- Email: [email protected]
Effects of time-restricted access to protein and of oral-sensory cues on protein selection
Physiology & Behavior,Vol. 55, No. 4, pp. 659-664, 1994 Copyright © 1994 ElsevierScienceLtd Printed in the USA. All fights reserved 0031-9384/94 $6.00 + .00
Pergamon
Effects of Time-Restricted Access to Protein and of Oral-Sensory Cues on Protein Selection MARK
D. H O L D E R *l A N D D A V I D
DIBATTISTAt
*Okanagan University College, Kelowna, British Columbia, Canada and "pBrock University, St. Catharines, Ontario, Canada R e c e i v e d 7 J u n e 1993 HOLDER, M. D. AND D. DIBA'Iq'ISTA. Effects of time-restricted access to protein and of oral-sensory cues on protein selection. PHYSIOL BEHAV 55(4) 659-664, 1994.--The effects on protein consumption of restricting access to protein and of varying the oral-sensory properties of protein diets were measured. During the initial phase of the study, rats were maintained on a selfselection diet in which three different macronutrient sources (carbohydrate, fat, and either soy-based or casein-based protein diets) were continuously available. For the remaining 9 days of the study, half of the rats were protein deprived for 23 h each day and the other half continued to receive the same protein diet during this 23-h period. The remaining 1 h of each day was a test period in which all rats had access to a protein diet that was either the same as or different from the one they had received in the initial phase. Compared to the nonrestricted rats, the protein-restricted rats consumed more than twice as much of the available protein diet during 1-h test periods. For the nonrestricted rats, those that received a different protein diet during the 1-h test periods consumed 60% more of the protein diet than did those that received the same protein diet. These results indicate that increases in protein consumption following protein deprivation can be attributed, at least in part, to the oral-sensory properties of diets and not necessarily to a specific protein appetite. Protein selection
Protein restriction
Rat
W H E N rats are allowed to select among separate sources of protein, fat, and carbohydrate, their total energy intake and the proportion of total energy derived from each dietary source are typically quite stable (20,26). Furthermore, self-selecting rats will often make appropriate adjustments in nutrient intake in response to challenges such as streptozotocin-induced diabetes (15), or exogenous administration of insulin (17) and 2-deoxyglucose (16). One type of challenge that has been studied extensively relates to experimentally induced shortages in particular dietary nutrients. Numerous studies have demonstrated that rats will selectively increase their intake of diets containing substances such as sodium (5,19) and thiamine (29,30) following a period of time during which these substances have been unavailable for consumption. It has similarly been observed that when access to dietary protein is restricted for a period of time, rats selectively increase their consumption of a protein-containing diet when it once again becomes available (6,7,22,25). Findings of this sort are often interpreted as evidence for the existence of a protein selection mechanism, that is, a specific appetite for protein that may be either innate (6) or acquired (3,11). However, changes in the availability of a protein-containing diet are usually confounded with changes in the availability of oral-sensory stimuli.
An alternative explanation of the selective increase in protein consumption following restricted access to dietary protein focuses on the importance of oral-sensory stimuli. It is well established that there is an important sensory-specific component to satiety. For example, rats generally prefer the less recently tasted of two food items (12). In another study, rats were given a 2-h meal consisting of one diet (21). When the diet was presented every 30 min with a different flavour added to it (i.e., four different flavours in total), rats consumed two to three times more of the diet than when only one flavour was used. Also, rats that have consumed a concentrated sugar solution to satiety will vigorously ingest either lab chow or powdered sugar when it is offered, but will refuse to drink more of the sugar solution (23). Furthermore, both rats (4,28) and humans (27,31) consume larger meals when presented with an assortment of different food items than when presented with a single food item. Thus, the selective increase in consumption of a protein-containing diet following protein restriction may be attributable, at least in part, to the novelty of the oral-sensory properties of the protein-containing diet rather than to the existence of a specific appetite for protein. This alternative explanation seems plausible in light of related research on carbohydrate selection. A recent theory proposed that carbohydrate selection is modulated by a negative feedback loop
Requests for reprints should be addressed to M. D. Holder, Okanagan University College, Department of Psychology, 3333 College Way, Kelowna, British Columbia, Canada V 1Y 1V7.
659
H()IA)ER AND I)IBATTISTA
660
involving brain serotonin (1,2,8,22). However, much of the research cited in support of this theory contk)unded the macronutrient composition of diets with their oral sensory properties. More recent studies have controlled the oral-sensory properties of the diets with conditioning techniques (11) or by using carbohydrate diets that differed in their oral-sensory properties (13). These studies have indicated that results that were interpreted as showing carbohydrate selection might better be interpreted as showing the selection of specific oral-sensory properties. The study reported here was conducted to evaluate the importance of the oral-sensory properties of a protein-containing diet in causing the selective increase in protein intake that is observed in the chronic protein-restriction paradigm (7). Briefly, rats were permitted to self-select from three separate diets: a carbohydrate diet, a fat diet, and a protein diet that contained either casein or soy protein. The carbohydrate and fat diets were continuously available throughout the experiment. After a baseline period during which all three diets were freely available, proteinrestricted animals were allowed access to one of the protein diets during a l-h test period each day, but bad no access to dietary protein during the remaining 23 h/day. Nonrestricted animals had access to dietary protein for 24 h/day; the original protein diet of the baseline period was available for 23 h/day, and one of the protein diets was available during the l-h test period. For both protein-restricted and nonrestricted rats, the protein diet presented during l-h test periods was either the same as or different from the protein diet available to the animal during the baseline period. METHOD
Subjects Sixty-nine Sprague-Dawley, albino male rats weighed an average of 258 g at the start of the experiment. The rats were housed individually in clear Plexiglas cages. All procedures were conducted in these cages. Room lights were on from 0700 to 1900 h in the animal colony. Water was continuously available.
Apparatus The rats were maintained on a self-selection diet consisting of separate carbohydrate, fat, and either protein-based or soybased protein diets, each supplemented with vitamins and minerals. The composition of each diet is described in Table 1. Pilot work had shown that the two protein diets differed in oral-sensory cues; a conditioned flavor aversion acquired to one of the protein diets did not generalize to the other protein diet. The diets were presented in spill-resistant containers that were placed at the front of each cage. The location of each individual diet varied each day. The containers were made by gluing a 250-ml glass jar inside a larger glass food cup. Food was placed in the glass jar with spillage being caught in the surrounding cup. A flat piece of aluminum (12 × 12 × 0.08 cm) with rounded comers was glued to the bottom of the cup to prevent tipping of the food container. The aluminum was secured to the home cage with tape. Spillage that occurred despite these precautions was carefully noted.
Procedu re Throughout the study, all rats had continuous access to the fat and carbohydrate diets. Access to the two protein diets, soy and casein, varied across different groups of rats and different phases. Phase 1, which lasted 10 days, was an adaption phase
TAB Lt:'~ l ( OMPOSITION OF IEXPERIMENTAI. I)IE3S ( P E R C E N T A G E OF T O T A I W E I G t l T I
So 5 Protein Soy protein* Caseint Vegetable
Casein Prolein
('arbohydrate
92 -
92
shortening$
Sucrose§ Cornstarch', Methioninet Vitamin mixture1[ Mineral mixture# Alphacel~ Caloric density (kcal/g)
|:at
--
94 .
