- Email: [email protected]
Evaluation of Satiety Sensations and Food Intake After Different Preloads
app p066
12-07-95 15:10:46
Appetite, 1995, 25, 17–30
Evaluation of Satiety Sensations and Food Intake After Different Preloads
MARISA PORRINI, ROBERTA CROVETTI and GIULIO TESTOLIN DISTAM, sez. Nutrizione, Universita’ degli Studi di Milano, via Celoria 2, 20133 Milano, Italia
SANTINO SILVA USSL 32 Treviglio, P. le Ospedale, 1, 24047 Treviglio (BG), Italia
The reproducibility of three questions, related to fullness, satiety and desire to eat, rated on an unmarked triangle was verified. In four sessions 12 volunteers ate pasta with tomato sauce (520 kcal) and were asked to rate the sensations felt. There was no difference in rating scores of the replications so the proposed questionnaire provides a stable measure of sensations related to satiety. Subsequently three satiety conditions were studied. Two foods, one rich in carbohydrate, pasta (baked macaroni) and the other in protein, polpette (meatballs), were used as loads at two calorie levels and as preload before an ‘‘ad libitum’’ meal. All the three questions proved useful in discriminating between the different satiety conditions. The food intake underlines the specificity of satiety: subjects, after eating a preload which previously had satiated them, ate other foods in different amounts depending on the kind of preload eaten. Food intake was significantly higher after the pasta preload, furthermore ‘‘fullness’’ and ‘‘satiety’’ ratings were significantly highest after the meatball preload, suggesting that in our experimental conditions, meatballs were more satiating than pasta. In conclusion, this study highlights the validity of using several quite different questions to study hunger and satiety, together with the actual food intake. 1995 Academic Press Limited
I The evaluation of hunger and satiety sensations is difficult as it appears that there are no universally accepted measurements for such sensations (Louis-Sylvestre et al., 1989). Two rating methods are often used to quantify psychological and physiological sensations experienced by subjects: fixed-point scales and visual-analogue scales. For a detailed description and discussion of such methods the reader is referred to two reviews by Blundell and Hill (1988) and Teghtsoonian et al. (1981). One of the problems faced in trying to rate sensations of hunger and satiety is the existence of sensory-specific satiety. Sensory-specific satiety has been studied by Research was supported by Grant n.91.00223.PF41 from the National Research Council, Targeted Project ‘‘Prevention and Control Disease Factors’’, Subproject ‘‘Nutrition’’. Address correspondence to: Marisa Porrini, Department of Food Science and Technology and Microbiology, via Celoria 2, 20133 Milan, Italy. 0195–6663/95/040017+14 $12.00/0
1995 Academic Press Limited
18
M. PORRINI ET AL.
Rolls et al. (1981a, 1981b) who demonstrated that if several foods, different in taste, appearance and texture, are offered in succession, the consumption of the meal will be higher than when only one food is given. Thus, overeating may occur when a wide variety of foods is readily available. Difficulty discriminating between sensations felt in a ‘‘sensory-specific satiety’’ condition and those felt in a ‘‘general satiety’’ state could explain, in part, the poor correlation between sensation ratings accompanying ingestion and food consumption reported by some authors (Blundell & Rogers, 1980; Silverstone & Stunkard, 1968; Wardle, 1987). As data from psychophysical methods are not univocal, nor is the information complete, descriptive and operational studies of eating should include both the subject’s evaluation of hunger and satiety and assessment of calorie consumption. The use of different questions concerning both the sensation of satiety and the perspective consumption could lead to more complete information about the real state of the subject. The primary purpose of our study was to verify whether three different questions (related to fullness, satiety and desire to eat) rated on an unmarked triangle would be useful to scale perceived satiety after both a single food and a complete meal. Two foods, different in protein and carbohydrate content as well as in sensory properties, were tested to verify the consistency of the method in evaluating their satiating power. M Subjects Twelve healthy males, 23–26 years old, of normal body weight took part in the study. Their body mass index [(weight in kg) (height in m)−2 ] was between 19·7 and 23·0. Subject Selection The subjects were recruited from a graduate lecture course. Subjects receiving medication, smokers and dieters were excluded. The potential subjects were asked to complete a questionnaire on their eating habits and their preferences (none, normal and high liking) for 92 foods that were listed in the questionnaire. The selected subjects had a normal preference for the foods included in the study. Satiety Ratings Questions asked for information on sensations related to hunger and satiety were: ‘‘How satiated do you feel?’’, ‘‘How full do you feel?’’ and ‘‘How great is your desire to eat?’’. The subjects were asked to rate their sensations by drawing a line, parallel to the baseline, across an isosceles triangle (height 15 cm, base 3 cm), oriented horizontally on the paper with the base to the right. The triangle was unbroken and was marked with a word anchor at the apex to indicate the minimum (not at all) of the experienced sensation and at the baseline for the maximum (extremely). The ratings were expressed in centimetres squared. Procedure The subjects were instructed not to eat after 2200 hrs the previous evening and to arrive at the laboratory at 1230 hrs for a prepared lunch. The instruction to miss
19
SATIETY SENSATIONS AND FOOD INTAKE
T 1 Nutrient composition of foods (referred to 100 g of product) Energy (kcal)
Fat (g)
Protein (g)
CHO (g)
Fibre (g)
Pasta Meatballs
132 146
3·9 4·1
5·7 20·2
18·4 7·1
1·4 0·4
Mozzarella Cracker Tart
249 425 412
19·0 11·7 15·4
19·5 8·9 4·3
0·0 70·9 64·1
0·0 2·9 0·9
T 2 Physical characteristics of preloads Firmness (kg) 1∗ Pasta Meatballs
2∗
Volume (ml) Weight (g)
3∗
21·8 17·6 13·1 36·7 31·3 27·2
950 560
727 600
∗ Successive measures on the same sample.
breakfast was designed to ensure that all of the subjects would be similarly deprived at the start of the experiment and that they would be readily motivated to eat. On arrival at the university dining room, where professors eat lunch, the subjects were seated at small, individual and well-separated tables, thus minimizing distraction. The subjects were asked to be silent and to remain at their respective tables for the duration of the experiment. The investigation was divided into two principal experiments. Experiment I This was performed to verify the reproducibility of the answers to the three questions used to rate satiety. In four different sessions on non-consecutive days, the subjects were asked to eat pasta with tomato sauce (520 kcal). Before starting the session, blood tests were taken to measure fasting blood glucose in order to ensure the same physiological state for each replication. Immediately before lunch, and immediately after, the subjects were asked to rate their ‘‘satiety’’, ‘‘fullness’’ and ‘‘desire to eat’’ on the questionnaire being tested. After eating they also rated the pleasantness of the food on a nine-point hedonic scale. To avoid any possibility of memorizing the ratings the sessions were mixed with other experiments which were in no way connected with the present experiment. Experiment II In this experiment, three different satiety conditions were studied. Two different foods were used: one rich in carbohydrate, pasta pasticciata (baked macaroni), and the other in protein, polpette (meatballs seasoned with tomato sauce). Nutrient composition, determined by chemical analysis, is reported in Table 1; physical characteristics are shown in Table 2.
