点击显示 收起
1 From the Research Department of Human Nutrition and the Centre for Advanced Food Studies, The Royal Veterinary and Agricultural University, Frederiksberg, Denmark, and the Department of Biochemistry and Nutrition, Technical University of Denmark, Lyngby.
2 Supported by The Danish Research and Development Program for Food Technology through the LMC Centre for Advanced Food Studies. 3 Reprints not available. Address correspondence to T Tholstrup, Research Department of Human Nutrition, The Royal Veterinary and Agricultural University, 30 Rolighedsvej, DK-1958 Frederiksberg, Denmark. E-mail: tine.tholstrup{at}fhe.kvl.dk.
ABSTRACT
Background: There is increasing evidence that postprandial triacylglycerol-rich lipoproteins may be related to atherogenic risk.
Objective: The objective was to investigate the effect of individual fatty acid intakes on postprandial plasma lipoprotein triacylglycerol and cholesterol concentrations, plasma fatty acids, and preheparin lipoprotein lipase and cholesterol ester transfer protein (CETP) activities.
Design: Six test fats high (43% by wt) in stearic acid, palmitic acid, palmitic + myristic acid, oleic acid, elaidic acid (trans 18:1), and linoleic acid were produced by interesterification. After having fasted for 12 h, 16 healthy young men were served the individual test fats incorporated into meals (1 g fat/kg body wt) in random order on different days separated by washout periods. Blood samples were drawn before and 2, 4, 6, and 8 h after the meals.
Results: Different responses to the test-fat meals were observed for plasma lipoprotein triacylglycerol and cholesterol concentrations, plasma fatty acid concentrations, and lipoprotein lipase and CETP activities (diet x time interaction: 0.001 < P < 0.05). Intake of the long-chain saturated fatty acids stearic and palmitic acids resulted in a relatively lower lipemic response than did intake of the unsaturated fatty acids, probably because the saturated fatty acids were absorbed less and at a lower rate; therefore, the lipemic response took longer to return to postabsorptive values.
Conclusions: Fatty acid chain length and degree of saturation appear to affect the extent and duration of lipemia and affect hepatic output indirectly. These effects may not be mediated via effects on lipoprotein lipase and CETP activities.
Key Words: Postprandial lipemia • lipoprotein lipase • cholesterol ester transfer protein • CETP • stearic acid • oleic acid • trans fatty acids • monounsaturated fatty acids • MUFA • polyunsaturated fatty acids • PUFA • saturated fatty acids • SFA
INTRODUCTION
Dietary fatty acid intakes currently recommended to help prevent cardiovascular disease (CVD) are based mainly on the well-known effects of fatty acid type on fasting cholesterol concentrations in lipoproteins (1). There is, however, increasing evidence that an elevated concentration of triacylglycerol-rich lipoproteins (TRLs), especially in the postprandial state, may also be atherogenic (2–6). The magnitude of the postprandial response appears to play a role in the etiology and progression of CVD (2) and a slow return to the postabsorptive state may be specifically associated with an increased risk of CVD (2, 4).
It is well known that increased intakes of long-chain (LC) n-3 fatty acids decrease the incidence of postprandial lipemia (7–10). Postprandial studies have focused on comparisons of the effect of a single fatty meal (mostly cream) on the lipemic responses of CVD patients and healthy individuals (11–13). When natural fats or blends were compared, test fats could not be matched to contain an identical amount of other fatty acids besides the target fatty acid. In addition, the distribution of fatty acids in the 3 positions of the triacylglycerol molecule of test fats usually differed as well (6, 7, 14–18). Thus, knowledge of the postprandial effect of fatty acids frequently found in the Western diet is scarce.
