GASTROENTEROLOGY 2000;119:943–948

The Effects of Intestinal Infusion of Long-Chain Fatty Acids on

Food Intake in Humans

STEPHEN J. FRENCH,* CATHRYN A. CONLON,* SANDRA T. MUTUMA,* MYRTHA ARNOLD,‡

NICK W. READ,* GERT MEIJER,§ and JOHN FRANCIS*
*Centre for Human Nutrition, University of Sheffield, Northern General Hospital, Sheffield, England; and ‡Institute for Animal Sciences,
Physiology and Animal Husbandry, Zurich, Switzerland; and §Lipton, Englewood Cliffs, New Jersey

Background & Aims: Dietary fat intake is related to the C18). McCaffery et al.10 showed that increasing the chain
degree of obesity, but the specific mechanisms by which length of medium-chain fatty acids increased the satiat-
fats regulate food intake in humans are unclear. We ing potency in sham feeding rats. In humans, intraduo-
compared food intake suppression, plasma triglyceride denal infusion of long-chain fatty acids (sodium oleate)
appearance, and cholecystokinin (CCK) response after was a potent inhibitor of food intake, whereas infusion of
intestinal infusion of oils enriched with C18 fatty acids of
medium-chain fatty acids (sodium caprylate) had no
increasing unsaturation. Methods: Food intake and ap-
effect on intake.11
petite changes after upper intestinal infusion of 0.9%
saline, 20% Intralipid, and 20% emulsions of oils en-
The effect of the degree of unsaturation of the fatty
riched with stearic, oleic, and linoleic acids were tested acid chains on food intake has been less well studied.
in 10 healthy male volunteers. Plasma triglyceride ap- Lewis et al.12 found that upper-intestinal infusion of
pearance and CCK release were tested separately in 7 solutions of the sodium salt of long-chain fatty acids of
additional volunteers. Results: Intralipid and linoleic acid the same chain length (C18) led to markedly different
infusions significantly reduced food intake compared reductions in food intake in rats, linoleic acid being more
with saline infusion (P < 0.05). No changes were ob- than twice as potent than oleic acid or linolenic acid.
served in appetite ratings. There were no differences in The mechanisms by which the products of fat diges-
plasma triglyceride response over the initial 75 minutes tion elicit satiating effects are unclear. Hopman et al.13
of intestinal infusion. Plasma CCK concentration in- showed that the peptide hormone cholecystokinin
creased after all lipid infusions (P < 0.001), Intralipid (CCK), a proposed satiety signal,14 is potently released by
infusion produced the highest increase in plasma CCK
long-chain, but not by medium-chain, triglycerides in
(P < 0.05), and CCK response was similar between the
humans. However, species differences are evident because
3 enriched oil emulsions. Conclusions: These results
indicate marked differences in the ability of C18 fatty medium-chain triglycerides are the most potent stimu-
acids to reduce food intake that appear not to be related lators of CCK release in rats.15 Beardshall et al.16 com-
to rate of absorption but may partially be explained by pared the effects of fats containing high proportions of
CCK release. the saturated, monounsaturated, and diunsaturated C18
fatty acids on CCK release in normal human volunteers.
They found that the diunsaturated oil was the most
here is a close relationship between the intake of
T dietary fat and both total energy intake and the
degree of overweight and obesity.1–3 Understanding the
potent releaser of CCK, followed by the monounsatu-
rated oil. Only the saturated fat produced a nonsignifi-
cant increase in plasma CCK levels. Similar results were
mechanisms by which specific macronutrients influence found in the same study by using free fatty acids on both
food intake is therefore important to enable formulation CCK release and gallbladder contraction.
of dietary strategies to combat excessive intake of fat and, The effects of long-chain fatty acids, differing only in
hence, total energy. their degree of unsaturation, on the control of appetite
Several human and animal studies have shown that the mechanisms in humans is not known. The aim of this
presence of lipid in the small intestine is a potent inhib- study was to compare the effectiveness of 3 oils modified
itor of food intake.4 – 8 Relatively few studies, however,
have investigated the responses of specific fatty acids on
food intake and, to date, have focused predominantly on Abbreviations used in this paper: Apo A-IV, apolipoprotein A-IV; CCK,
effects of chain length. Maggio and Koopmans9 found no cholecystokinin.
© 2000 by the American Gastroenterological Association
difference in the satiating potency of gastric infusions of 0016-5085/00/$10.00
fatty acids of differing chain length in rats (range, C6 – doi:10.1053/gast.2000.18139
944 FRENCH ET AL. GASTROENTEROLOGY Vol. 119, No. 4