20
M. PORRINI ET AL.
In order to decide the caloric levels of the meals a preliminary session was done. On two different days, subjects were instructed to eat pasta and meatballs until they felt ‘‘comfortably full’’ (‘‘specific’’ satiety condition). The mean pasta intake was 960 kcal, that of meatballs 880 kcal. After this test, the subjects with consumption higher or lower than two standard deviations from the mean were omitted. Ten of the 12 subjects were recruited for this study. Only completed data sets were included in the statistical analysis. The experiment consisted of three sessions for each food, which the subjects consumed in random order: first session, the food at the calorie level corresponding to their ‘‘specific’’ satiety (pasta at 960 kcal and meatballs at 880 kcal); second session, the food at half the satiating calorie level (pasta at 480 kcal and meatballs at 440 kcal); third session, the food at the larger calorie level followed by an ad libitum meal (test meal) of mozzarella cheese, crackers and fruit tart. The subjects had no knowledge of the calorie level of the different meals nor of the criteria followed in choosing the food. Test meals, together with 300 ml of water, were served on individual trays to all the subjects at the same time. Immediately before, immediately after and 2 hr after the meal the subjects were asked to rate their ‘‘satiety’’, ‘‘fullness’’ and ‘‘desire to eat’’ on the questionnaire they found on the table. Glucose Determination Blood (200 ll), taken by pricking the finger tip with a lancet, was collected in microtubes containing heparin. Glucose was measured using a Glucose and LLactate Analyzer (model 2300 STAT, Yellow Springs Instrument Co, U.S.A.). Physical Determination Texture was determined by means of an Instron Universal Machine (model 4301) equipped with a share press. To simulate mastication three successive measures on about 50 g of food were made on the same sample. Data reported in Table 2 are the means of four determinations. Volume was determined by displacement: the food was put into a cylinder full of water and the increase in volume noted. Weight was measured by a digital scale (Hanson LDT, U.K.) corrected to 2 g. Data Analysis Data were analysed using a repeated-measures design. In the first experiment, the reproducibility of the rating scores for the three questions were tested by analysis of covariance with replications as factors and blood glucose and palatability as covariates. Such an approach allowed the adjustment of rating values for variables that could not be standardized. To confirm the reproducibility of the questionnaire correlation analysis was employed. In the first and second sessions of the second experiment the changes in ‘‘desire to eat’’, ‘‘fullness’’ and ‘‘satiety’’ were analysed separately using a three-factor analysis of variance with food type (meatballs or pasta), calorie level (specific satiated or half satiated) and time as factors. Means of the different scores were compared using the Newman–Keuls multiple range test to highlight any significant difference. In the
SATIETY SENSATIONS AND FOOD INTAKE
21
third session of the second experiment changes in satiety sensations were analysed using an analysis of variance with the type of preload (meatballs or pasta pasticciata) and time as factors. The consumption of each item in the ad libitum meal and the test meal calorie intake were compared with paired t-test. The paired t-test was also used to compare the satiety sensation ratings expressed in the first and third sessions for each assessment time. Relationships among the dependent variables were analysed by means of simple linear regression.
F 1. Means (SD) of (A) ‘‘Desire to eat’’, (B) ‘‘Fullness’’, and (C) ‘‘Satiety’’ ratings, (Φ) before and (Ε) after lunch, in the four replications.
22
M. PORRINI ET AL.
R Experiment I Blood glucose and palatability ratings Mean fasting glucose levels (mg dl−1) were 71·8 (±5·4), 74·0 (±6·2), 73·5 (±4·4) and 74·8 (±5·0) in the four replications. Palatability ratings were respectively 7·7 (±0·9), 7·7 (±0·8), 7·4 (±1·0) and 7·1 (±0·9). Subjective ratings In Fig. 1 means and standard deviations of ‘‘desire to eat’’, ‘‘fullness’’ and ‘‘satiety’’ ratings obtained in the four replications are reported. There was no difference in the rating scores either on the change between before and after lunch or considering only data obtained after eating (significance levels always higher than 0·5). These results were confirmed by correlation analysis. The correlation coefficients were highly significant between the same sensation score in the different replications (Table 3). Subjective ratings expressed after eating were always highly intercorrelated. The correlation coefficients (r) were −0·6253 between ‘‘desire to eat’’ and ‘‘fullness’’, −0·6147 between ‘‘desire to eat’’ and ‘‘satiety’’ and 0·9315 between ‘‘satiety’’ and ‘‘fullness’’ (sample size=48, p<0·0001). Experiment II Food intake in the preliminary session As previously reported, the mean consumption of pasta in the preliminary session was 960 kcal (SD=151) and that of meatballs was 880 kcal (SD=185). This difference T 3 Correlation coefficients and significance levels between scores for each replication Desire to eat
1
Fullness
2
3
4
0·8468 0·0005
0·8373 0·0007 0·8732 0·0002
0·7712 0·0033 0·8642 0·0003 0·9460 0·0000
2 3
Satiety
1 2
2
3
4
0·7632 0·0039
0·7938 0·0021 0·9759 0·0000
0·8995 0·0001 0·9017 0·0001 0·9072 0·0000
3 1, 2, 3, 4=replications’ number. Sample size=12.
1 2 3
2
3
4
0·7138 0·0091
0·7178 0·0086 0·9445 0·0000
0·9051 0·0001 0·8364 0·0007 0·8494 0·0005
SATIETY SENSATIONS AND FOOD INTAKE
23
was not significant. All subjects rated both preloads as pleasant, and the scores were not statistically different (7·5 ±0·7 and 7·1 ±0·6 respectively for pasta and meatballs). Satiety ratings in the first and second sessions Fasting ratings for each of the three questions ‘‘How great is your desire to eat?’’, ‘‘How full do you feel?’’ and ‘‘How satiated do you feel?’’ obtained in the different sessions were compared to establish that baseline levels were not different. No significant difference was identified. In Fig. 2 the ratings for each question were plotted as a function of time. As can be noted, the effect of time on all the satiety sensations was similar for all the loads. In fact the mean changes for each score from before and after each load showed a significant main effect of time, for ‘‘desire to eat’’, F(2,120)=106·148, p<0·0001, for fullness ratings, F(2,120)=70·881, p<0·0001 and for satiety ratings, F(2,120)=63·609, p<0·0001. For ‘‘desire to eat’’, analysis of variance showed a significant effect of calorie levels, F(1,120)=24·260, p<0·0001. Analysis by Newman–Keuls confirmed that following the smaller energy load the ‘‘desire to eat’’ was significantly higher than following the larger one. There was no effect of the type of food eaten on this sensation. There was no significant interaction between the three experimental factors (food type, calorie level and time). As regards ‘‘fullness’’ there was a significant interaction calorie×time, F(1,120)= 13·221, p<0·0001; furthermore a significant effect of food eaten was present, F(1, 120)=5·175, p=0·0249. Newman–Keuls analysis of overall changes in ‘‘fullness’’ showed that this sensation was significantly higher after a larger calorie load. Following the meatball load, ‘‘fullness’’ was significantly higher than when pasta was eaten. As regards the question ‘‘How satiated do you feel?’’, the interaction calorie×time was still significant, F(1,120)=12·424, p<0·0001, but no significant effect of type of food was present. In Table 4, the correlation coefficients and significance levels between subjective scores for each assessment time are shown. As can be noted the ratings were always highly intercorrelated for both preloads. Food intake in the third session Figure 3 shows the subjects’ consumption of the different foods and calorie intake in the ad libitum meal. The subjects significantly altered their intake according to the type of preload eaten: they had greater intake following pasta (811±226 kcal) than following meatballs (438±185 kcal). Paired comparison analysis demonstrated significant differences across the two test meals for energy intake, F(1,15)=13·100, p=0·0028, mozzarella cheese, F(1,15)=8·199, p=0·0125, crackers, F(1,15)=7·203, p=0·0178, and apricot jam tart, F(1,15)=5·283, p=0·0375. Subjective ratings in the third session In Fig. 4 the ratings for each question were plotted as a function of time. As expected, the ratings of ‘‘desire to eat’’, ‘‘fullness’’ and ‘‘satiety’’ expressed after the two ad libitum meals were not different. There was a significant change for each sensation with time: respectively, F(2,48)=101·781, p<0·0001, F(2,48)=192·592, p<0·0001, F(2,48)=114·364, p<0·0001 for desire to eat, fullness and satiety ratings.
