significantly fewer calories (275+23 kcal) than those who rated :fullness as more important (424±65 kcal; see Fig. 2). Frequency of selecting a second course was equivalent in both groups (about half in each), thus of those who rated fullness as important in terminating the first course, about half were still able to consume a second course. Energy intake in the second course was not different between the fatigue/hedonics group (238.2±19.2 kcal) and the fullness group (182.0±33.9 kcal). Overall, there was no difference in energy intake between the two groups.
DISCUSSION
When participants were given a test meal to consume and asked to rate the reasons for terminating eating, getting tired of the food was rated as important by
Subjective Ratings 80 ¸ [] fatigue/hedonics • fullness
6O
*p < 0.05
rating
(mm) 4O 20
hunger
fullness
desireto eat
FIG. 5. Mean (±SEM) ratings of hunger, fullness and desire to eat by group.
116
HETHERINGTON
more consumers than fullness after a first course. However, of those who selected a second course, fullness was ranked as the most important reason for stopping. It may be argued that fatigue experienced by participants reflects sensory fatigue and that this may contribute to the development of sensory-specific satiety. Since participants who rated fullness as the most important reason for terminating the meal consumed more calories in the first course, this index of satiety may be relatively more crude than sensory or hedonic variables. However, since the energy intake overall was not different between groups, perhaps differences in assigning reasons for meal termination are a matter of eating style. Participants who rated fullness as most important were hungrier and it could be that, since they started off more hungry and ate more, they were paying more attention to the development of fullness. Participants who rated getting tired of the food as important were less hungry and ate less, perhaps they were attending more to the food itself rather than hunger and appetite. Both groups demonstrated strong sensory-specific satiety, and so, although they selected different reasons for meal termination, both experienced a decrease in the pleasantness of the taste of the eaten food relative to the uneaten food. Both groups tended to select a second course in equal numbers and both were more likely to choose a different food in the second course rather than the same food. This selection of a different food could be due to the stimulation presented by a food which is different from the one eaten before, or to the decline in pleasantness of the eaten food. Indeed, it is likely that the decline in the pleasantness of the eaten food contributes to the salience and the desire to eat the uneaten food. However, it is puzzling that half of those who ranked fullness as the most important reason for stopping eating nevertheless went on to eat a second course. Thus the participants got tired of the food, they reported this alongside changes in the hedonic assessment of the eaten food, and then after the second course fullness was selected as the most important reason for termination. This sheds some light on the findings by Mook and Votaw (14) that, if asked at the
end of a large meal (i.e., one containing two courses), then the most likely reason for meal termination is indeed fullness. However, if asked at the end of a first course, as satiety is developing, the main reason for stopping is not fullness, but rather that the subject tires of eating that food. Clearly, asking participants to rate their reasons for stopping eating during a meal will yield different answers to asking participants to rely upon their memory. Changes in pleasantness (hedonic reasons) and getting tired of that food may be relatively transient and not remembered, compared to the longer-lasting feeling of fullness. Moreover, lack of awareness of the contribution of hedonic changes to meal termination does not imply that these changes are not important in the process of satiation. In the present study, the majority of participants who selected a second course chose a different food and not the same food again. This demonstrates a tendency to select a food which differs in sensory characteristics to that just eaten (15). In behavioral terms, their actions suggested a preference for the food which was not eaten, and, in terms of pleasantness ratings, their judgements demonstrated a decline with time. It is likely that the two phenomena are connected. As satiety to the eaten food develops, so the pleasantness and desire for that food declines and when a second course is selected, a different food is eaten. In conclusion, conscious awareness of hedonic shifts may not be necessary for such shifts to influence intake of the food which has been eaten to satiety. Rather, participants were aware that they were tiring of the food and that following a second course they felt full. Thus, the development of satiety may be in two discrete stages for some consumers--satiety to a specific food, and then general satiety for the meal. This proposal has yet to be examined systematically. However, since changes in pleasantness can occur before absorption takes place, this suggests that hedonic shifts occur early on in the meal, and this may explain why the participants in the study by Mook and Votaw (14) remembered fullness rather than changes in pleasantness as important in meal termination. Future studies should examine these issues more closely, paying particular attention to individual eating styles and stages of satiety as it develops.
REFERENCES 1. Le Magnen, J. Palatability: concept, terminology, and mechanisms. In: Boakes, R. A.; Popplewell, D. A.; Burton, M. J., eds. Eating habits: food, physiology and learned behaviour. Chichester: Wiley; 131-154. 2. Rozin, P.; Kalat, J. W. Specific hungers and poison avoidance as adaptive specializations of learning. Psychol. Rev., 78:459--486; 1971. 3. Rolls, E. T. The neural control of feeding in primates. In: Booth, D. A., ed. Neurophysiology of ingestion. Oxford: Pergamon; 1993:137-169. 4. Roils, B. J. Sensory-specific satiety. Nutr. Rev. 44:93-101. 5. Hetherington, M.; Burley, V. J.; Rolls, B. J. The time course of sensory-specific satiety. Appetite 12:57-68; 1989. 6. Bellisle, F.; Lucas, F.; Amrani, R.; Le Magnen, J. Deprivation, palatability and the microstructure of meals in human subjects. Appetite 5:85-94; 1984. 7. Berridge, K. C. Modulation of taste affect by hunger, caloric
8. 9. 10.
11. 12.
satiety, and sensoryspecific satiety in the rat. Appetite 16:103-120; 1991. Swithers, S. E.; Hall, W. G. Does oral experience terminate ingestion? Appetite 23:113-138; 1994. Rolls, E. T.; Treves, A. The relative advantages of sparse versus distributed encoding for associative neuronal networks in the brain. Network 1:407--421; 1990. Hetherington, M. M.; Rolls, B. J. Sensory-specific satiety: theoretical frameworks and central characteristics. In: Capaldi, E. D.; Powley, T. L., eds. The psychology of eating. Washington, DC: American Psychological Society, in press. Berridge, K. C. Puzzles of reward: palatability, appetite, brain and consciousness. Paper presented to Society for the Study of Ingestive Behavior, McMaster University, 16 August; 1994. Blundell, J. E.; Hill, A. J. On the mechanism of action of dexfenfluramine: effect on alliesthesia and appetite motivation in lean and obese subjects. Clin. Neuropharmacol. 11:S121-S134; 1988.
SENSORY-SPECIFIC SATIETY 13. Fantino, M.; Hosotte, J.; Apfelbaum, M. An opioid antagonist, naltrexone, reduces preference for sucrose in humans. Am. J. Physiol. 251:R91-R96, 1986. 14. Mook, D. G.; Votaw, IV[. C. How important is hedonism? Reasons given by college students for ending a meal. Appetite
117 18:69-75; 1992. 15. Rolls, B. J.; Hetherington, M. M. The role of variety in eating and body weight regulation. In: Shepherd, R. ed. Handbook of the psychophysiology of human eating. Chichester: Wiley; 1989:57-84.