to contain enriched proportions of stearic, oleic, and Table 1. Percentage Fatty Acid Composition of Infused Oils
linoleic acids by using a study design that we have Stearic/
previously shown to be sensitive to manipulations of fat Oleic (% of Linoleic oleic (% of Intralipida
Fatty acid total) (% of total) total) (% of total)
dose.4 These oils were compared with a control infusion of
0.9% saline and a positive control of 20% Intralipid emul- C14:0 0.1 0.1 0.1
C16:0 4.1 7.1 4.4 9
sion (Pharmacia Laboratories Ltd., Milton Keynes, UK). C17:0 0 0 0.1
C18:0 3.7 2.6 34.1 7
Materials and Methods C20:0 0.3 0.3 1.1
C22:0 0.9 0.3 0.4
Patients C24:0 0.4 0.2 0.2
C18:1 trans 0.1 0 0.1
Studies were performed on healthy men aged 18 –35
C18:1 cis 74.7 14.2 44.2 29
years, with a body mass index of 20 –25. All patients were C18:2 trans
screened before entry to the study to exclude smokers or any isomers 0.1 1.1 0.2
subjects who had reported any serious medical conditions or C18:2 9c, 12c 14.6 72.7 14.3 54
recent illness. Dietary restraint was assessed by using the Other isomers ⱕ0.1 ⱕ0.1 ⱕ0.1
Total 99.4 99.3 99.5
Three Factor Eating Questionnaire17; this questionnaire
measures 3 factors associated with abnormal eating behav- NOTE. Other lipid components contained in oil emulsions: phospho-
ior: disinhibition (loss of control of eating and possible lipids (⬍0.01%), vitamin E (0.1%– 0.4%), and plant sterols (0.2%–
binging behavior), restraint (conscious control over food 0.4%).
aEstimated composition based on figures for pure soybean oil.
intake/dieting behavior), and abnormal hunger. Any patient
scoring higher than 10 on the eating restraint section of the
questionnaire was excluded from the study (mean value,
3.14; range, 1– 8). Patients were also asked to taste the acids (Table 1). Ninety minutes after the beginning of the
yoghurt drink test meal, and any patients who indicated infusion, patients were given a test meal of 1.5 L strawberry
that they did not like the drink were excluded from par- drinking yoghurt (composition: energy, 77 kcal/100 g; pro-
ticipating. All patients gave written informed consent to tein, 2.9 g/100 g; carbohydrate, 12.4 g/100 g; and fat, 1.8
participate in the study, and the study protocol was ap- g/100 g; Yoplait Dairy; Crest Ltd., Telford, England), pre-
proved by the Local Research Ethics Committee of the weighed in a large, covered container, and were instructed to
Northern General Hospital, Sheffield. drink until comfortably full. Patients were told that they could
request more if desired. The amount of yoghurt consumed was
Protocol assessed by weighing the food container again after the end of
Study 1. Ten patients participated in 5 studies, the experiment.
identical in design, differing only in the nature of the During the study, patients were asked to rate a range of
infusion. Studies were separated by at least 3 days and were subjective feelings including hunger, fullness, prospective con-
given in a randomized order with the patient blind to the sumption, and general well-being (e.g., nausea, dizziness) on
nature of the infusion. Patients were instructed to fast from 100-mm visual analogue scales. Questionnaires were com-
10 PM on the evening before studies and to arrive at the pleted before the beginning of the infusion, at 15-minute
center at 8 AM on each test day. On arrival, patients were intervals throughout the study, and before and after the pre-
intubated with a polyvinyl nasogastric tube (Merck Ltd., sentation of the test meal.
Alton, England) through the nose and into the stomach. Study 2. Seven additional patients participated in a
The tube was then allowed to pass through the pylorus into study of identical design for infusion conditions, as described
the duodenum under the force of gastric contractions. The earlier. After insertion of the nasogastric tube, a peripheral
position of the tube was monitored, and its position in the forearm vein was cannulated (Venflon 2; 18G, Ohmeda, Swe-
small intestine confirmed, by the measurement of the pat- den) to allow repeated blood sampling. Once the tube was in
tern of intraluminal pressure activity. The detailed methods position, a baseline blood sample (10 mL) was taken, the
of this procedure have been described previously.4,18 Once infusion was begun, and samples were taken at 15-minute
in the duodenum, the tube was allowed to progress until the intervals for 75 minutes. These were subsequently analyzed for
infusion port was 75 cm from the nasal opening, at which plasma CCK and triglyceride concentrations.
point the tube was fixed to the patient’s face with tape to
prevent further movement. If the tube was not positioned Preparation of Triglyceride Emulsions
into the duodenum by 10 AM on any study day, the test was Oils modified to contain varying proportions of stearic,
abandoned and rescheduled. oleic, and linoleic acids (see Table 1 for composition of oils)
Duodenal infusions were administered at a rate of 1 mL/min were prepared by Unilever Research (Vlaardingen, The Neth-
over a 100-minute period (calories infused, 180 kcal; total fat erlands). Oils were blended from the following: high-oleic
infused, 20 g). The 5 infusions consisted of 0.9% saline sunflower oil (oleic, 92.6%; linoleic, 1.4%; stearic/oleic,
(control), 20% Intralipid (positive control), and three 20% 30.2%); safflower oil (oleic, 7.4%; linoleic, 98.6%; stearic/
emulsions containing oils rich in oleic, linoleic, and stearic oleic, 12.8%); and fractionated shea nut oil (oleic, 0%; linoleic,
October 2000 FATTY ACIDS AND APPETITE 945

