European Journal of Clinical Nutrition (1999) 53, 535±541
ß 1999 Stockton Press. All rights reserved 0954±3007/98 $12.00
http://www.stockton-press.co.uk/ejcn
Effects of butter oil blends with increased concentrations of
stearic, oleic and linolenic acid on blood lipids in young adults
CC Becker1, P Lund1, G Hùlmer1*, H Jensen2 and B SandstroÈm2
1
Department of Biochemistry and Nutrition, Center for Food Research, Technical University of Denmark, Denmark; and 2Research
Department of Human Nutrition, Center for Food Research, The Royal Veterinary and Agricultural University, Frederiksberg, Denmark
Objective: The aim of this present project was to evaluate a more satisfactory effect on plasma lipoprotein
pro®le of spreads based on dairy fat.
Design: This study was designed as a randomised cross-over experiment with a three-week treatment separated
by a three-week wash-out period. Sixty ®ve grams of the fat content of the habitual diets was replaced by either
butter=grapeseed oil (90 : 10) (BG); butter oil and low erucic rapeseed oil (65 : 35) (BR) or butter blended in a
1 : 1 ratio with a interesteri®ed mixture of rapeseed oil and fully hydrogenated rapeseed oil (70 : 30) (BS).
Subjects: Thirteen healthy free-living young men (age 21±26 y) ful®lled the study.
Interventions: At the beginning and end of each diet period two venous blood samples were collected.
Triacylglycerol and cholesterol concentrations in total plasma and VLDL, LDL, IDL and HDL fractions were
measured, as were apo A-1 and apo B concentrations. Fatty acid composition of plasma phospholipids, plasma
cholesterol ester and platelets was also determined.
Results: Signi®cantly (P < 0.05) lower total and LDL-cholesterol concentrations were observed after the BR and
BS period, compared to BG. The effect of BR and BS did not differ. BG and BR resulted in equal concentrations
of HDL-C, but signi®cantly higher than BS. Consequently, a signi®cantly lower LDL-C=HDL-C ratio was seen
after the BR treatment compared to BG and BS. Apo A-1 concentrations were not signi®cantly different, but Apo
B was signi®cantly increased after BG.
Conclusions: Partially replacing milk fat with rapeseed oil seems to yield a more healthy spread. Stearic acid
had a HDL-C lowering effect compared to milk fat, but did not affect LDL-C signi®cantly. The addition of
stearic acid did not improve the plasma lipoprotein pro®le for young men with low cholesterol levels.
Sponsorship: Danish Food Research Programme (FéTEK I) and Danish Dairy Research Foundation.
Descriptors: butter oil; rapeseed oil; cholesterol; apoproteins; platelets; plasma phospholipids; plasma
cholesterol esters; stearic acid; oleic acid; linolenic acid
Introduction
The effect of individual fatty acids on blood lipids and
lipoproteins and their potential atherogenicity depends on
chain length and degree of unsaturation (Mensink & Katan,
1992; Yu et al, 1995; Howell et al, 1997). Fatty acids with
fewer than twelve carbon atoms do not seem to in¯uence
plasma cholesterol concentrations more than carbohydrate
does (Hill et al, 1990), whereas saturated fatty acid with
chain lengths of 12, 14, and 16 carbon atoms have shown a
hypercholesterolemic effect in humans (Hegsted et al,
1965; Keys et al, 1965; Yu et al, 1995). Early studies
(Ahrens et al, 1957; Hegsted et al, 1965; Keys et al, 1965),
as well as more recent ones (Kris-Etherton et al, 1993; Yu
et al, 1995) have shown that stearic acid, although a
*Correspondence: Prof G Hùlmer, Department of Biochemistry and
Nutrition, Bldg. 224, Technical University of Denmark, DK-2800 Lyngby,
Denmark.
Contributors: G Hùlmer and B SandstroÈm initiated the project, which was
part of the PhD study of CC Becker. All the authors were involved in
carrying out the experimental work. The food registration was conducted
by H Jensen and B SandstroÈm. CC Becker made the statistical analysis. CC
Becker, P Lund and G Hùlmer wrote the article and B SandstroÈm took part
in its ®nalisation. Guarantor: G Hùlmer.
Received Received 30 November 1998; revised 4 February 1999; accepted
13 February 1999
saturated fatty acid, seems to have an effect on plasma
lipids that more closely resembles that of oleic acid than the
other principal saturated fatty acids. The reason for the
observed neutrality of stearic acid is unknown, but it has
been suggested that it is poorly absorbed (Grande et al,
1970), or quickly desaturated to oleic acid (Bonanome &
Grundy, 1989). It seems, however, that desaturation is not
extensive enough to explain the non-cholesterolemic effect
of this fatty acid in humans (Rhee et al, 1997; Pai & Yeh,
1997).
Generally, oleic acid and linoleic acid consumption
leads to a lowering of plasma cholesterol (C) and LDL-C
concentrations when replacing saturates with 12±16 carbons (SFA) iso-energetically (Hegsted et al, 1965; Mensink
& Katan, 1992; Yu et al, 1995). In a meta-analysis of 27
clinical studies, Mensink & Katan (Mensink & Katan,
1992) concluded that polyunsaturated fatty acids (mostly
linoleic acid) and monounsaturated cis fatty acids (mostly
oleic acid) have similar effects on the concentration of
HDL-C. Another meta-analysis of 18 studies by Yu et al,
1995, found that monounsaturated cis fatty acids raised
HDL-C three times as much as polyunsaturated fatty acids.
A technological advantage of oleic acid over linoleic acid is
less susceptibility to oxidation, giving products with a
longer shelf life. Furthermore, the susceptibility to oxidation and the atherogenicity of plasma lipoproteins decreases
Blood lipids and milk fat based spreads
CC Becker et al
536
with increasing saturation of its constituent fatty acids,
which re¯ect the diet (Aviram & Eias, 1993).
The hypercholesterolemic effect of dairy products and
butter oil, has been reported often in both clinical (Wardlaw
& Snook, 1990; Kris-Etherton et al, 1993) and epidemiological studies (Renaud et al, 1991). Butter fat usually
contains about 60% saturates (by weight) of which approximately 10±12% is stearic acid and another 12% are short
chain fatty acids.
