You are on page 1of 150

HEALTH ASPECTS OF DIETARY TRANS FATTY ACIDS

August 1985
Prepared for
CENTER FOR FOOD SAFETY AND APPLIED NUTRITION
FOOD AND DRUG ADMINISTRATION
DEPARTMENT OF HEALTH AND HUMAN SERVICES
WASHINGTON, D.C. 20204
under
Contract Number FDA 223-83-2020
edi ted by
Frederic R. Senti, Ph.D.
(
LIFE SCIENCES RESEARCH OFFICE
FEDERATION OF AMERICAN SOCIETIES
FOR EXPERIMENTAL BIOLOGY
9650 Rockville Pike
Bethesda, Maryland 20814
,-------
u.s. ~ ~ : ~ ~ ~ ~ ~ ~ ~ : ~ ~ e r c e tfl
National Technicallnfonnation Servil;e
Springfield, Virginia 22161
'-------__-l
,
Health Aspedtsof Dietary
Trans Fatty Acids
r
BIBL.IOGRAPHIC DATA
SHEET
4. Tide and Subtitle
1
1. Report No.
FDA/ CFSAN-8 6/ 1
3. Recipient's Accession No.
PBB h 119 03 9 I At:,.
5. Report Date
August 1985
6.
7. Author(s)
ad hoc Review Panel on Trans Fatty Acids
9. PerJg.rmil).S Qrganizatiq.(l Name an,d
L1Te )ClenCeS KeSearCn OfTlce
Federation of American Societies for
Experimental Biology
9650 Rockville Pike, Bethesda MD 20814
12. Sponsoring Organization Name and Address
Center for Food Safety and Applied Nutrition
Food and Drug Administration
Department'of Health &Human Services, Washington, DC
15. Supplementary Note s
16. Abstracts
20204
B. Performing Organization Rept.'
No.
10. Project/Task/Work Unit No.
Task Order #2
11. Contract/Grant No.
223-83-2020
13. Type of Report & Period
Covered
Technical Report
14.
This report of an ad hoc Review Panel examines the available information on the
health aspects of dietary trans fatty acids. The report concludes that the available
scientific information suggests reason for concern with the. safety of dietary
trans fatty acids at present levels of and at future expected consumption
levels. The report suggests areas of research to clarify several unanswered question
(
17. Key Words and Document Analysis, 17a. Descriptors
Dietary Fat
Hydrogenated Oils
Diet
Lipids
Triglycerides
Identifiers/Open-Ended Terms
bo1ism
Depos it i on
Carcinogenicity
Atherosclerosis
Toxicology
.:;r
COSATI Fie ld/Group (
.
. . "
,.'
.'
'0."''':'', :
.. " .' ;,-..
-
I
lB. Availability Statement
Approved for public release; distribution
unlimited.
FORM NTI5-3S (REV. 1073) ENDORSED BY ANSI AND UNESCO.
19. Security Class (This
Report)
'lJNrl A
20. Security Class (This
Page
UNCLASSIFIED
THIS FORM MAY BE REPRODUCED
21. 'No. of Pages
Ilf"
22. Price
USCOMM DC 8265P74
FOREWORD
The Life Sciences Research Office (LSRO), Federation
of American Societies for Experimental Biology (FASEB), provides
scientific assessments of topics in biomedical sciences. Reports
are based on comprehensive literature reviews and the scientific
opinions of knowledgeable investigators engaged in work in relevant
areas of biology and medicine.
This report was developed for the Center for Food
Safety and Applied Nutrition, Food and Drug Administration (FDA)
in accordance with the provisions of Task Order #2 of Contract
No.- 223-83-2020. The Report was prepared and edited by Frederic
R. Senti, Ph.D., Senior Scientific Consultant, LSRO, FASEB, under
the direction of an ad hoc Review Panel on Trans Fatty Acids.
Scientists selected as members of the Panel were chosen for their
qualifications, experience, and jUdgment with due consideration
for balanc& and breadth in the appropriate professional disciplines.
Members of the Panel and others who assisted in preparation of the
report are identified in Section VIII.
As announced by FASEB and by FDA in the Federal Register,
the ad hoc Panel met on February 21, 22, 1985 and provided an
opportunity for any interested person to appear before the ad hoc
Review Panel at an open meeting on February 21, 1985 to make oral
presentations of data, information, and views on trans fatty acids
as food ingredients. Individuals who made presentations are iden-
tified in Section IX. The data, information, and views presented
at the open meeting were considered by the ad hoc Panel in reaching
its concllisions.
The ad hoc Panel's evaluation of available information
and data was made independently of FDA and any other group,
governmental or nongovernmental. The ad hoc Panel members and
LSRO accept responsibility for the accuracy of the report; how-
ever, listing of these individuals in Section VIII does not imply
that they specifically endorse each study conclusion. This report
was reviewed and approved by the LSRO Advisory Committee (which
consists of representatives of each constituent Society of FASEB)
under authority delegated by the Executive Committee of the
Federation Board. Upon completion of these review procedures,
the report was approved and transmitted to FDA by the Executive
Director, FASEB.
While this is a report of the
Societies for Experimental Biology, it
reflect the opinion of each individual
constituent Societies.
Federation of American
does not necessarily
member of the FASEB
Preceding Page Blank
iii
,
SUMMARY
This report provides a review of available information on
the health aspects of trans fatty acids and presents conclusions
reached by an ad hoc Review Panel. As estimated by the Review
Panel from publ-ished information, and independently by the edible
oil processing industry, trans fatty acid availability in the U.S.
food supply has been essentially constant at about 8 g per capita
daily, or about 5.6% of the fat in the food supply, over the past
- 2 decades. Based on fat consumption of individuals reported in
the 1977-78 USDA Nationwide Food Consumption Survey, and assuming
5.6% trans fatty acids in the fat consumed, maximum trans isomer
consumption by sex/age group was 6.5 g/d for males aged 15-18
years and 4.5 g/d for females aged 12-14 years. Maximum trans
fatty acid consumption reported in the literature was 13.7 g/d
for lactating women fed diets in which hydrogenated fat products
were used in the preparation of all meals.
Continuation at present revels or a modest decrease in
both total fat and trans fatty acid consumption is predicted for
the next 5 to 10 years.
Animal studies show that there is no discrimination
against the absorption of elaidic and linoelaidic acid isomers
and that the rate and extent of oxidation of these isomers to C02
are essentially the same as those of their cis isomer counterparts.
Dietary trans fatty acids are incorporated into most tissues of
experimental animals and humans but not at the same levels. Brain
tissue and fetal tissues appear resistant to incorporation. Among
lipid classes, only triglycerides in different tissues have similar
concentrations of trans isomers, with the distribution of trans
isomers similar to that in the diet. Certain positional octa-
decenoate isomers, both cis and trans, accumulate in the phospho-
lipids of most tissues of:rats fed partially hydrogenated fats.
Rat tissues also discriminate against the incorporation of certain
trans and cis positional isomers. The distribution of trans
positional octadecenoate isomers in the lipids of human tissues
taken at autopsy is similar but not identical to the average
distribution in commercial hydrogenated fats. The comparison
also indicates that there are similarities in the accumulation
of positional isomers in rat and human tissues but it appears that
accumulation is less in human tissues and that incorporation of
trans isomers in the phospholipids of human tissues is generally
less than their relative percentages in the diet.
Dietary trans acids perturb the metabolism of linoleic
acids in rats, the magnitude of the perturbation decreasing as the
ratio of trans isomer to linoleate in the diet decreases. At the
ratio present in the average U.S. diet, the metabolic perturbation
in rats had no observed adverse effect. In vitro studies with rat
liver microsomes show that several of the ~ 3 to ~ 1 6 trans and cis
Preceding Page Blank ,
v
octadecenoate isomers inhibit the desaturation of linoleic and
eicosanoic acids. Positional isomers of trans octadecenoic acid
also are desaturated by liver microsomes in vitro to cis, trans-
and cis, cis acids having unusual structures which,
in vivo, might serve as precursors for eicosanoids of unusual
structure and properties. However, no evidence for such unusual
structures or adverse effects that might be associated with their
formation has been reported in animal studies with hydrogenated
vegetable oils. Also, no eicosenoic acid isomers have been found
in human tissues at autopsy nor have deuterium-labeled octadeca-
dienoic or eicostetrenoic positional isomers been found in blood
lipids after the administration of deuterium-labeled trans or cis
positional isomers of octadecenoic acid.
Long-term and mUltigeneration rodent and rabbit feeding
studies with hydrogenated vegetable oils have not shown carcino-
genicity or other histopathological effects attributable to trans
fatty acids. Teratogenicity tests have revealed no adverse effects.
Dietary elaidic acid and partially hydrogenated vegetable oils are
cholesterolemic but not atherogenic in rabbits, swine, and monkeys.
Human studies indicate elaidic acid or partially hydrogenated
vegetable oil are no more, or 'little more, cholesterolemic than
oleic acid. No conclusions can be made with respect to the
effects of dietary trans fatty acids on immune function.
Gas-liquid chromatographic (GLC) methods are the most
accurate methods for the determination of trans isomer content
of fats, oils and lipid extracts. Argentation thin layer chroma-
tography is an effective adjunct to GLC in the analysis of the
distribution of geometrical positional isomers by preseparation
into the cis and trans 'fractions. Certain positional isomers not
resolved by GLC can be separated by reverse phase high performance
liquid chromatography (HPLC) and relatively complete separation
of the positional isomers in partially hydrogenated fats has
been reported using argentation HPLC. Ozonolysis continues to
be the most definitive method for the identification of positional
isomers. Limited studies indicate natural abundance 13C NMR is
an effective and rapid method for determination of the trans
content of intact triglycerides.
The report includes suggestions for future research
with geometrical and positional fatty acid isomers.
vi
TABLE OF CONTENTS
Page
Foreword iii
Summary
I. Introduction.
A.
B.
Scope of the Study
Background

v

1
.

1

1
1.
2.
3.
4.
Formation of isomeric fatty acids in
hydrogenation processes
Formulation of margarines . .
Formulation of shortenings .
Salad and cooking oils
1
6
9
9
II. Analytical Methodology . 11
A.
B.
C.
D.
E.
IR Absorption Spectrometry
UV Absorption Spectrometry
Mass Spectrometry
Nuclear Magnetic Resonance
Chromatographic Methods
;F
11
12
12
14
16
1. Adsorption chromatography . . 16
2. Argentation chromatography 16
3. Partition chromatography. . 17
4. High performance liquid chromatography 19
F.
G.
Chemical Methods
Enzymatic Methods
20
20
I I I. Consumer Exposure Data 23
A.
B.
C.
D.
Beef and Milk Fat
Margarines
Shortenings
Salad and Cooking Oils
vii
23
25
25
25
1. Composition of tissues obtained
at autopsy . . . . . . . 52
2. In vivo studies . . . .. . 58
I. Atherosclerosis . . . 60
1. Epidemiological studies . . . . 60
2. Dietary studies . . . . . . . . 61
3. Animal studies . . . . 65
viii
J. Competitive Inhibition of Linoleic Acid
Metabolism by Isomeric Unsaturated Acids
Page
67
1. Inhibition of the 65- and 66-desaturase
enzymes . 67
2. Desaturation of trans-18:1 isomers. 68
K. Action of Enzymes In Vitro 68
d
.
1. S-oxidation 68
L.
M.
Trans Fatty Acids and Immune Function
l. Modulation of the lymphocyte response
in vitro by exogenous trans fatty
acids . . . .

.

2. Modulation of macrophage function
in vitro by exogenous fatty acids

3. Modulation of immune status by dietary
trans fatty acids

Effects of Isomeric Fats on Membrane
Properties and Cell Function
69
70
72
72
73
V. Conclusions 77
A.
B.
C.
D.
E.
Consumer Exposure
Toxicological Effects
Metabolism
1. Absorption and catabolism

2. Deposition in tissues

3. Atherosclerosis

. .

4. Interference of 18:1 and 18:2 isomers
with linoleate metabolism

5. Immune function

.

