Short-term effects
and carbohydrates
of different
on
Cees
Jan A Weststrate,
de GraaJ
Tome
and Pauline
normal-weight
varied
in energy
and
macronutrient
and
1 .67
MJ
there
Besides
a zero
energy levels [0.42,
combined
with
three dominant
sources of macronutrients
(99% of energy from
CHO, 92% ofenergy
from fat, and 77% ofenergy
from protein).
After breakfast
the subjects
were not allowed
to eat or drink
(except
water) for 3.5 h. They then recorded
their voluntary
food intake
for the remainder
of the day. Subjects
also rated
their subjective
feelings concerning
food intake
on five different
types ofappetite.
The results showed that neither energy content
nor macronutrient
composition
ofthe liquid breakfasts
had any
effect on energy and macronutrient
intake during lunch and the
remainder
of the day. Ratings
of different
types
of appetite
showed
an increasing
satiating
effect with increasing
energy content ofthe breakfasts.
Proteins,
fats, and CHOS had similar effects
on appetite.
Am J C/in Nuir l992;55:33 -8.
1 .05,
[0.3 Mi (8 kcal)],
contents.
250,
(100,
were three
and
400
kcal)1
of protein,
Jas
Studies
bohydrates
WORDS
carbohydrate,
Appetite,
fat, dose
satiating
response,
efficiency,
caloric
energy,
protein,
compensation
Introduction
offood,
energy appears
to be the variable
for the regulation
ofappetite
(I). Because
the three main sources ofenergy
in our food are carbohydrates,
fats, and proteins,
it is important
to know whether
or not there
are differences
in the satiating
efficiency
of these three macronutrients.
The term satiating
efficiency
originates
from Kissileff
(2). He argued
that the satiating
effects of various
dimensions
of food (eg, volume,
weight,
viscosity)
should
be investigated
systematically
by manipulating
the level of one dimension
of
food while holding
the other dimensions
constant.
In such an
experiment,
subjects get different
preloads
with a varying level
ofa particular
food dimension.
Subsequently,
subjects get a test
meal. Kissileff(2)
defined the satiating
efficiency
ofa particular
food dimension
as the extent to which that dimension
is capable
of reducing
subsequent
energy intake. He applied
this strategy
for assessing the satiating
efficiency
ofliquid
and solid foods (3)
and various types ofsoups
vs a combination
ofcrackers,
cheese,
and apple juice (4). In this paper the satiating
efficiency
of a
Ofall
primarily
food
item
the properties
responsible
is defined
as the
is capable of influencing
ergy intake.
Am
J C/in Nuir
l992;55:33-8.
extent
motivational
Printed
to which
ratings
in USA.
a certain
food
or subsequent
© 1992 American
item
en-
Society
on the satiating
efficiency
of proteins,
fats, and carhave fixed the energy content
of a preload
at one
level
of energy
these
preloads
while
sessing
ratios
efficiency
on
food
within
was established
intake
and by as-
intake.
to motivational
ratings,
et al (7) found
of different
macronutrient
ratings
energy
respect
Barkeling
efficiencies
the
satiating
motivational
subsequent
With
and
varying
(5-1 1). The
by measuring
Geliebter
no differences
macronutrients.
between
Hill
and
(5), Driver (6),
the satiating
Blundell
(8, 9)
and Rolls et al (10) found that the protein
preloads
were more
satiating
than the carbohydrate
or fat preloads.
With respect to
subsequent
energy
intake,
Geliebter
(5) found no differences
whereas
Barkeling
et al (7), Hill and Blundell
(9), and Rolls et
al (10) found that energy intake in a subsequent
meal was lower
after the high-protein
preload
than after high-fat
or high-carbohydrate
preloads.
for calorie,
protein
Booth
was
et al (1 1) also
more
satiating
concluded
than
that
were
fat
calorie
or carbo-
hydrates.
The
may
KEY
fats,
differences
have
in the results
resulted
preloads
used.
ficiency
of the
partly
from
For an adequate
macronutrients
of the aforementioned
the
different
description
it is appropriate
energy
studies
levels
of the
of the satiating
to assess
efdose-
response
relationships
(2). When dose-response
relationships
are
assessed,
it should be possible
to express the magnitude
of the
effect of each macronutrient
per unit of weight or energy. This
enables
a sound comparison
between
the satiating
efficiencies
of different
food dimensions.