24
M. PORRINI ET AL.
F 2. Mean (+SEM) (A) ‘‘Desire to eat’’, (B) ‘‘Fullness’’ and (C) ‘‘Satiety’’ ratings just before (0), 30 and 120 min after eating the four different loads. Μ, larger pasta load; Α, smaller pasta load; Χ, larger meatball load; Β, smaller meatball load.
Satiety ratings expressed in height and area Relationships between intake of different amounts of foods studied (between 440 and 1500 kcal) and satiety ratings obtained 30 min after eating and expressed by both height and area are reported in Table 5. A good linear relation was always present.
25
SATIETY SENSATIONS AND FOOD INTAKE
T 4 Correlation coefficients between subjective scores for each assessment time
Desire to eat−Fullness Desire to eat−Satiety Satiety−Fullness
0 min
30 min
120 min
−0·5926 −0·7254 0·6383
−0·8179 −0·8061 0·8895
−0·7619 −0·7548 0·9679
Sample size=32. p always lower than 0·0001.
T 5 Correlation coefficients between satiety scores and calorie intake
Desire to eat−Calorie Fullness−Calorie Satiety−Calorie
Height
Area
−0·7707 0·7095 0·7140
−0·7268 0·7204 0·7291
Sample size=48. p always lower than 0·0001.
F 3. Test meal intake (+SEM) and consumption (+SEM) of the three ad libitum foods after pasta (Φ) and meatball (∆) preloads.
Comparison between ratings in the first and third sessions Paired t-test conducted between satiety sensations expressed after the larger meatball load and after the test meal with meatballs as preload did not show differences. The same comparison conducted for pasta indicated that ‘‘satiety’’ was higher after the test meal, 21·2±1·3 vs. 14·3±1·5, t(8)=2·37, p=0·0390, whereas
26
M. PORRINI ET AL.
F 4. Mean (+SEM) ratings just before (0), and 30 and 120 min after eating the two larger preloads, pasta (Α) and meatball (Β), plus test meal.
‘‘desire to eat’’ was higher after the larger pasta load, 1·1±0·3 vs. 0·1±0·0, t(8)= 3·26, p=0·0085. There was no statistically significant difference for fullness sensation between the two pasta sessions. C From the results of the first experiment, the subjective scores of ‘‘desire to eat’’, ‘‘fullness’’ and ‘‘satiety’’, assessed from the proposed questionnaire, provided a stable
SATIETY SENSATIONS AND FOOD INTAKE
27
measure of perceived sensations. Particular care was given to choosing the rating scale. Fixed-point scales were rejected so as to avoid any recall of previous rates and end-effect. As reported by Teghtsoonian et al. (1981), numbers are easily remembered and might be used by the subject to reflect not only his feelings of satiety, but also his preconception of what he should feel during a meal. Lengths, and consequently areas, are hard to remember so each length- or areasatiety match would be more directly perceivable. However instead of the usual line of visual-analogue scales we preferred to use a triangle to quantify the sensation felt. The fact that in the triangle the area increases helps subjects to better quantify hunger and satiety sensations which have an increasing and decreasing trend (for example the sensation of filling the stomach). To analyse whether any distortion of the scaling of the sensation occurred by using the area of the triangle instead of its height, we assessed the relationships between different amounts of foods studied in relation to satiety scores (see Table 5). As with both height and area there was a good linear relation between intake and satiety ratings we think that using area could help in quantifying sensation felt without distorting the results. On the whole, the questionnaire was helpful for both comparing the satiating effects of different foods and studying different sated conditions. In particular, subjects interpreted the scoring method very well and were able to rate, in a reproducible way, the sensations they felt after eating the food put before them. The ratings obtained for satiety, fullness and desire to eat were shown to be significantly affected by time and are therefore influenced by deprivation time. Furthermore, unlike Wooley et al. (1975), who found that subjective ratings vary predictably with deprivation time but fail to differentiate between larger and smaller energy preloads, we were able to discriminate between the sensations felt after two loads of the same food with different energy content. Although it is difficult to understand the validity of using several subjective ratings to study the different sated conditions, an analysis of the results has led to some considerations. In the condition where the subjects were not fully satiated, like that of the first experiment, the fullness and satiety ratings were negatively correlated with ‘‘desire to eat’’, whereas the correlation between ‘‘fullness’’ and ‘‘satiety’’ was positive. It may be possible that fullness and satiety are perceived as the same sensation. Similar correlations were also found in the first and second sessions of the second experiment where the subjects were not completely satiated. However a significant effect of the kind of food eaten was present for ‘‘fullness’’, but not for ‘‘satiety’’ and ‘‘desire to eat’’. The ratings obtained in the third session could not be intercorrelated as at the measurement extremes there was insufficient variability in the ratings. As expected, there was no difference between pasta and meatballs session sensations. On the other hand comparison of the data for sensation in the ‘‘specific’’ satiety condition (first session) with data in the ‘‘general’’ satiety condition (third session) indicates that ‘‘satiety’’ and ‘‘desire’’ ratings were different in the two conditions, at least as far as pasta was concerned. The differences in the sensations felt were confirmed by the food intake data. After the larger pasta preload, the subjects ate more than after the larger meatball preload, confirming that the differences reflected more than an occasional finding.
