Effect of the physical state of a food on subsequent intake in human subjects

~/~petire. 1991. Ici. 17-24

Effect of the Physical State of a Food on Subsequent in Human Subjects

Intake

A. TOURNIER and J. LOUIS-SYLVESTRE Laboratoire de Neurobiologie

de la Nutrition,

Universit6 Pierre et Marie Curie, Paris

state of food on further intake was studied. Human subjects were served two experimental foods at lunchtime 1 week apart. In a first experiment, the two-food meals had the same caloric content, weight. volume. composition and temperature but differed only in their physical form (solid or liquid). As compared to the solid meal, the 24-h total caloric intake following the liquid food was higher. This difference could be attributed to cognitive cues from the form of food or to the lack of masticatory movements while ingesting the liquid meal. The last hypothesis was tested in a second experiment, “In which the test-meals appeared absolutely identical to the senses and were composed of a liquid and a solid item; however, the major part of calories was either in the liquid or in the solid part. When most of the calories had to be drunk, the total subsequent caloric intake was higher than when the most of the calories had to bc eaten. It seems that calories ingested in a liquid form are not well taken in account and could induce a subsequent overconsumption, at least until satiety was conditioned to the fluid. The effect of the physical

The effect of the physical state of a food on further food intake has been barely studied. Pliner (1973) showed that, in normal weight subjects, the same caloric preload taken as a liquid or as a solid food was identically compensated during the meal ingested 1 h later. The foods used differed in composition, volume and mass so that it was difficult to draw conclusions about the regulatory processes. Kissileff et al. ( 1980) used yogurt with fruit lumps or the same yogurt homogenised and showed that subjects in the same conditions ate the same quantities of the two foods. Another observation made with very different foods (tomato soup and a combination of crackers, cheese and apple juice) led the author to the conclusion that liquid food may be more satiating (Kissileff et al, 1984). Some experiments on animals showed that rats on diets made liquid by addition of water ingested more calories and gained more weight than other rats on the same diet in solid form (Keane et al., 1963; Adkins et al.. 1967; Ramirez, 1987). These results were obtained with various diets, rich in sugars or starch (Sclafani & Xenakis. 1984: Ramirez, 1987a, b), high fat (Ramirez 19876) or others (Rogers & Harper, 196.5; Adkins et al., 1967). In all these studies the difference in texture involved a difference in caloric density with consequences on digestive and absorptive mechanisms. There were also differences in hedonic value. There could have been an interaction with body fluid regulation too.

Address reprint requests to: A. Tournier. Laboratorie de Neurobiologie Pierre et Marie Curie, 4 place Jussieu, 75252 Paris C‘edex 05. France. 11195-6663’91/010017+08

$03,00:0

de la Nutrition.

( 1991 Academic

I ~niverslt6 Press l.lmlred

18

A. TOURNIER

AND J. LOUIS-SYLVESTRE

Two studies reported here were carried out to test the long-term effect of the ingestion of food identical in all respects except texture. In the first study, 24-h intake was measured following two identical test-meals (same composition. weight and volume) differing only in their physical state (solid or liquid). In a second study, the test-meals were absolutely identical and composed of a liquid and a solid item; however, most calories were either in the liquid or in the solid part.

GENERAL

METHODS

Subjects

Subjects were six women and seven men in the first experiment, and seven women and six men in the second one. They were 22-25 years old with normal body weight and without particular knowledge about nutrition. Ratings

Foods were ingested and their palatability then rated on IOO-mm lines whose extremities were anchored on “extremely unpleasant” and “extremely pleasant”. Ratings were converted into numbers from 0 to 100. Hunger was also rated on TOO-mmlines. The words used as anchors at either end were “null” and “extreme”. Food htake Recordings

Subjects were provided with booklets and food-weighing scales. They had to record time at which the meal, snack or drink was consumed, a detailed description of the food or drink consumed and the amount ingested expressed as a weight (from the provided scale). Procedure

In both experiments, two tests (A and B) were performed 1 week apart. The subjects had a free-choice breakfast which had to be identical on test days. At 1200 hrs, subjects were given one of the experimental meals. They were told to ingest all of the offered food. Subjects had water ad Zibitum. Water consumption was measured only in the first experiment. Subjects had to rate their hunger before lunch, immediately after and every 30 min until 1600 hrs. They also rated the pleasantness of the food immediately after consumption. Subjects were told not to eat anything before 1600 hrs and then to eat ad Iibitum and to record all intakes up to the next breakfast. One half of subject performed test A on the first test-day and B 1 week later. The order of tests was reversed for the others. Data Analysis

Energy intakes were computed from the diary records with reference to standard food composition tables and manufacturers’ data (Renaud et al., 1976; AstierDumas, 1983; Klepping et al., 1988).