0%; stearic/oleic, 57%). These were emulsified by using 5%

food grade Tween 80 (Quest International, Zwijndrecht, The
Netherlands) at a concentration of 20% (vol/vol; 2 kcal/mL)
with sterile water in a sonicator bath (Kerry Ultrasonics Ltd.,
Hitchin, UK) for 45 minutes. Solutions were kept at approx-
imately 37°C before infusion in a heated water bath (Grant
Instruments, Cambridge, England) and during infusion by
using a heat pad.

Processing of Blood Samples

Blood samples (10 mL after discarding 2 mL) were
drawn into a Serum Separator tube (5 mL; Vacutainer SST)
and a Li-heparin tube (5 mL; Vacutainer) containing 5000
U/tube aprotinin (Trasylol; Bayer PLC, Berkshire, UK). All
tubes were precooled. SST tubes were rotated and left to
stand at 4°C for 40 minutes according to the manufacturer’s
instructions. Tubes were centrifuged at 3000 rpm for 10
minutes, and the plasma and serum were removed and
stored at ⫺70°C for later analysis. Serum samples were
analyzed for triglyceride concentration spectrophotometri-
cally by using a standard enzymatic kit (Triglyceride 334- Figure 1. Test meal intake after upper intestinal infusion of saline
UV; Sigma-Aldrich Ltd., Dorset, UK) in a method auto- and lipid infusions. Values are expressed as mean ⫾ SEM. *P ⬍ 0.05
mated for use on the Cobas bioautoanalyzer (Roche Products compared with saline infusion.
Ltd., Hemel Hempstead, UK). Plasma samples were ana-
lyzed for CCK concentration by using a commercial radio- Keuls post hoc tests revealed that food intake after
immunoassay (Euro-Diagnostica AB, Malmö, Sweden). The Intralipid and linoleic acid infusion was significantly
detection limit of the assay was 0.3 pmol/L, the intra-assay
lower than saline condition (Figure 1).
variation was 7%, and the interassay variation was 7.8%.
Visual analogue scale ratings of appetite and
Statistical Analysis well-being. Two-way repeated ANOVA showed no sig-
nificant difference between conditions throughout the
Data were analyzed by using SPSS for Windows v.6.0
period of the infusion for hunger (F[4,36], 1.24; P ⬎ 0.1;
(SPSS Inc., Chicago, IL). Results are expressed as mean. In all
Figure 2), fullness (F[4,36], 0.62; P ⬎ 0.1; Figure 2),
tests, probability (P) values of ⬍0.05 were considered statis-
tically significant. prospective food intake (F[4,36], 1.73; P ⬎ 0.1), or
Test meal intake was analyzed by using 1-way repeated nausea (F[4,36], 2.23; P ⬎ 0.1).
measures of analysis of variance (ANOVA), with infusion Study 2
condition as the main factor. Student–Newman-Keuls post
hoc tests were used to determine which conditions differed Plasma triglyceride. There were no differences in
significantly. Visual analogue ratings, plasma triglyceride, and plasma triglyceride levels over the 75 minutes of blood
CCK concentrations were analyzed by repeated measures with sampling in any of the conditions (F[4,24], 0.96; P ⫽
2-way ANOVA with time and infusion condition as within- 0.45; Figure 3).
patient factors. When significant overall effects were found Plasma CCK. The infusion condition signifi-
post hoc, Student–Newman-Keuls tests were used to reveal cantly affected plasma CCK concentration (F[4,24],
