Associations of Fish Intake and Dietary n

1205
Associations of Fish Intake and Dietary
n-3 Polyunsaturated Fatty Acids
With a Hypocoagulable Profile
The Atherosclerosis Risk in Communities (ARIC) Study
Eyal Shahar, Aaron R. Folsom, Kenneth K. Wu, Barbara H. Dennis, Tomoko Shimakawa,
Maureen G. Conlan, C.E. Davis, and O. Dale Williams, for the ARIC Study Investigators
Downloaded from http://atvb.ahajournals.org/ by guest on October 13, 2016
Recent epidemiological evidence indicates that the hemostatic profile is an important predictor of
cardiovascular disease, yet its dietary determinants are not well established. An important question is
whether dietary fatty acid intake influences blood levels of coagulation proteins. We examined potential
dietary determinants of six hemostatic factors —fibrinogen,factor VII, factor VIII, von Willebrand factor
(vWF), protein C, and antithrombin III —in four population-based samples totaling over 15 000
participants, blacks and whites, in the Atherosclerosis Risk in Communities (ARIC) Study. Usual dietary
intake was assessed by a food frequency questionnaire. Cross-sectional associations were explored using
multiple linear regression analysis, adjusting for gender, race, age, body mass index, smoking status,
alcohol use, diabetes, and field center. Dietary intake of n-3 polyunsaturated fatty acids (PUFAs) showed
negative associations with fibrinogen, factor VIII, and vWF (blacks and whites) and a positive association
with protein C (whites only). Fish intake, the major source of dietary n-3 PUFAs, was similarly related to
the hemostatic profile: a 1 serving per day greater fish intake was associated with the following predicted
differences (95% confidence interval): fibrinogen, - 2 . 9 mg/dL (-6.3, 0.5); factor VIII, -3.3% (-5.4,
-1.3); vWF, -2.7% (-5.2, -0.1) (blacks and whites); and protein C, +0.07 ^ig/mL (0.03, 0.11) (whites
only). Other nutrients or foods were variably associated with the hemostatic factors. These populationbased associations, although cross-sectional, suggest that increases in n-3 PUFA intake from fish may
modify the blood levels of several coagulation factors. (Arteriosclerosis and Thrombosis 1993:13:1205-1212)
KEY WORDS • blood coagulation factors • fibrinogen • factor VIII • von Willebrand factor •
protein C
R
ecent epidemiological evidence indicates that
hemostatic factors play an important role in
cardiovascular disease. Two coagulation proteins, fibrinogen and factor VII, have been shown in
prospective studies to be independent predictors of
cardiovascular events. 13 The contribution of other coagulation factors such as factor VIII, antithrombin HI
(AT-II1), and protein C to cardiovascular disease has
not been established prospectively but is quite plausible
in view of their well-known roles in thrombogenesis.
Dietary determinants of the hemostatic profile have
attracted scientific attention over the last decade, with
specific interest in the possible influence of various
Received November 12, 1992; revision accepted May 19, 1993.
From the Division of Epidemiology, School of Public Health,
University of Minnesota, Minneapolis (E.S., A.R.F.); the Division
of Hematology-Oncology, University of Texas Medical School,
Houston (K.K.W., M.G.C.); the Collaborative Studies Coordinating Center, University of North Carolina, Chapel Hill (B.H.D.,
C.E.D.); the Division of Epidemiology and Clinical Applications,
National Heart, Lung, and Blood Institute, Bethesda, Md (T.S.);
and the University of Alabama at Birmingham, School of Public
Health (O.D.W.).
Correspondence to Division of Epidemiology, School of Public
Health, University of Minnesota, Suite 300, 1300 South Second
Street, Minneapolis, MN 55454-1015 (Dr Shahar).
forms of fat intake.4 It was hypothesized that the
reduced platelet aggregability in Greenland Eskimos
might be attributed to their unique diet, which is largely
composed of marine fish rich in n-3 polyunsaturated
fatty acids (PUFAs).5 Different levels of coagulation
proteins across population samples of Japanese and
Caucasians may be related to dietary habits that presumably differ in quantity and type of fat consumption.6
Vegetarians, whose total fat intake is low, also reportedly have alterations of the hemostatic profile.7
Several prospective studies have suggested that fish
intake is inversely associated with coronary heart disease. 811 Inspired by these observations, numerous studies have examined the effect of fish oil supplements on
a variety of blood constituents. Effects on platelet
function are well established.12 Effects on the coagulation system, principally fibrinogen and AT-III, however,
have been inconsistent. Several investigators reported a
lowering of fibrinogen with fish oil supplementation,1320
but others did not corroborate this finding.2128 Conflicting results have also been reported for AT-III, ranging
from a lowering212829 to an elevating22-30 to no effect at
a ll 15.16 p e w data are available on other important
hemostatic factors. In addition, inferences from these
supplementation studies to the usual diet of populations
are limited, since the vast majority of such studies have
1206
Arteriosclerosis and Thrombosis
Vol 13, No 8 August 1993
used supplemental doses that exceed by several fold the
typical n-3 PUFA content of the Western diet.
Only a few attempts have been made to explore the
associations between Western dietary habits and hemostatic factors at the population level.3133 We had the
opportunity to pursue these issues in a large populationbased sample of participants in the Atherosclerosis Risk
in Communities (ARIC) Study.
Methods
Study Design
The overall design and operational characteristics of
the ARIC Study, which is a longitudinal investigation of
the etiology and natural history of atherosclerosis, were
described earlier.34 The ARIC Study includes two components: a community surveillance component, monitoring the occurrence of hospitalized myocardial infarction and coronary heart disease death in four US
communities, and a longitudinal cohort component established in these communities.
Downloaded from http://atvb.ahajournals.org/ by guest on October 13, 2016
Cohort Component
Cohorts of about 4000 persons aged 45 to 64 years
and including both men and women were recruited
between 1986 and 1989 in four US communities: Forsyth Co, NC; Jackson, Miss; a collection of suburbs of
Minneapolis, Minn; and Washington Co, Md. Each
cohort was a probability sample of age-eligible residents
of the local community, except that only black subjects
were eligible for selection in Jackson, Miss.
This component of the study involves repeated comprehensive clinic examinations of the members of the
cohort at 3-year intervals. The baseline evaluation
undertaken between 1986 and 1989 included medical
history, dietary questionnaire, anthropometric and
blood pressure measurements, blood sampling for lipid
and hemostatic profiles, pulmonary function studies,
and B-mode ultrasound scanning of the carotid arteries.