.
2 1.5 3.5 1 3.62
.
.
. . 2 1.5 3.5 l 3.83
-1.5 3.5 I 8.54
47 47 1.5 3.5 I 3.84
* ICN Canada. Isolated soy protein contains 92% protein, 4.1% ash, 1% fat, plus fibre and moisture. ICN Canada. $ Crisco, Procter and Gamble. § Atlantic Sugar, Montreal, Quebec. AIN Vitamin Mixture 76, ICN Canada. Contains 97.3% sucrose. # AIN Mineral Mixture 76, ICN Canada. Contains 11.8% sucrose. that allowed each rat to become familiar with the different diets. In addition to the fat and carbohydrate diets, half the rats had continuous access to the soy diet and the remaining rats had continuous access to the casein diet. Consumption of all diets was measured for the last 3 days of phase 1. Based on their consumption of the protein diet, each rat that had access to soy was assigned to one of four equal-sized groups such that consumption was similar across groups. The rats that had access to casein were similarly assigned to one of four groups in preparation for phase 2. During phase 2 of the study, which lasted 9 days, all rats had continuous access to the fat and carbohydrate diets, but access to the protein-containing diets varied across groups. The rats never received access to more than one protein diet at one time. Animals had access to dietary protein either for 24 h/day [nonrestricted (NRES)] or only during a l-h test period from 1730 to 1830 h each day [restricted (RES)]. During this test period, all rats received access to a protein diet, and the consumption of all diets (protein, fat, and carbohydrate) during this l-h period was measured. The protein diet available during the l-h test period was either the same as that received during phase 1 (same) or different from it (different). For the NRES rats, which had continuous access to protein, the protein diet available during the 23-h nontest period was always the same as that given during phase 1. Table 2 summarizes the eight groups created with this 2 (phase 1 protein diet: soy or casein) × 2 (protein access: RES or NRES) x 2 (protein likeness: test period protein diet either same as or different from phase 1 protein diet) factorial design.
Statistical Analyses Missing data points resulting from spillage, etc., were estimated using the procedure outlined by Kirk (18). Protein dietderived calories (PDC) and total calories derived from all dietary sources (TC) were calculated for both 1-h test periods and 24-h periods. The SPSS-PC+ statistical package, version 4.0, was used to perform analyses of variance; degrees of freedom were appropriately adjusted for within-subjects variables (14).
PROTEIN SELECTION
661
TABLE 2 SUMMARY OF RESEARCH DESIGN
Group
Phase 1 Protein Diet*
Protein Diet Available During Nontest Periods?t
Test Period Protein Diets
Soy-NRES-Same Soy-NRES-Diff Soy-RES-Same Soy-RES-Diff Casein-NRES-Same Casein-NRES-Diff Casein-RES-Same Casein-RES-Diff
Soy Soy Soy Soy Casein Casein Casein Casein
Yes Yes No No Yes Yes No No
Same (soy) Different (casein) Same (soy) Different (casein) Same (casein) Different (soy) Same (casein) Different (soy)
* Either the soy or the casein protein diet was continuously available during phase 1. 1-During phase 2, NRES rats had access during 23-h nontest periods to the same protein diet that was available during phase 1. RES rats had no access to either protein diet during the 23-h nontest periods. $ The protein diet available during 1-h test periods was either the same as (Same) or different from (Diff) the protein diet available during phase 1.
1-h intake of the casein-based diet being greater than that of the soy-based diet (8.62 _+ 0.88 vs. 5.42 _+ 0.48 kcal/h, respectively). If the oral-sensory cues provided by the diet are an important factor in the selection of protein diets during the 1-h test period, then consumption of the different protein diet should be greater than that of the same protein diet for the NRES rats. To determine whether this was the case, the protein access × protein likeness interaction for the 1-h PDC data, F(1, 54) = 2.92, p < 0.10, was investigated further. Results of this analysis are described in Table 3, which provides a summary of the 1-h PDC data averaged across both the nine test periods and the two test period protein diets. The 1-h PDC intake of NRES rats proved to be significantly greater (p < 0.05) when the test period protein diet was different from, rather than the same as, the phase l protein diet, which was available to NRES rats during the remaining 23 h of the day. Thus, although the 1-h PDC intake of RES rats was significantly greater than that of NRES rats regardless of the nature of the diet (p < 0.05), novelty of the protein diet presented during l-h test periods enhanced the 1-h PDC intake of NRES rats by some 60%. Figure 3 shows TC intake daring the 1-h test periods. A fourway A N O V A (test period protein diet × protein access × protein likeness × days) revealed only three significant main effects. As Fig. 3 suggests, rats generally tended to increase their 1-h TC intake during phase 2 [main effect of days: F(8, 301) = 8.65, p
RESULTS
Intake During 24-h Periods Figure 1 shows TC intake during consecutive 24-h periods. These data were analyzed using a four-way A N O V A with three between-subjects variables (phase 1 protein diet: soy or casein; protein access: RES or NRES; protein likeness: test period protein diet either same as or different from phase 1 protein diet) and one within-subjects variable (days). This analysis revealed a significant main effect of the protein access variable, F(I, 50) = 5.57, p < 0.05; over the course of the experiment, RES rats ingested about 6% fewer total calories than did NRES rats (mean - SE: 105.5 _ 2.35 vs. 112.5 ± 2.06 kcal/day, respectively). The greatest difference in TC intake between these groups occurred during the first few days of phase 2, as indicated by a significant protein access x days interaction, F(11,383) = 2.25, p < 0.05. In addition, it was found that daily TC intake over the course of the study was greater when rats were maintained on the soy-based diet (113.0 ± 2.40 kcal/day) rather than on the casein-based diet [105.2 _ 1.92 kcal/day; main effect of phase 1 protein diet, F(1, 50) = 7.03, p < 0.05].
