sign in sign up
<<
Home , Glyceride, Myristic acid, Saturated fat, Stearic acid, Stearin

European Journal of Clinical Nutrition (2001) 55, 88±96 ß 2001 Nature Publishing Group All rights reserved 0954±3007/01 $15.00 www.nature.com/ejcn

A -rich diet improves thrombogenic and atherogenic risk factor pro®les in healthy males

FD Kelly1, AJ Sinclair1*, NJ Mann1, AH Turner2, L Abedin1 and D Li1

1Department of Science, RMIT University, Melbourne, Australia; and 2Medical Laboratory Science, RMIT University, Melbourne, Australia

Objective: To determine whether healthy males who consumed increased amounts of dietary stearic acid compared with increased dietary exhibited any changes in their platelet aggregability, platelet pro®les, platelet morphology, or haemostatic factors. Design: A randomized cross-over dietary intervention. Subjects and interventions: Thirteen free-living healthy males consumed two experimental diets for 4 weeks with a 7 week washout between the two dietary periods. The diets consisted of  30% of energy as (66% of which was the treatment fat) providing  6.6% of energy as stearic acid (diet S) or  7.8% of energy as palmitic acid (diet P). On days 0 and 28 of each dietary period, blood samples were collected and anthropometric and physiological measurements were recorded. Results: Stearic acid was increased signi®cantly in platelet on diet S (by 22%), while on diet P palmitic acid levels in platelet phospholipids also increased signi®cantly (8%). Mean platelet volume, coagulation factor FVII activity and plasma concentrations were signi®cantly decreased on diet S, while platelet aggregation was signi®cantly increased on diet P. Conclusion: Results from this study indicate that stearic acid (19 g=day) in the diet has bene®cial effects on thrombogenic and atherogenic risk factors in males. The food industry might wish to consider the enrichment of with stearic acid in place of palmitic acid and trans fatty acids. Sponsorship: Grant from Research Corporation, Australia and donated by Meadow Lea Foods Ltd, Australia. Descriptors: stearic acid; palmitic acid; platelets; fatty acids; haemostatic factors; lipoproteins; platelet aggrega- tion; mean platelet volume European Journal of Clinical Nutrition (2001) 55, 88±96

Introduction (Hegsted et al, 1965; Keys et al, 1965). This was attributed to the stearic acid in the cocoa . Since then, various The relationship between dietary fatty acids and blood predictive equations incorporating individual saturated fatty levels was reported in the late 1950s when acid contributions to plasma cholesterol changes show Keys et al (1957) found that saturated fatty acids (SFA) stearic acid as neutral (Kris-Etherton & Mustad, 1994). from 12 to 18 carbon atoms were potent cholesterol-raising The risk of coronary events however is not solely agents. Subsequently, it was found that the predictive dependent on plasma cholesterol levels. Carefully con- equations for dietary fatty acids on plasma cholesterol ducted autopsy studies have demonstrated that thrombotic levels were inadequate when was used obstruction of the coronary artery is involved in most cases of sudden cardiac death (Davies & Thomas, 1984). Arterial *Correspondence: AJ Sinclair, Department of Food Science, RMIT thrombosis involves clot formation based around existing University, GPO Box 2476V, Melbourne, Victoria 3001, Australia. plaque and arterial lesions with the ®rst step involving E-mail: [email protected] Guarantor: AJ Sinclair. platelet aggregation, followed by ®brin formation through Contributors: FDK, AJS, NJM and AHT initiated the study. FDK prepared activation of the coagulation pathways. Any factor leading the drafts of the paper, did dietary analysis, collected data, helped in to increased platelet activity or up-regulation of the coagu- laboratory assays and did the statistical analysis. AJS selected the study site, lation cascade will thus increase the risk of thrombotic supervised the project and secured the funding. LA and DL helped in events. laboratory assays. All authors contributed to the drafts of the paper. Received 13 June 2000; revised 6 October 2000; Early studies suggested that stearic acid was prothrom- accepted 9 October 2000 botic based on three levels of evidence. Firstly, Connor Stearic acid-rich diet FD Kelly et al 89 (1962) reported that stearic acid was particularly effective at Subjects were counselled prior to commencement of the thrombus formation time during in vitro studies ®rst intervention period on which foods were to be excluded using citrated blood and of fatty acids. Sec- from the diet (included in a detailed information booklet). ondly, Renaud and Gautherton (1975) reported that diets Advice was provided on suitable low-fat alternatives to rich in stearic acid were associated with increased coagula- substitute for foods normally eaten, such as regular fat tion in New Zealand rabbits. This change was speculated to dairy products, pastries and cakes etc. Subjects were given result from an enhanced activity of platelet factor 3 (PF3) training in recording their food intakes and completed a 7 due to an increase in stearic acid in the platelet phospholi- day weighed food record of their habitual diet prior to the pids. Thirdly, Renaud et al (1978) found that study. Seven day weighed food records were also obtained and especially stearic acid were signi®cantly correlated with during each intervention period and washout phase. Dietary clotting activity and platelet aggregation in several popula- composition was determined using Diet 1 (Version 4, tion comparisons in man. These data alerted researchers to NUTTAB 95, Xyris Software Pty Ltd, Queensland, Austra- potentially negative effects of stearic acid on thrombotic lia) based on Composition of Foods, Australia (National factors and resulted in several short-term studies being Food Authority, 1995). The database was modi®ed by the conducted in humans (Tholstrup et al, 1994, 1996; Schoene inclusion of fatty acid data, for a wider variety of foods, for et al, 1992; Mutanen & Aro, 1997; Dougherty et al, 1995). C12:0 to C18:0 (lauric, myristic, palmitic and stearic acids), However, few of these have examined more than one (C16:ln-7) and (C18:ln-9), lino- measure of thrombosis tendency. Because there are often leic acid (18:2n-6) and alpha linolenic acid (18:3n-3) interactions involving more than one component of the obtained from Dr M James, University of Adelaide. haemostatic system, there is no single screening test During the two intervention stages, the percentage which can be used to identify a thrombotic tendency. energy derived from fat, carbohydrate and was Therefore a range of screening tests should be adopted to kept constant for each individual, based on habitual food detect markers known to be associated with cardiovascular records. The high stearic and high palmitic acid test risk (Giddings & Yamamoto, 1995). were supplied by Meadow Lea Foods (Sydney, Australia). The aim of this study was to compare the thrombosis Test fats were supplied to subjects in the form of and potential of diets rich in stearic acid with palmitic acid-rich spreading margarines and incorporated into biscuits, cakes diets by measuring platelet aggregation, platelet volume and and muf®ns given as snacks and in salad rolls supplied for other key components of the haemostatic pathways. weekday lunches. Breakfasts, dinners and weekend lunches were left to the subjects' discretion following dietary modi®cation advice. The remaining fat intake was obtained Methods from commercially available foods low in fat.