Analytical Methods
General .
77
78
78
78
79 .
80
80
81
81
83
VI.
VII.
VIII.
Suggestions for Future Consideration
Literature Cited.
Study Participants
85
87
109
IX. Open Meeting Participants
Appendi'l:
ix
. . . . . . . 111
. 113
A.
I. INTRODUCTION
SCOPE OF THE STUDY
The Food and Drug Administration (FDA), in requesting
this study on the health aspects of trans fatty acids as food
ingredients, asked that scientific information and data on three
broad issues be reviewed: (a) the level of trans fatty acids in
the U.S. food supply and the consumption of these isomers on a per
capita basis; (b) the toxicological effects and the nutritional
biochemistry and physiology of these substances; and (c) methods
for measuring trans fatty acids.
In regard to the first issue, FDA asked that three
specific questions be addressed: (a) what information is avail-
able about the level of trans fatty acids currently in the U.S.
food supply and on the levels of the past 10 to 20 years? (b)
what trends can be predicted about levels of trans fatty acids
in the U.S. food supply over the next 5 to 10 years? (c) what
estimates can be made regarding the U.S. daily dietary intake
of trans fatty acids?
In addressing the second issue, FDA requested that
information on the following questions be included: (a) what
are the pathways in the metabolism of trans fatty acids? (b) in
what sites are trans fatty acids deposited in the lipid classes
of various tissues? (c) do relatively high dietary levels of
trans fatty acids interfere with the absorption and metabolism
of the counterpart cis fatty acids? (d) what information is
available concerning the effects of trans fatty acids on the
immune system?
In further defining the scope of the third issue, it
was requested that the underlying principles and effectiveness
of analytical methods available for measuring trans fatty acids
of fats, oils, and foods be discussed.
These and related questions were addressed by the" ad hoc
Review Panel and are discussed in this report. A listing of the
Panel members is given in Section VIII, p . l 1 1 ~ "
B. BACKGROUND
1. Formation of isomeric fatty acids in hydrogenation
processes
Trans fatty acids are common constituents of the
American diet, both as components of commercially hydrogenated
vegetable oils used in the formulation of margarines, shortenings,
salad and cooking oils, and in ruminant fats including milk fat
1
and the fats consumed as a part of beef and lamb meats. Trans
fatty acids in ruminant fats are formed primarily by the bio-
hydrogenation of ingested linoleic and linolenic acids by micro-
organisms in the rumen.
The principal unsaturated fatty acids in domestic
vegetable oils that are substrates for hydrogenation are oleic,
linoleic, and linolenic. All are 18 carbon chain compounds. The
cis configuration of the double bond in oleic acid, also designated
9c-18:1, and the trans configuration of the double bond after
isomerization to elaidic acid (9t-18:1) are shown in Figure 1.
In the cis configuration, hydrogen atoms are on the same side of
the double bond; in the trans configuration, the hydrogen atoms
are on opposite sides. The cis double bonds in oleic, linoleic,
and linolenic acids introduce bends in the configuration of the
extended chain, as contrasted with trans double bonds which
produce a linear extended chain simil.r to that of stearic acid.
The packing of the linear extended trans chains in the crystalline
state results in a higher melting temperature than that of the
crystalline cis isomer.
In the hydrogenation process, hydrogen atoms are added
to double bonds of the fatty acids in the presence of a nickel
catalyst. Other reactions also occur. Some double bonds migrate
from their original positions in the fatty acid chain (positional
isomerism) and many of these are changed from the cis to the trans
configuration (geometrical isomerism). As many a s ~ trans and
cis positional isomers in 18:1 fatty acids have been reported in
partially hydrogenated vegetable oils (Carpenter and Slover, 1973;
Sampugna et al., 1982; Scholfield et al., 1967; Slover et al.,
1985). A similar number is found in butter and beef fat (Parodi,
1976). The major trans isomer in ruminant fats is vaccenic acid,
11t-18:1.
Double bond positions in the trans-18:1 isomers in
hydrogenated vegetable oils are reported to range from carbon
atoms 6 to 16 with most of the trans double bonds at carbon
atoms 9 to 12 (Carpenter and Slover, 1973; Sampugna et al.,
1982; Scholfield et al., 1967). Double bonds in the cis-18:1
isomers are also found at carbon atoms 6 to 16 but mosr-cis
double bonds are at positions 9 to 12 with a relatively large
proportion remaining in the original position at carbon atom 9.
Bar graphs of the distribution of trans- and cis-18:1 isomers
by double bond position in a commercially hydrogenated fat are
presented in Figures 2 and 3, respectively (Sampugna et al.,
1982).
Although this report is concerned only with the health
effects of trans fatty acids, the Review Panel recognized that the
properties of the cis fatty acid isomers also should be considered.
Emken (1981) estimated that the daily per capita consumption of
cis-18:1 isomers (other than 9c-18:1) was about one-third that
2
Oleic acids, cis
Elaidic acid, trans
Figure 1. Cis and trans isomers of octadecenoic acid, 9-18:1.
3
trans
25
--
Q)
c::
20
Q)
0
c::
0
e
~ I
15
~
10
0
fiil.
5
t.6 1:.7 1:.8 1:.9 t.l0 tJ.l lU2 tJ.3 tJ.4 t.15 tJ.6
double bond position
Figure 2. Double bond positions in trans monoene fractions of a
commercially hydrogenated fat product (Sampugna et al., 1982).
4
70
cis
60
Ql
C
Ql
0
50
c
0
8
~ I
40
'1-4
0
30
~
20
10
Do6 ~ 7 DoS Do9 alO Doll Do12 tl,13 614 Do15 Do16
double bond position
Figure 3. Double bond positions in cis monoene fractions of a
commercially hydrogenated fat producr-(Sampugna et al., 1982).
5
of trans-18:1 isomers. The data of Slover et ale (1985) indicate
that the ratio of trans-18:1 to cis-18:1 (other than 9c-18:1)
isomers in margarines is about The biological properties
of the cis isomers have been demonstrated to differ from those
of 9c-18:1 (Emken, 1983; Holman et al., 1983). Thus, in inter-
preting results of studies in which hydrogenated fats have been
fed, consideration must be given to possible effects of cis as
well as trans fatty acids.
Geometrical and positional isomers in 18:2 fatty acids
are also formed in the hydrogenation of vegetable oils. Three
geometrical isomers are possible: the cis, trans; trans, cis;
and the trans, trans. The principal positional isomers present
in commercially hydrogenated products have double bonds at carbon
atoms 9 and 12 positions). Conjugated dienes, e.g"., 9c,11t-
and may be formed in the hydrogenation of linoleate
acid but these isomers are relatively highly reactive and residual
levels in commercial hydrogenated oils are small (Gunstone and
Norris, 1983). The proportion of 18:2 isomers present in commer-
cial hydrogenated products is much less than that of the trans-
18:1 isomers (Lanza and Slover, 1981; Slover and Lanza, 1979).
In the hydrogenation of vegetable oils, reduction of
improves flavor stability and, at higher degrees
of hydrogenation, the trans double bonds formed increase the
melting range and impart plastic properties desired in the manu-
facture of margarine. Selectivity of hydrogenation for the more
rapid reduction of linolenate to linoleate relative to the rate
of reduction of linoleate to oleate, or oleate to stearate, depends
upon hydrogenation conditions, particularly with respect to the
catalyst and its concentration, pressure of hydrogen, temperature,
and agitation (Dutton, 1979; Weiss, 1983). In the production of
salad and cooking oils from soybean oil, it is considered desirable
to hydrogenate under selective conditions to reduce linolenate, a
flavor unstable component, to linoleate with minimal reduction of
linoleate and oleate.
2. Formulation of margarines
In order to obtain the desired plasticity and fatty
acid composition, margarines are formulated from one or more vege-
table oils hydrogenated to different degrees, or from blends of
hydrogenated and unhydrogenated oils (Brekke, 1980; Latondress,
1981; Weiss, 1983; Wiedermann, 1978). A stick (hard) margarine
(quarter pound units packaged in 1 pound cartons) over
a wide temperature range may be prepared by blending soybean oils
hydrogenated to two or three different degrees, thereby giving
components that melt over a range of temperatures. Margarines
with higher percentages of polyunsaturated fatty acids are
formulated from blends of unhydrogenated and hydrogenated oils.
Tub margarines, softer than stick margarines, may contain 75-85%
6
unhydrogenated oil. Another type of tub margarine is
formulated with 80% lightly hydrogenated soybean oil and
20% highly hydrogenated soybean oil. Both stick and tub
margarines are marketed in whipped form.
Federal identity standards (Office of the Federal
Register, 1984) require that products labeled as margarine
contain no less than 80% fat. There are no identity standards
for lower calorie products marketed as imitation margarines
or spreads; the fat content of these products is not fixed
but the former usually contain 40% fat and the latter about
60% fat (Brekke, 1980). Spreads are formulated with the oils
or oil combinations used either in stick or tub margarines;
tub margarine oils are generally used in formulating imitation
margarines (Applewhite, 1985).
Slover et al. (1985) reported 18 different blends. of
soybean, corn, cottonseed, safflower, sunflowerseed, and palm
oil, hydrogenated or unhydrogenated, in the labeled composition
of 40 brands of stick and tub margarines andmargarine-like
products that they analyzed. Among 40 margarines studied by
Enig et al. (1983), there were nine additional blends.
The trans fatty acid composition of 10 brands of stick
margarine and 9 brands of tub margarine and margarine-like
products, samples of each collected by Slover et al. (1985) in
one or more of five major U.S. cities, is given in T ~ b l e 1. A
total of 38 products was collected and analyzed. Ten different
blends of hydrogenated and unhydrogenated vegetable oils were
used in formulating these products. In the stick margarines, the
average percentage of trans-18:1 isomers in the different blends
ranged from 18.53 to 27.63%, and from 10.89 to 18.27% in the tub
margarines and margarine-like products (Table 1). Average per-
centage of cis-18:1 isomers other than 9c-18:1 in the blends of
stick margarines ranged from 8.06 to 14.44%, and from 5.25 to
9.34% in the tub products (not shown in Table 1).
Average percentages of tc- and ct-18:2 isomers were
similar in a given blend and ranged from D.11 toA.27% (Table 1).
The highest values were found in a blend of hydrogenated soybean,
cottonseed, sunflowerseed, and corn oils. Highest percentages
of the t,t-18:2 isomer in stick and tub products (1.28 and 1.33%,
respectively) were also found in products formulated from this
blend. The relatively high levels of these isomers probably
reflects hydrogenation conditions rather than the particular
blend of vegetable oils. Average t,t-18:2 was less than 0.5%
in the other blends. Linolenic acid was found in all products
at concentrations in the range 0.26-3.76%. Linolenic isomers, if
present, were at concentrations below the GLC limit of detection
which was estimated to be 0.002-0.003 gj100 sample (Slover, 1985).
7
Table 1. ..Fatty Acid Content In Margarines and Margarine-like Products Formulated
from Various Blends of Hydrogenated and Unhydrogenated Vegetable 01 Is. Samples of Each Product
Were Collected In One or More of Five U.S. CIties In 1981 (adapted from data of Slover et al.,
1985>
Blend and
Package*
C/CH
Stick
Tub
C/SBH
Stick
SBH
Stick
Tub
SB/SBH
Stick
Tub
SB/SBH/CSH
Stick
Tub
SBH/SB/CSH
Stick
SBH/CSH
Stick
Tub
SBH/P
Tub
No.
Products
t
2
5
2
4
7
1
2
1
6
No.
Samples
12
29
10
15
24
4
14
4
2
10
6
19
2 (40%)
18: It
20.62
12.26
18.87
27.63
15.02
19.19
10.89
21.85
23.33
16.05
tc
0.11
0.33
0.48
0.56
0.49
0.23
0.84
0.18
0.89
1.66
18:2
ct
0.28
0.39
0.67
0.62
0.50
0.38
0.86
0.41
0.88
1.68
2.31
tt
0.08
0.06
0.15
0.34
0.14
0.09
0.05
0.07
Tr
0.09
0.18
0.33
0.46
Total
Trans
21.09
13.04
20.17
31.14
17.57
20.46
11.93
19.21
16.14
22.53
25.28
19.72
22.23
9c,12c-18:2
33.27
42.27
10.78
27.90
27.44
36.50
26.30
32.14
24.75
22.41
26.75
26.21
Overall average content In stick and tub type margarines
SBH/CSH/SNH/CH
Stick
Tub
SN/SBH/CSH
Stick
Tub
Margarine type
Stick 14
Tub 24
4
3
4
2
69
95
25.49
18.27
22.04
14.64
22.85
14.53
3.36
3.96
0.69
0.26
1.01
1.20
0.82
0.42
0.95
1.24
1.28
1.33
0.28
Tr
0.26
0.21
33.89
27.83
23.83
15.32
25.07
17 18
14.37
20.52
35.49
46.39
23.31
31.89
*
Key to abbreviations: C=corn 01 I; H=hydrogenated; SB=soybean 01 I; CS=cottonseed 01 I;
P=pa 1m 011; SN=sunf lower seed 011.
t
Products are dIfferentIated by brand name and, for a given brand and package, by
type of product, I.e., margarine, whipped margarine, spread, or ImItation margarine.
8
3. Formulation of shortenings
General purpose shortenings are usually formulated
by adding 10-15% of an essentially fully hydrogenated fat to a
partially hydrogenated vegetable oil, interesterified lard, or
a blend of the three (Latondress, 1981; Weiss, 1983). The formula-
tion of these shortenings is a compromise between having the best
frying fat and the best pastry and cookie formulation, and one
also suitable for use in some cake formulations. In addition
to the general purpose shortenings which have wide use, special
purpose shortenings are formulated for cake baking, pie crusts,
crackers, and continuous bread baking (Weiss, 1983). An important
compositional variable in these shortenings is the type and amount
of emulsifier compounds. Enig et al. (1983) reported the trans
fatty acids composition of five household shortenings purchased
at supermarkets, two obtained from wholesale brokers, and -one
from a restaurant. They also extracted and determined the trans
fatty acid content of fats from french fried potatoes (purchased
at supermarkets and restaurants) and a variety of bakery products.
Products analyzed and average trans fatty acid content are given
in Table 2. Overall average trans acid content, giving equal
weight to each type of product, and representing both household
and food industry shortenings is 15.4%.
4. Salad and cooking oils
The principal salad and cooking oils used in
the U.S. are soybean, cottonseed, corn, peanut, and olive (U.S.
Department of Agriculture, 1984). Sunflowerseed and safflower
oils are also used but most of the U.S. production of these oils
is exported. Soybean oil is only hydrogenated to an extent to
reduce its content of linolenic acid and thereby increase its
flavor stability especially when used as a cooking oil. The
other salad oils are generally not hydrogenated. Both cottonseed
oil and hydrogenated soybean oil are "winterized"; this consists
of a cold treatment to remove high melting components that would
form a cloud at refrigerator temperatures (Weiss, 1983).
9
Table 2. ~ Fatty Acid Content of Fat Extracted from Bakery Products, Snack Chips,
and French Fried Potatoes ObtaIned from Supermarkets and Restaurants, and In Household
Shortenlngs*
No. Trans Fatty Acids In Fat, %
Product Samp les t-18: 1 1-18:2+ Total c.c-18:2
SHORTENINGS
Bread and rolls 10 8.7 1.4 10.1 32.4
\ ~
Breading mixes and fried crusts 4 19.6 3.2 22.8 32.4
Cake 4 12.8 1.1 13.9 21.0
Cookl es 24 17.4 2.1 19.5 15.4
Crackers 19 11.4 1.4 12.6 12.4
Pastries and pastry crusts 17 9.9 0.9 10.8 11 .5
PI z za crusts and pretzels 4 13.2 2.0 15.2 33.5
Household shortenings 7 21.7 3.6 25.3 12.4
AVERAGE 14.3 2.0 16.3 21.4
FRY ING 0 I LS
French fr led potatoes 11 10.7 1.3 12.0 10.9
Snack chips 13 10.1 1.8 11.9 31.8
AVERAGE 10.4 1.6 12.0 21.4
* Computed from data of En Ig et a I. ( 1983)
t
1-18:2 represents the sum of ct-, tc- and tt- Isomers of 18:2 fatty acids
10
II. ANALYTICAL METHODOLOGY
In this section, the application of various analytical
methods to the determination of geometrical and positional isomers
in fats and oils is discussed. Physical methods include infrared
(IR), ultraviolet (UV), mass (MS), and nuclear magnetic resonance
(NMR) spectrometry, and chromatographic separation procedures
including gas-liquid (GLC) , thin-layer (TLC), and high performance
liquid chromatography (HPLC). A chemical method, ozonolysis,
is used in the determination of positions of double bonds. An
enzymatic method is specific for cis methylene-interrupted double
bonds. Principles of these methods, limitations, and examples
of their application are presented. Methods for the analysis of
isomeric fatty acids have been reviewed by Scholfield (1979) and
Christie (1982). .
A. IR ABSORPTION SPECTROMETRY
IR absorption spectrometry is one of the most widely
used methods for determining the trans isomer content of fats and
oils. Isolated double bonds in the trans configuration exhibit
an IR absorption at approximately 967 cm-
1
Measurement of this
absorption band under controlled conditions and comparison with
the absorption of an appropriate standard is the basis for the
official method of the American Oil Chemists' Society. (ADCS,
1980), by the Association of Official Analytical
Chemists (AOAC, 1984), for the determination of the trans isomer
content in edible fats and oils. The method can be used with
either intact glycerides or methyl esters, but the values obtained
with triglycerides are higher, and those with methyl esters lower,
than true values measured in spiked samples. Errors are most evi-
dent in samples with trans contents less than 15%, and equations
for calculating correction factors are included in the official
method. Madison et ale (1982) have described a procedure which
appears to give more accurate results. They used a two-component
calibration standard consisting of a mixture of methyl elaidate
(0.000 to 0.100 g) and methyl linoleate (0.500 to 0.375 g) in
carbon disulfide and obtained baseline-corrected absorbance at
967 cm-
1
for each level of methyl elaidate. Baseline corrected
absorbance plotted against grams elaidic acid equivalent gave the
calibration curve for the translation of measured absorbances of
samples of methyl esters into grams elaidic acid equivalents. A
similar calibration curve was prepared from mixtures of trielaidin
with tristearin. Analytical recovery of low levels of added trans
isomers in refined and bleached vegetable oils showed applicability
of the method for trans levels below 15%. Trans isomer content
determined with the proposed method was in good agreement with
values obtained by capillary column gas-liquid chromatography
for a number of salad oils, shortenings, and margarines. However,
most of the samples analyzed had a trans isomer content
11
Allen (1969) showed that the ratio of the absorbance
at 965 cm-
1
to that at 1163 cm-
1
in methyl esters of fatty acids
has a linear relationship with the trans unsaturation content.
Isolated trans double bonds absorb at 965 cm-
1
; carbon-oxygen
bonds of the ester group absorb at 1163 cm-
1
. The method has
the advantage that the sample does not need to be weighed or made
up to a known volume. Comparison of the method with the official
IR AOCS procedure indicated it to be less accurate for samples
containing 0-20% isolated trans components (Huang and Firestone,
1971).
B.UV ABSORPTION SPECTROMETRY
Conjugated double bonds absorb strongly in the ultra-
violet spectrum. This is the basis for the official method of
the AOCS (1980), also adopted by the AOAC (1984), for the deter-
mination of conjugated polyunsaturated fatty acids in fats and
oils. Conjugated dienes, trienes, tetraenes, and pentaenes can
be determined from their absorptivities at their characteristic
wavelengths with appropriate corrections for overlapping of absorp-
tion bands. Before the development of GLC, concentrations of the
;various polyunsaturated acids in fats and oils were commonly deter-
mined by conjugating the double bonds systems with strong alkali
and measuring UV absorptions at their characteristic wavelengths.
The method remains useful as a sensitive technique for determining
percentage conjugation in partially hydrogenated edible fat
products. The specific extinction coefficient for the conjugated
18:2 isomers, 9c,11t- and 10t,12c-, is approximately 95 (Nichols
et al., 1951), and relatively low concentrations can be determined
with precision. For example, the average value S.D. reported
in a collaborative study (eight participants) of conjugated diene
in cottonseed oil was 0.32 0.01 (O'Connor et al., 1955).
C. MASS SPECTROMETRY
In mass spectrometry, positive ions of a sample substance
having different mass-to-charge ratios are separated as they pass
through combinations of magnetic and electrostatic fields (Beynon
et al., 1968). P o s i t i v ~ ions are formed in an ionization chamber
by collisions between electrons and sample molecules in the gaseous
state. Removal of a single electron forms a positive molecular
ion having the molecular weight of the sample substance. Addi-
tional energy imparted to the molecular ion as vibrational energy
by the bombarding electrons results in fragmentation into a variety
of ions of different masses. This fragmentation and associated
rearrangements forms the basis of the mass spectrometric analysis
of organic compounds. The ions formed in the ionization chamber
are accelerated in a strong electrostatic field and pass into the
mass analyzer portion of the spectrometer where they are separated
according to molecular weight and are focused on an ion detector.
12
The mass spectra of unsaturated fatty acids using conven-
tional electron impact ionization are unsuitable for determination
of the position of the double bond or geometrical isomers. Their
spectra are complicated by geometrical and positional isomerization
which may occur before fragmentation of the molecular ion (Beynon
et al., 1968). Derivatives prepared by introduction of functional
groups at the double bond such as the epoxide, methoxy, vicinal
diol, or ethers formed from the diol have been used on a limited
basis for determining positions of unsaturation (Aplin and Coles,
1967; Niehaus and Ryhage, 1968; Ryhage and Stenhagen, 1960).
Instead of using a derivative at the double bond, et ale
(1974) found that the N-pyrrolidine amide of unsaturated fatty
acids could be used to locate the position of the double bond.
None of the above fatty acid derivatives allows the
analysis of complex mixtures of positional isomers or the
detection of geometrical isomers. McCloskey and
(1965) showed that geometrical isomers of monounsaturated fatty
acid esters could be differentiated in mass spectrometry by using
the isopropylidine derivatives of diols prepared from the monoene
and comparison with the spectra of closely related standards. A
variation of this procedure employing the hexafluorolsopropylidine
derivative has been used to determine both the position and the
geometrical configuration of double bonds in olefinic hydrocarbons
(Johnson and Taylor, 1972). Dommes et ale (1976) reported the
analysis of mixtures of methyl esters of unsaturated fatty acids
by GLC/MS. Derivatives of the mixture were by gas
chromatography and the individual compounds analyzed by mass
spectrometry for chain length and positions and numbers of
double bonds.
More recently, chemical ionization mass spectrometry has
been used in determining positional isomerism in unsaturated fatty
acids. In this technique, a reaction gas is introduced into the
ionization chamber at a relatively high pressure as compared to
that of the sample compound. Electron impact on the reaction gas
produces a set of ions that react with the sample compound forming
a derivative having a characteristic fragmentation pattern. Chai
and Harrison (1981) were able to determine the double, bond location
in underivatized monounsaturated fatty acids using vinyl methyl
ether as the reactant gas in admixture with nitrogen or nitrogen/
carbon disulfide. However, Cervilla and Puzo (1983) noted that
this method is not specific for double bonds because adduct ions
are also observed with saturated compounds. These investigators
applied the mass analyzed ion kinetic energy spectrometry/collision
induced dissociation (MIKES/CID) technique to determine double
bond positions in monoenoic acids. Amino alcohol derivatives of
unsaturated fatty acid esters were prepared and were subjected to
chemical ionization with ammonia as the reagent gas. This "soft
ionization" mode produced molecular (M)+ or pseudomolecular ions
(M+H)+ of each molecular species present. Two types of spectra
of these precursor ions are produced: (a) MIKES spectra from
13
their decomposition in the second field-free area between the
magnetic field and electrostatic analyzers; and, (b) MIKES/CID
spectra by increasing the helium pressure in the collision cell
in the second field-free area. This method was used to identify
a number of monounsaturated fatty acids from Mycobacterium phlei
(Cervilla and Puzo, 1983).
Determination of the position of double bonds in poly-
unsaturated fatty acids is more complicated than in monoenes,
but limited success has been reported for pure compounds using
chemical ionization (Suzuki et al., 1981).
The use of mass spectrometry for determining the structure
of unsaturated fatty acids has limitations. Derivatives must be
prepared either prior to analysis or by interaction with a reaction
gas in the ionization chamber of the spectrometer. Because of the
complexity of mass spectra, the method has been used more often to
determine the structure of a single compound than for the quantita-
tive analysis of mixtures. For this reason, mass spectrometry is
frequently used in conjunction with gas chromatography.
D. NUCLEAR MAGNETIC RESONANCE
The nuclei of certain isotopes,. including 1H and 13C,
possess an intrinsic mechanical spin and an associated angular
momentum. Because atomic nuclei also have an electric charge,
this spin gives rise to an associated magnetic moment (Jackman,
1959). The nuclear spin isa quantized property; the allowed spin
numbers determine the number of orientations and associated energy
levels that the nucleus can assume in an applied magnetic field.
Energy differences between states are proportional to the strength
of the applied field. Appropriate frequencies in the radio fre-
quency range can cause transitions between adjacent energy levels.
The energy absorbed at a resonance frequency can be detected,
amplified, and displayed on a chart. The trace of a compound,
showing the variation in the intensity of the resonance signal
with increasing applied magnetic field, is the NMR spectrum of
that compound.
The frequency with which any given hydrogen atom resonates
depends on the extent of shielding of its nucleus from the applied
field by the surrounding electron distribution. The precise fre-
quency at which a given hydrogen atom resonates thus depends on
its molecular environment and is subject to "chemical shifts" by
the presence of adjacent atoms. Magnetic interactions arising
from spin-spin coupling with neighboring hydrogen atoms introduce
fine structure into the absorption bands. The number of hydrogen
atoms of a given type (i.e., those in the same magnetic environ-
ment) is proportional to the intensity of their absorption band.
The data on chemical shifts, absorption band intensity, and fine
structure may be used to deduce the arrangement of hydrogen atoms
14
j
and often the complete structure of an unknown compound. Similar
shielding and magnetic interactions affect the resonance frequencies
of 13C atoms, resulting in 13C-NMR spectra that are characteristic
of specific types of molecular structures.
For NMR analysis, compounds must be in solution, prefer-
ably in solvents such as carbon tetrachloride or deuterochloroform.
Positions of absorption peaks, i.e., chemical shifts, are not
expressed in absolute units but by the difference in frequency
relative to an absorption peak of an internal standard. Tetra-
methylsilane has a single absorption peak and is generally used
for this purpose. The difference in frequency x 10
6
divided by
the applied oscillator frequency gives the chemical shift in parts
per million.
Application of 1H NMR spectroscopy using 220 MHz equip-
ment to determine position of double bonds in fatty acids, and
differentiation of cis and trans isomers, has been reported by
Frost and Barzilay (1971a,b). IH NMR spectra of 143 nonconjugated
alkenoic and alkynoic acids and esters have been published and
features characteristics of geometrical and positional isomers
identified (Frost and Gunstone, 1975). For both cis and trans
octadecenoic acids, it is possible to identify the to
and to of single fatty acids indicates
double bond position). Differences in the spectra of the
and are small and identification is less
certain.
13C NMR spectra also can be used to study geometrical
and positional isomerism in unsaturated fatty acids. The ratio
of methyl oleate to methyl elaidate in mixtures was determined
by comparing the cis- and trans-C9 and C10 olefinic carbon peaks
(Barton et al., 1975). However, the cis-trans ratio in partially
hydrogenated fats could not be determined by this method because
of effects due to positional.isomerism. The reported greater
chemical shift of the allylic carbons makes 13C NMR a more
suitable method for determining the geometrical isomer content.
This method has been applied to a number of complex lipid mixtures
containing geometrical isomers of dienes and trienes as well as
monoenes (Pfeffer et al., 1977). The results agreed very well
with values obtained by GLC and IR spectrometry. The analysis
can be made on intact glycerides, which eliminates time-consuming
sample preparations required by the other methods. An added
advantage is that the compositional analysis of all the major
lipid components can also be determined. Cost of the instru-
mentation with these capabilities is likely to inhibit its use
for the routine determination of the geometrical isomers and
compositional analysis of edible fats and oils. 13C NMR is
unable to determine positional isomers except in selected cases.
15
E. CHROMATOGRAPHIC METHODS
There are two basic types of chromatographic methods for
analyzing lipid composition based on the separation principles
involved: adsorption chromatography and partition chromatography
(Christie, 1982; Stein and Slawson, 1967). Adsorption chromatog-
raphy includes TLC, liquid chromatography, and HPLC. Partition
chromatography includes GLC and liquid-liquid chromatography.
Separations in HPLC reverse-phase chromatography may be considered
to result from partition between two liquid phases. Discussion
of these methods will relate principally to their applicability
to the analysis of trans fatty acid isomers. Phenomenological
and thermodynamic bases of chromatographic separations are dis-
cussed by Horvath and Melander (1983).
1. Adsorption chromatography
In adsorption chromatography, fractionation of
lipids depends on differences in the degree to which the lipid
components are adsorbed on a solid support relative to their
solubility in an appropriate solvent (Christie, 1982; Horvath
and Melander, 1983). The polarity of the functional groups
in a lipid is the major factor determining the strength of its
adsorption on the solid support. The hydrocarbon chains of
lipids are nonpolar and are weakly adsorbed as compared to
carboxyl, hydroxyl, or the polar head groups of complex lipids.
Lipids are eluted from an adsorbent bed bypassing solvents of
increasing p ~ l a r i t y through the bed. The adsorbent, usually
finely divided silica gel (also referred to as silica or silicic
acid) may be deposited as a thin layer on a glass plate (TLC) or
held in glass columns (column chromatography). TLC is generally
used as an analytical tool to fractionate lipid extracts from
plant or animal tissues into their component classes. For animal
tissues these are triacYlglycerols, cholesterol esters, cholesterol,
free fatty acids, and phospholipids (Wood, 1979a). The phospho-
lipids can be further resolved by TLC. Because positional and
geometrical isomers have relatively small effects on adsorption
and elution of unsaturated fatty acid chains on silica, these
components of a lipid class do not separate under the usual
conditions of TLC. Similar separation of lipid classes can
be achieved With column chromatography on silica gel. However,
TLC is more rapid and sensitive, gives better resolution, and
detection of separated fractions is simpler (Christie, 1982).
2. Argentation chromatography
The incorporation of silver ions into the adsorbent
layer allows separation of certain lipid classes (acylglycerols,
fatty acid esters) according to the degree of unsaturation,
and fatty acid esters also according to geometric configuration
16
(Morris, 1966). Basis of the separation is the formation of
reversible polar complexes of silver ions with double bonds of
the aliphatic chains (Dewar, 1951). Difference in the strength
of this complex with cis and trans geometric isomers of fatty
acid esters is sufficient to allow their separation by argentation
adsorption chromatography in columns or thin layers. In TLC, all
cis positional isomers of 18:1 methyl esters have lower RF values
than the trans isomers except for the ~ 2 cis ester (Gunstone
et al., 1967; Morris et al., 1967). Differences in RF values
among cis or trans positional isomers as separate classes are
sufficient to make some separations possible by multiple solvent
development of TLC plates, e.g., ~ 6 , ~ 9 , and ~ 1 2 (Gunstone et al.,
1967; Morris et al., 1967), but generally such separations do not
occur in TLC separations of fatty acid methyl esters (Wood, 1983).
Argentation TLC is commonly employed in connection with
other analytical techniques for the separation and identification
of positional isomers in fatty acid esters. Saturated fatty
esters, cis monoenes, trans monoenes, dienes, and other lipids
resolved by argentation TLC are each removed from the TLC plate,
eluted from the adsorbent, internal standards added, and quantita-
tion of the fractions made by GLC. Some resolution of cis or
trans positional isomers may be achieved by glass capillary GLC or
by a combination of GLC and HPLC. More complete resolution can be
obtained by ozonolysis of the fractions and identification of the
degradation products by GLC (Johnston et al., 1978; Scholfield,
1979). It should be noted that the fractions separated by TLC may
not be pure and may possibly be contaminated with other classes of
lipids, or cis and trans fractions may be cross-contaminated with
fatty acid esters having a higher degree of unsaturation.
3. Partition chromatography
In partition chromatography lipids are separated
according to differences in their partition coefficients between
two immiscible liquids (liquid-liquid chromatography) or between
an inert gas and a liquid (GLC). Cis and trans fatty acid esters
are not readily separated by liquid-liquid chromatography (Christie,
1982) and this technique will not be discussed further. In GLC,
the lipids to be separated are volatilized by heating and passed
into a stream of inert gas through a heated, temperature con-
trolled column in which a high boiling-point liquid is coated
as a thin film on a solid supporting material (Christie, 1982;
Slover, 1983). The partition of lipid components between the
liquid and gas phases depends upon their relative volatilities
and their solubilities in the liquid phase. As a consequence,
components are retarded to different degrees as they pass through
the column. Separated lipid components emerge as peaks from the
column and are commonly detected with a flame ionization detector.
Response of the detector to homologous series of bompounds is
linearly dependent on the mass of each component over a wide
17
range of molecular weights. Resolution of components depends
on gas flow rate, length of column, and amount and nature of
the liquid phase.
Two types of GLC columns are used: packed columns in
which the solid supporting materials for the liquid phase are
generally diatomaceous earths treated to minimize adsorptive
effects, and open-tubular or capillary columns, the inner wall
of which is coated with the liquid phase. Packed columns, 1.5
to 2 m in length, containing a support with 10-15 weight percent
liquid phase are reported to have an efficiency of 3,000-5.000
theoretical plates. Glass capillary columns, 0.25 rom internal
diameter and 100 min length, may have an efficiency of 20,000-
300,000 theoretical plates and can resolve some of the positional
isomers of cis and trans monoenoic fatty acids (Christie, 1982;
Slover, 1983).
GLC is the method of choice for the analysis of most
lipids and derivatives of lipid hydrolysis products. Analytical
packed columns are used routinely to determine fatty acid com-
position according to hydrocarbon chain length and degree of
unsaturation. Separation of cis-18:1 isomers from trans-18:1
isomers in margarine fatty acid methyl esters as two incompletely
resolved peaks was reported with 6.1 m packed OV-275 columns
(Ottenstein et al., 1977; Perkins et al., 1977) and also with
15 ft packed Silar 10C columns (Conacher and Iyengar, 1978;
Heckers et al., 1977a). Similarly the 9,12-18:2 isomers were
incompletely resolved. Perkins et ale (1977) found a mean
deviation of 2.29 1.7% in the trans content of 12 margarines
measured by the GLC and the AOCS official infrared method. A
collaborative study of the GLC method using a 6.1 m x 2 rom id
column packed with 15% OV-275 (Perkins et al., 1977) showed good
recoveries of total trans acids; both between-laboratory and
within-laboratory CVs were consistently better for determination
of total trans acids than the AOAC IR method (Gildenberg and ,
Firestone, 1985).
Sampugna et ale (1982) reported the analysis of geometric
isomers of partially hydrogenated fats using relatively short
(15 m) glass capillary columns coated with SP-2340. Because
resolution of cis and trans isomers in the fatty acid methyl
esters of margarines was incomplete, correction factors were
derived based on chromatographs of mixtures containing known
quantities of cis and trans isomers. An advantage of the short
column was the relatively short time (15 minutes) required to
chromatograph a sample of hydrogenated food fat. The procedure
was used in the analysis of the purified fatty acid methyl esters
of 220 food fats (Enig et al 1983). A disadvantage of both
the 15 m capillary and the 6.1 m packed columns is the incomplete
baseline separation of the 9,12-18:2 isomers and unidentified
compounds which elute at similar times (Slover et al., 1985).
18
The contribution of these unidentified substances may lead to a
substantial overestimation of the content of the trans-9,12-18:2
isomers, particularly the 9t,12t isomer.
Most successful resolution of cis and trans positional
isomers of fatty acids has been made wirs-100 m or longer glass
capillary columns. However, even with highly polar liquid phases,
the resolution of positional octadecenoate isomers of either the
cis or trans configuration is not complete (Lanza and Slover,
1981; Slover, 1983; Slover and Lanza, 1979; Wood, 1983). The
octadecenoate isomers with the double bond in the 66, 67, and
68 positions are the most difficult to separate. These isomers
are present at low levels in most cis- and trans-octadecenoates
of partially hydrogenated fats and oils and other methods such
as ozonolysis and GLC of the degradation products are required
when these isomers need to be quantitated. The octadecadienoates
normally present in partially hydrogenated fats and oils, namely
the 69, 612 geometrical isomers, can be quantitated by capillary
GLC (Slover, 1983). Slover et ale (1985) used a glass capillary
column 100 m x 0.25 mm coated with OV-275 in the analysis of cis-
and trans-18:1 and 18:2 positional isomers in multiple samples-0f
83 brands of margarine. It should he noted, however, that identi-
fication and quantitation of the 9,12-18:2 geometrical isomers
can be difficult in some samples because of the presence of other
components that have similar retention times. A more general
problem is the lack of commercially available pure standards
for many of the isomeric 18:1 and 18:2 fatty acids. i,
4. High performance liquid chromatography
HPLC, i.e., high efficiency, rapid separation of
components has been achieved by using high pressure on the eluting
solvent and small particle size column packings (Hamilton and
Sewell, 1982). In most studies on the separation of fatty acid
derivatives, the packing material consists of a long chain hydro-
carbon chemically bonded onto microparticles of silica gel. This
provides a nonpolar stationary phase of essentially liquid hydro-
carbon from which the fatty acid derivatives by
partition into a polar solvent. Such separations are referred
to as reverse phase chromatography (Stein and Slawson, 1967).
The recent use of fatty acid derivatives that absorb in the
ultraviolet permits much improved sensitivity and stability
of detection over the previously used refractive index detector
(Wood, 1984; Wood and Lee, 1983). Geometrical isomers of monoenes
with the double bond at a specific location are easily resolved on
C-18 reversed phase columns (Wood, 1983). The cis, cis- and trans,
trans-9,12-18:2 isomers are resolved and both are-separated from
the cis, trans and trans, cis isomers that elute together (Wood
and Lee, 1983). Many of the positional isomers of either the cis
or trans octadecenoates, especially those having the double bond
19
between carbon 6 and the carboxyl group, can be resolved but
resolution of positional isomers with double bonds near the
middle of the hydrocarbon chain is poor (Wood, 1984).
A more complete HPLC-separation of 18:1 isomers has been
reported using columns packed with silver nitrate treated porous
silica (Battaglia and Frohlich, 1980). With this packing (adsorp-
tion chromatography), cis and trans 18:1 fractions of the methyl
esters of margarines were separated without overlapping. At the
optimal silica content of silver nitrate (20%), the individual
positional isomers (65-612) of both geometrical isomer fractions
were separated as identifiable peaks but not with baseline separa-
tion. This technique deserves further study.
F. C H E M ~ C A L METHODS
Despite the advances in technology and instrumentation,
ozonolysis remains the most generally applicable method for deter-
mining the positionsOof double bonds in a partially hydrogenated
fat. Methyl esters are usually prepared and geometrical isomers
separated by argentation TLC. Monoenes, the predominant fraction
in most commercial hydrogenated fats, are subjected to ozonolysis
(Privett and Nickell, 1963, 1966). The ozonides are cleaved by
pyrolysis or chemical reaction and the resulting aldehyde-esters
are quantitated by GLC. Several modifications of this procedure
exist, each having advantages that suit particular applications
and equipment (Beroza and BierI, 1967; Bitner et al., 1969;
Privett and Nickell, 1963, 1966; Spence, 1970). The diene and
triene fractions separated by TLC are usually analyzed intact by
capillary GLC or HPLC. Generally, the analysis of the double bond
positions is unnecessary because the diene and triene fractions
in most commercial hydrogenated fat products consist of the geo-
metrical isomers of the 9,12-18:2 and 9,12,15-18:3 fatty acids.
The positions of the double bonds in the diene and triene fractions
can be established by ozonolysis of the monoenes resulting from
partial hydrogenation (Privett and Nickell, 1966). However, the
double bonds at different positions are not reduced at the same
rate which makes the results only qualitative. An alternate
method has been reported which analyzes the alcohol fragments
formed by reduction of the ozonides (Johnston et al., 1978).
This method requires computer resolution of a matrix of linear
simultaneous equations.
G. ENZYMATIC METHODS
The specificity of enzyme action has been applied to
several problems in lipid analysis. Pancreatic lipase and
related enzymes selectively hydrolyze the ester bond at the
primary hydroxyl functions of triacylglycerols and phospholipids,
20
thus enabling the identification of the acyl group at position 2
and, in the case of phospholipids, the acyl group at position 1
(Breckenridge, 1978).
The enzyme lipoxidase (linoleate: oxygen reductase,
EC 1.13.13) is an oxidative enzyme that is specific for substrates
that contain. cis methylene-interrupted double bonds (MacGee, 1959).
Lipoxidase substrates that may be present in edible fats and oils
or in animal tissues are linoleic, linolenic, and arachidonic .
acids. The enzyme acts on these substrates to form conjugated
diene hydroperoxides that absorb strongly at 234 nm. This is
the basis of the lipoxidase method for the estimation of total
cis methylene-interrupted polyenoic acids. In the official method
of the AOCS (1980), a standard curve is prepared using weighed
quantities of high purity trilinolein. The method has been used
to detect the maximum content of linoleate in margarines, recog-
nizing that any residual linolenate, or other source of a cis
methylene-interrupted polyunsaturated acid would contribute to
the analysis (Beare et al., 1967; Beare-Rogers et al . , 1979;
Nazir et al., 1976).
21
III. CONSUMER EXPOSURE DATA
The major sources of trans fatty acids in the U.S. diet
are hydrogenated vegetable oil products and the fats of ruminant
animals, principally beef and milk fats. Published reports indi-
cate that there is variation in the trans fatty acid content in
both types of products. In ruminant fats there is a seasonal
variation depending on the composition of the ingested feed;
in hydrogenated vegetable oil products there has been a long-
term variation resulting from changes in processing technology
to develop products with improved physical properties and higher
ratios of polyunsaturated to saturated fatty acids.
In the following discussion, pUblished reports on
ruminant fats and hydrogenated vegetable oils are reviewed
and an estimate is presented of current per capita daily trans
fatty acid consumption. The limited pUblished data on the trans
content of hydrogenated vegetable oil products and the market
share of specific products, limit the accuracy of estimates
based on pUblished information. Based on their unpublished data
on composition and market volume of hydrogenated vegetable oil
products, the vegetable oil processors submitted estimates to
the ad hoc Review Panel of trans fatty acid availability in the
food supply for 1963, 1970, 1975, 1980, and 1984 (Hunter, 1985).
A summary of their estimates is presented in this section of the
report; the complete report appears in Appendix A.
A. BEEF AND MILK FAT
Slover et ale (1985) reported the average trans fatty
acid content of the fat of the separable lean portion of 14 retail
cuts of beef, raw and after cooking, to be 3.21 and 3.83%, respec-
tively (Table 3). The raw lean meat had the cover and seam fat
removed before analysis; the cooked lean meat had these fats
removed after cooking and before analysis. Part of the increase
in fat content after cooking is attributable to moisture loss
(71.14% moisture before cooking, 57.72% after cooking), but much
of the increase must have resulted from migration of fat from the
cover and seam fat. Trans content was 6.55% in raw, and 5.47% in
cooked separable (cover and seam) fat, levels considerably higher
than those in lean tissue fat.
Also listed in Table 3 are values reported for the trans
fatty acid content in butterfat and in lightly hydrogenated lard.
Highest values in butterfat were found in butter from milk produced
in early spring and summer, and lowest from milk produced in winter,
reflecting seasonal variation in the fat composition of the feed
of dairy cows (Parodi, 1976).
[ ~ e c e d i l i g Page BIa";'J 23
Table 3. ~ Fatty Acid Content In the Fat of Animal Products
Fat
No. Content Trans Content of Fat, %
Product Samples % 18: 1 18:2 Total Reference
Beef, reta II cuts
Lean portIon
raw* 111 6.46 3.18 3.21 Slover a I (1985 )
~
et
cooked* 109 11.03 3.65 3.85 Slover et a I (1985)
Separated fat
raw 12 70.89 6.51 6.55 Slover et a I. (1985)
braised 6 70.32 5.39 5.47 Slover et a I (1985 )
Beef, separated fat
raw 1.8 1.8 Enlg et al. (1983)
Beef, lean muse let
raw 3 3.2 2.5 Wood (1983 )
Butter 5 brands 81.0 1.8 Sml th et at. (1978)
Butter ( 12 f actor Ies 4.3-7.6 Parod 1 ( 1976)
samp Ied month Iy
throughout year)
Cream fat
ear Iy spr Ing milk 3.7 Hay and Morrison (1970)
Butter 3 3.4 En Ig et al. (1983)
Lard
part la I IY hydrogenated 0.3 En Ig et al. (1983)
* Fourteen cuts of beef from three prime, one choice, three good, and one USDA standard grade
carcass were analyzed raw and after cooking. Cover and seam fat were removed from raw cuts
before""fat extraction, but from cooked cuts after cooking.
t Sterno mandlbularls muscle from three feedlot anlma,ls with carcass weights of 722-832 lb.
All visible fat removed before analysis.
24
B. MARGARINES
Trans fatty acid composition reported for domestic
margarine sold at retail for home use is given in Table 4. Rice
et ale (1962) samples of all major brands of margarine
in five different regions centering about Atlanta, Chicago, Fort
Worth, Los Angeles, and New York City. The brands represented
over 90% of the margarine sold in 1961. Slover et ale (1985)
analyzed samples of major brands of stick and tub margarines and
margarine-like products (10 and 9 brands, respectively) collected
in Atlanta, Chicago, Houston, Washington-Baltimore, and Boston in
1981 (Table 1, p.8). They also reported analyses of 24 brands of
stick margarines collected in the Washington-Baltimore areas in
1980, and 15 brands of stick margarines and spreads in four major
cities in 1980. Enig et ale (1983) reported the analysis of 24
stick margarines and 16 tub margarines collected in the
D.C. metropolitan area in 1978-1980. Overall average trans con-
tent of the stick margarines analyzed by Slover et ale (1985) and
Enig et ale (1983) is 23.9% and for tub margarines is 16.2%. The
values reported in Table 4 indicate that the average trans content
of stick margarine has decreased since the 1960's and this is
probably also true of the tub margarines which were introduced
in the late 1960's.
C. SHORTENINGS
The average trans content in four shortenings collected
in 1965 by Scholfield et ale (1967) was 22.5% (Table 4). Mattson
et ale (1975) stated that shortenings typical of 95% of the house-
hold vegetable shortenings contained 19% trans fatty acids. As
discussed in the Background section, Enig et ale (1983) analyzed
shortenings purchased at supermarkets, obtained from wholesale
brokers, and extracted from a large number of bakery and other
products purchased at restaurants and supermarkets. Average
trans fatty content of these products is 16.3%.
D. SALAD AND COOKING OILS
Within the limitations posed by the relatively few samples
analyzed prior to 1980, the available data do not indicate a sig-
nificant change in the trans fatty acid content of hydrogenated
vegetable oils used in the manufacture of salad and cooking oils
(Table 4). The major or sole component of the oils collected by
Enig et ale (1983) was hydrogenated soybean oil. The major change
in salad and cooking oils over the past 10 or more years has been
the increased use of soybean oil, increasing from 3.1 billion lbs.
in 1973 to 4.7 billion lbs. in 1983 (U.S. Department of Agriculture,
1975, 1984), and now constituting about 83% of all salad oils.
25
Table 4. ~ Fatty Acid Content In the Fat of Vegetable 011 Products
Product
Date of Sample
Collection
No. Brands
(and Products*)
Analyzed
Margarine
Trans Content, %
18: 1 18:2 Total Reference
Stick
StIck
Tub
Stick
Tub
Stlck
t
StickS
Stick'
Tub'
1961
1970
1970
1978-1980
1978-1980
1980
1980
1981
1981
26
6
4
24
16
24 (0)
15 (17)
10 (14)
9 (24)
Shortening
22.4
12.7
21.75
22.27
22.85
14.53
1.9
1.7
1.21
1.81
2.22
2.65
34.8
28.0
24.3
14.6
23.0
24.1
25.1
17 .2
Rice et al. (1962)
Carpenter and
Slover (1973)
Carpenter and
Slover (1973)
En I g et a I. (1983)
Enlg et al. (1983)
Slover et al. (1985)
Slover et al. (1985)
Slover et al. (1985)
Slover et al. (1985)
1965
1975
1978-1980
4"
Typlcal**
89
tt
Salad and Cooking 01 I
14 5
14.3 2.0
22.5
19
16.3
Scholfield et al. (1967)
Mattson et al. (1975)
Enlg et al. (1983)
Hydrogenated
Hydrogenated
Unhydrogenated
Hydrogenated
Un hydrogenated
1965
1972
1972
1978-1980
1978-1980
2
5
9
28
B
14
7.9
Trace
10.1
Trace
1.8
1.9
Trace
9.4
9.7
Trace
12.0
Trace
Scholfield et al. (1967)
Carpenter et al. (1976)
Carpenter et al. (1976)
En Ig et al. (1983)
En I g et a I. (1983)
*
t

,
**
tt

Products for a given brand name may Include margarine, whipped margarine, spread, and
Imitation margarIne
Samples were collected In the Washington, D.C. metropolitan area
Samples were collected In four major U.S. cities
Samples of each brand were collected In one or more of five major U.S. cities
Values stated to be typical of 95% of household shortenings
Number of samples, see Table 2
Includes frying oils from Table 2
26
E. 'TRENDS IN CONSUMPTION OF FAT AND TRANS FATTY ACIDS
An estimate of the trend in per capita consumption of
fat and its constitutent trans fatty acids can be based on data
provided by the U.S. Department of Agriculture on the apparent
domestic consumption of fat from animal and vegetable sources
as indicated by production and sales data. These data for 1965,
1972, and 1983 (Raper, 1985; Rizek et al., 1974) are given in
Table 5. It will be noted that apparent daily fat consumption
per capita increased from 145 g in 1965 to 166 g in 1983. Animal
fat consumption was essentially constant, whereas vegetable fat
consumption increased by 22.7 gjcapitajday. Much of this increase
is accounted for by greater consumption of salad and cooking oils
(13.4 g) and vegetable shortenings (6.1 g).
Apparent per capita daily trans fatty acid consumption
in 1965, 1972, and 1983, was calculated from the data on apparent
fat consumption in Table 5 and the trans content of animal and
vegetable fats given in Tables 3 and 4. The quantity of
fat in Table 5 represents the total fat in retail cuts, lean plus
separable fat. The trans fatty acid content in this composite fat
was calculated to be 5.8% from the data of Slover et al. (1985) in
Table 3 and from the proportions of lean and separable fat in the
USDA grades of beef (Watt and Merrill, 1975). The average value
(3.4%) of Enig et al. (1983) was selected for the trans content
of butterfat. This value is outside (below) the range reported
by Parodi (1976) to apply to both milk and beef fat, whereas the
value selected for beef fat, 5.8% is at the of the range
(4.3-7.6%).
Values for the percentage of trans fatty acids in the
hydrogenated vegetable oil products were selected from Table 4 on
the basis of dates that the samples were collected. For margarines
in 1983, the average of values reported by Slover et al. (1985)
and Enig et al. (1983) was selected. The quantities of stick and
tub margarine available in 1983 were based.on information provided
by the National Association of Margarine Manufacturers. The same
relative amounts were used in computing an average trans content
of margarine in 1972 from the data of Carpenter and Slover (1973).
The percentage of trans in salad and cooking oils for
1983 is based on the assumption that only the soybean oil component
was hydrogenated, and that soybean oil constituted 83% of all salad
and cooking oils (U.S. Department of Agriculture, 1984). Applying
this factor to the average trans content of the hydrogenated salad
and cooking oil analyzed by Enig et al. (1983) gives a trans con-
tent of 10.0%.
As shown in Table 5, apparent daily per capita consumption
of trans fatty acids in 1965, 1972, and 1983 was 9.7, 9.9, and
10.2 g, respectively, or 6.7, 6.2, and 6.1%.of total apparent daily
fat consumption. These figures suggest that there has been little
change in the apparent daily per capita consumption of trans fatty
acids over the past 18 years.
27
Table 5. Fats and ~ Fatty AcIds Available In the U.S. Food Supply, 1965, 1972, 1983,
Per Capita/Day.
Item
Total fat
AnImal
Vegetable
Fat
g
145
96
49
1965
Trans Acids
g
Fat
g
156
97
59
1972
Trans Acids
g
Fat
g
1983
Trans AcIds
g
Selected sources
Salad & cooking oll.s
Shortening (veg)
Margarine (veg)
Hard
Soft
Butter
Lard
DIrect
Indirect
Ed IbIe be.ef fat
6.4
11.4
8.0
3.5
2.4 6.55
0.22
0.16
22.9
14.6
10.3
4.9
8.6
4.0
3.0
3.0
9.7
19
26.7
6.55
2.22
2.77
2.75
0.17
0.20
30.8
18.3
6.37
2.59
5. I
3.9
2.2
1.7
5.9
10.0
16.3
6.55
3.08
2.98
1.52
0.42
0.17
0.38
Meat, poultry, fish
Beef
Pork
Fat cuts
Lean pork
Pou Itry
DaIry products
Excluding butter
Whole fluid milk
Low fat ml Ik
Fluid cream
Cheese
Frozen desserts
TOTAL TRANS FATTY ACIDS
PERCENT OF TOTAL FAT
16.8
14.7
10.4
3.0
21.2
11.7
5.8 0.97
3.4 0.72 ..
19.6
29.3
17 .2
12. I
3.6
5.8 1.14
0.68
18. I
27.9
11.7
16.2
8.0
18.3
5.4
2.0
1.3
7.2
2.4
5.8
0.62
AvaIlability of fat In the U.S. food supply as IndIcated by USDA productIon and sales data.
As discussed In the text, these fIgures do not represent actual per capIta consumption of fats.
t Assumes 1I g/caplta/d of frying 01 Is used by food processors (Rlzek et al. 1983) and 75% of
separable tat of retail cuts of beef were dIscarded.
28
Although food disappearance data (Table 5) provide a
basis for calculating the trend in trans fatty acid consumption,
these data overestimate both fat consumption and per capita trans
fatty acid consumption. Not taken into account are losses in food
processing and food service industries, especially the discard of
used deep frying fats. Similar losses of frying fats occur in the
home preparation of food. Plate waste including the discard of
meat fat represents another important loss. A correction for
discard of deep frying fats by food processors and trimming the
separable fat from retail cuts of beef has been made in Table 5
for the 1983 estimate of trans fatty acid consumption. Rizek
et ale (1983) reported an estimate of 11 g/capita/day for the
discard of industrial deep frying fats. Discard of beef fat is
assumed to be 75% of the separable fat. These corrections reduced
total per capita daily fat consumption to 141 g, trans fatty acid
consumption to 8.3 g, and percentage of trans fatty acids in total
fat (141 g) to 5.9%.
The Institute of Shortening and Edible Oils in coopera-
tion with the Association of Margarine Manufactures (Hunter, 1985)
estimate that .trans fatty acid availability in the diet is about
7.6 g/capita/day and has changed little during the past 15 to
20 years (Table 6). Hunter (1985) used a different approach
in making its estimate of consumption of trans fatty acids from
that just discussed. Estimates of fat availability were based
on A.C. Nielsen Company data on retail sales (i.e., household use)
of shortening and salad and cooking oils, published estimates of
quantities used by the food service industry, and estimates of
amounts used by bakeries and other food processors (industrial
use). It was assumed that much of the fats and oils used in the
food industry was used in deep frying and was ultimately discarded.
Trans fatty acid contents of the various hydrogenated fats were
averages based on industry compositional data and market shares.
Also included are estimates of trans fatty acids consumed in beef'
and dairy fats. A detailed description of the bases of the Insti-
tute's estimates is given in the statement (Appendix A) presented
by J.E. Hunter at the Open Meeting of the ad hoc Review Panel.
Although the approaches used in deriving the estimates
of trans fatty acid availability in the food supply differed, the
estimates given in Table 5 (8.3 g/capita/day) and Table 6 (7.6 g/
capita/day) are in remarkably good agreement. The poundage of
deep frying fats and margarines on which the values in Table 6
are based is not cited in the Institute's statement (Hunter,
1985), but if it is assumed that they are the same as used for
Table 5, and that total fat disappearance was 141 g/capita/day,
then percentage of trans fatty acid in total available fat would
be 7.7/141 = 5.4% based on the Institute's data, somewhat less
than the value given in Table 5.
Emken (1981) estimated the average per capita daily
trans fatty acid consumption to be 6.8 g in the U.S. This value
was based on an estimated daily consumption of 34 g hydrogenated
29
T
a
b
l
e
6
.
T
r
a
n
s
F
a
t
t
y
A
c
i
d
L
e
v
e
l
s
I
n
t
h
e
D
i
e
t
1
9
6
3
-
1
9
8
4
I
n
g
/
p
e
r
s
o
n
/
d
a
y
(
H
u
n
t
e
r
,
1
9
8
5
)
S
a
l
a
d
a
n
d
H
o
u
s
e
h
o
l
d
F
o
o
d
S
e
r
v
i
c
e
I
n
d
u
s
t
r
i
a
l
M
e
a
t
&
D
a
i
r
y
Y
e
a
r
C
o
o
k
i
n
g
O
i
l
s
S
h
o
r
t
e
n
i
n
g
s
M
a
r
g
a
r
i
n
e
s
F
a
t
s
&
O
i
l
s
F
a
t
s
&
O
i
l
s

P
r
o
d
u
c
t
s
T
o
t
a
l
s
1
9
6
3
0
.
0
6
1
.
0
1
2
.
4
8
N
A
t
I
1
.
5
4
1
9
7
0
0
.
3
8
0
.
7
6
2
.
7
3
1
.
3
5
1
1
.
4
8
7
.
7
0
1
9
7
5
0
.
4
9
0
.
7
5
2
.
7
3
N
A
t
1
1
.
4
1
1
9
8
0
0
.
3
5
0
.
6
0
2
.
7
3
1
.
5
4
1
1
.
3
3
7
.
5
5
1
9
8
4
0
.
3
1
0
.
5
5
2
.
8
5
N
A
t
1
1
.
3
8
c
.
u
0