Intake During 1-h Test Periods Figure 2 illustrates PDC intake during the 1-h test periods of phase 2 of the experiment. The PDC intake data were analyzed using a four-way A N O V A with three between-subjects variables (test period protein diet: soy or casein; protein access: RES or NRES; protein likeness: test period protein diet either same as or different from phase I protein diet) and one within-subjects variable (days). As expected, the main effect of the protein access variable was highly significant, F(1, 54) = 53.3, p < 0.001; thus, the 1-h PDC intake of RES rats was more than twice that of NRES rats (9.76 ± 0.68 vs. 4.44 + 0.52 kcal/h, respectively). In additiom as Fig. 2 suggests, the 1-h PDC intake of RES rats, but not of NRES rats, tended to increase during phase 2, as indicated by a significant protein access x days interaction, F(8, 257) = 2.25, p < 0.05. A significant main effect of the test period protein diet variable was also found, F(1, 54) = 19.7, p < 0.001, with
--~-
RES-SAME
~
RES-DIFFERENT
NRES-SAME
o
NRES-DIFFERENT
Phase 1 protein diet: Soy 140
2
120
~ _c
A
8
100
80
--
L 1 I
£
•
__L
2 3 - - I Phase 1
1
£
_J
2
3
J
1
4
5
I
6
L
I~
7
8
£
J 9 I
Phase
2
Phase 1 protein diet: Casein 140 e
120
•~
loo
1
2 Phase
3
1
2
3
1
4
5
6
7
8
9
Pt~ase 2
Days
FIG. 1. Mean total energy derived from all dietary sources during 24-h periods. All rats had continuous access to fat and carbohydrate diets throughout the experiment. Phase 1: rats had continuous access to either the soy-based or the casein-based protein diet. Phase 2: protein-restricted rats (RES) had access to a protein-containing diet only during a 1-h test period each day; this diet was either the same as (SAME) or different from (DIFFERENT) the protein-containing diet available during phase I. Nonrestricted rats (NRES) had access to dietary protein for 24 h/day; the protein-containing diet of phase 1 was available for 23 h/day, and the protein diet available during the 1-h test period was either the same or different.
662
ttOLDER AND D~BATTISTA
'~
RES-SAME
o
RES-DIFFERENT
~'~
NRFS-SAME
~
NRES-DIFFERENT
TABLE 3 MEAN
Test-period protein diet: Soy
ENERGY
CONTAINING
DERIVED
DIETS
DURING
FROM
PROTI/IN
I-H TEST
PERIO[)S
Nature of Test Diet
~12i
- -
()
41
'
o
,^.
i
A
A
A
3
4
5
c~ ~,
o A
A
7
8
2
6
9
Test-period protein diet: Casein
~12
~8 --
c~ A
o
4 i
0! 1
2
3
4
5
6
7
Same
Different
Restricted Nonrestricted
10.05" (1.10) 3.37 (0.50)
9.52 10.g5) 5.40~: (0.83)
Values are means (kcal), with SE shown in parentheses. * Protein-restricted rats had access to a protein-containing diet only during a l-h test period each day; this diet was either the same as (Same) or different from (Different) the protein-containing diet available during phase 1. Nonrestricted rats had access to dietary protein for 24 h/day; the protein-containing diet of phase 1 was available for 23 h/day, and the protein diet available during the l-h test period was either the Same or Different. f Significantly greater than other value in same column, p < 0.05. $ Significantly greater than other value in same row, p < 0.05.
0 ~
1
Protein Access
8
Test periods
FIG. 2. Mean energy derived from protein-containing diets during 1-h test periods of phase 2. All rats had continuous access to fat and carbohydrate diets throughout the experiment. Protein-restricted rats (RES) had access to a protein-containing diet only during a 1-h test period each day; this diet was either the same as (SAME) or different from (DIFFERENT) the protein-containing diet available during phase 1. Nonrestricted rats (NRES) had access to dietary protein for 24 h/day; the protein-containing diet of phase 1 was available for 23 h/day, and the protein diet available during the I -h test period was either the same or different.
< 0.001]. It was also found that rats having access to the caseinbased diet during test periods consumed more TC than did rats having the soy-based diet [17.2 ± 1.29 vs. 12.1 ± 0.67 kcal/h, respectively; main effect of test period protein diet: F(1, 51) = 13.99, p < 0.001 ]. Finally, RES rats consumed significantly more calories during 1-h test periods than did NRES rats [16.8 _ 1.06 vs. 12.6 ± 1.05 kcal/h, respectively; main effect of protein access: F(I, 51) = 10.25, p < 0.01]. None of the effects of the protein likeness variable reached significance [e.g., protein access × protein likeness, F(1, 51) = 1.36, p > 0.20]. Because the l-h TC intake of RES rats was significantly greater than that of NRES rats, the increased 1-h PDC intake of RES rats may have been a by-product of an increased consumption of all diets during the 1-h test periods and may not have been specific to protein-containing diets. To investigate this possibility, 1-h PDC and TC intakes were separately added across test days for each subject, and the percentage of total calories derived from the protein-containing diet during 1-h test periods was calculated. (It was not possible to carry out this procedure using the data from individual 1-h periods because l - h TC intake was sometimes zero.) As expected, a three-way A N O V A (test period protein diet × protein access × protein likeness) revealed a significant main effect of the protein access variable, F(1, 51) = 36.5, p < 0.001, with RES animals deriving a substantially greater percentage of total calories from the protein diet during 1-h test periods than did NRES animals (59.4 ± 2.47 vs. 36.2 ± 3.17, respectively). However, neither the protein likeness main effect, F ( 1 , 5 1 ) = 1.99, p > 0.10, nor the protein access × protein
likeness interaction, F(I, 51) = 1.47, p > 0.10, proved significant. DISCUSSION
This experiment replicates the earlier finding that rats having time-limited access to protein demonstrate substantial and seleco
RES-SAME
~
RES-DIFFERENT
~x
NRES*SAME
o
NRES-DIFFERENT
Teat-period protein diet: Soy 30
o 20
== ¢o
_¢1o
1
A
¢:
':
2
3
4
5
6
7
8
9
6
7
8
9
Test-period protein diet: Casein 3O
A
~2o ClO
1
2
3
4
5
Test periods
FIG. 3. Mean total energy derived from all dietary sources during l-h test periods of the experimental phase. All rats had continuous access to fat and carbohydrate diets throughout the experiment. See legend of Fig. 2 for further details.