Subjects Test fats Thirteen free-living healthy males were recruited following The stearic acid-enriched baking was produced advertisement of the study throughout RMIT University. by an interesteri®ed blend of 35% hardened canola (100% Subjects were aged 35 Æ 12 y (mean Æ s.d.), of normal body hydrogenated) and 65% Sunola1 (a high oleic acid variety weight (body mass index (BMI) ˆ 26.0 Æ 3.3 kg=m2. Sub- of sun¯ower seed ), and the stearic acid-enriched spread- jects had no known metabolic, endocrine or haematological ing margarine was a blend of 30% hardened canola and 70% diseases, were non-smokers, not on any form of medication Sunola1. The palmitic acid-enriched baking margarine was and had a moderate activity level. Subjects were advised not produced by interesterifying a blend of 55% palm , to take any form of non-steroidal anti-in¯ammatory drugs 20% palm olein and 25% Sunola1. The palmitic acid- 14 days prior to commencement and during the study. enriched spreading margarine was produced as a mixture Subjects were counselled to maintain their usual intake of of interesteri®ed palm stearine (50%) and Sunola1 (50%). and to abstain from consuming alcohol 24 h prior to Both stearic acid and palmitic acid levels in their respective blood sampling. All subjects were required to give written margarines were maximized within the limits of physical consent to participate and were free to withdraw at any time. property characteristics conducive to normal use. The oleic The study protocol was approved by the Human Research acid content in the stearic acid-rich fats and biscuits was Ethics Committee of RMIT University (approval number higher than for the palmitic acid-rich fats and biscuits in 17=96). order for stearic acid products to be spreadable at room temperature. The fatty acid pro®le of the margarines used is Experimental design shown in Table 1. The dietary intervention consisted of subjects consuming both a high stearic (diet S) and a high palmitic acid diet (diet Physiological measurements P) for 4 weeks in a random crossover design, with a 7 week Subjects had their weight, height and body mass index wash-out (habitual diet) period between the two phases. (BMI, kg=m2) determined, and percentage body fat mea- Subjects consumed the equivalent of approximately two- sured using a bioimpedence fat analyser=scale (TBF-501 thirds of their habitual fat intake, as the test fats, during the Tanita Corporation, Illinois, USA). Systolic and diastolic intervention periods. blood pressures and pulse were also measured at each

European Journal of Clinical Nutrition Stearic acid-rich diet FD Kelly et al 90 Table 1 Fatty acid composition and fat content of test fats and biscuits Platelet and plasma fatty acids Platelet and plasma were extracted with : Fatty acid content (as percentage total fatty acids) 1:1 (C:M, v=v) containing 10 mg=l of butylated Stearic acid-enriched Palmitic acid-enriched hydroxytoluene (Labco, Vic, Australia), and 10 mg=lof C17:0 (La- diheptadeca- Spreading Baking Spreading Baking noyl, Sigma-Aldrich, Pty. Limited, NSW, Australia) as margarine margarine Biscuit margarine margarine Biscuit internal standard, as reported previously (Sinclair et al, C16:0 4.3 4.4 4.7 31.3 42.1 41.7 1987). The total platelet PL and total plasma PL fractions C18:0 29.1 33.8 31.1 4.5 4.5 4.5 were separated by thin-layer chromatography. The methyl C18:1 56.2 51.8 50.9 53.5 43.1 42.4 of the fatty acids were prepared by saponi®cation C18:2 7.1 6.6 7.7 7.8 7.6 8.6 using KOH (0.68 mol=l in methanol) followed by transes- C18:3 0.7 0.6 0.8 0.5 0.3 0.5 teri®cation with 20% boron tri¯uoride (BF3) in methanol Fata 82 82 24 79 79 22 and the fatty acid compositions were determined by gas ± liquid chromatography as previously described (Sinclair ag=100 g edible fat or biscuit. et al, 1987). appointment on a Lumiscope Digital Auto In¯ate Blood Pressure Monitor (Lumiscope Company Inc., NJ, USA). Full blood examination A full blood cell count (including platelet count (Plt) and mean platelet volume (MPV)) was performed on an auto- Blood sampling and preparation mated haematological analyser (Coulter Counter STKR, Fasting venous blood samples (  10 h) were taken from the Coulter Electronics Inc., Hialeah, USA) within 2 h of subjects before the commencement (day 0) and at the end blood collection. (day 28) of each study period by a quali®ed nurse at the RMIT University Health Service using standard vacutainer collection tubes. Blood for lipid analysis and cell counts was Plasma haemostatic factors collected into EDTA containing vacutainers (Greiner Labor- Prothrombin time (PT), activated partial thromboplastin technik, Austria). Blood for platelet fatty acid analysis was time (APTT), ®brinogen, factor VII coagulant activity collected into CTAD (citric acid, theophylline, adenosine, (FVII:C), plasminogen and antithrombin III (ATIII) levels dipyridamole) vacutainers (Becton Dickinson Ltd, Cowley, were determined using a centrifugal analyser (ACL 200, UK). Blood for whole blood platelet aggregation and the Coulter IL, Ltd) with commercially available kits (Coulter, coagulation and ®brinolytic factors was collected into sepa- Instrumentation Laboratory, Milano, Italy). rate sodium citrate vacutainers (Greiner Labortechnik). EDTA and citrated plasma were obtained from blood spun  Agonist induced ex vivo whole blood platelet aggregation at 3000 rpm at 4 C for 15 min. Plasma phospholipid (PL) Platelet aggregation was determined in whole blood using fatty acids, total cholesterol, HDL cholesterol and triacyl- a two-channel whole blood impedance aggregometer were determined from EDTA plasma stored at   (Chrono-Log Aggregometer, Model 540V5, Chrono-Log 20 C. Citrated plasma was stored at 70 C for subsequent Corporation, Havertown, Philadelphia, USA). Agonists analysis of coagulation and ®brinolytic factors. Following  used were collagen (2 mg=l), (AA, blood collection, CTAD tubes were maintained at 37 Cto 1.0 mmol=l) and adenosine diphosphate (ADP) (8 and prevent platelet activation and used within 1 h for platelet 17 mmol=l). Aggregation was measured as rate of aggrega- aggregation by the method of Castaldi and Smith (1980). tion (slope, O=min) as reported by Ingerman-Wojenski and Silver (1984). Plasma lipoprotein lipids Total cholesterol (TC) and triacylglycerol (TAG) concen- Statistical analyses trations in plasma were determined with a centrifugal Statistical analysis was performed using Statistical Packages autoanalyser (Hitachi Autoanalyser System 705, Japan) for Social Scientists (SPSS version 9.0, 1998, Chicago, IL). using commercially available enzymatic kits, CHOD-PAP Comparisons between the dietary groups involved a three- and GPO-PAP for TC and TAG, respectively (Boehringer way repeated measures analysis of variance using a general Mannheim, Germany). High- lipoprotein cholesterol linear model (GLM), where diet type (palmitic or stearic (HDL-C) was measured after all other plasma lipoproteins rich) and time were within-subject factors and dietary order were precipitated with polyethylene glycol 6000 (PEG- was the between-subject factor (P < 0.05 was regarded as 6000, BDH Chemicals, Kilsyth, Victoria). Low density signi®cant). Comparisons within individual diets (between lipoprotein cholesterol (LDL-C) was calculated from the days 0 and 28) were conducted on parameters where a values obtained for TC, HDL-C and TAG for each subject signi®cant time  treatment interaction or time effect was by using the Friedewald formula as developed by DeLong observed using paired t-tests and Bonferroni correction et al (1986). (P < 0.025 was regarded as signi®cant).