which means differed significantly. The area under the CCK 14.55; P ⬍ 0.001; Figure 4A and B). Integrated CCK
response curve was also calculated by using the trapezoid rule
response was significantly greater for all lipid infusion
(Sigma plot for Windows 5.0; SPSS Inc.) and compared by
conditions compared with saline (F[4], 15.08; P ⬍
using 1-way ANOVA with conditions as factors; post hoc
comparisons were made by using the Student–Newman-Keuls 0.001); post hoc analysis showed that area under the
test. curve CCK response was significantly greater after In-
tralipid infusion than all modified fatty acid composition
oil conditions (P ⬍ 0.05). No differences were found
Results between modified fatty acid composition oil conditions.
Study 1
Food intake. One-way repeated ANOVA indi- Discussion
cated a highly significant effect of infusion type on food The results of this study show that emulsions of
intake (F[4,36], 5.74; P ⫽ 0.001). Student–Newman- oils with different saturation indices, despite being iso-
946 FRENCH ET AL. GASTROENTEROLOGY Vol. 119, No. 4

oral stimulation (provided by the test meal) in humans.

Previous studies have reported a reduction of pre–test
meal ratings of hunger during infusion of oils, but these
reductions have been at much higher infusion concentra-
tions (4.9 kcal/mL for 75 minutes).5,7 It is unlikely that
in the present study significant amounts of the test meal
would have emptied from the stomach before the end of
the infusion.19,20 Thus, the satiating effect of the infused
oils is likely to be signaled by a combination of stimu-
lation of the upper intestine by constituents of the oils
with the gastric load produced by ingestion of the test
meal.
The most potent reduction in food intake was ob-
served after Intralipid infusion; this was used as a posi-
tive control in this experiment based on previous find-
ings.4 Intralipid is manufactured from purified soybean
oil and, thus, predominantly contains a mixture of lino-
leic (C18:2) and linolenic (C18:3) acids (COMP-EAT
dietary analysis software; Lifeline Nutritional Services
Ltd., London, England). The linoleic acid– containing oil
was also a potent inhibitor of food intake in this study,
whereas stearic/oleic– and oleic acid– containing oils did
not significantly reduce food intake compared with saline
infusion. Thus, it seems that the specific properties of the
C18 oils lead to differences in the biological responses. It
is unlikely that the lack of a significant effect of either
oleic or stearic/oleic infusions is caused by a type II error
because the calculated power of the present study is
0.803. One potential confounding factor concerning the
Intralipid infusion is that this contains different emulsi-
fiers from the 3 other oil emulsions. However, as stated
earlier, this was included as a positive control and the
Figure 2. The effect of upper intestinal infusions (F, control; ■,
Intralipid; Œ, linoleic; , oleic; }, stearic/oleic) on visual analogue
major findings of this study concern the differences in
scale ratings of (A) hunger and (B) fullness. Values are expressed as responses found between the linoleic, oleic, and stearic/
mean ⫾ SEM. oleic acid– containing oils.