PROTEIN SELECTION
663
tive increases in consumption of a protein-containing diet during the daily 1-h periods in which this diet is available (7). As Table 3 indicates, the protein intake of protein-restricted rats during 1h test periods was almost three times greater than that of nonrestricted rats that had continuous access to a single protein-containing diet for 24 h/day. Results of this sort have often been interpreted as providing evidence for the existence of a specific appetite for protein (6,7,22,25). The present results also draw attention to the role that the oral-sensory properties of diets may play in diet selection. As mentioned above, protein-restricted rats consumed much more protein during 1-h test periods than did nonrestricted rats. However, nonrestricted rats consumed 60% more protein during 1-h test periods when the protein-containing diet presented during the test period was different from the protein diet that was available during the remainder of the day (see Table 3). Of course, for the protein-restricted rats, the protein diet presented during test periods was always novel relative to the fat and carbohydrate diets, which were continuously available. Thus, it appears that some of the increase in protein consumption occurring in the protein-restriction paradigm results from the relative novelty of the protein-containing diet that is presented during test periods and is not attributable to a specific appetite for protein. This result is consistent with previous findings that there is an important stimulus-specific component to satiety (4,23,28), and it serves as a reminder that both the oral-sensory properties and macronutrient content of diets must be considered in the interpretation of diet selection studies. The results of this experiment do not address the nature of the other factors that may contribute to the selective increase in protein intake that occurs in response to time-restricted access to protein. There is evidence to suggest that both innate factors (6) and learning (3,11) may be involved in protein selection. Although the results presented here do not speak to the relative importance of these factors, they are consistent with a protein appetite interpretation. In contrast to related theories that describe
specific mechanisms for carbohydrate selection, the existence of a protein selection mechanism seems more likely for several resasons. First, in normally nourished rats and humans there are no nutritional requirements for carbohydrates (9). In addition, the evolution of food intake regulation may not have involved selective pressures that favoured mechanisms limiting carbohydrate intake (13). Finally, animals cannot store protein efficiently, yet protein accounts for almost half of the dry weight of a typical cell and must be obtained in the diet on a continuing basis to fulfill a variety of vital functions (24). Nevertheless, the results of the present study do not provide strong evidence that protein-restricted rats are specifically responding to the macronutrient (i.e., protein) content of the test diet. For example, the possibility remains that chronic protein restriction will cause rats to consume substantial amounts of any diet that is available on an intermittent basis, regardless of the macronutrient content of this diet. According to this interpretation, protein-restricted rats in this experiment selectively increased their consumption of the protein-containing diet that was presented during 1-h test periods because it was relatively novel, and they would have demonstrated this increased intake even if the test period diet had contained no protein at all. This type of neophilia has been observed in other situations in which rats have been subjected to chronic shortages of necessary nutrients, and in some situations it allows rats to cope with dietary imbalances through learning (10). The selective increase in protein intake that occurs in response to time-restricted access to protein may depend on the learning of a preference for the diet that alleviates the protein deficit, but further research is needed to evaluate this possibility. ACKNOWLEDGEMENTS This research was supported by grants from the Natural Sciences and Engineering Research Council of Canada to M. D. Holder (A 1221) and to D. DiBattista (OGP0003340). The authors wish to thank Sun-mi Shin for her assistance in carrying out the research.
REFERENCES 1. Anderson, G. H. Diet, neurotransmitters and brain function. Br. Med. Bull. 37:95-100; 1981. 2. Anderson, G. H.; Johnston, J. L. Nutrient control of brain neurotransmitter synthesis and function. Can. J. Physiol. Pharmacol. 61:271-281; 1983. 3. Baker, B. J.; Booth, D. A.; Duggan, J. P.; Gibson, E. L. Protein appetite demonstrated: Learned specificity of protein-cue preference to protein need in adult rats. Nutr. Res. 7:481-487; 1987. 4. Clifton, C.; Burton, P. G.; Sharp, C. Rapid loss of stimulus-specific satiety after consumption of a second food. Appetite 9:149-156; 1987. 5. Denton, D. The hunger for salt. New York: Springer-Verlag; 1982. 6. Deutsch, J. A.; Moore, B. O.; Heinrichs, S. C. Unlearned specific appetite for protein. Physiol. Behav. 46:619-624; 1989. 7. DiBattista, D. Effects of time-restricted access to protein and to carbohydrate in adult mice and rats. Physiol. Behav. 49:263-269; 1991. 8. Fernstrom, J. D. Dietary effects on brain serotonin synthesis: Relationship to appetite regulation. Am. J. Clin. Nutr. 42:1072-1082; 1985. 9. Fernstrom, J. D. Food-induced changes in brain serotonin synthesis: Is there a relationship to appetite for specific macronutrients. Appetite 8:163-182; 1987. 10. Garcia, J.; Hankins, W. G.; Coil, J. D. Koalas, men, and other conditioned gastronomes. In: Milgram, N.; Krames, L.; Alloway, T., eds. Food aversion learning. New York: Plenum Press; 1977. 11. Gibson, E. L.; Booth, D. A. Acquired protein appetite: Dependence on a protein specific need state. Experientia 42:1003-1004; 1986.
12. Hill, W. F. Effects of mere exposure on preferences in nonhuman mammals. Psychol. Bull. 85:1177-1198; 1978. 13. Holder, M. D.; Huether, G. Role of prefeedings, plasma amino acid ratios and brain serotonin levels in carbohydrate and protein selection. Physiol. Behav. 47:113-119; 1990. 14. Huynh, H.; Feldt, L. S. Estimation of the Box correction for degrees of freedom from sample data in randomized block and split-plot designs. J. Educ. Stat. 1:69-82; 1976. 15. Kanarek, R. B.; Ho, L. Patterns of nutrient selection in rats with streptozotocin-induced diabetes. Physiol. Behav. 32:639-645; 1984. 16. Kanarek, R. B.; Marks-Kaufman, R. Increased carbohydrate consumption by rats as a function of 2-deoxy-D-glucose administration. Physiol. Behav. 30:47-50; 1983. 17. Kanarek, R. B.; Marks-Kaufman, R.; Lipeles, B. J. Increased carbohydrate intake as a function of insulin administration in rats. Physiol. Behav. 25:779-782; 1980. 18. Kirk, R. E. Experimental design, 2nd ed. Monterey, CA: Brooks/ Cole; 1982. 19. Krieckhaus, E. E.; Wolf, G. Acquisition of sodium by rats: Interaction of innate mechanisms and latent learning. J. Comp. Physiol. Psychol. 68:197-201; 1968. 20. Larue-Achagiotis, C.; Martin, C.; Verger, P.; Louis-Sylvestre, J. Dietary self-selection vs. complete diet: Body weight gain and meal pattern in rats. Physiol. Behav. 51:995-999; 1992. 21. LeMagnen, J. Hyperphagie provoqure chez le rat blanc par l'altrration du mrchanisme de satiet6 pEriphrrique, C. R. Soc. Biol. (Paris) 147:1753-1757; 1956.
664
22. Li, E. T. S.; Anderson, G. H. Meal composition influences subsequent food selection in the young rat. Physiol. Behav. 29:779 783; 1982. 23. Mook, D. G.: Dreifuss, S.; Keats, P. H. Satiety for glucose solution in the rat: The specificity is postingestive. Physiol. Behav. 36:897901; 1986. 24. Pike, R. 1,.: Brown, M. L. Nutrition: An integrated approach, 3rd ed. New York: Macmillan; 1986. 25. Piquard, F.: Shaefer, A.; Habery, P. Influence of fasting and protein deprivation on food self-selection in the rat. Physiol. Behav. 20:771 778; 1978. 26. Richter, C. Total self-regulatory functions in animals and human beings. Harvey Lect. 38:63 103; 1943. 27. Rolls, B. J.; Rowe, E. A.; Rolls, E. T.; Kingston, B.: Megson, A.;
ttOLI)ER
28.
29. 30.
31.
AND DIBA'FTISTA
Gunary, R. Variety in a meat enhances h)od intake in man. Physiol. Behav. 26:215 221: 1981. Rolls, B. J." van Duijvenw)order, P. M.: Rowe, I~i. A. Variety in the diet enhances intake in a meal and contributes t~) the development of obesity in the rat. Physiol. Beha~. 3 1 : 2 1 - 2 7 : 1983. Rozin, P. Specific aversions as a component o | specilic hungers. J. Comp. Physiol. Psychol. 64:237--242; 1967. Rozin, P.: Rogers, W. L. Novel diet preferences in vitamin deficient rats and rats recovered from vitamin deficiencies. J. Comp. Physiol. Psychol. 63:421 428; 1967. Spiegel, T. A.: Stellar, E. Effects of variety on li~od intake of underweight, normal-weight, and overweight women, Appetite 15:4761 : 1990.