European Journal of Clinical Nutrition Stearic acid-rich diet FD Kelly et al 91 Results diet P. There was a signi®cant decrease in fat intake (%TE, P < 0.025) compared with baseline during diet P. Compared All subjects stayed within 2 kg of their initial body weights with baseline, there was a signi®cant decrease (P < 0.025) in throughout the dietary phases of the study. There were no the percentage of total energy (%TE) from saturated and signi®cant changes in BMI, waist-to-hip ratio, or blood polyunsaturated fatty acids (PUFA) during diet S and a pressure during either dietary period. signi®cant decrease (P < 0.025) in PUFA during diet P. The mean intakes for all subjects at baseline and Compared with their respective baseline intakes, there was during the two dietary phases are presented in Table 2. a signi®cant increase (P < 0.025) in MUFA (%TE) during There were no signi®cant differences in total energy intake both diet S and diet P. during either dietary period compared with baseline or Therewasa12g=day increase in stearic acid between the two dietary phases. The changes in the percen- (P < 0.025) and a 9 g=day decrease (P < 0.025) in pal- tage of total energy (%TE) of protein, fat and carbohydrate miticacidintakeondietS.OndietP,therewasa during diet S were not signi®cantly different to those during 7g=day increase (P < 0.025) in palmitic acid and a

Table 2 Daily nutrient intakes of subjects at baseline (habitual diet) and during the 4 week stearic acid (diet S) and palmitic acid (diet P) intervention periods (n ˆ 13)

Diet S a Diet P a P b

Nutrient Day 0 c Day 28 d Day 0 c Day 28 d TTÂ D

Total energy (kJ) 10259Æ 1676 11001Æ 1680 10335Æ 1537 10694Æ 1561 NS NS Protein (%TE) 17.1Æ 1.5 16.0Æ 1.6 16.8Æ 2.2 16.9Æ 1.3 NS NS Fat (%TE) 29.3Æ 5.1 27.8Æ 3.8 29.9Æ 5.3 27.4Æ 4.1* < 0.05 NS SFA (%TE) 11.5Æ 2.5 10.0Æ 1.5* 11.5Æ 2.5 10.3Æ 1.3 < 0.05 NS MUFA (%TE) 10.3Æ 1.5 13.0Æ 2.1* 10.5Æ 2.1 12.3Æ 2.2* < 0.05 NS PUFA (%TE) 4.7Æ 1.5 2.9Æ 0.3* 4.6Æ 1.8 2.9Æ 0.4* < 0.05 NS Carbohydrate (%TE) 50.4Æ 5.4 54.4Æ 4.1* 50.7Æ 4.9 53.5Æ 4.0 < 0.05 NS Alcohol (%TE) 3.2Æ 2.9 2.1Æ 2.1 2.6Æ 2.3 2.2Æ 2.6 < 0.05 NS Dietary ®bre (g) 25Æ 730Æ 827Æ 829Æ 9NSNS Cholesterol (mg) 244Æ 59 172Æ 42* 236Æ 72 185Æ 50 < 0.05 NS

aValues are meanÆ s.d., based on 7 day weighed food records. bP, level of signi®cance; T, signi®cant time effect, P < 0.05; T  D, signi®cantly different change in dietary intake between diets, P < 0.05 (time  treatment interaction); GLM Ð Repeated Measures ANOVA. cBaseline dietary intake. dDietary intake during the study. %TE ˆ percentage of total energy intake. *Signi®cantly different from baseline within diet, P < 0.025 (paired t-test Bonferroni correction), applied to parameters where a signi®cant time  treatment interaction or time effect was observed.

Table 3 Daily fatty acid intakes of subjects at baseline (habitual diet) and during the 4 week stearic acid (diet S) and palmitic acid (diet P) intervention periods (n ˆ 13)

Diet S a Diet P a P b

Dietary fatty acid Day 0 c Day 28 d Day 0 c Day 28 d TTÂ D

Total fat (g) 79Æ 16 81Æ 15 81Æ 17 77Æ 14 NS NS C12:0 (g) 1.3Æ 0.8 0.2Æ 0.1* 1.4Æ 0.8 0.2Æ 0.1* < 0.05 NS C14:0 (g) 29.Æ 1.0 0.7Æ 0.3* 3.0Æ 1.0 1.1Æ 0.2* < 0.05 NS C16:0 (g) 15.4Æ 3.3 6.6Æ 1.4* 15.5Æ 3.3 22.5Æ 5.3* < 0.05 < 0.05 C18:0 (g) 7.3Æ 1.9 19.4Æ 4.5* 7.6Æ 2.1 4.4Æ 0.8* < 0.05 < 0.05 C16:1 (g) 1.3Æ 0.3 0.4Æ 0.2* 1.3Æ 0.3 0.6Æ 0.2* < 0.05 NS C18:1 (g) 25.2Æ 5.1 37.3Æ 7.9* 26.0Æ 6.5 33.1Æ 6.7* < 0.05 < 0.05 C18:2 (g) 11.1Æ 3.9 7.3Æ 1.3* 11.7Æ 5.1 7.7Æ 1.3* < 0.05 NS C18:3 (g) 1.1Æ 0.5 0.8Æ 0.3* 1.1Æ 0.5 0.7Æ 0.2* < 0.05 NS

aValues are meanÆ s.d., based on 7 day weighed food records. bP, level of signi®cance; T, signi®cant time effect, P < 0.05; T  D, signi®cantly different change in dietary intake between diets, P < 0.05 (time  treatment interaction); GLM Ð repeated measures ANOVA. cBaseline dietary intake. dDietary intake during the study. *Signi®cantly different from baseline within diet, P < 0.025 (paired t-test Bonferroni correction), applied to parameters where a signi®cant time  treatment interaction or time effect was observed.