E
s
t
i
m
a
t
e
d
t
N
A
-
N
o
t
a
v
a
i
l
a
b
l
e
soybean oil containing 20% trans-18:1 and the assumption that
the quantities of other hydrogenated vegetable oils contribute
negligible amounts of trans acids to the average diet.
Enig et al. (1978) calculated a daily per capita intake
of 12.1 g of trans fatty acids in the U.S. population. The details
of their calculations were not given. However, the Institute of
Shortening and Edible Oils (Hunter, 1985) pointed out that the
major source of hydrogenated oil in the food supply is soybean
oil, and to achieve a per capita intake of 12 g/day, 80% of total
soybean oil would need to be partially hydrogenated to an average
~ trans level of 30%, quantities and levels that are not consistent
with industry practice.
F. TRANS FATTY ACIDS IN HUMAN ADIPOSE TISSUE AS A MEASURE
OF DIETARY TRANS FATTY ACIDS
Based on the slow turnover rates of fatty acids in human
adipose tissue, several investigatorsbave proposed that the fatty
acid composition of adipose tissue reflects ,the long-term average
fatty acid composition of the diet (Beynen et al., 1980; Dayton
et al., 1966; Hirsch et al., 1960). Because trans fatty acids
are not synthesized by humans, their distribution in the adipose
tissue should indicate their average level and source in. the diet
consumed over a period of 2 to 3 years. Thomas et ale (1981) found
an average trans fatty acid content of 5.4% in the postmortem
adipose tissue of 231 subjects in the United Kingdom as compared
to 5.2% calculated for fat in the average diet. Heckers et ale
(1979) reported 1.16 to 4.20% trans-18:1 in the sUbcutaneous fat
of 16 individuals in Germany as compared to an estimated 2.34 to
3.39% in the diet. In the U.S., Johnston et al. (1957) found 2.4
to 12.2% trans isomers in the adipose tissue of 24 sUbjects and
Ohlrogge et ale (1981) reported 2.0 to 5.8% in the adipose tissue
of eight sUbjects, consistent with estimates of the percentage of
trans acids inthe diet.
Ohlrogge et ale (1981) also reported an analysis of the
distribution of double bond positions in the cis and trans octa-
decenoate fractions from several human tissueS:- They found that
the average distribution of trans double bonds in the adipose
tissue of the eight sUbjects was similar to the average dis-
tribution in commercial hydrogenated vegetable oil products as
contrasted to the distribution in ruminant fats, indicating that
the former were the major source of trans isomers in the diet.
G. OTHER DATA ON FAT AND TRANS FATTY ACID CONSUMPTION
The per capita daily fat consumption of 36,142 individuals
interviewed in the USDA 1977-78 Nationwide Food Consumption Survey
was 82.8 g (Rizek et al., 1983), considerably less than 141 g com-
puted from food disappearance data adjusted for discard of deep
31
frying fats and the separable fat of beef. Consumption was based
on a 1-day dietary recall followed by a 2-day dietary record.
Maximum fat consumption was 116.5 g/d for males aged 15 to 18
years and 81.1 g/d for females aged 12 to 14 years. Assuming
the average trans acid content of the fats consumed in 1977 was
5.6%, per capita daily trans acid consumption was 4.6, 6.5, and
4.5 g/d, respectively, for the three population groups.
Goor et al. (1985) reported the fat intake of adult
populations aged 2 0 ~ 5 9 years at nine locations in the U.S. based
on 1-day dietary recalls. Data were collected in 1972-1975.
Numbers of individuals interviewed ranged from 287 to 765 per
location, about equally divided between sexes. Males aged 20-24
years had maximum fat intakes, ranging from an average of 110 to
150 g/d by location. Maximum mean intake by location for females
also was for the group aged 20-24 years, and ranged from 90 to
110 g/d.
Kim et al. (1984) reported an average daily fat intake
of 112 g for 13 males and 68 g for 16 females aged 20-53 years,
in a 1-year study in which food intakes from Self-selected diets
were recorded daily. While this study involved relatively few
.subjects, the reported intakes are consistent with those found
in the USDA 1977-78 National Food Consumption Survey (Rizek et al.,
1983). Average daily fat intake determined by analyses of duplicate
food collections for 1 week during each of the four seasons was
96 g for males and 58 g for females. Daily intakes of saturated
fatty acids as well as oleic and linoleic acids were reported,
but not the intake of trans fatty acids.
Three recent studies have been reported in which the
trans fatty acid content in duplicate portions of the diet was
determined by analysis. Aitchison et al. (1977) analyzed the
trans fatty acid content of duplicate 3-day food collections of
self-selected diets of 11 healthy nursing mothers. Total trans
isomer content of ~ h e individual diets ranged from 1.30-8.27% of
total fatty acids. Average for the group was 5.03%. Trans-18:2
isomers were detected in the diets of five sUbjects; average
levels in the dietary fatty acids of these individuals was 0.20%.
Trans fatty acid intake of eight white healthy girls
(ages 12 to 15 years) was determined in weighed samples of food
identical to that consumed in self-selected diets in a 7-day
period (van den Reek et al., 1985). Total fat averaged 52 g/d.
Trans 18:1 content in the diets ranged from 3.5 to 8.2% with an
average of 5.3% of total fatty acids. Total dietary trans fatty
acids including ct, tc-18:2 isomers averaged 6.5% of total fatty
acids. --
In a study of the effect of hydrogenated fats in the diet
of nursing mothers on lipid composition and 'prostaglandin content
of human milk, Craig-Schmidt et al. (1984) prepared duplicate sets
32
of meals for 5-day periods. The two sets were identical except
that sources of hydrogenated fats were used in the meals for one
period and nonhydrogenated fats in the other. Sources of the
hydrogenated fats were not specified. Mean total fat in the diet
containing hydrogenated fat was 82.7 g/d. Trans fatty acid con-
sumption per day over the 5-day period as determined by analyses
of duplicate portions was 11.04, 10.24, 6.68, 6.91, and 13.72 g
with a mean value of 9.72 g/d. Trans-18:1 isomers constituted
11.8% of the total fatty acids in the hydrogenated fat diet,
considerably higher than the estimate of the average trans fatty
acid content of fat in the available food supply which includes
animal fats as well as unhydrogenated salad oils (Table 5).
H. PREDICTED LEVELS OF TRANS FATTY ACIDS IN THE U.S. FOOD
SUPPLY OVER THE NEXT 5 TO 10 YEARS
A forecast of total fat and trans fatty intake over the
next 5 to 10 years by representatives of the edible fat and oil
industry was presented at the Open Meeting of the ad hoc Review
Panel (Ace-Federal Reporters, 1985). They predicted that fat
intake is unlikely to increase and that a modest decrease, perhaps
of the order of several percent will probably occur. Two factors
support this forecast. One is the recommendations by numerous
government and other organizations that Americans decrease their
total fat consumption for health reasons. These organizations
include the American Heart Association, American Cancer Society,
National Cancer Institute, the U.S. Department of Agriculture,
and the Department of Health and Human Services. The other is
that cost-savings can be made by industry by reducing the extent
of hydrogenation, and hence trans isomer content of hydrogenated
products, because hydrogenation is an expensive operation.
Improved processing and handling operations have led to consumer
acceptance of progressively higher levels of unhydrogenated soy-
bean oil in margarines and salad oils. This trend is unlikely
to reverse. The ad hoc Review Panel agrees with this forecast.
I LINOLEIC ACID IN THE U.S. FOOD SUPPLY
Because of the relationship of the physiological proper-
ties of trans fatty acids exhibited in animal feeding studies to
the level of essential fatty acids in the diet, data on the trend
of availability of linoleic acid in the U.S. diet are presented
in Figure 4. Availability of linoleic acid has increased over the
period 1909-1982 with visible fats and oils, primarily vegetable
oil products, providing the increased availability. The linoleic
acid content of a number of margarines and spreads is given in
Table 1 (p.8); content in shortenings and frying oils is presented
in Table 2 (p.10).
33

ell
't:l
.......
.:
30

(J)
0.
.......
rn
E
ell

b.O
'-' 20

M
""""
M
.0
ell
"""" M 10
ell
:::-
<e:
40
13.6
9.0
I.
1--.,(,:,i;i:.i:i
19
:i:i
4
:o:.J
7
_l49
7.8
19.7
13.0
1967-69

To t al
:.:.:..:.:..

From Fats
and Oils
25.5
18.6
1982
Figure 4.
1909-1982.
Availability of linoleic acid in the U.S. food supply
Adapted from Marston and Welsh (1984).
34
'.
IV. BIOLOGICAL STUDIES
The biological properties of trans fatty acids and
hydrogenated vegetable oils have been studied extensively.
Reports have included studies of their metabolism and their
nutritional and physiological effects. Investigations have
been conducted with both animals and humans fed diets containing
specific trans isomers or partially hydrogenated vegetable oils.
In addition, in vitro studies of the reactions of specific enzymes
of various tissues with both positional and geometrical isomers
of octadecenoic acid have been reported. Several reviews of
these studies have appeared in the last few years (Applewhite,
1981; Beare-Rogers, 1983; Emken, 1984a;Gottenbos, 1985; Kinsella
et al., 1981).
A. ABSORPTION AND CATABOLISM
Several animal studies have compared the absorption and
oxidation of elaidic and linoelaidic acids with their respective
cis isomers, oleic, and linoleic acids. Ono and Fredrickson
(1964) administered 1-14C-labeled oleic, elaidic, linoleic, and
linoelaidic acids to Sprague-Dawley rats as nonesterified fatty
acids with simultaneously administered 3H-palmitic acid as a
"metabolic" marker. No difference was found in the intestinal
absorption of the isomeric pairs as determined from , ~ a t e s of
appearance relative to palmitic acid in the thoracic duct lymph
of cannulated adult male rats administered the fatty acids by
stomach tube. Rates of excretion of 14C0
2
in air after intra-
venous injection of the albumin-bound labeled fatty acids in
groups of four adult rats was the same for the four isomers.
All unsaturated isomers were removed at similar rates from the
plasma of a fasting dog in which the labeled compounds had been
injected intravenously.
The 1-14C-labeled trans isomers of linoleic acid fed to
groups of six Holtzman rats as components of randomly rearranged
soybean oil were catabolized to a greater extent than 1-14C-labeled
linoleic acid, also fed as a component of rearranged soybean oil
(Coots, 1964a). Initial rates of excretion of labeled C02 were
the same; however, after 10 h the excretion rate for the group
of rats fed linoleic acid fell behind and at 51 h, 64% of the
recovered 14C had been eliminated in the C02 of the linoleic acid
group while 72% of the recovered 14C had been eliminated in the
C02 of the groups fed either the trans, trans or the mixture of
cis, trans and trans, cis isomers. The difference (statistically
significant only at p <0.1, was attributed by the investigator
to the lack of essential fatty acid activity of the trans isomers
which would make them more available than the cis, cis isomer as
an energy source. Overall absorption, as indicated by residual
14C in the gastrointestinal tract of the rats at 51 h and in the
35
feces collected, exceeded 98% for all groups. These findings were
confirmed by studies in which uniformly 14C-Iabeled geometrical
isomers of 9,12-18:2 were fed to groups of eight rats as pom-
ponents of randomly rearranged soybean oil (Anderson and Coots,
1967).
In similar studies in which 1-14C-Iabeled oleic and
elaidic acid were fed to groups of three or seven Holtzman rats
as components of rearranged soybean oil, the rate and extent of
excretion as respiratory C02 were essentially the same for the
two isomers (Coots, 1964b). However, in similar studies with
uniformly 14C-labeled oleic and elaidic acids, also fed to groups
of seven or eight rats as components of rearranged soybean oil,
oleic acid was catabolized to C02 to a slightly greater extent
(70%) than was elaidic acid (65%) 51 h after feeding (Anderson
and Coots, 1967).
In human studies, no discrimination was found against
the absorption of 9t-, 12t-, 12c-, 13t-, or 13c-18:1 isomers as
compared to 9c-18:1 (Emken et al., 1979a,b, 1980). Young adult
males were fed mixtures of three triglycerides each containing
one of three different isomers labeled with a different number
of deuterium atoms. Blood samples were drawn at various times
after feeding and the isomer composition of the chylomicron
triacylglycerols was determined by mass spectrometry.
B. SHORT-TERM ANIMAL FEEDING STUDIES
This section reviews a number of short-term feeding
studies with diets containing trans fatty acids in which effects
on signs of essential fatty acid deficiency, growth rates, tissue
pathology, or interference with linoleate metabolism and prosta-
glandin production are reported. Short-term feeding studies of
the effects of dietary trans isomers on the fatty acid composition
of tissues are reviewed in the section "Deposition of Fatty Acid
Isomers in Animal Tissues" (see p.52).
Studies demonstrating that trans isomers of oleic (9t-
18:1) and linoleic acid (9c,12t-; 9t,12c-; and 9t,12t-18:2) do ,.
not have essential fatty acid (EFA) activity have been reported
by several investigators (Blank and Privett, 1963; Holman, 1951;
Holman and Aaes-Jrgensen, 1956; Privett and Blank, 1964; Privett
et al., 1955). In these studies, EFA-deficient rats were fed fat-
free diets supplemented with a trans isomer. Improvement in skin
condition was the primary criterion for EFA activity but body weight
gain as compared to controls was also considered. Negative control
animals were fed the basal diet supplemented with hydrogenated
coconut oil; positive controls received linoleate.
36
Based on body weight gain data, several investigators
concluded that trans isomers did not inhibit metabolism of
linoleate. Alfin-Slater et ale (1957a) reported equivalent
mean body weight gains in groups of 8-15 EFA-deficient male rats,
16 weeks old, fed fat-free diets supplemented with methyl linoleate
alone (20 or 50 mg), or with the same quantity of methyl linoleate
,plus partially hydrogenated triolein (250 or 500 mg) containing
32.8% trans isomers, for 8 wk. Johnston et ale (1958a) found
no evidence of growth inhibition in weanling female rats, 10 per
group, fed a basal diet containing 10% of a margarine stock (40%
trans acids) and 2% soybean oil as a source of EFA for 3 months.
Mean weight gain did not differ significantly from the control
group fed the basal diet containing 10% olive oil (7% EFA).
Mattson (1960) reported that groups of 10 male EFA-
deficient rats fed an EFA-deficient diet, supplemented with
50 mgjratjday of a trans fatty acid ester (9t-18:1; 9c,12t- and
9t,12c-18:2;, or 9t,12t-18:2) and an equal amount of linoleate
ester, had weight gains equal to those of the group receiving
only the linoleate supplement. 'He concluded that the trans
fatty acid esters did 'riot interfere with EFA utilization.
Other studies have indicated that the elaidate and
linoelaidate may interfere with the utilization of linoleate.
Privett et ale (1977) compared body weight gains of groups of
six male EFA-deficient Sprague-Dawley rats fed diets containing
(1) 5% safflower oil (72% linoleic acid) and 5% ethyl elaidate,
(2) 5% safflower oil and 5% ethyl linoelaidate, or (3) 5%
safflower oil and 5% hydrogenated coconut oil. After 24 wk,
groups 1 and 2 gained 155 37 and 158 13 g, respectively.
This was significantly less (p <0.05) than the weight gain,
202 36 g, of group 3, indicating that the trans acids lowered
growth response to linoleate in EFA-deficient rats. The fatty
acid composition of liver and serum of group 2 indicated that
linoelaidate lowered the conversion of linoleate to arachidonic
acid.
Anderson et ale (1975) investigated the effect of the
ratio t,t-18:2 to c,c-18:2 or ct,tc-18:2 to c,c-18':2 'in the diets
of rats on the conversion of linoleate to arachidonate in the
liver of rats. Ratios of 0, 0.5, 1, 2, 4, and 8 were fed in the
diets of weanling male Sprague-Dawley rats (five per diet) for
6 weeks. The trans isomers were incorporated into triglycerides
and levels in the diet (19% total fat) ranged from 0 to 18.7% of
total dietary triglycerides. No significant differences in body
or epididymal fat pad weights were observed at 6 weeks. Increasing
levels of dietary t,t-linoleic acid caused decreased levels of
arachidonate in the liver, indicating that elongation of c,c-
linoleic acid is somewhat inhibited by high ratios of t,t-18:2
to c,c-18:2 even in rats that do not have an EFA deficiency. The
investigators noted, however, that their data indicate that the
levels of t,t- and c,c-18:2 present in human diets would not be
37
expected to have an appreciable effect on arachidonic acid syn-
thesis. Levels of arachidonate in the liver were not affected
in rats fed diets containing ct,tc-18:2, but a trans-20:4 acid
was formed, presumably by elongation of c,t-18:2.
Hill et ale (1982) conducted rat feeding studies designed
to test the effects of partially hydrogenated soybean oil (PHSO)
and saturated fat on linoleate metabolism as measured by growth
rate, fatty acid composition of liver and heart phospholipids,
and 65, and 66 desaturase activities in liver microsomes. The
65 and 66 desaturases are involved in the conversion of linoleic
acid to arachidonic acid, the precursor of prostaglandins,
thromboxane, and prostacyclin. Two studies were conducted.
In study 1, PHSO providing 10.5% total trans isomers and 4.6%
total cis-18:1 isomers in the diet, was compared with hydrogenated
coconut oil (HCNO) containing all saturated fat. Both were incor-
porated in the respective diets at the 20% level. A third diet
containing 0.15% corn oil as the sole source of fat was fed as a
low-fat control. All diets were EFA-deficient and contained only
18% of the linoleate requirement. ' These diets were fed to groups
of 12 male weanling Sprague-Dawley rats for 31 wk.
In study 2, seven dietary levels (0.5-7.5% of calories)
of linoleic acid were fed at each of two levels of PHSO (5 and
10%). One group received no PHSO but linoleate at 1% of calories
to serve as a low-fat control. Diets were made isocaloric at 15%
fat by adjustment with hYdrogenated coconut oil. These diets were
fed to groups of eight male weanling rats for 14 wk.
In study 1, growth and dermal scores indicated that the
rats fed HCNO and PHSO diets were more EFA-deficient than the
group fed the low-fat diet. Levels of linoleate in liver phospho-
lipids of the PHSO group and the low-fat group were equal but total
metabolic products derived from linoleate were lower in the PHSO
group than in the HCNO and low-fat groups. In vitro 65 desaturase
assays showed no significant differences in specific activities of
liver microsomes from the three dietary groups. Activity of the
66 and 69 desaturases was significantly lower (p <0.05) in liver
microsomes from the PHSO group than in the HCNO group.
In study 2, body weights of the rats increased stepwise
as the level of linoleic acid in the diet increased. Enhancement
of physical signs of EFA deficiency occurred at linoleate levels
below 1% of calories, the dietary requirement, but w6 fatty acid
[fatty acids with the first double bond between the 6th and 5th
carbon from the methyl group carbon; included are linoleic acid
(18:2w6) and arachidonic acid (20:4w6)] metabolism was inhibited
by PHSO at linoleate levels higher than 1%. This was indicated
by the significantly lower content of w6 fatty acids in heart
phosphatides in rats fed 10% PHSO than in those fed 5% PHSO at
38
six of the seven supplementary levels of linoleate. The magnitude
of the inhibition is indicated by the 5% reduction in w6 acids in
rats fed 10% of PHSO and 7.5% linoleate calorie level as compared
to animals fed the 5% PHSO at the same linoleate level.
(1958) reported histopathological examina-
tion of organ tissues and effects on growth and dermal conditions
in rats fed diets containing 1% conjugated 9c,11t-18:2 ethyl
ester; conjugated 9t,11t-18:2 ethyl ester; 9c,12c-18:2 ethyl
ester; 9c,11t,13t-18:3; or 9t,11t,13t-18:3. Four percent ethyl
palmitate was included in the diets giving a total fat content
of 5%. Reference diets contained 5% ethyl palmitate, cottonseed
oil, hydrogenated coconut oil, or an additional 5% sucrose and
no fat. Groups of six male rats were given the diets ad libitum
for 17 wk. There were no significant differences average body
weight gains of the groups fed the conjugated fatty acids, 5%
ethyl palmitate, or the fat-free diet. However, weight gains
of these groups were all significantly (p <0.01) less than those
of the groups receiving ethyl linoleate or cottonseed oil. A
steady increase in skin scali ness was seen in all groups except
those fed ethyl linoleate or cottonseed oil. There was almost
no difference between the average skin scores of rats fed the
conjugated isomers and those fed the fat-free diet. Histological
examination of the heart, aorta, and kidney tissues of animals fed
the conjugated isomers revealed no abnormalities. Random, very
slight degeneration of the spermatogen epithelium characteristic
of EFA deficiency was seen in the testes. In the of
rats fed the conjugated trienoic fatty acids, a small number of
multinuclear cells was observed, and the number of spermatozoa
appeared to be somewhat reduced.
The effects of feeding relatively high levels of
trans, trans linoleate on weight gain, 65- and 66-desaturase
activity in liver microsomes, hematological parameters, and
serum thromboxane B2 levels were investigated by Kinsella and
associates (Bruckner et al., 1983; Shimp et al., 1982; Yu et al.,
1980). Yu et al. (1980) fed groups of 12 weanling male Sprague-
Dawley rats diets containing 5.5% fat consisting of hydrogenated
coconut oii (0.3% linoleate), trans, ester,
cis, cis-linoleic acid methyl ester, or a 50:50 mixture of lino-
elaidate and linoleate as the sole source of dietary fat. After
12 wk, rats chosen randomly from each group were killed, blood
withdrawn, and hearts removed. Average body weights of rats
receiving the four diets were 309 29, 228 26, 334 54, and
373 26 g, respectively at 12 wk. Weight gain of the group fed
the trans, trans-linoleate diet was lowest and less than that of
the group fed the EFA-deficient hydrogenated coconut oil diet.
Gains on the trans-linoleate diets and the hydrogenated coconut
oil diet were significantly less (p <0.05) than those of rats
fed the cis, cis-linoleate diet. There were no significant
differences in-average heart weight per gram body weight among
39
r
!
groups fed the four diets. The hearts from rats of the group
fed trans, trans-linoleate had the highest amount of DNA/g but
the lowest amount of protein/g.
Shimp et ale (1982) prepared diets that contained 0,
5, 10, 20, or 50% trilinoelaidate in the fat component. Each
of these diets contained 1.1% of total calories as trilinoleate
to prevent EFA deficiency, and a variable amount of hydrogenated
tallow to provide a total of 5.5% dietary fat. Diets composed
of basal diet plus 5.5% hydrogenated tallow as the sole source of
fat, and another consisting of a commercial laboratory diet were
fed for comparative purposes. These diets were fed to groups of
six weanling male Sprague-Dawley rats for periods of 4, 8, and
11 wk. Average body weights of the groups receiving diets con-
taining 1.1% trilinoleate did not differ significantly at 4 or
11 wk, nor did liver weights differ significantly among treatments
at 4 and 11 wk. The desaturase activity in liver microsomes
of the groups fed 10, 20, and 50% of trilinoelaidate in their
fat was 97, 75, and 51% of the rats fed no trans acids;
desaturase activity was unaffected by trans acid level in the
diet.
Bruckner et ale (1983) reported effects of dietary tri-
linoelaidate on hematological parameters in rats following the
same dietary protocol used by Shimp et ale (1982) and described
in the preceding paragraph. Bleeding time, platelet and red
blood cell (RBC) counts, packed RBC volume, and RBC hemolysis
were unaffected by increasing levels of t,t-18:2 up to 50% of
the dietary fat. At the 50% level of t,t-18:2, there was a
significant decrease in myeloid/erythroid marrow ratios and a
notable decrease in RBC calcium content. Also at this dietary
level of t,t-18:2, serum thromboxane B2 levels were significantly
decreased compared to controls with no apparent change in 6-keto-
prostaglandin Fla.
The results of the above studies by Kinsella and asso-
ciates indicate that the level of trans, trans linoleate found in
edible hydrogenated vegetable oils 0.5%) would not be expected
to cause any adverse effects in view of the levels of linoleate
in these hydrogenated products.
C. LONG-TERM ANIMAL FEEDING STUDIES
The most extensive animal feeding study of partially
hydrogenated fats was started by Deuel et ale (1945, 1950) and
continued by Alfin-Slater and her colleagues (Alfin-Slater et al.,
1957a,b, 1970, 1973, 1976). A diet containing 9.24% margarine
oil (trans fatty acid content not reported) as the sole source
of dietary fat was fed ad libitum to 24 generations of rats
(Wistar derived strain) with 10 rats per group (Deuel et al.,
1945, 1950). Other components of the diet were ground whole
40
,
wheat, skim milk powder, and sodium chloride. Good nutritional
status was maintained over the 24 generations as evidenced by
reproduction and lactation performances, growth rate, and longevity.
No differences were observed between rats of the 21st generation
maintained on' the margarine oil diet and rats born of stock animals
and fed a diet containing an equal level of milk fat. The margarine
oil presumably contained a substantially greater amount of trans
fatty acids than the milk fat in the control diet.
Alfin-Slater et ale (1957b) continued the work of Deuel
et ale (1945, 1950) and reported the results of feeding margarine
to rats through the 46th generation. Whole margarine was fed
at a level of 11.2% to provide 9.2% fat in the diet. This fat
contained 19% saturated fatty acids, 3.7% essential fatty acids
(increased to 6.9% beginning with the 45th generation) and 35.3%
trans fatty acids. The rats were monitored through the 46th
generation. No adverse effects were observed in body weight gain,
tibia length, reproduction performance, lactation, longevity, or
in the histopathological examination of the liver, kidney, stomach,
small and large intestines, spleen, heart, lungs, pancreas, gonads,
brain, and adrenals of animals surviving at the end of a 2-year
longevity study of the 27th generation.
The study described above was continued through 75 genera-
tions fed margarines with fat containing 35% trans acids and 6.5%
EFA through the 60th generation and 20% trans acids and 27% EFA
for generations 61-75 (Alfin-Slater et al., 1970, 1 9 7 ~ ) . Both
margarines were fed at levels that provided 11.2% fat in the
diets. No adverse findings were reported in reproductive or
lactation performance or growth rate of animals of the 47th-75th
generations, or in the histopathological examination of organs
of rats of the 75th generation that survived at 100 wk.
Two additional 25-generation feeding studies with rats
were reported by Alfin-Slater et ale (1973). Purpose of the
studies was to investigate the health effects of margarines with
relatively high levels of polyunsaturates. In the first study,
a margarine containing unhydrogenated cottonseed oil as the major
component was fed for 14 generations, then replaced by a margarine
with unhydrogenated soybean oil as the principal component. In
the second study, margarine with unhydrogenated corn oil as the
major component was fed. All three margarines contained about
20% trans acids and 27.5% EFA, and were incorporated in the diet
at a level that provided 16% dietary fat. The corn oil margarine
was also fed at the 11.2% dietary fat level for 15 generations.
Gross and histopathological examinations of organs of generations
2 and 5 were conducted at 104 wk on rats fed the cottonseed oil
diet. Gross pathology also was examined for the 10th generation
at 52 and 88 weeks . Survivors of the 15th generation at 100 weeks
also were SUbjects of histopathological examination. The investi-
gators reported no significant adverse findings that could be
attributed to the diets.
41
No significant difference was noted in weight gain or
liver weight of rats fed six different fats (15% level) with
varying contents of trans monounsaturated fatty acids (1-45%)
and trans diunsaturated fatty acids (0-6%) (Alfin-Slater et al.,
1976). Female rats were fed for two generations and males for
one.
No adverse effects attributable to diet were reported in
long-term studies with mice and rats fed diets containing soybean
oil, three hydrogenated soybean oils containing different levels
of trans isomers (15, 19, and 61%), coconut oil, or butterfat
(VIes and Gottenbos, 1972a,b). In each diet, 54% of the calories
was provided by the experimental fat and 6% by soybean oil. The
soybean oil assured that the content of essential fatty acids was
adequate. In the study with mice, each dietary group consisting
of 40 male and 40 female weanling SPF-Swiss mice was fed ad libitum
for 18 months. At 18 months, no significant difference was observed
among the dietary groups in mortality; growth rate; liver abnormal-
ities; weights of kidneys, spleens, and testicles; tumor frequency;
or type or site of neoplasms. Incidence of non-neoplastic diseases
and changes was randomly distributed among groups. Livers of
mice receiving the high-trans soybean oil diet were significantly
~ e a v i e r than the livers of mice fed the soybean oil diets, but
histopathological examination showed no abnormalities.
In the study with rats, diets were fed ad libitum to
groups of 39 to 40 each of male weanling SPF-Wistar rats, (VIes
and Gottenbos, 1972a). Differences in growth between the various
dietary groups were not significant for males or females. Differ-
ences in the total population of neoplastic tumor bearers (males
plus females) between dietary groups were not statistically sig-
nificant at p <0.05. However, among the males, the group fed the
15% trans isomer diet had a significantly higher number of tumor
bearers whereas females fed this diet had a significantly lower
number of tumors. Other pathological changes were not signifi-
cantly different among groups.
VIes et ale (1977) fed female Viennese x Alaska rabbits
diets containing the same fats (except butterfat) used in the rat
and mice studies just described. The diets contained 22.5 calorie
percent experimental fat plus 2.5 calorie percent soybean oil as
a source of EFA. Diets were fed for life of the rabbits - up to
10 years. Differences in lifespan among dietary groups were not
significant. There were no significant differences among dietary
groups in total tumor incidence. Analysis of the incidence in the
various groups as a function of time showed that dietary treatment
did not affect the induction time of tumors. Atherosclerotic
lesions were more severe in rabbits fed the diets with the lowest
amounts of linoleic acid, i.e., coconut oil and hydrogenated soy-
bean oil (61% trans, 0% linoleic acid).
42