Pergamon
Effects of Time-Restricted Access to Protein and of Oral-Sensory Cues on Protein Selection MARK
D. H O L D E R *l A N D D A V I D
DIBATTISTAt
*Okanagan University College, Kelowna, British Columbia, Canada and "pBrock University, St. Catharines, Ontario, Canada R e c e i v e d 7 J u n e 1993 HOLDER, M. D. AND D. DIBA'Iq'ISTA. Effects of time-restricted access to protein and of oral-sensory cues on protein selection. PHYSIOL BEHAV 55(4) 659-664, 1994.--The effects on protein consumption of restricting access to protein and of varying the oral-sensory properties of protein diets were measured. During the initial phase of the study, rats were maintained on a selfselection diet in which three different macronutrient sources (carbohydrate, fat, and either soy-based or casein-based protein diets) were continuously available. For the remaining 9 days of the study, half of the rats were protein deprived for 23 h each day and the other half continued to receive the same protein diet during this 23-h period. The remaining 1 h of each day was a test period in which all rats had access to a protein diet that was either the same as or different from the one they had received in the initial phase. Compared to the nonrestricted rats, the protein-restricted rats consumed more than twice as much of the available protein diet during 1-h test periods. For the nonrestricted rats, those that received a different protein diet during the 1-h test periods consumed 60% more of the protein diet than did those that received the same protein diet. These results indicate that increases in protein consumption following protein deprivation can be attributed, at least in part, to the oral-sensory properties of diets and not necessarily to a specific protein appetite. Protein selection
Protein restriction
Rat
W H E N rats are allowed to select among separate sources of protein, fat, and carbohydrate, their total energy intake and the proportion of total energy derived from each dietary source are typically quite stable (20,26). Furthermore, self-selecting rats will often make appropriate adjustments in nutrient intake in response to challenges such as streptozotocin-induced diabetes (15), or exogenous administration of insulin (17) and 2-deoxyglucose (16). One type of challenge that has been studied extensively relates to experimentally induced shortages in particular dietary nutrients. Numerous studies have demonstrated that rats will selectively increase their intake of diets containing substances such as sodium (5,19) and thiamine (29,30) following a period of time during which these substances have been unavailable for consumption. It has similarly been observed that when access to dietary protein is restricted for a period of time, rats selectively increase their consumption of a protein-containing diet when it once again becomes available (6,7,22,25). Findings of this sort are often interpreted as evidence for the existence of a protein selection mechanism, that is, a specific appetite for protein that may be either innate (6) or acquired (3,11). However, changes in the availability of a protein-containing diet are usually confounded with changes in the availability of oral-sensory stimuli.
An alternative explanation of the selective increase in protein consumption following restricted access to dietary protein focuses on the importance of oral-sensory stimuli. It is well established that there is an important sensory-specific component to satiety. For example, rats generally prefer the less recently tasted of two food items (12). In another study, rats were given a 2-h meal consisting of one diet (21). When the diet was presented every 30 min with a different flavour added to it (i.e., four different flavours in total), rats consumed two to three times more of the diet than when only one flavour was used. Also, rats that have consumed a concentrated sugar solution to satiety will vigorously ingest either lab chow or powdered sugar when it is offered, but will refuse to drink more of the sugar solution (23). Furthermore, both rats (4,28) and humans (27,31) consume larger meals when presented with an assortment of different food items than when presented with a single food item. Thus, the selective increase in consumption of a protein-containing diet following protein restriction may be attributable, at least in part, to the novelty of the oral-sensory properties of the protein-containing diet rather than to the existence of a specific appetite for protein. This alternative explanation seems plausible in light of related research on carbohydrate selection. A recent theory proposed that carbohydrate selection is modulated by a negative feedback loop
Requests for reprints should be addressed to M. D. Holder, Okanagan University College, Department of Psychology, 3333 College Way, Kelowna, British Columbia, Canada V 1Y 1V7.
659
H()IA)ER AND I)IBATTISTA
660
involving brain serotonin (1,2,8,22). However, much of the research cited in support of this theory contk)unded the macronutrient composition of diets with their oral sensory properties. More recent studies have controlled the oral-sensory properties of the diets with conditioning techniques (11) or by using carbohydrate diets that differed in their oral-sensory properties (13). These studies have indicated that results that were interpreted as showing carbohydrate selection might better be interpreted as showing the selection of specific oral-sensory properties. The study reported here was conducted to evaluate the importance of the oral-sensory properties of a protein-containing diet in causing the selective increase in protein intake that is observed in the chronic protein-restriction paradigm (7). Briefly, rats were permitted to self-select from three separate diets: a carbohydrate diet, a fat diet, and a protein diet that contained either casein or soy protein. The carbohydrate and fat diets were continuously available throughout the experiment. After a baseline period during which all three diets were freely available, proteinrestricted animals were allowed access to one of the protein diets during a l-h test period each day, but bad no access to dietary protein during the remaining 23 h/day. Nonrestricted animals had access to dietary protein for 24 h/day; the original protein diet of the baseline period was available for 23 h/day, and one of the protein diets was available during the l-h test period. For both protein-restricted and nonrestricted rats, the protein diet presented during l-h test periods was either the same as or different from the protein diet available to the animal during the baseline period. METHOD
Subjects Sixty-nine Sprague-Dawley, albino male rats weighed an average of 258 g at the start of the experiment. The rats were housed individually in clear Plexiglas cages. All procedures were conducted in these cages. Room lights were on from 0700 to 1900 h in the animal colony. Water was continuously available.
Apparatus The rats were maintained on a self-selection diet consisting of separate carbohydrate, fat, and either protein-based or soybased protein diets, each supplemented with vitamins and minerals. The composition of each diet is described in Table 1. Pilot work had shown that the two protein diets differed in oral-sensory cues; a conditioned flavor aversion acquired to one of the protein diets did not generalize to the other protein diet. The diets were presented in spill-resistant containers that were placed at the front of each cage. The location of each individual diet varied each day. The containers were made by gluing a 250-ml glass jar inside a larger glass food cup. Food was placed in the glass jar with spillage being caught in the surrounding cup. A flat piece of aluminum (12 × 12 × 0.08 cm) with rounded comers was glued to the bottom of the cup to prevent tipping of the food container. The aluminum was secured to the home cage with tape. Spillage that occurred despite these precautions was carefully noted.
Procedu re Throughout the study, all rats had continuous access to the fat and carbohydrate diets. Access to the two protein diets, soy and casein, varied across different groups of rats and different phases. Phase 1, which lasted 10 days, was an adaption phase
TAB Lt:'~ l ( OMPOSITION OF IEXPERIMENTAI. I)IE3S ( P E R C E N T A G E OF T O T A I W E I G t l T I
So 5 Protein Soy protein* Caseint Vegetable
Casein Prolein
('arbohydrate
92 -
92
shortening$
Sucrose§ Cornstarch', Methioninet Vitamin mixture1[ Mineral mixture# Alphacel~ Caloric density (kcal/g)
|:at
--
94 .