European Journal of Clinical Nutrition Stearic acid-rich diet FD Kelly et al 92 Table 4 Lipoprotein concentration of subjects at baseline (habitual diet) and during the 4 week stearic acid (diet S) and palmitic acid (diet P) intervention periods (n ˆ 13)

Diet S a Diet P a P b

Parameter Day 0 Day 28 Day 0 Day 28 T T Â D

TC (mmol=l) 4.67Æ 1.03 4.21Æ 0.94* 4.71Æ 1.04 4.44Æ 1.00{ < 0.05 NS HDL-C (mmol=l) 1.22Æ 0.32 1.09Æ 0.24* 1.20Æ 0.28 1.13Æ 0.31{ < 0.05 NS LDL-C (mmol=l) 3.03Æ 0.90 2.67Æ 0.79* 3.13Æ 0.98 2.84Æ 0.90 < 0.05 NS TAG (mmol=l) 1.15Æ 0.64 1.20Æ 0.62 1.02Æ 0.43 1.15Æ 0.64 NS NS

aValues are meanÆ s.d. bP, level of signi®cance; T, signi®cant time effect, P < 0.05; T  D, signi®cantly different change in dietary intake between diets, P < 0.05 (time  treatment interaction); GLM Ð repeated Measures ANOVA. *Signi®cantly different from baseline within diet, P < 0.025 (paired t-test Bonferroni correction), applied to parameters where a signi®cant time  treatment interaction or time effect was observed. {Signi®cant dietary order effect, P < 0.05 (GLM Ð repeated measures ANOVA).

3g=day decrease (P < 0.025) in stearic acid. The changes The full blood examination and haemostatic parameters in the amount of stearic, palmitic and oleic acid intake are reported in Table 5. There was a signi®cant difference on diet S were signi®cantly different (P < 0.05) to those (P < 0.05) in the change in MPV on diet S compared with on diet P. On both diets there were signi®cant changes diet P. The MPV decreased signi®cantly following diet S (relative to baseline) for other fatty acids, particularly an (P < 0.025) while there was no change in MPV on diet P. increase in oleic acid and decrease in (Table The diet order effect on diet P was due to one group 3). While these changes were not planned, they were the showing an increased MPV value. The FVII:C activity result of efforts to maintain energy intake constant while was signi®cantly decreased (P < 0.025) during diet S com- removing certain fat sources from the diet and replacing pared with baseline. The FVII:C activity and ®brinogen them with the test fats. levels of two samples taken during diet S were excluded The results of lipoprotein concentrations are reported in from calculations due to suspected pre-activation of clotting Table 4. Diet S resulted in a signi®cant decrease in total- (related to venipuncture technique). There were no signi®- cholesterol (TC), HDL-cholesterol (HDL-C) and LDL- cant changes in PT, APTT, plasminogen, ®brinogen or cholesterol (LDL-C, P < 0.025) compared with baseline, ATIII on either diet. however, there were no signi®cant changes in plasma Fatty acid compositions of platelet PL are shown in Table lipoprotein lipids on diet P. A diet order effect was present 6. There were signi®cant differences in the changes in in the changes in TC and HDL-C (ie affected by whether platelet stearic and palmitic acid levels on diet S compared the subjects started on diet S or diet P ®rst). TC and HDL-C with diet P (P < 0.05). Diet S resulted in signi®cantly fell in both groups on diet P but fell signi®cantly more in increased stearic-acid levels and signi®cantly decreased one group than the other on diet S. palmitic acid levels in platelet PL compared with baseline

Table 5 Full blood examination and haematological parameters of subjects at baseline and the end of the 4 week stearic acid (diet S) and palmitic acid (diet P) intervention periods (n ˆ 13)

Diet S a Diet P a P c

Parameter Day 0 Day 28 Day 0 Day 28 T T Â D

WBC (109=l) 6.6Æ 1.0 6.5Æ 1.2 6.3Æ 0.9 6.7Æ 1.1 NS NS RBC (1012=l) 5.1Æ 0.3 5.0Æ 0.2 5.0Æ 0.2 5.0Æ 0.3 NS NS Hgb (g=l) 153Æ 9 152Æ 7 151Æ 9 152Æ 12 NS NS PLT (109=l) 223Æ 40 225Æ 33 228Æ 38 226Æ 35 NS NS MPV (¯) 8.9Æ 0.6 8.4Æ 0.5* 8.8Æ 0.6 9.0Æ 0.6{ < 0.05 < 0.05 Fibrinogen (g=l)b 2.4Æ 0.4 2.5Æ 0.3 2.4Æ 0.4 2.4Æ 0.5 NS NS FVII (% activity) b 94Æ 24 85Æ 18* 96Æ 21 92Æ 24 < 0.05 NS PT (s) 14.2Æ 0.8 14.6Æ 0.8 14.2Æ 1.0 14.9Æ 1.6 < 0.05 NS APTT (s) 33.9Æ 3.2 33.8Æ 3.8 33.8Æ 5.2 35.6Æ 6.4 NS NS ATIII (%) 90Æ 17 93Æ 13 92Æ 11 92Æ 11 NS NS Plasminogen (%) 88Æ 10 85Æ 885Æ 11 86Æ 9NSNS

aValues are meanÆ s.d., bn ˆ 11. cP, level of signi®cance; T, signi®cant time effect, P < 0.05; T  D, signi®cantly different change in parameter between diets, P < 0.05 (time  treatment interaction); GLM Ð repeated measures ANOVA. *Signi®cantly different from baseline within diet, P < 0.025 (paired t-test Bonferroni correction), applied to parameters where a signi®cant time  treatment interaction or time effect was observed. {Signi®cant dietary order effect, P < 0.05 (GLM Ð repeated measures ANOVA).