Soybean oil and partially hydrogenated soybean oil
value 108) were included in a 2-year chronic toxicity
test of commercially used frying fats (Nolen et al., 1967).
An iodine value of 108 corresponds to a isomer content
of about 20% (Weiss, 1983). Each fat was fed at the 15% level
in a semipurified diet to 50 male and 50 female Sprague-Dawley
weanling rats. Growth as evidenced by weight gain at 2, 12, and
21 months did not differ for the unhydrogenated and hydrogenated
oils. Histological examination of 10 rats of each sex for each
diet after 2 years of feeding showed no difference attributable
to diet in incidence of fatty liver, chronic pyelonephritis,
kidney tubule mineralization, adrenal telangiectasis, alveolar
foam cells, mammary or other tumors. Liver cholesterol and kidney
sodium concentration were higher for the unhydrogenated soybean
oil diet. .
D. CARCINOGENICITY
Reviews of numerous epidemiological studies report a
positive correlation between increased dietary fat intake and
breast and colon cancer (Carroll, 1975; Doll and Peto, 1981;
Hankin and Rawlings, 1978; Miller et al., 1978). Epidemiological
data correlating cancer deaths with the type of fat in the U.S.
diet were reported by Enig et al. (1978, 1979). Significant
positive correlations (p <0.05) were found between total, breast,
and pancreatic cancer mortality and the amount of vegetable fat,
but not animal fat, available in the U.S. food supply for the
period 1909-1972. The significant positive correlation for
vegetable fat could not always be explained by the effects
of total unsaturated fatty acids, individual unsaturated fatty
acids, or total saturated fatty acids. However, incorporating
the authors' estimates of trans fatty acid consumption in multiple
partial correlation calculations showed a "significant consistent
role for trans fatty acids in positive correlations, for vegetable
fats". The inference that the correlations show that trans fatty
acids have a role in cancer deaths 'was challenged by Applewhite
(1979), Bailar (1979), and Meyer (1979). The major points raised
'were that the epidemiological data identify assocations, not
cause and effect, that most cancers are multifactorial in origin,
and that evidence for a role of trans fatty acids in
is no better, and no worse, than that for countless other dietary
components.
The long-term studies described in the previous section
have shown no carcinogenic effects of partially hydrogenated vege-
table oils as compared with the unhydrogenated product. However,
the report of Enig et al. (1978) apparently stimulated studies
comparing the tumor promoting properties of cis and trans fatty
acid isomers. Awad (1981) investigated their possible role in .
the development of Ehrlich tumor cells in inbred CBA mice. Groups
of 5-10 mice were fed diets containing either 5% free elaidic acid
43
or 5% olive oil (64% oleic acid), each plus 2% corn oil, for
4 wk prior to treatment with Ehrlich ascites tumor cells. Cells
were injected intraperitoneally to produce the ascites form of
the tumor, and intramuscularly to give the solid form. Tumor
cells grown in animals fed the trans diet took up more 3H-labeled
thymidine at various time intervals after cell injection than
those in animals Ithat received the cis diet. In two experiments
with mice bearing solid tumors, the mean survival time was less
for the groups fed the trans diets than for those fed the cis
diet. However, the difference was not ,statistically significant
at p <0.05. The investigators suggested that trans fatty acids
may be considered as promoters of tumor DNA synthesis. The
results, however,are not free of uncertainty because of the
low statistical significance and a flaw in experimental design
(Hunter, 1982). In the trans diet, the trans isomer was free
elaidic acid whereas in the cis diet, the cis isomer was oleic
acid in triacylglycerol form-.-- ---
A study by Hogan and Shamsuddin (1984) was designed
to determine the promotional effects of high dietary levels of
elaidic acid on cardinogenesis in the large intestine initiated
by intramuscular injections of azoxymethane (AOM). Groups of
40 female F344 rats were fed diets containing 25% free elaidic
or oleic acid. Both diets were deficient in linoleic acid.
Supplemental EFA was not given. Groups of 10 rats were fed a
control diet consisting of commercial laboratory feed (4.5% fat).
Thirty of the animals on each experimental diet received weekly
injections of AOM; the remaining 10 animals were injected with
saline. Twelve animals, chosen at random from each dietary group,
were killed after the 11th injection. The remaining animals were
killed at least 8 wk after the 12th injection. No tumors were
seen in any of the saline injected animals. In the trans fatty
acid group, 36.7% developed invasive carcinomas of the large
intestine compared to 23% in the cis group and 15% in the group
fed the control diet. However, differences among groups were
not significant at p <0.05.
Line 168 mammary tumor cells were transplanted into
female BALB/cAnN mice fed diets containing high levels of either
cis or trans fatty acids to investigate the effect of geometrical
isomers on the growth and experimental metastasis of mammary tumors
(Erickson et ai., 1984). The experimental fats were similar in
fatty acid composition except one contained 41.7% trans isomers.
The fats were fed to groups of 10 weanling mice at dietary levels
of 4.4 and 19.3% with 0.6 and 0.7% added corn oil, respectively,
to provide minimum EFA. A fifth group was fed 5% corn oil as
the source of dietary fat. Tumor cells were transplanted sub-
cutaneously to induce tumor growth; radiolabeled tumor cells were
injected in the lateral tail vein to assess metastasis. Animals
were killed 3 weeks after treatment. No difference in latency
or rate of primary tumor growth was observed among groups fed the
diets containing cis or trans fatty acids. At day 7 after i.v.
44
,
.. ~ i ~ W ~
~ -
injection of radiolabeled tumor cells, livers and spleens from
animals fed both 5 and 20% cis diets contained significantly more
viable tumor cells than those-fed the trans diet. It was concluded
that trans fatty acids behave similarly to the cis fatty acids
with respect to promotion of transplantable mammary tumor growth,
but trans fatty acids are less effective than cis fatty acids in
promoting blood-borne implantation and distant survival of the
tumor cells.
Selenskas et ale (1984) fed female rats, 25 per group,
diets containing cis or trans fatty acids to test the effect of
fat level and geometrical configuration on induced mammary tumori-
genesis. The trans fat contained 38% trans f a ~ t y acids, primarily
trans monoenes; the cis fat had a similar fatty acid composition
but contained no trans-isomers. A diet containing corn oil as the
dietary fat was included in the study. All" fats were fed at 5 and
20% dietary levels. Tumor "induction was achieved by administering
5 mg of dimethylbenzanthracene (DMBA) intragastrically in the rats
at 50 days of age. Rats were killed 24 weeks after administering
the DMBA. There was no significant difference (p <0.05) in tumor
incidence and yield in animals fed the cis and trans diets and
both diets were much less effective than the corn oil diets in
promoting the development of mammary neoplasia at either the 5"
or 20% fat level. The authors suggested that trans fat behaves
much like a saturated fat in the modification of tumorigenesis.
The effects of quantity, degree of unsaturation, and
trans content of dietary fat on the incidence of spontaneous
and induced tumors were compared in a study with CD-l Swiss mice
(Brown, 1981). Five well characterized vegetable fats each at
5% and 17% of the diet were fed ad libitum". Sucrose and fat were
exchanged on an isocaloric basis in formulating the diets. The
five fat types were predominately saturated fat (58.9% saturates),
monounsaturated fat (60% monoenes), diunsaturated fat (59.2%
linoleate), triunsaturated fat (4% linolenate, 57.1% linoleate)
or trans monounsaturated fat (35.4% trans monoenes). Forty wean-
ling CD-l mice of each sex and 60 weanling female C3H mice were "
assigned to each diet group. Tumors were induced in the mice by
weekly subcutaneous injection of 0.25 mg 1,2-dimethylhydrazine
hydrochloride (DMH) for 10 wk. Spontaneous development of tumors
was studied in C3H mice. Liver and mammary lesions were assessed
after feeding the diets for 17 months. Food consumption and weight
gain were not influenced significantly by dietary fat level or by
type of dietary fat. The type of dietary fat caused no particular
pattern of tumor incidence in mammary glands (spontaneous) or livers
(induced). Thus, there was no indication that diets containing
trans fat caused a different incidence of tumors than the diets
containing other kinds of fat.
45
E. ACTIVATION OF HEPATIC MICROSOMES
The S-9 fractions of livers from SK H-hr-1 mice fed diets
containing 4 or 12% corn oil, .or 12% partially hydrogenated corn
oil (60% hydrogenated) with and without supplementation with 2%
antioxidants, were used in the Ames mutagenicity assay (Salmonella
typhimurium TA-1538 strain) to test the effect of fat and
antioxidants on the hepatic activation of N-2-fluorenylacetamide
(Black and Gerguis, 1980). The antioxidant consisted of a mixture
of 1.2% ascorbic acid, 0.5% butylated hydroxytoluene (BHT), 0.2%
dl-a-tocopherol, and 0.1% reduced glutathione. Diets were fed
for 38 wk. No effect was observed with diets not containing anti-
oxidants. Significant increases in activation that did not differ
with dietary lipid were observed with S-9 fractions from mice fed
the antioxidant-supplemented diets. Differences with antioxidant-
supplemented diets resulted with mice injected intraperitoneally
with 100 mg/kg body weight of N-2-fluorenylacetamide 1 week before
sacrifice; mutagenic frequency was significantly increased with
S-9 fractions from animals fed 4 or 12% corn oil but not from
those fed hydrogenated corn oil.
In a similar study with Sprague-Dawley rats using S.
typhimurium TA 98 as the tester strain, no effects were observed
on benzo(a)pyrene (BP), 2-acetylaminofluorene (AAF), or 2-amino-
fluorene (AF) activation with S-9 fractions from animals fed, for
5-6 wk, 15% fat diets consisting of hydrogenated fat (43% trans)
or unhydrogenated high oleic safflower oil (Ponder and Green,
1985). Inclusion of 0.5% BHT in the diet increased the mutagenic
potential of AAF and AF but not BP, independently of dietary fat.
The increase for AAF was significantly greater (p <0.05%) for S-9
fraction from rats fed the hydrogenated fat diet as compared with
the high oleic safflower oil diet. In view of the formation of
trans isomers at the expense of linoleate in partially hydrogenated
fats, and the potentiating effect of oils high in linoleate con-
tent, (Black and Gerguis, 1980; Castro et al., 1978) it would be
informative to compare partially hydrogenated and unhydrogenated
oil with respect to the potentiation of hepatic microsomal activity
induced by BHT.
F. TERATOGENICITY
The teratogenic effects of fresh hydrogenated soybean oil
(iodine value 107), a similar fat used 56 hr for deep frying, and
an unhydrogenated mixture of fats and oils with a fatty acid com-
position similar to the hydrogenated soybean oil but containing
no geometric isomers, were studied by Nolen (1972). Test groups
of 25 male and 25 female Charles River rats were fed diets con-
the fats at the 15% level for two generations. The first
two litters of each generation were born naturally. During the
third pregnancy of each generation, one-half of the females were
sacrificed on day 13 of gestation and inspected for early embryonic
46
'\
-,.
~ .
deaths. The remaining females were sacrificed on day 21 of
gestation, and fetuses were examined for skeletal and soft-tissue
abnormalities. No evidence of teratogenic effects or adverse
effects on reproductive performance was associated with feeding
the fresh or used hydrogenated soybean oils.
G. DEPOSITION OF FATTY ACID ISOMERS IN ANIMAL TISSUES
In this section, selected data on the deposition of geo-
metrical and positional monoene isomers and 9,12-18:2 geometrical
isomers are presented. The focus is on the levels of isomeric
fatty acids deposited in various tissues and lipid classes, and
evidence for their accumulation or exclusion.
1. Monoene isomers
a. Incorporation and turnover in tissues
Dietary trans fatty acids have been shown to
be incorporated into most tissues and fluids of several animal
species. Except in brain, where the level of incorporation is
very low (Wood, 1979a), the degree of deposition has been found
to depend upon the tissue, dietary concentrations and, in some
cases, duration of dietary exposure. Rats fed 15% dietary fat
containing 18 or 50% trans fatty acids for 3 months incorporated
trans fatty acids into plasma, liver, kidney, heart, adipose
tissue, and red blood cells (Moore et al., 1980). Except for
triacylglycerols of kidneys, testes, and red blood cells which
required 12 weeks, the levels of trans-18:1 acids in the various
lipid fractions of all tissues from animals decreased to control
values within 8 weeks after trans fatty acids were removed from
the diets. Disappearance from adipose tissue was much slower and
control levels had not been reached at 12 weeks.
b. Effect of dietary trans acids on lipid class
levels
Dietary trans fatty acids affect the concentra-
tion of the lipid classes in animal tissues to varying degrees.
Egwim and Sgoutas (1971a,b) and Egwim and Kumrnerow (1972), con-
ducted a series of experiments in which rats were fed partially
hydrogenated soybean oil (10% or 20% dietary levels) for 10, 15,
and 20 weeks. Trans acids accounted for 48% of the partially
hydrogenated oil. Total lipids, cholesteryl ester (CE), and
triacylglycerols (TG) concentrations in the hearts of rats 15
weeks old were significantly greater than in hearts of tallow-fed
controls. CE was the only lipid class elevated in adrenal lipids
from animals fed the 10% dietary level of partially hydrogenated
fat compared to the level in controls fed 10% corn oil diets. The
cholesteryl ester level was reduced when 2% corn oil was added to
the experimental diet.
47
Liver cholesterol levels in groups of seven or eight male
Wi star rats fed a low trans diet (10% fat, 23% trans-18:1) or a
high trans diet (10% fat, 41% trans-18:1) were unaffected by the
diets either with or without added cholesterol (Sugano et al.,
1983). Liver TG concentrations were lower in animals consuming
both experimental diets than in control animals fed diets con-
taining 10% unhydrogenated soybean oil. Wood (1979a) observed
that only cholesterol and TG concentrations in" the livers of rats
fed a number of dietary fats, including partially hydrogenated
safflower oil, showed significant concentration changes compared
with levels in animals fed a commercial laboratory feed (4-6%
fat). Concentrations of TG and phosphblipids in other tissues
(brain, heart, kidney, lung, muscle, spleen, and adipose tissue)
in animals fed diets containing 15% partially hydrogenated fatty
acids (51% trans monoene) were not different from concentrations
in animals fed the laboratory chow control diet (Wood, 1979b;
Wood et al., 1979).
Studies with rats suggest that when partially hydrogenated
fats are fed at levels of 10-15% in the diet, plasma or'serum lipid
class concentrations are perturbed initially, but then gradually
return to initial concentrations over a period of a few weeks
(Wood, 1979a). This also may be true for other tissues, but not
enough data are available to reach that conclusion at the present
time.
c. Levels of trans isomers in lipids and lipid
classes of different tissues
The data from several studies show that the
trans monoene fatty acids are not incorporated into the lipids of
all tissue at the same level (Bonaga et al., 1980; Reichwald-Hacker
et al., 1979a; Wood, 1979a; Wood et ale 1977). For example, Bonaga
et ale (1980) reported that the trans fatty acid content in the
total lipid of adipose tissue was 9.6%, brain 2.3%, heart 8.5%,
kidney 3.3%, and liver 7.7% in rats fed 10% hydrogenated sunflower
oil (32% trans) in their diets for 40 days.
The levels of trans octadecenoates within lipid classes
also differ and depend upon both the tissue and the lipid class
(Moore et al., 1980; Reichwald-Hacker et al., 1979b; Wood, 1979b;
Wood and Chumbler, 1978; Wood et al., 1977). The notable exception
appears to be triacylglycerols (TG). Wood (1979b) observed that
the concentration of trans-18:1 acids in TG was approximately the
same for heart, kidney, lung, muscle, spleen, and adipose tissue
but not for liver. Reichwald-Hacker et ale (1979a) found that
the percentages of trans-18:1 in liver, heart, and serum TG were
similar. Heart and liver tissues differed from other tissues in
their relatively high contents of trans isomers in the phospha-
tidylethanolamine (PE) and phosphatidylcholine (PC) classes which
were greater than the trans content of adipose tissue (Wood, 1979b).
48
",--
" :.
.. .' 1-
1
,
d. Distribution of trans isomers among glycerol
positions
Trans fatty acids are not equally distributed
among the three positions of the glycerol moiety of TG. Reichwald-
Hacker et ale (1979b) found that trans acids are incorporated pri-
marily in the 1- and 3-positions. They also reported that trans
fatty acids were incorporated preferentially at the I-position of
liver, heart, and serum PC. These data are in agreement with the
earlier work of Wood and Chumbler (1978) that showed trans-18:1
represented more than 75% of the octadecenoates at the I-position
of liver PC and PE whereas the octadecenoates at the 2-position
of these two classes contained 10% or less trans-18:1.
e. Distribution of positional isomers among lipid
classes
As noted in the Background section, the cis
and trans octadecenoates in partially hydrogenated vegetable-0ils
contain positional isomers with the double bond located in the
66-616 positions. Wood (1979b), Wood and Chumbler (1978), and
Wood et ale (1977, 1979) determined the distribution of positional
isomers of the cis and trans octadecenoates and hexadecenoates
in the TG, PC, and PE of eight tissues of groups of three to six
Buffalo strain rats fed diets containing 15% partially hydrogen-
ated safflower fatty acids (51% trans). Reichwald-Hacker et ale
(1979a,b) reported the positional analysis of the cis and trans
octadecenoates in the total lipids of liver, adipose
tissue, testes and adrenals, and the TG and PC from liver, heart,
and serum of groups of 12 male Wistar rats fed 18% partially
hydrogenated soybean oil (12.3% trans) in their diet for 12 weeks.
The results of the two studies are similar and may be summarized
as follows: 1) Hexadecenoafes isolated from the lipid classes -in
the Wood studies contained trans isomers which were not present
in the diet. 2) The 8t-16:1 isomer represented 40-50% of the hexa-
decenoates, indicating preferential chain shortening of the 10t-18:1
isomer. 3) The distribution of the cis and trans octadecenoate
isomers in TG was similar in all tissues, with the distribution
of the trans isomers similar to that in the diet. 4) In contrast,
the percentage distribution of the trans octadecenoates in PC and
PE differed markedly from that in the diet and was dependent on
both the tissue and the lipid class. 5) The 612-, 613-, and 614-
trans octadecenoates in ,PC and PE; 611-, 612-, 613-, and 614-cis
octadecenoates in PC; and 611-, 613-, and 614-cis octadecenoates
in PC and PE of most tissues appeared in higher percentages than
in the diet, suggesting accumulation. 6) The 10c-18:1 isomer was
not incorporated to any significant extent into any tissue, but
was found in serum; there was discrimination against 10t- and
11t-18:1 in most tissues and strong discrimination against 10t-
18:1 in heart and liver tissues.
49
and (1979) reported the incorporation of
positional isomers into the liver mitochondrial total lipids
from groups of six Wistar rats fed dietary levels of 15, 20,
and 25% partially hydrogenated peanut oil (54% trans) for 3,
6, or 10 weeks. The results were similar to those of the two
studies just described: 10c-18:1 was almost completely excluded
from the mitochondrial fatty acids. The concentration of 10t-
18:1 was much lower, and that of 12t-18:1 much higher, than their
relative concentrations in the diet.
f. Distribution of 14C-labeled isomers 4-8 hours
after gavage
14C-labeled oleic or elaidic acid was admin-
istered to rats by gavage and tissue distributions of the isomer
was determined 4 hours after intubation (Guo and Alexander, 1978).
Higher levels of elaidic acid than oleic acid were found in liver,
plasma, and adrenals both in rats that had been fed either corn
oil or hydrogenated coconut oil diets. Ratios of elaidic to oleic
acid specific activities for the three organs were 2.7, 1.5, and
1.5, respectively, for rats fed the corn oil diet. Similar results
were reported by Searcey and Arata (1972) who fed the 14C-labeled
18:1 isomers to threonine-imbalanced rats.
g. Incorporation of trans isomers in membrane
lipids
Several studies indicate that dietary trans
fatty acids are incorporated into membrane lipids. and
(1981) reported the incorporation of dietary isomers into
the membranes of liver mitochondria. Alam et ale (1985) found 18%
trans-18:1 incorporated into submandibular gland plasma membranes
of rats fed diets containing 20% margarine stock (51% trans-18:1).
Decker and Mertz (1967) reported that trans-18:1 acids represented
13% of the total fatty acids in liver mitochondria and erythrocyte
stroma from animals fed isomerized olive oil (55% trans isomers).
Blomstrand and Svensson (1983) observed that trans fatty acids
displaced approximately 25% of the saturated fatty acids from
the 1-position of PC and PE isolated from cardiac mitochondria
of animals fed diets containing either partially hydrogenated
peanut oil (43% trans) or partially hydrogenated fish oils.
Dietary trans monoene fatty acids have been shown to be
incorporated into neoplastic cells and affect their growth. Trans
monoenes are incorporated into host grown hepatomas, but not with
the same distribution into lipid classes as in liver lipids (Wood,
et al . , 1977).
50
2. Diene isomers
Studies with diene fatty acid isomers have been
almost exclusively limited to the geometrical isomers of linoleic
acid. Concentrations of these isomers in partially hydrogenated
vegetable oil are much lower than those of the monoene isomers,
but they are important because of their possible involvement in
prostaglandin biosynthesis.
Approximately 15% trans isomers were found in the TG
and CE of the aortas of rabbits fed 6 g per day of elaidinized
linoleic acid and 1 g cholesterol for 84 days (Weigensberg and
McMillan, 1964). The aortic arch from animals fed this diet
contained no more atherosclerotic lesions than control animals
fed linoleic acid. Selinger and Holman (1965) reported that
dietary tt-18:2 was incorporated into liver PC; unlike cc-18:2
which was esterified at the 2-position, 80% of the tt-18:2 was
found in the 1-position.
Privett etal. (1966) determined the distribution of
tt-18:2, ct-18:2, and cc-18:2 in the liver TG and PC of rats fed
5% of these isomers in their diets for 18'to 20 days. cc-18:2
and ct-18:2 were esterified predominantly at the 2-position, while
the tt-18:2 isomer was found at 'the 1-position of PC and TG. The
positional arrangement was influenced not only by the position of
the glycerol molecule to which the isomer was directed, but also
by composition of other fatty acids directed to the same position.
The effects of dietary tt-1S:2 on the composition of rat heart
lipids have been reported by Yu et al. (1980). Hearts from rats
fed tt-18:2 contained less PC than rats fed an EFA-deficient diet.'
The fatty acid composition data of lipids from the hearts of animals
fed various diets suggested that tt-18:2 suppressed elongation and
desaturation of oleate, which accumulated in the cardiac lipid
along with 16:1 and tt-18:2.
Chern et al. (1984) fed groups of six Sprague-Dawley male
rats diets containing 10 or 50% of dietary fat (5.5%) as linoelaidate
and the remainder as hydrogenated tallow. Linoelaidate level in
liver microsomal phospholipid was 8.4% of total fatty acids after
feeding the higher treatment level for 11 weeks. There was a con-
comitant increase in linoleic acid and a decrease in arachidonic
acid levels in microsomal lipids consistent with the inhibiton of
Phospholipase activity toward endogenous phospho-
lipids was not affected.
At low intracellular concentrations (1.25 of lino-
lenic acid in cultured rat kidney cells, activity
was depressed by linoelaidic acid at concentrations of 25-200
in the culture medium (Chern and Kinsella, 1983). a-Linolenic,
y-linolenic, and dihomo-y-linolenic acids had similar but lesser
effects. However, suppression of arachidonic acid synthesis was
readily overcome by increasing the concentration of linoleic acid.
51
H. DEPOSITION OF FATTY ACID ISOMERS IN HUMAN TISSUES
Information on the incorporation of fatty acid isomers
in human tissues is derived mainly from the analysis of tissues
taken at autopsy or obtained ,incidental to a surgical procedure.
The levels found reflect the dietary levels to which free-living
subjects ~ n g e s t i n g self-selected diets were exposed in contrast
to animal studies in which dietary levels of trans isomers were
controlled and were known. Exceptions are studies of levels in
various blood components taken from human subjects administered
diets containing radiolabeled isomers.
1. Composition of tissues obtained at autopsy
Information on the trans fatty acid isomer content
of total lipid extracts from various human organs is summarized
in Table 7. The table is divided into parts: one lists studies
in which the trans isomer content was determined by the IR method,
the other by GLC. With the exception of blood components, all
tissues were taken at autopsy. Included are trans isomer levels
found in tissues collected from sUbjects in five countries.
Adipose tissue is over 90% TG and reflects the long term
average dietary fatty acid composition (see Consumer Exposure,
p.23). Total trans content in adipose tissue samples from U.K.,
U.S., German, and Israeli subjects ranged from 0.27-6.6% as
determined by GLC. The range in values for subjects of various
nationalities is similar which suggests similar levels of dietary
trans acids. The trans content of adipose samples analyzed by
IR spectrometry range from 1.5-12.2%. These higher values are
probably a reflection of higher trans values generally associated
with IR methods.
Levels of trans isomers, measured by GLC, in heart,
aorta, myocardium, liver, RBC, and plasma lipids are all in the
same range (0.27-4.1%). Trans values for these samples, which
contained about 60% phospholipids, are lower than values for the
adipose tissue samples. These data indicate that incorporation
of trans isomers into phospholipids is generally lower than the
levels consumed in the diet based on analysis of adipose tissue.
The percent trans in total brain lipid extract was 0.12-
0.88% and indicates a marked discrimination against incorporation
of trans isomers in brain as compared with other tissues. The
range in percent trans (Table 7) for various tissue samples prob-
ably reflects variation in individual diets. No individual has
been identified whose tissues contained exceptionally high trans
values.
52
T
a
b
l
e
7
.
T
r
a
n
s
I
s
o
m
e
r
s
C
o
n
t
e
n
t
o
f
L
i
p
i
d
s
o
f
H
u
m
a
n
T
i
s
s
u
e
s
~
J
N
o
.
o
f
T
r
a
n
s
J
N
o
.
o
f
S
a
m
p
l
e
b
y
I
R
S
u
b
j
e
c
t
s
C
o
u
n
t
r
y
R
e
f
e
r
e
n
c
e
b
y
G
L
C
S
U
b
j
e
c
t
s
C
o
u
n
t
r
y
R
e
f
e
r
e
n
c
e
A
d
I
p
o
s
e
1
.
8
5
-
1
1
.
6
2
3
1
U
K
T
h
o
m
a
s
e
t
a
l
.
(
1
9
8
1
)
1

I
-
5
.
9
2
3
1
U
K
T
h
o
m
a
s
e
t
a
l
.
(
1
9
8
3
b
)
A
d
I
p
o
s
e
1
.
5
-
6
.
8
6
U
S
J
o
h
n
s
t
o
n
e
t
a
l
.
(
1
9
5
8
b
)
1
.
0
-
4
.
3
1
6
G
e
r
m
a
n
y
H
e
c
k
e
r
s
e
t
a
l
.
(
1
9
7
9
)
A
d
i
p
o
s
e
2
.
4
-
1
2
.
2
2
4
U
S
J
o
h
n
s
t
o
n
e
t
a
l
.
"
(
1
9
5
7
)
1
.
9
-
6
.
6
8
I
s
r
a
e
l
E
n
l
g
e
t
a
l
.
(
1
9
8
4
)
A
d
i
p
o
s
e
1
.
6
-
9
.
3
1
5
G
e
r
m
a
n
y
K
a
u
f
m
a
n
&
M
a
n
k
e
l
(
1
9
6
4
)
2
.
0
-
5
.
8
8
U
S
O
h
l
r
o
g
g
e
e
t
a
l
.
(
1
9
8
1
)
H
e
a
r
t
4
.
6
-
9
.
3
2
4
U
S
J
o
h
n
s
t
o
n
e
t
a
l
.
(
1
9
5
7
)
1
.
2
-
4
.
1
8
U
S
O
h
l
r
o
g
g
e
e
t
a
l
.
(
1
9
8
1
)
H
e
a
r
t
1
.
0
-
8
.
2
1
4
G
e
r
m
a
n
y
K
a
u
f
m
a
n
&
M
a
n
k
e
l
(
1
9
6
4
)
0
.
4
-
1
.
2
-
5
3
F
r
a
n
c
e
R
o
c
q
u
e
l
l
n
e
t
8
1
.
(
1
9
8
4
)
A
o
r
t
a
2
.
3
-
8
.
8
2
4
U
S
J
o
h
n
s
t
o
n
e
t
a
l
.
(
1
9
5
7
)
0
.
6
7
-
1
.
5
3
2
3
G
e
r
m
a
n
y
H
e
c
k
e
r
s
e
t
a
l
.
(
1
9
7
7
a
)
A
o
r
t
a
2
.
0
-
8
.
0
5
G
e
r
m
a
n
y
K
a
u
f
m
a
n
&
M
a
n
k
e
l
(
1
9
6
4
)
1
.
4
-
3
.
9
8
U
S
O
h
l
r
o
g
g
e
e
t
a
l
.
(
1
9
8
1
)
M
y
o
c
a
r
d
I
u
m
0
.
2
7
-
1
.
5
3
2
3
G
e
r
m
a
n
y
H
e
c
k
e
r
s
e
t
a
l
.
(
1
9
7
7
a
)
C
J
1
c
..u
L
i
v
e
r
4
.
0
-
1
4
.
4
2
4
U
S
J
o
h
n
s
t
o
n
e
t
a
l
.
(
1
9
5
7
)
1
.
5
-
2
.
5
8
U
S
O
h
l
r
o
g
g
e
e
t
a
l
.
(
1
9
8
1
)
L
I
v
e
r
1
.
0
-
6
.
I
1
3
G
e
r
m
a
n
y
K
a
u
f
m
a
n
&
M
a
n
k
e
I
(
1
9
6
4
)
K
i
d
n
e
y
1
.
I
-
8
.
6
1
3
G
e
r
m
a
n
y
K
a
u
f
m
a
n
&
M
a
n
k
e
l
(
1
9
6
4
)
B
r
a
i
n
0
.
1
2
-
0
.
4
6
1
0
C
a
n
a
d
a
M
a
c
B
e
a
t
h
&
C
o
o
k
(
1
9
7
7
)
B
r
a
i
n
.
0
.
2
3
-
0
.
8
8
8
U
S
O
h
l
r
o
g
g
e
e
t
a
l
.
(
1
9
8
1
)
R
e
d
B
l
o
o
d
C
e
l
l
s
1
.
6
-
2
5
G
e
r
m
a
n
y
H
e
c
k
e
r
s
e
t
a
l
.
(
1
9
7
7
b
)
R
e
d
B
I
o
e
d
C
e
I
I
s
0
.
4
-
2
.
1
7
U
S
O
h
I
r
o
g
g
"
e
(
1
9
8
3
)
,
-
,
;
-
~
~
P
l
a
s
m
a
L
I
p
i
d
s
0
.
2
5
-
2
.
3
1
2
U
S
O
h
I
r
o
g
g
e
(
1
9
8
3
)
-
T
o
t
a
l
p
h
o
s
p
h
o
l
i
p
I
d
s
<
-
.
~
-
+
\
Data of Ohlrogge et al. (1982) on levels of trans
isomers in lipid classes found in RBC. plasma. liver. and heart
are summarized in Table 8. Tissues were obtained during autopsy
of nine adults. six males and three females. most nf whom died
of traumatic injuries. Blood samples were obtained from healthy
adult male volunteers. Individual lipid classes were isolated
from total lipid extracts of liver and heart by TLC or silica
gel column chromatography. Total percent trans values for all
phospholipid classes were equal or less than percent trans values
for triacylglycerols. These data again suggest that trans isomers
are not selectively incorporated into phospholipids. Total trans
values are similar within lipid classes regardless of the tissue
source. These data indicate little organ specificity for trans
isomer incorporation. An exception may be the low trans content
of plasma CE compared,with CE from organs and RBC.
Levels of trans isomers in placental. fetal.liver. and
adipose tissue of premature infants were present in zero to trace
amounts by IR ana+ysis (Johnston et al 1958b) and 0.1 to 0.9%
by capillary GC analysis (Ohlrogge et al 1982). These data
indicate that the placental membrane discriminates against trans-
port of trans isomers into fetal tissue. No data are available
for the cis positional isomers.
GLC analysis indicates that levels of isomers of linoleic
acid in human tissues are below 1%. Examples are: RBC-phospho-
lipids (trace) (Heckers et al 1977b); adipose tissue (0.2-0.8)
(Enig et al 1984); milk (0.4-0.7) (Craig-Schmidt et al 1984);
various major organs (0.1-0.8) (Adlof and Emken. 1985). Isomers
identified in these studies were the geometrical isomers of 9.12-
18:2.
The distribution patterns of cis- and trans-18:1 posi-
tional isomers in total lipid extracts-of adipose. heart. liver.
and aorta tissue are summarized in Figure 5 (Ohlrogge et al
1981). Tissues were collected from eight adults during routine
autopsies performed within 4 hours of death. Cis and trans mono-
enoic fractions of the methyl esters of lipids extracted from each
tissue were separated by argentation TLC. purified by preparative
GLC. and double bond distribution determined by ozonolysis.
Similar data for liver TG. liver PC. and heart PC are shown in
Figure 6 (Ohlrogge et al 1982). The distribution pattern for
hydrogenated vegetable oil represents the average of patterns
for 20 commercial margarines and cooking oils. The distribution
patterns for the trans positional isomers in tissue lipids and
hydrogenated fats are similar but not identical (Figure 6).
The amount of 10t-18:1 in heart and liver PC is about
10 percentage points less than in dietary hydrogenated oils; 11t-
18:1 rather than 10t-18:1 is the most abundant trans isomer and
both 14t- and 15t-18:I isomers are present in greater percentages
than in hydrogenated oil. However. levels for all individual trans
positional isomers are low and indicate no major accumulations.
54
T
a
b
l
e
8
.
T
r
a
n
s
I
s
o
m
e
r
C
o
n
t
e
n
t
o
t
L
I
p
I
d
C
l
a
s
s
e
s
t
r
o
m
T
i
s
s
u
e
s
o
f
N
i
n
e
U
.
S
.
A
d
u
l
t
s
(
O
h
l
r
o
g
g
e
e
t
a
l
.
,
1
9
8
2
)
L
i
p
I
d
C
l
a
s
s

R
B
C
A
v
e
(
N
)
t
R
a
n
g
e
P
l
a
s
m
a
A
v
e
(
N
)
R
a
n
g
e
L
I
v
e
r
A
v
e
(
N
)
R
a
n
g
e
H
e
a
r
t
A
v
e
(
N
)
R
a
n
g
e
T
G
2
.
1
0
(
6
)
0
.
9
-
4
.
1
1
.
7
0
(
1
1
)
0
.
6
-
3
.
1
1
.
6
0
(
9
)
1
.
0
0
-
2
.
1
2
.
2
0
(
7
)
0
.
8
0
-
3
.
0
C
E
0
.
9
2
(
6
)
0
.
1
-
3
.
2
0
.
2
5
(
9
)
0
.
0
-
0
.
7
1
.
8
0
(
4
)
1
.
3
0
-
2
.
1
1
.
5
0
(
4
)
0
.
7
5
-
2
.
3
P
C
0
.
8
5
(
6
)
0
.
3
-
1
.
4
1
.
3
0
(
1
2
)
0
.
1
-
2
.
6
0
.
7
5
(
8
)
0
.
3
0
-
1
.
6
0
.
7
7
(
6
)
0
.
3
0
-
1
.
2
c
.
n
c
.
n
P
E
0
.
9
2
(
6
)
0
.
3
-
2
.
0
0
.
9
6
(
1
1
)
0
.
2
-
1
.
7
1
.
4
0
(
9
)
0
.
2
7
-
2
.
9
0
.
6
6
(
7
)
0
.
2
6
-
1
.
2
P
S
/
P
I
1
.
3
0
(
6
)
0
.
4
-
2
.
2
0
.
9
0
(
3
)
0
.
4
0
-
1
.
2
.
0
.
7
1
(
3
)
0
.
4
0
-
1
.
3
L
P
C
0
.
9
3
(
3
)
0
.
5
-
1
.
2
1
.
0
0
(
1
2
)
0
.
1
-
2
.
3
S
M
0
.
3
7
(
6
)
0
.
2
-
0
.
6
1
.
1
0
(
1
0
)
0
.
2
-
2
.
7
F
F
A
1
.
7
0
(
6
)
0
.
8
-
2
.
8
2
.
3
0
(
1
1
)
1
.
0
-
3
.
6

T
G
,
t
r
l
a
c
y
l
g
l
y
c
e
r
o
l
s
;
C
E
,

e
s
t
e
r
;
P
C
,
p
h
o
s
p
h
a
t
l
d
y
l
c
h
o
l
l
n
e
;
P
E
,
p
h
o
s
p
h
a
t
l
d
y
l
e
t
h
a
n
o
l
a
m
l
n
e
;
P
S
,
p
h
o
s
p
h
a
t
l
d
y
l
s
e
r
l
n
e
;
P
I
,
p
h
o
s
p
h
a
t
l
d
y
l
l
n
o
s
l
t
o
l
;
L
P
C
,
I
y
s
o
p
h
o
s
p
h
a
t
l
d
y
l
c
h
o
l
l
n
e
;
S
M
,
s
p
h
I
n
g
o
m
y
e
l
I
n
;
F
F
A
,
t
r
e
e
t
a
t
t
y
a
c
I
d
t
N
=
n
u
m
b
e
r
o
f
S
U
b
j
e
c
t
s
,
tmJ ciJ
lID
30
110 ~
DB

&
4
2
I
I
I
4
2
I
I
&
4
2
~
I;
8 'S
'a
4 1
ltt
2 .i!
D
--
311
..l!!!.
I
2D &
4
II
2
D a
31
.!!!!.

211
8
4
I.
I
D a
3D
!!!!.

2D I
4
II
2
--r
"-
I
14 I. 7 9 11 13
IIlIIIIII illIlIII PasiliaI
Figure 5. Average double bond distribution in cis and trans
octadecenoate fraction of total fatty acids from human tissues,
from hydrogenated vegetable oils, and from butter (Ohlrogge
et al., 1981).
56
I,
1'3M CIS
12
10
8
6
4
2
o
12
10
8
6
4
cD
2 -