.
2 1.5 3.5 1 3.62
.
.
. . 2 1.5 3.5 l 3.83
-1.5 3.5 I 8.54
47 47 1.5 3.5 I 3.84
* ICN Canada. Isolated soy protein contains 92% protein, 4.1% ash, 1% fat, plus fibre and moisture. ICN Canada. $ Crisco, Procter and Gamble. § Atlantic Sugar, Montreal, Quebec. AIN Vitamin Mixture 76, ICN Canada. Contains 97.3% sucrose. # AIN Mineral Mixture 76, ICN Canada. Contains 11.8% sucrose. that allowed each rat to become familiar with the different diets. In addition to the fat and carbohydrate diets, half the rats had continuous access to the soy diet and the remaining rats had continuous access to the casein diet. Consumption of all diets was measured for the last 3 days of phase 1. Based on their consumption of the protein diet, each rat that had access to soy was assigned to one of four equal-sized groups such that consumption was similar across groups. The rats that had access to casein were similarly assigned to one of four groups in preparation for phase 2. During phase 2 of the study, which lasted 9 days, all rats had continuous access to the fat and carbohydrate diets, but access to the protein-containing diets varied across groups. The rats never received access to more than one protein diet at one time. Animals had access to dietary protein either for 24 h/day [nonrestricted (NRES)] or only during a l-h test period from 1730 to 1830 h each day [restricted (RES)]. During this test period, all rats received access to a protein diet, and the consumption of all diets (protein, fat, and carbohydrate) during this l-h period was measured. The protein diet available during the l-h test period was either the same as that received during phase 1 (same) or different from it (different). For the NRES rats, which had continuous access to protein, the protein diet available during the 23-h nontest period was always the same as that given during phase 1. Table 2 summarizes the eight groups created with this 2 (phase 1 protein diet: soy or casein) × 2 (protein access: RES or NRES) x 2 (protein likeness: test period protein diet either same as or different from phase 1 protein diet) factorial design.
Statistical Analyses Missing data points resulting from spillage, etc., were estimated using the procedure outlined by Kirk (18). Protein dietderived calories (PDC) and total calories derived from all dietary sources (TC) were calculated for both 1-h test periods and 24-h periods. The SPSS-PC+ statistical package, version 4.0, was used to perform analyses of variance; degrees of freedom were appropriately adjusted for within-subjects variables (14).
PROTEIN SELECTION
661
TABLE 2 SUMMARY OF RESEARCH DESIGN
Group
Phase 1 Protein Diet*
Protein Diet Available During Nontest Periods?t
Test Period Protein Diets
Soy-NRES-Same Soy-NRES-Diff Soy-RES-Same Soy-RES-Diff Casein-NRES-Same Casein-NRES-Diff Casein-RES-Same Casein-RES-Diff
Soy Soy Soy Soy Casein Casein Casein Casein
Yes Yes No No Yes Yes No No
Same (soy) Different (casein) Same (soy) Different (casein) Same (casein) Different (soy) Same (casein) Different (soy)
* Either the soy or the casein protein diet was continuously available during phase 1. 1-During phase 2, NRES rats had access during 23-h nontest periods to the same protein diet that was available during phase 1. RES rats had no access to either protein diet during the 23-h nontest periods. $ The protein diet available during 1-h test periods was either the same as (Same) or different from (Diff) the protein diet available during phase 1.
1-h intake of the casein-based diet being greater than that of the soy-based diet (8.62 _+ 0.88 vs. 5.42 _+ 0.48 kcal/h, respectively). If the oral-sensory cues provided by the diet are an important factor in the selection of protein diets during the 1-h test period, then consumption of the different protein diet should be greater than that of the same protein diet for the NRES rats. To determine whether this was the case, the protein access × protein likeness interaction for the 1-h PDC data, F(1, 54) = 2.92, p < 0.10, was investigated further. Results of this analysis are described in Table 3, which provides a summary of the 1-h PDC data averaged across both the nine test periods and the two test period protein diets. The 1-h PDC intake of NRES rats proved to be significantly greater (p < 0.05) when the test period protein diet was different from, rather than the same as, the phase l protein diet, which was available to NRES rats during the remaining 23 h of the day. Thus, although the 1-h PDC intake of RES rats was significantly greater than that of NRES rats regardless of the nature of the diet (p < 0.05), novelty of the protein diet presented during l-h test periods enhanced the 1-h PDC intake of NRES rats by some 60%. Figure 3 shows TC intake daring the 1-h test periods. A fourway A N O V A (test period protein diet × protein access × protein likeness × days) revealed only three significant main effects. As Fig. 3 suggests, rats generally tended to increase their 1-h TC intake during phase 2 [main effect of days: F(8, 301) = 8.65, p
RESULTS
Intake During 24-h Periods Figure 1 shows TC intake during consecutive 24-h periods. These data were analyzed using a four-way A N O V A with three between-subjects variables (phase 1 protein diet: soy or casein; protein access: RES or NRES; protein likeness: test period protein diet either same as or different from phase 1 protein diet) and one within-subjects variable (days). This analysis revealed a significant main effect of the protein access variable, F(I, 50) = 5.57, p < 0.05; over the course of the experiment, RES rats ingested about 6% fewer total calories than did NRES rats (mean - SE: 105.5 _ 2.35 vs. 112.5 ± 2.06 kcal/day, respectively). The greatest difference in TC intake between these groups occurred during the first few days of phase 2, as indicated by a significant protein access x days interaction, F(11,383) = 2.25, p < 0.05. In addition, it was found that daily TC intake over the course of the study was greater when rats were maintained on the soy-based diet (113.0 ± 2.40 kcal/day) rather than on the casein-based diet [105.2 _ 1.92 kcal/day; main effect of phase 1 protein diet, F(1, 50) = 7.03, p < 0.05].
Intake During 1-h Test Periods Figure 2 illustrates PDC intake during the 1-h test periods of phase 2 of the experiment. The PDC intake data were analyzed using a four-way A N O V A with three between-subjects variables (test period protein diet: soy or casein; protein access: RES or NRES; protein likeness: test period protein diet either same as or different from phase I protein diet) and one within-subjects variable (days). As expected, the main effect of the protein access variable was highly significant, F(1, 54) = 53.3, p < 0.001; thus, the 1-h PDC intake of RES rats was more than twice that of NRES rats (9.76 ± 0.68 vs. 4.44 + 0.52 kcal/h, respectively). In additiom as Fig. 2 suggests, the 1-h PDC intake of RES rats, but not of NRES rats, tended to increase during phase 2, as indicated by a significant protein access x days interaction, F(8, 257) = 2.25, p < 0.05. A significant main effect of the test period protein diet variable was also found, F(1, 54) = 19.7, p < 0.001, with
--~-
RES-SAME
~
RES-DIFFERENT
NRES-SAME
o
NRES-DIFFERENT
Phase 1 protein diet: Soy 140
2
120
~ _c
A
8
100
80
--
L 1 I
£
•
__L
2 3 - - I Phase 1
1
£
_J
2
3
J
1
4
5
I
6
L
I~
7
8
£
J 9 I
Phase
2
Phase 1 protein diet: Casein 140 e
120
•~
loo
1
2 Phase
3
1
2
3
1
4
5
6
7
8
9
Pt~ase 2
Days
FIG. 1. Mean total energy derived from all dietary sources during 24-h periods. All rats had continuous access to fat and carbohydrate diets throughout the experiment. Phase 1: rats had continuous access to either the soy-based or the casein-based protein diet. Phase 2: protein-restricted rats (RES) had access to a protein-containing diet only during a 1-h test period each day; this diet was either the same as (SAME) or different from (DIFFERENT) the protein-containing diet available during phase I. Nonrestricted rats (NRES) had access to dietary protein for 24 h/day; the protein-containing diet of phase 1 was available for 23 h/day, and the protein diet available during the 1-h test period was either the same or different.