caloric, suppress food intake in normal-weight men with

markedly different potencies. The greatest reduction in
test meal intake followed Intralipid infusion (225 kcal),
and secondly, by linoleic acid infusion (153 kcal). These
results support previous findings from animal studies12
and suggest that the properties of oils containing C18
long-chain fatty acids, differing only in the degree of
unsaturation, lead to marked differences in the satiating
potencies in humans. No significant changes in visual
analogue reports of hunger or fullness were observed
during the infusion periods. These results have been
reported previously in studies conducted in our labora-
tory,4 but in both studies we have observed significant
effects on food intake. This may suggest that for intes-
Figure 3. The effect of upper intestinal infusions (F, control; ■,
tinal stimulation by fats (and possibly other nutrients), Intralipid; Œ, linoleic; , oleic; }, stearic/oleic) on plasma triglyceride
to signal satiety requires an interaction with gastric and concentration. Values are expressed as mean ⫾ SEM.
October 2000 FATTY ACIDS AND APPETITE 947

saturated and transunsaturated fatty acids are less well ab-

sorbed under certain conditions.23,24 The melting point will
also affect the ease of emulsification of the oils, which will
determine the ease of dispersion in emulsion both before
infusion and by bile salts in the small intestine. In the
present study, however, after slow infusion of oils into the
upper small intestine, there was no difference in the rate of
early triglyceride appearance in the plasma. We recognize
that this measure does not wholly represent triglyceride
absorption and that peak plasma triglyceride levels do not
usually occur until 4 hours after a meal.25 The present
results do not, however, show gross differences in the rate of
early absorption and suggest that the likely site of feedback
responses elicited is preabsorptive. This argument is sup-
ported by work showing that the appearance of radio-
labeled fats in the plasma does not occur until after sup-
pression of feeding in rats.26,27 At present, we do not know
whether differences in plasma triglyceride levels would be
observed if the time of measurement was extended.
It has been shown that the potency of physiologic
responses from the small intestine in dogs after infusion
of fats is related to the area of the small intestine exposed,
presumably caused by the number of putative receptors
recruited.28 Fatty acids, as opposed to the triglyceride
molecule, are the critical stimulus for many physiologic
responses to fat arising from the small intestine, includ-
ing the satiety responses and CCK release.26,29,30 CCK,
within the physiologic range, is a signal to reduce food
intake in humans.31,32 The order of potency of the dif-
ferent oils in suppressing food intake in this study was
similar to that of a study that investigated the effects of
the degree of unsaturation of oils on CCK response after
oral ingestion.16 Corn oil (predominantly 18:2-linoleic
acid) was the most potent oil tested, followed by olive oil
Figure 4. Fatty acid–induced CCK secretion. (A) Time course of CCK
responses to fatty acid infusion. Peak CCK levels were reached within (predominantly 18:1-oleic acid), and lastly beef suet
20 minutes after the commencement of infusion for all lipid infusions (predominantly 18:0 and 16:0-stearic and palmitic ac-
(F, control; ■, intralipid; Œ, linoleic; , oleic; }, stearic/oleic). CCK ids); thus, the changes in CCK concentration were di-
response after Intralipid infusion was significantly different from sa-
line at all time points from 20 minutes onward and from all other fatty rectly related to the degree of unsaturation. However, in
acid infusion conditions from 20 minutes onward apart from time 30. the present study there was not a direct relationship
*P ⬍ 0.05 vs. control and modified oil emulsions; †P ⬍ 0.05 vs. between plasma CCK concentration and food suppression
control. (B) Area under plasma CCK curve during control and lipid
infusions. Total CCK response after Intralipid infusion was signifi- (although it is difficult to make a direct comparison
cantly greater than both control and all modified oil emulsions, total because these were conducted in separate experiments).
CCK response after modified oil emulsions was significantly greater Although Intralipid did produce the greatest suppression
than control infusion, and there were no differences in CCK response
between modified oil emulsion infusions. *P ⬍ 0.05 vs. control and in food intake and the highest CCK response, there was
modified oil emulsions; †P ⬍ 0.05 vs. control. Values are expressed not a distinction between the release of CCK after infu-
as mean ⫾ SEM. sion of the other oils enriched with specific fatty acids.
These findings suggest that other mediators, in addi-
The degree of unsaturation of fatty acids of the oils will tion to CCK, may be involved in the satiety responses
lead to differences in a number of physicochemical charac- observed in the present study. One possible mediator of
teristics (e.g., melting point, boiling point, pH level).21,22 satiety potentially affected by fatty acid composition is
These differences can have a major effect on the rate of apolipoprotein A-IV (apo A-IV). Apo A-IV is secreted by
absorption of lipids.22 In particular, it has been noted that the small intestine in response to fat feeding and may
948 FRENCH ET AL. GASTROENTEROLOGY Vol. 119, No. 4