28
M. PORRINI ET AL.
In summary, ‘‘fullness’’ was influenced by food but not by the different satiation conditions. However, ‘‘satiety’’ and ‘‘desire to eat’’ were not influenced by the kind of food eaten but were different in the ‘‘specific’’ and ‘‘general’’ satiety conditions. Clearly, many variables influenced subjects in rating their sensations. Thus, after a monotonous meal, the desire to eat remained at a high level even though the amount consumed (as regards both calories and weight) had induced the subjects to stop eating. Satiety sensation could unconsciously be influenced by the variety of the meal or by other characteristics of the food itself, such as its rheological properties. Consequently the three sensations did not completely overlap in describing the satiation condition, especially when, in addition to physiological variables, other variables are present. Few studies ask ‘‘How satiated do you feel?’’ (Blundell & Rogers, 1980; Kissileff et al., 1984). Our results indicate the importance of rating each sensation accompanying consumption separately. It would seem that the use of a questionnaire alone is inadequate to have really precise and complete results when studying hunger and satiety; consequently, we think that it should also be important to determine the actual intake by weighing the food consumed. In this study the results of food intake confirmed the specificity of satiety. After the consumption of a specific food of a given calorie level, which was sufficient to satiate the subjects in a preliminary session, the same subjects ate other foods in different amounts depending on the kind of preload eaten. In the experimental conditions used, meatballs were more satiating than pasta. Food intake was significantly higher after the pasta preload for all the three foods eaten (cheese, crackers and tart). The tart, which had high hedonistic characteristics and whose consumption could be independent of a real hunger stimulus, was consumed in higher amounts after pasta, confirming the different effect of the two preloads. Subjective ratings to the three questions confirmed the fact that meatballs were more satiating than pasta, in fact both in the ‘‘specific satiated’’ and in the ‘‘half satiated’’ conditions, fullness and satiety ratings are highest after meatballs. The two foods have the same fat content but reciprocal amounts of protein and carbohydrate (starch). Furthermore the meatball preload had a lower weight, similar calories, but lower volume and higher firmness. It follows, that the higher satiating effect of meatballs could be ascribed to the nutrient composition and/or the consistency. To clarify these differences, additional studies are needed involving food different in nutritive density and rheological parameters (volume vs. calorie, consistency vs. calorie etc.). There continues to be scientific debate over the satiating properties of macronutrients. In some studies high protein foods produce satiety more readily than protein-poor foods (Blundell & Hill, 1985; Booth et al., 1970). High carbohydrate foods appear to stimulate appetite and hunger more than high fat or high protein foods (Geiselman & Novin, 1982; Rodin et al., 1985). In contrast, other studies found no differences between the satiating properties of protein and carbohydrates (Geliebter, 1979; Panksep, 1971; Sunkin & Garrow, 1982), while they suggested that fat may have the greatest satiating value due to its slow gastric emptying time and greater effect on palatability (Rogers, 1990). Contrary to what has been shown by Rolls et al. (1988), our data showed no evidence of macronutrient-specificity in the development of satiety. In fact, following
SATIETY SENSATIONS AND FOOD INTAKE
29
the meatball preload, there was lower consumption of all the three foods presented (and consequently of all the macronutrients) than after eating pasta. In conclusion, satiety is difficult to rate objectively. Thus, we are in agreement with Blundell et al. (1987) and Kissileff (1984), that it is important to use standardized conditions in which only one dimension at a time is changed in order to compare the results. Furthermore, it would be useful to study a group of subjects who are both homogeneous in physiological characteristics and well-trained in evaluating sensations. Such ‘‘standardization’’ could lead to more precise and complete information that could then be applied to population groups.
R Blundell, J. E. & Hill, A. J. (1985) Analysis of hunger: interrelationships with palatability, nutrient composition and eating. In T. B. Van Itallie & J. Hirsch (Eds.), Recent advances in obesity research. Pp. 118–29. London: Libbey. Blundell, J. E. & Hill, A. J. (1988) Descriptive and operational study of eating in humans. In B. J. Blinder, B. F. Chaitin & R. Goldstein (Eds.), The eating disorders. Pp. 65–85. London: PMA Publishing. Blundell, J. E. & Rogers, P. J. (1980) Effects of anorexic drugs on food intake, food selection and preferences, hunger motivation and subjective experiences: pharmacological manipulation as a tool to investigate human feeding processes. Appetite, 1, 151–65. Blundell, J. E., Rogers, P. J. & Hill, A. J. (1987) Evaluating the satiating power of foods: implications for acceptance and consumption. In J. Solms et al. (Eds.), Food acceptance and nutrition Pp. 205–19. London: Academic Press. Booth, D. A., Chase, A. & Campbell, A. T. (1970) Relative effectiveness of protein in the late stages of appetite suppression in man. Physiology and Behavior, 5, 1299–302. Geiselman, P. J. & Novin, D. (1982) The role of carbohydrates in appetite, hunger and obesity. Appetite, 3, 203–23. Geliebter, A. (1979) Effects of equicaloric loads of protein, fat and carbohydrate on food intake in the rat and man. Physiology and Behavior, 22, 267–73. Hill, A. J., Magson, L. D. & Blundell, J. E. (1984) Hunger and palatability: tracking ratings of subjective experience before, during and after the consumption of preferred and less preferred food. Appetite, 5, 361–71. Kissileff, H. R. (1984) Satiating efficiency and a strategy for conducting food loading experiments. Neuroscience and Biobehavioral Reviews, 8, 129–35. Kissileff, H. R., Gruss, L. P., Thornton, J. & Jordan, H. A. (1984) The satiating efficiency of foods. Physiology and Behavior, 32, 319–32. Louis-Sylvestre, J., Tournier, A., Verger, P., Chabert, N., Delorme, B. & Hossenlopp, J. (1989) Learned caloric adjustment of human intake. Appetite, 12, 95–103. Panksep, J. (1971) Effects of fats, proteins and carbohydrates on food intake in rats. Psychonomic Sciences, 4, 85–95. Rodin, J., Wack, J., Ferrannini, E. & De Fronzo, R. A. (1985) Effect of insulin and glucose on feeding behavior. Metabolism, 34, 826–31. Rogers, P. J. (1990) Dietary fat, satiety and obesity. Food Quality and Preference, 2, 103–10. Rolls, B. J., Rowe, E. A., Rolls, E. T., Kingston, B., Megson, A. & Gunary, R. (1981a) Variety in a meal enhances food intake in man. Physiology and Behavior, 26, 215–21. Rolls, B. J., Rolls, E. T., Rowe, E. A. & Sweeney, K. (1981b) Sensory specific satiety in man. Physiology and Behavior, 27, 137–42. Rolls, B. J., Hetherington, M. & Burley, V. J. (1988) The specificity of satiety: the influence of foods of different macronutrient content on the development of satiety. Physiology and Behavior, 43, 145–53. Rolls, B. J., Fedoroff, I. C., Guthrie, J. F. & Laster, L. J. (1990) Foods with different satiating effects in humans. Appetite, 15, 115–26. Silverstone, J. T. & Stunkard, A. J. (1968) The anorexic effect of dexamphetamine sulphate. British Journal of Pharmacology and Chemotherapy, 33, 513–22.