The purpose of this study was to investigate the effects
of enriching butter with oleic and linolenic acids (from
rapeseed oil) or with stearic acid (from fully hydrogenated
rapeseed oil) on the plasma concentrations of lipids, and
apolipoproteins. The re¯ections of dietary fatty acids on
fatty acid distribution in blood lipids as well as platelets
were also examined.
Through dilution of the butter fat with vegetable oils, a
product, similar to butter in appearance and taste, but with
less ability to raise cholesterol concentrations in the blood
stream, can be obtained. Rapeseed oil was selected for this
study, as it is a rich source of both oleic, and linolenic
acids. Linolenic acid seems to have an indirect inhibitory
effect on platelet aggregation (Renaud, 1990) and thrombosis as n-3 fatty acids give rise to other prostanoids than
do n-6 fatty acids.
There is however a limit for the addition of vegetable oil
if a solid spread, at room temperature, should be obtained.
Increasing the content of stearic acid might be a bene®cial
way of solving this problem instead of using partially
hydrogenated oils, which may have unfavourable effects
on blood lipids (Aro et al, 1997).
Subjects and methods
Design
Thirteen young men participated serving as their own
controls in this intervention trial. A three-period threetreatment cross-over design was used to minimise the
effect of variation between subjects. The study consisted
of three three-week dietary periods interrupted by a threeweek wash-out period. The subjects were randomly
assigned to one of the three treatment groups in the ®rst
period and subsequently randomly assigned to one of the
remaining two treatments. No intercurrent illnesses
occurred and the concentrations of C-reactive proteins
were below 5 mg=L throughout the experiment, except in
two cases where it was 12 and 40 mg=L. Values below 10
were considered normal so 12 mg=L is a little greater than
normal, while 40 mg=L is elevated and indicative of a
minor infection (Kessler et al, 1994). However, the lipid
values for these persons were normal, and they were not
excluded.
Subjects
Participants were adult male volunteers in good health
recruited among students from the University campus.
Applicants were selected according to age, the results of
a blood test and an interview. Eligible were non-smoking
male subjects in their twenties who had total serum cholesterol concentrations greater than 4.0 mmol=L, and who
exercised moderately or not at all. Candidates who were
diabetic or taking medication for any reason were eliminated. All participants agreed to maintain their pre-study
weight and level of physical activity. They were 21±26 y
old, and weighed between 61 and 88 kg at the outset of this
study.
The study protocol was reviewed and approved by the
regional ethical committee (Frederiksberg) and was in
accordance with the Helsinki declaration of 1983. Informed
written consent was obtained from each participant.
Test fats
Three test fats were prepared by Aarhus Oil, Aarhus,
Denmark. They were BG: butter=grapeseed oil (90 : 10);
BR: butter and low-erucic rapeseed oil (65 : 35); and BS:
butter blended in a 1 : 1 ratio with an interesteri®ed mixture
of rapeseed oil and fully hydrogenated rapeseed oil
(70 : 30). All fats were made to contain about 8.5%
(w=w) linoleic acid, to ensure an adequate supply thereof.
The fatty acid compositions of the dietary test fats are
given in Table 1. Regiospeci®c analyses were carried out
according to (Becker et al, 1993). A single lot of each test
fat was prepared to cover the needs of the entire study.
Fatty acid composition of the test fats was determined by
the gas chromatographic (GC) analysis of fatty acid methyl
esters, obtained by KOH catalysed transmethylation
according to (Christofferson & Glass, 1969). Fatty acid
methyl ester analyses were carried out on Hewlett-Packard
5880A gas chromatograph with a 30 m60.32 mm fused
silica capillary column coated with a 0.2 mm ®lm of SP2380
(Supelco Inc., Bellefonte, PA, USA). Initial oven temperature was maintained at 35 C for 3 min before rising to
155 C at a rate of 20 C=min. After 2 min the temperature
was increased at a rate of 10 C=min to 225 C and maintained there for 7 min.
Diets
The habitual dietary intake was assessed by a seven-day
weighed food record. This assessment was repeated for
seven days in each dietary period. Energy and nutrient
intake were calculated with the use of a national food
database `Dankost-2' (Mùller, 1989). During the test periods the subjects were given 65 g of test fat per day. Thirty
grams of test fat was delivered as spread and the remaining
Table 1 Fatty acid composition of the triacylglycerols (TG) and 2monoacylglycerols (2-MG) of the test fats. (wt%)
BG
C4:0
C6:0
C8:0
C10:0
C12:0
C14:0
C14:1 n-5
C15:0
C16:0
C16:1 n-7
C17:0
C18:0
C18:1 n-9a
C18:1 n-7
C18:2 n-6
C18:2 conj
C18:3 n-3
C20:0
BR
BS
TG
2-MG
TG
2-MG
TG
2-MG
5.0
2.6
1.4
3.0
3.6
10.4
1.4
1.1
26.1
0.9
0.7
9.9
23.0
0.6
8.7
1.0
0.5
0.1
0.0
0.0
0.0
0.8
4.2
18.9
2.3
1.6
34.8
2.7
0.5
5.0
18.5
0.3
8.9
0.6
0.5
0.1
3.7
1.9
1.1
2.2
2.6
7.6
1.1
0.8
20.0
1.3
0.4
7.3
34.3
1.5
8.5
1.2
4.3
0.4
0.0
0.0
0.0
0.7
2.5
13.2
1.6
1.1
25.5
2.0
0.4
3.9
26.5
0.4
13.6
0.4
7.9
0.2
2.8
1.5
0.8
1.8
2.0
6.0
0.8
0.6
16.8
1.0
0.4
19.2
30.9
1.5
8.2
1.0
3.9
0.6
0.0
0.0
0.0
0.8
2.6
11.9
1.4
1.0
23.7
1.8
0.3
14.5
26.4
1.5
8.4
0.4
4.5
0.6
BG: 90% butter oil ‡ 10% grapeseed oil; BR: 65% butter oil ‡ 35%
rapeseed oil; BS: 50% butter oil ‡ 50% interesteri®ed (70% rapeseed
oil ‡ 30% fully hydrogenated rapeseed oil). aIncluding trans 18:1 isomers.