PHYSICAL

10

STATE OF FOOD

Data were compared using two- or three-way analysis of variance with repeated measurements according to standard formula (var 111, Rouanet & Lepine, 19771.

EXPERIMENT I

Method rest A

The liquid food was composed of cooked white beans, ham, green peas and maltodextrins (Glucidex 12, Roquette France) mashed together; it had to be drunk from a bowl. A glass containing lOOmI of warm tomato consommk with added gelatine was also presented. Test B The experimental food was a solid composed of the same ingredients mashed separately. Maltodextrins were added to the green pea purke. Gelatine was added to the three purCes which were set in layers so as to obtain an attractive appearance. This terrine had to be eaten with knife and fork (see Table 1 for compositions). A glass containing lOOm1 of warm tomato consommi without added gelatine was also presented.

RESULTS

Ratings of hunger before ingesting test foods did not differ between the two test-days [F(1,12) =2.1 (Figure l)]. There was no difference either in the hedonic evaluation of the two foods [F&96) = 1.O].Over 4 h following ingestion there was no difference in time course of hunger ratings after the liquid food (test A) or the solid food (test B). Water intakes during the test-meals were not significantly different (Table 2). Intakes at the afternoon snack, dinner and breakfast were slightly but not significantly increased after the liquid food.

TABLE 1

Composition and caloric content of experimental foods (Experiment I

TEST A soup Tomato Juice + Gelatine Total TEST B Soup + Gelatine Tomato Juice Total

Weight (g)

Caloric content (kcal)

Protein (kcal)

Fat (kcal)

430.0 100.0 8.8 538.8

621.5 29.0 28.5 679.0

89.0 0.5 28.5 128.0

188.0 3.0 0.0 191.0

0.0 360.0

430-o 8.8 100.0 538.8

621.5 28.5 29.0 679.0

89.0 28.5 0.5 128.0

188.0 0.0 3.0 191.0

334.5 0~0 75.5 360.0

Carbohydrate (kcal)

334.5 zf;5

20

AND J. LOUIS-SYLVESTRE

A. TOURNIER

Total intake after the liquid food was (mean +SD) 118.4+ 8.7% of the total intake after the solid food. This result just falls short of significance [F(1,12)=4.46, p ~0.061. Only two of the 13 subjects ate more after the solid experimental lunch, and each had an exceptional dinner that evening arising from social commitments.

DISCUSSION

The two meals had the same weight, volume, temperature, caloric content, water content and composition; they were equally appreciated. In one case, the food was drunk from a bowl; in the other case it was eaten with knife and fork. What factors in these conditions could explain the possibly higher consumption after the liquid food than after the solid one? The liquids that adult humans ingest are usually low in caloric content; for example, a normal soup has a caloric density of 0.5 kcal/g and a sweet chocolate beverage has 0.8 kcal/g. In the present study, the experimental food, whether liquid or solid, was 1.45 kcal/g so it could be hypothesized that an inadequate metering of ingested energy may have occurred because of lack of experience with such a calorific fluid. The role of cognitive cues has been underlined by Wooley et al. (1972, 1975) and further evidence was provided by Booth et al. (1982). Another possibility, though, is that the lack of masticatory movements when ingesting the liquid food could contribute to overconsumption. Therefore, we conducted another experiment to determine if this factor could lead to greater intake.

EXPERIMENT

2

Method

Each of the two test-meals was composed of a solid food that was either normocaloric (507 kcal) or hypocaloric (56 kcal) served with a liquid food, respectively either hypocaloric (56 kcal) or normocaloric (507 kcal). The liquid and solid foods had the same weight. The normocaloric food was a mixture containing 1.45 kcal/g. It was made of canned concentrated vegetable soup, powered oxtail, plain

TABLE 2

Pleasantness ratings and water and caloric intakes following tests A and B (Experiment 1; n=13/

TEST B

TEST A SOUP (liquid) SEM Means Hedonic value Water (ml) Snack (kcal) Dinner (kcal) Breakfast (kcal) Total (kcal) “No difference

45.2 254 334 781 417 1548 was statistically

significant.

4.5 43 102 72 88 215

TERRINE Means 42.3 336 254 769 321 1344

(solid) SEM 5.3 58 66 II3 74 165

F(l,l2)” of difference 0.53 l-58 I.73 0.02 2.48 5.56

‘I

PHYSICAL STATE OF FOOD

_ 70

-

60 -

P i

5Ok T

$40c ; 3O20 -

IO-

/

I

I

Before meal

I

After meal

I

1300

I

1330

/

1400

I

1430

I500

1

!

