Exercise training modifies nutrient self-selection in rats

Physiology & Behavior,Vol. 56, No. 2, pp. 367-372, 1994 Copyright © 1994 ElsevierScienceLtd Printed in the USA. All rights reserved 0031-9384/94 $6.00 + .00

Pergamon 0031-9384(94)E0092-I

Exercise Training Modifies Nutrient Self-Selection in Rats CHRISTIANE

LARUE-ACHAGIOTIS,

1 NATHALIE

RIETH AND

JEANINE

LOUIS-SYLVESTRE

Neurobiologie de la Nutrition, Bdt B, Universitd Pierre et Marie Curie, 4, place Jussieu 75252 Paris Cedex 05, France R e c e i v e d 30 N o v e m b e r 1993 LARUE-ACHAGIOTIS, C., N. RIETH AND J. LOUIS-SYLVESTRE. Exercise training modifies nutrient self-selection in rats. PHYSIOL BEHAV 56(2) 367-372, 1994.--Few studies have examined the effects of exercise training on macronutrient selfselection in rats. It has been observed that trained rats decreased carbohydrate and increased fat and protein intakes. In the present experiment, total energy intake and macronutrient self-selection were examined in adult male rats placed on a self-selection regimen and submitted to 2 h of treadmill exercise daily for 20 days at the beginning of the nocturnal period. Two control groups of rats were examined during the same time: a sedentary group that was food and water deprived during the same 2 h while trained rats were exercising, and a control group that was examined only for body weight gain and 24-h food intake. Food intakes of sedentary and trained rats were continuously recorded. At the end of the experiment, body weight of trained rats was lower than that of sedentary and control rats. The 24-h cumulative intake of trained rats was significantly reduced; this reduction was due to a decrease in fat intake, whereas carbohydrate intake was increased. In sedentary rats, 24-h intake was not modified but fat intake was significantly increased from the beginning to the end of the experiment. During the first 6 h of the night, protein and fat intakes of trained rats were reduced, and carbohydrate ingestion remained the same. Daytime food intake represented only 8.7% of the 24-h intake. Exercise training significantly increased this intake. It is noteworthy that during the middle part of the day (39 h) trained rats significantly enhanced protein and carbohydrate ingestion. In sedentary rats during the 3 h preceding fasting, fat intake increased. The present results could be ascribed to an anticipatory phenomenon. In trained rats, an increased carbohydrate intake, early in the light period preceding exercise, might improve exercise endurance, whereas sedentary rats anticipate an obligatory induced fast by increasing fat intake. Exercise training

Self-selection

Fasting

M A N Y studies have examined the effects of exercise training on food intake and body weight in laboratory animals. Conflicting results have been obtained due to the large variability in experimental protocols in terms of exercise: type [swimming vs. treadmill (20)], intensity (13), duration, time of the day, etc., subjects [male vs. female (4)], and environmental conditions [temperature, time of the year, etc. (23)]. Overall, it seems that in a similar exercise program, female rats maintain their body weight and do not modify their food intake, whereas male rats decrease food intake and lose weight (1,21). Few studies, on the other hand, have investigated the effect of training on macronutrient self-selection in rats. Andik et al. (3) reported that male rats given a choice of high carbohydrate, fat, and protein diets and trained for 40 days ( 6 - 7 km/day) at low intensity (0.75 km/h) ate more fat and protein and less carbohydrate. Total energy intake was therefore augmented. More recently, Pariskova and Stankova (18) observed that trained rats (18 m/min; 50 min/day) increased carbohydrate and decreased fat intakes, while body weight remained unchanged; however, body composition changed. In a previous study (15), examining the effects of an acute exercise performed at the end of the day-

time period on self-selection, we observed that total energy intake was significantly reduced. This was mainly due to an early reduction in carbohydrate and protein intakes, occurring just after the exercise, whereas a fat intake decrease appeared later, in the last part of the night. To explore the effect of training on body weight gain and macronutrient selection, this study was performed in adult male rats placed on a self-selection regimen and submitted to a daily 2-h training exercise at the beginning of the night (i.e., the beginning of the activity period for the rat), for 20 days. Two other groups of rats were examined. A sedentary group of rats was food and water deprived during the 2 h while the experimental rats were trained. The food intakes of these two groups were continuously recorded. A control group was also examined, but only for body weight gain and 24-h food intake (protein, fat, and carbohydrate). METHOD