mediate the anorectic effect of a lipid meal.33,34 Not only 16. Beardshall K, Frost G, Morarji Y, Domin J, Bloom SR, Calam J.
Saturation of fat and CCK release: implications for pancreatic
will the rate of digestion and absorption of lipids affect carcinogenesis. Lancet 1989;2:1008 –1010.
the release of apolipoproteins,35 but differing degrees of 17. Stunkard AJ, Messick S. The three factor eating questionnaire to
unsaturation of fatty acids may lead to binding of dif- measure dietary restraint, disinhibition and hunger. J Psychosom
Res 1985;29:71– 83.
ferent amounts of the apo A-IV molecule.36 Thus, the
18. French SJ, Castiglione KE, Francis J. Techniques for intestinal
final pattern of released apo A-IV may be markedly nutrient infusion in humans. In: Wellman PJ, Hoebel B, eds.
altered depending on the nature of the fatty acid infu- Ingestive behaviour protocols. New York: SSIB, 1997: 81– 86.
sion. The effects of the present oils on release of apo A-IV 19. Bracco U. Effect of triglyceride structure on fat absorption. Am J
Clin Nutr 1994;60(suppl):1002S–1009S.
and their relation to differences in the satiety effect 20. Brener W, Hendrix TR, McHugh PR. Regulation of gastric emptying
remain to be tested. of glucose. Gastroenterology 1983;85:76 – 82.
In conclusion, our results show marked differences in 21. McHugh PR, Moran TH, Wirth JB. Postpyloric regulation of gastric
emptying in rhesus monkeys. Am J Physiol 1982;243:R408 –R415.
the potency of suppression of food intake by emulsions of 22. Small DM. The effect of glyceride structure on absorption and
oils in humans; linoleic acid was the most potent sup- metabolism. Annu Rev Nutr 1991;11:413– 434.
pressor of food intake of the specific oils tested. These 23. Constantin MJ, Savary P. On the incorporation of stearic acid in
the triglycerides of rat-lymph chylomicrons. Biochim Biophys Acta
effects may be caused by the different physicochemical
1965;106:248 –266.
properties of the oils affecting CCK response and other 24. Bernard A, Echinard B, Carlier H. Differential absorption of two
possible mediators that remain as yet untested. fatty acid isomers: elaidic and oleic acids. Am J Physiol 1987;
253:G751–G759.
25. Cohn JS, McNamara JR, Cohn SD, Ordovas JM, Schaefer EJ.
References Postprandial plasma lipoprotein changes in human subjects of
1. Bray GA, Popkin BA. Dietary fat intake does affect obesity. Am J different ages. J Lipid Res 1988;29:469 – 479.
Clin Nutr 1998;68:1157–1173. 26. Greenberg D, Smith GP. The controls of fat intake. Psychosom
2. Blundell J, Macdiarmid J. Fat as a risk factor for overconsump- Med 1996;58:559 –569.
tion: satiation, satiety and patterns of eating (suppl). J Am Diet 27. Greenberg D, Kava RA, Lewis DR, Greenwood MRC, Smith GP.
Assoc 1997;97:S63–S69. Time course for entry of intestinally infused lipids into blood of
3. Stubbs RJ, Harbron CG, Murgatroyd PR, Prentice AM. Covert rats. Am J Physiol 1995;269:R432–R436.
manipulation of dietary fat and energy density: effect on sub- 28. Lin HC, Doty JE, Reedy TJ, Meyer JH. Inhibition of gastric empty-
strate flux and food intake in men eating ad libitum. Am J Clin Nutr ing by sodium oleate depends on length of intestine exposed to
1995;62:316 –329. nutrient. Am J Physiol 1990;259:G1031–G1036.
4. Castiglione KE, Read NW, French SJ. Food intake responses to 29. Meyer JH. Motility of the stomach and gastroduodenal junction.
upper gastrointestinal lipid infusions in humans. Physiol Behav In: Johnson LR, ed. Physiology of the gastrointestinal tract. 2nd