662
ttOLDER AND D~BATTISTA
'~
RES-SAME
o
RES-DIFFERENT
~'~
NRFS-SAME
~
NRES-DIFFERENT
TABLE 3 MEAN
Test-period protein diet: Soy
ENERGY
CONTAINING
DERIVED
DIETS
DURING
FROM
PROTI/IN
I-H TEST
PERIO[)S
Nature of Test Diet
~12i
- -
()
41
'
o
,^.
i
A
A
A
3
4
5
c~ ~,
o A
A
7
8
2
6
9
Test-period protein diet: Casein
~12
~8 --
c~ A
o
4 i
0! 1
2
3
4
5
6
7
Same
Different
Restricted Nonrestricted
10.05" (1.10) 3.37 (0.50)
9.52 10.g5) 5.40~: (0.83)
Values are means (kcal), with SE shown in parentheses. * Protein-restricted rats had access to a protein-containing diet only during a l-h test period each day; this diet was either the same as (Same) or different from (Different) the protein-containing diet available during phase 1. Nonrestricted rats had access to dietary protein for 24 h/day; the protein-containing diet of phase 1 was available for 23 h/day, and the protein diet available during the l-h test period was either the Same or Different. f Significantly greater than other value in same column, p < 0.05. $ Significantly greater than other value in same row, p < 0.05.
0 ~
1
Protein Access
8
Test periods
FIG. 2. Mean energy derived from protein-containing diets during 1-h test periods of phase 2. All rats had continuous access to fat and carbohydrate diets throughout the experiment. Protein-restricted rats (RES) had access to a protein-containing diet only during a 1-h test period each day; this diet was either the same as (SAME) or different from (DIFFERENT) the protein-containing diet available during phase 1. Nonrestricted rats (NRES) had access to dietary protein for 24 h/day; the protein-containing diet of phase 1 was available for 23 h/day, and the protein diet available during the I -h test period was either the same or different.
< 0.001]. It was also found that rats having access to the caseinbased diet during test periods consumed more TC than did rats having the soy-based diet [17.2 ± 1.29 vs. 12.1 ± 0.67 kcal/h, respectively; main effect of test period protein diet: F(1, 51) = 13.99, p < 0.001 ]. Finally, RES rats consumed significantly more calories during 1-h test periods than did NRES rats [16.8 _ 1.06 vs. 12.6 ± 1.05 kcal/h, respectively; main effect of protein access: F(I, 51) = 10.25, p < 0.01]. None of the effects of the protein likeness variable reached significance [e.g., protein access × protein likeness, F(1, 51) = 1.36, p > 0.20]. Because the l-h TC intake of RES rats was significantly greater than that of NRES rats, the increased 1-h PDC intake of RES rats may have been a by-product of an increased consumption of all diets during the 1-h test periods and may not have been specific to protein-containing diets. To investigate this possibility, 1-h PDC and TC intakes were separately added across test days for each subject, and the percentage of total calories derived from the protein-containing diet during 1-h test periods was calculated. (It was not possible to carry out this procedure using the data from individual 1-h periods because l - h TC intake was sometimes zero.) As expected, a three-way A N O V A (test period protein diet × protein access × protein likeness) revealed a significant main effect of the protein access variable, F(1, 51) = 36.5, p < 0.001, with RES animals deriving a substantially greater percentage of total calories from the protein diet during 1-h test periods than did NRES animals (59.4 ± 2.47 vs. 36.2 ± 3.17, respectively). However, neither the protein likeness main effect, F ( 1 , 5 1 ) = 1.99, p > 0.10, nor the protein access × protein
likeness interaction, F(I, 51) = 1.47, p > 0.10, proved significant. DISCUSSION
This experiment replicates the earlier finding that rats having time-limited access to protein demonstrate substantial and seleco
RES-SAME
~
RES-DIFFERENT
~x
NRES*SAME
o
NRES-DIFFERENT
Teat-period protein diet: Soy 30
o 20
== ¢o
_¢1o
1
A
¢:
':
2
3
4
5
6
7
8
9
6
7
8
9
Test-period protein diet: Casein 3O
A
~2o ClO
1
2
3
4
5
Test periods
FIG. 3. Mean total energy derived from all dietary sources during l-h test periods of the experimental phase. All rats had continuous access to fat and carbohydrate diets throughout the experiment. See legend of Fig. 2 for further details.
PROTEIN SELECTION
663
tive increases in consumption of a protein-containing diet during the daily 1-h periods in which this diet is available (7). As Table 3 indicates, the protein intake of protein-restricted rats during 1h test periods was almost three times greater than that of nonrestricted rats that had continuous access to a single protein-containing diet for 24 h/day. Results of this sort have often been interpreted as providing evidence for the existence of a specific appetite for protein (6,7,22,25). The present results also draw attention to the role that the oral-sensory properties of diets may play in diet selection. As mentioned above, protein-restricted rats consumed much more protein during 1-h test periods than did nonrestricted rats. However, nonrestricted rats consumed 60% more protein during 1-h test periods when the protein-containing diet presented during the test period was different from the protein diet that was available during the remainder of the day (see Table 3). Of course, for the protein-restricted rats, the protein diet presented during test periods was always novel relative to the fat and carbohydrate diets, which were continuously available. Thus, it appears that some of the increase in protein consumption occurring in the protein-restriction paradigm results from the relative novelty of the protein-containing diet that is presented during test periods and is not attributable to a specific appetite for protein. This result is consistent with previous findings that there is an important stimulus-specific component to satiety (4,23,28), and it serves as a reminder that both the oral-sensory properties and macronutrient content of diets must be considered in the interpretation of diet selection studies. The results of this experiment do not address the nature of the other factors that may contribute to the selective increase in protein intake that occurs in response to time-restricted access to protein. There is evidence to suggest that both innate factors (6) and learning (3,11) may be involved in protein selection. Although the results presented here do not speak to the relative importance of these factors, they are consistent with a protein appetite interpretation. In contrast to related theories that describe
specific mechanisms for carbohydrate selection, the existence of a protein selection mechanism seems more likely for several resasons. First, in normally nourished rats and humans there are no nutritional requirements for carbohydrates (9). In addition, the evolution of food intake regulation may not have involved selective pressures that favoured mechanisms limiting carbohydrate intake (13). Finally, animals cannot store protein efficiently, yet protein accounts for almost half of the dry weight of a typical cell and must be obtained in the diet on a continuing basis to fulfill a variety of vital functions (24). Nevertheless, the results of the present study do not provide strong evidence that protein-restricted rats are specifically responding to the macronutrient (i.e., protein) content of the test diet. For example, the possibility remains that chronic protein restriction will cause rats to consume substantial amounts of any diet that is available on an intermittent basis, regardless of the macronutrient content of this diet. According to this interpretation, protein-restricted rats in this experiment selectively increased their consumption of the protein-containing diet that was presented during 1-h test periods because it was relatively novel, and they would have demonstrated this increased intake even if the test period diet had contained no protein at all. This type of neophilia has been observed in other situations in which rats have been subjected to chronic shortages of necessary nutrients, and in some situations it allows rats to cope with dietary imbalances through learning (10). The selective increase in protein intake that occurs in response to time-restricted access to protein may depend on the learning of a preference for the diet that alleviates the protein deficit, but further research is needed to evaluate this possibility. ACKNOWLEDGEMENTS This research was supported by grants from the Natural Sciences and Engineering Research Council of Canada to M. D. Holder (A 1221) and to D. DiBattista (OGP0003340). The authors wish to thank Sun-mi Shin for her assistance in carrying out the research.