a
Q'
t2 =
...
...
10 .E
8
6
4
2
o
t2
to
8
6
4
2
o
Hydrogenated Vegetable Oil
Heart Phosphatidvlcholine
igO !J16
8 10 12 14' 16 6 8 10 12 14
Double Bond Position
30
25
20
15
10
5
0--.......--=--
30 Liver Triglycerides
25 85Q
20
.15
- 10
cD
:; 5
0 ===-
C5 30
=t: 25
;;
Q" 20
15
10
5
a __=-
30
25
20
15
10
5
o6
Figure 6. Average double bond distribution of cis and trans
octadecenoate fraction of hydrogenated vegetable-0ils and from
human tissue lipid classes (Ohlrogge et al., 1982).
57
The distribution patterns of cis-18:1 positional isomers
are difficult to compare because of the high 9c-18:1 content in
all tissues (Figures 5 and 6 ) ~ The levels of 11c-18:1 in tissue
lipids are much higher than in hydrogenated fats because signifi-
cant amounts of this isomer are normally present in dietary animal
fats and ,because it can be synthesized endogenously.
Compared with the 10c-18:1 levels in hydrogenated oils,
levels found in tissue lipids are two to three times lower
(Figure 5). These data for 10c-18:1'are consistent with findings
from rat studies (see preceding section, p.47). The distributions
of positional isomers in the total 18:1 fraction from two adipose
tissue samples analyzed by Jacob and Grimmer (1967) were similar
to the cis-18:1 data in Figure 5 for adipose tissue. Excluding
the 9- and 11-18:1 esters, levels of the other positional isomers
ranged from 0.2-1.7% of the total 18:1 methyl esters.
2. In, vivo studies
a. Incorporation of isomeric fatty acids in plasma
lipids
The incorporation into human plasma lipids
of positional 18:1 isomers relative to 9c-18:1 has been reported
by Emken (1984b, 1985) and Emken et al. (1979a,b, 1980, 1983).
Mixtures of triacylglycerols containing deuterated pairs of
positional isomers (e.g., 10c- and 10t-18:1) were fed to two
young male adults and blood samples were drawn at various times
thereafter. Plasma lipid classes were then isolated and analyzed
by GLC-MS. Studies were conducted with 9t-, 12t-, 12c-, 13c-, and
13t-18:1 isomers. The findings are summarized in Table 9, where
the incorporation of each isomer in various plasma lipid classes
is compared with that of 9c-18:1 (included in the triacylglycerols
of each experiment as an internal standard). Values are percent
differences and negative values indicate less incorporation, and
positive values greater incorporation, of the isomer than of 9c-
18:1. For all isomers, levels in plasma TG were less than that of
9c-18:1; this probably indicates more rapid removal of the isomers.
Most of the values in Table 9 are negative and indicate discrimina-
tion against incorporation. PE tended to selectively incorporate
more 9t- and 1 1 t - 1 8 ~ 1 , and PC incorporated more 9t- and 12t-18:1
isomers than 9c-18:1. There was, however, strong selective incor-
poration of both cis and trans positional isomers into the I-acyl
position of PC, and selective incorporation of 12c-18:1 into the 2-
acyl position of PC. Relative turnover rates for the 18:1 isomers
compared to 9c-18:1 are similar and indicate no accumulation would
be expected. These data for plasma lipids are generally consistent
with the organ tissue data (Table 7) which indicate that the trans
content of PL is lower than that in adipose tissue. No statistical
analyses were performed because of the limited number of subjects
and observations.
58
Table 9.
Compared
1979a,b,
Percent Difference In Incorporation of 16:1-
2
H PositIonal Isomers
to 9c-16:1-
2
H InHuman Plasma Lipids (Emken, 1964a; Emken et-al.,
1960, 1963)
LipId trans-16: 1* cls-16: 1*
Class 9 10 11 12 13 10 11 12 13
TG -12 -14 -51 -95 -76 -26 -17 -41 -51
CE -425 -247 -1565 -390 -659 -425 -124 -45 -224
PE +95 -10 +41 +2 -17 -12 +26 +29 -99
PC +35 -66 0.0 +46 -10 +29 +15 +260 -14
PC-l +263 +119 +216 +437 +401 +296 +62 +224 +169
PC-2 -62 -336 -641 -694 -3661 -41 -17 +347 -56
* Negative values IndIcate percent discriminatIon and positive values
IndIcate percent selective uptake of the 16: I Isomer compared to 9c-16:1.
c_
~
59
b. Trans fatty acids in human milk
Values for the trans fatty acid content of
human milk fat have been reported to range from 0.5-6.7% with an
average of about 3.4% (Clark et al., 1980; Craig-Schmidt et al.,
1984; Hundrieser et al., 1983; Picciano and Perkins, 1977). Trans
isomer content in human milk varies with the level in the diet
(Aitchison et al., 1977). Milk samples from eight nursing mothers
receiving hydrogenated vegetable oils (11.8% trans 18:1 isomers)
in their diets for a 5-day period had 6.5% trans isomers in their
milk lipids (Craig-Schmidt et al., 1984). The level dropped to
1.8% within 5 days after hydrogenated vegetable oil products were
removed from their diets.
I . ATHEROSCLEROSIS
1. Epidemiological studies
Thomas et ale (1981, 1983a,b) determined the levels
of trans fatty acids in body fat collected at the autopsies of
231 male residents of 10 regions of the United Kingdom. Of these
patients, 136 died of ischemic heart disease (IHD) and 95 died
of other causes. In the first study (Thomas et al., 1981), fatty
acid composition according to chain length and degree of unsatura-
tion was determined by GLC of the methyl esters of the fatty acids
of the adipose tissue samples. Total trans acids were determined
by IR analysis of the methyl esters. In the second study (Thomas.
et al., 1983a,b), the analyses were repeated using a 40 ft x 1/8
inch column packed with OV-275 that gave complete separation of
trans-16:1 + 17:0 anteiso- and good but not baseline separation
of trans- and cis-18:1 acids. Difference in total trans content
of adipose tissues of IHD cases and controls was not statistically
significant. Overall average trans content was 5.33%; difference
between means was 0.23%. However, the adipose tissue from IHD
cases had a significantly higher concentration of fatty acids with
chain lengths less than C18 [14:1, 15:0, 15:0 (branched), 15:1,
16:0 br., 17:0, 17:1] than that taken from controls, but the
concentration of fatty acids with chain length greater than C18
(20:0, 20:1, 20:2, 20:3, 22:0, 22:1) did not differ. The former
group of fatty acids are designated by the letter L, the latter
by the letter H. The L fatty acids in the U.K. diet came largely
from ruminant fats, whereas the H fatty acids were derived from
marine oils (partially hydrogenated) that are widely used as a
component in U.K. margarines. Partially hydrogenated vegetable
oils are also used as a component. in U.K. margarines. Partially
hydrogenated marine oils contribute a larger proportion of the
trans-16:1 and trans-18:1 fatty acids (t fatty acids) in marga-
rines than do partially hydrogenated vegetable oils. Partially
hydrogenated marine oils also contribute trans acids of chain
length C20 and C22.
60
The ratio, tiL for butter fat is about 0.8; for average
U.K. partially hydrogenated vegetable oils tiL is about 60. Hence,
consumption of margarines would be expected to increase the ratio
tiL in body fat. For patients, the mean value of tiL in body fat
was 2.24 0.90 and for controls, 2.06 0.79 . The values were
not significantly different (p = 0.14) suggesting no differences
in the consumption of margarines. However, tiL for both cases
and controls increased linearly with H,i.e., with margarine con-
sumption, assuming that H is a measure of consumption. Analysis
of covariance of the regression of tiL on H for cases and controls.
showed a statistically significant difference (p <0.01) between
adjusted means of tiL for cases and controls. The authors con-
cluded that the cases consumed a higher amount of hydrogenated
fats relative to ruminant fats than did the controls. Coronary
heart disease, however, is generally recognized to be of multi-
factorial origin, including such factors as hypertension, cigarette
smoking, obesity, and physical activity. These factors were not
considered by Thomas et al. (1981, 1983a,b) in their analyses.
In a study of the composition of the lipids of tissue
samples the myocardium, jejunum, and aorta taken at autopsy from
15 men with stage III and eight men with stage O-I atherosclerosis
of the aorta and coronary arteries, Heckers et al. (1977c) found
no statistically significant differences in the content of trans
fatty acids. The men were of normal weight and were 52-78 years
old. Methyl esters of the tissue fatty acids were analyzed by
GLC using a 30 m glass capillary column.
2. Dietary studies
The primary objective of these studies has been
to relate the consumption of trans fatty acids to heart disease
indirectly by determining the effects of the isomers on the serum
or plasma levels of cholesterol and other lipids. One of the
earliest studies was conducted by Anderson et al. (1961) who
compared the effects of feeding hydrogenated and unhydrogenated
vegetable oils to groups of 23 to 27 male schizophrenics that were
otherwise in good health. In one experiment, the men were divided
into four groups after 21 d on an American type diet; two received
diets which compared hydrogenated (35% trans acids) and unhydrog-
enated safflower oil and the other two, hydrogenated (37% trans
acids) and unhydrogenated corn oil. Each oil (100 g) provided
about 30% of total calories and was added to a low-fat basal diet
containing no hydrogenated fats. After 21 d, the serum cholesterol
concentrations were significantly higher (p <0.001) in groups fed
hydrogenated oils as compared with unhydrogenated oils. However,
the difference was less than that predicted by Keys' equations
(Keys et al., 1965) which relate changes in serum cholesterol to
differences in the content of saturated ,and polyunsaturated fatty
acids in the diets. The results suggested a hypocholesterolemic
effect of the trans isomers.
61
In another experiment with the same group of men, Anderson
et ale (1961) added either 98.5 g of hydrogenated corn oil (36.1%
trans acids) or a special formulated fat to the low-fat basal diet
containing no hydrogenated fats. The special fat was formulated
to have the same fatty acid composition as the hydrogenated corn
oil except that it contained no isomeric fats. After feeding
these diets for 21 d, the mean serum cholesterol concentration
(209 9.0) in the group receiving the hydrogenated corn oil diet
was significantly higher than in the group receiving the reference
fat (188 6.8 mg per 100 ml) that contained no isomers. The
increase in cholesterol. level was associated with the trans acid
content of the hydrogenated corn oil. However, although total
dienes including trans isomers in the fats of the two diets were
approximately equal, the hydrogented corn oil contained only 12.8%
EPA as compared with 22.9% EPA in the special fat. Mattson et al.
(1975) pointed'"out that the hydrogenated oil contained 12.3% con-
jugated and nonconjugated t,t-18:2; a level much higher than found
0.5%) in commercial edible hydrogenated vegetable oils. .
McOsker et ale (1962) found no significant differences in
serum cholesterol levels of groups of six male subjects, 25-44 years
old, who were fed diets containing cottonseed oil, two partially
hydrogenated soybean oils, or two blends of partially hydrogenated
cottonseed and soybean oils. Total trans acids ranged from 15.2
to 20.6% in the hydrogenated oils. The fats were fed as a com-
ponent of a liquid diet and provided 41% of the daily calories.
No statistically significant difference in plasma choles-
terol levels was found in adult males, 29 to 45 years of age, fed
liquid formula diets containing 8 or 10.6% of total fatty acids
as trans acids as compared with diets of similar fatty acid com-
position that contained no trans acids (Erickson et al., 1964).
Diets were fed for 5-week periods and each group of six men served
as its own control. Also, no significant difference in cholesterol
levels was observed when egg yolk, supplying 714 mg cholesterol per
day, was added to the diets. In the absence of added cholesterol,
plasma triacylglycerols were significantly higher in the group fed
trans acids. Average prestudy plasma cholesterol concentration
was 218.5 mg/dl. Dietary trans acids were provided as component
blends of partially hydrogenated and unhydrogenated fats.
de Iongh et ale (1965) concluded from studies of the
effects of hydrogenated dietary fats containing different levels
of trans isomers and linoleic acid that trans fatty acids have
little effect on serum cholesterol levels. The hydrogenated fats
included two soybean oils (10 and 70% trans, respectively), two
whale oils (18 and 50% trans, respectively) one fish oil (58%
trans), and- a margarine (20% trans). A series of dietary fats
with trans isomer contents of a (safflower oil), 6, la, 18, 20,
35, 50, and 70% were prepared by blending the hydrogenated oils
with safflower and/or soybean oil as appropriate. The dietary
fats were fed at the 33 calorie percent level to 72 physically
62
healthy.male subjects with oligophrenia, 20 to 50 years old, over
a period of 4 to 6 weeks. At the end of each experimental period
fasting blood samples were taken from all sUbjects and serum lipids
determined. Mean total cholesterol level was 159 mg/dl for the
six periods that the subjects received safflower oil; the level
ranged from 164 to 208 for the experimental fats. Calculation of
the expected cholesterol levels with the equation of Keys et ale
(1965) based ~ n the content of saturated and dienoic fatty acids
in the experimental fats showed good agreement with observed levels
indicating that the trans isomer content had little effect on serum
cholesterol level.
The response of serum cholesterol and triacylglycerol
levels in male subjects, 24 to 41 years old, to a hydrogenated
fat containing a relatively high level of trans fatty acid (34%
trans-18:1, 9% cis, trans-18:2, and 1% trans, trans-18:2) was
investigated by Mattson et ale (1975). Two liquid formula diets
were fed: one containing the hydrogenated fat (high trans diet),
the other alike in fatty acid composition, fat, protein, carbohy-
drate, cholesterol content, and caloric density but containing
only cis unsaturated fatty acids (high cis diet). Experimental
fats constituted 18.4% of dry solids in the diets and provided
35% of the calories; other constituents were dried egg white,
dried egg yolk, dextrose, sodium chloride, and flavoring. The
egg yolk provided 104 mg cholesterol per 100 g dry dietary solids
or approximately 500 mg per day for a 70 kg man. Before giving
the experimental diets, the average plasma cholesterol level. of
the 33 males who participated in the study was 187.8 and triacyl-
glycerol level was 108.4 mg/dl. Over a period of 7 days, the
entire group was changed to the high cis diet, and then m a i n t a i n e d ~
on this diet for 21 days. Seventeen sUbjects continued on this,
diet; the remaining 16 subjects received the high trans diet for
28 days. Blood samples were drawn twice weekly after an overnight
fast. The group receiving the high trans diet did not have a
significant change in plasma cholesterol or triacylglycerol con-
centrations relative to their period on the high cis diet, or
relative to the group continuously maintained on the high cis
diet.
Vergroesen and Gottenbos (1975) reported two studies in
which the effect of dietary trans-18:1 on serum cholesterol levels
in adults was studied by feeding liquid formula diets of similar
compositions with and without trans-18:1. Each of these diets
was also formulated with and without added egg as a source of
cholesterol. Fat provided 38% of the calories and contained
approximately 11% 16:0, 3% 18:0, 73% oleic or 37% oleic and
35% elaidic acid, and 10% linoleic acid. Fats were formulated
from coconut oil, olive oil, hydrogenated olive oil, and safflower
oil. The liquid formula diets were fed to groups of eight or nine
Trappist monks and nuns as the main source of food for 20 days.
Mean serum cholesterol level in the group fed the all cis diet
was essentially unchanged at 192 mg/dl, whereas, the level in
63
the group fed the trans diet fell from 192 to 176 mg/dl at the end
of 20 days. Addition of egg to the cis diet depressed the serum
cholesterol level about 2 mg percent, whereas, it increased the
level in the group fed the trans diet about 9 mg percent. The
difference in the changes in serum cholesterol levels were not
significant at p <0.05. The authors noted that "the course of
this experiment was not flawless - for unknown reasons the initial
serum cholesterol levels fluctuated more than normally, causing
rather large variations in these concentrations at the start of
the actual experiment."
In a second study, a liquid formula diet with 40% dietary
energy provided by the experimental fats was fed for 4 weeks to
groups of 11-12 Trappist monks and nuns (Vergroesen and Gottenbos,
1975). The fats were formulated from partially hydrogenated and/or
oil, coconut, and olive oil. Dried egg yolk
was added to provide 115 mg cholesterol per 1000 kcal. The experi-
mental fats contained 14% palmitic and stearic acids, 78% oleic or
41%, oleic and 35% elaidic, and 10% linoleic acids. Initial choles-
terol level of both groups was approximately 197 mg/dl. After
receiving the diet for 4 weeks, mean serum cholesterol level in
the trans group was 190 mg/dl, whereas that of the cis group was
185 mg/dl, a difference of about 3 mg percent. Two other groups
were fed similar diets except that the linoleic acid content of
the fat was increased to 34% at the expense of oleic acid. The
initial cholesterol level of the trans group was) 202 mg and that
of the cis group 196 mg/dl. After 4 weeks, mean serum cholesterol
level or-the trans group was 194 mg and that of the cis group was
188 mgjdl. In both groups serum cholesterol levels fell about
8 mgjdl after receiving the experimental fats.
,
Two studies with university students fed diets containing
commercial margarines (12-47.8 trans acids) did not indicate a
hyperlipidemic effect of these products (Beveridge and Connell,
1962). In the first study, 88 university students were fed a
basal diet containing 22.5% of the calories as butter for 8 days.
The diet was then modified by substituting 22.5% of carbohydrate
calories with butter, corn oil, or light margarines containing
7-26% 18:2, 11.8-23.8% 16:0, and 12-47.8% trans fatty acids.
At the end of 16 days, the plasma cholesterol of the groups fed
margarine diets did not differ significantly from each other or
from the level in the corn oil group.
In the second study of Beveridge and Connell (1962),
78 men and 19 women were fed a fat-free liquid formula diet for
8 days. They were then divided into 10 groups and fed the liquid
formula diet modified by the substitution of fat (eight margarines,
corn oil, or butter) for 45% of carborhydrate calories. After
8 days, all margarine-fed groups had significantly higher serum
cholesterol- levels than the corn oil groups, but lower than the
groups fed butter. Only three margarines led to a significant
increase in serum cholesterol compared with the level observed
64
with the fat-free diets., Average trans and saturated fatty acid
contents of these three margarines were the same as the other five
but content of 18:2 acids was somewhat less (10.2 vs. 16.1%).
The effects of feeding diets containing high levels of
trans isomers of linoleic acid were investigated by Mishkel and
Spritz (1969). Six patients, two normal, three with type II and
one with type IV hyperlipoproteinemia were fed triacylglycerols
of trans linoleic acid (trans-LLL prepared by S02 isomerization)
containing approximately 60% of the trans, trans isomer and 40%
of the cis, trans and trans, cis isomers. Trans-LLL was fed in
liquid formula diets in amounrs-varying from 25 to 100% of the
dietary fat (40% of calories), the remainder being made up of corn ~
oil or lard. Diets were fed for 7 days. Serum triacylglycerol
concentration increased in all subjects at the highest level of
trans acids fed (75 or 100% of dietary fat). Serum cholesterol
levels were less consistent. Four SUbjects had elevated levels;
one had an increased level at a low trans-LLL level but 'not at
higher levels; anp one had no change in cholesterol level.
3. Animal studies
The effect of trans fatty acids on serum cholesterol
and triacylglycerol levels in rabbits has been studied using diets
comparing elaidic acid and oleic acid (Weigensberg et al., 1961),
elaidinized olive oil and olive oil (McMillan et al., 1963),
partially hydrogenated vegetable oils and unhydrogenated oils
(Kritchevsky et al., 1954; Ruttenberg et al., 1980, 1983), and
elaidinized linoleic acid and linolenic acid (Weigensberg and
McMiilan, 1964). The general practice has been to feed rations
providing about 6 g of trans fatty acid or trans containing fat
and 1 g cholesterol per rabbit (New Zealand or Dutch Belted) daily
for 2-5 months. Rabbits fed rations containing trans-18:1 isomers
exhibited higher serum cholesterol and triacylglycerol levels
compared with control animals but the overall severity of athero-
sclerosis was the same. Ruttenberg et al. (1983) ~ o u n d no signif-
icant differences in activity of five hepatic enzymes: glucose-
6-phosphatase (microsomal), fatty acid synthease (cystolic),
malate dehydrogenase, a-hydroxybutyrate dehydrogenase, and mono-
amine oxidase (mitochondrial). Studies with elaidinized linoleic
acid indicated it to be no more cholesterolemic or atherogenic in
rabbits than linoleic acid.
The cholesterolemic and atherogenic effects of dietary
trans fatty acids in swine have been reported by several groups.
Kummerow et al. (1974, 1978) reported that groups of 12 6-month
old crossbred swine fed rations containing hydrogenated fats
(trans content 50% or 20%) for 8 months had higher serum choles-
terol levels than groups fed butterfat, beef tallow, corn oil, or
a rearranged completely hydrogenated soybean-cottonseed oil. All
fats were fed at 16.7% level in the diet., Corn in the basal diet
65
provided an additional 3% f a t ~ The groups fed partially hydrogen-
ated fat had the highest serum cholesterol level, however, only
the serum cholesterol level for animals fed corn oil was signif-
icantly lower at p <0.05. The groups fed hydrogenated fat also
had the largest amount of atherosclerotic involvement in the
abdominal aorta (expressed as a percentage of the weight of the
total intimal surface), but not significantly more than that of
the other groups. This study has been criticized because of the
marginal amounts of EFA in the trans fatty acid diets and because
the basal ration was low in selenium and protein and contained
high levels of vitamin 0 (Heckers et al 1979; Kaunitz, 1976;
Royce et al., 1984).
Jackson et al. (1977) fed 8-week-old Yorkshire piglets
rations containing 11.4% hydrogenated fat (13% trans isomers) with
and without 0.4% added cholesterol, or 11.4% beef tallow. The
basal ration contained 3% corn oil. After 4 months the animals
were sacrificed and the aorta examined. Grossly visible fatty
streaks and fibrous plaques were not found. Light and electron
microscopy revealed atherosclerotic lesions in the distal abdominal
aorta but no significant differences were observed among groups
fed the different diets.
In an experiment designed to study the effect of inter-
actions among saturated, unsaturated, and cis and trans fatty
acids on tissue lipids, seven fat blends were fed to 5-6 week old
cross-bred pigs (Duroc-Yorkshire-Hampshire) for 10 months (Elson
et al., 1981; Rapacz et al., 1977). The blends contributed 17%
fat to the rations; an additional 2% was present in the components
of the basal rations. Three blends contained 0, one 16.7, two 25.
and one 50% trans fatty acids in the form of hydrogenated soybean
oil. After 10 months, serum cholesterol levels in all groups were
elevated about 50%. However. examination of the aortas showed
that only five of 64 swine had aortic intimal lipidosis in excess
of 2% of the surface area. The five swine were from groups fed
blends contain{n 0, 25, or 50% trans isomers. Each of the animals
carried the Lppp gene. These animals were demonstrated to have a
tendency to develop more extensive fatty streaking when they were
fed a high-fat ration low in cis unsaturation (Rapacz et al 1977).
The investigators concluded that the results of their study pro-
vided little support for the suggestion that dietary trans fatty
acids cause the development of aortic intimal fatty streaking and
. fibrosis.
In a recently reported study (Royce et al., 1984), weanling
male Poland-China swine were fed, for 27 weeks, either safflower
oil, lard, or an 82:18 blend (31.8% trans) of hydrogenated with
unhydrogenated soybean oil. The experimental fats contributed
20% fat to the ration; ground corn contributed an additional 1.4%.
Examination at necropsy revealed no significant differences at
p <0.05 among dietary groups in the extent of atherosclerosis
in coronary arteries or in the synthesis of thromboxane by blood
components.
66
The atherosclerotic effect of dietary trans fatty acids
in Vervet monkeys was reported by Kritchevsky et ale (1977). The
monkeys were fed semipurified diets (14% fat) containing either
of two partially hydrogenated fats (23 or 43% trans), or a 72:28
mixture of olive oil and corn oil (control diet). After 12 months,
serum cholesterol levels in the groups fed the low trans, high
trans, and control diets were 134 12, 163 13, and 166
8 mg/dl, respectively. There were no significant differences
in incidence or severity of atherosclerosis.
J. COMPETITIVE INHIBITION OF LINOLEIC ACID METABOLISM
BY ISOMERIC UNSATURATED ACIDS
1. Inhibition of the 65- and 66-desaturase enzymes
Individual positional isomers (63-616) of trans-18:1
acids have been demonstrated to inhibit the desaturation of linoleic
to-y-linolenic acid (66-desaturase) and eicosa-8,11,14-trienoic.
to arachidonic acid (65-desaturase) by rat microsomes from
EFA-deficient rats (Mahfouz et al., 1980a). To compare the effect
of double bond desaturation of linoleic acid with 66-
desaturase was measured in presence of each trans-18:1 isomer
at an inhibitor/substrate ratio of 3:1. Similar studies were con-
ducted with 65-desaturase at an inhibitor/substrate ratio of 6:1.
At these ratios, some of the trans-18:1 isomers gave about 50%
inhibition. Comparison of inhibitions was made with controls
in which the total 18:2w6 or 20:3006 substrate concentrations
(no inhibitor) were equal to the sum of the inhibitor and sub-
strate concentration used in the inhibition experiment. All
trans-18:1 positional isomers inhibited 66-desaturase; most
strongly were the trans-63, -67, and -615-18:1
isomers. For the desaturation of 20:3 to 20:4, the 63-, 69-,
613-, and 615-trans isomers were the most effective inhibitors
Whereas the 64-, 65-, 68-, 612-, 614-,. and 616-trans
isomers had little or no inhibitory effe6t.
Individual cis positional isomers of octa-
decenoic acid also were found to be inhibitory to the 66- and
65-desaturases in rat liver microsomes from EFA-deficient rats
(Mahfouz et al., 1981). The experimental protocol was similar
to that used by Mahfouz et ale (1980a) in their study of trans
positional isomers. Maximum levels of inhibition for the
desaturation of linoleic acid (66-desaturase) and eicosa-
8,11,14-trienoic acid (65-desaturase) were obtained with the
67- through 611-cis-18:1 isomers. The strongest inhibitor for
both desaturases-was the 8c-18:1 isomer.
67
2. Desaturation of trans-18:1 isomers
Mahfouz et ale (1980b) demonstrated the desaturation
of positional isomers of trans-18:1 fatty acids to cis,trans- and
cis,cis-18:2 acids by the liver microsomes of EFA-deficient rats.
Trans-18:1 isomers (e.g., Xt-18:1) were desaturated by
to form the 9c,Xt-18:2 fatty acids. Conversion rates were highest
for and isomers. Rates for
and were one-half or less of those of tHe former group.
The desaturation to cis, trans isomers 'was accompanied by a trans
to cis isomerization yielding cis, cis-18:2 from trans-18:1 isomers.
The and isomers gave the highest rates of cis, cis
diene formation; rates for the through and for-
the were very low.
K. ACTION OF ENZYMES IN VITRO
1. a-oxidation
Rate of in vitro oxidation of elaidic acid by rat
heart mitochondria was reported to be lower than that of oleic
acid (Hsu and Kummerow, 1977). These investigators also reported
that heart mitochondria from Holtzman rats fed diets containing
hydrogenated soybean oil (46.8% trans-18:1) had lower respiratory
activity than mitochondria from rats fed corn oil diets.
Willebrands and Van der Veen (1966) concluded that rat
hearts perfused with albumin complexes of oleic or elaidic acid
removed oleic acid from the perfusate more rapidly than elaidic
acid suggesting more rapid oxidation of the cis isomer.
The amount of uniformly 14C-Iabeled oleate oxidized to
14C02 and soluble intermediates by rat liver mitochondria exceeded
the amount of uniformly labeled elaidate oxidized at
all incubation times in experiments reported by Anderson (1967).
Studies with the oxidation of the acid isomers
labeled in the C-l or C-10 positions indicated that the difference
in oxidation rates was due to a slower oxidation of the alkyl chain
on the methyl side of the trans double bond of elaidic acid.
Lawson and Holman (1981) found the oxygen uptake rates
of the trans positional isomers through acid
to be lower than that of oleic acid in studies with isolated rat
heart and liver mitochondria. Hearts and livers were removed from
adult Sprague-Dawley rats fed a nutritionally adequate, semipuri-
fied diet containing 2% corn oil. The fatty acids were converted
to their CoA esters prior to oxidation. Liver mitochondria oxi-
dized most even-positional trans isomers significantly more rapidly
than adjacent odd-positional isomers. This pattern was observed
for heart mitochondria only for isomers with double bond positions
through Heart mitochondria oxidized nearly all the trans
68
"":'L
isomers significantly more slowly than stearoyl-CoA; in contrast,
liver mitochondria oxidized the even-positional trans isomers
nearly as rapidly as stearoyl-CoA. These results indicate that
all trans fatty acids are not equivalent to the corresponding
saturated fatty acid in their metabolism rates in rats.
In contrast to the results of studies with isolated heart
mitochondria, Menon and Dhopeshwarkar (1983) found that homogenates
of rat hearts oxidized oleic and elaidic acids at equal ~ a t e s .
This relationship was observed for hearts from Wistar rats fed
diets containing either 5% hydrogenated corn oil (52.2% trans-
18:1), 5% commercial oleic acid (67% cis-18:1; 8.6% 9c,12c-18:2),
or a stock diet. Oxidative activity for the 18:1 isomers of
heart homogenates from rats fed the trans diet was greater than
homogenates ,from rats fed the cis or stock diets.
Lanser et al. 1985) reported that homogenates of human
heart tissue converted 4C-labeled oleic and elaidic acids into
S-oxidation intermediates at equal rates. Human heart tissue
was obtained from five males and six females during heart bypass
surgery. The supernatant obtained from the centrifugation of
tissue homogenates at 500-600 g for 5 minutes was used in the
oxidation rate studies. Homogenates from five male and six
female rats prepared in identical manner, oxidized oleic acid
35-40% faster than elaidic acid.
L. TRANS FATTY ACIDS AND IMMUNE FUNCTION
Altering the fatty acid composition of lymphocytes
in culture has been shown to modulate immune function (Meade
and Mertin, 1978; Traill and Wick, 1984). Both enhancement
and suppression of B- and T-cell responses h ~ v e been reported
depending upon how fatty acids were introduced into the culture.
Numerous direct observations of cell cultures support the hypoth-
esis that fatty acids, especially polyunsaturates, take part in
the immunoregulatory mechanisms and that the conc,entration of
these fatty acids could affect the interaction of lipids and
binding glycoproteins of the membrane of cells involved in
immune responsiveness.
Additional observations also support the concept that
certain fatty acids may play an imrnunomodulatory role. First,
lymphocyte stimulation by antigen or mitogen alters the membrane
fatty acid composition (Ferber et al., 1975). Activation of the
acyl CoA lysolecithin acyltransferase enzymes increases the incor-
poration of unsaturated fatty acids, such as arachidonic acid,
into phospholipids (Trotter et al., 1982). Second, changes in the
levels of dietary fats fed to rodents are re.flected by changes in
the fatty acid composition of lymphocytes (Erickson et al., 1983;
Meade et al., 1978; Tsang et al., 1976). Third, malignant lympho-
cytes grown ,in media containing fatty acids incorporate those fatty
acids into the cell membrane (Mandel and Clark, 1978; Mandel et al.,
69
1978). However, addition of fatty acids in organic solvents to
lymphocyte cultures as has been frequently done is nonphysiological;
moreover, changes in immune response after manipulation of fatty
acids in vitro may not relate to changes in vivo because of the
complexity of cellular and humoral interactions observed in vivo.
In view of these considerations, experiments have been conducted
to assess the influences of dietary fat concentration and satura-
tion on immune response. Most of the diets used have not been
reported to contain trans fatty acids. Generally in those experi-
ments, high levels of dietary fat, particularly cis unsaturated
fats, have suppressed lymphocyte functions when essential fatty
acid requirements were met whereas low levels of fat augmented
those immune responses. Essential fatty acid deficiency, however,
may also suppress selected lymphocyte functions.
Compared with the number of studies in which cis unsatu-
rated and saturated fatty acids have been used, very few-studies
of immune function have been conducted with trans fatty acids.
Moreover, most of the latter studies have been with cell cultures.
Nevertheless, a'review of in vitro trans fatty acid 'effects on
lymphocytes and macrophages should be helpful for hypothesis
formulation and understanding how dietary trans fatty acids may
influence immune responsiveness.
1. Modulation of the lymphocyte response in vitro
by exogenous trans fatty acids
Trans fatty acids are incorporated into cultured
lymphocytes including the plasma membrane (Klausner et al.,
1980a,b; Mandel and Clark, 1978; Mandel et al., 1978). Lympho-
cytes, however, maintain a relatively constant degree of fatty
acid saturation in the membrane phosphat ides when grown in the
presence of unsaturated fatty acids (Mandel and Clark, 1978).
For example, no difference was observed in electron spin resonance
(ESR) motional parameters between trans-18:2 substituted plasma
membranes and unsubstituted controls suggesting that incorporation
of exogenous fatty acids induces compensatory changes in the mem-
brane lipid composition (Poon and Clark, 1982). ESR, however, may
not be sufficiently sensitive to detect changes due to modifications
of the acyl chain composition in complex biological membranes.
Although controversial, the use of 1,6-diphenyl-1,3,5-hexatriene
(DPH) allows assessment of environmental constraints upon the
rotation of the DPH molecule inserted between the acyl chains
of membrane phospholipids. In contrast to cis unsaturated fatty
acids which incorporate into the lymphocyte membrane and reduce
the polarization of DPH, trans (trans-18:1) and saturated (18:0
and 19:0) fatty acids had no effect on DPH polarization (Klausner
et al., 1980a). Based on differences in fluorescent lifetimes,
Klausner and colleagues (1980a,b) believe the cis unsaturated
free fatty acids partition preferentially into the fluid-phase
70
lipids whereas trans unsaturated and saturated fatty acids
partition into the gel-phase lipid. However, McVey et ale (1981)
have shown that extensive changes in the fatty acid composition of
malignant murine T-cells did not lead to any biophysical changes
that could be detected by DPH. They concluded that fatty acid
alterations caused dramatic changes in the function of plasma
membrane-associated enzyme systems.
Cross-linking and rearrangement of cell surface antibody
receptor molecules may be an event that triggers an immune response.
Receptor proteins in the plasmalemma are mobile and may be cross-
linked by divalent ligands such as antibodies. These form aggrega-
tions called patches which coalesce into a cap which is shed or
internalized by the cell. Cap formation may be an energy-requiring
process that involves microfilaments. Although low levels of cis-
free fatty acids intercalated into the lymphocyte plasma membrane
were associated with inhibition of capping, no effect was observed
with trans (trans-18:1) or saturated (19:0) fatty acids (Klausner
et al., 1980b). Corps et ale (1980) believe that the inhibition
of capping may be accounted for by quantitative effects of the
fatty acids on ATP levels. However, under the same conditions
as capping inhibition, there appeared to be neither ATP inhibition
nor respiratory uncoupling (Hoover et al., 1982).
Trans fatty acids in culture media may influence some
immunological responses. For example, fatty acid-altered lympho-
cyte membranes lead to changes in the cytolytic capacity of
restimulated memory cells. Unsaturated fatty acids, regardless
of the position or geometry of the isomer, led to an increase in
cytolytic capacity whereas saturated fatty acids led to decreased ,.
cytotoxic function (Gill and Clark, 1980). The fatty acid a l t e r a - ~
tion of cytolytic capacity did not seem to be simply altering the :.
general metabolic activity of the lymphocytes such as rates of DNA
and protein synthesis. These effects occurred in the absence of
cell division; therefore, the net result of fatty acid alteration
was an increase (unsaturated fatty acids) or decrease (saturated
fatty acids) in the frequency population of cytotoxic T-Iymphocytes
capable of lysing target cells (Bialick et al., 1984).
Tumor target cells cultured with certain fatty acids in
the media may be more susceptible to complement-mediated cytolysis
than control cells cultured without fatty acid supplements. Gener-
ally, fluidity of the tumor cell target regulates the cell's sus-
ceptibility to complement attack; increased susceptibility corre-
lates with decreased cholesterol/phospholipid ratios. Hepatoma
cells which were enriched with cis-18:2 and cis-18:1 and contained
an increased amount of those fatty acids in their cellular lipids
were more susceptible to complement-mediated cytolysis than con-
trol cells (Schlager, 1979; Schlager and Ohanian, 1979; Yoo et al.,
1980). In contrast, culture of tumor target cells in media con-
taining other fatty acids such as 19:0, trans-18:1, and cis-18:3
71
did not affect cytolysis by antibody and complement (Mandel
et al., 1978). However, substitution of those same fatty acids
into the plasma membrane of the tumor target influenced the rate
of patching of histocompatability (H-2) surface antigens. There-
fore, alterations of the fluid state of the membrane that affect
lateral movements of the surface proteins may not affect the cyto-
lytic process.
2. Modulation of macrophage function in vitro
by exogenous fatty acids
Macrophages are a highly diverse population of cells
capable of executing or modifying a number of important biological
functions. One major function is to protect the host from infec-"
tious agents or their products, particularly before the establish-
ment of a specific immune response. In addition, the macrophage
cooperates in all antigen-specific lymphocyte responses. ThUS,
all specific immune responses seem to require macrophages to
be present; they function in presenting antigens to specific
lymphocytes.
Unsaturated fatty acids can be rapidly incorporated
into macrophages in vitro and can selectively alter the fatty
acyl composition of the macrophage phospholipids (Schroit and
Gallily, 1979). Although trans-18:1 and 19:0 may be incorporated,
no significant alterations in the ratio of saturated to unsaturated
fatty acids (Erickson et al., 1983) or cholesterol. to phospholipid
have been observed (Mahoney et al., 1980; Schroit et al., 1976).
Incorporation of either trans-18:1 or 19:0 fatty acids decreased
flUid-phase pinocytosis and receptor-mediated phagocytosis (Mahoney
et al., 1977). These investigators have suggested that the degree
of lipid flUidity of the macrophage membrane influenced receptor-
mediated phagocytosis as well as the ability to the cell to interi-
orize its plasma membrane., Other workers (Schroit et al., 1976)
have found that macrophages containing a high proportion of cis
unsaturated fatty acids (cis-18:1) have higher phagocytic poten-
tials than macrophages containing higher proportions of trans-
unsaturated fatty acids (trans-18:1). In contrast, they found
no correlation between nonreceptor-mediated macrophage phagocytic
activity and membrane fluidity.
3. Modulation of immune status by dietary trans
fatty acids
Very few studies have directly addressed the effect
of dietary trans fatty acids on any parameters of immune function.
In one stUdy (Clifford et al., 1983), rats were fed 8% triacyl-
glycerols including partially hydrogenated soybean oil (trans-
18:1, 43.2%; trans-16:1, 0.1%) and trielaidin, (trans-18:1, 35.2%).
Differences in mitogen-induced lymphocyte transformation among
72
rats fed the various dietary triacylglycerols were unrelated to
the saturation of the lipids; the response correlated negatively
with total lipid absorbed. Mitogen responses, however, were
measured in serum-free medium with a 72 hour assay. Because
it takes about 24 hours for lymphocytes to regain their char-
acteristic fatty acid profiles after the surrounding lipid con-
centration is changed, assessments of in vivo lipid modulation
may be only reflected in the first 24 hours of the assay (Mandel
and Clark, 1978; Mandel et al., 1978). Moreover, changing the
serum in culture media may lead to changes in the fatty acid pro-
files of the lymphocytes (Loomis et al., 1983). In contrast to
a lack of differences in mitogen responses, serum IgG was lower
in rats fed trielaidin and partially hydrogenated soybean oil
than rats fed corn oil. Except for tripalmitin and tristerarin,
dietary lipids significantly altered the fatty acid profiles
of total lipids in plasma, spleen, and thymus.
M. EFFECTS OF ISOMERIC FATS ON MEMBRANE PROPERTIES
AND CELL FUNCTION
The previous section discusses the effects of isomeric
unsaturated fatty acids on immune function; this section reviews
studies with microorganisms, tissue cultures, and physical measure-
ments of membrane properties as affected by isomeric fatty acids.
Numerous approaches have been used to investigate the
impact of isomeric fatty acid structures on cell function and
membrane properties. A series of studies indicate cell growth
a ~ d division are inhibited when microorganisms or liver cells
are cultured in media containing various single cis or trans
positional fatty acid isomers (Graff and Lands, 1976; Ohlrogge
et al., 1976; Steele and Jenken, 1972; Tsao and Lands, 1980;
Vandenhoff et al., 1975; Walenga and Lands, 1975; Wennerstrom
and Jenken, 1976). However, there appears to be no correlation
of specific physical properties such as melting point or double
bond position of various 18:1 isomers with loss of cell viability.
Other studies indicate that membrane permeability of
single cell organisms is affected by the introduction of fatty
acid isomers (McElhaney et al., 1970). Although 9t- and 11t-18:1
as fat sources inhibited the growth of Escherichia coli, addition
of higher levels of cyclic AMP to the culture medium restored
normal growth. This indicates that the trans-18:1 isomers pro-
vided sufficient membrane fluidity to support cell growth and
suggests that the isomers may have influenced cell metabolism
rather than only membrane properties (Tsao and Lands, 1980).
Single cell organisms appear to be able to regulate the incor-
poration of isomers into phospholipids so as to achieve desired
membrane physicochemical properties (Saito and McElhaney, 1978).
Hill et al. (1983) showed that 9t-18:1 ~ i s similar to 18:0 and 16:0
in its modulation of glucose-6-phosphate translocase activity.
73
Decker and Mertz (1967) reported increased swelling rates
of liver mitochondria and increased rates of induced hemolysis of
erythrocytes from rats fed dietary trans acids (elaidinized olive
oil, 43.5% elaidic acid, 0.4% linoleic acid) as compared with
erythrocytes from rats fed diets containing olive oil. A more
recent study indicated that dietary trans acids (partially
hydrogenated soybean oil containing 43.5% trans 18:1, 10.2%
linoleic acid) did not significantly (p <0.05) increase rat RBC
osmotic fragility or membrane permeability to water as compared
with RBC from rats fed high oleic acid safflower or corn oil
diets (Benga et al., 1984). Osmotic fragility of RBC from rats
fed either partially hydrogenated soybean oil or high oleic acid
safflower oil was greater (p <0.05) than that of RBC from rats
fed corn oil.
FluidIty of submandibular gland plasma membranes of rats
fed 20% margarine stock (51% trans) as measured by fluorescence
polarization of 1,6-diphenyl-1,3,5-hexatriene, was less than that
in rats fed 20% corn oil (Alam et al., 1985). However, distribu-
tion of fatty acids in the membranes was similar to that in rats
fed EFA-deficient diets. Incorporation of 1% corn oil in the
margarine stock markedly changed the fatty acid composition of the
plasma membranes and increased membrane fluidity. Other studies
with mitochondria (Combe et al., 1980), and tissue cultures (Saito
and McElhaney, 1978) indicate membrane permeability is increased
by the incorporation of isomeric unsaturated fatty acids.
Various physical measurements have been made on intact
cell membranes and model membranes by NMR (GaIly et al., 1979;
Seelig and Waespe-Sarcevic, 1978), differential scanning colorim-
etry (Baldassare et al., 1976), electron spin resonance (Guyer
and Bloch, 1983), and x-ray diffraction (Esfahani et al., 1971).
NMR data indicate that 9t-18:1 incorporated into phospholipids did
not influence the molecular organization of lipid by layers (GaIly
et al., 1979; Seelig and Waespe-Sarcevic, 1978). Esfahani et al.
(1971) used x-ray diffraction to show that liquid crystalline-to-