Blood lipids and milk fat based spreads
CC Becker et al
in the form of two daily snacks (a cake and a bun) prepared at
the Department. The test fats were weighed in individual daily
portions and delivered weekly. Individual dietary instructions, based on the food records, and aimed to reduce habitual
fat intake by an amount similar to the provided test fat, were
given by a dietician. All visible fat in the form of spreads was
excluded as well as cheese. High fat dairy products were
replaced by low-fat products. Low-fat meat cuts, marmalade
and honey were delivered free from the Department in
amounts requested by the subjects for bread meals, which
comprise a relatively large part of the Danish diet. The
subjects were also instructed to keep their food habits as
similar as possible in the three test periods and to avoid
excessive food and alcohol intake. Duplicate portions of the
test foods and fats were directly analysed to check their
conformity with the planned fat intake.
Blood collection and analysis
At the beginning and end of each diet period two venous
blood samples were collected on consecutive days after
10 min supine rest in the morning with minimal stasis
(Davis et al, 1990). Subjects were fasting ( > 12 h) and
had abstained from drinking alcohol ( > 12 h) and heavy
physical activity ( > 48 h). Twelve millilitre samples were
drawn into siliconised evacuated tubes containing EDTA
(plasma analysis) or citrate (platelet analysis). Plasma was
separated by centrifugation (10006g for 15 min) within the
hour, and aliquots were isolated for the immediate analysis
of apoproteins A-1 and B, and for plasma C, HDL-C and
TG measurement. VLDL, LDL, IDL and HDL fractions
were isolated by ultracentrifugation (L7-55, Beckman
Instruments, Palo Alto, CA) of fresh plasma in a 50.4 Ti
rotor (Beckman Instruments, Palo Alto, CA).
VLDL was isolated by adjusting the density of 3 mL
plasma to 1.006 g=L and centrifugation for 16 h at
150 0006g. The tubes were sliced 30 mm from the
bottom. The top VLDL fraction and the bottom fraction
containing IDL, LDL and HDL were each brought to a total
volume of 5 mL. Likewise IDL and VLDL were isolated by
adjusting the density of 3 mL plasma to 1.019 g=L and
centrifugation for 16 h at 150 0006g (top fraction:
IDL ‡ VLDL; bottom fraction: LDL ‡ HDL) and HDL
was isolated by adjusting the density of 3 mL plasma to
1.063 g=L and centrifugation for 18 h at 170 0006g (top
fraction: VLDL ‡ IDL ‡ LDL; bottom fraction: HDL).
Enzymatic methods were used to assess concentrations
of cholesterol and triacylglycerols in plasma and lipoprotein subfractions of fresh plasma (Boehringer Mannheim
GmbH, Mannheim, Germany). Measurements were performed on a robotic analyser (Cobas Mira, Hoffmann la
Roche, Basle, Switzerland). Serum HDL-cholesterol was
enzymatically determined after the precipitation of apo B
containing lipoproteins by phosphor-wolfram acid-MgCl2
reagent. The concentration of serum C-reactive protein was
determined by an immunochemical method (Orion Diagnostica, Espoo, Finland). The concentrations of apolipoproteins A-1 and B were determined manually on fresh
plasma. Assay kits and reference standards were purchased
from Boehringer Mannheim (Mannheim, Germany). The
average coef®cient of variation (CV) for total C, HDL-C
and TG determinations was less than 2%. The mean CV for
apoprotein measurements was less than 3%.
The fatty acid pro®les of plasma phospholipids (PL) and
cholesterol esters (CE) were determined at the end of each
dietary period. The plasma lipids were extracted into
chloroform=methanol, separated by thin layer chromatography, isolated and methylated (transmethylation catalysed by
BF3) as previously described (Hùy & Hùlmer, 1988).
The fatty acid composition was determined using a
HP5880A gas chromatograph with split injection, FID
and a 30 m60.32 mm fused silica column with a 0.2 mm
®lm of SP2380 (Supelco Inc. Bellefonte, PA). The oven
temperature was programmed to rise at a rate of 2 C=min
from 140±160 C where it was maintained for 2 min before
rising again at 3 C=min to 200 C.
Platelets
Platelet-rich plasma (separated by centrifugation 2006g
for 10 min) was centrifuged at 10006g for 20 min. After
the removal of platelet-poor plasma, the platelet pellet was
washed twice with aqueous NaCl (0.15 mol=L). The platelets were stored at 7 40 C under methanol containing
0.01% BHT to avoid fatty acid oxidation. The lipids were
later extracted with chloroform=methanol (1 : 1 v=v) and
freed from protein and water by ®ltration through a layer of
anhydrous Na2SO4. After evaporating the solvent the lipid
was dissolved in chloroform=methanol (95 : 5) containing
0.005% BHT. Fatty acid methyl esters were prepared by
BF3 catalysed transmethylation and analysed by gas chromatography using a HP5890 gas chromatograph with a FID
and on-column injection. The column was a
50 m60.25 mm fused silica capillary coated with 0.2 mm
CP-Sil 88 (Chrompack, Middleburg, the Netherlands). The
temperature was programmed to rise from 90±160 C at a
rate of 40 C=min, and then to 200 C at 4 C=min, and
®nally after 10 min to 220 C at a rate of 4 C=min.
Statistical analysis
Statistical analyses were performed using Statgraphics
statistical software (Manugistics Inc., Rockville, MD) or
a validated spreadsheet template (Excel). Tests were carried out at the 5% level of signi®cance, unless otherwise
noted. A two tailed test with a signi®cance level of 1% was
used in tables of fatty acid compositions.
While the results are tabulated as the means of multiple
subjects, the actual statistical tests were carried out on
individual differences using Students t-test for paired
observations. Therefore, each subject served as his own
control and the variation between individuals was minimised to make the effects between diets more easily
detectable. The values shown in the tables are combined
data from all subjects because statistical analysis showed
no signi®cant effect (P < 0.05) of time or sequence in
feeding the diets.
Results
Subjects
Of the 15 selected participants, 13 completed the protocol,
and two people left the study for personal reasons. The
body mass of the participants did not change signi®cantly
throughout the experiment. Participants used checklists to
keep daily records of all food eaten extramurally and these
were discussed during informal interviews. The key data
are summarised in Table 2. Average cholesterol intake was
9% higher during the BG period compared to the BR
period. Other variables differed little. There was no indication of a preference for any one of the three test fats among
the subjects and an independent taste panel later found the
test fats to be equally palatable (unpublished results).
