The effects of saturated fat and n-6 polyunsaturated fat on postprandial lipemia and hemostatic activity

ATHEROSCLEROSIS

Atherosclerosis

103 (1993)

I - 11

The effects of saturated fat and n-6 polyunsaturated fat on postprandial lipemia and hemostatic activity V. Salomaa* a, V. Rasi”, J. Pekkanenb, M. Jauhiainen”, E. Vahterae, H. Korhonena, K. Kuulasmaaa, C. Ehnholm”

P. Pietinend,

‘Dept. of Epidemiology and Health Promotion, ‘Dept. of Environmental Epidemiology, “Dept. of Biochemistry, ‘Dept. of Nutrition, National Public Health Institute. Elimiienkatu 25 A, 6th Floor, FIN-00510 Helsinki, Finland ‘Coaguiaiion iaboraiory. Finnish Red Goss Biood Transfusion Service, iieisinki. Finiand (Received

20 November

1992; revision

received 4 May 1993; accepted

11 May 1993)

Abstract The effect of different fat loads on postprandial lipemia and hemostatic activity was examined in 10 middle-aged men using 3 different meals. One meal was rich in saturated fatty acids (cream), the other rich in n-6 polyunsaturated fatty acids (sunflower oil) and the third was fat-free containing only carbohydrates. Lipoprotein lipids and hemostatic parameters were measured during fasting and 2, 4, 6 and 8 h after the test meal. In fasting samples, several hemostatic parameters were significantly associated with lipoprotein lipids. Most notable were the strong associations of librinolysis parameters tissue plasminogen activator antigen and plasminogen activator inhibitor activity (PAI-1) with total and very low density lipoprotein’ (VLDL) triglycerides. During lipemia, the associations were approximately similar or slightly weaker than in the fasting state. Both fat loads resulted in similar postprandial lipid responses: VLDL and high density lipoprotein (HDL) triglycerides reached maximum at 4 h after the meal. VLDL cholesterol also increased 4 and 6 h after the fat loads. HDL, cholesterol declined after the fatty meals but no change was observed in the HDL, fraction. The fat-free meal gave no significant lipid changes during the time course studied. Factor VII activity (F VILC) increased 6 and 8 h after the fatty meals, whereas a decrease was observed after the fat-free meal The changes (f SD.) at 8 h after cream, sunflower oil and fat-free meal were 5.2 f 3.3, 3.3 f 4.2 and -5.8 f 7.9 percentage points, respectively, and the effect of the meal on the changes was statistically significant (F ,x,99j= 2.99, P = 0.0048). F VII antigen (F VII:Ag) tended to decline during the day but there was no difference between the meals. Factor VIII activity (F VIII:C) was highest after the polyunsaturated fat meal and lowest after the fat-free meal. PAI,a,.,.,:_,.J A....:_.,. l.._ ua.y A,... dllU ,._,J rl_,. A_..,:^,. ,,...A,.A f^ _._____* dllC1 ..rr_.. &L_ r___ -__-:-_ .cc-. “1 -L-*1.UCUI‘ IE;”Uu‘ Ulg ?? UK LUG“CLIIIIG LGIIUCU LOL_ “C >lCCptX UK r..r laL-I,ce Ill”, mng -_-I mea,. lY__ 1 ne e,,eci me __.-I ,,,ea, on the changes in lipoprotein lipids and hemostatic factors varied significantly between individuals. In conclusion, postprandial lipemia after a single fatty meal was associated with procoagulatory change in F VII:C but there was no difference between saturated fat and n-6 polyunsaturated fat.

Key words: Hemostatic sclerosis

* Corresponding

author.

activity;

Postprandial

lipemia;

Saturated

fats; Polyunsaturated

Fax: +358-O-146 2185.

0 1993 Elsevier Scientific 0021-9150/93/$06.00 SSDI 002 I-9 150(93)05084-I

Publishers

Ireland

Ltd. All rights reserved.

fats; Thrombosis;

Athero-

2

V. Salomaa et al. /Atherosclerosis

1. Introduction Most measurements of coronary heart disease (CHD) risk factors have been performed in fasting subjects, aithough most time is spent in me postprandial state. Zilversmit [l] has suggested that postprandial lipoproteins are atherogenic and later angiographic studies have supported this idea [2,3]. Furthermore, it has been recognized that the coagulation activity is closely associated with the concentration of triglyceride-rich lipoproteins in the blood [4-61. The activity of Factor VII, which is a strong risk factor of myocardial infarction, follows serum triglyceride (Tg) concentration with a delay of about 2-4 h [7]. These findings together with epidemiological I observations ]8;9] have shown that dietary fat ;is associated with blood coagulability. Most studies performed in this field have been based on long-term dietary manipulations and usually focused on a single clotting factor. We conducted a study on short-term effects of dietary fat in the form of a single fatty meal on postprandial lipemia and hemostatic activity. Furthermore, we investigated if the effect of a meal containing saturated fatty acids (cream) is different from the effect of a meal containing n-6 fatty acids (sunflower oil). 2. Subjects and methods

103 (1993) I-II

Table 1 Laboratory parameters” Variableb Total cholesterol (mmol/l) LDL-cholesterol (mmolfl) VLDL-cholesterol (mmoliI) HDL-cholesterol (mmol/l) HDL+holesterol (mmol/l) HDL+holesterol (mmoliI) Total triglycerides (mmol/l) VLDL-triglycerides (mmol/I) LDL-triglycerides (mmol/l) HDL-triglycerides (mmol/l) Apolipoprotein B (g/I) Apolipoprotein Al (g/I) Lipoprotein(a) (mg/ml) Fibrinogen (g/l) F VII:Ag (%) F VILC (%) F VIII:C (%) vWF:Ag (%) t-PA:Ag (&ml) PAI-I (Au/ml) AT III BTG (ng/ml)c PF4 (@ml) Platelet Count (109/1)