Animals Male adult Wistar rats (Ardenay, France) were used. They were housed individually in square Plexiglas cages and placed in

To whom requests for reprints should be addressed. 367

368

a quiet, temperature-controlled room (23 ± 1°C) with a 12-12 h light-dark cycle (0300-1500 h light). Rats had continuous access to water and food unless otherwise indicated (i.e., training time: 1500-1700 h). They were provided with three separate food cups for the three macronutrients: protein, fat, or carbohydrate. The feeding patterns for each food cup (or item) were continuously recorded on a chart recorder by means of pens connected to electric strain gauge microbalances, one chanel for each item, as previously described (14).

Diets The three macronutrient diets used were proteins, fats, and carbohydrates. The protein diet (metabolizable energy, 3.43 kcal/ g) was composed of 93% casein, 2% cellulose powder, 4% salt mixture (U.A.R, No. 205, Villemoisson/Orge), and 1% vitamins (U.A.R, No. 200). The fat diet (7.88 kcal/g) consisted of 91% lard, 2% sunflower oil, 2% cellulose powder, 4% salt mixture, and 1% vitamins. The carbohydrate diet (3.45 kcal/g) was composed of 85% pregelatinized corn starch, 8% commercial grade sucrose, 2% cellulose powder, 4% salt mixture, and 1% vitamins. The protein and carbohydrate diets were in powdered form, although the fat one was semisolid.

Exercise Exercise was performed in a motor-driven, five-channel treadmill with 0° inclination at a speed of 1.2 km/h. A weakly electrified grid (50 V DC, 60 mA) was positioned at the rear of each compartment to ensure that the animals would continue to run throughout the exercise period. Adaptation to the treadmill was realized over 3 mornings during the food adaptation period. The speed and duration of the exercise were gradually increased (from 5 to 15 m/rain and 5 to 10 rain). Virtually no shocks were needed in trained rats. In addition, there were no visible signs of trauma produced by the running or the shocks.

LARUE-ACHAGIOTIS. RIETH AND LOLilS-.~YI VESTR~

didymal, retroperitoneal, mesenteric, and inguinal, were carefull3 removed and weighed.

Statistical Analysis Results were analyzed in terms of 24-h total energy intakes, night or day energy intakes, protein, fat, and carbohydrate intakes. Food intake was also expressed in terms of five periods of 4 successive days (P1 = days 1-4; P2 = days 5-8; P3 = days 9-12; P4 = days 13-16; P5 = days 17-20) and was compared to the 4-day baseline intake; four successive night and daytime food intakes were also cumulated and used for statistical comparisons. These cumulative 4-day periods were used to minimize the interday variations. All data are expressed as mean _+ SEM. Intragroup values were compared using analysis of variance (ANOVA) for repeated measures. Posteriori comparisons between means were made using Dunnetrs t-test. Differences between the trained and sedentary groups were tested using Student's t-test. RESULTS

Body Weight Body weights at the beginning of the experimental period were not significantly different (C = 259 _+ 12.1 g; S = 267.4 _+ 4.5 g; T = 265.4 _+ 6g ; F = 0.72, NS). The body weights of the trained group became significantly lower than that of the sedentary group on day 10 and therafter for all subsequent days. Sedentary rats were slightly but not significantlyheavier than rats of the control group. Regarding body weight gains (Fig. 1), a significant lower body weight gain was evident in trained rats compared to sedentary rats from day 5 to the end of the training. Moreover, at the end of the experiment, body weight gain of trained rats was significantly lower than that of control rats. The difference in body weight between sedentary and control rats did not reach statistical significance (t = 1.97, NS).