537
Blood lipids and milk fat based spreads
CC Becker et al
538
Table 2 Nutrient and energy content of the diets consumed during
experimental periodsa
BG
BR
BS
Energy (MJ)
13.1 2.2
12.9 1.8
13.2 2.3
Protein (g)
88.5 20.6
90.5 18.6 93.8 24.4
Fat (g)
104.5 15.8 105.5 15.8 107.4 16.6
Saturated fatty acids (g)
50.1 12.1
39.7 5.3
44.1 7.4
Monounsat. fatty acids (g)
30.6 10.2
37.0 5.4
34.2 5.6
Polyunsat. fatty acids (g)
14.7 4.4
15.9 3.0
15.3 2.5
Cholesterol (mg)
448 91
411 82
427 121
Carbohydrates (g)
412.7 88.2 400.5 79.5 408.6 90.1
Fibre (g)
30.8 6.2
29.5 4.6
28.8 7.5
Vitamin C (mg)
93.6 105.7 61.6 44.7 46.6 38.4
Calcium (mg)
1025 443
948 367 1045 411
Iron (mg)
13.7 2.3
13.7 3.1
14.0 3.0
Zinc (mg)
12.5 2.9
12.5 2.2
12.8 3.1
Protein (% of energy)
11 2
12 2
12 2
Fat (% of energy)
31 4
32 4
31 4
Carbohydrates (% of energy)
54 6
54 5
52 5
Alcohol (% of energy)
4 3
3 2
4 3
BG: 90% butter oil ‡ 10% grapeseed oil, n ˆ 13; BR: 65% butter
oil ‡ 35% rapeseed oil, n ˆ 12; BS: 50% butter oil ‡ 50% interesteri®ed
(70% rapeseed oil ‡ 30% fully hydrogenated rapeseed oil), n ˆ 13.
a
Calculated from weighed food records and analysis of test fats. Values
are means s.d.
Test fat composition
The total, as well as the regiospeci®c, compositions of the
test fats are given in Table 1. The concentration of stearic
acid in BS, was 1.9 times greater than in BG and 2.6 times
greater than in BR. BR contained more oleic and linolenic
acid than BG, due to the incorporation of rapeseed oil, but
was comparable to BS in that respect. BG had a greater
concentration of the saturated fatty acids with 16 or fewer
carbon atoms than the other test fats. The sn-2 position of
the test fats did not contain any of the characteristic short
and medium chain fatty acids found in dairy fats, but was,
as expected, enriched in myristic and palmitic acids. All the
test fats contained approximately 8.5% linoleic acid. More
than half of the linoleic acid (C18:2 n-6) and linolenic acid
(C18:2 n-3) of BR was concentrated in the sn-2 position,
while these were more evenly distributed in the other fats.
Plasma lipid and lipoprotein changes
Table 3 presents the concentrations of plasma C, plasma
TG, LDL-C (calculated according to Friedewald (Friedewald et al, 1972)), HDL-C (precipitated) and apoproteins
A-1 and B, as well as the concentrations of C and TG in
VLDL, LDL, IDL and HDL fractions isolated by ultracentrifugation. To convert cholesterol or triacylglycerol values
to mg=dL, multiply by 38.61 or 88.5.
A signi®cant (P < 0.009) decrease of 0.32 and
0.30 mmol=L (7%) occurred in total plasma cholesterol
concentrations (total C) when BG was replaced with BR
or BS. This reduction was almost entirely due to decreased
concentrations of LDL-C. Estimated (Friedewald et al,
1972) concentrations of LDL-C fell by 0.31 (12%,
P < 0.019) and 0.21 mmol=L (8%, P < 0.002), respectively,
while LDL-C isolated by ultracentrifugation was decreased
by 0.41 (16%, P < 0.014) and 0.27 mmol=L (11%,
P < 0.005). BG and BR gave similar HDL-C values,
which were 0.10±0.13 mmol=L greater than the HDL-C
concentration after BS (P < 0.045). These differences were
highly signi®cant for the results obtained by ultracentrifugation (P < 0.005). Hence, the total C=HDL-C ratio and the
LDL-C=HDL-C ratios determined either after precipitation
or ultracentrifugation were signi®cantly lower (8%, 15%
and 17% respectively) after BR.
The test fats resulted in apo A-1 concentrations which
were not signi®cantly different. BR and BS gave concentrations of apo B which were not signi®cantly different, but
were signi®cantly lower (9%, P < 0.05) than the apo B
concentration after BG. Following BR and BS intake, the
ratios of apo B to apo A-1 were not signi®cantly different
but were signi®cantly lower (P < 0.05) than the BG value.
Triacylglycerol concentrations were unaffected by the
experimental diets except for a decrease in LDL-TG after
BS compared to BG.