5.1 f 1.2 3.1 * 1.0 0.35 * 0.15 1.24 i 0.25 0.96 zt 0.30 0.48 f 0.09 1.19 f 0.46 0.69 f 0.26 0.35 zt 0.18 0.07 f 0.02 0.85 i 0.35 1.46 f 0.20 0.26 f 0.38 2.59 f 0.29 85.3 zt 10.3 86.9 zt 9.9 103.9 * 30.5 89.5 f 29.2 6.60 f 3.02 10.86 f 6.75 100.7 f 5.4 23.2 f 5.9 4.7 f 1.5 249.6 zt 50.0

“laboratory parameters were estimated using the averages of 3 fasting measurements, n = 10. Values are mean f SD. bAbbreviations: LDL, low density lipoprotein; VLDL, very low density lipoprotein; HDL, high density lipoprotein; F VII:Ag, factor VII antigen; F VII:C, factor VII activity; F VIII:C, factor VIII activity; vWF:Ag, von Willebrand factor antigen; t-PA:Ag, tissue plasminogen activator antigen; PAII, plasminogen activator inhibitor I; AT III, Antithrombin III activity; BTG, beta-thromboglobulin; PF4, platelet factor 4. ‘a = 9 since values of I subject were excluded as artificially high due to ex-vivo platelet activation.

The study population consisted of 10 white middle-aged men from the research staff of the participating institutes. They all reported themselves as healthy non-smokers and did not use any S.D.) was medications. Their mean age ( ?? 40.8 f 8.4 years and body mass index (BMI) 23.3 f 2.7. Serum lipid levels and hemostatic parameters were within the normal range

not to engage in any strenuous physical activity.

(T2bie

The int~nral 1 II., 111bW1 . u. hetuwm “IC..WVII

i).

Each subject had 3 test meals in random order. The meals were given at 08:OOh, immediately after a 12-h fasting blood sample was drawn, and consumed within the next 10 min. Subsequently, blood samples were drawn every 2 h up to the 8-h sample at 16:00 h. A light fat-free lunch was given immediately after the 12:00 h blood sample was drawn. Between blood drawings the subjects were allowed to pursue their usual work activities but

th,= CI1Wt,=Qt CIYCmc=.calc LllVUlY \u.ar ..U” i> 1 1 WPPL ..1W1.)

and during that time the subjects followed their usual diets. To find out its composition each subject kept estimated food records for 1 week. The main composition of the usual diet calculated on the basis of these records is given in Table 2. Daily intakes of foods and nutrients were computed using the software system developed at the National Public Health Institute, Helsinki, Finland. The food composition data are a mixture of values ob-

V. Salomaa et al. /Atherosclerosis

Table 2 Habitual

daily dietary

103 (1993)

I-11

3

and special attention was paid to obtaining a good blood flow. The first milliliters of blood were discarded. Lipoprotein lipids were analyzed in the Department of Biochemistry, National Public TT__l,tiT-_&I&-.&_ rT_,_:._l.l__> ~_____~r~l_--Ix--^ nedlin InsLiLuLe, ne:IsinKi dnu nernos~d~ic: parame-

intake”

Variable Energy (MJ) Cholesterol (mg) Fat (% of energy) Saturated fatty acids (“/u of energy) Monounsaturated fatty acids (%I of energy) Polyunsaturated fatty acids (%I of energy) W-6 PUFA (% of energy) W-3 PUFA (Y/Uof energy) Carbohydrates (“/Dof energy) Proteins (% of energy)

mean f S.D.

f zt f zt

1.51 60.1 4.5 2.6

10.4

f

1.5

4.6 3.4 0.99 48.6 15.1 15.5 3729

Alcohol (g) Sodium (mg) “Values are (n = 10).

9.49 324.2 32.9 10.7

based

on

7-day

zt ??

f f f f f

1.3 1.2 0.21 5.5 1.6 11.2 740

food

records

tamed from Finnish food analyses and values taken from international food composition tables. The fat-free meal consisted of a banana (150 g), 200 g of fat-free milk and orange juice as needed to give an energy content of 16.7 kJ/kg ( = 4 kcal/kg) of body weight. The saturated fat meal was otherwise similar to the fat-free meal but cream was used instead of fat-free milk to give a fat load of 1 g/kg of body weight. The polyunsaturated fat meal was prepared from the fat-free meal by adding sunflower oil to the fat-free milk to form a milk shake, which gave 1 g of fat/kg of body weight. Egg yolk was added to the milk shake to give the same amount of cholesterol as contained by the cream. Lunch was the same as the fat free meal. The fiber content of the meals was negligible derived only from banana. Water, but Il,..,,A ,i....;,, no other extra food or &ii&, iWS &IUWGU UUIIII~ the 8 h study period. The test meals were well tolerated and no gastro-intestinal side-effects were reported. 2.1. Laboratory methods The blood samples were drawn from antecubital veins using repeated venipunctures with the subject in supine position. Vacutainer tubes, an 18-G needle, and a minimum amount of stasis was used