30
M. PORRINI ET AL.
Sunkin, S. & Garrow, J. S. (1982) The satiety value of protein. Human Nutrition, 36A, 197–201. Teghtsoonian, M., Becker, E. & Edelman, B. (1981) A psychophysical analysis of perceived satiety: its relation to consummatory behavior and degree of overweight. Appetite, 2, 217–29. Wardle, J. (1987) Hunger and satiety: a multidimensional assessment of responses to caloric loads. Physiology and Behavior, 40, 577–82. Wooley, O. W., Wooley, S. C. & Woods, W. A. (1975) Effect of calories on appetite for palatable food in obese and nonobese humans. Journal of Comparative and Physiological Psychology, 9, 619–25. Received 5 August 1992, revision 20 May 1994
12-07-95 15:10:46
Appetite, 1995, 25, 17–30
Evaluation of Satiety Sensations and Food Intake After Different Preloads
MARISA PORRINI, ROBERTA CROVETTI and GIULIO TESTOLIN DISTAM, sez. Nutrizione, Universita’ degli Studi di Milano, via Celoria 2, 20133 Milano, Italia
SANTINO SILVA USSL 32 Treviglio, P. le Ospedale, 1, 24047 Treviglio (BG), Italia
The reproducibility of three questions, related to fullness, satiety and desire to eat, rated on an unmarked triangle was verified. In four sessions 12 volunteers ate pasta with tomato sauce (520 kcal) and were asked to rate the sensations felt. There was no difference in rating scores of the replications so the proposed questionnaire provides a stable measure of sensations related to satiety. Subsequently three satiety conditions were studied. Two foods, one rich in carbohydrate, pasta (baked macaroni) and the other in protein, polpette (meatballs), were used as loads at two calorie levels and as preload before an ‘‘ad libitum’’ meal. All the three questions proved useful in discriminating between the different satiety conditions. The food intake underlines the specificity of satiety: subjects, after eating a preload which previously had satiated them, ate other foods in different amounts depending on the kind of preload eaten. Food intake was significantly higher after the pasta preload, furthermore ‘‘fullness’’ and ‘‘satiety’’ ratings were significantly highest after the meatball preload, suggesting that in our experimental conditions, meatballs were more satiating than pasta. In conclusion, this study highlights the validity of using several quite different questions to study hunger and satiety, together with the actual food intake. 1995 Academic Press Limited
I The evaluation of hunger and satiety sensations is difficult as it appears that there are no universally accepted measurements for such sensations (Louis-Sylvestre et al., 1989). Two rating methods are often used to quantify psychological and physiological sensations experienced by subjects: fixed-point scales and visual-analogue scales. For a detailed description and discussion of such methods the reader is referred to two reviews by Blundell and Hill (1988) and Teghtsoonian et al. (1981). One of the problems faced in trying to rate sensations of hunger and satiety is the existence of sensory-specific satiety. Sensory-specific satiety has been studied by Research was supported by Grant n.91.00223.PF41 from the National Research Council, Targeted Project ‘‘Prevention and Control Disease Factors’’, Subproject ‘‘Nutrition’’. Address correspondence to: Marisa Porrini, Department of Food Science and Technology and Microbiology, via Celoria 2, 20133 Milan, Italy. 0195–6663/95/040017+14 $12.00/0
1995 Academic Press Limited
18
M. PORRINI ET AL.
Rolls et al. (1981a, 1981b) who demonstrated that if several foods, different in taste, appearance and texture, are offered in succession, the consumption of the meal will be higher than when only one food is given. Thus, overeating may occur when a wide variety of foods is readily available. Difficulty discriminating between sensations felt in a ‘‘sensory-specific satiety’’ condition and those felt in a ‘‘general satiety’’ state could explain, in part, the poor correlation between sensation ratings accompanying ingestion and food consumption reported by some authors (Blundell & Rogers, 1980; Silverstone & Stunkard, 1968; Wardle, 1987). As data from psychophysical methods are not univocal, nor is the information complete, descriptive and operational studies of eating should include both the subject’s evaluation of hunger and satiety and assessment of calorie consumption. The use of different questions concerning both the sensation of satiety and the perspective consumption could lead to more complete information about the real state of the subject. The primary purpose of our study was to verify whether three different questions (related to fullness, satiety and desire to eat) rated on an unmarked triangle would be useful to scale perceived satiety after both a single food and a complete meal. Two foods, different in protein and carbohydrate content as well as in sensory properties, were tested to verify the consistency of the method in evaluating their satiating power. M Subjects Twelve healthy males, 23–26 years old, of normal body weight took part in the study. Their body mass index [(weight in kg) (height in m)−2 ] was between 19·7 and 23·0. Subject Selection The subjects were recruited from a graduate lecture course. Subjects receiving medication, smokers and dieters were excluded. The potential subjects were asked to complete a questionnaire on their eating habits and their preferences (none, normal and high liking) for 92 foods that were listed in the questionnaire. The selected subjects had a normal preference for the foods included in the study. Satiety Ratings Questions asked for information on sensations related to hunger and satiety were: ‘‘How satiated do you feel?’’, ‘‘How full do you feel?’’ and ‘‘How great is your desire to eat?’’. The subjects were asked to rate their sensations by drawing a line, parallel to the baseline, across an isosceles triangle (height 15 cm, base 3 cm), oriented horizontally on the paper with the base to the right. The triangle was unbroken and was marked with a word anchor at the apex to indicate the minimum (not at all) of the experienced sensation and at the baseline for the maximum (extremely). The ratings were expressed in centimetres squared. Procedure The subjects were instructed not to eat after 2200 hrs the previous evening and to arrive at the laboratory at 1230 hrs for a prepared lunch. The instruction to miss
19
SATIETY SENSATIONS AND FOOD INTAKE
T 1 Nutrient composition of foods (referred to 100 g of product) Energy (kcal)
Fat (g)
Protein (g)
CHO (g)
Fibre (g)
Pasta Meatballs
132 146
3·9 4·1
5·7 20·2
18·4 7·1
1·4 0·4
Mozzarella Cracker Tart
249 425 412
19·0 11·7 15·4
19·5 8·9 4·3
0·0 70·9 64·1
0·0 2·9 0·9
T 2 Physical characteristics of preloads Firmness (kg) 1∗ Pasta Meatballs
2∗
Volume (ml) Weight (g)
3∗
21·8 17·6 13·1 36·7 31·3 27·2
950 560
727 600
∗ Successive measures on the same sample.
breakfast was designed to ensure that all of the subjects would be similarly deprived at the start of the experiment and that they would be readily motivated to eat. On arrival at the university dining room, where professors eat lunch, the subjects were seated at small, individual and well-separated tables, thus minimizing distraction. The subjects were asked to be silent and to remain at their respective tables for the duration of the experiment. The investigation was divided into two principal experiments. Experiment I This was performed to verify the reproducibility of the answers to the three questions used to rate satiety. In four different sessions on non-consecutive days, the subjects were asked to eat pasta with tomato sauce (520 kcal). Before starting the session, blood tests were taken to measure fasting blood glucose in order to ensure the same physiological state for each replication. Immediately before lunch, and immediately after, the subjects were asked to rate their ‘‘satiety’’, ‘‘fullness’’ and ‘‘desire to eat’’ on the questionnaire being tested. After eating they also rated the pleasantness of the food on a nine-point hedonic scale. To avoid any possibility of memorizing the ratings the sessions were mixed with other experiments which were in no way connected with the present experiment. Experiment II In this experiment, three different satiety conditions were studied. Two different foods were used: one rich in carbohydrate, pasta pasticciata (baked macaroni), and the other in protein, polpette (meatballs seasoned with tomato sauce). Nutrient composition, determined by chemical analysis, is reported in Table 1; physical characteristics are shown in Table 2.