Procedure Rats (180-200 g at the beginning of the experiment) were allowed to adapt to the diets for a 2-week period. The positions of the food cups were changed daily. It has been observed that the introduction of food triggers some feeding activity in rats; therefore, food cups were refilled at the beginning of the dark period, during the training or fasting sessions. Body weight was recorded just before training. Following the 2-week adaptation period, food intake was continuously recorded. The last 4-day food intake records were used as a baseline. Thereafter, three groups of rats were used: a control group (C, n = 5 rats), a sedentary group (S, n = 10 rats), and a trained group (T, n = 10 rats). Trained rats were exercised while the sedentary ones were placed in new cages without food and water near the treadmill, so that the fasting and environmental conditions would be the same in the two groups of rats. Rats were run every day at the beginning of the nocturnal cycle throughout 20 days. During the first 4 days, duration and running speed were increased progressively to achieve a final intensity and duration of 20 m/rain, 2 h/day. This level was achieved on day 5 and continued throughout the duration of the experiment. Water was provided ad lib except during the 2-h running or fasting. Food intake of control rats was not recorded; however, food cups were weighed every day. At the end of the experiment, rats were sacrified by an overdose of pentobarbital. Four depots of white adipose tissue, epi-

White Adipose Tissue The total dissectable white adipose tissue was significantly reduced in trained rats (13.9 _+ 3.1 g) compared to sedentary (22.6 __+ 1.7 g) or control rats (18.6 -+ 2.3 g), F(2, 23) = 3.56, p < 0.05,

Body weight gain

g 100

-

.o-

0

80-

40.-0-- Control

20-

Sedentary

41- Trmed 0

-4

1

5

10

15

2 0 days

FIG. 1. Body weight gain in control, sedentary, and trained rats. p < 0.05 for trained rats vs. sedentary rats. p < 0.05 for trained rats vs. control rats.

TRAINING AND SELF-SELECTION IN RATS

Control rats

kcal 80-

369 Cumulative 24-h intake of trained rats was reduced significantly by training, F(20, 209) = 1.97, p < 0.05. This reduction was due to a decrease in fat ingestion, F(20, 209) = 1.65, p < 0.05, whereas carbohydrate intake was augmented, F(20, 209) = 1.75, p < 0.05. Protein intake was not modified. At the end of the experiment, when expressed in percent of total daily intake (Fig. 3), control rats had increased fat ingestion, decreased carbohydrate, and protein intake was maintained; sedentary rats had increased their fat ingestion and decreased protein and carbohydrate intakes. In contrast, trained rats had decreased significantly their fat intake (periods 4 and 5), maintained their protein intake, and increased their carbohydrate intake.

o @

~lf

70-

-- =

50-

,

40i

30 -!

10-

I

o base

II 5

10

.15

Ii

20 days

Nighttime Food Intake

Sedentary rats (2 hfainting)

koal 8070-

~'

.o

j iim

50-

3o • o-

:

I

I

j i

i

2O 10 o base

koal 70-

/I

I 5

10

15

20 days

Exercised-trained rats

605040302010-

0[ ~ e

5

10 I

Proteins

15 ~

Fete

In sedentary rats, the comparison of baseline intake with that of the five experimental periods indicates that total nocturnal energy intake was not significantly enhanced. Protein and carbohydrate intakes were not modified; however, fat ingestion was significantly reduced during the first period of fasting and then increased in the last one (Table 1). In trained rats, total energy intake was significantly decreased as soon as the first period of training. This reduction affected fat intake most significantly. The comparison between the two groups of rats indicated that, in the trained group compared to the sedentary one, there was a significant decrease in total energy intake from day 5 to the end of the experiment. This reduction was due to a significant decrease in fat intake during the same time, and later on, to a reduction of protein intake. Night intake was further analyzed in two parts of the 6-h cumulative intake (Fig. 4). In sedentary rats, over the first part ( 0 6 h), energy intake was significantly reduced in the first period of 4 days, due to fat intake reduction. On the other hand, during the last period of 4 days, fat intake was increased. Protein and carbohydrate intakes were not modified. In trained rats, total energy intake was significantly reduced compared to baseline. This reduction affected both protein intake from day 9 to the end and fat intake from day 5 to the end; carbohydrate intake was not modified. The comparison between the two groups of rats indicated that, in trained rats compared to sedentary rats, there was a significant