ters in the Coagulation Laboratory, Finnish Red Cross Blood Transfusion Service, Helsinki. Lipid and apolipoprotein analyses. Total cholesterol (Chol) was determined enzymatically [lo] using a commercial CHOD-PAP method (Boehringer-Mannheim, Germany). HDL-Chol was determined after @-lipoprotein precipitation with dextran sulfate/magnesium chloride [ Ill. Serum Tg concentration was determined enzymatically [12]. Quality control results have recently been published [13;14]. Apolipoproteins (apo) A-I and B were determined by immunoturbidometry [ 151. Ultracentrifugal separation of serum lipoproteins was performed in a type TFT 45.6 rotor (Kontron Instruments, Zurich, Switzerland) using the Beckman Model L preparative ultracentrifuge (Beckman Instruments, Inc. Palo Alto, CA, USA) according to Ref. 16. Very low density lipoprotein (VLDL), low density lipoprotein (LDL), and HDL2 were isolated at densities d = 1.006 g/ml, d = 1.063 g/ml, and d = 1.125 g/ml, respectively. The fraction with d > 1.125 g/ml was used to assess HDL,. Analyses of hemostatic factors. For coagulation assays 9 ml of venous blood was mixed with 1 ml of 0.129 M trisodium citrate. The samples were spun at room temperature for 30 min at 1400 x g. For assays of beta-thromboglobulin (BTG) and platelet factor 4 (PF4), polyvinyl tubes were filled to contain 6 ml when mixed with 0.2 ml of 10% EDTA, 0.2 ml of 5.4 mg/ml theophylline, and 0.2 ml of 1 pG/ml prostaglandin E,, and samples prepared as described previously [ 171. Samples 4-A.. I,,, +I..,.. 1 ,,_,... LA-,.., .rm..n ,+,,,A aL nt -,v7nof-L ,“I ws,c; 3L”IGU 1E;XY Llla‘l 1 yku “Cl”lC analyses. Antithrombin III (AT III) was measured as heparin cofactor activity (Coatest Antithrombin III, Kabi Diagnostica, Sweden), BTG by radioimmunoassay [18], fibrinogen with the ACL method based on the measurement of light scattered from the reaction mixture [19]. Factor VII procoagulant activity (F VII:C) was measured with the one stage method [20] by determining the ability of the test

4

V. Salomaa et al. /Atherosclerosis

sample to correct the clotting time of human F VII deficient plasma. The plasma levels were expressed as percent activity compared with a pool of 20 normal plasmas taken as 100%. Factor VII antigen (F VII:Ag) was measured with an ELISA technique (Asserachrom F VII:Ag, Diagnostica Stago, France), Factor VIII coagulation activity (F VIII:C) with an amidolytic method (Coatest F VIII, Kabi Diagnostica, Sweden), von Willebrand factor antigen (vWF:Ag) with rocket immunoelectrophoresis [21], platelet factor 4 (PF4) with radioimmunoassay (PF4 RIA Abbot, Diagnostics Division, Germany), tissue plasminogen activator antigen (t-PA:Ag) by enzyme linked immunosorbent assay (Imulyse-5 t-PA, Biopool Ab, Sweden), and fast-acting plasminogen activator inhibitor activity (PAI-1) with an amidolytic method (Coatest PAI, Kabi Diagnostica, Sweden). Each analysis was performed in duplicate and the mean is reported. 2.2. Statistical methods The effect of the different morning meals on the change in lipoprotein lipid or in hemostatic factor during the observation period was investigated using the following analysis of variance model: AYgk = ai + bj +

Ck +

djk

•I

e$,

where AYiik is the change in the lipoprotein lipid or hemostatic factor being considered, for subject i (i=l,... ,lO) between baseline and the jth ,4) after meal k measurement o’=l,... (k= 1,. . . ,3). a, b and c are the main effects of person, time and meal, respectively, d is the interaction between time and meal, allowing for reasonable time for developing a change after the meal, and e is the error term. The difference of the effects of the different meals was tested using the F-statistic [22] for the null-hypothesis H,: ck = djk = 0 for every meal and time. The test gives an overall comparison of the effects of the meals. Graphical presentation is used to show the nature of the possible effects of the meals. To check the consistency of the effect of the meal among the subjects the model with an additional interaction term for subject and meal was also considered.

IO3 (1993)

1-11

The associations of the fasting levels of hemostatic factors with fasting lipoprotein lipids, age, BMI and selected variables of background diet were examined using Spearman rank correlation coefficients. The averages of the 3 individual fasting measurements of the laboratory parameters were used for calculating the correlation coefficients. The associations of F VII:C concentration with total Tg and VLDL-Tg concentrations in different time points were investigated separately after each meal using Spearman rank correlation coefficients. Statistical computations were performed using Statistical Analysis System (SAS) [231.

There were 8 observations (5.3% of all blood drawings) where both BTG and PF4 were considerably elevated being clear outliers, and suggesting a problem with venipuncture and platelet activation ex vivo [24]. These values with BTG 1 78 and/or PF4 1 13 were excluded from the analyses of platelet activation peptides. Scrutinization of other hemostatic values from these blood drawings did not reveal any exceptions from the usual pattern for that particular person considering the time of day and the meal given. Furthermore, in 23 observations (15%) PAIvalues were below the lower detection limit of our laboratory, 1.3 Au/ml. This rather large amount probably reflects the selected nature of our study population with healthy life-styles. These observations occurred at 4-, 6- and 8-h time points and 5 of them occurred after the saturated fat, 10 after polyunsaturated fat and 8 after the fat-free meal. In these cases the value of PAI- was arbitrarily set as 1, which was chosen as a biologically plausible alternative. 3. Results In the fasting state, fibrinogen was positively associated with BMI and negatively with HDL and HDL2-Chol (Table 3). F VII:C was positively associated with total and LDL-Chol, apo B and LDL-Tg but its association with total fasting Tg (r = 0.61, P = 0.06) or VLDL-Tg (r = 0.54, P = 0.11) did not reach statistical significance in this small sample. Several other hemostatic parameters were significantly associated with lipoprotein lipids. Most notable were the strong

V. Salomaa et al. /Atherosclerosis

Table 3 Spearman

rank correlations”

Age Body mass index Cholesterol LDL-cholesterol VLDL-cholester01 HDL-cholester01 HDLz-cholester01 HDL,-cholester01 Total triglycerides VLDL-triglycerides LDL-triglycerides HDL-triglycerides Apohpoprotein B Apolipoprotein A-I Lipoprotein(a)

103 (1993)

Fibrinogen

F VII:Ag

F VII:C

F VIII:C

0.40 ^ ^_.