20
M. PORRINI ET AL.
In order to decide the caloric levels of the meals a preliminary session was done. On two different days, subjects were instructed to eat pasta and meatballs until they felt ‘‘comfortably full’’ (‘‘specific’’ satiety condition). The mean pasta intake was 960 kcal, that of meatballs 880 kcal. After this test, the subjects with consumption higher or lower than two standard deviations from the mean were omitted. Ten of the 12 subjects were recruited for this study. Only completed data sets were included in the statistical analysis. The experiment consisted of three sessions for each food, which the subjects consumed in random order: first session, the food at the calorie level corresponding to their ‘‘specific’’ satiety (pasta at 960 kcal and meatballs at 880 kcal); second session, the food at half the satiating calorie level (pasta at 480 kcal and meatballs at 440 kcal); third session, the food at the larger calorie level followed by an ad libitum meal (test meal) of mozzarella cheese, crackers and fruit tart. The subjects had no knowledge of the calorie level of the different meals nor of the criteria followed in choosing the food. Test meals, together with 300 ml of water, were served on individual trays to all the subjects at the same time. Immediately before, immediately after and 2 hr after the meal the subjects were asked to rate their ‘‘satiety’’, ‘‘fullness’’ and ‘‘desire to eat’’ on the questionnaire they found on the table. Glucose Determination Blood (200 ll), taken by pricking the finger tip with a lancet, was collected in microtubes containing heparin. Glucose was measured using a Glucose and LLactate Analyzer (model 2300 STAT, Yellow Springs Instrument Co, U.S.A.). Physical Determination Texture was determined by means of an Instron Universal Machine (model 4301) equipped with a share press. To simulate mastication three successive measures on about 50 g of food were made on the same sample. Data reported in Table 2 are the means of four determinations. Volume was determined by displacement: the food was put into a cylinder full of water and the increase in volume noted. Weight was measured by a digital scale (Hanson LDT, U.K.) corrected to 2 g. Data Analysis Data were analysed using a repeated-measures design. In the first experiment, the reproducibility of the rating scores for the three questions were tested by analysis of covariance with replications as factors and blood glucose and palatability as covariates. Such an approach allowed the adjustment of rating values for variables that could not be standardized. To confirm the reproducibility of the questionnaire correlation analysis was employed. In the first and second sessions of the second experiment the changes in ‘‘desire to eat’’, ‘‘fullness’’ and ‘‘satiety’’ were analysed separately using a three-factor analysis of variance with food type (meatballs or pasta), calorie level (specific satiated or half satiated) and time as factors. Means of the different scores were compared using the Newman–Keuls multiple range test to highlight any significant difference. In the
SATIETY SENSATIONS AND FOOD INTAKE
21
third session of the second experiment changes in satiety sensations were analysed using an analysis of variance with the type of preload (meatballs or pasta pasticciata) and time as factors. The consumption of each item in the ad libitum meal and the test meal calorie intake were compared with paired t-test. The paired t-test was also used to compare the satiety sensation ratings expressed in the first and third sessions for each assessment time. Relationships among the dependent variables were analysed by means of simple linear regression.
F 1. Means (SD) of (A) ‘‘Desire to eat’’, (B) ‘‘Fullness’’, and (C) ‘‘Satiety’’ ratings, (Φ) before and (Ε) after lunch, in the four replications.
22
M. PORRINI ET AL.
R Experiment I Blood glucose and palatability ratings Mean fasting glucose levels (mg dl−1) were 71·8 (±5·4), 74·0 (±6·2), 73·5 (±4·4) and 74·8 (±5·0) in the four replications. Palatability ratings were respectively 7·7 (±0·9), 7·7 (±0·8), 7·4 (±1·0) and 7·1 (±0·9). Subjective ratings In Fig. 1 means and standard deviations of ‘‘desire to eat’’, ‘‘fullness’’ and ‘‘satiety’’ ratings obtained in the four replications are reported. There was no difference in the rating scores either on the change between before and after lunch or considering only data obtained after eating (significance levels always higher than 0·5). These results were confirmed by correlation analysis. The correlation coefficients were highly significant between the same sensation score in the different replications (Table 3). Subjective ratings expressed after eating were always highly intercorrelated. The correlation coefficients (r) were −0·6253 between ‘‘desire to eat’’ and ‘‘fullness’’, −0·6147 between ‘‘desire to eat’’ and ‘‘satiety’’ and 0·9315 between ‘‘satiety’’ and ‘‘fullness’’ (sample size=48, p<0·0001). Experiment II Food intake in the preliminary session As previously reported, the mean consumption of pasta in the preliminary session was 960 kcal (SD=151) and that of meatballs was 880 kcal (SD=185). This difference T 3 Correlation coefficients and significance levels between scores for each replication Desire to eat
1
Fullness
2
3
4
0·8468 0·0005
0·8373 0·0007 0·8732 0·0002
0·7712 0·0033 0·8642 0·0003 0·9460 0·0000
2 3
Satiety
1 2
2
3
4
0·7632 0·0039
0·7938 0·0021 0·9759 0·0000
0·8995 0·0001 0·9017 0·0001 0·9072 0·0000
3 1, 2, 3, 4=replications’ number. Sample size=12.
1 2 3
2
3
4
0·7138 0·0091
0·7178 0·0086 0·9445 0·0000
0·9051 0·0001 0·8364 0·0007 0·8494 0·0005
SATIETY SENSATIONS AND FOOD INTAKE
23
was not significant. All subjects rated both preloads as pleasant, and the scores were not statistically different (7·5 ±0·7 and 7·1 ±0·6 respectively for pasta and meatballs). Satiety ratings in the first and second sessions Fasting ratings for each of the three questions ‘‘How great is your desire to eat?’’, ‘‘How full do you feel?’’ and ‘‘How satiated do you feel?’’ obtained in the different sessions were compared to establish that baseline levels were not different. No significant difference was identified. In Fig. 2 the ratings for each question were plotted as a function of time. As can be noted, the effect of time on all the satiety sensations was similar for all the loads. In fact the mean changes for each score from before and after each load showed a significant main effect of time, for ‘‘desire to eat’’, F(2,120)=106·148, p<0·0001, for fullness ratings, F(2,120)=70·881, p<0·0001 and for satiety ratings, F(2,120)=63·609, p<0·0001. For ‘‘desire to eat’’, analysis of variance showed a significant effect of calorie levels, F(1,120)=24·260, p<0·0001. Analysis by Newman–Keuls confirmed that following the smaller energy load the ‘‘desire to eat’’ was significantly higher than following the larger one. There was no effect of the type of food eaten on this sensation. There was no significant interaction between the three experimental factors (food type, calorie level and time). As regards ‘‘fullness’’ there was a significant interaction calorie×time, F(1,120)= 13·221, p<0·0001; furthermore a significant effect of food eaten was present, F(1, 120)=5·175, p=0·0249. Newman–Keuls analysis of overall changes in ‘‘fullness’’ showed that this sensation was significantly higher after a larger calorie load. Following the meatball load, ‘‘fullness’’ was significantly higher than when pasta was eaten. As regards the question ‘‘How satiated do you feel?’’, the interaction calorie×time was still significant, F(1,120)=12·424, p<0·0001, but no significant effect of type of food was present. In Table 4, the correlation coefficients and significance levels between subjective scores for each assessment time are shown. As can be noted the ratings were always highly intercorrelated for both preloads. Food intake in the third session Figure 3 shows the subjects’ consumption of the different foods and calorie intake in the ad libitum meal. The subjects significantly altered their intake according to the type of preload eaten: they had greater intake following pasta (811±226 kcal) than following meatballs (438±185 kcal). Paired comparison analysis demonstrated significant differences across the two test meals for energy intake, F(1,15)=13·100, p=0·0028, mozzarella cheese, F(1,15)=8·199, p=0·0125, crackers, F(1,15)=7·203, p=0·0178, and apricot jam tart, F(1,15)=5·283, p=0·0375. Subjective ratings in the third session In Fig. 4 the ratings for each question were plotted as a function of time. As expected, the ratings of ‘‘desire to eat’’, ‘‘fullness’’ and ‘‘satiety’’ expressed after the two ad libitum meals were not different. There was a significant change for each sensation with time: respectively, F(2,48)=101·781, p<0·0001, F(2,48)=192·592, p<0·0001, F(2,48)=114·364, p<0·0001 for desire to eat, fullness and satiety ratings.