Fatty acid composition of plasma lipids and platelets
The fatty acid composition of the plasma PL and CE (Table
4) was only slightly affected by the differences between the
test fats. This might be due to dilution by the lipid that was
Table 3 Concentrations of cholesterol and triacylglycerol in plasma, VLDL, IDL, LDL, HDL and apoproteins A-1 and Ba
BG
Total cholesterol, mmol=L
VLDL cholesterol, mmol=Lb
LDL cholesterol (Friedewald) mmol=Lc
LDL cholesterol, mmol=Lb
IDL cholesterol, mmol=Lb
HDL cholesterol (precipitated) mmol=L
HDL cholesterol, mmol=Lb
Plasma triacylglycerol, mmol=L
VLDL triacylglycerol, mmol=Lb
LDL triacylglyerol, mmol=Lb
IDL triacylglycerol, mmol=Lb
HDL triacylglycerol (precipitated) mmol=L
HDL triacylglycerol, mmol=Lb
Apo A-1, mg=L
Apo B, mg=L
C=HDL-C (precipitated)
Apo B=Apo A-1
LDL-C=HDL-Cd
LDL-C=HDL-Cb
4.23 0.18
0.18 0.02
2.61 0.19
2.55 0.17
0.07 0.01
1.27 0.07
1.43 0.06
0.77 0.05
0.48 0.05
0.13 0.01
0.04 0.00
0.15 0.01
0.13 0.01
1374 22
1034 63
3.33 0.24
0.76 0.05
2.17 0.22
1.84 0.16
BR
BS
{{{
3.91 0.19
0.17 0.03
2.30 0.18{{
2.14 0.18{{
0.08 0.01{
1.28 0.05§
1.41 0.06§§§
0.73 0.06
0.42 0.06
0.12 0.01
0.04 0.00
0.16 0.01
0.13 0.01
1417 46
957 53{
3.06 0.18{{§§§
0.68 0.04{
1.83 0.16{{§§§
1.52 0.12{§§
3.93 0.15{{
0.22 0.03
2.40 0.13{{{
2.28 0.11{{{
0.14 0.02{{
1.18 0.07{{{{
1.30 0.06{{{{{{
0.77 0.05
0.50 0.05
0.11 0.01{
0.04 0.00
0.15 0.01
0.12 0.01
1364 32
930 44{
3.33 0.20{{{
0.69 0.04{
2.13 0.18{{{
1.81 0.12{{
BG: 90% butter oil ‡ 10% grapeseed oil; BR: 65% butter oil ‡ 35% rapeseed oil; BS: 50% butter oil ‡ 50% interesteri®ed (70% rapeseed oil ‡ 30% fully
hydrogenated rapeseed oil). aMean s.e.m. n ˆ 13. bIsolated by ultracentrifugation. cCalculated according to the formula: LDL-C ˆ TC 7 (HDLC ‡ Tot.TG=2.2). dLDL (Friedewald) and precipitated HDL. {Differs from BG. {(P < 0.05); {(P < 0.02); {{(P < 0.005). {Differs from BR. {(P < 0.05);
{{
(P < 0.02); {{{(P < 0.005). §Differs from BS. §(P < 0.05); §§(P < 0.02); §§§(P < 0.005).
Blood lipids and milk fat based spreads
CC Becker et al
Table 4
539
Major fatty acids in plasma phospholipid (PL) and cholesterol ester (CE). (wt%)a
PL
C14:0
C16:0
C16:1 n-7
C18:0
C18:1 n-9
C18:1 n-7
C18:2 n-6
C18:3 n-3
C20:3 n-6
C20:4 n-6
C20:5 n-3
C22:6 n-3
CE
BG
BR
BS
BG
BR
BS
0.6 0.1
26.1 0.2§
0.5 0.0
13.2 0.3
9.0 0.2{
1.4 0.0{
22.7 0.6
0.2 0.0{
0.5 0.1
8.7 0.3
1.4 0.2
4.7 0.3
0.7 0.1
25.7 0.2§
0.5 0.0
12.8 0.4
9.8 0.2{
1.6 0.1{§
21.8 0.5
0.4 0.0{
0.3 0.1
9.0 0.3
1.6 0.2
4.5 0.3
0.7 0.1
24.6 0.3{{
0.5 0.0
13.8 0.2
9.6 0.3
1.4 0.1{
22.5 0.4
0.4 0.0{
0.4 0.1
8.5 0.2
1.5 0.2
4.1 0.3
1.1 0.1
11.4 0.2§
2.2 0.2
1.7 0.1
16.2 0.4{
1.1 0.0{
49.4 0.8
0.7 0.0{§
0.7 0.0
5.2 0.2
0.9 0.2
0.6 0.1
1.0 0.0
11.4 0.2§
2.1 0.2
1.7 0.1
17.4 0.4{
1.3 0.0{§
47.8 0.9
1.1 0.0{
0.6 0.0
5.3 0.3
0.9 0.1
0.6 0.0
0.9 0.1
10.5 0.1{{
2.0 0.2
1.6 0.1
17.3 0.4
1.1 0.0{
48.9 0.9
1.1 0.0{
0.6 0.0
5.2 0.2
1.0 0.1
0.6 0.1
BG: 90% butter oil ‡ 10% grapeseed oil; BR: 65% butter oil ‡ 35% rapeseed oil; BS: 50% butter oil ‡ 50% interesteri®ed (70% rapeseed oil ‡ 30% fully
hydrogenated rapeseed oil). aMean s.e.m. n ˆ 13. {Signi®cantly different from BG (P < 0.01). {Signi®cantly different from BR (P < 0.01). §Signi®cantly
different from BS (P < 0.01).
in the self selected part of the diets. However, compliance
was con®rmed by the higher amount of linolenic acid in
both PL and CE after the rapeseed oil containing test fats.
Oleic and palmitic acid concentrations also varied in a
dose-dependent manner, with the lowest concentrations of
palmitic acid after BS (P < 0.003). The concentration of
oleic acid was higher after BR than BG with the value for
BS being intermediate and insigni®cantly different from the
two.
The fatty acid pro®le of the platelets (Table 5) largely
re¯ected the composition of the test fats. The concentration
of stearic acid was signi®cantly higher (P < 0.01) after BS
compared to the other test fats. The concentration of
palmitic acid, on the other hand was signi®cantly lower
(P < 0.01) after BS compared to the other test fats
(P < 0.004). After all diet periods the sum of palmitic and
stearic acid concentrations were between 31.7 and 32.3%,
which is not statistically different. The concentrations of
C18:3 n-3 and C18:1 n-9 were lower (P < 0.002) after BG
than after the other test fats, between which there was no
Table 5
C14:0
C16 A
C16:0
C16:1 n-7
C18:0 A
C18:1 A
C18:0
C18:1 n-9
C18:1 n-7
C18:2 n-6
C20:0
C18:3 n-3
C20:1 n-9
C20:3 n-6
C20:4 n-6
C20:5 n-3
C22:4 n-6
C22:5 n-6
C22:5 n-3
C22:6 n-3
Major fatty acids in platelet lipids. (wt%)a
BG
BR
BS
1.5 0.2
2.5 0.3
16.7 0.5§
0.8 0.1
3.6 0.4
1.0 0.3
15.1 0.4§
13.1 0.4{§
2.1 0.4§
5.8 0.1
0.8 0.1
0.2 0.0{§
0.7 0.0{
1.3 0.1
20.6 0.6
0.7 0.1
1.8 0.1
0.2 0.0§
1.5 0.1
2.1 0.1
1.3 0.2
2.0 0.4
16.5 0.5§
0.8 0.1
2.7 0.5§
1.2 0.3
15.4 0.5§
14.5 0.4{
2.2 0.4§
5.6 0.1
0.9 0.1
0.3 0.0{
0.9 0.0{§
1.2 0.1
20.8 0.4
0.7 0.0
2.0 0.2
0.2 0.0
1.5 0.1
2.0 0.1
1.3 0.1
2.2 0.3
15.3 0.3{{
0.7 0.0
4.1 0.5{
1.1 0.4
16.4 0.3{{
14.7 0.4{
1.2 0.3{{
5.8 0.1
0.9 0.0
0.3 0.0{
0.7 0.0{
1.3 0.0
21.4 0.2
0.8 0.1
1.8 0.1
0.1 0.0{
1.6 0.1
2.0 0.1
BG: 90% butter oil ‡ 10% grapeseed oil; BR: 65% butter oil ‡ 35%
rapeseed oil; BS: 50% butter oil ‡ 50% interesteri®ed (70% rapeseed
oil ‡ 30% fully hydrogenated rapeseed oil). aMean s.e.m. n ˆ 13.