-0.05 ^ ^_ -0.38

0.51 ^ ^^ -U.UY

0.19 ^ .^ -U.IX

0.17 0.24 0.39

0.27 0.34 0.20

U.XX*'

5

of age, body mass index and serum lipids with hemostatic

0.76* 0.83** 0.59

variables

vWF:Ag

t-PA:Ag

PAI-I

AT III

BTGb

0.30 ^ ^^_

0.68*

0.48

0.42

0.47

0.26 0.31

-0.25

-0.08 -0.13 -0.15

0.10 0.08 0.13

0.83** 0.77** 0.96**

0.81** 0.75: 0.93**

0.53 0.41

-0.UU6

0.44

0.22

PF4

Platelet count

0.47 0.12

-0.16 0.54

0.47 0.40 0.47

0.79** 0.70: 0.89**

-0.30 -0.45 -0.05

0.67*

0.09

-0.006

-0.64*

-0.59

-0.61

0.19

-0.59

-0.52

-0.46

-0.03

-0.56

0.05

-0.22

-0.45

-0.14

0.03

-0.36

0.61

-0.01

0.14

0.61

-0.22

0.07

0.95**

0.92**

0.52

0.33

0.85**

-0.07

0.16

0.54

-0.26

0.03

0.95**

0.87**

0.46

0.30

0.83**

0.07

0.24

0.43

0.71*

-0.07

0.27

0.62

0.71*

0.93;;

-0.35

0.45

-0.52

0.31

-0.15

0.13

0.30

0.60

0.50

0.23

-0.02

0.32

-0.24

0.29

0.27

-0.10

0.09

0.81**

0.79**

0.51

0.40

0.71*

-0.37

-0.32

-0.08

0.01

0.19

0.07

-0.42

-0.41

-0.15

-0.22

-0.47

-0.47

-0.19

0.36

0.38

0.37

0.43

0.15

0.27

0.26

0.37

0.27

-0.01

-0.71*

-0.28

-0.51

0.19

-0.03

-0.67*

-0.25

-0.38

0.38

-0.16

-0.19

-0.14

0.46

0.24

0.52

0.27 0.78**

“The correlation coefficients were calculated b,? z 9 , “_.. EPP fnntnnte nf ..,U..I”.I “. Tzahle . &&“._ I. *P < 0.05. **p < 0.01. Abbreviations:

I-11

See footnote

of Table

using the average

of 3 fasting

measurements

(n = IO).

I.

associations of fibrinolysis parameters t-PA:Ag and PAIwith total Tg, VLDL-Tg and also VLDL-Chol and apo B. Dietary constituents calculated from the 7-day food records were only weakly and nonsignificantly associated with fasting levels of hemostatic factors. Alcohol intake showed; however, quite strong positive association with tPA:Ag (1. = 0.77, P c 0.01) and PAI- (r = 0.77, P < 0.01) concentrations. Also, the proportion of energy from n-6 fatty acids was inversely associated with fasting BTG concentration (Y = -0.83, P < 0.01) but not with fasting PF4 concentration (r = -0.38, P = 0.28). During lipemia, the associations were approx-

imately similar or slightly weaker than in the fasting state. The correlation coefficients did not show any consistent relation between total or VLDL-Tg concentrations and F VILC concentrations taken 2 or 4 h later. The fatty meals caused an increase in VLDL-Tg, which was at its maximum at 4 h after the mea! and almost returned to fasting level at 8 h (Fig. 1). The curves for total Tg were similar to those for VLDL-Tg. VLDL-Chol also increased after the fatty meals, remaining still elevated 6 h after the meals but decreasing thereafter rapidly back to the fasting level. HDL-Tg remained at the fasting level up to 2 h after the fatty meals but increased rapidly thereafter reaching its maximum at 4 h. LDL-Tg

V. Salomaa et al. /Atherosclerosis

6

VLDL-TG

mmoli

103 (1993)

I-11

VLDL-CHOL mmol/l

1.2 c

0.25

T

0.20 0,15 -

.0,05 0

-0,lO I-J 0

2

I 4

6

a

6

8

Hours

HDL-TG

mmoll

HDL,-CHOL

mmoll 0,04 0,02 -

_.,_.._.~_...._.._._.._..

0.02 -

...0,oo -A&

0.01 -

T

--__

.0,02 -0,04 -0,06 -

0

2

4

-0,oa

6

' 0

2

Hours

Cream

Sunflower ____

oil

Fat-free .....__

4 Hours

Fig. 1. Changes in VLDL-Tg, VLDL-Chol, HDL-Tg and HDL&hol after saturated fat meal (cream), n-6 polyunsaturated fat meal (sunflower oil) and fat-free meal. n = 10. Bars depict SE. The changes in each parameter are significantly different (P < 0.0001 for the overall difference between the meals).

did not change significantly after any of the meals. Total HDL-Chol decreased slightly and insignificantly after the fatty meals. This was due to the HDLs subfraction, whereas the HDL2 subfraction did not change. Apolipoproteins A-I and ,B were measured during fasting and at 4 and 8 h after the meals. They showed no postprandial changes. The changes in lipoprotein lipids were similar after the 2 fat loads, whereas the fat-free meal gave no postprandial response. F VII:Ag decreased during 4 h following the meals and this change was similar after all 3 test meals (Fig. 2). F VII:C remained unchanged during the first 4 h but increased 6 and 8 h after the fatty meals, whereas after the fat-free meal a decrease was observed. The effect of the meal on the change in F VII:C was statistically significant (Fcs,99, = 2.99, P = 0.0048). As shown in Fig. 2,

the main difference was between the fat-free and fatty meals, whereas the changes after the fatty meals were similar to each other. The mean values (*SD.) 8 h after the test meals were 91.7% f 13.1%, 90.4% f 12.2% and 81.3% f 10.5% after cream, sunflower oil and fat-free meal, respectively. T-PA:Ag decreased almost linearly during the 8 h following the meals and there was no difference between meals (Fig. 2). The decrease of PAIactivity tended to be less steep after fatty meals than after the fat-free meal but the effect of the meal was not statistically significant (Fc8,99) = 0.47, P = 0.87). The mean PAI-1 (*S.D.) at 8-h time point was 5.64 f 4.00 Au/ml after cream, 4.03 f 3.60 Au/ml after sunflower oil and 3.92 f 2.38 Au/ml after the fat-free meal. F VIII:C was highest after the polyunsaturated