European Journal of Clinical Nutrition Stearic acid-rich diet FD Kelly et al 93 (P < 0.025). Diet P resulted in a signi®cant increase Agonist-induced whole blood platelet aggregation results (P < 0.025) in the palmitic acid level in platelet PL. These are reported in Table 7. On diet S, there were no signi®cant changes were of a lesser magnitude compared with the changes in the rate of aggregation while during diet P, the change on diet S. Both dietary interventions resulted in a rate of aggregation was signi®cantly increased (P < 0.025) signi®cant decrease in linoleic acid and a signi®cant in response to collagen and ADP (8 and 17 mm) compared to increase in oleic acid (P < 0.025) compared with baseline. baseline. On diet S there was a signi®cant increase (P < 0.025) in the stearic acid proportion in the plasma PL (13.6 Æ 1.2% to 16.5 Æ 2.3% of plasma PL fatty acids) and a signi®cant Discussion decrease (P < 0.025) in the level of palmitic acid (28.1 Æ 1.7% to 25.9 Æ 2.4%). Diet P led to a signi®cant This study examined the effect of diets rich in stearic acid increase (P < 0.025) in the level of palmitic acid relative to palmitic acid on thrombotic risk factors asso- (28.3 Æ 1.7% to 29.9 Æ 1.2%). The changes in the proportion ciated with platelets and haemostasis, with platelet aggrega- of stearic acid and palmitic acids on diet S were signi®cantly tion as an important outcome. The test fats were prepared by different (P < 0.05) to those on diet P (data not shown). interesterifying blends of particular fats as described, which

Table 6 Fatty acid composition of total platelet phospholipids (percentage fatty acid) of subjects at baseline and the end of the 4 week stearic acid (diet S) and palmitic acid (diet P) intervention periods (n ˆ 13)

Diet S a Diet P a P b

Nutrient Day 0 c Day 28 d Day 0 c Day 28 d TTÂ D

C14:0 0.2Æ 0.1 0.1Æ 0.0* 0.2Æ 0.1 0.1Æ 0.0* < 0.05 NS C16:0 15.4Æ 0.9 13.0Æ 1.4* 15.5Æ 0.9 16.7Æ 1.1* < 0.05 < 0.05 C17:0 0.5Æ 0.1 0.4Æ 0.1 0.4Æ 0.1 0.4Æ 0.0 < 0.05 NS C18:0 20.1Æ 1.6 24.6Æ 2.2* 21.4Æ 1.1 20.1Æ 1.7 < 0.05 < 0.05 C16:ln-9 0.3Æ 0.1 0.3Æ 0.1 0.3Æ 0.0 0.3Æ 0.1 NS NS C16:ln-7 0.2Æ 0.1 0.2Æ 0.1* 0.2Æ 0.1 0.2Æ 0.1 < 0.05 < 0.05 C18:ln-9 14.0Æ 0.8 14.9Æ 0.8* 14.2Æ 0.9 15.2Æ 0.7* < 0.05 NS C18:ln-7 1.0Æ 0.2 1.0Æ 0.1 1.0Æ 0.2 1.0Æ 0.2 NS < 0.05 C18:2n-6 5.1Æ 0.5 4.5Æ 0.4* 5.1Æ 0.5 4.5Æ 0.4* < 0.05 NS C20:0 0.9Æ 0.1 1.0Æ 0.2* 0.9Æ 0.1 0.8Æ 0.1* < 0.05 < 0.05 C20:1 0.6Æ 0.1 0.6Æ 0.1 0.6Æ 0.1 0.7Æ 0.1* NS < 0.05 C20:3n-6 1.6Æ 0.4 1.5Æ 0.4 1.5Æ 0.3 1.5Æ 0.4 NS NS C20:4n-6 25.1Æ 1.3 25.8Æ 0.9 25.1Æ 1.2 24.9Æ 1.0 NS NS C20:5n-3 0.4Æ 0.2 0.3Æ 0.1 0.4Æ 0.2 0.3Æ 0.1 NS NS C22:4n-6 2.4Æ 0.4 2.5Æ 0.4 2.4Æ 0.4 2.4Æ 0.4 NS NS C22:5n-6 0.3Æ 0.1 0.3Æ 0.1 0.3Æ 0.1 0.2Æ 0.1 NS NS C22:5n-3 1.9Æ 0.3 1.8Æ 0.2 1.9Æ 0.4 1.7Æ 0.2* < 0.05 NS C22:6n-3 1.7Æ 0.4 1.7Æ 0.4 1.7Æ 0.4 1.7Æ 0.4 NS NS

aValues are meanÆ s.d. bP, level of signi®cance; T, signi®cant time effect, P < 0.05; T  D, signi®cantly different change in parameter between diets, P < 0.05 (time  treatment interaction); GLM Ð repeated measures ANOVA. *Signi®cantly different from baseline within diet, P < 0.025 (paired t-test Bonferroni correction), applied to parameters where a signi®cant time  treatment interaction or time effect was observed.

Table 7 Agonist-induced whole blood platelet aggregation in subjects at baseline and the end of the 4 week stearic acid (diet S) and palmitic acid (diet P) intervention periods

Diet S a Diet P a Pd

Nutrient Day 0 Day 28 Day 0 Day 28 T T Â D

Collagen (2 mg=ml)c 6.6Æ 1.1 7.0Æ 1.5 7.4Æ 2.0 8.4Æ 1.5* < 0.05 NS Arachidonate (1.0 mM)b 7.4Æ 2.3 7.0Æ 3.0 7.5Æ 1.5 8.3Æ 1.4 NS NS ADP (8 mM)c 3.8Æ 1.5 4.0Æ 2.2 3.8Æ 1.4 5.2Æ 2.3* < 0.05 NS ADP (17 mM)c 4.7Æ 1.6 4.9Æ 1.9 5.1Æ 1.6 6.4Æ 2.4* < 0.05 NS

aValues are meanÆ s.d.; bn ˆ 11; cn ˆ 12. dP, level of signi®cance; T, signi®cant time effect, P < 0.05; T  D, signi®cantly different change in dietary intake between diets, P < 0.05 (time  treatment interaction); GLM Ð repeated measures ANOVA. *Signi®cantly different from baseline within diet, P < 0.025 (paired t-test Bonferroni correction), applied to parameters where a signi®cant time  treatment interaction or time effect was observed.