{
Signi®cantly different from BG (P < 0.01). {Signi®cantly different from
BR (P < 0.01). §Signi®cantly different from BS (P < 0.01). A: Aldehyde.
difference. There was no signi®cant difference between the
effects of the experimental fats on the linoleic acid composition of the platelets, probably because all test fats
contained the same amount of C18:2 n-6.
Discussion
The subjects went through the study without signi®cant
changes in the intake of cholesterol, total fat or energy, or
any other micro or macro nutrient (Table 2). Each treatment
period lasted three weeks. Previous experiments have
shown that this duration is suf®cient to achieve steady
state (SandstroÈm et al, 1992). Plasma and platelet lipid
(Tables 4 and 5) composition varied with the diet, an
indication of compliance.
The principal result (Table 3) of the experiment was that
the rapeseed oil (oleic and linolenic acids) enriched fats
(BR and BS, respectively) appeared to be equally effective
in reducing plasma cholesterol and LDL cholesterol compared to BG. We observed a signi®cant (P < 0.009)
decrease of 7% in plasma cholesterol by replacing BG
with BR or BS. The reduction was almost entirely due to
decreased concentrations of LDL-C. BS contained 15%
more stearic acid than BR, and correspondingly less milk
fat. When no change in total C and LDL-C was seen, it may
indicate that stearic acid and butter oil fatty acids are
equally cholesterolemic. A separate effect of linoleic acid
was ruled out as this fatty acid was present in the same
concentration in all fats. The few studies (Singer et al,
1986; Abbey et al, 1990; Mantzioris et al, 1994) published
that compare linoleic and linolenic acid, report no difference in effect on lipoproteins.
A number of authors have carried out meta-analyses of
previously published studies and condensed these results
into equations obtained by multiple regression. Two of
these take stearic acid into consideration as an independent
variable (Derr et al, 1993; Yu et al, 1995) and predict
decreases in total C, and LDL-C concentrations of 0.14±
0.18 mmol=L when BG is replaced by BR or BS. We
observed reductions in total C and LDL-C of approx.
0.30 mmol=L and 0.35 mmol=L. This discrepancy between
the observed and the expected change in total C may be due
to the high concentration of myristic and lauric acids in
milk fat. LDL-C is derived from VLDL, and cleared from
the bloodstream mainly by the liver through a receptor
dependent mechanism. The rate of clearance depends on
Blood lipids and milk fat based spreads
CC Becker et al
540
the amount of cholesterol ingested, the type and amount of
fatty acids in the diet as well as the activity of the LDLreceptors. Myristic acid suppresses the removal of LDL
cholesterol from the bloodstream, while 2±7 E% (Hayes &
Khosla, 1992) linoleic acid has the opposite effect in a
dose-dependent way. Oleic and stearic acids appear neutral
in their effect on the LDL receptor (Woollett et al, 1992a,
1992b). The BR and BS diet contained about 5 E% linoleic
and linolenic acid. At the same time the dietary intake of
cholesterol was around 400 mg=d. This is not enough to
severely depress LDL receptor activity, since the subjects
still have low plasma C, but apparently enough to make
them sensitive to the kind of fatty acid present in diet. The
equations probably underestimated the effect of the BG to
BR transition as only few of the studies, upon which they
are based, used diets with appreciable amounts of butter,
which is a concentrated source of the most cholesterolemic
fatty acid: myristic acid (Derr et al, 1993; Yu et al, 1995).
BG and BR had very similar effect on HDL-C, and
resulted in values which were signi®cantly greater than the
HDL-C concentration after BS. This is in accordance with
results published earlier by others (Tholstrup et al, 1994;
Yu et al, 1995; Dougherty et al, 1995; Aro et al, 1997),
who also found that when stearic acid is exchanged for
C12±14 saturates, a moderate decrease in HDL-C was
observed.
Concentrations of arachidonic, eicosapentaenoic, and
docosahexaenoic acids did not change signi®cantly in
platelets or in plasma PL and CE after BR and BS, relative
to BG, even though the rapeseed oil containing diets
supplied linolenic acid. This observation could be
explained by the competition between the n-3 and n-6
families for desaturases. All diets supplied substantial
amounts of linoleic acid, to which was added the contribution of the back-ground diet. The conversion of linolenic
acid into long chain metabolites is strongly dependent on
the concentration of linoleic acid (Emken et al, 1994).
In platelets the sum of palmitic and stearic acids
remained constant, suggesting that one was substituted
for the other in cell membranes. Replacing butter fatty
acids (15% of test fat) with stearic acid had little in¯uence
on the fatty acid compositions of platelets and plasma PL
and CE. Possibly, this is because of preferential incorporation of stearic acid into other fatty acid pools, such as
adipose tissue. But stearic acid could also be desaturated to
oleic acid, chain-shortened or fully catabolised. There has
been some discussion regarding the rate of absorption of
stearic acid from the intestine, relative to other fatty acids
with lower melting points. BS contains 19% stearic acid,
but due to interesti®cation the amount of tristearin present
would be negligible, and the absorption of stearic acid from
mixed triglycerides, is comparable to that of other fatty
acids (Tholstrup et al, 1996; Haumann, 1998).