The objective
of this study was to investigate
the satiating
efficiency
of proteins,
fats,
and
carbohydrates
by giving
the
sub-
jects liquid breakfasts
containing
different
ratios of the three
macronutrients.
The energy content
was varied within each type
of macronutrient.
The description
of different
motivational
concepts
with respect
to food intake
was derived
from unpublished observations
(W Vaessen,
iA Weststrate,
and B de Boer,
1989) that feelings regarding
food intake and appetite
could be
divided into four different types: 1) appetite
for a meal; 2) appetite
for a snack; 3) feeble, weak with hunger;
and 4) a feeling
of
1 From
the Department
of Human
Nutrition,
Wageningen
Agricultural
University,
Wageningen,
and Unilever
Research
Laboratory,
Vlaardingen,
The Netherlands.
2 Address
reprint requests to C de Graaf, Department
of Human
Nutrition, Wageningen
Agricultural
University,
P0 Box 8129, 6700 EV
Wageningen,
The Netherlands.
Received
March 18, 1991.
Accepted
for publication
July 3, 1991.
for Clinical
Nutrition
33
Downloaded from www.ajcn.org at Univ of Colorado Libraries Serials Dept on August 15, 2007
This study investigated
the satiating efficiencies
fats, and carbohydrates
(CHOs). Twenty-nine
female,
subjects each received
10 liquid breakfasts,
which
ABSTRACI’
ofproteins,
condition
Huishof
amounts
34
DE
oversatiety
or
fullness.
For
“appetite
for a snack”
sweet”
and “appetite
Subjects
and
the
purpose
was divided
for something
of the
present
into “appetite
savory.”
GRAAF
ET
AL
TABLE
study,
29 female
kg/m2)
volunteers
varied
aged
mass
indices
(in
mean
(±SD)
of 2 1 . 1 ± 1 .9. All the subjects
99 en% CHO
1 en% PRO
0 en% FAT
between
1 7.3
19-23
and
were
y. Body
25. 1 with
students
of hunger
and
satiety.
treatment
on
by the Medical
of Human
They
were
not
aware
a particular
day.
Ethical
Committee
of receiving
feeling
conditions
in energy
dition
were
and
consisted
nutrient
factor,
energy,
250,
and
many
tein,
and
The
to one
was
added
preloads
was one
zero
con-
of the
1 .05,
factor,
protein,
ofthe
and
1.67
macronutrient,
and fat.
carbohydrate,
part
in a mixture
commercially
to disguise
available
textural
I
Table
pro-
Zoetermeer,
of nine
tomato
parts
juice,
tap
The
carbohydrate
available
diet
product
The
Netherlands).
In the
Dutch
for the
Energy
rated
Uden,
bitterness,
reported
to consist
of88%
milk
the preloads
protein
(Slagroom,
Melkunie,
Woerden,
The
Netherlands).
on
All subjects
naire
completed
the Dutch
developed
by van
Eating
Strien
(12),
for restraint
eating, external
eating, and emotional
The subjects were asked to express their feelings
a slash on five 150-mm
visual-analogue
meal, 2) appetite
for something
sweet,
Question-
producing
eating.
by means
of
One of each of these
the five visual-analogue
or 5) feeble, weak with hunger.
terms were placed above the center of
scales, which were anchored
at the left
and
the terms
savory;
right
The
4) satiety
hand
subjects
(fullness),
sides
were
with
orally
weak
instructed
and
that
strong,
appetite
nutrient
respectively.
for a meal
and coded
table
the
is not
by experienced
di-
by using
perceived
and
intensity
saltiness,
freshness.
Ratings
of
sourness,
were
made
scales.
preloads
were offered
subject.
Each
day,
resulting
The
first
analysis.
The
in a random
subject
received
in a within-subjects,
experimental
day
was
sequence
each
whole
study
lasted
and
preload
on
repeated-measures
regarded
as a practice
subjects
to the experimental
from this day were excluded
procedure.
from further
6 w.
Subjects were invited to come to the laboratory
between 0800
and 0900. They were asked not to eat or drink after 2300 the
previous
evening.
The preload
was consumed
through
a straw
placed in a milk-shake
beaker.
During
consumption
subjects
wore a noseclip,
preventing
them from detecting
any differences
in the odors of the different
preloads.