European Journal of Clinical Nutrition Stearic acid-rich diet FD Kelly et al 94 gave stearic acid or palmitic acid-rich products; however, in enhanced collagen-induced platelet aggregation (but not order to make the stearic products spreadable, the stearic with ADP) compared with a high trans fatty acid diet acid products contained a higher level of oleic acid than the (8.7% total energy). In comparison, a study by Kwon et al palmitic acid products. Thus, it is conceivable that some of (1991) incorporating a 3 week controlled SFA diet, demon- the results found could be in¯uenced by the stearic acid strated using whole blood that stearic acid in platelet content as well as the oleic acid content. In reality, the oleic phospholipids was associated (r ˆ0.69) with decreased acid intake=day=subject on the stearic acid diet was collagen-induced platelet aggregation, suggesting that 37 g=day compared with 33 g=day on the palmitic acid stearic acid may not be prothrombotic. diet and the difference was probably not of biological Tholstrup et al (1996) studied the acute effects of stearic signi®cance. We do not know of any evidence which and myristic acids on ADP- and collagen-induced platelet shows that this small difference in oleic acid between the aggregation. Both diets reduced platelet aggregation post- two diets would produce the changes in platelet aggregation, prandially after a high-fat meal compared with fasting MPV or Factor VII observed. levels, with signi®cantly lower platelet aggregation at 24 h There were no effects of diet S on collagen or ADP after consumption of stearic acid than . It was agonist-induced in vitro platelet aggregation, despite the hypothesized that a mechanism for this decrease in aggrega- fact that the stearic acid level in platelets increased sig- tion was the coating of the platelets with chylomicrons, ni®cantly (P < 0.025). In contrast, diet P led to a signi®cant which may interfere with the platelet ± collagen interac- increase in the level of palmitic acid in the platelets which tion in the initial stage of platelet aggregation (platelet was associated with a signi®cant increase in agonist- activation). induced platelet aggregation in response to collagen and In the present study, MPV signi®cantly decreased on diet ADP. This ®nding is novel since no previous studies have S by 6% relative to baseline and by 7% relative to diet P. compared platelet aggregation on diets rich in stearic acid vs This ®nding is supported by data from Schoene et al (1992), palmitic acid. who reported a similar decrease in MPV in 10 subjects fed In vitro platelet aggregation is based on the assumption in excess of 20 g stearic acid=day from compared that an increased response indicates an increased tendency with a high palmitic acid diet. The MPV is regarded as an for thrombogenesis in vivo. Interpretation of results, how- index of platelet activation and has been shown to be an ever, is dif®cult due to many factors affecting in vitro independent risk factor for recurrent measurements (Schoene, 1997). To date, the data reported (Martin et al, 1991; Schultheiss et al, 1994). Platelets by various authors on the effect of dietary stearic acid on normally circulate as thin discs and smaller platelets are platelet aggregation are con¯icting. It has been demon- thought to be less active and, thus, in a quiescent state are strated that, for a particular dietary fatty acid, the aggrega- less likely to be involved in thrombotic conditions (Schoene tory response to the same agonist can give opposing results et al, 1992). Activated platelets produce changes in the (Turpeinen et al, 1998; Kwon et al, 1991), due partly to the platelet microtubules, causing them to undergo a disc-to- medium in which platelets have been assessed. There is sphere shape transformation (Laufer et al, 1979), resulting argument as to whether whole blood or platelet-rich plasma in an increase in volume, and these larger platelets have (PRP) samples are more appropriate for determination of ex been shown to have enhanced activity (Abbate & Boddi, vivo aggregation, or as a model to represent the in vivo 1987; Sharp et al, 1994). Flow cytometry techniques have situation. Whole blood aggregation testing used in the demonstrated that this greater reactivity of larger platelets is present study may represent more closely physiological in associated with more ®brinogen receptors being exposed, vivo conditions compared with PRP used in other studies initiating arterial thrombi formation to a greater extent than (Renaud et al, 1981; 1986a, b). It has been demonstrated, smaller platelets (Giles et al, 1994; Michelson, 1996). using whole blood in vitro aggregation techniques, that Further studies on the biological signi®cance in the reduc- platelet aggregates contain both leukocytes and erythrocytes tion of the MPV following stearic acid-rich diets are (Joseph et al, 1989). Leukocytes can down-regulate platelet warranted. reactivity via synthesis while erythrocytes can Prospective studies have demonstrated that coagulation enhance platelet activation via ADP release, the active take- factor VII is an independent predictor of total and fatal CHD up of adenine nucleotides and the preferential binding of in the ®rst 5 y of follow up of the Northwick Park Heart prostacylin. These factors which can in¯uence platelet Study (NPHS) (Meade et al, 1980, 1986). This observation function are excluded in the process of aggregation in has also been found in fatal CHD in longer follow-up PRP as opposed to whole blood techniques which allow periods of the NPHS (Meade et al, 1993) and The Prospec- the study of platelets in their natural milieu (Joseph et al, tive Cardiovascular Munster (PROCAM) study (Assmann et 1989; Marcus & Sa®er, 1993; Homstra, 1989). Despite this, al, 1996). It has also been suggested that decreasing the total some researchers still believe that PRP is more appropriate dietary fat content reduces factor VII coagulant activity since it removes such factors giving a better picture of the (Marckmann et al, 1990, 1993, 1994; Miller et al, 1989). changes in platelets (Turpeinen et al, 1998). FVII:C signi®cantly decreased on diet S relative to baseline, Turpeinen et al (1998) demonstrated using PRP that a suggesting favourable effects on haemostasis while no high stearic acid containing diet (9.3% total energy) for 5 signi®cant change was observed on diet P. Tholstrup et al, weeks in a group of 80 volunteers led to a signi®cantly (1994) also observed a 13% decrease in factor VII on a