24
M. PORRINI ET AL.
F 2. Mean (+SEM) (A) ‘‘Desire to eat’’, (B) ‘‘Fullness’’ and (C) ‘‘Satiety’’ ratings just before (0), 30 and 120 min after eating the four different loads. Μ, larger pasta load; Α, smaller pasta load; Χ, larger meatball load; Β, smaller meatball load.
Satiety ratings expressed in height and area Relationships between intake of different amounts of foods studied (between 440 and 1500 kcal) and satiety ratings obtained 30 min after eating and expressed by both height and area are reported in Table 5. A good linear relation was always present.
25
SATIETY SENSATIONS AND FOOD INTAKE
T 4 Correlation coefficients between subjective scores for each assessment time
Desire to eat−Fullness Desire to eat−Satiety Satiety−Fullness
0 min
30 min
120 min
−0·5926 −0·7254 0·6383
−0·8179 −0·8061 0·8895
−0·7619 −0·7548 0·9679
Sample size=32. p always lower than 0·0001.
T 5 Correlation coefficients between satiety scores and calorie intake
Desire to eat−Calorie Fullness−Calorie Satiety−Calorie
Height
Area
−0·7707 0·7095 0·7140
−0·7268 0·7204 0·7291
Sample size=48. p always lower than 0·0001.
F 3. Test meal intake (+SEM) and consumption (+SEM) of the three ad libitum foods after pasta (Φ) and meatball (∆) preloads.
Comparison between ratings in the first and third sessions Paired t-test conducted between satiety sensations expressed after the larger meatball load and after the test meal with meatballs as preload did not show differences. The same comparison conducted for pasta indicated that ‘‘satiety’’ was higher after the test meal, 21·2±1·3 vs. 14·3±1·5, t(8)=2·37, p=0·0390, whereas
26
M. PORRINI ET AL.
F 4. Mean (+SEM) ratings just before (0), and 30 and 120 min after eating the two larger preloads, pasta (Α) and meatball (Β), plus test meal.
‘‘desire to eat’’ was higher after the larger pasta load, 1·1±0·3 vs. 0·1±0·0, t(8)= 3·26, p=0·0085. There was no statistically significant difference for fullness sensation between the two pasta sessions. C From the results of the first experiment, the subjective scores of ‘‘desire to eat’’, ‘‘fullness’’ and ‘‘satiety’’, assessed from the proposed questionnaire, provided a stable
SATIETY SENSATIONS AND FOOD INTAKE
27
measure of perceived sensations. Particular care was given to choosing the rating scale. Fixed-point scales were rejected so as to avoid any recall of previous rates and end-effect. As reported by Teghtsoonian et al. (1981), numbers are easily remembered and might be used by the subject to reflect not only his feelings of satiety, but also his preconception of what he should feel during a meal. Lengths, and consequently areas, are hard to remember so each length- or areasatiety match would be more directly perceivable. However instead of the usual line of visual-analogue scales we preferred to use a triangle to quantify the sensation felt. The fact that in the triangle the area increases helps subjects to better quantify hunger and satiety sensations which have an increasing and decreasing trend (for example the sensation of filling the stomach). To analyse whether any distortion of the scaling of the sensation occurred by using the area of the triangle instead of its height, we assessed the relationships between different amounts of foods studied in relation to satiety scores (see Table 5). As with both height and area there was a good linear relation between intake and satiety ratings we think that using area could help in quantifying sensation felt without distorting the results. On the whole, the questionnaire was helpful for both comparing the satiating effects of different foods and studying different sated conditions. In particular, subjects interpreted the scoring method very well and were able to rate, in a reproducible way, the sensations they felt after eating the food put before them. The ratings obtained for satiety, fullness and desire to eat were shown to be significantly affected by time and are therefore influenced by deprivation time. Furthermore, unlike Wooley et al. (1975), who found that subjective ratings vary predictably with deprivation time but fail to differentiate between larger and smaller energy preloads, we were able to discriminate between the sensations felt after two loads of the same food with different energy content. Although it is difficult to understand the validity of using several subjective ratings to study the different sated conditions, an analysis of the results has led to some considerations. In the condition where the subjects were not fully satiated, like that of the first experiment, the fullness and satiety ratings were negatively correlated with ‘‘desire to eat’’, whereas the correlation between ‘‘fullness’’ and ‘‘satiety’’ was positive. It may be possible that fullness and satiety are perceived as the same sensation. Similar correlations were also found in the first and second sessions of the second experiment where the subjects were not completely satiated. However a significant effect of the kind of food eaten was present for ‘‘fullness’’, but not for ‘‘satiety’’ and ‘‘desire to eat’’. The ratings obtained in the third session could not be intercorrelated as at the measurement extremes there was insufficient variability in the ratings. As expected, there was no difference between pasta and meatballs session sensations. On the other hand comparison of the data for sensation in the ‘‘specific’’ satiety condition (first session) with data in the ‘‘general’’ satiety condition (third session) indicates that ‘‘satiety’’ and ‘‘desire’’ ratings were different in the two conditions, at least as far as pasta was concerned. The differences in the sensations felt were confirmed by the food intake data. After the larger pasta preload, the subjects ate more than after the larger meatball preload, confirming that the differences reflected more than an occasional finding.