Subjects
five types
were instructed
ofappetitejust
to rate their feelings
before
the consumption
a sweet
peanuts,
pleasantness
of the preload
were rated
of the preload.
Subjects
were requested
Satiety
This
meal.
calculated
sweetness,
sandiness,
consist
be accurately
(13).
and
attributes:
visual-analogue
were
then
Appetite
for something
savory is appetite
for
cocktail
nuts, French fries, or a savory dessert.
refers to a feeling of having eaten too much (a
evening
intakes
refers to appetite
for a complete
meal,
either
a hot meal or a
sandwich
meal.
Appetite
for something
sweet is appetite
for a
sweet snack,
eg, a cookie,
chocolate,
a candy
bar, sweet pie, or
dessert.
cheese,
(fullness)
of stan-
usually
can
measures.
at the
the pleasantness
day to accommodate
The data obtained
scales:
1) appetite
for a
3) appetite
for something
items
weight
household
for each
design.
scores
weights
by means
snack
the
drinks
composition
The different
Behavior
and
were checked
viscosity,
a different
(DEBQ)
Soehnle-
The
Procedures
varied
Measurements
scalina,
FRG).
estimated
of which
on seven
150-mm
lunch
and
and
food
To obtain
an acceptable
product
from a sensory
point
of view, the protein
was mixed
with carbohydrates
and
fat. The fat preload
contained
commercially
available
cream
(Soehnle
Each
food diary
for each day
five pages.
The five pages
were
standard
foods
food diaries
etitians.
Subjects
Netherlands)
diet
items,
by using
case
The
1 2% water.
scales
Murrhardt/WUrtt,
were
with
weak
by means ofa food
were weighed
to an
meal
electronic
consumed
food
the Dutch
The
Feeble,
evening
household
measures.
an instruction
and
of standard
the
to
for the
& Co,
foods
the package
the product
consisted
in weight of 90% maltodextrine, 7% maltose,
and 3% glucose.
The protein
preload
contained
a commercially
available
diet product
(Protein
88, Wander,
and
at Christmas).
meant to record the amount
ofwater
consumed
in the morning
(page 1) and the foods and drinks
consumed
at lunch
(page 2),
in the afternoon
(page 3), at the evening
meal (page 4), and in
the evening
(page 5). The scales were used for the evening
meal
(Fan-
According
of 2 g on
estimated
which
differences.
a commercially
Nutricia,
5 en% CHO
3 en% PRO
92 en% FAT
experience
used
GMBH
other
only.
dissolved
foods
dards
for
contained
nine conditions
factorial
design.
0.42,
second
carbohydrate,
composition
were
contained
tomalt,
levels:
The
accuracy
Waagen
fat preloads.
nutrients
water
three
kcal).
also had three
levels:
shows the macronutrient
There
people
The
that
solutions
and the remaining
to a three-by-three
had
400
of550-mL
contents.
[0.03
Mi (8 kcal)]
composed
according
The first
Mi (100,
27 en% CHO
70 en% PRO
3 en% FAT
hunger is a strong urge to eat.
Energy and nutrient
intakes were recorded
Preloads
varied
Fat
a
The experiment
was
ofthe
Department
Nutrition.
10 preload
Protein
S Each
type of preload
was given at three energy levels, 0.42, 1.05,
and 1.67 MJ (100, 250, and 400 kcal). The 10th condition consisted of
a zero preload that had an energy content
of 0.03 MJ (8 kcal).
diary.
The
of preloads
Ratings
only
in the
every
water
morning
30 mm
until
after
3 h and
were
the
done
first
45 mm
after
rating.
after
concerning
the
ofthe
preload.
1 5, 30, and
45 mm
and
Sensory perception
and
just after consumption
not to eat, and drink
the first
rating,
just
before
Downloaded from www.ajcn.org at Univ of Colorado Libraries Serials Dept on August 15, 2007
particular
approved
types
a
at the
Agricultural
University
and were paid for their participation.