Cross-sectional data from that baseline examination
were used for the present analysis.
Assessment of Diet
Usual dietary habits, defined as the average intake
over the last year, were estimated in ARIC by a
semiquantitative food-frequency questionnaire, a modified version of a 61-item instrument developed and
validated by Willett et al.35 Four principal modifications
were made: (1) A few items were split into detailed
subcategories, for example, a single question addressing
fish consumption was separated into three specific fish
items used in a later version of this questionnaire. (2)
Several food items (eg, donuts and biscuits) were added.
(3) Questions regarding wine, beer, and hard liquor
were asked in another format. (4) The questionnaire
was converted from a self-administered format to interviewer administered. The final questionnaire contained
a total of 66 items.
Trained interviewers administered the questionnaire,
generally using direct data entry into a microcomputer.
Participants were asked how often on average they had
consumed a specified portion size of each food (eg, 8-oz
glass of whole milk) during the preceding year. Nine
response categories were coded ranging from "almost
never" to "more than 6 times per day." These nine
categories were transformed to servings per day using
the following weights: "almost never"=0; "1 to 3 per
month"=0.066; "1 per week"=0.14; "2 to 4 per
week"=0.43; "5 to 6 per week"=0.79; "1 per day" = 1.0;
"2 to 3 per day"=2.5; "4 to 6 per day"=5.0; and "more
than 6 per day" = 7.0. For food group analysis, daily
servings of the food items within the group were
summed, regardless of differences in portion sizes
among component foods. Alcoholic beverage consumption was evaluated by asking participants to report their
average weekly servings of wine, beer, and hard liquor.
Daily nutrient intake was calculated by multiplying
the nutrient content of the specified portion of each
food item by the frequency of its daily consumption and
summing over all items. Nutrient values for each item
were obtained from a nutrient data base.35 Nutrient
values for total fat were obtained by summing the values
of animal fat and vegetable fat.
Hemostatic Factor Measurements
Participants were asked to fast for 12 hours before the
clinic visit. A blood sample was drawn from an antecubital vein with minimal stasis. The specimen for analysis
of hemostatic factors was collected into 4.5-mL vacuum
tubes containing 1/1 Oth volume of 3.8% sodium citrate.
After centrifugation at the field center laboratory at
3000g for 10 minutes at 4°C, supernatant liquid aliquots
were stored at -70°C until final analysis. Samples were
shipped to the Central Hemostasis Laboratory at the
University of Texas at Houston, where they were assayed in a blinded fashion within 2 weeks of blood
collection. Detailed descriptions of the blood collection
and processing techniques have been published.36
Six hemostatic factors —fibrinogen,factor VII, factor
VIII, von Willebrand factor (vWF), protein C, and
AT-III—were measured. Plasma fibrinogen level was
measured by the thrombin time titration method described by Gauss. 37 Reagents for the fibrinogen assay
were obtained from General Diagnostics (OrganonTeknika Co, Morris Plains, NJ). Factor VII and factor
VIII coagulant activity were measured by determining
the ability of the testing sample to correct the clotting
time of human factor VII- or factor VIH-deficient
plasma obtained from George King Biomedical, Inc
(Overland Park, Kan). The plasma levels were then
expressed as percent activity by relating the clotting
time to a calibration curve constructed for each batch of
samples. vWF and protein C were measured by enzymelinked immunosorbent assays using kits from American
Bioproducts (Parsippany, NJ). AT-III activity was measured by the chromogenic substrate method.38 Different
batches of Universal Coagulation Reference (Thromboscreen, Pacific Hemostasis) were used for calibration
and quality control of hemostasis assays. Reliability
coefficients estimated from split specimens sent 1 week
apart to the laboratory were AT-III, 0.42; protein C,
0.56; vWF, 0.68; fibrinogen, 0.72; factor VII, 0.78; and
factor VIII, 0.86.39 These coefficients are generally
higher than those reported in other studies.39
Other Measurements
Body mass index was computed from height and
weight (in kilograms per meter squared). Height was
measured to the nearest centimeter using a vertical
metal rule. Weight was measured on a calibrated scale
Shahar et al
Fish Intake and the Coagulation Profile
1207
TABLE 1. Means and Percentiles for Estimated Dietary Intake of Selected Nutrients Among 14 571
Participants in the ARIC Study, 1986-1989
Percentile
Nutrient
Mean
25th
50th
Total fat
60.0
40.3
55.5
74.3
Saturated fat
Monounsaturated fat
Polyunsaturated fat
22.0
14.4
20.2
23.1
15.2
21.3
27.3
28.9
9.0
5.9
8.2
11.2
Animal fat
n-3 Fatty acids
36.2
0.30
23.2
0.12
33.4
0.22
0.09
0.03
0.07
45.6
0.39
0.12
0.18
0.08
0.01
0.13
0.02
Eicosapentaenoic acid
Docosahexaenoic acid
Docosapentaenoic acid
Cholesterol
Dietary fiber
Crude fiber
Carbohydrates
0.03
0.253
0.228
75th
0.23
0.04
0.314
17.2
0.162
11.4
15.9
4.3
2.8
3.9
5.3
195.4
135.3
181.5
239.5
Downloaded from http://atvb.ahajournals.org/ by guest on October 13, 2016
Caffeine
0.288
0.048
0.168
Alcohol
Total energy (kcal/d)
5.6
0.0
0.0
1610
1175
1517
21.3
0.380
5.5
1938
ARIC, Atherosclerosis Risk in Communities. Values are given in grams per day.
to the nearest pound with the subject wearing a scrub
suit and no shoes. Smoking status and alcohol use were
each categorized into four levels: current, former, never,
or unknown. Prevalent diabetes was defined as nonfasting glucose >200 mg/dL, fasting glucose >140 mg/dL,
and/or a history of or treatment for diabetes.
Statistical Methods
The cohort that completed the baseline evaluation
included 15 800 participants. This report is limited to
participants who reported their race as either white or
black. We excluded from the analysis 835 participants
who reported that they were currently taking "blood
thinning" medications. To improve the validity of the
dietary data, we excluded 329 subjects whose daily
energy intake was below the first percentile or above the
99th percentile of the sex-specific distribution, on the
assumption that the information provided was less valid.
These values were 591 and 4173 kcal/d for men and 488
and 3608 kcal/d for women. We also excluded a few
subjects with 10 or more missing responses on the
questionnaire. These exclusions left 14 571 subjects with
data available for this report. We did not exclude aspirin
users because hemostatic factor levels were unrelated to
aspirin usage.