20 days ~

24 h food intake

GHO

FIG. 2. Protein, fat, carbohydrate, and 24-h cumulative intakes in the three groups of rats throughout the experimental period. ~rp < 0.05, significantly different from baseline.

loo% .

Control . . . . . . .

which accurately reflects changes in macronutrient intake and body weight gains. 24-Hour Food Intake 26'k . . . . . .

Average baseline food intake consisted of 25 % proteins, 55% fats, and 20% carbohydrates (Fig. 2). In control rats, total food ingestion was unchanged, F(20, 104) = 1.3, NS, throughout the experiment; fat intake increase did not reach statistical significance (F = 1.41, NS). In sedentary rats, the 24-h cumulative food intake was not affected by the 2-h fasting at the beginning of the nocturnal period, F(20, 209) = 0.86, NS. However, fat intake increased significantly on days 15, 16, and 17, F(20, 209) = 1.63, p < 0.05.

. . . . . . . . . . .

o%

Imme 1 2

3

4 6

belle 1 2 [~

a

4 6

Proteins

ti bliie I I---"I Fst4

,

4 6 ~

CHO

FIG. 3. Protein, fat, or carbohydrate intakes (expressed in percent of total intake) during baseline and throughout the experimental periods..kp < 0.05, significantly different from the same baseline intake.

370

L A R U E - A C H A G I O T I S , RIETH AND L O U I S - S Y I VESTR[-.

TABLE I PROTEIN, FAT, CARBOHYDRATE, AND TOTAL INTAKES IN SEDENTARY AND TRAINED RATS DURING NIGHT AND DAYTIME ~)Xc'E~ BASELINE AND THE FIVE SUCCESSIVE PERIODS OF TRAINING

Base (4 days)

P1 (Days 1-4)

P2 (Days 5 - 8 )

P3 (Days 9 - 1 2 )

P4 (Days 13-16)

P5 (Days 17-20)

ANOVA

I"t5. 5t~!

Night Intakes (kcal) Sedentary rats Protein Fat Carbo. Sum Trained rats Protein Fat Carbo. Sum

69.2 132.5 39.4 241.1

_+ 7.7 - 13.2 a _+ 7.9 _+ 11.3

60.3 110.7 40.5 211.5

- 5.1 _+ 15b _+ 6.9 _+ 11.1

58.3 _+ 6.9 127.5 ± 13.5 ~ 36.0_+ 6.7 221.8 _+ 17.9

59.3 138.2 41.2 239.7

+_ 5.6 _+ 11.5a _+ 7.9 _+ 13.5

62.7 147.0 39.4 249.1

_+ 6.2 _+ 15.6 a ± 7.8 _+ 12.7

64.6 147.7 34.4 246.0

± 7.1 _+ 16.6~ _+ 8.8 _+ 11

1.04 3.91 0.54 0.94

NS p < 0.01 NS NS

52.4 122.4 56.7 231.5

_+ 5.3 ___ 13.2 ~ ± 13.5 + 24.4 a

44.2 97.8 43.5 185.5

_ 6.2 ___ 12.3b -+ 10.6 4- 15.4 b

44.5 81.1 47.3 172.9

42.5 77.9 48.0 168.4

± 3.6 -+ 8.3 b _+ 7.9 _+ 12.6 b

45.8 73.8 50.8 170.4

_+ 4.1 -+ 10.7b ± 8 ± 12.2 b

43.7 84.4 55.6 183.7

_+ 4.4 ± 12.5b -+ 8.4 _+ 11.7 b

0.19 3.66 1.02 3.13

NS p < 0.01 NS p
-+ ± _ -+

4.7 6.1 b 8.5 10.4 b

Daytime Intakes (kcal) Sedentary rats Protein Fat Carbo. Sum Trained rats Protein Fat Carbo. Sum