European Journal of Clinical Nutrition Stearic acid-rich diet FD Kelly et al 95 stearic acid-rich diet (from shea butter) compared with a present study. It has been shown there is a high proportion high palmitic acid diet. In contrast, Mutanen & Aro (1997) of stearic acid in platelet and phospha- demonstrated no signi®cant difference in the level of FVII:C tidylinositol, which are important phospholipids for cell in 40 subjects fed stearic acid-rich diets (9.3% total energy) membrane signalling events (Mori et al, 1987). Perhaps for 5 weeks. Some authors have suggested that shea fat has a the increased stearic acid in the platelet membranes is unique ability in lowering FVII:C which may be unrelated speci®c for one of these two phospholipids or alternatively to its fatty acid composition, but more related to the non- the increased stearic acid might in¯uence the function of components such as and hydrocarbons receptors through changes in membrane (Tholstrup et al, 1994; Mutanen & Aro, 1997). ¯uidity (Litman & Mitchell, 1996). Total cholesterol, LDL-C and HDL-C levels in the This study supports other studies which show that diets plasma were all signi®cantly decreased on diet S. The enriched in stearic acid do not contribute to an increase in total cholesterol and LDL-C results observed are consistent classical cardiovascular risk factors and those related to with data from Bonanome and Grundy (1988), Tholstrup thrombosis. These data might appear to be in contradiction et al (1994) and Dougherty et al (1995), while signi®cant with the recent report by Hu et al (1999), who suggest that decreases in HDL-C concentrations on high stearic acid there is no basis for a distinction between stearic acid and diets compared with baseline have been previously reported other saturated fatty acids in `normal' diets, because of the by Tholstrup et al (1994) and Dougherty et al (1995). The high correlation between these two factors. Since stearic lack of effect on the plasma TAG concentration is consistent acid levels in `normal' diets are half those of palmitic plus with other data on stearic acid rich diets (Tholstrup et al, myristic acid levels, it is not a major saturated fatty acid in 1994; Dougherty et al, 1995; Bonanome & Grundy, 1988). our diet. Our data on a modi®ed diet, containing approxi- The mechanisms by which stearic acid lower plasma chol- mately 2.5 times more stearic acid than usual, raises the esterol levels are still uncertain; however, they may in part possibility of using an increased amount of stearic acid in be due to the regio-speci®c location of the saturated fatty the food supply in place of those fats rich in palmitic and acids, which in turn in¯uences their absorption and further myristic acids or trans fatty acids. The stearic acid could metabolism. These properties have been attributed, ®rstly, to satisfy the requirements for particular physical properties in the lower gastrointestinal absorption rates of stearate, espe- the food while appearing to have either neutral or bene®cial cially when found in the sn-1 and sn-3 positions of the effects in relation to risk factors for . triacylglycerol (Kritchevsky, 1994; Bracco, 1994). This is largely due to its being above body tempera- Acknowledgements ÐL Johnson and M Francis provided technical assis- ture and its ability to form calcium (Small, 1991). tance. We gratefully acknowledge Dr Gerald R Elsworth from the Centre Secondly, it is possible that stearic acid is rapidly converted for Program Evaluation, The University of Melbourne, Victoria, Australia to oleic acid in the liver (Grundy, 1994), thus reducing its for advice on the statistical analysis of the data. potential impact on saturated fatty acid levels in the body. In natural lipid sources such as cocoa butter and shea butter (Dougherty et al, 1995), the stearic acid is predominantly References located in the sn-1 and sn-3 positions with minimal amounts in the sn-2 position (Padley et al, 1994). However, when fats Abbate R & Boddi M (1987): Which test(s) for platelet aggregation? In Antiplatelet Therapy: Twenty Years' Experience, ed. GVR Born, GG are interesteri®ed, as with the fats used in the present study, Neri Serneri, pp 91 ± 102. Florence: Exerpta Medica. a greater proportion of stearic acid is located in the sn-2 Assmann G, Cullen P, Heinrich J & Schulte H (1996): Haemostatic position, which is absorbed into the mucosal cells as a variables in the prediction of coronary risk: results of the 8 year monoacylglycerol. Further studies to show the absorption of follow-up healthy men in the Munster Heart Study (PROCAM). Isr. J. stearic acid from the sn-2 position could include chylo- Med. Sci. 32, 364 ± 370. Bonanome A & Grundy SM (1988): Effect of dietary stearic acid micron studies postprandially after stearic feeding and also on plasma cholesterol and lipoprotein levels. New Engl. J. Med. 318, faecal studies to determine apparent absorption of these 1244 ± 1248. formulated fats. Bracco U (1994): Effect of triglygeride structure on fat absorption. Am. J. The use of stearic acid-rich lipid sources could be of Clin. Nutr. 60(Suppl), 1002S ± 1009S. Castaldi PM & Smith IL (1980): The effect of platelets on the in vitro advantage to the food industry if it displaces other saturated response to prothrombin complex concentrates in FVIII inhibitor plasma. and trans fatty acids from the foods, giving them nutritional Pathology 12, 111 ± 118. and physiological advantages over natural lipid sources. Connor WE (1962): The acceleration of thrombus formation by certain fatty Furthermore, if the stearic acid-rich lipid source is inter- acids. J. Clin. Invest. 41, 1199 ± 1205. esteri®ed, a greater proportion of stearic acid could be Davies MJ & Thomas A (1984): Thrombosis and acute coronary- artery lesions in sudden cardiac ischemic death. New Engl. J. Med. randomly distributed to the sn-2 position which in turn 310, 1137 ± 1140. could increase the amounts absorbed compared with natural DeLong DM, DeLong ER, Wood PD, Lippel K & Rifkind BM (1986): A lipid sources. comparison of methods for the estimation of plasma low- and very low- In this study, the platelet stearic acid level increased by density lipoprotein cholesterol. The LRC prevalence study. J. Am. Med. 22% on diet S, which is consistent with the extent of Assoc. 256, 2372 ± 2377. Dougherty RM, Allman MA & Iacono JM (1995): Effects of diets contain- increase reported by Turpeinen et al (1998). Plasma phos- ing high or low amounts of stearic acid on plasma lipoprotein fractions pholipid stearic acid levels also increased by 21% in the and fecal fatty acid excretion of men. Am. J. Clin. Nutr. 61, 1120 ± 1128.