Nutrient and food intake was described by computing
means and percentiles for the final sample. Potential
associations between nutrient or food intake and hemostatic factors were explored using linear regression
models with the hemostatic factor as the dependent
variable and the nutrient or food as a predictor. Each
model included the following predetermined list of
covariates: age, gender, race, body mass index, smoking
status, alcohol use, diabetes, and field center. All are
known to be associated with the hemostatic profile in
the ARIC Study,40-41 and their relation to dietary habits
is very plausible. The data were also analyzed by
analysis of covariance. In this approach, adjusted mean
levels of hemostatic factors were compared across quartiles of nutrient or food intake.
Initially, energy intake was considered as a potential
confounder in this analysis. Univariately, total energy
intake was inversely related to most of the hemostatic
factors. However, when these relationships were
adjusted for the predetermined covariates, only a weak
association (P<.05) of calories with vWF remained. We
concluded that energy intake was not an independent
predictor of the hemostatic factors in question, and its
potentially confounding effect was properly taken into
account by including other covariates in the models.
The associations between nutrients or foods and the
hemostatic factors may differ across race and/or gender
strata. To investigate such a possible effect modification,
two-way interaction terms involving race or gender were
tested in the linear regression models. Whenever a
significant (f<.05) interaction was found, race (or
gender)-specific associations were examined and only
those that were statistically significant (P<.05) are
reported, SAS was used for computations.42
Results
The distribution of dietary intake of selected nutrients among the study sample subjects is shown in Table
1. All of the distributions were skewed to the right to
some extent. As expected, the contribution of n-3
PUFAs to total fat intake was very small: eicosapentaenoic acid (EPA, 20:5), docosahexaenoic acid (DHA,
22:6), and docosapentaenoic acid (DPA, 22:5) together
accounted for only 0.5% of total fat intake. The Pearson
correlation coefficients between these three fatty acids
ranged from 0.87 (DHA vs DPA) to 0.97 (DHA vs
EPA).
Table 2 shows the distribution of fish and other
seafood intake. The average consumption of any fish
item was only 2 servings per week. Dark meat fish,
particularly rich in n-3 PUFAs,43 accounted for less than
25% of total fish intake. Mean±SD levels of the six
1208
Arteriosclerosis and Thrombosis Vol 13, No 8 August 1993
TABLE 2. Means and Percentiles for Dietary Fish and Other
Seafood Intake Among 14 571 Participants in the ARIC Study,
1986-1989
Food item (serving size)
Any fish
Canned tuna fish (3-4 oz)
Dark meat fish (3-5 oz)*
Other fish (3-5 oz)t
Shrimp, lobster, scallops
Percentile
50th
75th
Mean
25th
2.0
0.9
0.9
1.4
0.0
0.5
1.0
0.4
0.0
0.0
0.5
0.7
0.3
0.0
0.0
0.5
0.0
1.0
0.5
2.4
ARIC, Atherosclerosis Risk in Communities. Values are given
in servings per week.
•For example, salmon, mackerel, swordfish, sardines, bluefish.
tFor example, cod, perch, catfish.
Downloaded from http://atvb.ahajournals.org/ by guest on October 13, 2016
hemostatic factors were fibrinogen, 303 ±65 mg/dL;
factor VII, 119±29%; factor VIII, 131±39%; AT-III,
111±22%; protein C, 3.17±0.62 Mg/mL; and vWF,
118±48%.
Three n-3 PUFAs, namely EPA, DHA, and DPA,
showed negative associations with plasma levels of
fibrinogen, factor VIII, and vWF and positive associations with protein C level (Table 3). These relationships
were almost always consistent in direction and usually
consistent in magnitude across the four race and sex
strata. However, the associations of EPA and DHA with
protein C were limited to whites. There was no association of n-3 PUFAs with factor VII or AT-III. A 100-mg
greater daily intake of EPA predicted a 1.1-mg/dL lower
fibrinogen, a 0.9% lower factor VIII, a 0.9% lower vWF,
and a 0.02-^.g/mL higher protein C level. The predicted
differences in these factors for a 100-mg greater DHA
intake were about half as great as for EPA. The
associations of DPA with these factors were strong but
should be interpreted cautiously in light of the extremely low consumption of DPA among the study
population. This limitation of DPA is reflected in the
wide confidence limits of the predicted differences.
Linolenic acid (18:3), the parent compound of n-3
PUFAs, was essentially unrelated to the hemostatic
factors (data not shown).
Since fish is the major source of dietary n-3 PUFAs,44
we further explored the relationships at the food-item
level. Table 4 shows the predicted differences in fibrinogen, factor VIII, vWF, and protein C levels associated
with a 1 serving per day greater intake of different fish
items. The findings paralleled those for the nutrient
data, but most associations (P<.05) were limited to the
"other fish" item and the "any fish" variable. As before,
the relationships were almost always consistent across
the four race and sex strata, although the relations with
protein C were again restricted to whites.
In another analytic approach, participants were categorized according to quartiles of fish consumption.
Adjusted mean levels of the hemostatic factors were
computed for each quartile from an analysis of covariance model. Although a graded relationship was not
always apparent, participants in the upper consumption
quartile had the lowest levels of fibrinogen, factor VIII,
and vWF and the highest protein C level (Figure).
Arachidonic acid (20:4), an n-6 PUFA whose physiology and metabolism sometimes compete with those of
the n-3 PUFAs, was not associated with any hemostatic
factor (data not shown). Likewise, neither total fat, total
saturated fatty acids, nor total monounsaturated fatty
acids were associated (/)<.O5) with any hemostatic
factor in the study sample. However, among white men
(n=4834) total fat intake as well as saturated fatty acid
and total energy intake were positively albeit weakly
related to factor VII. A 10-g/d greater fat intake
predicted a 0.3% higher factor VII level (95% confidence interval, 0.1% to 0.5%).
Associations of several other nutrients with hemostatic factors are presented in Table 5. Statistically
significant (P<.05) associations are identified. Cholesterol intake was positively associated with fibrinogen,
factor VII, factor VIII, and vWF, although most of these
associations were restricted to men. Animal fat intake
showed a similar pattern. The data suggested modest
inverse relationships of crude fiber intake with factor
VII and factor VIII levels. Similar associations with
factor VII and factor VIII were found for total fiber
intake and consumption of all fruits and all vegetables
(data not shown).