3.0 +13.9_-_ 3.4___ 20.3 _+

1.1 3 1.4 4.7

4.3 __. 23.2 ___ 4.1_ 31.6 _+

1.1 4.7 1.5 5.4

3.6 -+ 0.9 25.4___ 6.9 4 . 9 _+ 2.6 33.9 _+ 7.5

4.2 _* 24.1 _+ 2.6_+ 31.0 _+

1.1 5.2 1 5.6

3.8 ___ 23.6___ 5.1 _+ 32.5 _+

1.5 5.8 3.4 8

2.7 22.4 3.3 28.4

± ± _+ +

1.6 4.7 1.8 5.5

0.41 1.4 0.72 1.24

NS NS NS NS

3.0 12.9 4.9 20.8

0.9 a 3.5 1.2~ 4.7 ~

6.3 22.6 6.2 35.1

1.7ab 5.4 1.1~c 6.8b

6.8 +-19.7 _ 8.2 _+ 34.6_+

9.4 19.8 8.6 37.8

1.6b 4.8 1.7~cd 5.4b

10.4 18.7 12.6 41.7

2.6b 4.6 1.8~d 5.4 b

11.5 15.9 14.3 41.7

+__ _+ + +

2.4 b 4.5 2.4t~ 4.8 b

4.06 1.61 6.32 5.37

p < 0.01 NS p < 0.01 p < 0.01

±4-+ ___

+-+ 4_+

1.7ab 4.3 2.6"~d 4.1 b

± ± _+ _+

± _ _+ _+

Means on the same line sharing a common superscript are not significantly different from one another (p < 0.05).

reduction in total energy intake from day 5 to the end of the experiment. Fat intake was reduced as soon as the second period of 4 days, whereas protein intake was significantly reduced in the third, fourth, and fifth periods. During the second part of the night ( 6 - 1 2 h), total energy intake of both groups was not significantly modified. However, in the trained compared to the sedentary group, a significant decrease in fat intake was observed in the second period of 4 days, and a significant increase in carbohydrate intake in the last 4 day period.

Daytime Food Intake During baseline, the daytime intake represented only 8.7 +1.5% of the overall 24h intake (Table 1). In sedentary rats, the total and the three nutrient intakes were not significantly modified by 2 h of food deprivation at the beginning of the night period. In contrast, in trained rats daytime energy intake was increased as soon as the first period of training. At the end of training, rats were ingesting twice their baseline daytime intake. Comparisons of daytime intakes between the two groups did not show any significant differences throughout the experiment. However, protein and carbohydrate intakes in trained rats were higher from day 9 to the end of the experiment. Daily energy intake, to better define the temporal nature of the nutrient intakes, was analyzed in terms of three time periods: 0 - 3 h, 3 - 9 h, 9 - 1 2 h (Fig. 5). In the first part o f the day, no significant differences were observed in the two groups. In sedentary rats, over the second part o f the day, total energy and the three nutrient intakes were

identical throughout the experiment. By contrast, in trained rats, due to increased protein and carbohydrate intakes, the 6-h food intake was significantly increased from the beginning to the end of training; fat ingestion was not significantly modified. The comparison between trained and sedentary rats indicated a significantly increased intake in the second, fourth, and fifth 4day periods in trained rats. This increase affected protein and carbohydrate ingestion only. Regarding the last part of the day, trained animals had a significantly reduced energy intake during the second period of training. In contrast, sedentary rats increased (but not significantly) the 3-h intake. The comparison between the two groups of rats indicated, in sedentary rats, a significantly increased fat intake. DISCUSSION