The subjects
were informed
that the experiment
was being
carried out to investigate
the effect ofdifferent
nutrients
on feelings
of different
Carbohydrate
Subjects
were
composition
Type of preload
methods
The subjects
1
Macronutrient
for something
SATIATING
EFFICIENCY
OF
35
MACRONUTRIENTS
Macronutrient
Energy
0;00
115
0:30
Time
1:45
after
2:16
first
2:46
rating
3:16
3:46
0:00
0:30
(hour:min)
1:16
Time
1:46
after
first
2:16
2:46
rating
3:16
3:45
(hour:min)
FIG 1. Net responses
to appetite
for a meal as a function
oftime
after the first rating. The first rating was made just
before consumption
ofa liquid breakfast.
Left panel: curves for each level ofenergy
ofthe preloads averaged
over type
of macronutrient.
Right panel: curves for the different
types of macronutrient
in the preloads
averaged
over energy
levels. Filled symbols denote a significant
difference
between
ratings, P < 0.05.
consumption
of the preload.
structed
to eat whatever
they
the
food
Data
they
consumed
After that time
felt like eating.
for the
remainder
subjects
Subjects
of the
were inrecorded
day.
to the
on the 150-mm
dimensions
were
(Famngton
and were
read
responded
to the
responded
to the right
automatically
left side
analysis
and macronutrient
out with
visual-analogue
scales
for the appetite
by an Optical
Mark
Reader
Data Processing
BY, Haarlem,
The Netherlands)
converted
into scores from 1 to 25. A score of 1 cor-
Statistical
P
of the
scale
(weak),
a score
of 25 cor-
side of the scale (strong).
for
the
intakes
effects
of the
preloads
and motivational
repeated-measures
analysis
on
ratings
of variance
energy
was carried
(ANOVA)
(14)
by using the 8V module
ofthe
BMDP
statistical
software
package
(15). The time and type of preload
were fixed within
subject
factors.
Subjects
were treated
as a random
factor.
With respect
to characteristics
on eating behavior,
subjects
were divided
by
a median
split into groups with low and high scores of restraint
and external and emotional
eating. In this case, group was treated
as a fixed between-subjects
factor.
Relationships
between
motivational
ratings and energy intake
and between
sensory ratings and motivational
ratings were determined
by using linear-regression
analysis.
five
Results
The first
at zero
ratings
and
before
for each of the
the type of preload
that
a meal
(F19,2521
(fullness)
hunger
one
other.
meal
Because
was
tailed
types
larger
analysis
the
consumption
of the
preload
were
set
different
conditions.
ANOVA
showed
had a significant
effect on appetite
for
appetite
for something
for something
savory
sweet
(F[9,2521
of appetite
than
variance
the
is given
left-hand
P
were
strongly
in the
ratings
to the
types
a de-
The
of appetite
were
similar
to
the net responses
to
macronutrient
showed
that the
5.02, P < 0.01).
showed
at 1.75,
mean
the different
showed
responses
energy
that
the
to appetite
levels,
effect
also
of the
significant
(F12,561 = 0.25,
time yielded
no significant
P
bohydrates,
fat did
protein,
and
panels
ofFigure
for each
(including
the zero
=
12. 1, P < 0.001)
0.50).
differences.
>
not
the
the
and
of Figure
averaged
preload.
1
over
ANOVA
of macronutrient
was
Separate
ANOVAs
The net responses
not
at each
to car-
differ.
to appetite
as a function
of the
macronutrients
with a different
that the main
preload)
but not
Appetite
content
excluding
type
panel
for a meal
that
2.25,
2 show the net responses
of the
P > 0.5) and fat (F13.841
analyses
the existence
ofa
can be inferred.
a higher
energy
time
responses
separately
the
for each
to
attributes,
only.
1 shows
similar
for a
other
dimension
of Figure
out
Factor
to appetite
responses
to appetite
for a meal were different
2.75 h after the first rating.
The right-hand
shows
carried
0.001),
<
correlated.
were
appetite
for a meal averaged
over the different
preloads,
excluding
the zero preload.
ANOVA
effect ofenergy
level was significant
(F12,561 =
ANOVAs
P
0.02).
<
dimensions
variance
for this
panel
3.55,
=
2.27,
=
on the subjective
ratings
yielded
>
1 , indicating
that the responses
and results
for the other
four
those
for appetite
for a meal.
The
(F19,2521
(F19,2521
to the five appetite
time after the first rating,
energy.