The fatty acids in the sn-2 position are mostly retained
during absorption of dietary fat (Mattson & Volpenhein,
1964) and may exert a more direct in¯uence on cholesterol
and lipoprotein metabolism than the fatty acids esteri®ed to
the 1,3 position of the glycerol, which may undergo betaoxidation in the enterocytes, or be incorporated into other
lipid classes, and are less likely to reach the liver. Therefore, one could hypothesise that regiospeci®c structure
would in¯uence the effect of the triacylglycerols independently of fatty acid composition. In BS the polyunsaturated
fatty acids are evenly distributed across the three positions
of the glycerol backbone, while in BR half of these fatty
acids are concentrated in the sn-2 position. Christophe et al
(1978) reported that feeding six subjects 85 g of interesteri®ed butter oil, with a lower palmitic acid concentration in
the sn-2 position, caused a decrease in plasma cholesterol,
compared to native butter oil. In a number of other experiments, however, interesteri®cation of cocoa butter (Grande
et al, 1970) and high-palm oil blend (Zock et al, 1995;
Nestel et al, 1995) did not elicit a different cholesterolemic
response compared to the native oil. Previously we have
shown that BR and interesteri®ed BR elicited the same
response in a similar group of subjects (to be published).
Therefore, we believe that structure did not play an important role in the response to the dietary challenges in this
study although it was not addressed speci®cally.
Conclusions
Dairy spreads enriched in oleic and linolenic acids (BR) or
additionally enriched in stearic acid (BS) gave signi®cant
improvements compared to BG, when predicted cardiovascular risk is considered, as they gave signi®cantly lower
plasma C, LDL-C, apo B, and apo B=apo A-1 values.
However, if the ratio of plasma C to HDL-C or LDLC=HDL-C, are considered as well, adding stearic acid to
BR did not further improve the nutritional value of the
blend. On the contrary BR gave higher HDL-C values than
BS.
The 7% reduction in plasma C after BR compared to BG
is noteworthy, as it is commonly assumed that one percent
decrease in plasma cholesterol leads to a two percent
decrease in CDH risk (Lipid Research Clinics Program,
1984). This demonstrates the great potential that moderate
changes of a component of the diet may have on indices for
heart disease risk. It is likely that the effect would have
been even greater in hypercholesterolemic subjects, who
are more sensitive to the dairy fat than healthy young men.
Acknowledgements ÐThis study was a collaboration between the Center
for Food Research, The Technical University of Denmark, and the Danish
Dairy Research Foundation and ®nanced as a part of the Danish Food
Research Programme (FéTEK I). We thank IL Grùnfeldt and the staff of
the metabolic kitchen and gratefully acknowledge the laboratory assistance
of technicians K JoÈrgensen, J Bidstrup, and M Rosenberg. We are grateful
to L Bandholm of Aarhus Olie A=S (Aarhus, Denmark) who kindly
delivered the test fats.
References
Abbey M, Clifton PM, Kestin M, Belling GB & Nestel PJ (1990): Effect of
®sh oil on lipoproteins, lecithin : cholesterol, acyltransferase, and lipid
transfer protein activity in humans. Atherosclerosis 10, 85±94.
Ahrens EH, Hirsch J, Insull W, Tsaltas TT, Blomstrand R & Peterson ML
(1957): The in¯uence of dietary fats on serum-lipid levels in man.
Lancet I, 943±953.
Aro A, Jauhiainen M, Partanen R, Salminen I & Mutanen M (1997):
Stearic acid, trans fatty acids, and dairy fat: effects on serum
and lipoprotein lipids, apolipoproteins, lipoprotein(a), and lipid
transfer proteins in healthy subjects. Am. J. Clin. Nutr. 65, 1419±
1426.
Aviram M & Eias K (1993): Dietary olive oil reduces low-density lipoprotein
uptake by macrophages and decreases the susceptibility of the lipoprotein
to undergo lipid peroxidation. Ann. Nutr. Metab. 37, 75±84.
Becker CC, Hùlmer G & Rosenquist A (1993): Regiospeci®c analysis of
triacylglycerols using allyl magnesium bromide. Lipids 28, 147±149.
Bonanome A & Grundy SM (1989): Intestinal absorption of stearic acid
after a consumption of high fat meals in humans. J. Nutr. 119, 1556±
1560.
Christofferson SW & Glass RL (1969): Preparation of milk fat methyl
esters by alcoholysis in an essentially nonalcoholic solution. J. Dairy
Blood lipids and milk fat based spreads
CC Becker et al
Sci. 52, 1289±1290.
Christophe A, Matthys F, Geers R & Verdonk G (1978): Nutritional
studies with randomized butter. Arch. Int. Physiol. Biochem. 83, 413±
415.
Davis CE, Rifkind BM, Brenner H & Gordon DJ (1990): A single
cholesterol measurement underestimates the risk of coronary heart
disease. J. Am. Med. Assoc. 264, 3044±3046.
Derr J, Kris-Etherton PM, Pearson TA & Seligson FH (1993): The role of
fatty acid saturation on plasma lipids, lipoproteins and apolipoproteins:
II. The plasma total and low density lipoprotein cholesterol response of
individual fatty acids. Metabolism 42, 130±134.
Dougherty RM, Allman MA & Iacono JM (1995): Effects of diets
containing high or low amounts of stearic acid on plasma lipoprotein
fractions and fecal fatty acid secretion of men. Am. J. Clin. Nutr. 61,
1120±1128.
Emken EA, Adlof RA & Gulley RM (1994): Dietary linoleic acid
in¯uences desaturation and acylation of deuterium labelled linoleic
and linolenic acids in young adult males. Biochim. Biophys. Acta. 1213,
277±288.
Friedewald WT, Levy RI & Fredrickson DS (1972): Estimation of the
concentration of low-density lipoprotein cholesterol in plasma, without
use of the preparative ultracentrifuge. Clin. Chem. 18, 499±502.
Grande F, Anderson JT & Keys A (1970): Comparison of effects of
palmitic and stearic acids in the diets on serum cholesterol in man. Am.
J. Clin. Nutr. 23, 1184±1193.
Haumann BF (1998): Stearic acid: a ``different'' saturated fatty acid.