The present results show that, in rats, treadmill training reduces body weight gain and modifies total energy intake as well as nutrient selection. The 24-h energy intake was significantly reduced during training. Nighttime intake was mainly decreased during the first period of the night, affecting protein and fat intakes, whereas carbohydrate ingestion was not modified. Daytime energy intake was increased. It is remarkable that trained rats significantly enhanced protein and carbohydrate intakes during the middle part o f the daytime. In a similar fashion, during the last 3-h preceding fasting, sedentary rats increased fat intake. Reduced body weight gain in exercise-trained male rats is well documented (4,8). This reduction is mostly due to a decrease in body fat, with only a small or no change in lean body mass.

TRAINING AND SELF-SELECTION IN RATS

371

Nocturnal food intake (O-6h)

(12,19); moreover, repeated episodes of food restriction were associated with greater increases in fat intake (9). We observed here for the first time that a short but repeated fast leads to the same phenomenon. The reduction in food intake observed just after exercise has already been observed in male rats trained for a long time (3,18), but the modifications in intakes of the various nutriments were different. Andik et al. (3) observed increased fat and protein intakes and decreased carbohydrate intake. On the contrary, Pariskova and Stankova (18) reported increased carbohydrate and decreased fat intakes. Our results agree with the last study. However, our exercise duration was longer although intensity was nearly the same; the energy expenditure was therefore greater. In a recent study (15) following an acute exercise, we observed the same reduction of protein intake just after exercise, but thereafter

kcal 140 . . . . . 120---

-

200

S

T

base

kc~

1 -A

T

8

T

P.1

8

T

P.2

8

i_,t T

P.3

8

T

P.4

$

T

P.5

kcal

Nocturnal food intake (6-12h)

Diurnal food intake (O-3h)

8-

'

!

2-

~

base

P.1

kcal

P,2

~

P.3

L I

:

P.4

P.5

r~i, lrn=l fn~rl inf=kA /."4-Qh~l

3026 . . . .

-~lf

2624-

8 T bue

S T P.1

S r P.2

ST P.3 Pro~n=

81" P.4 [Z~ Fstl

~

8 r P.5

22-

GHO

18 16 14

FIG. 4. Nocturnal food intake during the first part of the night (upper graphs) or last part of the night (lower graphs). *p < 0.05, significantly different from sedentary group values during the same period. S = sedentary rats, T = trained rats. The arrow indicates the beginning of training or short 2-h fasting.

20

12

10 8 S 4 2

We have shown in this study that total dissectable adipose tissue was lower in trained rats. This result agrees with that of Stern et al. (22). The difference observed in body weight gain between sedentary and control rats could be due to the fact that sedentary rats are fasted every day for 2 h, at the beginning of the night, at a time where those rats usually ingested mainly fats (50% of their 2-h intake or 14% of their 24-h fat intake). They compensate the short fast by increasing their fat intake during the first part of the night. This increase could be dependent on the level of corticosterone, as previously observed with longer fasting periods [(24 h or 48 h (5)]. Control rats also increased their fat intake, but this increase did not reach statistical significance. It is well known that rats, as they get older, do tend to alter their macronutrient intake, increasing fat but decreasing carbohydrate intakes (12). These trends are observed here in the two groups of sedentary and control rats, but to a greater extend in the sedentary group. Moreover, the sedentary animals anticipate the short fast by increasing their fat intake at the end of the day, thus leading to a significant increase of fat intake during the 24 h. This excessive diurnal fat ingestion at a time when, in rats, lipolysis is predominant leads to enhanced lipid storage. It has already been observed that following longer periods of food deprivation (24 h or more), rats increased fat consumption

0

base

P.1

kcal

P.2

P.3

P.4

P.5

Diurnal food intake (9-12h)

=

:-== S

T

base

=S

iJlii!i aa~

%

=-* T

P,1

S

T

P.2 ~1

::m S

~

T

T

T

P.4 ~

ii~!i~i_.S.:

m

S

P.3 Proteins

H~iiii :

Fats

P.5 ~

CHO

FIG. 5. Protein, fat, and carbohydrate intakes over the three parts of the day during baseline and the five successiveperiods of training or fasting. *p < 0.05, significantlydifferent from the same intake in sedentaryrats. The arrow indicates the beginning of training or fasting.