Caffeine was inversely related to factor VII and factor
VIII (blacks and whites) and to vWF (whites only).
Alcohol intake was negatively associated with plasma
fibrinogen in men and women and with factor VIII and
vWF in women only (Table 5). Of note, ARIC data did not
suggest any important dietary determinants of AT-III.
Discussion
Despite much scientific effort over the last decade, the
relationship of diets and dietary fat in particular to
coagulation factors has not been clearly established. The
TABLE 3. Predicted Differences and 95% Confidence Intervals* in Hemostatic Factor Levels Associated With a 100-mg Greater Daily
Intake of n-3 Polyunsaturated Fatty Acids
Fatty acid
Fibrinogen
(mg/dL)
Eicosapentaenoic
acid
-1.1 (-2.2,-0.1)t
Docosahexaenoic
acid
-0.6 (-1.2,0.0)
Docosapentaenoic
acid
-3.9 (-7.6,-0.2)t
•From linear regression adjusted for
t/'<.05.
tWhites only.
Hemostatic factor
von Willebrand
factor (%)
Factor VII
Factor VIII
0.2 (-0.3,0.7)
-0.9 (-1.6,-0.3)t
0.1 (-0.2,0.4)
-0.6 (-0.9,-0.2)t
Protein C
Antithrombin III
-0.9 (-1.7,-0.1)t
0.02 (0.01, 0.04)tt
0.0 (-0.4, 0.4)
-0.5 (-0.9,0.0)t
0.01(0.01,0.02)^
0.0 (-0.2,0.2)
0.5 (-1.2, 2.2) -3.6 (-5.8, -1.4)t -3.7 (-6.4,-0.9)t 0.05 (0.02,0.09)t
-0.4 (-1.7, 0.9)
age, gender, race, body mass index, smoking status, alcohol use, diabetes, and field center.
Shahar et al
Fish Intake and the Coagulation Profile
1209
TABLE 4. Predicted Differences and 95% Confidence Intervals* in Hemostatic Factor Levels Associated With a 1- Serving Greater Daily
Intake of Fish and Other Seafood
Fibrinogen
(mg/dL)
Food item
(serving size)
Hemostatic factor
Factor VIII
von Willebrand
)
factor (%)
Factor VII
Any
fish
-2.9 (-6.3, 0.5) 0.5 (-1.1, 2.0) -3.3 (-5.4, -1.3)t
Canned tuna fish
(3-4 oz)
-1.3 (-6.7, 4.0) 0.2 (-2.3, 2.7) -3.0 (-6.2, 0.2)
Dark meat fish
(3-5 oz)
-9.2 (-18.7, 0.3) 1.6 (-2.7, 5.9) -3.4 (-9.1, 2.4)
Other fish
(3-5 oz)
-5.1 (-12.2, 2.0) 0.6 (-2.7, 3.9) -7.1 (-11.4, -2.8)t
Shrimp, lobster,
scallops
-6.4 (-20.3, 7.4) 1.7 (-4.5,7.9) -4.7 (-13.0,3.6)
*From linear regression adjusted for age, gender, race, body mass index,
Protein C
(fig/mh)
Antithrombin III
-2.7 (-5.2,-0.1)t
0.07 (0.03, 0.11 )tt -0.3 (-1.4, 0.9)
-0.8 (-4.8, 3.2)
0.04 (-0.01, 0.10) -0.5 (-2.3, 1.4)
-5.7 (-12.8, 1.4)
0.19 (0.07, 0.31)t*
-6.9 (-12.3, -1.6)t
0.08 (0.01, 0.16)t
1.7 (-1.5, 5.0)
-1.3 (-3.8, 1.2)
-4.5 (-14.9,5.9)
-0.11 (-0.25,0.02) -4.4 (-9.1,0.3)
smoking status, alcohol use, diabetes, and field center.
t/J<.05.
^Whites only.
supplementation of the diet with various doses of fish
oil. A dose-response relation has also been suggested.20
Few feeding studies have examined the effect of n-3
PUFAs on vWF or factor VIII. Lowering of the vWF
level was reported in two trials20'48; a single study
reported lowering of the factor VIII level.28 To our
knowledge, this is the first report corroborating these
associations within a population-based sample consuming a Western diet.
A positive association of n-3 PUFA intake with
protein C has not been reported previously. Three
feeding studies1516'29 did not find any effect, and another28 suggested that EPA and DHA feeding lowered
protein C level. The associations in this study were
statistically significant but were limited to whites. We
suspect that the race difference may be due to the
smaller sample size in blacks. Reanalysis by the four
an
o
c
\m H SBi
o
120
115
ifiaflg*
mm
MM I
mm
mm
IM
130
125
i.
fact
310"
305"
300"
uSBEKBBa
295"
HPH HBII fmmm
290"
285"
280- ^ • D B E B L M B B B B B L
1
2
3
4
Willc ranc
E
£
315"
105
t
•
4C
^
*
%
•K
j»
*
*.
1 "
• * *
>
:
•*
*
1
2
3
4
Quartile of Fish Intake
an
3
• •
*
i
t
y
100
^^
an •
*'
*'
J
«
a
Quartile of Fish Intake
145
140
135
130
125120
115
110
i
110
von
at
JZU
/ml
Downloaded from http://atvb.ahajournals.org/ by guest on October 13, 2016
study data reported here provide insights into these relationships at the population level. Dietary n-3 PUFAs,
derived primarily from fish, were found to be negatively
associated with the three hemostatic factors fibrinogen,
factor VIII, and vWF. They were positively associated with
the coagulation inhibitor protein C, although the clinical
implication of this finding is uncertain. Heterozygous
protein C deficiency did not seem to confer an increased
risk of venous thrombosis,45 and the role of protein C in
arterial thrombosis is controversial.4647 Overall, the observed relationships with the four factors suggest an
association of these fatty acids with reduced blood coagulability. The inverse relationship with fibrinogen, albeit
weak, could be considered favorable with respect to the
risk of cardiovascular disease.
Numerous human feeding studies13-20 but not all21-28
have reported a lowering of plasma fibrinogen after
i
e
j
.5
3.4
3.3
3.2
HSBH wSSmi
IWI
JHLJBL
1
2
3
C
'v
3.1
o
3.0
mmm
4
Quartile of Fish Intake
2.9
1
2
3
4
Quartile of Fish Intake
Adjusted mean (SE) hemostatic
factor levels according to quartile of
fish consumption. Mean levels were
adjusted for gender, race, age, body
mass index, smoking status, alcohol use, diabetes, and field center
using analysis of covariance. Protein C results are for whites only.