1

1530

1600

Time

FIGURE 1. Mean hunger ratings. Experiment 1 (0) are SEM.

powered milk, maltodextrins juice (0.16 kcal/g).

I

test A: liquid; (0) test B: solid. Bars

and water (Table 3). The hypocaloric

food was a tomato

Test A The normocaloric food was served in a liquid form (without gelatine) and was drunk from a bowl, the hypocaloric food was served in a solid form (with 12 g gelatinc added) and was eaten with knife and fork.

TABLE 3 Composition

and caloric content of experimental

foods

(Experiment

2 f

Fat (kcal)

(kcal)

Weight (g)

Caloric content (kcal)

346

507

78

144

7x5

346 12 704

56 40 603

0 40 118

0 0 144

%l (I 141

346 12

507 40

78 40

144 0

‘X5 0

346 704

56 603

0 II8

0 144

56 341

Protein (kcal)

Carbohydrate

TEST A Normocaloric liquid

mixture Hypocaloric solid

tomato juice + gelatine Total

TEST B Normocaloric solid

mixture + gelatine HJlpocaloric liquid

tomato juice Total

22

A. TOURNIER AND J. LOUIS-SYLVESTRE -.-_70 -

60 -

7

I

I

Before meal

FIGURE2.

I

I

I

After meal

1300

I

1330

I

1400

1430

I

I

1500

1530

I

1600

Ttme

Mean hunger ratings in Experiment 2: (0) test A: most of calories liquid; (0) test B: most of calories solid. Bars are SEM.

Test B The hypocaloric food was served in a liquid form (without gelatine) normocaloric one in a solid form (with 12g gelatine added; Table 3).

and the

RESULTS On the two test-days hunger ratings before ingesting test-foods were not different (F(8,96)= 1.6 (Figure 2)]. The two types of foods in both forms were equally was observed in hedonic ratings (Table 4) appreciated, i.e. no difference

TABLE 4

Pleasantness ratings and caloric intakes following tests A and B (Experiment (n = 13)

2),

TEST B

TEST A

Difference Mean

SEM

Mean

SEM

F(1,12)

P

5.2

0.6

ns

3.1

0.1

ns

5.1 13.5 1.7 8.4

< 0.05 CO.01

Hedonic values (mm)

Normocaloric

Food

Hypocaloric Food Intakes (kcal) Snack Dinner Breakfast Total

Liquid 40.6 Solid 32.0

4.4 3.1

Solid 43.1 Liquid 33.0

331

94

506

1,135 322 1,788

135 67 191

725 271 1,504

64 104 49 152

ns -Co.01

PHYSICAL

STATE OF FOOD

23

[F(1,12) = 1.81. There was no difference between tests A and B in time course of hunger rating over 4 h following ingestion. Following test A (liquid normocaloric and solid hypocaloric foods) food intake over 24 h was significantly higher than after test B (Table 4) by (280 kcal) or about 20% of the subsequent intake. Compared to test B, the snack following test A was significantly smaller but the dinner was significantly larger, and there was no difference in the breakfast intakes as recorded.

DISCUSSION

In this experiment, the two meals were exactly the same in weight, temperature, caloric content, composition and quantities eaten and drunk. Therefore, stomach contents were presumably identical. Moreover, in both tests, subjects had to perform the same number of chewing and swallowing movements. It is therefore a puzzling fact that, following consumption with most of the calories as liquid, immediate snack intake was smaller (by about 175 kcal) whereas dinner was considerably large (by about 400 kcal). No clear explanation can be provided. All that can be noted is that in one test most of the calories were in the food to be eaten and chewed, and in the other test most of the calories were in the food to be drunk.