372

L A R U E - A C H A G I O T I S , RIETH AND L O U I S - S Y 1 VESTRt!

rats maintained this decrease throughout the following 24 h and to a lesser extent in the following nycthemeral cycle. Here rats had a reduced protein intake during the night, but they increased it rapidly during the following daytime, as early as the first period of training. Thus, the 24-h protein ingestion was maintained. The nocturnal inhibition of protein intake could be due to an adrenergic action (10). It has been demonstrated (6,16) that catecholamines can decrease the plasma ratio of tryptophan relative to large neutral amino acids. Anderson et al. (2) reported that an inverse relationship exists between the amount of protein consumed by self-selecting rats and this ratio. This sequence of events could be responsible for the strong reduction in protein intake just after the exercise. Regarding carbohydrate ingestion, there was no significant reduction of intake during the night. On the contrary, a significantly increased carbohydrate intake was observed during the day: it appeared very soon into the second period of training, and this intake reached, at the the end of training, three times that of baseline or that of sedentary rats during the same period. Indeed, in the first days of training, rats increased their fat intake (75% in the first period, 50% in the second period) during the day. We observed that running was a little difficult for some of the rats. Thereafter, when they increased carbohydrate intake during the middle part of the day, they ran without any further problem right up to the end of the experiment.

Decreased fat intake in trained rats was significant over the 24 h but there was also a significant difference in the two parts of the nycthemeral cycle. This decrease appeared as soon as the first period of training. In the acute experiment (15), fat intake inhibition appeared later at night, and daytime fat intake was identical to baseline, as observed in the present experiment. So. we can consider that this early fat intake reduction (i.e., just after exercise) in the present study was an effect of training per se. The present results appear to show an anticipatory phenomenon that has an effect earlier in the day in trained rats and later in the day in sedentary rats. The fast-induced anticipation in sedentary rats has already been observed in rats eating a complete diet (17). In the present experiment, sedentary rats were able to choose the nutrient that they preferred. They choose the fat diet that was more energetic. On the contrary, in trained rats the anticipation appeared earlier and concerned mainly carbohydrates. It has been observed in humans (7,11) that the consumption of a carbohydrate supply 4 h before exercise greatly improves exercise endurance. It seems that, in the present study, rats learn to anticipate this intake and also are able to improve performance. Confirmation of this hypothesis is under investigation. ACKNOWLEDGEMENTS The authors are grateful to Marie Claude Laury for preparing the white adipose tissue. We thank Simon Thornton for his kind contribution to the English correction.