1210
Arteriosclerosis and Thrombosis
Vol 13, No 8 August 1993
TABLE 5. Predicted Differences and 95% Confidence Intervals* in Hemostatic Factor Levels Associated With a Specified Unit Greater
Daily Intake of Various Nutrients
Hemostatic factor
Nutrient
(unit)
Fibrinogen
(mg/dL)
Factor VII
Factor VIII
(%)
von Willebrand
factor (%)
Protein C
(jtg/mL)
Antithrombin
III (%)
Cholesterol
0.7 (0.2,1.1 )tt
0.6 (0.0,1.2)tt
0.9 (0.2,1.7)tt
0.00 (-0.01, 0.01) 0.1 (-0.2, 0.3)
(100 mg)
0.9(0.1,1.7)t
Animal fat
0.8 (0.2, 1.3)tt
0.00 (-0.01, 0.01) 0.1 (-0.1, 0.3)
0.4(0.1,0.7)tt
0.1 (-0.3, 0.4)
(10 g)
0.7 (0.1,1.4)t§
Crude fiber
-0.1 (-0.4, 0.3)
0.00 (-0.01, 0.00) -0.1 (-0.3, 0.0)
-0.3 (-0.6, -0.1)t -0.5 (-0.7, -0.2)t
(lg)
0.0 (-0.5,0.5)
Caffeine
-0.6 (-0.9, -0.3)t§ 0.00 (-0.01, 0.00) 0.1 (-0.1, 0.2)
-0.2 (-0.4, -0.1)t -0.4 (-0.7, -0.2)t
(100 mg)
0.3 (-0.1, 0.7)
Alcohol
-2.8 (-4.1, -1.6)t|| -4.1 (-5.6, -2.5)t|| 0.00 (0.00, 0.00)
0.0 (0.0, 0.0)
(10 g)
-2.5 (-3.3,-1.6)t
0.2 (-0.2, 0.5)
*From linear regression adjusted for age, gender, race, body mass index, smoking status, alcohol use, diabetes, and field center. Alcohol
use was not included as a covariate in models for alcohol intake. Gender (or race)-specific models did not include the gender (or race)
variable.
t/ > <.05.
Downloaded from http://atvb.ahajournals.org/ by guest on October 13, 2016
JMen only.
§Whites only.
||Women only.
race and sex strata revealed that the associations were
similar for white men, white women, and black women
but not for black men. The latter group contained by far
the smallest number of participants (n=1501).
Intake of fish, particularly dark meat fish, is the main
source of dietary n-3 PUFAs43-44 and is strongly correlated with plasma levels of EPA and DHA.49 Dietary
n-3 PUFAs, as estimated by food-frequency questionnaires similar to the ARIC dietary questionnaire, also
correlate with the levels of these fatty acids in plasma
phospholipids50 and adipose tissue.51 Not surprisingly,
the relations with hemostatic factors were consistent for
both fish consumption and n-3 PUFA intake. Statistical
significance (P<.05) for the most part, however, was
limited to the "other fish" and "any fish" variables
rather than to the "dark meat fish" item. Although
"dark meat fish" are rich in n-3 PUFAs, participants
reported to have consumed them far less frequently
than "other fish" (Table 2). The contribution of the
"other fish" category to total EPA, DHA, and DPA
intake was as great as that of "dark meat fish" (about
30%) and was particularly substantial to DPA intake
(about 50%). A relatively small contribution of "other
seafood," also known to be rich in n-3 PUFAs, to n-3
PUFA intake might explain the lack of any significant
association of this item with the hemostatic factors.
ARIC data did not suggest any relation between dietary
n-3 PUFAs and AT-III. Although Eskimos, who consume
large amounts of EPA and DHA, reportedly have elevated AT-III levels,52 results of human feeding studies
have varied.1516-21'22-28-30 A rather large variability in the
laboratory measurement of AT-III may account for the
inconsistencies. The chromogenic substrate method that
was used to measure AT-III in this and many other studies
did not prove to be highly reliable; the reliability coefficient in ARIC was only 0.42, the lowest of all the six
hemostatic factors measured.39
Cross-cultural53 and prospective811 studies have suggested an inverse relationship between fish intake and
coronary heart disease. Suggested mechanisms include
an improved lipid profile,54 decreased platelet aggrega-
bility,55 and enhanced fibrinolytic activity.56 Another
potential mechanism may be hypocoagulability. Plasma
fibrinogen level has been reported to be an independent
predictor of cardiovascular events 13 ; factor VIII plays
an important role in the intrinsic coagulation pathway57;
and vWF regulates platelet adhesion to the vessel wall.57
All of these factors were inversely related to fish intake.
The biological mechanisms underlying the associations remain to be elucidated. Both fibrinogen and
protein C are synthesized by hepatocytes, whereas vWF
is synthesized in extrahepatic sites, mostly in endothelial
cells.57 The exact site of synthesis of the active component of factor VIII is unknown. Recently it was reported
that fibrinogen production by hepatoma cells was inhibited in fish oil media.58 Animal studies suggest that
ingestion of EPA and DHA increases their content in
liver mitochondrial and microsomal membranes59-60 and
renders the latter more susceptible to lipid peroxidation.60 The resultant alterations in cellular membranes
may influence the production of fibrinogen and protein
C in the hepatocyte. The effect of n-3 PUFAs on vWF
synthesis in endothelial cells has not been studied.
However, fish oils were found to modulate production
of other substances by endothelial cells, such as plateletderived growth factor-like protein61 and endotheliumderived relaxing factor.62
With the possible exception of animal fat intake,
other dietary fats generally did not prove to be important determinants of the hemostatic factors in question.
Dietary cholesterol, however, was positively associated
with plasma fibrinogen, a finding consistent with a
previous report of higher fibrinogen levels in subjects on
a habitual high-cholesterol diet.63
Two feeding studies suggested that total fat intake is
positively related to factor VII coagulant activity,64-65
but a more recent report did not corroborate this
finding.66 ARIC data showed only a weak association
between total fat intake and factor VII, which was
restricted to white men. The single other populationbased study reporting this relation was conducted in
white men as well.67 However, it is uncertain whether fat
Shahar et al
Downloaded from http://atvb.ahajournals.org/ by guest on October 13, 2016
consumption per se or total energy intake primarily
accounted for this association. Both were related to
factor VII level in ARIC white men and, as expected,
were highly intercorrelated (r=.9). Adjustment for caloric intake using the nutrient residual method68 suggested that total energy intake rather than fat consumption was the main determinant of factor VII.