ANOVA
showed
satiety
(F19,2521 = 3.21, P < 0.01)
=
2.30, P < 0.05), appetite
with
different
responses
for a meal
ofhunger
satiety
weak
The
The three
Feelings
0.01),
<
feeble,
analysis
with varimax
rotation
one factor with an eigenvalue
analysis
Ratings
7.70,
=
and
curve
effect
for each level of
of energy
level
was significant
for the protein
for the carbohydrate
(F13.841
0.5, P >
dose-response
0.5)
preloads.
relationship
for a meal is lower
than after a preload
after
with
(F13,1
=
0.36,
From
these
for protein
a preload
with
a lower energy
Downloaded from www.ajcn.org at Univ of Colorado Libraries Serials Dept on August 15, 2007
-15
36
DE
Protein
040
aft.r
ET
AL
Carbohydrates
uS
Tim.
GRAAF
i
*4S
fIrst ratIng
$dS
$S
050
0.10
TIm.
(hourmk)
US
after
%41
Fat
L5
first rating
141
$,IS
.4$
0.00
0.30
(hourrnin)
Time
b$
after
4$
first
L$
rating
t,41
$.$
(houunk)
content.
With
respect
to carbohydrates,
Figure
2 shows
that
with
energy
level
and
on the
energy
0. 10; nutrient:
F12,561 = 0.25, P > 0.50).
4 shows the amounts
ofdifferent
macronutrients
conafter the different
preloads
during
the whole day. No
Energy
and macronutrient
intakes
as afunction
of preload
3 shows the mean energy intake (±SD) for lunch and
the remainder
of the day as a function
of characteristics
of the
preload.
Note that the energy intake of the preload
itself is cxcluded. ANOVA
showed that there were no differences
in energy
intake
for the
(F19,2521
not differ
remainder
either
=
Energy
intake
of the
0.76, P
(F19,2521 =
>
day
0.50).
I .57, P
Whole
Lunch
aft#{149}rpreload
between
the
preload
con-
Energy
intakes
at lunch
did
> 0. 10). A two-way
ANOVA
during
0.42
Type
1.05
CHO
1.67
of preload
remainder
of the
of nutrient
day
for any of the macronutrients.
fats, protein,
and alcohol
respectively,
and
during
(23-4
range
1 . 15, 0.90,
of SDs
the whole
1 ) for fats,
(energy:
>
0.90,
of the
weights
0.74
(all, P>
0.05).
The
were,
The average
of macronutrients
consumed
day
were 74 g (49-9 1) for carbohydrates,
22 g ( 16-29)
for protein,
and I 7 g ( 10-3
33 g
1 ) for
alcohol.
The macronutrient
content
of the lunch
was also independent
from the type of preload.
The F192521 ratios for carbohydrates,
fats, and proteins
were, respectively,
0.75, 1.34, and
1 .82 (all, P > 0.05).
No alcohol
was consumed
during
lunch.
day
(MJ)
t05
0.03
ZERO
the
or type
effect of preload
was found
F19.2521 ratios
for carbohydrates,
and
Figure
ditions
sumed
intake
ofenergy
(zero
F12,561 =
Figure
P
no effect
as factors
The responses
to the l.67-MJ
were lower than the responses
loads at 1.75, 2.25, and 2.75
hand panel of Figure 2 shows
energy levels of the fat preloads
2.30,
showed
of macronutrient
condition
(400-kcal)
carbohydrate
preload
after the other carbohydrate
preh after the first rating.
The rightthat the responses
to the different
were almost equal.
excluded)
type
compared
with the zero preload
the responses
to appetite
for a
meal were lower for the first 2 h, but were higher
after 2.75 h.
0.42
1.05
PRO
1.67
(MJ and macronutrient)
0.42
1.05
FAT
1.67
FIG 3. Energy intake of 29 subjects at lunch and during the whole
day after consumption
of liquid breakfasts
with different
energy and
macronutrient
contents.
Subjects were not allowed to eat until 3.5 h after
breakfast.
The CHO-, PRO-, and FAT-preload
types contained
99 en%
carbohydrates,
77 en% proteins,
and 92 en% fat, respectively.
I ± SD.
t67
CHO
Type of preload
O.4z
#{149}
5
t67
F.,)
0.42
1
Ft,.
(MJ and macronutrient)
FIG 4. Mean macronutrient
intake of29 subjects during the remainder
of the day after consumption
of liquid breakfasts
with different
energy
and macronutrient
contents.
Subjects were not allowed to eat until 3.5
h after breakfast.