REFERENCES 1. Ahrens, R. A.; Bishop, C. L.; Berdanier, C. D. Effect of age and dietary carbohydrate source on the responses of rats to forced exercise. J. Nutr. 102:241-248; 1972. 2. Anderson, G. H.; Glanville, N. T.; Li, E. T. S. Amino acids and the regulation of quantitative and qualitative aspects of food intake. In: Blackburn, G. L.; Grant, J. P.; Young, V. R., eds. Amino acids: Metabolism and medical applications. Boston: John Wright; 1983:225-238. 3. Andik, I.; Bank, J.; Moring, I.; Szegvari, G. Y. The effects of exercise on the intake and selection of food in the rat. Acta Physiol. Hung. 5:457-461; 1954. 4. Applegate, A.; Upton, D. U.; Stern, J. S. Food intake, body composition and blood lipids following treadmill exercise in male and female rats. Physiol. Behav. 28:917-920; 1982. 5. Bligh, M. E.; DeStefano, M. B.; Kramlik, S. K.; Douglass, L. W.; Dubuc, P.; Castonguay, T. W. Adrenal modulation of the enhanced fat intake subsequent to fasting. Physiol. Behav. 48:373-381; 1990. 6. Chaouloff, F.; Elgozi, J. L.; Guezermec, Y.; Laude, D. Effects of conditioned running on plasma, liver and brain tryptophan and on 5-hydroxytryptamine metabolism in the rat. Br. J. Pharmacol. 86:33-41; 1985. 7. Costill, D. L. Carbohydrate nutrition before, during and after exercise. Fed. Proc. 44:364-368; 1985. 8. Crews, E. L.; Fuge, K. W.; Oscai, L. B.; HoUoszy, J. O.; Shark, R. E. Weight, food intake and body composition: Effect of exercise and protein deficiency. Am. J. Physiol. 216:359-363; 1969. 9. Gerardo-Gettens, T.; Miller, G. D.; Horwitz, B. A.; et al. Exercise decreases fat selection in female rats during weight cycling. Am. J. Physiol. 260:R518-R524; 1991. 10. Guilland, J. C.; Moreau, D.; Genet, J. M.; Klepping, J. Role of catecholamines in regulation by feeding of energy balance following chronic exercise in rats. Physiol. Behav. 42:365-369; 1988. 11. Hargreaves, M.; Costill, D. L.; Coggan, A.; Fink, W. J.; Nishibbata, I. Effect of carbohydrate feedings on muscle glycogen utilization and exercise performance. Med. Sci. Sports Exerc. 16:219-222; 1984.

12. Kanarek, R. B. Factors influencing fat consumption in diet selection studies. In: Mela, D. J., ed. Dietary fats: Determinants of preference selection and consumption. New York: Elseviers Applied Science Publishers; 1991:139-165. 13. Katch, V. L.; Martin, R.; Martin, J. Effects of exercise intensity on food consumption in the male rat. Am. J. Clin. Nutr. 32:1401 - 1407; 1979. 14. Larue-Achagiotis, C.; Martin, C.; Verger, P.; Louis-Sylvestre, J. Dietary self-selection or standard complete diet: Body weight gain and meal pattern in rats. Physiol. Behav. 51:995-999; 1992. 15. Larue-Achagiotis, C.; Martin, C.; Verger, C.; Chabert, M.; LouisSylvestre, J. Effects of acute treadmill exercise and delayed access to food on food selection in rats. Physiol. Behav. 53:403-408; 1993. 16. Leibowitz, S. F. Neurotransmetteurs centraux et contrfle des apprtits sprcifiques pour les macronutrients. Ann. Endocrinol. (Paris) 49:133-140; 1988. 17. Le Magnen, J. Etude d'un phrnom~ne d'apprtit provisionnel. C. R. Acad. Sci. 249:2400-2402; 1954. 18. Parizkova, J.; Stankova, L. Influence of physical activity on a treadmill on the metabolism of adipose tissue in rats. Br. J. Nutr. 18:325332; 1964. 19. Schutz, H. G.; Pilgrin, F. J. Changes in the self-selection pattern for purified dietary components by rats after starvation. J. Comp. Physiol. Psychol. 47:444-449; 1954. 20. Stevenson, J. A. F.; Box, B. M.; Feleki, V.; Beaton, J. R. Bouts of exercise and food intake in the rat. Am. J. Physiol. 21:118-122; 1966. 21. Stern, J. S. Is obesity a disease of inactivity? In: Stunkar, A. J.; Stellar, E., eds. Eating and its disorders. New York: Raven Press; 1984:131-139. 22. Stern, J. S.; Titchenal, C. A.; Johnson, P. R. Obesity: Does exercise make a difference? In: Berry, E. M.; et al., eds. Recents advances in obesity research, vol. 52. London: John Libbey; 1987:352-364. 23. Titchenal, C. A. Exercise and food intake. What is the relationship? Sports Med. 6:135-145; 1988.