Several limitations of this study should be mentioned.
First, cross-sectional relationships do not necessarily
imply causal associations. Fish intake, for example, may
be a surrogate measure for a yet-undetermined factor
that influences the hemostatic profile. Nonetheless, the
causal nature of many of the associations revealed is
supported by previous experimental evidence. Second,
n-3 PUFAs were highly intercorrelated in this data set.
Theoretically, only one of them might have been causally related to the hemostatic profile, whereas the other
two might have been merely surrogate measurements.
Finally, the range of n-3 PUFA consumption among the
study participants was remarkably narrow. Predictions
outside this range must be made cautiously.
Many of the observed associations were weak. However, it should be remembered that the strong association between dietary lipids and serum cholesterol level
demonstrated consistently in feeding studies is undetectable in most cross-sectional studies.69 Dietary habits
as well as hemostatic factors are measured with considerable error, which is likely to attenuate existing associations. Thus, the true magnitude of the relationships
between dietary habits and hemostatic factors could be
greater than apparent from these analyses. Based on
risk estimates from the Northwick Park Heart Study,2
increasing fish consumption to 1 serving per day might
reduce plasma fibrinogen and thereby lower the risk of
ischemic heart disease death by 1% to 2%. The true risk
reduction could be greater, given the attenuation effect
of measurement error. The clinical significance of the
observed relationships with factor VIII, vWF, and protein C is yet to be determined.
Acknowledgments
Supported by National Heart, Lung, and Blood Institute
contracts N01-HC-55015, N01-HC-55016, N01-HC-55018,
N01-HC-55019, N01-HC-55020, N01-HC-55021, and N01HC-55022. The authors thank Walter Willett and Laura
Sampson for their permission to use and assistance with the
dietary questionnaire. The authors also thank the following
ARIC personnel: Alfredo H. Figueroa, Jane G. Johnson,
Robert Smith, and Jessie M. Spencer of the Jackson Field
Center; Joel Hill, Audrey Harrell, Patricia Hawbaker, and
Trudy Littenberg of the Washington Co Field Center; Myra
Carpenter of the Collaborative Studies Coordinating Center,
University of North Carolina; Donna J. Neal and Patricia A.
Smith of the ARIC Project Office, National Heart, Lung, and
Blood Institute; and Laura Kemmis and Mike Wickersheimer
for manuscript preparation.
References
1. Meade TW, North WRS, Chakrabarti R, Stirling Y, Haines AP,
Thompson SG. Haemostatic function and cardiovascular death:
early results of a prospective study. Lancet. 1980;l:1050-1054.
2. Meade TW, Mellows S, Brozovic M, Miller GJ, Chakrabarti RR,
North WRS, Haines AP, Stirling Y, Imeson JD, Thompson SG.
Haemostatic function and ischaemic heart disease: principal
results of the Northwick Park Heart Study. Lancet. 1986;2:533-537.
3. Kannel WB, Wolf PA, Castelli WP, D'Agostino RB. Fibrinogen
and risk of cardiovascular disease: the Framingham Study. JAMA.
1987;258:1183-1186.
Fish Intake and the Coagulation Profile
1211
4. Nordoy A, Goodnight SH. Dietary lipids and thrombosis: relationships to atherosclerosis. Arteriosclerosis. 1990;10:149-163.
5. Dyerberg J, Jorgensen KA. Marine oils and thrombogenesis. Prog
LipidRes. 1982;21:255-269.
6. Iso H, Folsom AR, Wu KK, Finch A, Munger RG, Sato S, Shimamoto T, Terao A, Komachi Y. Hemostatic variables in Japanese
and Caucasian men: plasma fibrinogen, factor VI Ic, factor VIIIc,
and von Willebrand factor and their relations to cardiovascular
disease risk factors. Am J Epidemiol. 1989;130:925-934.
7. Haines AP, Chakrabarti R, Fisher D, Meade TW, North WRS,
Stirling Y. Haemostatic variables in vegetarians and nonvegetarians. ThrombRes. 1980;19:139-148.
8. Kromhout D, Bosschieter EB, Coulander C de L. The inverse
relation between fish consumption and 20-year mortality from
coronary heart disease. N Engl J Med. 1985;312:1205-1209.
9. Shekelle RB, Missell LV, Paul O, Shryock AM, Stamler J. Fish
consumption and mortality from coronary heart disease. N Engl J
Med. 1985;313:820-824.
10. Norell SE, Ahlbom A, Feychting M, Pedersen NL. Fish consumption and mortality from coronary heart disease. Br Med J. 1986;
293:426.
11. Burr ML, Fehily AM, Gilbert JF, Rogers S, Holliday RM, Sweetnam PM, Elwood PC, Deadman NM. Effects of changes in fat, fish,
and fibre intakes on death and myocardial reinfarction: Diet and
Reinfarction Trial (DART). Lancet. 1989;2:757-761.
12. Leaf A, Weber PC. Cardiovascular effects of n-3 fatty acids. N Engl
J Med. 1988;318:549-557.
13. Hostmark AT, Bjerkedal T, Kierulf P, Flaten H, Ulshagen K. Fish
oil and plasma fibrinogen. Br Med J. 1988;297:180-181.
14. Saynor R, Gillott T. Fish oil and plasma fibrinogen. Br Med J.
1988;297:1196.
15. Radack K, Deck C, Huster G. Dietary supplementation with lowdose fish oils lowers fibrinogen levels: a randomized, double-blind
controlled study. Ann Intern Med. 1989;111:757-758.
16. Radack K, Deck C, Huster G. The comparative effects of n-3 and
n-6 polyunsaturated fatty acids on plasma fibrinogen levels: a
controlled clinical trial in hypertriglyceridemic subjects. J Am Coll
Nutr. 1990;9:352-357.
17. Haglund O, Wallin R, Luostarinen R, Saldeen T. Effects of a new
fluid fish oil concentrate, ESKIMO-3, on triglycerides, cholesterol,
fibrinogen and blood pressure. J Intern Med. 1990;227:347-353.
18. Flaten H, Hostmark AT, Kierulf P, Lystad E, Trygg K, Bjerkedal
T, Osland A. Fish-oil concentrate: effects on variables related to
cardiovascular disease. Am J Clin Nutr. 1990;52:300-306.