The CHO-, PRO-, and FAT-preload
types
contained
99 en% carbohydrates,
77 en% proteins,
and 92 en% fat, respectively.
For information
on standard
deviations,
see text.
Downloaded from www.ajcn.org at Univ of Colorado Libraries Serials Dept on August 15, 2007
FIG 2. Net responses
to appetite for a meal as a function
of time after the first rating, with a separate curve for each
level of energy. The first rating was made just before consumption
of a liquid breakfast.
Left panel: protein preloads
(77 en% PRO); middle panel: carbohydrate
preloads
(99 en% CHO); and right panel: fat preloads
(92 en% fat). For
reference,
the responses
to the zero preload are shown in each of the three panels. Filled symbols denote a significant
difference
between
ratings, P < 0.05.
SATIATING
Energy
intake as afunction
Compared
ulation,
ofeating-behavior
with van Strien’s
the
median
value
(12) Dutch
for
external
EFFICIENCY
characteristics
female
eating
reference
for
the
popsubjects
on the
emotional,
external,
and
restraint
factors.
None
had a significant
effect on the energy intake
the whole day (emotional:
F11,261 = 0.40, P > 0.5; external:
=
0.60, P > 0.5; restraint:
F11,261 = 0.07, P > 0.5).
during
factors
ofthese
Sensory
ratings
ofihe
F1 1,261
intake
at
lunch
and
during
the
whole
day.
The
motivational
ratings
at 2.25 h were chosen because
the differences
in ratings
between
conditions
were largest 2.25 h after the first rating.
The ratings ofpleasantness
did not correlate with energy intake
at lunch (r12881 = 0.08, P > 0.05) or with energy intake during
the remainder
of the day (r12881 = -0.02,
P > 0.05). Sweetness
ratings also did not correlate
with energy intake at lunch (r12881
=
-0.03,
P > 0.05) or with energy intake during
the remainder
of the day (r12881 = 0.03, P > 0.05). The only significant
correlations between
sweetness,
pleasantness,
and the motivational
ratings were those between
sweetness
and satiety (r12881 = 0.17,
P < 0.01) and between
pleasantness
and appetite
for something
sweet (r12881 = -0.20,
P < 0.01).
Motivational
ratings
before lunch and energy
intake
at lunch
About 96% ofthe lunches were consumed
between
1200 and
Depending
on the time of the breakfast,
subjects
were
allowed to eat again between
1 145 and 1245. The responses
to
the motivational
ratings 3.75 h after the first rating,
which were
in most cases just before lunch, correlated
significantly
with energy intake at lunch. The correlations
were 0.32 for appetite
for
a meal, 0.28 for appetite
for something
sweet, 0.40 for appetite
for something
savory, -0.23
for satiety (fullness),
and 0.33 for
1400.
feeble,
ability
weak
<
with
hunger.
All correlation
coefficients
had
a prob-
0.01.
intake.
tion
(8 kcal)]
than
after
a preload
with
a high
energy
content
did
did
MJ
[1.67
MJ (400 kcal)]. The nutrient
composition
of the preloads
had
no influence
on the energy intake during lunch and the remainder of the day. No shift occurred
in the consumed
amounts
of
carbohydrates,
Differences
between
were
disappeared
proteins,
and
fats. Thus,
after
a high-carbohydrate
preload,
the subjects
consumed
just as much carbohydrates
after a high-protein
or a high-fat
preload.
Similar
results
found for protein and fat consumption.
as
were
3.75
appearance
explains
no effect
h after
ratings
on intake,
of this
between
consumption
ofdifferences
results
emerged
highest
they
did
of the
were found
study
generally
I h after
have
consump-
1.75 and 2.75
in motivational
why no differences
The
preload.
ratings
after
in energy
intake
agree
with
the
h, but
The
dis-
3.75
mm
at lunch.
results
of
other studies on the satiating
efficiency
of the different
macronutrients
(5-1 1, 18, 19). However,
some specific results differ.
The present
results concur
with those of Geliebter
(5), Driver
Barkeling
et al (7), who also found
that the three
machad similar
effects on feelings
of hunger
and satiety.
In contrast,
the studies of Hill and Blundell
(8, 9) and of Rolls
et al (10) found that protein
had a larger effect on motivational
ratings
than did fats and carbohydrates.