28
M. PORRINI ET AL.
In summary, ‘‘fullness’’ was influenced by food but not by the different satiation conditions. However, ‘‘satiety’’ and ‘‘desire to eat’’ were not influenced by the kind of food eaten but were different in the ‘‘specific’’ and ‘‘general’’ satiety conditions. Clearly, many variables influenced subjects in rating their sensations. Thus, after a monotonous meal, the desire to eat remained at a high level even though the amount consumed (as regards both calories and weight) had induced the subjects to stop eating. Satiety sensation could unconsciously be influenced by the variety of the meal or by other characteristics of the food itself, such as its rheological properties. Consequently the three sensations did not completely overlap in describing the satiation condition, especially when, in addition to physiological variables, other variables are present. Few studies ask ‘‘How satiated do you feel?’’ (Blundell & Rogers, 1980; Kissileff et al., 1984). Our results indicate the importance of rating each sensation accompanying consumption separately. It would seem that the use of a questionnaire alone is inadequate to have really precise and complete results when studying hunger and satiety; consequently, we think that it should also be important to determine the actual intake by weighing the food consumed. In this study the results of food intake confirmed the specificity of satiety. After the consumption of a specific food of a given calorie level, which was sufficient to satiate the subjects in a preliminary session, the same subjects ate other foods in different amounts depending on the kind of preload eaten. In the experimental conditions used, meatballs were more satiating than pasta. Food intake was significantly higher after the pasta preload for all the three foods eaten (cheese, crackers and tart). The tart, which had high hedonistic characteristics and whose consumption could be independent of a real hunger stimulus, was consumed in higher amounts after pasta, confirming the different effect of the two preloads. Subjective ratings to the three questions confirmed the fact that meatballs were more satiating than pasta, in fact both in the ‘‘specific satiated’’ and in the ‘‘half satiated’’ conditions, fullness and satiety ratings are highest after meatballs. The two foods have the same fat content but reciprocal amounts of protein and carbohydrate (starch). Furthermore the meatball preload had a lower weight, similar calories, but lower volume and higher firmness. It follows, that the higher satiating effect of meatballs could be ascribed to the nutrient composition and/or the consistency. To clarify these differences, additional studies are needed involving food different in nutritive density and rheological parameters (volume vs. calorie, consistency vs. calorie etc.). There continues to be scientific debate over the satiating properties of macronutrients. In some studies high protein foods produce satiety more readily than protein-poor foods (Blundell & Hill, 1985; Booth et al., 1970). High carbohydrate foods appear to stimulate appetite and hunger more than high fat or high protein foods (Geiselman & Novin, 1982; Rodin et al., 1985). In contrast, other studies found no differences between the satiating properties of protein and carbohydrates (Geliebter, 1979; Panksep, 1971; Sunkin & Garrow, 1982), while they suggested that fat may have the greatest satiating value due to its slow gastric emptying time and greater effect on palatability (Rogers, 1990). Contrary to what has been shown by Rolls et al. (1988), our data showed no evidence of macronutrient-specificity in the development of satiety. In fact, following
SATIETY SENSATIONS AND FOOD INTAKE
29
the meatball preload, there was lower consumption of all the three foods presented (and consequently of all the macronutrients) than after eating pasta. In conclusion, satiety is difficult to rate objectively. Thus, we are in agreement with Blundell et al. (1987) and Kissileff (1984), that it is important to use standardized conditions in which only one dimension at a time is changed in order to compare the results. Furthermore, it would be useful to study a group of subjects who are both homogeneous in physiological characteristics and well-trained in evaluating sensations. Such ‘‘standardization’’ could lead to more precise and complete information that could then be applied to population groups.
R Blundell, J. E. & Hill, A. J. (1985) Analysis of hunger: interrelationships with palatability, nutrient composition and eating. In T. B. Van Itallie & J. Hirsch (Eds.), Recent advances in obesity research. Pp. 118–29. London: Libbey. Blundell, J. E. & Hill, A. J. (1988) Descriptive and operational study of eating in humans. In B. J. Blinder, B. F. Chaitin & R. Goldstein (Eds.), The eating disorders. Pp. 65–85. London: PMA Publishing. Blundell, J. E. & Rogers, P. J. (1980) Effects of anorexic drugs on food intake, food selection and preferences, hunger motivation and subjective experiences: pharmacological manipulation as a tool to investigate human feeding processes. Appetite, 1, 151–65. Blundell, J. E., Rogers, P. J. & Hill, A. J. (1987) Evaluating the satiating power of foods: implications for acceptance and consumption. In J. Solms et al. (Eds.), Food acceptance and nutrition Pp. 205–19. London: Academic Press. Booth, D. A., Chase, A. & Campbell, A. T. (1970) Relative effectiveness of protein in the late stages of appetite suppression in man. Physiology and Behavior, 5, 1299–302. Geiselman, P. J. & Novin, D. (1982) The role of carbohydrates in appetite, hunger and obesity. Appetite, 3, 203–23. Geliebter, A. (1979) Effects of equicaloric loads of protein, fat and carbohydrate on food intake in the rat and man. Physiology and Behavior, 22, 267–73. Hill, A. J., Magson, L. D. & Blundell, J. E. (1984) Hunger and palatability: tracking ratings of subjective experience before, during and after the consumption of preferred and less preferred food. Appetite, 5, 361–71. Kissileff, H. R. (1984) Satiating efficiency and a strategy for conducting food loading experiments. Neuroscience and Biobehavioral Reviews, 8, 129–35. Kissileff, H. R., Gruss, L. P., Thornton, J. & Jordan, H. A. (1984) The satiating efficiency of foods. Physiology and Behavior, 32, 319–32. Louis-Sylvestre, J., Tournier, A., Verger, P., Chabert, N., Delorme, B. & Hossenlopp, J. (1989) Learned caloric adjustment of human intake. Appetite, 12, 95–103. Panksep, J. (1971) Effects of fats, proteins and carbohydrates on food intake in rats. Psychonomic Sciences, 4, 85–95. Rodin, J., Wack, J., Ferrannini, E. & De Fronzo, R. A. (1985) Effect of insulin and glucose on feeding behavior. Metabolism, 34, 826–31. Rogers, P. J. (1990) Dietary fat, satiety and obesity. Food Quality and Preference, 2, 103–10. Rolls, B. J., Rowe, E. A., Rolls, E. T., Kingston, B., Megson, A. & Gunary, R. (1981a) Variety in a meal enhances food intake in man. Physiology and Behavior, 26, 215–21. Rolls, B. J., Rolls, E. T., Rowe, E. A. & Sweeney, K. (1981b) Sensory specific satiety in man. Physiology and Behavior, 27, 137–42. Rolls, B. J., Hetherington, M. & Burley, V. J. (1988) The specificity of satiety: the influence of foods of different macronutrient content on the development of satiety. Physiology and Behavior, 43, 145–53. Rolls, B. J., Fedoroff, I. C., Guthrie, J. F. & Laster, L. J. (1990) Foods with different satiating effects in humans. Appetite, 15, 115–26. Silverstone, J. T. & Stunkard, A. J. (1968) The anorexic effect of dexamphetamine sulphate. British Journal of Pharmacology and Chemotherapy, 33, 513–22.
30
M. PORRINI ET AL.
Sunkin, S. & Garrow, J. S. (1982) The satiety value of protein. Human Nutrition, 36A, 197–201. Teghtsoonian, M., Becker, E. & Edelman, B. (1981) A psychophysical analysis of perceived satiety: its relation to consummatory behavior and degree of overweight. Appetite, 2, 217–29. Wardle, J. (1987) Hunger and satiety: a multidimensional assessment of responses to caloric loads. Physiology and Behavior, 40, 577–82. Wooley, O. W., Wooley, S. C. & Woods, W. A. (1975) Effect of calories on appetite for palatable food in obese and nonobese humans. Journal of Comparative and Physiological Psychology, 9, 619–25. Received 5 August 1992, revision 20 May 1994