19. Cobiac L, Clifton PM, Abbey M, Belling GB, Nestel PJ. Lipid,
lipoprotein, and hemostatic effects of fish vs fish-oil n-3 fatty acids
in mildly hyperlipidemic males. Am J Clin Nutr. 1991;53:1210-1216.
20. Schmidt EB, Varming K, Ernst E, Madsen P, Dyerberg J. Doseresponse studies on the effect of n-3 polyunsaturated fatty acids on
lipids and haemostasis. Thromb Haemost. 1990;63:l-5.
21. Sanders TAB, Vickers M, Haines AP. Effect on blood lipids and
haemostasis of a supplement of cod-liver oil, rich in eicosapentaenoic and docosahexaenoic acids, in healthy young men. Clin Sci.
1981;61:317-324.
22. Mortensen JZ, Schmidt EB, Nielsen AH, Dyerberg J. The effect of
n-6 and n-3 polyunsaturated fatty acids on hemostasis, blood lipids
and blood pressure. Thromb Haemost. 1983;50:543-546.
23. Rogers S, James KS, Butland BK, Etherington MD, O'Brien JR,
Jones JG. Effects of a fish oil supplement on serum lipids, blood
pressure, bleeding time, haemostatic and rheological variables: a
double blind randomised controlled trial in healthy volunteers.
Atherosclerosis. 1987;63:137-143.
24. Tulleken JE, Limburg P, van Rijswijk MH. Fish oil and plasma
fibrinogen. Br Med J. 1988;297:615-616.
25. Schmidt EB, Kristensen SD, Dyerberg J. The effect of fish oil on
lipids, coagulation and fibrinolysis in patients with angina pectoris.
Artery. 1988;15:316-329.
26. Gans ROB, Bilo HJG, Schouten JA, Rauwerda JA. Fish oil and
plasma fibrinogen. Br Med J. 1988;297:978-979.
27. Hansen J-B, Olsen JO, Wilsgard L, Osterud B. Effects of dietary
supplementation with cod liver oil on monocyte thromboplastin
synthesis, coagulation and fibrinolysis. J Intern Med. 1989;
225(suppl 1):133-139.
28. Lox CD. The effects of dietary marine fish oils (omega-3 fatty acids)
on coagulation profiles in men. Gen Pharmacol. 1990;21:241-246.
29. Schmidt EB, Kristensen SD, Sorensen PJ, Dyerberg J. Antithrombin III and protein C in stable angina pectoris: influence of dietary
supplementation with polyunsaturated fatty acids. Scand J Clin
Lab Invest. 1988;48:469-473.
1212
Arteriosclerosis and Thrombosis
Vol 13, No 8 August 1993
Downloaded from http://atvb.ahajournals.org/ by guest on October 13, 2016
30. Stoffersen E, Jorgensen KA, Dyerberg J. Antithrombin III and
dietary intake of polyunsaturated fatty acids. Scand J Clin Lab
Invest. 1982;42:83-86.
31. Fehily AM, Milbank JE, Yarnell JWG, Hayes TM, Kubicki AJ,
Eastham RD. Dietary determinants of lipoproteins, total cholesterol, viscosity, fibrinogen, and blood pressure. Am J Clin Nutr.
1982;36:890-896.
32. Yarnell JWG, Fehily AM, Milbank J, Kubicki AJ, Eastham R,
Hayes TM. Determinants of plasma lipoproteins and coagulation
factors in men from Caerphilly, South Wales. J Epidemiol Community Health. 1983;37:137-140.
33. Rogers S, Yarnell JWG, Fehily AM. Nutritional determinants of
haemostatic factors in the Caerphilly study. EurJ Clin Nutr. 1988;
42:197-205.
34. The ARIC Investigators. The Atherosclerosis Risk in Communities
(ARIC) Study: design and objectives. Am J Epidemiol. 1989;129:
687-702.
35. Willett WC, Sampson L, Stampfer MJ, Rosner B, Bain C, Witschi
J, Hennekens CH, Speizer FE. Reproducibility and validity of a
semiquantitative food frequency questionnaire. Am J Epidemiol.
1985;122:51-65.
36. Papp AC, Hatzakis H, Bracey A, Wu KK. ARIC hemostasis
study: I. development of blood collection and processing systems
suitable for multicenter hemostatic studies. Thromb Haemost.
1989;61:15-19.
37. Clauss A. Gerinnungsphysiologische Schnellmethode zur Bestimmung des Fibrinogens. Ada Haematol. 1957; 17:237-246.
38. National Heart, Lung, and Blood Institute. Atherosclerosis Risk in
Communities (ARIC) Study manual of operation No. 9: Hemostasis
Determinations. Chapel Hill, NC: ARIC Coordinating Center,
School of Public Health, University of North Carolina.
39. Chambless LE, McMahon R, Wu K, Folsom A, Finch A, Shen
Y-L. Short-term intraindividual variability in hemostasis factors:
the ARIC Study. Ann Epidemiol. 1992;2:723-733.
40. Folsom AR, Wu KK, Davis CE, Conlan MG, Sorlie PD, Szklo M.
Population correlates of plasma fibrinogen and factor VII, putative
cardiovascular risk factors. Atherosclerosis. 1991 ;91:191 -205.
41. Folsom AR, Wu KK, Conlan MG, Finch A, Davis CE, Marcucci
G, Sorlie PD, Szklo M. Distributions of hemostatic variables in
blacks and whites: population reference values from the Atherosclerosis Risk in Communities (ARIC) Study. Ethnicity Dis. 1992;
2:35-46.
42. SAS Institute Inc. SAS/STAT* User's Guide, version 6, ed 4, vol 2.
Cary, NC: SAS Institute Inc; 1989:1-846.
43. Hepburn FN, Exler J, Weihrauch JL. Provisional tables on the
content of omega-3 fatty acids and other fat components of
selected foods. J Am Diet Assoc. 1986;86:788-793.
44. Nettleton JA. w-3 fatty acids: comparison of plant and seafood
sources in human nutrition. J Am Diet Assoc. 1991;91:331-337.
45. Miletich J, Sherman L, Broze G Jr. Absence of thrombosis in
subjects with heterozygous protein C deficiency. N Engl J Med.
1987;317:991-996.
46. O'Connor NT, Brockmans AW, Bertina RM. Protein C values in
coronary artery disease. BrMedJ. 1984;289;1192.