V. Salomaa et al. /Atherosclerosis

103 (1993)

7

I-II

F VII:Ag

F VII:c

% 2

% st

I-T

T

-6 -0

1,

1

0

2

4

6

8

Hours

t-PA:Ag

PAI-

nglml

Au/ml

“h -0,5

-

.I

-

-1,5

-

._.._ _.......

-6 -

. . . .. .

-7 -2 -

-8 ,J_

-2,5

’ 0

2

4

6

HOlIE

0

0 _

-.,

Ueamsuntlower .__.

2 -

4

6

8

HOtUS

011t-at-tree _.....

Fig. 2. Changes in F VII:Ag, F WC, t-PA:Ag and PAI-I activity after saturated fat meal (cream), n-6 polyunsaturated fat meal (sunflower oil) and fat-free meal. n = 10. Bars depict S.E. y-axes of F VIi:C and F Vii:Ag give absolute diflerences in percentage activity and percentage antigen concentration of F VII. The changes in F VII:C are significantly different (P = 0.0048 for the overall difference between the meals).

fat meal and lowest after the fat-free meal (Table 4). The effect of the meal on the changes from the fasting level was of borderline statistical significance (F(s,%)= 2.01, P = 0.053). At 6 h after the polyunsaturated fat meal the mean BTG concentration tended to be higher than after the fat-free meal (Table 4). The overall

The main finding of our study was the increase of F VII:C after the fatty meals and the decrease

c-honnm ~~1cau~vo

nfit aftnrttrn f"t_f*,PmPol l-ha ,i:+Ta..a..n- hot..7.-.h.. "1 1L UIIbI Lllcl&al-‘1L.d 111LQ1. 1L‘G uIII=IGiILG V~LW~~,,

Aw;nn UU11116

thp Cllr

nhom-vat;nn ““.JCI “UCIVII

net;nA pbllvu

,I,P..P ““tiI”

nnt U”,)

however, significantly different between the meals (+?30) = 1.11, P = 0.37). No changes were observed in PF4 (Table 4) or in platelet count. Furthermore, no postprandial changes were observed in AT III, vWF:Ag or fibrinogen. The effect of the meal type on the postprandial changes was significantly different in different individuals as demonstrated for VLDL-Tg, F VII:C and PAI- in Table 5. In fact, the meal by person

interaction-term was statistically significant for each lipoprotein lipid studied as well as for each hemostatic factor except t-PA:Ag. 4. Discussion

the saturated fat meal and the fat-free meal at 8 h was 10.4 percentage points (12.0% of the mean fasting value), which is large enough to have biological significance. In the Northwick Park Heart Study [25] F VII:C was an independent risk factor of CHD incidence and its effect was stronger than the effect of cholesterol. Accordingly, the 12.0% difference in F VII:C encountered in this study may influence the risk of CHD events.

8 Table 4 F VIII:C, meala

V. Salomaa ef al. /Atherosclerosis

beta-thromboglobulin

104.4 105.8 106.6 103.0 103.9

f f f f f

Beta-thromboglobulin 0 23.8 f 2 22.5 f 4 24.5 * 6 25.1 f 8 21.6 zt Platelet factor 4 0 4.4 2 4.1 4 4.4 6 4.7 8 4.1

factor

4 by test

Polyunsaturated fat

Fat-free

35.3 32.1 32.1 33.4 35.3

103.7 109.2 109.2 106.2 106.8

f f zt f f

26.6 28.0 30.1 28.0 28.8

103.5 106.9 103.1 102.6 100.9

9.0 7.2 6.1 8.1 10.2

22.0 24.1 25.4 26.3 24.4

zt f f f f

9.2 9.9 8.4 9.7 9.5

23.3 19.8 19.0 18.2 20.3

f f * f f

6.5 4.9 6.0 3.3 5.2

2.4 2.2 2.1 1.8 2.6

4.9 4.9 5.1 5.1 3.8

f f f f f

3.1 2.2 2.1 1.7 1.9

4.6 3.6 4.2 4.1 3.9

f zt zt f f

1.9 1.0 1.4 1.8 2.1

Time (h) Saturated fat F VIII:C 0 2 4 6 8

and platelet

f * f zt +

zt 32.1 zt 31.3 ?? 31.9 zt 30.4 f 31.5

“Values are mean f SD., n = IO for F VIII:C for betathromboglobulin and platelet factor 4 the outliers are excluded (see Statistical methods, Section 2.2. in the text).