A key factor in postprandial lipid metabolism is the activity of lipoprotein lipase, which plays a role in the clearance of chylomicrons derived from dietary fat (19). The quantity, degree of saturation, and chain length of fatty acids are suggested to affect lipoprotein lipase activity (20–23). The proposed mechanism for this effect is a fatty acid feedback system in which the accumulation of fatty acids obstructs lipoprotein lipase hydrolysis by disassociating the enzyme from its binding sites (24). The regulation of TRL hydrolysis by lipoprotein lipase is, however, not fully understood (25). It was suggested that adipose tissue is the site of a set of mechanisms involved in the regulation of lipoprotein lipase (26). Another compound that plays a role in lipid metabolism is cholesterol ester transfer protein (CETP), which exchanges cholesteryl esters and triacylglycerol between TRLs and HDLs and LDLs. CETP activity in the late postprandial phase may determine to what extent impaired clearance of plasma TRLs is atherogenic. An increase in CETP activity is suggested to be associated with the degree of lipemia (27), as seen in dyslipemic plasma (28). Fasting CETP activity was shown to be affected by dietary fatty acid composition (27, 29, 30). Postprandial studies indicated an increase in CETP activity after intake of fatty meals (31), the increase being higher after polyunsaturated fatty acid (PUFA) than after monounsaturated fatty acid (MUFA) intakes (32).
In the present study we wanted to determine whether dietary fatty acid quality (ie, dietary fatty acid composition) affects the postprandial lipid profile. We chose 6 fatty acids frequently found in the Western diet that have different cholesterolemic characteristics. To produce test fats that were as similar as possible in fatty acid composition, positional distribution of the fatty acid in the triacylglycerol molecule, and content of nonglyceride components (plant sterols), except for the test fat, we esterified commercially available pure triacylglycerol with the same batch of high–oleic acid sunflower oil.
SUBJECTS AND METHODS
Subjects
Sixteen healthy young men who ranged in age from 21 to 28 y (
The test meals were served in random order: 16 subjects received meals high in the test fats stearic, palmitic, oleic, and linolenic acids; 15 subjects received meals high in the test fat elaidic acid (trans 18:1); and 8 subjects received meals high in the test fat palmitic + myristic acid. [We originally intended to test 5 test fats randomly in 16 persons; however, because of speculation that the effect of palmitic acid is modified by that of myristic acid, we included a sixth test fat—palmitic + myristic acid. Because there was a problem with the delivery of trans 18:1 (for which reason 15 and not 16 subjects received meals high in trans fatty acids), we extended the study so that 8 of these subjects also received palmitic + myristic acid.] The different intervention periods were separated by a washout period of 3 wk, during which time the subjects consumed their habitual diets. The high-fat test meals were served in the morning, after the subjects had fasted for 12 h, and were eaten within 15 min of the serving time.
Diets
To minimize any effect of the diet eaten before the study days, we provided the subjects with food items for consumption on the 2 d before each experimental day. The fatty acid composition of this pretest diet was standardized to approximate the mean composition of the current Danish diet (33): 40% of total fat was saturated fatty acid (SFA), 41% was MUFA, and 19% was PUFA. The food consisted of margarine, bread, ready-made dinners, and cakes. The subjects were told 1) to refrain from eating high-fat products such as cheese, chips, ice cream, chocolate, and sausages on the 2 d before the experimental days; 2) to report all foods eaten, the amounts of all food eaten, and the duration of intakes on the 2 d before the experimental days; and 3) to standardize and report physical activities on the 3 d before the experimental days.
The meals were prepared and weighed as individual servings at the experimental kitchen of the Research Department of Human Nutrition, The Royal Veterinary and Agricultural University, Frederiksberg, Denmark. The meals consisted of mashed potatoes, in which the test fats were incorporated, and juice. The fat intake of each test meal was fixed at 1 g/kg body wt (range: 65–85 g). The energy content of the meals was 7 MJ for a person with a body weight of 75 kg. The test meals contained 50.6% of energy from fat, 43.0% from carbohydrate, and 6.4% from protein.
Six dietary fats dominated by stearic, palmitic, palmitic + myristic, oleic, trans 18:1, and linoleic acids were from Aarhus Olie (Oils and Fats Division, Aarhus, Denmark). The experimental fats were produced by interesterification of tristearin, tripalmitin, trimyristin (Hüls, Marl, Germany), high–oleic acid sunflower oil (TRISUN 80; SVO Enterprise, Eastlake, OH)—a fat rich in trans 18:1 (produced by Aarhus Olie by hydrogenation of high–oleic acid sunflower oil specifically for this study)—and high–linoleic acid sunflower oil (Aarhus Olie) with high–oleic acid sunflower oil. The target fatty acids made up 41–47% by wt. The aim was to keep the amount of nonglyceride constituents as low as possible by using triacylglycerol (tristearin, tripalmitin, and trimyristin). Furthermore, to balance the content of these components, we used the same batch of sunflower oil for interesterification. The fatty acid composition of the test fats is given in Table 1.