gel transition temperatures of phospholipids were similar for
E. coli cultured in media containing elaidic or linoleic acid
as the source of fat.
Fatty acids are important structural constituents of cell
membrane phospholipids and are concentrated in separate fluid and
solid-like domains in the membrane rather than uniformly distributed
in a bilayer (Klausner et al., 1980a). Phospholipids generally
contain two different fatty acids rather than two fatty acids of
the same structure. An exception is 16:0-PC in lung tissue which
has a special role as a surfactant. In general, specific pairing
of different fatty acids (Baldassare et al., 1977), (e.g., 16:0/
18:2-PC) in phospholipids has an important role in conserving
certain membrane functions by ensuring fluidity. Thus, pairing
fatty acids with the same structure (e.g., 16:0/18:0-PC) results
in liquid crystalline interactions that are normally undesirable
74
(Baldassare et al., 1976). Hence, one of the reasons for selective
acylation of PC by phospholipid acyltransferases may be to ensure
that pairing does not occur by placing saturated fats in the I-acyl
position and polyunsaturated fats in the 2-acyl position of phos-
p h o l i p i d s ~ Indeed, changes in cell function associated with the
introduction of an isomeric fatty acid at a relatively high con-
centration may result ,from the forced pairing of that isomer.
However, at a level of 5% trans isomer, which is the level normally
found in phospholipids, the probability of pairing of the same
isomeric structure is near zero, especially since the trans-18:1
isomers consist of six or more positional isomers, and trans
isomers are strongly excluded from the 2-acyl phosphatidylcholine
position. From this viewpoint, the presence of 5% trans isomers
in organ phospholipid membranes provides a greater diversity of
fatty acid structures which can be utilized by acyltransferase
reacylation reactions for retailoring the physical properties of
membrane phospholipids.
75
v. CONCLUSIONS
A. CONSUMER EXPOSURE
1. Trans fatty acid availability in the U.S. food
supply has been essentially constant over the past 2 decades
at about 8 g per capita daily. The estimate of the Review Panel
and that of the edible oil processing industry, while derived
from different data, are in relatively close agreement.
2. The estimated average trans fatty acid content in
the U.S. food supply, 5.4-5.8% (fat basis), is consistent with the
trans fatty acid content found in the adipose tissues of humans.
Fatty acids in adipose tissue have low turnover rates and reflect
long-term average fatty acid composition in the diet.
3. Based on food consumption of individuals reported
in the USDA 1977-78 .Nationwide Food Consumption Survey, maximum
mean fat consumption by sex/age group was 116.5 g/d for males aged
15-18 years and 81.1 g/d for females aged 12-14 years. Assuming
5.6% trans isomer content in dietary fat, the corresponding trans
fatty acid consumption was 6.5 and 4.5 g/d, respectively.
4. Maximum trans fatty acid content reported in U.S.
diets was in experimental diets prepared for lactating women in
which hydrogenated fat products were used as the source of fats
in all meals prepared for a 5-day period. Average trans-18:1
isomer content of the meals was 11.8%; highest daily trans isomer
consumption was 13.7 g.
5. A continuation at present levels or modest decrease,
perhaps of the order of several percent, in total fat intake and
trans fatty acid intake is predicted for the next 5 to 10 years.
Recommendations of government agencies and medical associations
for decreased consumption of fats, and cost savings to industry
through increased use of unhydrogenated vegetable oils or oils
with lower degrees of hydrogenation, are the basis of this
forecast.
6. Although this report is concerned only with the
health aspects of trans fatty acids as constituents of hydrogen-
ated vegetable oil products and ruminant fats, cis octadecenoate
positional isomers are also present in significant amounts in
hydrogenated vegetable oils. Thus, it is essential to consider
the possible effects of cis as well as trans isomers in inter-
preting the biological properties of hydrogenated fats.
Preceding Page Blank 77
B. TOXICOLOGICAL EFFECTS
1. Long-term and multigeneration rodent and rabbit
feeding studies with diets containing hydrogenated vegetable oils
having relatively high trans fatty acid contents have not shown
carcinogenic properties or other pathological effects attributable
to their trans isomer content.
2. Examination of pregnant rats fed hydrogenated
vegetable oil for early embryonic deaths, and fetuses at birth
for malformations provides no evidence for teratogenicity of
trans fatty acid isomers.
3. Diets containing trans or cis fatty acids (18:1 or
predominantly 18:1 isomers) were found nor-to have statistically
significant differences (p <0.05) in promotional effects on the
development of colon or mammary tumors initiated by chemical
carcinogens, or on the incidence of spontaneous tumors in rodents.
No differences in latency or rate of primary tumor growth of trans-
planted mammary tumor cells were observed in mice fed these diets.
4. In mice, dietary corn oil but not hydrogenated corn
oil (60% hydrogenated), potentiated the activation of the S-9 liver
fraction by 2-acetylaminofluorene (AAF) in the Ames Salmonella
mutagenicity assay of AAF. Dietary hydrogehated fat (43% trans
isomers) potentiated the activation of the S-9 rat liver fraction
by butylated hydroxytoluene (BHT) in the Ames assay of AAF muta-
genicity but not benzo(a)pyrene or 2-aminofluorene mutagenicity.
The results suggest that hydrogenated fat may potentiate the
activation of the diacetylase system but not the mixed-function
oxidase system in rodent liver microsomes.
C. METABOLISM
1. Absorption and catabolism
a. Animal studies indicate there is no discrimin-
ation against the absorption of elaidic and linoelaidic acids as
compared with their cis counterparts. Human studies show no dis-
crimination against absorption of the 9t-, 12t-, or 13t-18:1 isomers
as compared with oleic acid.
b. Rate and extent of oxidation of elaidic and
linoelaidic acid isomers to C02 in rats are essentially the same
as those of their cis counterparts.
c. Although most in vitro studies show that rat
heart mitochondria oxidize elaidic acid more slowly than oleic
acid, experiments with homogenates of human heart tissue indicate
no differences in rates of oxidation to C02 and S-oxidation inter-
mediates.
78
2. Deposition in tissues
a. Dietary trans fatty acids are incorporated
into most tissues and fluids of experimental animals and humans.
The level of deposition depends upon the tissue, dietary concen-
tration, and duration of exposure.
b. Trans fatty acids are not incorporated into all
tissue lipids or lipid of rats at the same level. Brain
tissue appears resistant to incorporation of isomeric fatty
in both experimental animals and humans. Among lipid classes,
only triacylglycerols in different tissues have similar
tions of trans isomers. The distribution of trans isomers in this
class is similar to that in the diet.
c. The octadecenoates
appear to accumulate in the phosphatidylcholine and phosphatidyl-
ethanolamine in most tissues of rats fed partially hydrogenated
fats. The and octadecenoates also appear to
accumulate in these lipid classes, as does the isomer in
phosphatidylcholines.
d. Most rat tissues discriminate against the
incorporation of and octadecenoates; heart and
liver tissues discriminate especially strongly against incorpora-
tion of these isomers in their phospholipids. The octa-
decenoate is strongly discriminated against by all tissues.
e. Levels of trans fatty acids in the lipids of
adipose tissue of U.S. sUbjects ranged from 2.0-5.8% as measured
by GLC and reflect the long-term level in the diet.
f. The distribution of trans positional isomers
in the triacylglycerols of heart, liver, and adipose tissue of
humans taken at autopsy is similar but not identical to the esti-
mated average distribution in commercial dietary hydrogenated fats.
g. Trans isomer levels in the phospholipids of
human tissues are no greater, and are generally less, than in
adipose tissue indicating that incorporation of trans isomers
into phospholipids is generally lower than levels in the diet.
h. Liver and heart tissues of humans discriminate
against incorporation of 10t-18:1 in phosphatidylcholine. The 14t-
and 15t-18:1 isomers are present in greater proportions in whole
tissue lipids than in the average of commercial dietary hydrogenated
oils, but levels are relatively low and indicate no major accumula-
tion. Levels of 10c-18:l are two to three times lower in tissue
lipids than in the average of dietary hydrogenated fats.
i. Although data indicate some similarities in
the selectivity of rat and human tissues for incorporation and
exclusion of positional isomers of 18:1 fatty acids, present
79
evidence indicates less accumulation of to and trans
octadecenoates in human heart and liver tissues, and-relatively
weaker discrimination against incorporation of cis- and trans-
18:1 isomers. The data also indicate a much stronger preferential
incorporation of 9t-18:1 in plasma phospholipids in rats than in
" humans.
j. The position and geometrical configuration of
monoene double bonds affect the metabolic fate of the isomer and
it is incorrect to assume that other cis and trans 18:1 positional
isomers are deposited in tissues like oleate and elaidate, respec-
tively.
k. The low levels of trans isomers found in tissue
lipids of premature infants (0.1-0.9%) indicate that the human
placenta is a barrier to the transport of these isomers.
3.' Atherosclerosis
a. Studies with humans fed diets of similar fatty
acid composition except for the replacement of part of 9c-18:1 in
one diet with the 18:1 isomers in partially hydrogenated vegetable
oils in the other (each group serving as its own control) indicate
that the 18:1 isomers in partially hydrogenated vegetable oil are
no more cholesterolemic than 9c-18:1 in subjects with cholesterol
levels about 200 mg/dl. Similar studies in which elaidic and oleic
acid were fed to different experimental groups are not definitive;
differences in elevation of serum cholesterol level were not sig-
nificant at p (0.05.
b. Short-term studies with rabbits, swine, and
monkeys show that dietary elaidic acid or partially hydrogenated
vegetable oils are cholesterolemic, but not atherogenic, as com-
pared with oleic acid or the unhydrogenated oils, respectively.
4. Interference of 18:1 and 18:2 isomers with linoleate
. metabolism
a. Both dietary trans-18:1 and trans-9,12-18:2
isomers perturb the metabolism of linoleic acids in rats, the
magnitude of the perturbation decreasing as the ratio of trans
isomer to linoleic acid in the diet decreases. At the relative
per capita levels of trans isomers and linoleate available in
the U.S. diet, the perturbation of linoleate metabolism including
formation of serum eicosanoids, appears to be without adverse
effect.
b. In vitro studies with rat liver microsomes from
EFA-deficient rats show that several of the to trans- and
cis-18:1 positional isomers inhibit the desaturation of linoleic
80
to Y-linolenic acid and eicosa-8,11.14-trienoic to
arachidonic acid Liver microsomes also have been
shown to desaturate positional isomers of trans-18:1 fatty acids
to cis. trans- and cis,cis-18:2 isomers of unusual structures.
These isomers could-COmpete with the normal metabolism of linoleic
acid and potentially be precursors of prostaglandins, thromboxane,
and leukotrienes of unusual structure. However, neither octadeca-
dienoic acids nor eicosanoids with unusual structures have been
reported in tissues of animals fed hydrogenated vegetable oils
and no evidence has been found that effects of the trans-18:1
positional isomers on fatty acid metabolism are of physiologic
consequence.
5. Immune function
In vitro trans fatty acids incorporated into lympho-
cytes and macrophages influence selected biophysical characteristics,
generally paralleling effects observed with saturated fatty acids.
Limited in vitro studies preclude any definite conclusions with
respect to changes in immune function. It may be hypothesized
that trans fatty acids act similarly to saturated fatty acids in
activities such as antigen binding, receptor modulation, signal
transduction. and antibody secretion. However, because appro-
priately controlled studies have not been performed in vivo,
no conclusion can be made with respect to the effects of dietary
trans fatty acids on immune response.
D. ANALYTICAL METHODS
1. Two methods. infrared (IR) spectrometry and gas-
liquid chromatography (GLC) are best suited for routine deter-
mination of total trans fatty acid content of large numbers of
samples of fats, oils, or lipid extracts. The IR method has the
advantage that it can be applied directly to lipids without con-
version to methyl esters. It is also the official method of the
American Oil Chemists' Society and the Association of Official
Analytical Chemists. It has the disadvantage that results with
triacylglycerols tend to be higher, and with methyl esters lower.
than true values for trans isomer contents less than 15%. A
modification of the IR method proposed by Madison et al. (1982)
appears to allow more accurate determinations at low trans fatty
acid concentrations.
GLC using 6.1 m packed columns coated with OV-275, or
15 m glass capillary columns coated with SP-2340, have greater
accuracy than the IR method. and provide data on the content of
both trans monoenes and dienes. Disadvantages are that lipids
must be converted to the methyl esters, and there is incomplete
baseline separation of the 9,12-18:2 isomers and unidentified
compounds which elute at similar times. This may lead to over-
estimation of the content of trans-9.12-18:2 isomers.
81
The 100 m glass capillary column requires a somewhat
greater time to chromatograph a sample (ca 50 minutes) than the
15 m column (but not than the 6.1 m packed column) but has the
advantages that the isomers can be quantitated
free from interference of unidentified substances, and data are
provided for a number of specific cis and trans positional isomers
in addition to the content of totar-trans-18:1 isomers.
2. Ultraviolet spectrometry is a sensitive method for
determining the conjugated diene content in edible fats and oils.
The method does not distinguish positional or geometric isomers.
3. Argentation thin layer chromatography separates
cis from trans'monoenes and from dienes as lipid classes. This
preseparation is a valuable adjunct to GLC allowing better resolu-
tion of cis or trans positional isomers than can be obtained by
GLC of the mixture.
4. High performance liquid chromatography (HPLC) is
complementary to GLC in the resolution of positional isomers;
to isomers are resolved in reversed phase HPLC,
a region of poor resolution in GLC. Resolution of the
isomers has been reported using argentation HPLC.
5. Mass spectrometry is best suited to the characteri-
zation of single fatty acids. The technique provides data on the
molecular weight, number, and position of the double bond. However,
in order to determine the position of unsaturation, a derivative
must be formed at the double bond either prior to analysis or by
interaction with a reaction gas in the ionization chamber of the
spectrometer. Geometrical isomerism can be determined but the
result is less definitve than position determination.
6. Proton nuclear magnetic resonance (lH NMR) is limited
to the determination of geometrical isomers and certain positional
isomers in single fatty acids. Hence, this method, like MS, must
be used in conjunction with other separation techniques in the
analysis of fatty acid mixtures.
The natural abundance 13C NMR technique has been applied
to the analysis of the trans content of fatty acid mixtures with
results that agree well with the GLC method. Percentages of
monoene, diene, and triene unsaturation also can be determined.
An advantage of the method is that it can be applied to intact
acylglycerols. A disadvantage is cost relative to chromatographic
methods. .
82
7. Ozonolysis is the most generally applicable method for
determining the positions of double bonds in partially hydrogenated
fats. The method requires conversion to the methyl esters and
fractionation according to the extent of unsaturation. Monoenes
require further separation into cis and trans isomers if geometrical
positional isomers are to be determined. Analysis of mixtures of
positional diene isomers require special techniques.
8. The lipoxidase enzymatic method is specific for the
determination of cis methylene-interrupted pentadiene structures.
It provides a sensitive method for measuring the maximum content
of linoleic acid in partially hydrogenated fat (the measurement
includes any residual linolenic acid or isomers formed containing
cis methylene interrupted double bonds).
E. GENERAL
Although trans octadecenoate isomers are similar to
saturated fatty acids in some respects, e.g., in their prefer-
ential deposition in the 1 and 3 positions of triacylglycerols
and the 1 position of phospholipids, they differ in other respects,
e.g., there is discrimination against incorporation of some trans
positional isomers in animal tissues and accumulation of others.
Also, the trans fatty acids as consumed in hY9rogenated vegetable
oil appear to be the equivalent of oleic acid in their choles-
terolemic properties in humans. In this respect they are similar
to stearic acid, but dissimilar to palmitic, myristic, and lauric
acids. Hence, classification of trans fatty acids as saturated
acids should be limited to specific properties and specific trans ~
fatty acids.
83
VI. SUGGESTIONS FOR FUTURE CONSIDERATION
While the ad hoc Panel has concluded that the available
scientific information suggests little reason for concern with
the safety of dietary trans fatty acids both at their present and
expected levels of consumption and at the present and expected
levels of consumption of dietary-[inoleic acid, certain additional
studies would clarify several unanswered questions. The following
investigations would better define the physiological properties of
partially hydrogenated vegetable oils and their component isomeric
fatty acids.
Additional studies should be conducted on the isomeric
fatty acid composition of the adipose tissues of humans.
Studies should be designed in which adipose tissue com-
position can be correlated with dietary fat composition
as determined by dietary record and/or analysis. Results
of these studies would provide an improved database for
relating adipose tissue composition to dietary intake
of fatty acids and their isomers. .
Performance of further studies on the geometrical and
positional fatty acid isomers in various lipid classes of
human tissues would confirm and extend/present information
on the accumulation and discrimination against specific
isomers.
Study of the cholesterolemic properties of the major posi-
tional and geometrical isomers of 18:1 fatty acids present
in partially hydrogenated vegetable oils relative to oleic
acid-should be made employing diets and protocols similar
to those used by Mattson et ale (1975) in comparing the
cholesterolemic properties of partially hydrogenated vege-
table oil and oleic acid in humans. In these studies, the
relative concentrations of high-density and low-density
lipoproteins as well as total cholesterol should be
determined.
Although studies with dietary trans fatty acids have
not shown statistically significant (p <0.05) promotional
effects on the rate of primary tumor growth of trans-
planted tumor cells, or on tumors initiated by chemical
carcinogens, limitations in experimental design in some
of these studies, suggest that additional studies with
spontaneous, chemically induced, and transplanted tumors
should be conducted. The effects of trans fatty acids
with those of saturated and polyunsaturated fatty acids
should be compared.
Preceding Page Blank i
- - - - - - - - - - ~
85
Investigation of the effects of the major dietary geo-
metrical and positional isomers of fatty acids present
in hydrogenated vegetable oils should include studies on
the modulation of immune function in response to defined
antigens. These studies should investigate changes in
membrane fluidity, receptor availability and mobility,
enzyme activity, and levels of prostaglandins and
leukotrienes .
86
VII. LITERATURE CITED
Aaes-Jrgensen, E. 1958. Essential fatty acid deficiency. III.
Effects of conjugated isomers of dienoic and trienoic fatty acids
in rats. J. Nutr. 66:465-483.
Ace-Federal Reporters, Inc. 1985. Life Sciences Research Office
ad hoc review panel of trans fatty acids. Transcript of open
meeting proceedings, February 21, Bethesda, MD. 95p. Available
from: Ace-Federal Reporters, Inc., Washington, DC.
Adlof, R.O.; Emken, E.A. 1985. Fatty acid isomer levels in human
tissue lipids. Unpublished.
Aitchison, J.M.; Dunkley, W.L.; Canolty, N.L.; Smith, L.M.
Influence of diet on trans fatty acids in human milk. Am.
Clin. Nutr. 30:2006-2015.
1977.
J.
Alam, S.Q.; Alam, B.S.; Banerji, A. 1985. Incorporation of trans
fatty acids into submandibular salivary gland lipids. Lipids
20:16-23.
Alfin-Slater, R.B; Aftergood, L.; Bingemann, L.;
Deuel, H.J., Jr. 1957a. Effect of hydrogenated
utilization of essential fatty acids in the rat.
Exp. BioI. Med. 95:521-523.
Kryder, G.D.;
triolein on
Proc. Soc.
Alfin-Slater, R.B.; Wells, A.F.; Aftergood, L.; Deuel, H.J., Jr.
1957b. Nutritive value and safety of hydrogenated vegetable fats
as evaluated by long-term feeding experiments with rats. J. Nutr.
63:241-261.
Alfin-Slater, R.B.; Aftergood, L.; Whitten, T. 1976. Nutritional
evaluation of trans fatty acids. J. Am. Oil Chern. Soc. 53:468A
(Abstract).
Alfin-Slater, R.B.; Wells, P.; Aftergood, L. 1973. Dietary fat
composition and tocopherol requirement: IV. Safety of polyunsatu-
rated fats. J. Arn. Oil Chern. Soc. 50:479-484.
Alfin-Slater, R.B.; Wells, P.; Aftergood, L.A. 1970. Longevity
and multigeneration studies in rats fed hydrogenated fats of
increasing polyunsaturated fat content. J. Am. Oil Chern. Soc.
47:274 (Abstract).
Allen, R.R. 1969. A rapid method for the determination of trans
unsaturation in fats and derivatives. J.Arn. Oil Chern. Soc.
46:552-553.
American Oil Chemists' Society. 1980. Official methods and
recommended practices of the American Oil Chemists' Society.
Champaign, IL: American Oil Chemists' Society.
87
Anderson, B.A.; Heimermann, W.H.; Holman, R.T.1974. Comparison
of pyrrolidides with other amides for mass spectral determination
of structure of unsaturated fatty acids. Lipids 9:443-449.
Anderson, J.T.; Grande, F.; Keys, A. 1961.
the diet and lipids in the serum of man. J.
Hydrogenated fats in
Nutr. 75:388-394.
Anderson, R.L. 1967. Oxidation of the geometric isomers of
octadecenoic acid by rat-liver mitochondria. Biochim. Biophys.
Acta 144:18-24.
Anderson, R.L.; Coots, R.H. 1967. The catabolism of the geo-
metric isomers of 14C-labeled acid and
uniformly 14C-labeled -octadecadienoic acid by the fasting
rat. Biochim. Biophys. Acta 144:525-531.
Anderson, R.L.; Fullmer, C.S., Jr.; Hollenbach, E.J. 1975.
Effects of trans isomers of linoleic acid on the metabolism
of linoleic acid in rats. J. Nutr.
Aplin, R.T.; Coles, L. 1967. A simple procedure for localisation
of ethylenic bonds by mass spectrometry. J. Chern. Soc. Chern.
Commun. p.858-859.
Applewhite, T.H.
fats to cancer:
38:2435.
1979. Statistical "correlations" relating trans-
a commentary. Fed. Proc. Fed. Am. Soc. Exp. BioI.
Applewhite, T.H. 1981. Nutritional effects of hydrogenated soya
oil. J. Am. Oil Chern. Soc. 58:260-269.
Applewhite, T.H. 1985. Kraft, Inc., Glenview, IL. Personal
communication with F.R. Senti, January 25, Life Sciences Research
Office, Federation of American Societies for Experimental Biology,
Bethesda, MD.
Association of Official Analytical Chemists. 1984. Isolated
trans isomers in margarines and shortenings: infrared spectro-
scopic method. Final action. American Oil Chemists' Society
Method. In: Official methods of analysis of the Association
of Official Analytical Chemists. 14th ed. Arlington, VA:
Association of Official Analytical Chemists, Inc. p.518-519.
Awad, A.B. 1981. Trans fatty acids in tumor development and
host survival. J. Natl. Cancer Inst. 67:189-192.
Bailar, II 1.
further critique.
2436.
1979. Dietary fat and cancer trends: a
Fed. Proc. Fed. Am. Soc. Exp. BioI. 38:2435-
Baldassare, J.J.; Brenckle, G.M.; Hoffman, M.; Silbert, D.F. 1977.
Modification of membrane lipid. J. BioI. Chern. 252:8797-8803.
88
Baldassare, J.J.; Rhinehart, K.B.; Silbert, D ~ F . 1976. Modifica-
tion of membrane lipid: physical properties in relation to fatty
acid structure. Biochemistry 15:2986-2994.
Barton, F.E., II; Himmelsbach, D.S.; Burdick, D. 1975. Deter-
mination of the cis-trans composition of methyl oleate and methyl
elaidate by 13C NMR. J. Mag. Resonance 18:167-171.
Battaglia, R.; Frohlich, D. 1980.
trans monounsaturated fatty acids.
HPLC-separation of cis and
Chromatographia 13:428-431.
Beare, J.L.; Heroux, C.; Murray, T.K. 1967. Variation in poly-
unsaturates of special margarines. Can. Med. Assoc. J. 96:1575-
1576.
Beare-Rogers, J.L.
fatty acids. In:
research. Vol. 5.
p.171-200.
1983. Trans- and positional isomers of common
Draper, H.H., ed. Advances in nutritional
New York: Plenum Publishing Corporation.
Beare-Rogers, J.L.; Gray, L.M.; Hollywood, R. 1979. The linoleic
acid and trans fatty acids of margarines. Am. J. Clin. Nutr.
32:1805-1809.
Benga, G.; Travis, B.D.; Pop, V.I.; Popescu, 0.; Toader, S.;
Holmes, R.P. 1984. The effect of the saturation and isomerization
of dietary fatty acids on the osmotic fragility and water diffu-
sional permeability of rat erythrocytes. Biochim. Biophys. Acta
775:255-259.
Beroza, M.; BierI, B.A. 1967. Rapid determination of olefin
position in organic compounds in microgram range by ozonolysis
and gas chromatography. Anal. Chem. 39:1131-1135.
Beveridge, J.M.R.; Connell, W.F. 1962.
margarines on plasma cholesterol levels
Nutr. 10:391-397.
The effect of commercial
in man. Am. J. Clin.
Beynen, A.C.; Hermus, R.J.J.; Hautvast, J.G.A.J. 1980. A
mathematical relationship between the fatty acid composition
of the diet and that of the adipose tissue in man. Am. J. Clin.
Nutr. 33:81-85.
Beynon, J.H.; Saunders, R.A.; Williams, A.E. 1968. The mass
spectra of organic molecules. New York: Elsevier/North-Holland,
Inc. 519p.
Bialick, R.; Gill, R.; Berke, G.; Clark, W.R. 1984. Modulation
of cell-mediated cytotoxicity function after alteration of fatty
acid composition in vitro. J. Immunol. 132:81-87.
89
Bitner, E.D.; Davison, V.L.; Dutton, H.J. 1969. Integrated
system for microreactor gas chromatography. J. Am. Oil Chern.
Soc. 46:113-117.
Black, H.S.; Gerguis, J. 1980. Use of the Ames test in assessing
the relation of dietary lipid and antioxidants to N-2-fluorenyl-
acetamide activation. J. Environ. Path. Toxicol. 4:131-138.
Blank, M.L.; Privett, O.S. 1963. Studies on the metabolism of
cis,trans isomers of methyl linoleate and linolenate. J. Lipid
Res. 4:470-476.
Blomstrand, R.; Svensson, L. 1983. The effects of partially
hydrogenated marine oils on the mitochondrial function and mem-
brane phospholipid fatty acids in rat heart. Lipids 18:151-170.
Bonaga, G.; Trizzino, M.G.; Pasquariello, M.A.; Biagi, P.L. 1980.
Nutritional aspects of trans fatty acids. Note I. Their accumula-
tion in tissue lipids of rats fed with normolipidic diets containing
margarine. Biochem. Exp. BioI. 16:51-54.
Breckenridge, W.C. - 1978. Fatty acids and glycerides.
Kuskis, A., ed. Handbook of lipid research. Vol. 1.
Plenum Press. p.197.
In:
New York:
Brekke, O.L. 1980. Specifications for soybeen oil. In:
Erickson, D.R.; Pryde, E.H.; Brekke, O.L.; Mounts, T.L.; Falb,
R.L., eds. Handbook of soy oil processing and utilization.
Champaign, IL: American Oil Chemists' Society. p.377-437.
Brown, R.R. 1981. Effects of dietary fat on incidence of
spontaneous and induced cancer in mice. Cancer Res. 41:3741-
3742.
Bruckner, G.; Trimbo, S.; Goswami, S.; Kinsella, J.E. 1983.
Dietary tri1inoelaidate: effects on hematological parameters,
serum eicosanoids and tissue fatty acid composition in rats.
J. Nutr. 113:704-713.
D.L.; Lehmann, J.; Mason, B.S.; Slover, H.T. 1976.
Lipid composition of selected vegetable oils. J. Am. Oil Chern.
Soc. 53:713-718.
Carpenter, D.L.; Slover, H.T. 1973. Lipid composition of
selected margarines. J. Am. Oil Chern. Soc. 50:372-376.
Carroll, K.K. 1975. Experimental evidence of dietary factors
and hormone-dependent cancers. Cancer Res. 35:3374-3383.
Castro, C.E.; Felkner, I.C.; Yang, S.P. 1978. Dietary lipid-
dependent activation of the carcinogen N-2-fluorenylacetamide
in rats as monitored by Salmonella typhlmurium. Cancer Res.
38:2836-2841.
90
Cervilla, M.; Puzo, G. 1983. Determination of the double bond
position in monounsaturated fatty acids by mass analyzed ion
kinetic energy spectrometry/collision induced dissociation after
chemical ionization of their amino alcohol derivatives. Anal.
Chern. 55:2100-2103.
Chai, R.; Harrison, A.G. 1981. Location of double bo.nds by
chemical ionization mass spectrometry. Anal. Chern. 53:34-37.
Chern, J.C.; Bruckner, G.; Kinsella, J.E. 1984. Effects of
dietary trilinoelaidate on liver mitochondrial phospholipids and
endogenous phospholipase activity. Nutr. Res. 4:79-92.
Chern, J.C.; Kinsella, J.E. 1983. The effects of unsaturated
fatty acids on the synthesis of arachidonic acid in rat kidney
cells. Biochim. Biophys. Acta 750:465-471.
Christie, W.W.
Pergamon Press.
1982. Lipid analysis.
207p.
2nd ed. New York:
Clark, R.M.; Ferris, A.M.; Fey, N.; Hundrieser, .K.E.; Jensen, R.G.
1980. The identity of cholesteryl esters in human milk. Lipids
15:972-974.
Clifford, C.K.; Smith, L.M.; Erickson, K.L.; Hamblin, C.L.;
Creveling, R. K.;, Clifford A. J 1983. Effect of dietary tri-
glycerides on lymphocyte transformation in rats. J. Nutr.
113:669-679.
Combe, N.; Wolff, R.; Rietsch, J.; Entressangles, B. 1980.
Implications de l'incorporation d'isomeres trans d'acides gras
insatures au niveau des membranes cellulaires. Ann. Nutr. Alim.
34:305-316.
Conacher, H.B.S.; Iyengar, J.R. 1978. Gas-liquid chromatographic
determination of trans-unsaturation in fats and oils on packed
columns: effect of positional isomers. J. Assoc. Off. Anal.
Chern. 61:702-708.
Coots, R.H. 1964a. A comparison of the metabolism of cis,cis-
linoleic, trans ,trans-linoleic, and a mixture of cis,trans- and
trans,cis-linoleic acids in the rat. J. ~ i p i d ReS:- 5:473-476.
Coots, R.H. 1964b. A comparison of the metabolism of elaidic,
oleic, palmitic, and stearic acids in the rat. J. Lipid Res.
5:468-472.
Corps, A.N.; Pozzan, T.; Hesketh, T.R.; Metcalfe, J.C. 1980.
cis-Unsaturated fatty ,acids inhibit cap formation on lymphocytes
by depleting cellular ATP. J. BioI. Chern. 255:10566-10568.
91
}.
Craig-Schmidt, M.C.; Weete, J.D.; Faircloth, S.A.; Wickwire, M.A.;
Livant, E.J. 1984. The effect of hydrogenated fat in the diet of
nursing mothers on lipid composition and prostaglandin content of
human milk. Am. J. Clin. Nutr. 39:778-786.
Dayton, S.; Hashimoto, S.; Dixon, W.; Pearce, M.L. 1966. Composi-
tion of lipids in human serum and adipose tissue during prolonged
feeding of a diet high in unsaturated fat. J. Lipid Res. 7:103-
111.
Decker, W.J.; Mertz, W. 1967. Effects of dietary elaidic acid on
membrane function in rat mitochondria and erythrocytes. J. Nutr.
91:324-330.
de Iongh, H.; Beerthuis, R.K.; den Hartog, C.; Dalderup, L.M.;
van der Spek, P.A.F. 1965. The influence of some dietary fats
on serum lipids in man. Bibl. Nutr. Dieta 7:137-152.
Deuel, H.J.; Hallman, L.F.; Movitt, E. 1945. Studies on the
comparative nutritive value of fats. VI. Growth and reproduction
over 10 generations on Sherman diet B where butterfat was replaced
by a margarine fat. J. Nutr. 29:309-320.
Deuel, H.J., Jr.; Greenberg, S.M.; Savage, E.E.; Bavetta, L.A.
1950. Studies on the comparative nutritive value of fats.
J. Nutr. 42:239-255.
Dewar, J.S. 1951.
Soc. Chim. France
A review of the TT -complex theory.
18: C71-C79.
Bull.
Doll, R.; Peto, R. 1981. The causes of cancer: quantitative
estimates of avoidable risks of cancer in the United States today.
J. Natl. Cancer Inst. 66:1191-1308.
Dommes, V.; Wirtz-Peitz, F.; Kunau, W.-H. 1976. Structure deter-
mination of polyunsaturated fatty acids by gas chromatography-mass
. spectrometry: a comparison of fragmentation patterns of various
derivatives. J. Chromatogr. Sci. 14:360-366.
Dutton, H.J. -1979. Hydrogenation of fats and its significance.
,In: Emken, E.A.; Dutton, H.J., eds. Geometrical and positional
fatty acid isomers. Champaign, IL: American Oil Chemists'
Society. p .1-16. ~
Egwim, P.O.; Kummerow, F.A. 1972. Incorporation and distribution
of dietary elaidate in the major lipid classes of rat heart and
plasma lipoproteins. J. Nutr. 102:783-792.
Egwim, P.O.; Sgoutas, D.S. 1971a. Occurrence of eicosadienoic
acids in liver lipids of rats fed p a r t i a l ~ y hydrogenated soybean
fat. J. Nutr. 101:307-314.
92
Egwim, P.O.; Sgoutas, D.S. 1971b. The fatty acids of adrenal
lipids from rats fed partially hydrogenated soybean fat. J. Nutr.
101:315-322.
Elson, C.E.; Benevenga, N.J.; Canty, D.J.; Grummer, R.H.; Lalich,
J.J.; Porter, J.W.; Johnston, A.E. 1981. The influence of dietary
unsaturated cis and trans and saturated fatty acids on tissue
lipids of swine. Atherosclerosis 40: 115-137'.
Emken, E.A. 1981. Influence of trans-9-, trans-12-, and cis-12-
octadecenoic acid isomers on fatty acid composition of human-
plasma lipids. Prog. Lipid Res. 20:135-141.
Emken, E.A. 1983. Human studies with deuterium labeled octa-
decenoic acid isomers. In: Perkins, E.G.; Visek, W.J., eds.
Dietary fats and health. Champaign,' IL: American Oil Chemists'
Society. p.302-319.
Emken, E.A. 1984a. In vivo distribution of trans- and cis-10-
octadecenoic acid isomers in human plasma lipids. J.
Chern. Soc. 61:678.
Emken, E.A. 1984b. Nutrition and biochemistry of trans and
positional fatty acid isomers in hYdrogenated oils. Annu. Rev.
Nutr. 4:339-376.
Emken, E.A. 1985. Use of deuterium labelled fats to follow
incorporation and turnover of trans- and cis-11-octadecenoic acid
isomers in human plasma lipids. J. Am. Oil Chern. Soc. 62:647.
Emken, E.A.; Adlof, R.O.; Rohwedder, W.K.; GulleY,R.M. 1983.
.'1\,
Incorporation of deuterium-labeled trans- and cis-13-octadecenoic
acids in human plasma lipids. J. Lipid Res. 24:34-46.
Emken, E.A.; Dutton, H.J.; Rohwedder, W.K.; Rakoff, H.; Adlof,
R.O.; Gulley, R.M.; Canary, J.J. 1980. of deuterium-
labeled cis- and trans-12-octadecenoic acids in human plasma and
lipoprotein lipids. Lipids 15:864-871.
Emken,' E.A.; Rohwedder, W.K.; Dutton, H.J.; DeJarlais, W.J.;
Adlof, R.O.; Mackin, J.F.; Dougherty, R.M.; Iacono, J.M. 1979a.
Incorporation of deuterium-labeled cis- and trans-9-octadecenoic
acids in humans: plasma, erythrocyte, and platelet phospholipids.
Lipids 14:547-554.
Emken, E.A.; Rohwedder, W.K.; Dutton, H.J.; DeJarlais, W.J.;
Adlof, R.O.; Mackin, J.; Dougherty, R.; Iacono, J.M. 1979b.
Incorporation of deuterium labeled cis- and trans-9-octadecenoic
acid in humans: plasma, platelet neutral lipids.
Metabolism 28:575-583.
93
Enig, .M.G.; BUdowski, P.; Blondheim, S.H. 1984. trans-Unsaturated
fatty acids in margarines and human sUbcutaneous fat in Israel.
Hum. Nutr. Clin. Nutr. 38C:223-230.
Enig, M.G.; Munn, R.J.; Keeney, M. 1978. Dietary fat and cancer
trends: a critique. Fed. Proc. Fed. Am. Soc. Exp. BioI. 37:2215-
2220.
Enig, M.G.; Munn, R.J.; Keeney, M. 1979. Response. Fed. Proc.
Fed. Am. Soc. Exp. BioI. 38:2437-2439.
Enig, M.G.; Pallansch, L.A.; Sampugna, J.; Keeney, M. 1983.
Fatty acid composition of the fat in selected food items with
emphasis on trans components. J. Am. Oil Chern. Soc. 60:1788-1795.
Erickson, B.A.;. Coots, R.H.; Mattson, F.H.; Kligman, A.M. 1964.
The effect of partial hydrogenation of dietary fats, of the ratio
of polyunsaturated to saturated fatty acids, and of dietary choles-
terol upon plasma lipids in man. J. Clin. Invest. 43:2017-2025.
Erickson, K.L.; Adams, D.A.; McNeill, C.J. 1983. Dietary lipid
modulation of immune responsiveness. Lipids 18:468-474.
Erickson, K.L.; Schlanger, D.S.; Adams, D.A.; Fregeau, D.R.;
Stern, J.S. 1984. Influence of dietary fatty acid concentration
and geometric configuration on murine mammary tumorigenesis and
experimental metastasis. J. Nutr. 114:1834-1842.
Esfahani, M.; Limbrick, A.R.; Knutton, S.; Oka, T.; Wakil, S.J.
1971. The molecular organization of lipids in the membrane of
Escherichia coli: phase transitions. Proc. Natl. Acad. Sci. USA
68:3180-3184-.---
Ferber, E.; De Pasquale, G.G.; Resch, K. 1975. Phospholipid
metabolism of stimulated lymphocytes composition of phospholipid
fatty acids. Biochim. Biophys. Acta 398:364-376.
Frost, D.J.; Barzilay, J. 1971a. Proton magnetic resonance
identification of nonconjugated cis-unsaturated fatty acids and
esters. Anal. Chern. 43:1316-1318.
Frost, D.J.; Barzilay, J. 1971b. PMR cis/trans analysis of
double bonds using alkyl substituent effects. Reel. Trav. Chim.
Pays Bays 90:705-712.
Frost, D.J.; Gunstone, F.D. 1975. The PMR analysis of non-
conjugated alkenoic and alkynoic acids and esters. Chern. Phys.
Lipids 15:53-85.
GaIly, H.U.; Pluschke, G.; Overath, P.; Seelig, J. 1979. Struc-
ture of Escherichia coli membranes: phospholipid conformation
in model membranes and cells as studied by deuterium magnetic
resonance. Biochemistry 18:5605-5610.
94
Gildenberg, L.; Firestone, D. 1985. Gas chromatographic deter-
mination of trans unsaturation in margarine: collaborative study.
J. Assoc. Off. Anal. Chern. 68:46-51.
Gill, R.; Clark, W. 1980. Membrane structure-function relation-
ships in cell-mediated cytolysis. I. Effect of exogenously
incorporated fatty acids on effector cell function in cell-
mediated cytolysis. J. Immunol. 125:689-695.
Goor, R.; Hosking, J.D.; Dennis, B.H.; Graves, K.L.; Waldman,
G.T.; Haynes, S.G. 1985. Nutrient intakes among selected North
American populations in the Lipid Research Clinics Prevalence
Study: composition of fat intake. Am. J. Clin. Nutr. 41:299-311.
Gottenbos, J.J. 1983. Biological effects of trans fatty acids.
In: Perkins, E.G.; Visek, W.J., eds. Dietary fats and health.
Champaign, IL: American Oil Chemists' Society. p.375-390.
Graff, G.; Lands, W.E.M. 1976. A shift from phospholipid to tri-
glyceride synthesis when cell division is inhibited by trans-fatty
acids. Chern. Phys. Lipids 17:301-314.
Gunstone, F.D.; Ismail, I.A., Lie Ken Jie, M. 1967. Fatty acids,
Part 16. Thin layer and gas-liquid chromatographic properties of
the cis and trans methyl octadecenoates and of some acetylenic
esters7 Chern. Phys. Lipids 1:376-385.
Gunstone, F.D.; Norris, F.A.
biochemistry and technology.
1983. Lipids in foods: chemistry,
New York: Pergamon Press. 170p.
Guo, L.S.S.; Alexander, J.C. 1978. Incorporation of [10_
14
C]
oleic acid or [10_
14
C] elaidic acid into lipids of liver, adrenal
and plasma lipoproteins of normal and essential fatty acid-
deficient rats. Can. Inst. Food Sci. Technol. J. 11:169-172.
Guyer, W.; Bloch, K. 1983. Phosphatidylcholine and cholesterol
interactions in model membranes. Chern. Phys. Lipids 33:313-322.
Hamilton, R.J.; Sewell, P.A.
ance liquid chromatography.
. 248p.
1982. Introduction to high perform-
2nd ed." New York: Chapman and Hall
Hankin, J.H.; Rawlings, V. 1978. Diet and breast cancer:
a review. Am. J. Clin. Nutr. 31:2005-2016.
Hay, J.D.; Morrison, W.R. 1970. Isomeric monoenoic fatty acids
in bovine milk fat. Biochim. Biophys. Acta 202:237-243.
Heckers, H.; Dittmar, K.; Melcher, F.W.; Kalinowski, H.O. 1977a.
Silar10 C, Silar 9 CP, SP 2340 and OV-275 in the gas-liquid
chromatography of fatty acid methyl esters on packed columns.
J. Chromatogr. 135:93-107.
95
Heckers, H.; Melcher, F.W.; Schloeder, U. 1977b. SP 2340 in
the glass capillary chromatography of fatty acid methyl esters.
J. Chromatogr. 136:311-317.
Reekers, H.; Korner, M.; Tuschen, T.W.L.; Melcher, F.W. 1977c.
Occurrence of individual trans-isomeric fatty acids in human
myocardium, jejunum and aorta in relation to different degrees
of atherosclerosis. Atherosclerosis 28:389-398.
Heckers, H.; Melcher, F.W.; Dittmar, K. 1979. Daily consumption
of trans-isomeric fatty acids: a calculation based on composition
of commercial fats and of various human depot fats. Fette Seifen
Anstrichm. 81:217-226.
Hill, D.J.; Dawidowicz, E.A.; Andrews, M.L.; Karnovsky, M.J.
1983. Modulation of microsomal glucose-6-phosphate translocase
activity by free fatty acids: implications for lipid domain struc-
ture in microsomal membranes. J. Cell Physiol. 115:1-8.
Hill,
R.T.
acids
Nat!.
E.G.; Johnson,' S.B.; Lawson, L.D.; Mahfouz, M.M.; Holman,
1982. Perturbation of the metabolism of essential fatty
by dietary partially hydrogenated vegetable oil. Proc.
Acad. Sci. USA 79:953-957.
Hirsch, J.; Farquhar, J.W.; Ahrens, E.H., Jr.; Peterson, M.L.;
Stoffel, W. 1960. Studies of adipose tissue in man. Am. J.
Clin. Nutr. 8:499-511.
Hogan, M.L.; Shamsuddin, A.M. 1984. Large intestinal carcino-
genesis. I. Promotional effect of dietary fatty acid isomers
in the rat model. J. Natl. Cancer Inst. 73:1293-1296.
Holman, R.T. 1951. Metabolism of isomers of linoleic and
linolenic acids. Proc. Soc. Exp. BioI. Med. 76:100-102.
Holman, R.T.; E. 1956. Effects of trans fatty
acid isomers upon essential fatty acid deficiency in rats. Proc.
Soc. Exp. BioI. Med. 93:175-179.
Holman, R.T.; Mahfouz, M.M.; Lawson, L.D.; Hill, E.G. 1983.
etabolic effects of isomeric octadecenoic acids. In: Perkins,
E.G.; Visek, W.J., eds. Dietary fats and health. Champaign, IL:
American Oil Chemists' Society. p.320-340.
Hoover, R.L.; Klausner, R.; Karnovsky, M.J. 1982. Inhibition of
cap formation on lymphocytes by free fatty acids is not mediated
by a depletion of ATP. J. BioI. Chern. 257:2151-2154.
Horvath, C.; Melander, W.R. 1983. Theory of chromatography. In:
Heftmann, E., ed. Chromatography. Fundamentals and applications
of chromatographic and electrophoretic methods. Part A. Funda-
mentals and techniques. New York: Elsevier Scientific Publishing
Company. p.A28-A130.
96
H ~ y , C.-E.; H ~ l m e r , G. 1979. Incorporation of cis- and trans-
octadecenoic acids into the membranes of rat liver-mitochondria.
Lipids 4:727-733.
H ~ y , C.-E.; H ~ l m e r , G. 1981. Incorporation of cis-octadecenoic
acids into the rat liver mitochondrial membrane phospholipids and
adipose tissue triglycerides. Lipids 16:102-108.
Hsu, C.M.L.; Kummerow, F.A. 1977. Influence of elaidate and
erucate on heart mitochondria. Lipids 12:486-494.
Huang, A.; Firestone, D. 1971. Comparison of two infrared methods
for the determination of isolated trans unsaturation in fats, oils,
and methyl ester derivatives. J. Assoc. Off. Anal. Chern. 54:1288-
1292.
Hundrieser, K.E.; Clark, R.M.; Brown, P.B. 1983. Distribution of
trans-octadecenoic acid in the major glycerolipids of human milk.
J. Pediatr. Gastroenterol. Nutr. 2:635-639.
Hunter, J.E. 1982. Trans fatty acids in tumor development
. J. Natl. Cancer Inst. 69:319-320.
Hunter, J.E. 1985. Current, previous, and predicted levels
of trans fatty acids in the U.S. diet. Statement presented on
behalf of the Institute of Shortening and Edible Oils, Washington,
DC at the open meeting of the ad hoc Review Panel on Trans Fatty
Acids, February 21, Federation of American Societies for Experi-
mental Biology, Bethesda, MD. Appendix A, this report.
Jackman, L.M. 1959. Applications of nuclear magnetic resonance
spectroscopy in organic chemistry. New York: Pergamon Press.
134p. .
Jackson, R.L.; Morrisett, J.D.; Pownall, H.J.; Gotto, A.M., Jr.;
Kamio, A.; Imai, H.; Tracy, R.; Kummerow, F.A. 1977. Influence
of dietary trans-fatty acids on swine lipoprotein composition and
structure. J. Lipid Res. 18:182-190.
Jacob, J.; Grimmer, G. 1967. Occurrence of positional isomers of
octadecenoic and hexadecenoic acids in human depot fat. J. Lipid
Res. 8: 308-311.
Johnson, B.M.; Taylor, J.W. 1972. Hexafluoroacetone ketals as
derivatives for positional and geometrical characterization of
double bonds. Anal. Chern. 44:1438-1444.
Johnston, A.E.; Dutton, H.J.; Scholfield, C.R.; Butterfield, R.O.
1978. Double bond analysis of dienoic fatty acids in mixtures.
J. Am. Oil Chern. Soc. 55:486-490.
Johnston, P.V.; Johnson, O.C.; Kummerow, F.A.
of trans fatty acids in human tissue. Science
97
1957. Occurrence
126:698-699.
Johnston, P.V.; Johnson, O.C.; Kummerow, F.A. 1958a. Deposition
in tissues and fecal excretion of trans fatty acids in the rat.
J. Nutr. 65:13-23.
Johnston, P.V.; Kummerow, F.A.; Walton, C.H. 1958b. Origin of
the trans fatty acids in human tissue. Proc. Soc. Exp. BioI. Med.
99:735-736.
""
Kaufmann, H.P.; Mankel, G. 1964. Uber das Vorkommen von trans-
Fettsiuren. Fette Seifen Anstrichm. 66:6-13.
Kaunitz, H. 1976.
Z. Ernahrungswiss.
Biological effects of trans fatty acids.
15:26-33.
Keys, A.; Anderson, J.T.; Grande, F. 1965. Serum cholesterol
response to changes in the diet. IV. Particular saturated fatty
acids in the diet. Metabolism 14:776-787.
Kim, W.W.; Mertz, W.; JUdd, J.T.; Marshall; M.W.; Kelsay, J.L.;
Prather, E.S. 1984. Effect of making duplicate food collectIons
on nutrient intakes calculated from diet records. Am. J. Clin.
Nutr. 40:1333-1337.
Kinsella, J.E.; Bruckner, G.; Mai, J.; Shimp, J. 1981. Metabolism
of trans fatty acids with emphasis on the effects of trans,trans-
octadecadienoate on composition, essential fatty acid, and
prostaglandins: an overview. Am. J. Clin. Nutr. 34:2307-2318.
Klausner, R.D.; Kleinfeld, A.M.; Hoover, R.L.; Karnovsky, M.J.
1980a. Lipid domains in membranes. J. BioI. Chern. 255:1286-1295.
Klausner, R.D.; Bhalla, D.K.; Dragsten, P.; Hoover, R.L.;
Karnovsky, M.J. 1980b. Model for capping derived from inhibition
of surface receptor capping by free fatty acids. Proc. Natl.
Acad. Sci. USA 77:437-441.
Kritchevsky, D.; Davidson, L.M.; Kim, H.K.; Krendel, D.A.;
Malhotra, S.; Vander Watt, J.J.; duPlessis, J.P.; Winter, P.A.D.;
Ipp, T.; Mendelsohn, D.; Bersohn, I. 1977. Influence of semi-
purified atherosclerosis in African Green monkeys. Exp.