Inform 9, 202±208.
Hayes KC & Khosla P (1992): Dietary fatty acids and cholesterolemia.
FASEB 6, 2600±2607.
Hegsted DM, McGandy RB, Myers ML & Stare FJ (1965): Quantitative
effects of dietary fat on serum cholesterol in man. Am. J. Clin. Nutr. 17,
281±295.
Hill JO, Peters JC, Swift LL & Yang D (1990): Changes in blood lipids
during six days of overfeeding with medium or long chain triglycerides.
J. Lipid Res. 31, 407±416.
Howell WH, McNamara DJ, Tosca MA, Smith BT & Gaines JA (1997):
Plasma lipid and lipoprotein responses to dietary fat and cholesterol: a
meta-analysis. Am. J. Clin. Nutr. 65, 1747±1764.
Hùy C-E & Hùlmer G (1988): Dietary linoleic acid and the fatty acid
pro®les in rats fed partially hydrogenated marine oils. Lipids 23, 973±
980.
Kessler A, GruÈnert C & Wood WG (1994): The limitations and usefulness
of C-reactive protein and elastase-a1-proteinase inhibitor complexes as
analytes in the diagnosis and follow-up of sepsis in newborns and
adults. Eur. J. Clin. Chem. Clin. Biochem. 32, 365±368.
Keys A, Anderson JT & Grande F (1965): Serum cholesterol response to
changes in the diet. IV. Particular saturated fatty acids in the diet.
Metabolism 14, 776±787.
Kris-Etherton PM, Derr J, Mitchell DC, Mustad VA, Russell ME,
McDonnell ET, Salabsky D & Pearson TA (1993): The role of fatty
acid saturation on plasma lipids, lipoproteins, and apolipoproteins: I.
Effects of whole food diets high in cocoa butter, olive oil, soybean oil,
dairy butter, and milk chocolate on the plasma lipids of young men.
Metabolism 42, 915±917.
Lipid Research Clinics Program (1984): The lipid research clinics coronary primary prevention trial results II: the relationship of reduction in
incidence of coronary heart disease to cholesterol lowering. J. Am. Med.
Assoc. 251, 365±374.
Mantzioris E, James MJ, Gibson RA & Cleland LG (1994): Dietary
substitution with an a-linolenic acid-rich vegetable oil increases eicosapentaenoic acid concentration in tissues. Am. J. Clin. Nutr. 59, 1304±
1309.
Mattson FH & Volpenhein RA (1964): The digestion and absorption of
triglycerides. J. Biol. Chem. 239, 2772±2777.
Mensink RP & Katan MB (1992): Effect of dietary fatty acids on serum
lipids and lipoproteinsÐa meta-analysis of 27 trials. Arterioscler.
Thromb. 12, 911±919.
Mùller A (1989): Food Composition Tables 1989. National Food Agency
of Denmark, Sùborg, Denmark.
Nestel P, Noakes M, Belling GB, McArthur R & Clifton PM (1995): Effect
on plasma lipids of interesterifying a mix of edible oils. Am. J. Clin.
Nutr. 62, 950±955.
Pai T & Yeh Y-Y (1997): Desaturation of stearate is insuf®cient to increase
the concentrations of oleate in cultured rat hepatocytes. J. Nutr. 127,
753±757.
Renaud S (1990): Linoleic acid, platelet aggregation and myocardial
infarction. Atherosclerosis 80, 255±256.
Renaud S, Backer DEG, Thevenon C, Joossens JV, Vermylen J, Kornitzer
M & Verstraetes M (1991): Platelet fatty acids and function in two
distinct regions of Belgium: Relationship to age and dietary habits. J.
Intern. Med. 229, 79±88.
Rhee SK, Kayani AJ, Ciszek A & Brenna JT (1997): Desaturation and
interconversion of dietary stearic and palmitic acids in human plasma
and lipoproteins. Am. J. Clin. Nutr. 65, 451±458.
SandstroÈm B, Marckmann P & Bindslev N (1992): An eight-month controlled
study of a low-fat high-®bre diet: effects on blood lipids and blood pressure
in healthy young subjects. Eur. J. Clin. Nutr. 46, 95±109.
Singer P, Berger I, Wirth M, Godicke W, Jaeger W & Voigt S (1986):
Slow desaturation and elongation of linoleic and a-linolenic acid as a
rationale of eicosapentaenoic acid-rich diet to lower blood pressure and
serum lipids in normal, hypertensive and hyperlipidemic subjects.
Prostaglandins Leukot. Med. 24, 173±193.
Tholstrup T, Andreasen K & SandstroÈm B (1996): Acute effect of high-fat
meals rich in either stearic or myristic acid on hemostatic factors in
healthy young men. Am. J. Clin. Nutr. 64, 168±176.
Tholstrup T, Marckmann P, Jespersen J & SandstroÈm B (1994): Fat high in
stearic acid favorably affects blood lipids and factor VII coagulant
activity in comparison with fats high in palmitic acid or high in myristic
acid and lauric acids. Am. J. Clin. Nutr. 59, 371±377.
Wardlaw GM & Snook JT (1990): Effect of diets high in butter, corn oil,
or high-oleic acid sun¯ower oil on serum lipids and apolipoproteins in
men. Am. J. Clin. Nutr. 51, 815±821.
Woolett LA, Spady DK & Dietschy JM (1992a): Regulatory effects of the
saturated fatty acids 6 : 0 through 18 : 0 on hepatic low density lipoprotein receptor activity in the hamster. J. Clin. Invest. 89, 1131±1141.
Woollett LA, Spady DK & Dietschy JM (1992b): Saturated and unsaturated fatty acids independently regulated low density lipoprotein receptor activity and production rate. J. Lipid Res. 33, 77±88.
Yu S, Derr J, Etherton TD & Kris-Etherton PM (1995): Plasma cholesterol-predictive equations demonstrate that stearic acid is neutral and
monounsaturated fatty acids are hypocholesterolemic. Am. J. Clin. Nutr.
61, 1129±1139.
Zock PL, de Vries JHM, de Fouw NJ & Katan MB (1995): Positional
distribution of fatty acids in dietary triglycerides: effects on fasting
blood lipoprotein concentrations in humans. Am. J. Clin. Nutr. 61, 48±55.
541