Barkeling
et al (7), Hill and Blundell
(9), and Rolls et al (10)
found that subsequent
energy intake. was lower after a protein
preload
than after fat or carbohydrate
preloads.
These results
do not agree with the present findings.
It appears
that the present
findings are more in line with those ofGeliebter
(5) and Sunkin
and Garrow
(18), who also found that energy intake after a protein preload was not lower than after fat or carbohydrate
preloads.
Foltin et al (19) did not find any difference
in the effect on energy
intake between
high-fat and high-carbohydrate
lunches.
In that
study, however,
subjects ate less after a high-energy
lunch than
after a low-energy
lunch.
The present
finding
that different
macronutrient
preloads
did not result
in differences
in subsequent
macronutrient
consumption
concurs
with the results
of Rolls et
al (10).
The differences
between
the results of various studies may be
explained
by methodological
differences
in experimental
setup.
One issue is the time between preload and test meal or ad libitum
food intake. In the present study the subjects
were not allowed
to eat after consumption
of the breakfast
preload
until lunch
time. Two h after the preload
consumption
there were differences
in motivational
ratings;
for instance,
subjects
had a greater
appetite
for a meal
after
a low-energy
preload
than
after
a high-
energy
ratings
preload. The significant
correlations
between
motivational
before lunch and energy intake at lunch suggests that, if
subjects
had been allowed
to eat after 2 h, the result ofthe present
study might have been different.
In the experiments
of Rolls et
al (10) and Hill and Blundell
(9), the protein preload
was more
satiating
than
tiating
vational
efficiency
ratings
in the
The main conclusion
from these results is that subjects
not compensate
for the energy taken in at breakfast.
They
not eat more after a preload with a low energy content
[0.03
had
of the preload,
In these
preload.
Discussion
preloads
were
carbohydrate
of protein
was
and the subsequent
or fat preloads.
apparent
energy
experiments
the test meals
The time between
preload
present
study
was
almost
The
higher
from both the
intake
in a test
samotimeal.
were given 2-3 h after the
and ad libitum
food intake
4 h. As Figures
1 and
2 show,
this is an important
difference.
Apparently
the time between
the
preload
and test meal is a crucial variable
in determining
whether
or not caloric
compensation
occurs.
Similar
remarks
can be made with respect to the time of the
preload.
In the present
study the subjects
were given breakfast.
Would
the results have been different
if they had been given
lunch or dinner
instead
of breakfast?
The results of Barkeling
et al (7) and Foltin et al(19), in which subjects got lunch, suggest
that this can be an important
variable.
The present study lasted
only 1 d. If we had given the subjects
similar
preloads
for a
number
ofconsecutive
days, would the subjects then have com-
Downloaded from www.ajcn.org at Univ of Colorado Libraries Serials Dept on August 15, 2007
The sensory profile ofthe preloads
differed in all dimensions.
All F19,2521 ratios were > 2.50 with a probability
< 0.01.
The sweetness
and pleasantness
of preloads
may exert an influence on motivational
ratings and subsequent
energy intake
(16, 17). Therefore,
there was an investigation
as to whether
the
sweetness
and pleasantness
ratings correlated
with the motivational ratings 2.25 h after the first rating and with the energy
the
an effect on motivational
ratings. After a high energy intake in
the preload, ratings ofappetite
were lower than after a low energy
(6), and
ronutrients
preloads
37
MACRONUTRIENTS
Although
who participated
in the present
study could be classified
as average. The median values for restraint
and emotional
eating could
be classified
as above average.
An ANOVA
was carried out to
compare
the energy intakes
for the groups
with low and high
scores
OF
38
DE
GRAAF
pensated
for the amount
ofenergy
in the preload?
Ifso, at what
time would they have started compensating?
The results of LouisSylvestre
et al (20) suggest that subjects
learn about the consequences
of preloads
within 5 d. Similarly,
in the present
study
subjects
were given liquid, bland-tasting
artifical
preloads;
in
that sense our preloads
were similar to those ofGeliebter
(5). In
the other studies discussed,
subjects
were given normal
food
items. In the present
study subjects
did not compensate
for a
1.67
MJ (400-kcal)
difference
in preload.
At what
level
of energy
difference
between
the preloads
would they start to compensate?
These questions
show that the results and conclusions
of this
study might be limited to the specific experimental
conditions
determining
satiating
efficiencies
of
various
dimensions
We thank
Nijmegen
of
B
foods.