Our finding is in good agreement with studies by Miller and colleagues [7,26,27], and further shows that the effect of fat on F VII:C can be observed even after a single fat-containing meal. Moreover, the decline in F VII:C after the fat-free meal

Person Time Meal Mea!*Time Meal*Person Abbreviations:

Degrees freedom

of

9 3 2 6 18 see footnote

VLDL-tg

l-11

accounted for half of the difference between the fat loads and the fat-free meal suggesting that the values measured after a 12-h fast were still affected by fat ingested on the previous day. In the present study, the time difference between the maximal Tg response and the F VII:C response was longer than observed by Miller et al. [7]. We were also unable to confirm a consistent relation between VLDL-Tg or total Tg concentration and F VII:C 2 h later. These discrepancies are probably due to differences in the design of the experiments or in the timing of the sampling. The maximal values can easily be missed between the blood drawings in either of the trials. We found no difference in F VII:Ag concentration after the fat loads as compared with the fat-free meal, which supports the idea that the changes in F VII:C are due to an acute effect of Tg-rich lipoprotein particles on F VII reactivity [7]. It has previously been reported that PAI- has a strong diurnal variation (281 and that it is closely associated with serum Tg concentration [29]. It can also increase rapidly by release from endothelial cells or from platelet alpha granules [30,31]. In keeping with this theoretical framework, a somewhat steeper decline of PAI- was observed after the fat-free meal than after the fatty meals. This effect was consistent at 6 and 8 h after the meal, although statistically non-significant. The mean differences observed between the saturated fat meal and the fat-free meal were, however, as large as over one-fourth of the baseline level of PAI- (1.87 i 0.89 Au/ml and 1.72 f 0.99 Au/ml

Table 5 Analysis of variance of the effects of meal type, person, time and the interactions of VLDL-triglycerides, FVII:C and PAI-I during the follow-up of 8 h Model

103 (1993)

of meal type with time and person

on the changes

PAI-I

F VII:C

F

P

F

P

F

P

5.04 11.91 46.08 5.34 2.31

0.0001 0.0001 0.0001 0.0001 0.005

11.47 1.30 8.42 3.96 4.83

0.0001 0.28 0.0005 0.0016 0.0001

11.62 20.78 3.02 0.84 11.55

0.0001 0.0001 0.054 0.54 0.0001

of Table

I.

V. Salomaa et al. /Atherosclerosis

103 (1993)

I-11

at 6 and 8 h, respectively). Furthermore, the effect of the meal on the change in PAI-I activity was strongly dependent on the person, suggesting that in some individuals the amount and type of dietary fat may be significant determinants of fibrinoiytic activity. The assessment of the differences in PAIis obscured by the fact that in 23 observations the PAI- activity was below the detection limit (1.3 Au/ml) of our laboratory. These values were arbitrarily marked at 1.0 Au/ml, which is biologically plausible, but may lead to an underestimation of the true difference between the saturated fat meal and the other meals. Whether the observation that BTG was elevated after the meal-_ is ___- nolvmsaturated r__ ,___ - _______ - fat _- _____ __ nhvsinln&alT__,-_-__D__-_ ly meaningful remains unclear. It is biologically plausible that the amount and type of dietary fat can cause changes in platelet activity [lo]. On the other hand, the assessment of platelet activity in vivo on the basis of BTG and PF4 values is technically difficult because of the problem of platelet activation ex vivo. There are no universally accepted criteria for BTG and PF4 or their ratio to distinguish in vitro from in vivo activation [32-341. We chose to exclude blood drawings where both BTG and PF4 were elevated [24], which coincided well with the clinical information on minor difficulties with venipuncture. These exclusions were made before any statistical testing. In contrast, the proportion of energy from n-6 polyunsaturated fat in the usual diet as assessed from the 7-day food records was inversely associated with mean fasting BTG concentration. Obviously, more research using larger materials and more versatile measurements of platelet function are needed for proper assessment of short-term and long-term effects of dietary fat on platelet q&;.,;t., LIbL‘“ILJ. Zucker et al. [35] have described a decrease in the functional activity of AT III after a saturated fat meal. Meals rich in unsaturated fatty acids or carbohydrate alone did not affect functional AT III activity significantly. A possible reason for the discrepancy between the present data and their observations is that different people received different fat loads in the South African trial and the differences in changes may thus be due to factors

9

other than dietary fat. Also, longer term dietary experiments have given conflicting results on the effect of dietary fat on AT III activity [36]. One explanation for the controversies may be that the ,t.... F.. repearammy or me iaboratory anaiysis of AT iii is not always very good [37]. Since each parameter was analyzed in duplicate from 3 fasting blood samples, we were able to use the averages of 6 fasting measurements in the calculations of correlation coefficients between the lipoprotein lipids and hemostatic parameters. The coefficients are thus fairly free of intra-individual and laboratory variability, which in part explains their strength. In particular, the associations between triglyceride and tPA:Ag or PAIwere strongj clearly stronger than in previous studies [5,38,39]. The associations of F VII:C with serum cholesterol and the positive association of fibrinogen with BMI have been described previously [40], but the inverse association of fibrinogen with HDL or HDL*-Chol has not been consistently found. A recent report from the Atherosclerosis Risk In Communities (ARIC) Study [41] described the inverse association with total HDL-Chol but considered it as a residual effect of smoking. Our subjects were all nonsmokers suggesting that other mechanisms must be involved. A limitation of our study is that the fat load was fairly large corresponding approximately to the amount of fat these health-conscious men used to eat during the whole day on their usual diets. Hence, direct conclusions to the real life situation are not possible. Also, our experimental setting does not exclude the possibility that the different energy content of the meals would contribute to the changes observed. However, data from the literature suggest that a more plausible explanation ..,,...lrl -r\..+-..+ “1 ,.c +l., ?L 771 W”UlU r.0 “G tC, LIIC;r,+ 1ar c”UIeUI LIIC;,,..I” lll&;a,> r7 L,,L”,i,,. Another word of caution is needed regarding the number of statistical tests performed. When performing several statistical tests for the various lipids and clotting factors, the possibility of a low P-value by chance increases. Our finding on F VII:C is, however, strong, based on the predefined hypothesis, and consistent with the literature. On the other hand, borderline findings -- e.g.. on F VIII:C - need to be confirmed in future studies.