GENERALDISCUSSION

The results of both experiments indicate that overconsumption can follow the ingestion of a liquid food as compared to an equicaloric solid one. Nevertheless, in the second experiment chewings and swallowings were identical in the two tests and so the overconsumption cannot be attributed to any effect of mechanical activity. It is well known that solid foods empty more slowly from the stomach than liquid foods. Thus it could be suggested that liquid calories are absorbed more rapidly and so hunger returns more quickly and food intake is greater. However the present results have shown that the patterns of hunger rating development after the two tests do not differ. Moreover, very little time is required for the solid food to be liquefied in the stomach, according to a estimate in an in uirro test: the introduction of mouthfulsized amounts at a normal rate of ingestion to a container, maintained at 37°C in a water-bath, showed that fewer than 9 min were needed to reduce 438.8 g of terrine to the liquid state in this schematic stomach. Therefore it is reasonable to expect that gastric emptying and absorption occur under the same conditions after the two tests. This study underlines the role of cognitive interpretation of cues from the form of the diet: it seems that calories ingested in a liquid form are not well taken in account and so induce subsequent overconsumption. It has been well established that a conditioning mechanism can allow the adjustment of the ingestion of a food to the caloric content of the food (Booth et 01.. 1976, 1982). As a result of nutritional reinforcement, the sensory characteristics of the food are associated to post-ingestive effects within a compound conditioned stimulus (Booth & Toase, 1983). The subjects of these experiments were presumably accustomed to low-caloric drinks, and so had no learned satiety response. After some meals of the liquid caloric food, however, a learned adjustment would occur; thus. the effects seen in these experiments might not persist after repeated tests.

A. TOURNIER AND J.LOUIS-SYLVESTRE

24 The

present

data

suggest

that,

caloric adjustment on a short-term interpreted cautiously.

at least in human subjects, results concerning basis with liquid diets or liquid preloads have to be

REFERENCES Adkins, J. S.. Wertz, J. M., Boffman, R. H. & Hove, E. L. (1967) Influence of addition of water or “nonessential” nitrogen on growth of rats fed low levels of essential L-amino acids. Proceedings afSociety,for Experimental Biology and Medicine, 126, 500-504. Astier-Dumas, M. (1983) Valeur nutritionelle de quelques produits prets a Ctre consommes. Paris: Imprimerie Graphic Rouvet. Booth, D. A., Lee, M. & McAleavey, C. (1976) Acquired sensory control of satiation in man. British Journal af Psychology, 67, 137- 1476. Booth, D. A., Mather, P. & Fuller, J. (1982) Starch content of ordinary foods associatively conditions human appetite and satiation, indexed by intake and eating pleasantness of starch-paired flavours. Appetite, 3, 163- 184. Booth, D. A. & Tease, A. M. (1983) Conditioning of hunger/satiety signals as well as flavour cues in dieters. Appetite, 4, 2355236. Keane, K. W., Smutko, C. J., Krieger, C. H. & Denton, A. E. (1963) Water addition to purified diets. II Effects of water and carbohydrate on protein efficiency ration in rats. Journal af Nutrition, RI, 87-90. Kissileff, H. R., Klingsberg. F. & Yan Itallie. T. (1980) Universal eating monitor for continuous recording of solid or liquid consumption in man. American Journal a/’ Physiology. 238. R 1422. Kissileff, H. R., Gruss. L. P., Thornton, J. & Jordan, H. A., (1984) The satiating efficiency of foods. Physiology and Behavior, 32, 3 199332. Klepping, J., Guilland, J. C., Fuchs, F., Marcer, I. and Houard-Malval, M. (1988) Recueil de don&es sur la composition des aliments. Produits Roche, Neuilly-sur-Seine. Pliner. P. L. (I 973) Effect of liquid and solid preloads on eating behavior of obese and normal persons. Physiology and Behavior, II, 285-290. Ramirez, I. (1987a) When does sucrose increase appetite and adiposity? Appetite, Y, 1- 19. Ramirez, I. (I 9876) Practical liquid diets for rats: effects on growth. Physiology and Behavior. 39, 527-530. Renaud, S., Godsey, R.. Ortchanian, E., and Baudier, R (1976) Table de composition de.r aliments, Asnieres: Imprimerie Astra-Calve. Rogers, Q. R. & Harper, A. E. (1965) Amino acid diets and maximal growth in the rat. Journal of Nutrition, 87, 2677273. Rouanet, H. & Lepine, D. (1977) Introduction a l’analyse des comparaisons pour le traitement des donnees experimentales. lnformatique et Sciences humaines, 33, 34. Sclafani, A. & Xenakis, S. (1984) Influence of diet form on the hyperphagia-promoting effect of polysaccharide in rats. Lifr Sciences, 34, 1253-1259. Wardle, J. (1987) Hunger and satiety: a multidimensional assessment of responses to caloric loads. Physiology and Behavior. 40, 577- 582. Wooley, 0. W., Wooley. S. C. & Dunham. R. B. (1972) Can calories be perceived and do they affect hunger in obese and non-obese humans? Journal ofComparutive and P/ry.riological Psychology. 80, 25Om258. Wooley, 0. W., Wooley. S. C. & Woods, W. A. (1975) Effects of calories on appetite for palatable food in obese and non-obese humans. Journal af Comparative and Physiological Psychology,

8Y, 619-625.

Received

31 July 1989, revision 26 March

1990