Mol. Pathol. 26:28-51.
Kritchevsky, D.; Moyer, A.W.; Tesar, W.C.; Logan, J.B.; Brown,
R.A.; Davies, M.C.; Cox, H.R. 1954. Effect ot cholesterol
vehicle in experimental atherosclerosis. Am. J. Physiol.
178:30-32.
Kummerow, F.A.; Mizuguchi, T.; Arima, T.; Cho, B.H.S.; Huang, W.Y.
1978. The influence of three sources of dietary fats and choles-
terol on lipid composition of swine serum lipids and aorta tissue.
Artery 4:360-384.
98
Kumrnerow, F.A.; Mizuguchi, T.; Arima, T.; Yeh, s.c.; Cho, B.
1974. Swine as an animal model in studies on atherosclerosis.
Fed. Proc. Fed. Am. Soc. Exp. BioI. 33:235 (Abstract).
Lanser, A.C.; Emken, E.A.; Ohlrogge, J ~ B . [1985]. Oxidation
of oleic and elaidic acids by rat and human heart homogenates.
Biochim. Biophys. Acta In press.
Lanza, E.; Slover, H.T. 1981. The use of SP2340 glass capillary
columns for the estimation of the trans fatty acid content of
foods. Lipids 16:260-267.
Latondress, E.G. 1981. Formulation of products from soybean oil.
J. Am. Oil Chern. Soc. 58: 185-187."
Lawson, L.D.; Holman, R.T. 1981. S-Oxidation of the geometric
and positional isomers of octadecenoic acid by rat heart and liver
mitochondria. Biochim. Biophys. Acta 665:60-65.
Loomis, R.J.; Marshall, L.A., "Johnston, P.V. 1983.
acid effects on cultured rat spenocytes. J. Nutr.
Sera fatty
113:1292-1298.
MacBeath, L.S.; Cook, H.W.
in human brain. Clin. Res.
1977. Trans-octadecenoic fatty acids
25:709A (Abstract).
MacGee, J. 1959. Enzymatic determination of polyunsaturated
fatty acids. Anal. Chern. 31:298-302.
Madison, B.L.; Depalma, R.A.; D'Alonzo, R.P. 1982. Accurate
determination of trans isomers in shortenings and edible oils by ~
infrared spectrophotometry. J. Am. Oil Chern. Soc. 59:178-181.
Mahfouz, M.; Johnson, S.; Holman, R.T. 1981. Inhibition of
desaturation of palmitic, linoleic and eicosa-8,11,14-trienoic
acids in vitro by isomeric cis-octadecenoic acids. Biochim.
Biophys. Acta. 663:58-68.
Mahfouz, M.M.; Johnson, S.; Holman, R.T. 1980a. The effect
of isomeric trans-18:1 acids on the desaturation of palmitic,
linoleic and eicosa-8,ll,14-trienoic acids by rat liver micro-
somes. Lipids 15:100-107.
Mahfouz, M.M.; Valicenti, A.M.; Holman, R.T. 1980b. Desaturation
of isomeric trans-octadecenoic acids by rat liver microsomes.
Biochim. Biophys. Acta 618:1-12.
Mahoney,E.M.; Hamill, A.L.; Scott, W.A.; Cohn, Z.A. 1977.
Response of endocytosis to altered fatty acyl composition of macro-
phage phospholipids. Proc. Natl. Acad. Sci. USA 74:4895-4899.
Mahoney, E.M.; Scott, W.A.; Landsberger, F.R.; Hamill, A.L.; Cohn,
Z.A. 1980. Influence of fatty acyl substitution on the composition
and function of macrophage membranes. J. BioI. Chern. 255:4910-4917.
99
Mandel, G.; Clark, W. 1978. Functional properties of EL-4 tumor
cells with lipid-altered membranes. J. Immunol. 120:1637-1643.
Mandel, G.; Shimizu, S.; Gill, R.; Clark, W. 1978. Alteration
of the fatty acid composition of membrane phospholipids in mouse
lymphoid cells. J. Immunol. 120:1631-1636.
Marston, R.M.; Welsh, S.O. 1984. Nutrient content of the U.S.
food supply, 1982. Natl. Food Rev. 25:7-13.
Mattson, F.H. 1960. An investigation of the essential fatty acid
activity of some of the geometrical isomers of unsaturated fatty
acids. J. Nutr. 71:366-370.
Mattson, F.H.; Hollenbach, E.J.; Klingman, A.M. 1975. Effect of
hydrogenated fat on the plasma cholesterol and triglyceride levels
in man. Am. J. Clin. Nutr. 28:726-731.
McCloskey, J.A.; McClelland, M.J. 1965. Mass spectra of O-iso-
propylidene derivatives of unsaturated fatty esters. J. Am. Chern.
Soc. 87:5090-5093.
McElhaney, R.N.; De Gier, J.; Van Deenen, L.L.M. 1970. The
effect of alterations in fatty acid composition and cholesterol
content on the permeability of Mycoplasma laidlawii B cells and
derived liposomes. Biochim. Biophys. Acta 219:245-247.
McMillan, G.C.; Silver, M.D.; Weigensberg, B.I. 1963. Elaidinized
olive oil and cholesterol atherosclerosis. Arch. Pathol. 76:118-
124.
McOsker, D.E.; Mattson, F.H.; Sweringen, H.B.; Kligman, A.M
. 1962. The influence of partially hydrogenated dietary fats on
serum cholesterol levels. J. Am. Med. Assoc. 180:380-385.
McVey, E.; Yguerabide, J.; Hanson, D.C.; Clark, W.R. 1981.
The relationship between plasma membrane lipid composition and
physical-chemical properties. I. Fluorescence polarization
studies of fatty acid-altered EL4 tumor cell membranes. Biochim.
Biophys. Acta 642:106-118.
Meade, C.J.; Mertin, J . 1978. Fatty acids and immunity. Adv.
Lipid Res. 16:127-165.
Meade, C.J.; Mertin, J.; Sheena, H.; Hunt, R. 1978. Reduction by
linoleic acid of the severity of experimental allergic encephalo-
myelitis in the guinea pig. J. Neurol. Sci. 35:291-308.
Menon, N.K.; Dhopeshwarkar, G.A. 1983. Differences in the fatty
acid profile and S-oxidation by heart homogenates of rats fed cis
and trans octadecenoic acids. Biochim. Biophys. Acta 751:14-20.
100
Meyer, W.H.
Exp. BioI.
1979. Further comments.
38:2436-2437.
Fed. Proc. Fed. Am. Soc.
Miller, A.B.; Kelly, A.; Choi, N.W.; Matthews, V.; Morgan, R.W.;
Munan, L.; Burch, J.D.; Feather, J.; Howe, G.R., Jain, M. 1978.
A study of diet and breast cancer. Am. J. Epidemiol. 107:499-509.
Mishkel, M.A.; Spritz, N. 1969. The effects of trans isomerized
trilinolein on plasma lipids of man. In: Holmes, W.L.; Carlson,
L.A.; Paoletti, R., eds. Drugs affecting lipid metabolism. New
York: Plenum Press. p.355-364.
Moore, C.E.; Alfin-Slater, R.B.; Aftergood, L. 1980. Incorpora-
tion and disappearance of trans fatty acids in rat tissues. Am.
J. Clin. Nutr. 33:2318-2323.
Morris, L.J. 1966. Separations of lipids by silver ion chroma-
tography. J. Lipid Res. 7:717-732.
Morris, L.J.; Wharry, D.M.; Hammond, E.W. 1 9 6 7 ~ Chromatographic
behaviour of isomeric long-chain aliphatic compounds. II.
Argentation thin-layer chromatography of isomeric octadecenoates.
J. Chromatogr. 31:69-76.
Nazir, D.J.; Moorecroft, B.J.; Mishkel, M.A. 1976. Fatty acid
composition of margarines. Am. J. Clin. Nutr. 29:331-339.
Nichols, P.L., Jr.; Herb, S.F.; Riemenschneider, R.W. 1951.
Isomers. of conjugated fatty acids. I. Alkali-isomerized linoleic
acid. J. Am. Chern. Soc. 73:247-252.
Niehaus, W.G., Jr.; Ryhage, R. 1965. Determination of double ~
bond positions in polyunsaturated fatty acids by combination gas
chromatography-mass spectrometry. Anal. Chern. 40:1840-1847.
Nolen, G.A. 1972. Effects of fresh and used hydrogenated soybean
oil on reproduction and teratology in rats. J. Am. Oil Chern. Soc.
49:688-693.
Nolen, G.A.; Alexander, J.C.; Artman, N.R. 1967. Long-term rat
feeding study with used frying fats. J. Nutr. 93:337-348.
O'Connor, R.T.; Formo, M.W.; Herb, S.; Goldwasser, S.; McLaughlin,
J., Jr.; Rockwood, B.N.; Stillman, R.C.; Wolff, H. 1955. Report
of the spectroscopy committee, 1954-1955. J. Am. Oil Chern. Soc.
32:452-454.
Office of the Federal Register, General Services Administration.
1984. Code of Federal Regulations. Title 21: food and drugs,
parts 100-169 rev. Washington, DC: U.S. Government Printing
Office.
101
Ohlrogge, J.B. 1983. Distribution in human tissues of fatty acid
isomers from hydrogenated oils. In: Perkins, E.G.; Visek, W.J.,
eds. Dietary fats and health. Champaign, IL: American Oil
Chemists' Society. p.359-374.
Ohlrogge, J.B.; Barber, E.D.; Lands, W.E.M.; Gunstone, F.D.;
Ismail, I.A. 1976. Quantitative effects of unsaturated fatty
acids in microbial mutants. VI. Selective growth responses of
yeast and bacteria to cis-octadecenoate isomers. Can. J. Biochem.
54:736-745.
Ohlrogge, J.B.; Emken, E.A.; Gulley, R.M. 1981. Human tissue
lipids: occurrence of fatty acid isomers from dietary hydrogenated
oils. J. Lipid Res. 22:955-960.
Ohlrogge, J.B.'; Gulley, R.M., Emken, E.A. 1982. Occurrence of
octadecenoic fatty acid isomers from hydrogenated fats in human
tissue lipid classes. Lipids 17:551-557.
Ono, K.; Fredrickson, D.S. 1964. The metabolism' of 14C-labeled
cis and trans isomers of octadecenoic and octadecadienoic acids.
~ B i o l . Chern. 239:2482-2488.
Ottenstein, D.M.; Wittings, L.A.; Walker, G.; Mahadevan, V.; Pelick,
N. 1977. trans Fatty acid content of commercial margarine samples
determined by gas liquid chromatography on OV-275. J. Am. Oil Chern.
Soc. 54:207-209. .
Parodi, P.W. 1976.
acids in milk fat.
Distribution of isomeric octadecenoic fatty
J. Dairy Sci. 59:1870-1873.
Perkins, E.G.; McCarthy, T.P.; O'Brien, M.A.; Kummerow, F.A. 1977.
The application of packed column gas chromatographic analysis to
the determination of trans unsaturation. J. Am. Oil Chern. Soc.
54: 279-281.
Pfeffer, P.E.; Luddy, F.E.; Unruh, J.; Shoolery, J.N. 1977.
Analytical 13C NMR: a rapid, nondestructive method for deter-
mining the cis,trans composition of catalytically treated
unsaturated lipid mixtures. J. Am. Oil Chern. Soc. 54:380-386.
"
Picciano, M.F.; Perkins, E.G. 1977. Identification of the trans
isomers of octadecenoic acid in human milk. Lipids -12:407-408.
Ponder, D.L.; Green, N.R. 1985. Effects of dietary fats and
butylated hydroxytoluene on mutagen activation in rats. Cancer
Res. 45:558-560.
Poon, R.; Clark, W.R. 1982. The relationship between plasma mem-
brane lipid composition and physical-chemical properties. III.
Detailed physical and biochemical analysis of fatty acid-substi-
tuted EL4 plasma membranes. Biochim. Biophys. Acta 689:230-240.
102
Privett, O.S.; Blank, M.L. 1964. Studies on the metabolism
of linoelaidic acid in the essential fatty acid-deficient rat.
J. Am. Oil Chern. Soc. 41:292-297.
Privett, O.S.; Nickell, E.C. 1963. Stereoisomer formation on the
ozonization of esters of monounsaturated fatty acids. J. Lipid
Res. 4:208-211.
Privett, O.S.; Nickell, E.C. 1966. Determination of the specific
positions of cis and trans double bonds in polyenes. Lipids
1:98-103. ---
Privett, O.S.; Nutter, L.J.; Lightly, F.S. 1966. Metabolism
of trans acids in the rat: influence of the geometric isomers
of linoleic acid on the structure of liver triglycerides and
lecithins. J. Nutr. 89:257-264.
Privett, O.S.; Phillips, F.; Shimasaki, H.; Nozawa, T.; Nickell,
E.C. 1977. Studies of effects of trans fatty acids in the diet
on lipid metabolism in essential fatty acid deficient rats. Am.
J. Clin. Nutr. 30:1009-1017.
Privett, O.S.; Pusch, F.J.; Holman, R.T.
fatty acid metabolism. VIII. Nonpotency
linoleate as essential fatty acid. Arch.
57:156-162.
1955. Polyethenoid
of cis-9,trans-12-
Biochem. Biophys.
Rapacz, J.; Elson, .C.E.; Lalich, J.J. 1977. Correlation of an
immunogenetically defined lipoprotein type with aortic intimal
lipidosis in swine. Exp. Mol. Pathol. 27:249-261.
Raper, N. 1985. Human Nutrition Information Service, U.S.
Department of Agriculture. Personal communication with F.R.
Senti, Life Sciences Research Office, Federation of American
Societies for Experimental Biology, Bethesda, MD.
Reichwald-Hacker, I.; Ilsemann, K.; Mukherjee, K.D. 1979a.
Tissue-specific incorporation of positional isomers of dietary
cis- and acids in the rat. J. Nutr. 109:
1051-1056.
Reichwald-Hacker, I.; Grosse-Oetringhaus, S.; Kiewitt, I.;
Mukherjee, K.D. 1979b. Incorporation of positional isomers of
cis- and trans-octadecenoic acids into acyl moieties of rat tissue
lipids. Biochim. Biophys. Acta 575:327-334.
Rice, E.E.; Weiss, T.J.; Mattil, K.F. 1962. Composition of modern
margarines. J. Am. Diet. Assoc. 41:319-322.
Rizek, R.L.; Friend, B.; Page, L.
supply: level of use and sources.
244-250.
103
1974. Fat in today's food
J. Am. Oil Chern. Soc. 51:
Rizek, R.L.; Welsh, S.O.; Marston, R.M.; Jackson, E.M. 1983.
Levels and sources of fat in the U.S. food supply and in diets
of individuals. In: Perkins, E.G.; Visek, W.J., eds. Dietary
fats and health. Champaign, IL: American Oil Chemists' Society.
p.13-43.
JRocquelin, G.L.; Justrabo, E.; Grynberg, A.; David, M. 1984.
Fatty acid composition of human heart phospholipids: data from
53 biopsy specimens. Unpublished.
Royce, S.M.; Holmes, R.P.; Takagi, T.; Kummerow, F.A. 1984. The
influence of dietary isomeric and saturated fatty acids on athero-
sclerosis and eicosanoid synthesis in swine. Am. J. Clin. Nutr.
39:215-222.
Ruttenberg, H.; Davidson, L.M.; Little, N.A.; Klurfeld, D.M.;
Kritchevsky, D. 1983. Influence of trans unsaturated fats on
experimental atherosclerosis in rabbits. J. Nutr. 113:835-844.
Ruttenberg, H.; Little, N.A.; DaVidson, L.M.; Kritchevsky, D.
1980. Influence of dietary trans fatty acid on atherosclerosis in
rabbits. Fed. Proc. Fed. Am. Soc. Exp. BioI. 38:1039 (Abstract).
Ryhage, R.; Stenhagen, E. 1960. Mass spectrometric studies. VI.
Methyl esters of normal chain oxo-, hydroxy-, methoxy- and epoxy-
acids. Arkiv Kemi 15:545-566.
Saito, Y.; McElhaney, R.N. 1978. The positional distribution
of a series of positional isomers of cis-octadecenoic acid in
phosphatidylglYcerol from Acholeplasma laidlawii B. Biochim.
Biophys. Acta 529:224-229.
Sampugna, J.; Pallansch, L.A.; Enig, M.G.; Keeney, M. 1982. Rapid
analysis of trans fatty acids on SP-2340 glass capillary columns.
J. Chromatogr. 249:245-255.
Schlager, S.I. 1979. Specific 125I-iodination of cell surface
lipids: plasma membrane alterations induced during humoral immune
attack. J. Immunol. 123:2108-2113.
Schlager, S.I.; Ohanian, S.H. 1979. A role for fatty acid com--
position of complex cellular lipids in the susceptibility of tumor
cells to humoral immune killing. J. Immunol. 123:146-152.
Scholfield, C.R. 1979. Analysis and physical properties of iso-
meric fatty acids. In: Emken, E.A.; Dutton, H.J., eds. Geometri-
cal and positional fatty acid isomers. Champaign, IL: American
Oil Chemists' Society. p.17-52.
Scholfield, C.R.; Davison, V.L.; Dutton, H.J. 1967. Analysis
for geometrical and positional isomers of fatty acids in partially
hydrogenated fats. J. Am. Oil Chem. Soc. 44:648-651.
104
Schroit, A.J.; Gallily, R. 1979. Macrophage fatty acid com-
position and phagocytosis: effect of unsaturation on cellular
phagocytic activity. Immunology 36:199-205.
Schroit, A.J.; Rottem, S.; Gallily, R. 1976. Motion of spin-
labeled fatty acids in murine macrophages: relation to cellular
phagocytic activity. Biochim. Biophys. Acta 426:499-512.
Searcey, M.T.; Arata, D.A. 1972. A comparison of the metabolism
of geometrical isomers of acid in threonine-
imbalanced rats. J. Nutr. 102:1429-1435.
Seelig, J.; Waespe-Sarcevic, N. 1978. Molecular order in
cis and trans unsaturated phospholipid bilayers. Biochemistry
17:3310-3315.
Selenskas, S.L.; Ip, M.M.; Ip, C. 1984. Similarity between trans
fat and saturated fat in the modification of rat mammary carcino-
genesis. Cancer Res. 44:1321-1326.
Selinger, Z.; Holman, R.T. 1965. The effects of trans, trans-
linoleate upon the metabolism of linoleate and linolenate and the
positional distribution of linoleate isomers in liver lecithin.
Biochim. Biophys. Acta 106:56-62.
Shimp, J.L.; Bruckner, G.; Kinsella, J.E. 1982. The effects of
dietary trilinoelaidin on" fatty acid and acyl desaturases in rat
liver. J. Nutr. 112:722-735.
Slover, H.T. 1983. Gas chromatography: packed and capillary.
In: Perkins, E.G.; Visek, W.J., eds. Dietary fats and health.
Champaign, IL: American Oil Chemists' Society. p.90-109.
Slover, H.T. 1985. Beltsville Human Nutrition Research Center,
Beltsville, MD. Personal communication with F.R. Senti, Life
Sciences Research Office, Federation of American Societies for
Experimental Biology, Bethesda, MD.
Slover, H.T.; Lanza, E. 1979. Quantitative analysis of food
fatty acids by capillary gas chromatography. J. Am. Oil Chern.
Soc. 56:933-943.
Slover, H.T.; Thompson, R.H., Jr.; Davis, C.S.; Merola, G.V.
1985. Lipids in margarines and margarine-like foods. J. Am.
Oil Chern. Soc. 62:775-786.
Smith, L.M.; Dunkley, W.L.; Franke, A.; Dairiki, T. 1978.
Measurement of trans and other isomeric unsaturated fatty acids
in butter and margarine. J. Am. Oil Chern. Soc. 55:257-261.
Spence, M.W. 1970. A simple method for determining the double-
bond position in monoenoic fatty acids. Biochim. Biophys. Acta
218:357-359.
105
Steele, W.; Jenkin, H.M. 1972. The effect of two isomeric octa-
decenoic acids on alkyl diacyl glycerides and neutral glycosphingo-
lipids of Novikoff hepatoma cells. Lipids 7:556-559.
Stein, R.A.; Slawson, V. 1967. Column chromatography of lipids.
In: Holman, R.T., ed. Progress in the chemistry of fats and other
lipids. New York: Pergamon Press. p.375-420.
Sugano, M.; Ide, T.; Kohno, M.; Watanabe, M.; Cho, Y.-J.; Nagata,
Y. 1983. Biliary and fecal steroid excretion in rats fed partially
hydrogenated soybean oil. Lipids 18:186-192.
Suzuki, M.; Ariga, T.; Sekine, M.; Araki, E.; Miyatake, T. 1981.
Identification of double bond positions in polyunsaturated fatty
acids by chemical ionization mass spectrometry. Anal. Chern.
53:985-988.
Thomas, L.H.; Jones, P.R.; Winter, J.A.; Smith, H. 1981. Hydrog-
enated oils and fats: the presence of chemically-modified fatty
acids in human adipose tissue. Am. J. Clin. Nutr. 34:877-886.
Thomas, L.H.; Winter, J.A.; Scott, R.G. 1983a. Concentration
of transunsaturated fatty acids in the adipose body tissue of
decedents dying of ischaemic heart disease compared with controls.
J. Epidemiol. Commun. Health 37:22-24.
Thomas, L.H.; Winter, J.A.; Scott, R.G. 1983b. Concentration
of 18:1 and 16:1 transunsaturated fatty acids in the adipose body
tissue of decedents dying of ischaemic heart disease compared with
controls: analysis by gas liquid chromatography. J. Epidemiol.
Commun. Health 37:16-21.
Traill, K.N.; Wick, G. 1984. Lipids and lymphocyte function.
Immunol. Today 5:70-76.
Trotter, J.; Flesch, I.; Schmidt, B.; Ferber, E. 1982. Acyl-
transferase-catalyzed cleavage of arachidonic acid from phospho-
lipids and transfer to lysophosphatides in lymphocytes and macro-
phages. J. BioI. Chern. 257:1816-1823.
Tsang, W.M.;
Zilkha, K.J.
linoleate in
39:767-771.
Belin, J.; Monro, J.A.; Smith, A.D.; Thompson, R.H.S.;
1976. Relationship between plasma and lymphocyte
multiple sclerosis. J. Neurol. Neurosurg. Psychiatr.
Tsao, Y.-K.; Lands, W.E.M. 1980. Cell growth with trans fatty
acids is affected by adenosine 3',5'-monophosphate and membrane
fluidity. Science 207:777-779.
u.S. Department of Agriculture. 1975. Fats and oils situation.
Available from:' Economic Research Service, Washington, DC; FOS-
280, October.
106
u.s. Department of Agriculture. 1984. Oil crops: outlook and
situation report. Available from: Economic Research Service,
Washington, DC; OCS-6, August.
van den Reek, M.; Craig-Schmidt, M.C.; Weete, J.D.; Clark, A.J.
1985. Isomeric fat in the diets of adolescent girls. Fed. Proc.
Fed. Am. Soc. Exp. BioI. 44:1523 (Abstract).
Vandenhoff, G.; Gunstone, F.D.; Barve, J.; Lands, W.E.M. 1975.
Inhibition of growth of microbial mutants by trans-octadecenoates.
J. BioI. Chern. 250:8720-8727.
Vergroesen, A.J.; Gottenbos, J.J. 1975. The role of fats in
human nutrition: an introduction. In: Vergroesen, A.J., ed.
The role of fats in human nutrition. New York: Academic Press.
p.1-41.
VIes, R.O.; Gottenbos, J.J. 1972a. Long-term effects of feeding
butterfat, coconut oil and hydrogenated or non hydrogenated soya-
bean oils. I. Eighteen-month experiment in mice. Voeding 33:
428-433.
VIes, R.O.; Gottenbos, J.J. 1972b. Long-term effects of feeding
butterfat, coconut oil, and hydrogenated or non hydrogenated soya-
bean oils. II. Life-span experiment in rats . Voeding 33:455-465.
VIes, R.O., Gottenbos, J.J.; van Pijpen, P.L. 1977. Aspects
nutritionnels des huiles de soja hydrogenees et de leurs acides
gras insatures isomeriques. Bibl. Nutr. Dieta 25:186-196.
Walenga, R.W.; Lands, W.E.M. 1975. Effectiveness of various
unsaturated fatty acids in supporting growth and respiration
in Saccharomyces cerevisiae. J. BioI. Chern. 250:9121-9129.
Watt, B.K.; Merrill, A.L. 1975. Composition of foods -- raw,
processed, prepared. Agriculture Handbook No.8. Available from:
U.S. Government P r i n ~ i n g Office, Washington, DC. 189p.
Weigensberg, B.I.; McMillan, G.C. 1964. Serum and aortic lipids
in rabbits fed cholesterol and linoleic acid stereoisomers. J.
Nutr. 83:314-324.
Weigensberg, B.I.; McMillan, G.C.; Ritchie, A.C. 1961. Elaidic
acid: effect on experimental atherosclerosis. Arch. Pathol.
72:358-366.
Weiss, T.J. 1983. Food oils and their uses. 2nd ed. Westport,
CT: AVI Publishing Company, Inc. 320p.
Wennerstrom, D.E.; Jenkin, H.M. 1976. The effect of two isomeric
octadecenoic acids on the lipid metabolism and growth of Novikoff
hepatoma cells. Biochim. Biophys. Acta 431:469-480.
107
1966. The metabolism of
Biochim. Biophys. Acta
Wiedermann, L.H. 1978. Margarine and margarine oil, formulation
and control. J. Am. Oil Chern. Soc. 55:823-829.
Willebrands, A.F.; Van der Veen, K.J.
elaidic acid in the perfused rat heart.
116:583-585.
Wood, R. 1979a. Distribution of dietary geometrical and
positional isomers in brain, heart, kidney, liver, lung, muscle,
spleen, adipose, and hepatoma. In: Emken, E.A.; Dutton, H.J.,
eds. Geometrical and positional fatty acid isomers. Champaign,
IL: American Oil Chemists' Society. p.213-281.
Wood, R. 1979b. Incorporation of dietary cis and trans octa-
decenoate isomers in the lipid classes of various rat tissues.
Lipids 14: 975,-982.
Wood, R. 1983. Geometrical and positional monoene isomers in
beef and several processed meats. In: Perkins, E.G.; Visek, W.J.,
eds. Dietary fats and health. Champaign; IL: American Oil
Chemists' Society. p.341-358.
Wood, R. 1984. High-performance liqUid chromotography analysis
of isomeric monoenoic and acetylenic fatty acids. J. Chromatogr.
287:202-208.
Wood, R.; Chumbler, F. 1978. Distribution of dietary octa-
decenoate isomers at the 1- and 2-positions of hepatoma and
liver phospholipids. Lipids 13:75-84.
Wood, R.; Chumbler, F.; Matocha, M.; Zoeller, A. 1979.
Geometrical and positional isomer content of monounsaturated
fatty acids from various rat tissues. Lipids 14:789-794.
Wood, R.; Chumbler " F.; Wiegand, R.. 1977. Incorporation of
dietary cis and trans isomers of octadecenoate in lipid classes
of liver and hepatoma. J. BioI. Chern. 252:1965-1970.
Wood, R.; Lee, T. 1983. High-performance liquid chromatography
of fatty acids: quanti tati ve analysis of saturated, monoenoic"
polyenoic and geometrical isomers. J. Chromatogr. 254:237-246.
Yoo, T.-J.; Chiu, H.C.;
G.M.; Lee, N.F. 1980.
cultured hepatoma cells
cytolysis. Cancer Res.
Spector, A.A.; Whiteaker, R.S.; Denning,
Effect of fatty acid modifications of
on susceptibility to complement-mediated
40:1084-1090.
Yu, P.H.; Mai, J.; Kinsella, J.E. 1980. The effects of dietary
trans, trans methyl octadecadienoate acid on composition and fatty
acids of rat heart. Am. J. Clin. Nutr. 33:598-605.
108
VIII. STUDY PARTICIPANTS
A. AD HOC REVIEW PANEL ON TRANS FATTY ACIDS
CO-CHAIRMEN
Owen Fennema, Ph.D.
Professor
Department of Food Sciences
University of Wisconsin
1605 Linden Drive
Madison, Wisconsin 53706
Frederic R. Senti, Ph.D.
Senior Scientific Consultant
Life Sciences Research Office
Federation of American Societies
for Experimental Biology
9650 Rockville Pike
Bethesda,Maryland 20814
CONSULTANTS
E.A. Emken, Ph.D.
Research Leader
Biochemistry and Biophysical
Properties Research
Oilseeds Crops Laboratory
USDA Agriculture Research Service
1815 North University Street
Peoria, Illinois 61604
Kent L. Erickson, Ph.D.
Associate Professor
Department of Human Anatomy
University of California
School of Medicine
Davis, California 95616
Ralph T. Holman, Ph.D.
Professor of Biochemistry
Executive Director
Hormel Institute
University of Minnesota
Austin, Minnesota 55912
J.E. Kinsella, Ph.D.
Professor and Chairman
Department of Food Science
Cornell University
114 Stocking Hall
Ithaca, New York 14853
Albert I. Mendeloff, M.D.
Professor of Medicine
Johns Hopkins University
School of Medicine
2109 North Cliff Drive
Baltimore, Maryland 21209
Samuel B. Tove, Ph.D.
Professor and Head
Department of Biochemistry
North Carolina State University
Raleigh, North Carolina 27650
Randall D. Wood, Ph.D.
Professor
Department of Biochemistry and Biophysics
Texas A&M University
College Station, Texas 77843
109
B. REVIEWING CONSULTANTS
F.H. Mattson, Ph.D.
Adjunct Professor
Department of Medicine
School of Medicine
University of California,
San Diego
La Jolla, California 92093
Hal T. Slover, Ph.D.
Research Chemist
Nutrient Composition Laboratory
Beltsville Human Nutrition
Research Center
USDA-ARS
Beltsville, Maryland 20705
C. LIFE SCIENCES RESEARCH OFFICE
Sue Ann Anderson, Ph.D.
Senior Staff Scientist
Harolyn B. Cohen
Secretary
Kenneth D. Fisher, Ph.D.
Director
Beverly R. Lea
Literature Retrieval/
Technical Report Specialist
110
Judith F. Miller
Administrative Assistant
Susan M. Pilch, Ph.D.
Staff Scientist
Stephen Simpson
Literature Technical Assistant
John M. Talbot, M.D.
Senior Medical Consultant
IX. OPEN MEETING PARTICIPANTS
A. SCHEDULED
Thomas H. Applewhite
Institute of Shortening
and Edible Oils
Washington, D.C.
Mary G. Enig
University of Maryland
College Park, Maryland
B. UNSCHEDULED*
Joyce Beare-Rogers
Department of National
Health and Welfare
Ottawa, Canada
J. Edward Hunter
Institute of Shortening
and Edible Oils
Washington, D.C.
Ronald Simpson
National Association of
Margarine Manufacturers
Washington, D.C.
Walter Meyers
Institute of Shortening
and Edible Oils
Washington, D.C.
Lembitu Reio
National Food Administration
Uppsala, Sweden
* Offered comments in response to invitation of Co-chairmen
of ad hoc Review Panel
111
.'
I,
APPENDIX A
Current, Previous, and Predicted Levels
of Trans Fatty Acids in the U.S. Diet
Presented by J.E. Hunter on behalf of the
Institute of Shortening and Edible Oils,
Washington, D.C. at the Open Meeting of the
ad hoc Review Panel on Trans Fatty Acids
February 21, 1985
Federation of American Societies for Experimental Biology
Bethesda, Maryland 20814
113
CURRENT, PREVIOUS, AND PREDICTED LEVELS OF
TRANS FATTY ACIDS IN THE U.S. DIET
This paper addresses (1) levels of trans fatty acids in the current U.S. food
supply and during the past 10 to 20 years; (2) levels of essential fatty acids
in the diet, 1967-1982; (3) levels of trans acid consumption predicted over the
next 5 to 10 years; and (4) estimates of the current U.S. dietary intake of
trans fatty acids.
1. Levels of trans fatty acids in the current U.S. food 'supply and during the
past 10 to 20 years.
There have been two major trends in fat consumption in the U.S. since 1960
(1). First, total per capita consumption of dietary fat (inclUding food
fats and oils as well as fats in meat, dairy products, and other fat
containing foods) increased from almost 119 pounds in 1960 to about 130
pounds in 1982. Second, there has been a shift from animal to vegetable
sources of .fats and o l l s ~ reflecting consumers' efforts to reduce saturated
and increase polyunsaturated fatty acids in their diets and expansion of
U.s. oilseed production. In 1960, fats and oils from animals represented
70$ of the total. By 1982, the share had fallen to 5 7 ~ (1) .
The shift to vegetable fats atthe expense of animal fats is reflected in
Nielsen sales data which, as shown in Figure 1, indicate increasing sales
of margarine and of salad and cooking oils during the past 20 years. Salad
and cooking oils are made exclusively from vegetable oils, and margarines
are made pt"imarily from vegetable oils. In contrast to sales of margarine,
-2-
sales of butter have declined markedly. Sales of shortenings, which also
are made primarily from vegetable oils, have declined slightly in recent
years. The USDA's disappearance data show trends 1n usage of salad and
cooking oils, margarine, butter, and shortening that are consistent with
the Nielsen sales. figures (1). In totaling the sales of these products, it
1s apparent that total usage of these so-called "visible fats" has
increased slightly but steadily during the past 20 years.
Because salad and cooking oils, margarines, and shortenings frequently
contain partially hydrogenated fats and oils and therefore are the
principal sources of trans acids in the diet, these data would suggest that
there has been an increase in trans acid intake since around 1960.
However, in assessing changes in trans acid intake, one also must consider
that the compositicn of many such products as well as their market shares
have changed over time, in part for reasons of consumer preference and
economics . Accordingly, in an effort to assess as accurately as possible
the current and previous levels of trans acids. in the U.S. food supply, the
Institute of and Edible Oils eISEO) has undertaken a study of
changes in market size and market share of various fat and oil products as
well as changes in fatty acid We believe that estimates of
trans acid intake based on nutritional surveys are inaccurate not only
because of errors inherent in collecting such data, but also because such
surveys usually fail to consider changes in market share and in product
composition over time.
-3-
I have obtained data member companies of the ISEO on market size,
market share, and compositional changes of various products made fron
partially hydrogenated fats and oils. This paper will focus on changes
related to the following four major product categories: (1) salad and
cooking oils; (2) household shortenings; (3) margar-ines; and (4) food
service fats and oils. In addition, changes in trans acid levels
contributed by meat and dairy products, in which trans acids occur
naturally, also will be considered.
a. Salad and Cooking Oils. First I consider the category of
household salad and cooking oils. Prior to the early 19605, most
salad and cooking oils were made from unhydrogenated vegetable oils and
therefore contained insignificant levels of trans fatty acids. The
principal source oils used for such products at that time were
-.',.
cottonseed oil and corn oil. In the early 1960s, partially
hydrogenated. soybean oil products were introduced, in due to
increasing demand for economical liquid oils with relatively high
..
levels of polyunsaturated fatty acids. Major benefits to the consumer
of partially hydrogenated soybean oils compared to unhydrogenated
soybean oil included improved oxidative stability (and thus longer
shelf life) and reduced likelihood of producing undesirable flavors and
odors on heating.
11"
""t. - "
Table 1 shows how trans acid levels of salad and cooking oils have"
changed since the early 1960s. As indicated earlier. the market size
(or total sales) of salad and cooking oils has increased continuously
since the early 1960s. In 1963, the total level of trans acids in the
food supply from such products was low reflecting the introduction of
partially hydrogenated soybean oil products at that time. As the
market for salad and cooking oils increased, availability of trans
acids increased in a correspotiding fashion through the mid to late
19708. "In the late 1970s, their production of
products which involved less hydrogenation. ihus the trans levels of
typical products began declining as did the total trans level of .
food supply contributed by these products, although the market size
continued to increase. Expressed on a per capita basis, since the mid
to late 1970s trans acid availability from salad and cooking oils
declined from about 0.5 to 0.3 g/day.
Extending these calculations to mayonnaise, we have estimated that
curently mayonnaise would contribute less than 0.05 g of trans
acids/person/day. Other salad dressings would contribute even "lower
levels of trans acids. Thus, mayonnaise and salad dressings make
relatively insignificant contributions to total availability of trans
acids compared to salad and cooking oils.
It is also important to note that despite increased availability of
-5-
trans acids during the 1960s and early 19705 from salad and cooking
oils, there has been no apparent in the availability of
polyunsaturated fatty acids from these products (Table 1). In fact,
increasing sales of salad and cooking oils. have resulted in increased
availability of polyunsaturates. Since polyunsaturates largely
fatty these data suggest that the increasing
availability of trans fatty acids did not adversely affect the
availability of essential fatty acids from salad and cooking oils.
'b. Household Shortenings. Table 2 shows how trans acid levels
contr;buted by household shortenings have changed over the years.
Overall, the size of the shortening market has not changed much since
1960, although there has been a slight decline in sales since the early
to mid 1970s. Similarly, the overall composition of household
shortenings has not changed very much, however" in the 1960s there was
. a slight increase in usage of vegetable fats at the expense of animal
fats.' Sales of household shortenings containing animal fats increased"
.....
somewhat in the early 1980s largely for price reasons. Typical trans
acid levels of shortenings containing animal fats have remained fairly
constant since 1960, whereas, typical trans levels of shortenings made
exclusively from vegetable fats" have declined somewhat. Overall, the
per capita availabiiity of trans acids from household shortenings has
declined frOM about 1 g/day in 1960 to about 0.6 g/day currently. Also
as noted in Table ?, the contribution of dietary polyunsaturates by
_ ...... ..... ,-
-6-
household shortenings has remained relatively unchanged since 1960.
c. Margarines. Turning now to margarines, as noted earlier, sales of
margarine increased during the 1960s and 1970s as margarine continued
to butter as the. principal table spread in the U.S. In the
19605, margarines were largely the stick variety with total trans acid
levels ranging 'from about 25 to Tub margarines were introduced in
the late 1960s. These products contained unhydrogenated as well as
partially hydrogenated vegetable oils, and they had lower total trans
acid revels than stick margarines, typically ranging from,about 13 to
The popularity of tub margarines continued to increase during the
1970s and, to a limited extent, sales of tub margarines replaced sales
of stick margarines.
Changes in trans acid levels in the U.S. food supply from margarines
since 1970 are shown in Table 3. These data take into account the
relative contributions of both stick and tub margarines. Overall the.
data indicate a slight increase in the per capita availability of trans
acids from margarine since 1970. Polyunsaturate levels of margarines
have increased to an even larger extent since 1970, indicating that
over the past 15 years margarines have continued to be a useful dietary
source of essential fatty acids.
d.' Food service fats and oils. Food ,service fats and oils include
II '1
<#', ~
-7-
butter, margarine, deep frying fats, pan and grill fats, cake and
pastry shortening, and bottled salad and cooking oils used by
restaurants. We were able to obtain data on usage of food service fats
and oils only for 1969 and 1979. Our usage data represent amounts of
various fats and oils sold into the food service industry. One must
k ~ e p in mind that. in the food service industry, unlike in the home,
there is frequently considerable wastage of fats and oils. This is
particularly true in the case of fats used for deep frying, since large
amounts of such fats (we estimate as much as 5 0 ~ ) are discarded after
use.
As indicated in Table 11, the total market for food service fats and
oils increased between 1969 and 1979, reflecting largely continued
growth of the fast food industry. In conjunction with this increase in
market size, there was also an increase in total trans acid
availability due to increased usage of partially hydrogenated oils in
some food service fats and oils. Trans acid levels of individual food
service fats and oils, however, generally remained unchanged or
decreased slightly during this period. Overall, between 1969 and 1919,
estimated per capita availability of trans acids from food service fats
and oils increased from 1.35 to 1.54 g/day.
An additional food category that uses partially hydrogenated fats and
oils and therefore contributes some trans acids to the diet is the
I;).. ()
--,
-8-
category of industrial fats and oils. These fats and oils are used
largely by bakeries for the production of a wide variety of baked goods
and by manufacturers of snack foods such as potato chips. Member
companies of the ISEO had'no data available on the usage of industrial
fats and oils. However, using data derived from the 1977 Census of
Manufacturing on ways in ,which visible fats and oils are used (1), we
, have estimated that approximately of total visible fats and oils
available for consumption may be used by the sector.
Considering that much of this fat is used in deep frying and thus is
discarded ultimately, we believe that a reasonable estimate of per
capita availability of trans acids from industrial fats and oils is
about 1 g/day, a level similar to that deri ved for food service fats
and oils.
e. Meat and dairy products. Small amounts of trans acids occur
naturally 1n foods such as meat, milk, butter, and tallow as a result
of microbial biohydrogenationin ruminants (2,3). Thus contributions
to the diet of trans acids from these sources also must be taken into
account. Considering that the average total trans level of these fats
is about 3% the capita availability of trans acids from these
sources can be calculated from USDA disappearance data (1,5). Results
of such calculations shown in Table 5 indicate that the per capita
availability of trans acids from meat and dairy products has" remained
relatively constant since 1960 at about to 1.5 g/day.
/.;2./
.'
-9-
f. Total trans acid levels in the diet. A summary of changes in trans
acid levels of salad and cooking oils, household shortenings, .
margarines, food service and industrial fats and oils, and meat and
dairy products between 1963 and 1984 is presented in Table 6. From
these data, total per capita availability -of trans acids can be
calculated to be about 7.70 g/day in 1970 and about 7.55 g/day in 1980
(assuming the data for food service fats and oils in 1970 and 1960 are
similar to those we for 1969 and 1979. respectively).
Recognizing 'that there have not been large variations in the fats used
in the food service industry since 1979. we estimate that total per
capita availability of trans has remained relatively unchanged at
about 7.6 g/day since 1970.
2. Levels of essential fatty acids in the diet. 1967-1982.
Although partially hydrogenated fats and oils are the principal sources of
trans acids in the diet, it is important to point out that they are also
. important 'dietary sources of the essential fatty acid, linoleic acid.
According to USDA disappearance data. the total per capita availability of
linoleic acid has increased continuously since 1909 (Figure' 2) (6). The
.-
amount of linoleic acid contributed by fats and oils also has increased
steadily since 1909 and in particular increased between 1967-69 and
1982," from 13.0 to 18.6 g/day (Figure 2) (6). The amount of monoenoic
fatty acids contributed by fatsand increased from 26.6 to
j
/
/
I
/.
I
3.
-10-
29.1 g/day during this period (6) . And, we. have shown that there has not
been a rampant increase in the per capita availability of trans acids
during the past 15 to 20 years, but there has been a substantial increase
in the availability of linoleic acid. Thus the increased usage of
processed fats and oils has had a positive effect on the essential fatty
acid level of the diet.
Levels of trans acid consumption predicted over the next 5 to 10 years.
In view of recommendations by numerous health advisory organizations
(including the American Heart Association, the National Academy of
Sciences, the American Cancer Society, the National Cancer Institute, and
. most recently, an advisory panel convened by the National Institutes of
Health) that Americans decrease their total fat consumption for health
reasons, we would speculate that fat intake will not likely increase over
the next 5 to 10 years, and probably will decrease somewhat. This likely
decrease in fat intake probably will apply to all sources of dietary fat,
including sources of visible fat which contain trans acids. However, we
would point out that fats are responsible for many of the desirable
flavors, aromas, and textures found in foods. Because of this, large
decreases in total fat consumption as recommended by various health
advisory organizations would be difficult for many people to achieve and
maintain.
-11-
In summary, while we would expect some decrease in total fat intake,
including trans fatty acid intake, to occur over the next 5 to 10 years, we
believe that such decreases will be modest, perhaps on the order of several
percent, but nowhere near the 25$ decrease in total fat intake recommended
the American Heart Association and some other organizations.
Estimated current U.S. dietary intake of trans fatty acids.
There are currently no reliable data on the of trans fat in-the
United States, however, it is possible to make reasonable estimates. In
part 1 of our comments above, we indicated that using market share and
product composition data one might calculate a per capita availability of
about 7.6 g/day. Using USDA disappearance Emken reported a similar
estimate (7 g/day) of the daily consumption of trans 18:1 isomers, the most
abundant trans isomers in human diets (7). This is based on an estimated
daily per capita consumption of 3q g of partially hydrogenated soybean oil
having an average total trans value of 20$. From another .point of view,
considering that the composition of unsaturated fatty acids of
tissue reflects that of the diet and that a range of 2.0 to trans
fatty acids has been reported in human tissue (8), an adult male
consuming his RDA of 2700 calories/day (9) of a diet providing calories
as fat (10) would ingest around 2.3 to 6.6 g of trans fatty acids per day.
This range is consistent with our estimate of 7.6 g/day. On the other
hand, Enig et a1. (11) have estimated a total trans fatty acid intake of
-12-
slightly more than 12 g/person/day and have stated that individual diets
have been found to contain up to 20 g/day (12). These estimates are not
..
supported by reliable data and would appear to be excessive.
.'
Enig et al.'s estimates of total trans acid intake also would seem
unrealistically high when considered from another perspective. Among the'
vegetable"'"61.1s used in the U.S., only two, namely soybean oil and corn oil,
are partially hydrogenated any degree. USDA disappearance data indicate
that for the period October 1, 1982, to September 30, 1983, total soybean
oil usage, was 9.3 billion pounds (1). Table 7 shows estimates of trans
acid availability assuming that 40, 50, or 60% of total soybean oil were
partially hydrogenated to an average trans acid level 20, 25, or 30$. For
example, if we. assumed that of total soybean oil were partially
hydrogenated to an average trans level of the expected per capita
, '
tra-ns acid availability, would' be about 3.9 g/day. Partially hydrogenated
corn oil is used only in margarine, and the per capita trans acid
contri bution from this source would be only about 0.2 g/day. Thus even ...
under the unlikely conditions that 60$ of total,soybean oil were partially
hydrogenated to an average trans level of trans acid availability
would be around 9 g/day, which is still less than the intake estimated
by Enig et ale of 12 g/day. In order to achieve a per capita intake of 12
g/day, one would need to assume that over of total soybean oil would be
partially hydrogenated to an average trans level of In actual
practice, manufacturers of fats and oils simply do not hydrogenate
-13-
sufficient quantities of soybean oil to allow for an average per capita
intake of trans 'acids as high as 12 g/day.
In conclusion, our review of market size, market share, and product
composition data suggests that despite increased usage of partially
hydrogenated vegetable fats and oils during the past 15 to 20 years, trans
acid availability has changed little during this period and currently is
around 7.6 g/person/day. Realistic levels of intake based on the
composition of human adipose tissue may be somewhat lower, perhaps on the
order of 3 t07 g/person/day. Partially hydrogenated soybean oil products
not only provide significant consumer benefits such as increased oxidative
stability (compared to unhydrogenated soybean oil products) but also
continue to be important dietary sources of essential fatty acids.
...:.
IJU665
(see IJH666 for tables) .
References
1. Bunch, K. and Hazera, J. Fats and oils: consumers use more, but different
kinds. National Food Rev., 26:18-21, 198q.
2. Kaufmann, H. P. and Mankel, Go, The occurrence of trans-fatty acids.
Fette, Sei,fen, Anstrichm., 1964.
3. Parodi,P. W., Distribution of isomeric octadecenoic fatty acids in milk
fat. J. Dairy Sci., 59:1870-1873, 1976.
q. Enig, M. G., Pallansch, L. A., Sampugna, J., and Keeney, M. -Fatty acid
composition of the fat in selected food items with emphasis on trans
components. J. Am. Oil Chern. Soc., 60:1788-1795, 1983.
5. Stucker, T. A. and Parham, K. D. Beef, pork, and poultry: our changing
consumption habits. National Food Rev., 25:20-22, 1984.
6. Marston, M. and Welsh, S. 0., Nutrient content of the U.S. food
supply, 1982. National Food Rev., 25:7-13, 1984.
7. Emken, E. A., Hetabolic aspects of positi9nal monounsaturated fatty acid
isomers. J. Am. Oil Chern. Soc., 58:278-283, 1981.
8. Ohlrogge, J. B., Emken, E. A., and Gulley, R. M. Human tissue lipids:
occurrence of fatty acid isomers from dietary hydrogenated oils. J. Lipid
Res.,22:955-960, 1981.
9. Food and Nutrition Board, National Research Council, Recommended Dietary
Allowances, 9th ed., Washington, D.C., National Academy of Sciences,
1980.
10. Kim, W. W., Mertz, W., Judd, Jo T., Harshall, W., Kelsay, J. L., and
Prather, E. S. .Effect of making duplicate food collections one'nutrient
intakes calculated from diet records. Am. J. Clin. Nutr., 40:1333-
1337, 198q.
11. Enig, M. G., Munn, R. J., and Keeney, H., Dietary fat and cancer trends-
a critique. Fed. Proc., 37:2215-2220, 1978.
12. Enig, M. G., BUdowski, P., and Blondheim, S. H., Trans-unsaturated
fatty acids in margarines and human subcutaneous fat in Israel.
Nutr.: Clin. Nutr., 38C:223-230, 19Sq.
IJH666
1J..1
FIG. 1
GROCERY STORE SALES
(PER CAPITA) OF FATS AND OILS
1960-1982
Totar...
20 ....--.,
15
Sales
(Ibs! 10
personl
year}
5