K Salomaa et al. /Atherosclerosis

10

The generalization of our results is problematic because our subjects were a selected group of research staff. Even in this relatively homogenous group the postprandial changes in lipoprotein lip:,1, ClllU “..A L--,.Y*-Lr,.,*,.., ._.^_^ “:--:rZ-,.-rl.. ueA1UJ UcIU”sLacIGlilGL”IS were slglllllcallrly pendent on person. The small size of our study precludes more detailed analysis of individual characteristics affecting the postprandial changes. It can be argued, however, that more obese and sedentary persons are likely to have more pronounced postprandial lipemia and procoagulatory changes. Postprandial lipemia and changes in hemostasis parameters were similar after the n-6 polyunsaturated fat meal and after the saturated fat meal. This finding is potentially important if the postprandial lipemia plays a major role in atherogenesis as has been suggested [ 11. It may be typical for short-term fat use only, however, since it has been described that long-term diets rich in n6 polyunsaturated fat [42,43] or fish oil [43,44] reduce the magnitude of postprandial lipemia compared with diet rich in saturated fat. In conclusion, our study showed that the postprandial lipemia following a single fatty meal is associated with a procoagulatory change in F VII:C. This change was significantly different in ..^^ ditterent individuais but there was no difference between saturated and n-6 polyunsaturated fat.

7

8 9

10

11

12

13

14

15

16

5. References 17 Zilversmit, D.B., Atherogenesis: a postprandial phenomenon, Circulation, 60 (1979) 473. Blankenhorn, D.H., Alaupovic, P., Wickham, E., Chin, H.P. and Azen S.P., Prediction of angiographic change in native human coronary arteries and aortocoronary bypass grafts, Circulation, 81 (1990) 470. Groot, P.H.E., van Stiphout, A.H.J., Krauss. X.H., Jansen, ii., van Toi, A., van Ramshorst, E., Cinin-On, A., Hofman, S.R., Cresswell, S.R. and Havekes, L., Postprandial lipoprotein metabolism in normolipidemic men with and without coronary artery disease, Arterioscl. Thromb., 11 (1991) 653. Simpson, H.C.R., Meade, T.W., Stirling, Y., Mann, J.I., Chakrabarti, R. and Woolf, L., Hypertriglyceridemia and hypercoagulabihty, Lancet, (i) (1983) 786. Crutchley, D.J., McPhee, G.V., Terris, M.F. and Canossa Terris, M.A., Levels of three hemostatic factors in relation to serum lipids, Arteriosclerosis, 9 (1989) 934. Mitropoulos, K.A., Hypercoagulability and factor VII in

18

19

20 21

22 23

103 (1993)

I-11

hypertriglyceridemia, Semin. Thromb. Hemostasis, 14 (1988) 246. Miller, G.J., Martin, J.C., Mitropoulos, K.A., Reeves, B.E A., Thompson, R.L., Meade, T.W., Cooper, J.A. and Cruickshank, J.K., Plasma F VII is activated by postprandial trigiyceridemia, irrespective of dietary fat composition, Atherosclerosis, 86 (1991) 163. Strom, A. and Jensen, R.A., Mortality from circulatory diseases in Norway 1940-1950, Lancet, (i) (1952) 126. Nordoy, A. and Goodnight, S.H., Dietary lipids and thrombosis. Relationships to atherosclerosis, Arteriosclerosis, 10 (1990) 149. Roschlau, P., Bernt, E. and Gruber, W., Enzymatische Bestimmung des Gesamtcholesterins in Serum, Z. Khn. Biochem., 12 (1974) 226. Kostner, G.M., Enzymatic determination of cholesterol in high density lipoprotein fraction prepared by polyanion precipitation, Chn. Chem., 22 (1976) 695. Wahlefeld, A.W., Triglycerides. Determination after enzymatic hydrolysis. In: Bergmeyer, H.V. (Ed.), Methods of Enzymatic Analysis, Academic Press, New York, 1974, p. 18-31. Salomaa, V.V., Jauhiainen, M., Pietinen, P., Korhonen, H.J., Kartovaara, L., Vartiainen, E. and Tuomilehto, J., Five-year trend in serum HDL-lipoprotein cholesterol in the Finnish population aged 25-64 years. A suggestion of an increase, Atherosclerosis, 31 (1991) 39. Valsta, L.M., Jauhiainen, M., Aro, A., Katan, M.B. and Mutanen, M., Effects of a monounsaturated rapeseed oil and a polyunsaturated sunflower oil diet on lipoprotein levels in humans, Arterioscl. Thromb., 12 (1992) 50. Riepponen, P., Marniemi, J. and Rautaoja, T., Determinntinn nf ~nnlinnnmt~inr A.1 r~r,,m Prslnrl ...lll...l”Ll “. ..~Yz.y”yL”.‘ L..” ‘ . . .anti . . . -R .in . . “IL”.ll) Y1....u. J. Clin. Lab. Invest., 47 (1987) 739. Havel, R.J., Eder, H.A. and Bragdan, J.H., The determination and chemical composition of ultracentrifugally separated lipoproteins in human serum, J. Clin. Invest., 34 (1955) 1345. Rasi, V. Beta-thromboglobulin in plasma: false high values caused by platelet enrichment of the top layer of plasma during centrifugation, Thromb. Res., 15 (1979) 543. Kuronen, B., Kauppinen, H-L., Rasi, V., Kekomaki, R. and Myllyhi, G., A coate-tube radioimmunoassay for beta thromboglobuhn, Med. Biol., 58 (1980) 264. Rossi, E., Mondonico, P., Lombardi, A. and Preda, L., Method for the determination of funciionai (ciottabie) fibrinogen by the new family of ACL coagulometers, Thromb. Res., 52 (1988) 453. Denson, K.W., The specific assay of Prower-Stuart factor Factor VII, Acta Haematol., 25 (1961) 105. Laurell, C-B., Quantitative estimation of proteins by electrophoresis in agarose gel containing antibodies, Anal. Biochem., 15 (1966) 45. Draper, N. and Smith, H., Applied regression analysis. 2nd Edn. Wiley, New York, 1989, p. 102. SAS Institute Inc., Users Guide: Statistics. Version 6, Fourth Edition, Volume 2, Cary, 1989. 846 pp.