REFERENCES 1. Anderson, G. H. Diet, neurotransmitters and brain function. Br. Med. Bull. 37:95-100; 1981. 2. Anderson, G. H.; Johnston, J. L. Nutrient control of brain neurotransmitter synthesis and function. Can. J. Physiol. Pharmacol. 61:271-281; 1983. 3. Baker, B. J.; Booth, D. A.; Duggan, J. P.; Gibson, E. L. Protein appetite demonstrated: Learned specificity of protein-cue preference to protein need in adult rats. Nutr. Res. 7:481-487; 1987. 4. Clifton, C.; Burton, P. G.; Sharp, C. Rapid loss of stimulus-specific satiety after consumption of a second food. Appetite 9:149-156; 1987. 5. Denton, D. The hunger for salt. New York: Springer-Verlag; 1982. 6. Deutsch, J. A.; Moore, B. O.; Heinrichs, S. C. Unlearned specific appetite for protein. Physiol. Behav. 46:619-624; 1989. 7. DiBattista, D. Effects of time-restricted access to protein and to carbohydrate in adult mice and rats. Physiol. Behav. 49:263-269; 1991. 8. Fernstrom, J. D. Dietary effects on brain serotonin synthesis: Relationship to appetite regulation. Am. J. Clin. Nutr. 42:1072-1082; 1985. 9. Fernstrom, J. D. Food-induced changes in brain serotonin synthesis: Is there a relationship to appetite for specific macronutrients. Appetite 8:163-182; 1987. 10. Garcia, J.; Hankins, W. G.; Coil, J. D. Koalas, men, and other conditioned gastronomes. In: Milgram, N.; Krames, L.; Alloway, T., eds. Food aversion learning. New York: Plenum Press; 1977. 11. Gibson, E. L.; Booth, D. A. Acquired protein appetite: Dependence on a protein specific need state. Experientia 42:1003-1004; 1986.
12. Hill, W. F. Effects of mere exposure on preferences in nonhuman mammals. Psychol. Bull. 85:1177-1198; 1978. 13. Holder, M. D.; Huether, G. Role of prefeedings, plasma amino acid ratios and brain serotonin levels in carbohydrate and protein selection. Physiol. Behav. 47:113-119; 1990. 14. Huynh, H.; Feldt, L. S. Estimation of the Box correction for degrees of freedom from sample data in randomized block and split-plot designs. J. Educ. Stat. 1:69-82; 1976. 15. Kanarek, R. B.; Ho, L. Patterns of nutrient selection in rats with streptozotocin-induced diabetes. Physiol. Behav. 32:639-645; 1984. 16. Kanarek, R. B.; Marks-Kaufman, R. Increased carbohydrate consumption by rats as a function of 2-deoxy-D-glucose administration. Physiol. Behav. 30:47-50; 1983. 17. Kanarek, R. B.; Marks-Kaufman, R.; Lipeles, B. J. Increased carbohydrate intake as a function of insulin administration in rats. Physiol. Behav. 25:779-782; 1980. 18. Kirk, R. E. Experimental design, 2nd ed. Monterey, CA: Brooks/ Cole; 1982. 19. Krieckhaus, E. E.; Wolf, G. Acquisition of sodium by rats: Interaction of innate mechanisms and latent learning. J. Comp. Physiol. Psychol. 68:197-201; 1968. 20. Larue-Achagiotis, C.; Martin, C.; Verger, P.; Louis-Sylvestre, J. Dietary self-selection vs. complete diet: Body weight gain and meal pattern in rats. Physiol. Behav. 51:995-999; 1992. 21. LeMagnen, J. Hyperphagie provoqure chez le rat blanc par l'altrration du mrchanisme de satiet6 pEriphrrique, C. R. Soc. Biol. (Paris) 147:1753-1757; 1956.
664
22. Li, E. T. S.; Anderson, G. H. Meal composition influences subsequent food selection in the young rat. Physiol. Behav. 29:779 783; 1982. 23. Mook, D. G.: Dreifuss, S.; Keats, P. H. Satiety for glucose solution in the rat: The specificity is postingestive. Physiol. Behav. 36:897901; 1986. 24. Pike, R. 1,.: Brown, M. L. Nutrition: An integrated approach, 3rd ed. New York: Macmillan; 1986. 25. Piquard, F.: Shaefer, A.; Habery, P. Influence of fasting and protein deprivation on food self-selection in the rat. Physiol. Behav. 20:771 778; 1978. 26. Richter, C. Total self-regulatory functions in animals and human beings. Harvey Lect. 38:63 103; 1943. 27. Rolls, B. J.; Rowe, E. A.; Rolls, E. T.; Kingston, B.: Megson, A.;
ttOLI)ER
28.
29. 30.
31.
AND DIBA'FTISTA
Gunary, R. Variety in a meat enhances h)od intake in man. Physiol. Behav. 26:215 221: 1981. Rolls, B. J." van Duijvenw)order, P. M.: Rowe, I~i. A. Variety in the diet enhances intake in a meal and contributes t~) the development of obesity in the rat. Physiol. Beha~. 3 1 : 2 1 - 2 7 : 1983. Rozin, P. Specific aversions as a component o | specilic hungers. J. Comp. Physiol. Psychol. 64:237--242; 1967. Rozin, P.: Rogers, W. L. Novel diet preferences in vitamin deficient rats and rats recovered from vitamin deficiencies. J. Comp. Physiol. Psychol. 63:421 428; 1967. Spiegel, T. A.: Stellar, E. Effects of variety on li~od intake of underweight, normal-weight, and overweight women, Appetite 15:4761 : 1990.