47. Cortellaro M, Boschetti C, Confrancesco E, Zanussi C, Catalano
M, de Gaetano G, Gabrielli L, Lombardi B, Specchia G, Tavazzi L,
Tremoli E, della Volpe A, Polli E, and the PLAT Study Group.
The PLAT Study: hemostatic function in relation to atherothrombotic ischemic events in vascular disease patients: principal results.
Arterioscler Thromb. 1992;12:1063-1070.
48. Miller ME, Anagnostou AA, Ley B, Marshall P, Steiner M.
Effect offish oil concentrates on hemorheological and hemostatic
aspects of diabetes mellitus: a preliminary study. Thromb Res.
1987;47:201-214.
49. Iso H, Sato S, Folsom AR, Shimamoto T, Terao A, Munger RG,
Kitamura A, Konishi M, Iida M, Komachi Y. Serum fatty acids and
fish intake in rural Japanese, urban Japanese, Japanese American
and Caucasian American men. Int J Epidemiol. 1989;18:374-381.
50. Silverman DI, Reis GJ, Sacks FM, Boucher TM, Pasternak RC.
Usefulness of plasma phospholipid n-3 fatty acid levels in predicting dietary fish intake in patients with coronary artery disease. Am
J Cardiol. 1990;66:860-862.
51. London SJ, Sacks FM, Caesar J, Stampfer MJ, Siguel E, Willett WC.
Fatty acid composition of subcutaneous adipose tissue and diet in
postmenopausal US women. Am J Clin Nutr. 1991;54:340-345.
52. Jorgensen KA, Nielsen AH, Dyerberg J. Hemostatic factors and
renin in Greenland Eskimos on a high eicosapentaenoic acid
intake: results of the Fifth Umanak Expedition. Ada Med Scand.
1986;219:473-479.
53. Armstrong BK, Mann Jl, Adelstein AM, Eskin F. Commodity
consumption and ischaemic heart disease mortality, with special
reference to dietary practices. J Chronic Dis. 1975;28:455-469.
54. Harris WS. Fish oils and plasma lipid and lipoprotein metabolism
in humans: a critical review. J Lipid Res. 1989;30:785-807.
55. von Schacky C, Weber PC. Metabolism and effects on platelet
function of the purified eicosapentaenoic and docosahexaenoic
acids in humans. J Clin Invest. 1985;76:2446-2450.
56. Barcelli U, Glas-Greenwalt P, Pollak VE. Enhancing effect of
dietary supplementation with cu-3 fatty acids on plasma fibrinolysis
in normal subjects. Thromb Res. 1985;39:307-312.
57. Ogston D. The Physiology of Hemostasis. Cambridge, Mass: Harvard University Press; 1983.
58. Zyglewska T, Merrill C, Conlan MG, Wu KK. n-3 fatty acids
selectively inhibit fibrinogen production by cultured human hepatoma (HepG2) cell lines. Thromb Haemost. 1991;65:857. Abstract.
59. Laganiere S, Yu BP, Fernandes G. Studies on membrane lipid
peroxidation in omega-3 fatty acid-fed autoimmune mice: effect of
vitamin E supplementation. Adv Exp Med Biol. 1990;262:95-102.
60. Garrido A, Garrido F, Guerra R, Valenzuela A. Ingestion of high
doses of fish oil increases the susceptibility of cellular membranes
to the induction of oxidative stress. Lipids. 1989;24:833-835.
61. Fox PL, DiCorleto PE. Fish oils inhibit endothelial cell production
of platelet-derived growth factor-like protein. Science. 1988;241:
453-456.
62. Shimokawa H, Lam JYT, Chesebro JH, Bowie EJW, Vanhoutte
PM. Effects of dietary supplementation with cod-liver oil on endothelium-dependent responses in porcine coronary arteries. Circulation. 1987;76:898-905.
63. Vorster HH, Silvis N, Venter CS, van Ryssen JJ, Huisman H, van
Eeden TS, Walker ARP. Serum cholesterol, lipoproteins, and
plasma coagulation factors in South African blacks on a high-egg
but low-fat intake. Am J Clin Nutr. 1987;46:52-57.
64. Miller GJ, Martin JC, Webster J, Wilkes H, Miller NE, Wilkinson
WH, Meade TW. Association between dietary fat intake and
plasma factor VII coagulant activity: a predictor of cardiovascular
mortality. Atherosclerosis. 1986;60:269-277.
65. Marckmann P, Sandstrom B, Jespersen J. Effects of total fat content and fatty acid composition in diet on factor VII coagulant
activity and blood lipids. Atherosclerosis. 1990;80:227-233.
66. Marckmann P, Sandstrom B, Jespersen J. Fasting blood coagulation and fibrinolysis of young adults unchanged by reduction in
dietary fat content. Arterioscler Thromb. 1992;12:201-205.
67. Miller GJ, Cruickshank JK, Ellis LJ, Thompson RL, Wilkes HC,
Stirling Y, Mitropoulos KA, Allison JV, Fox TE, Walker AO. Fat
consumption and factor VII coagulant activity in middle-aged men:
an association between a dietary and thrombogenic coronary risk
factor. Atherosclerosis. 1989; 78:19-24.
68. Willett W. Nutritional Epidemiology. New York, NY: Oxford University Press; 1990.
69. Jacobs DR Jr, Anderson JT, Blackburn H. Diet and serum cholesterol: Do zero correlations negate the relationship? Am J Epidemiol. 1979;110:77-87.
Downloaded from http://atvb.ahajournals.org/ by guest on October 13, 2016
Associations of fish intake and dietary n-3 polyunsaturated fatty acids with a
hypocoagulable profile. The Atherosclerosis Risk in Communities (ARIC)
Study.
E Shahar, A R Folsom, K K Wu, B H Dennis, T Shimakawa, M G Conlan, C E
Davis and O D Williams
Arterioscler Thromb Vasc Biol. 1993;13:1205-1212
doi: 10.1161/01.ATV.13.8.1205
Arteriosclerosis, Thrombosis, and Vascular Biology is published by the American Heart
Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1993 American Heart Association, Inc. All rights reserved.
Print ISSN: 1079-5642. Online ISSN: 1524-4636
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://atvb.ahajournals.org/content/13/8/1205
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles
originally published in Arteriosclerosis, Thrombosis, and Vascular Biology can be obtained via
RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the
online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further
information about this process is available in thePermissions and Rights Question and Answer
document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Arteriosclerosis, Thrombosis, and Vascular
Biology is online at:
http://atvb.ahajournals.org//subscriptions/