V. Salomaa et al. /Atherosclerosis 24

25

26

21

28

29

30 31

32

33

34

103 (1993)

l-11

Kaplan, K.L. and Owen, J., Plasma levels of beta tbromboglobulin and platelet factor 4 as indices of platelet activation in vivo, Blood, 57 (1981) 199. Meade, T.W., Mellows, S., Brozovic, M., Miller, G.J., Chakrabarti, R.R., North, W.R.S., Haines, A.P., Stirling, Y., Imeson, J.D. and Thompson, S.G., Hemostatic function and ischaemic heart disease: principal results of the Northwick Park Heart Study, Lancet, (ii) (1986) 533. Miller, G.J., Martin, J.C., Webster, J., Wilkes, H., Miller, N.E., Wilkinson, W.H. and Meade, T.W., Association between dietary fat intake and plasma factor VII coagulant activity a predictor of cardiovascular mortality, Atherosclerosis, 60 (1986) 269. Miller, G.J., Cruickshank, J.K., Ellis, L.J., Thompson, R.L., Wilkes, H.J., Stirling, Y., Mitropoulos, K.A., Allison, J.V., Fox, T.E. and Walker, E.D., Fat consumption and factor VII coagulant activity in middle-aged men. An association between a dietary and thrombogenic coronary risk factor, Atherosclerosis, 78 (1989) 19. Andreotti, F., Davies, G.J., Hackett, D.R., Khan, M.I., De Bart, A.C.W., Aber, V.R., Maseri, A. and Kluft, C., Major circadian fluctuations in fibrinolytic factors and possible relevance to time of onset of myocardial infarction, sudden cardiac death and stroke, Am. J. Cardiol., 62 (1988) 635. Juhan-Vague, I., Vague, P., Alessi, M.C., Badier, C., Valadier, J., Aillaud, M.F. and Atlan, C., Relationships between plasma insulin, triglyceride, body mass index, and plasminogen activator inhibitor 1, Diabete. Metab., 13 (1987) 331. Sprengers, E.D. and Kluft, C., Plasminogen activator inhibitors, Blood, 69 (1987) 381. Kruithof, E.K.O., Plasminogen activator inhibitor type 1: biochemical, biological and clinical aspects, Fibrinolysis, 2 (Suppl. 2) (1988) 59. Arocho-Pinango, C.L. and Ojeda, A., Does the BTG/PM ratio have any value? (Letter), Thromb. Haemostas., 53 (1985) 288. Fabris, F., Lazzatto, G. and Girolami,. A., The value of BTG/PF4 ratio in patients with high platelet count (Letter), Thromb. Haemostas., 53 (1985) 288. Viero, P., The value of BTGIPF4 ratio in patients with high platelet count - reply, Thromb. Haemostas., 53 (1985) 289.

II

35

Zucker,

36

Feesey, M., Kuschke, R. and Seftel, H., Antithrombin functional activity after saturated and unsaturated fatty meals and fasting in normal subjects and some disease states, Thromb. Res., 15 (1979) 37. Schmidt, E.B., Kristensen, S.D., Sorensen, P.J. and Dyerberg, J., Antithrombin III and protein C in stable angina pectoris - influence of dietary supplementation with polyunsaturated fatty acids, Stand. J. Clin. Lab. Invest., 48 (1988) 469.

37

38

39

40

41

42

43

44

M.L.,

Comperts,

E.D.,

Russel,

D., Joffe,

B.,

Chambless, L.E., McMahon, R., Wu, K., Folsom, A., Finch, A. and Sben, Y.-L., Short-term intraindividual variability in hemostasis factors: the ARIC Study. Ann. Epidemiol., 2 (1992) 723. Hamsten, A., Wiman, B., De Faire, U. and Blomback, M., Increased plasma levels of a rapid inhibitor of tissue plasminogen activator in young survivors of myocardial infarction, N. Engl. J. Med., 313 (1985) 1557. Mehrabiam, M., Peter, J.B., Bardard. R.J. and Lusis, A.J., Dietary regulation of fibrinolytic factors, Atherosclerosis, 84 (1990) 25. Meade, T.W., North, W.R.S., Chakrabarti, R., Stirling, Y., Haines, A.P. and Thompson, S.G., Haemostatic function and cardiovascular death: early results of a prospective study, Lancet, (i) (1980) 1050. Folsom, A.R., Wu, K.K., Davis, C.E., Conlan, M.G.. Sorlie, P.D. and Szklo, M., Population correlates of plasma fibrinogen and factor VII, putative cardiovascular risk factors, Atherosclerosis, 91 (1991) 191. Demacker, P.N.M., Rejnen, I.G.M., Katan, M.B., Stuyt. P.M.J. and Stalenhoef, A.F.H., Increased removal of remnants of triglyceride-rich lipoproteins on a diet rich in polyunsaturated fatty acids, Eur. J. Clin. Invest., 21 (1991) 197. Weintraub, M.S., Zechner, R., Brown, A., Eisenberg, S. and Breslow, J.L., Dietary polyunsaturated fats of the n-6 and n-3 series reduce postprandial lipoprotein levels, Chronic and acute effects of fat saturation on postprandial lipoprotein metabolism, J. Chn. Invest., 82 (1988) 1884. Brown, A.J. and Roberts, D.C.K., Moderate lish oil intake improves lipemic response to a standard fat meal. A study in 25 healthy men, Arterioscl. Thromb., I1 (1991) 457.