European Journal of Clinical Nutrition Stearic acid-rich diet FD Kelly et al 96 Giddings JC & Yamamoto J (1995): Changing concepts in investigations of Michelson AD (1996): Flow cytometry: a clinical test of platelet function. haemostasis. Clin. Lab. Haematol. 17, 85 ± 91. Blood 87, 4925 ± 4936. Giles H, Smith RE & Martin IF (1994): Platelet glycoprotein IIb-IIIa and Miller GJ, Cruikshank JFK, Ellis LJ, Thompson RL, Wilkes HG, Stirling Y, size are increased in acute Ins myocardial infarction. Eur. J. Clin. Invest. Mitropoulos KA, Allison JV, Fox TE & Walker AO (1989): Fat 24, 69 ± 72. consumption and factor VII coagulant activity in middle-aged men. Grundy SM (1994): In¯uence of stearic acid on cholesterol metabolism 78, 19 ± 24. relative to other long-chain fatty acids. Am. J. Clin. Nutr. 60(Suppl), Mori TA, Codde JP, Vandongen R & Beilin LJ (1987): New ®ndings in the 986S±990S. fatty acid composition of individual platelet phospholipids in man after Hegsted DM, McGandy RB, Myers ML & Stare FT (1965): Quantitative dietary ®sh oil supplementation. Lipids 22, 744 ± 750. effects of dietary fat on serum cholesterol in man. Am. J. Clin. Nutr. 17, Mutanen M & Aro A (1997): Coagulation and ®brinolysis factors in healthy 281 ± 295. subjects consuming high stearic acid or trans fatty acid diets. Thromb. Hornstra G (1989): The signi®cance of ®sh and ®sh-oil enriched food for Haemost. 77, 99 ± 104. the prevention and therapy of ischemic cardiovascular disease. In The National Food Authority (1995): Composition of Foods, Australia Can- Role of Fats in , 2nd edn, ed. AJ Vergroesen, M berra, Australian Government Publishing Service. Crawford, pp 151 ± 235. San Diego, USA: Academic Press. Padley EB, Harwood, JL & Gunstone FD (1994): Occurrence and char- Hu EB, Stampfer MJ, Manson JE, Ascherio A, Colditz GA, Speizer FE, acteristics of and lipids. In The Lipid Handbook 2nd edn., eds. FD Hennekens CH & Willett WC (1999): Dietary saturated fats and their Gunstone, JL Harwood & FB Padley, p 120. Cambridge: Chapman & food sources in relation to the risk of coronary heart disease in women. Hall. Am. J. Clin. Nutr. 70, 1001 ± 1008. Renaud S & Gautheron P (1975): In¯uence of dietary fats on atherosclero- Ingerman-Wojenski CM & Silver MI (1984): A quick method for screening sis, coagulation and platelet phospholipids in rabbits. Atherosclerosis for platelet dysfunctions using the whole blood lumi-aggregometer. 21, 115±124. Thromb. Haemost. 51, 154 ± 156. Renaud S, Dumont E, Godsey F, Suplisson A & Thevenon C (1978): Joseph R, Welch KMA, D'Andrea C & Riddle JM (1989): Evidence for the Platelet functions in relation to dietary fats in farmers from two regions presence of red and white cells within `platelet' aggregates formed in of France. Thromb. Haemost. 40, 518 ± 531. whole-blood. Thromb. Res. 53, 485 ± 491. Renaud S, Morazain R, Godsey F, Dumont E, Symington IS, Gillanders EM Keys A, Anderson JT & Grande F (1957): Prediction of serum cholesterol & O'Brien JR (1981): Platelet functions in relation to diet and serum responses of man to changes in fats in the diet. Lancet ii, 959 ± 966. lipids in British farmers. Br. Heart J. 46, 562 ± 570. Keys A, Anderson, JT & Grande F (1965): Serum cholesterol response to Renaud S, Godsey F, Dumont E, Thevenon C, Ortehanian E & Martin JL changes in the diet. IV. Particular saturated fatty acids in the diet. (1986a): In¯uence of long term diet modi®cation on platelet function and Metabolism 14, 776 ± 786. composition in Moselle farmers. Am. J. Clin. Nutr. 43, 136 ± 150. Kris-Etherton PM & Mustad VA (1994): feeding studies: a novel Renaud S, Morazain R, Godsey F, Dumont E, Thevenon C, Martin JL & approach for evaluating the plasma lipid effects of stearic acid. Am. J. Mendy F (1986b): , platelet function and composition in nine Clin. Nutr. 60(Suppl), 1029S ± 1036S. groups of French and British farmers. Atherosclerosis 60, 36 ± 38. Kritehevsky D (1994): Stearic acid metabolism and atherogenisis: history. Schoene NW (1997): Design criteria: tests used to assess platelet function. Am. J. Clin. Nutr 60(Suppl), 997S ± 1001S. Am. J. Clin. Nutr. 65(Suppl), 1665S ± 1668S Kwon JS, Snook JT, Wardlaw GM & Hwang DH (1991): Effects of diets Schoene NW, AlIman MA, Dougherty RM, Denvir E & Iacono JM (1992): high in saturated fatty acids, , or saf¯ower oil on platelet Diverse effects of dietary stearic and palmitic acids on platelet morphol- function, B2 formation and fatty acid composition of ogy. In Essential Fatty Acids and , ed. R Gibson, AJ Sinclair, platelet phospholipids. Am. J. Clin. Nutr. 54, 351 ± 358. pp 290 ± 292. Champaign, IL: AOCs. Laufer N, Grover NB, Ben-Sasson S & Freund H (1979): Effects of Schultheiss HF, Tschoepe D, Esser J, Schwippert B, Roesen P. Nieuwen- adenosine diphosphate, colchicine and temperature on size of human huis HK, Schmidt-Soltau C & Stauer B (1994): Larger platelets continue platelets. Thromb. Haemost. 41, 491 ± 497. to circulate in an activated state after myocardial infarction. Eur J. Clin. Litman RI & Mitchell DC (1996): A role for phospholipids in modulating Invest. 24, 243 ± 247. membrane protein function. Lipids 31, S193 ± 198. Sharp DS, Bath PMW, Martin JF, Beswick AD & Sweetnam PM (1994): Marckmann P, Sandstrom B & Jespersen J (1990): Effects of total fat Platelet and erythrocyte volume and count: epidemiological predictors of content and fatty acid composition in diet on factor VII coagulant impedance measured ADP-induced platelet aggregation in whole blood. activity and . Atherosclerosis 80, 227 ± 233. Platelets 5, 252 ± 257 Marckmann P, Sandstrom B & Jespersen J (1993): Favourable long-term Sinclair AJ, O'Dea K, Dunstan G, Ireland PD & Niall M (1987): Effect on effect of a low-fat/high ®bre diet on human blood coagulation and plasma lipids and fatty acid composition of very low fat diets enriched ®brinolysis. Arterioscler. Thromb. 13, 505 ± 511. with ®sh or kangaroo meat. Lipids 22, 523 ± 529. Marckmann P, Sandstrom B & Jespersen J (1994): Low fat-fat, high ®bre Small DM (1991): The effect of glyceride structure on absorption and diet favourably affects several independent risk markers of ischaemic metabolism. Ann. Rev. Nutr. 11, 413 ± 434. heart disease: observations on blood lipids, coagulation, and ®brinolysis Tholstrup T, Marckmann P, Jespersen J & Sandstrom B (1994): Fat high in from a trial of middle-aged Danes. Am. J. Clin. Nutr. 59, 935 ± 939. stearic acid favourably affects blood lipids and factor VII coagulant Marcus AT & Sa®er LB (1993): Thromboregulation: multicellular modula- activity in comparison with fats high in palmitic acid or high in myristic tion of platelet reactivity in hemostasis and thrombosis. FASEB 7, 515 ± and lauric acids. Am. J. Clin. Nutr. 59, 371 ± 377. 522. Tholstrup T, Andreasen K & Sandstrom B (1996): Acute effect of high-fat Martin JF, Bath PMW & Burr ML (1991): In¯uence of platelet size on meals rich in either stearic or myristic acid on haemostatic factors in outcome after myocardial infarction. Lancet 338, 1409 ± 1411. healthy young men. Am. J. Clin. Nutr. 64, 168 ± 176. Meade TW, Chakrabarti R, Haines AP, North WRS, Stirling Y & Thomp- Turpeinen AM, Wubert J, Aro A, Lorenz R & Mutanen M (1998): Similar son SG (1980): Haemostatic function and cardiovascular death: early effects of diets rich in stearic acid or trans-fatty acids on platelet func- results of prospective study. Lancet i, 1051 ± 1053. tion and endothelial prostacyclin production in humans. Arterioscler. Meade TW, Brozovic M, Chakrabarti R, Haines AP, Imeson JD, Mellows Thromb. Vasc. Biol. 18, 316 ± 322. S, Miller GJ, North WRS, Stirling Y & Thompson SG (1986): Haemo- static function and ischaemic heart disease: principal results of the Northwick Park Heart Study. Lancet ii, 533 ± 537. Meade TW, Ruddock V, Stirling Y, Chakrabarti R & Miller GJ (1993): Fibrinolytic activity, clotting factors, and long-term incidence of ischaemic heart disease in the Northwick Park Heart Study. Lancet 342, 1076 ± 1079.

European Journal of Clinical Nutrition