Karin
for their
Besselink
assistance
and Anita Nijnens
in data collection.
of the
1. Blundell
JE. Appetite
Drugs l990;39(suppl
disturbance
and the problems
7. Barkeling B, ROssner 5, Bj#{246}rvell
H. Efficiency ofa high-protein
meal
(meat)and
a high carbohydrate
meal (vegetarian)on
satiety measured
by automated
computerized
monitoring
ofsubsequent
food intake,
motivation
to eat and food preferences.
Int J Obes l990;14:743-51.
8. Hill AJ, Blundell
JE. Macronutrients
and satiety: the effects of a
high-protein
or high carbohydrate
meal on subjective
motivation
to
eat and food preferences.
Nutr Behav 1986;3:l33-44.
9. Hill AJ, Blundell JE. Comparison
of the action of macronutrients
on the expression
ofappetite
in lean and obese human subjects. Ann
N Y Acad Sci l990;580:529-31.
10. Rolls BJ, Hetherington
M, Burley VJ. The specificity
ofsatiety:
the
influence
of different
macronutrient
contents
on the development
of satiety. Physiol Behav l988;43: 145-53.
1 1. Booth DA, Chase A, Campbell
AT. Relative effectiveness
of protein
in the late stages of appetite
suppression
in man. Physiol Behav
l970;5: 1299-302.
12. van Strien T. Eating behavior,
personality
traits and body mass.
Lisse: Swets & Zeitlinger,
1990.
13. Nederlands
Voedingsstoffenbestand
(NEVO)
Dutch nutrient
database 1989/1990.
Zeist, The Netherlands:
Stichting
NEVO,
1989 (in
Dutch).
Hogeschool
References
of overweight.
3): 1-19.
2. KissileffHR.
Satiating
efficiency and a strategy for conducting
food
loading experiments.
Neurosci
Biobehav
Rev 1984;8:l29-35.
3. KissileffHR.
Effects ofphysical
state (liquid-solid)
offoods
on food
intake: procedural
and substantive
contributions.
Am J Clin Nutr
l985;42:956-65.
4. Kissileff HR, Gruss LP, Thornton
J, Jordan
HA. The satiating
efficiency of foods. Physiol Behav l984;32:3
19-32.
5. Geliebter
AA. Effects of equicaloric
loads
of protein,
fat and carbohydrate
on food intake in the rat and man. Physiol Behav l979;22:
14. Winer RI. Statistical
principles
in experimental
design. New York:
McGraw-Hill,
1971.
15. Dixon WJ, Brown MB, Engelman
L, Jennrich
RI. BMDP statistical
software manual 1990. Berkely: University
ofCalifornia
Press 1990.
16. Hill AJ, Magson LD, Blundell JE. Hunger and palatability:
tracking
ratings of subjective experience
before, during and after consumption
of preferred
and less preferred
food. Appetite
l984;5:361-7
1.
17. Blundell
JE, Rogers PJ, Hill AJ. Artificial
sweeteners
and appetite
in man. In: Birch GO, Lindley MG, eds. Low calorie products. London: Elsevier,
1988.
18. Sunkin 5, Garrow iS. The satiety value ofprotein.
Hum Nutr Appl
Nutr
1982;36A:
197-201.
19. Foltin RW, Fischman
MW, Moran TH, Rolls BJ, Kelly TH. Caloric
compensation
for lunches varying in fat and carbohydrate
content
by humans
in a residential
laboratory.
Am J Clin Nutr l990;52:
267-73.
969-80.
6. Driver
Cit. The efficiency of meal composition
on the degree of
satiation
following
a test meal and possible mechanisms
involved.
BrJ Nutr
AL
l988;60:44l-9.
20.
Louis-Sylvestre
Learned caloric
103.
J, Tournier
adjustment
A, Verger P. Chabert
M, Delorne
B.
ofhuman
intake. Appetite
l989;12:95-
Downloaded from www.ajcn.org at Univ of Colorado Libraries Serials Dept on August 15, 2007
used.
Regarding
the complexity
of the issue, it appears
that a
careful and systematic
evaluation
of all potential
influences
is
required
to assess the factors that regulate
appetite
and food
intake.
This idea is equivalent
to Kissileff’s
(2) strategy
of
ET