$.3101 d .and
. Oils.
fln In9
-------..--__-..:6:..:..utter
o "'- --J
19'00
1geo
-..
.,'
,
Source: A. C. Nielsen Co.
IJH666
-- -
-
o
.'-'-
=
-+--
=-

-
-
-
:
-
= -..
-
=
.-
:-
. -
=:E
G


) L
:z:
) ..
..

."
..
,...-'.


,-'
..:..:...
:::J:
...;..;
. .
'-l-
E:
..,...
.-+0
--'
::;.
:=:===
:i:
-
.-'-!- .,.,...;i.
.""'"
-
..... .:.....,
. """'-'
., ;


o
....,.
-
.,-
==- -==
.=
-
-. - =-
-


= ..
-
JJ..'f

_c::::::=::: ......
.
-
:::::
:::
"'""'"-:-'
.



, :'
-:::j.. " ;..L;
-
.==
,.....:...
.'
TABLE 1
Trans Acid Levels of Salad and Cooking Oils,
1963 - 1984
Typical Total Trans Per Capita Availability
,J.
Year Market Size [Trans] Acids Trans Acids Polyunsaturates
(millions of ( ~ ) (millions of . (g/person/day)
pounds) pounds)
1963 518 15 9
0.06 1.86
1970 763 15 .63 0.38 2. 01
1915 856 15 85 0.49 2009
1980 1032 10 63 0.35 2.66
1984 1190 8 59 0.31 3.03
Source: Institute of Shortening and Edible Oils
IJH666
/30
TABLE 2
Trans Acid Levels rif Household Shortenings, .
1960 - 1984
Year
1960
1970
1975
1980
1984
...
Market Typical [Trans] Total Trans Per Capita Availability
Size Animal Veg. Acids Trans Acids PolyunsaturatE
"'-
(millions ( ~ ) (millions (g/person/day)
of pounds) of pounds)
695 10 26 148 1.01 1.04
749 10 19 126 0.76- 1.04
824 10 17 130 0075 1.. 09
709 10 17 111 0.60 0 .. 88
684 10 17 105 0.55 0,,80
Source: Institute of Shortening and Edible Oils
IJH666
_ " " " ~ r .....'
/31
o. --. -
TABLE 3
Trans and Polyunsaturated Fatty Acid Levels of Margarines,
1970 - 1984
Year
Per Capita Availability
Trans Acids Polyunsaturates
(g/person/day)
1970
1980
1984
1.5"
2.5
2.6
. ~ .
..
Source: National Association of Margarine Manufacturers
IJH666
......_-_.-
TABLE 4
Trans Acid Levels of Food Service Fats and Oils,
1969 and 1979
Total Fat
..,
Total T
2
ans Per Capita Availability
Year Production' Acids of Trans Acids
(millions of pounds) (g/person/day)
'969
1397 221 1.35
1979 2352 280 1.54
for nonfat components (butter, margarine, salad dressings)
. Adjusted for estimated wastage of deep frying fat.
Source: Institute of Shortening and Edible Oils
IJH666
TABLE 5
Trans Acid Levels of Meat and Dairy Products,
1960 - 1984
Year
1960
1970 .
1975
1980
1984
Per Capita Availab\lity
of Trans Acids
(g/person/day)
1. 54
1'.48
1.41
1.33
1.38
1Calculated by ISEO using data from:
Bunch & Hazera, Nat!. Food Rev., 26:18, 1984
Stucker & Parham, Natl. Food Rev., 25:20, 1984
IJH666
T
A
B
L
E
6
T
r
a
n
s
A
c
i
d
L
e
v
e
l
s
i
n
t
h
e
D
i
e
t
,
1
9
6
3
-
1
9
8
4
S
a
l
a
d
a
n
d
H
o
u
s
e
h
o
l
d
F
o
o
d
S
e
r
v
i
c
e
I
n
d
u
s
t
r
i
a
l
1
H
'
e
a
t
&
D
a
i
r
y
Y
e
a
r
C
o
o
k
i
n
g
0
1
1
5
S
h
o
r
t
e
n
i
n
g
s
M
a
r
g
a
r
i
n
e
s
F
a
t
s
&
O
i
l
s
F
a
t
s
&
O
i
l
s
P
r
o
d
u
c
t
s
T
o
t
a
l
s
(
g
/
p
e
r
s
o
n
/
d
a
y
)
1
9
6
3
0
.
0
6
1
.
0
1
2
.
4
8
N
A
2
1
1
.
5
4
1
9
7
0
0
.
3
8
0
.
7
6
2
.
7
3
,
1
.
3
5
,
1
.
4
8
7
.
7
0
1
9
7
5
0
.
4
9
0
.
7
5
2
.
7
3
N
A
1
1
.
4
1
-
-
1
9
8
0
0
.
3
5
0
.
6
0
2
.
7
3
'
.
5
4
1
1
.
3
3
7
.
5
5
w
'
"
1
9
8
4
0
.
3
1
0
.
5
5
'
2
.
8
5
N
A
1
1
.
3
8
'
E
s
t
i
m
a
t
e
d
2
N
A
-
N
o
t
a
v
a
i
l
a
b
l
e
I
J
H
6
6
6
0
,
,
.
.
.
.
.
.
."
~
,
jo
'
t
TABLE 7
Estimated Trans Acid Availability f r ~ m Soybean Oil (S8O) Usage.
October 1. 1982 to September 30. 1983
"
Per Capita Availability of Trans Acids (g/day)
1
1: of sao
Partially Average Trans Level in SBO
Hydrogenated 1 0 ~ 25S 3 0 ~
40 3.9 11.9 5.9
50 11.9 6.1 7.11
60 5.9 7.q 8.8
'Estimated by ISEO using data from Bunch & Hazera. natl . Food Rev 26:18,19811_
IJH666
13 ~

You might also like