1998;64:141–145. ed. New York: Raven, 1987: 613– 629.
5. Drewe J, Gadient A, Rovati LC, Beglinger C. Role of circulating 30. Lewis LD, Williams JA. Regulation of cholecystokinin secretion by
cholecystokinin in control of fat-induced inhibition of food intake food, hormones, and neural pathways. Am J Physiol 1990;258:
in humans. Gastroenterology 1992;102:1654 –1659. G512–G518.
6. Greenberg D, Smith GP, Gibbs J. Intraduodenal infusions of fats elicit 31. Lieverse RJ, Jansen JBMJ, Masclee AAM, Lamers CHBW. Satiety
satiety in sham feeding rats. Am J Physiol 1990;259:R110–R118. effects of a physiological dose of cholecystokinin in humans. Gut
7. Welch IMCL, Sepple CP, Read NW. Comparisons of the effects on 1995;36:176 –179.
satiety and eating behaviour of infusion of lipid into the different 32. Ballinger A, McLoughlin L, Medbak S, Clark M. Cholecystokinin is
regions of the small intestine. Gut 1988;29:306 –311. a satiety hormone at physiological post-prandial plasma concen-
8. Woltman T, Reidelberger R. Effects of duodenal and distal ileal trations. Clin Sci 1995;89:375–381.
infusions of glucose and oleic acid on meal patterns in rats. Am J 33. Fujimoto K, Cardelli JA, Tso P. Increased apolipoprotein A-IV in rat
Physiol 1995;269:R7–R14. mesenteric lymph after lipid meals acts as a physiological signal
9. Maggio CA, Koopmans HS. Food intake after intragastric meals of for satiation. Am J Physiol 1992;262:G1002–G1006.
short-, medium-, or long-chain triglyceride. Physiol Behav 1982; 34. Fujimoto K, Fukagawa K, Sakata T, Tso P. Suppression of food
28:921–926. intake by apolipoprotein A-IV is mediated through the central
10. McCaffery J, French SJ, Greenberg D. Differential satiety effects nervous system in rats. J Clin Invest 1993;91:1830 –1833.
after duodenal infusion of medium-chain fatty acids in sham 35. Feldman EB, Russell BS, Hawkins GB, Forte T. Fatty acid compo-
feeding rats (abstr). Appetite 1994;23:315. sition of the diet and intestinal lipoproteins (abstr). Arterioscle-
11. Matzinger D, Drewe J, Ketterer S, Hildebrand P, Dübendorfer R, rosis 1981;1:84.
Beglinger C. Effect of digestion products of fat on food intake and 36. Feldman EB, Russell BS, Chen R, Johnson J, Forte T, Bennett
satiety in humans (abstr). Gastroenterology 1997;112:A892. Clark S. Dietary saturated fatty acid content affects lymph li-
12. Lewis DR, Philopena JM, Greenberg D. Increasing the saturation poproteins: studies in the rat. J Lipid Res 1983;24:967–976.
of a fatty acid increases its satiating potency in rats. Soc Neu-
rosci Abstr 1990;16:295.
13. Hopman PM, Jansen JBMJ, Rosenbusch G, Lamers BHW. Effect
of equimolar amounts of long-chain triglycerides and medium- Received November 2, 1999. Accepted May 10, 2000.
chain triglycerides on plasma cholecystokinin and gallbladder Address requests for reprints to: Stephen J. French, Ph.D., Centre for
contraction. Am J Clin Nutr 1984;39:356 –359. Human Nutrition, University of Sheffield, Northern General Hospital,
14. Gibbs J, Young R, Smith GP. Cholecystokinin decreases food Herries Road, Sheffield, S5 7AU, England.
intake in rats. J Comp Physiol Psychol 1973;84:488 – 495. Supported by a research grant from the Biotechnology and Biolog-
15. Douglas BR, Jansen JBMJ, de Jong AJL, Lamers CBHW. Effect of ical Sciences Research Council (to S.J.F).
various triglycerides on plasma cholecystokinin levels in rats. J The authors thank Dr. Gareth Evans and Prof. Wolfgang Langhans
Nutr 1990;120:686 – 690. for their critical review of the manuscript.