View this table:
TABLE 1. Major fatty acids in test fat triacylglycerols (TG) and 2-monoacylglycerol (2MAG)1
Blood sampling and analysis
The blood samples were taken in the morning on each test day after the subjects had fasted for 12 h and after 20 min of supine rest. Subjects were asked to refrain from alcohol intake and heavy physical activity for 24 h before the blood samples were taken. In addition, the subjects were told not to bike, run, or exercise in the morning on the test days so that they would be in a relaxed state at the time of blood sampling. Postprandial blood samples were taken 2, 4, 6, and 8 h after the meals began. Between blood samplings, the subjects pursued their usual study activities or walked leisurely within the Department. Blood for lipid and lipoprotein analyses was collected into tubes containing EDTA, which were immediately placed on ice and centrifuged at 3000 x g for 15 min at 4°C. Plasma for apolipoprotein and fatty acid analyses was stored at -80°C. Plasma for lipoprotein analysis was stored at 4°C and analyzed within 48 h. Chylomicrons [Svedberg flotation unit (Sf) > 400], VLDL + chylomicron remnants [density (d) < 1.006 kg/L], and LDL + HDL fractions (d > 1.006 kg/L) were separated by ultracentrifugation (100000 x g, 23 min, 20°C). Chylomicrons were separated according to the method of Dole and Hamlin (34) by carefully overlayering 3 mL plasma with 2.5 mL saline solution (d = 1.006 kg/L) in tubes (13 x 64 mm) and centrifuged at 100000 x g for 23 min at 20°C in an ultracentrifuge (L7-55; Beckman Instruments, Palo Alto, CA) with a fixed-angle rotor (96.1 mm, 50.4 Ti; Beckman Instruments). The top and bottom fractions were separated by slicing the tubes 45 mm from the bottom. The top fraction (Sf > 400) was transferred and adjusted to a total volume of 5 mL with saline solution. The bottom fraction (Sf < 400) was transferred to another tube, readjusted with saline solution to a d of 1.006 kg/L, and centrifuged at 170000 x g for 16 h at 4°C. After being tube-sliced, the top fraction (d < 1.006 kg/L, which contained VLDL and chylomicron remnants) and the bottom fraction (d > 1.006 kg/L, which contained LDL + HDL) were transferred to separate tubes and adjusted to a final volume of 5 mL. Cholesterol and triacylglycerol concentrations were assessed in plasma and various lipoprotein fractions by enzymatic procedures (Boehringer Mannheim GmbH, Mannheim, Germany) on a Cobas Mira analyzer (Roche, Basel, Switzerland). HDL- and HDL3-cholesterol concentrations were measured enzymatically after precipitation with polyethylene glycol (Quantolip; Immuno AG, Vienna) (35). HDL2 cholesterol was calculated as the difference between total HDL cholesterol and HDL3 cholesterol. LDL cholesterol was calculated as the difference in cholesterol in the infranate and HDL cholesterol. The concentrations of plasma apolipoproteins (apo) B and A-I were determined by immunoturbidimetry by using monospecific polyclonal antibodies against apo B and A-I (kit Unimate-3 apoA/apoB; Roche) in a Cobas Mira analyzer.
Blood for lipoprotein lipase analysis was collected in precooled tubes containing sodium heparin (29000 IU/L). The tubes were immediately placed in an ice bath and plasma was recovered for lipid analyses within 30 min by low-speed centrifugation (1750 x g, 20 min, 1°C). The plasma samples were frozen at -80°C within 1 h of blood sampling. At the time of analysis, the plasma samples were preincubated for 2 h on ice with 0.5 mL goat antibodies against hepatic lipase (delivered by the Department of Medical Biochemistry and Biophysics, Umeå University, Umeå, Sweden) to suppress hepatic lipase. Assay conditions for lipoprotein lipase activity were similar to the ones described elsewhere (34). We used [3H]oleic acid triolein emulsion (KABI3H-Intralipid; Pharmacia-Upjohn, Uppsala, Sweden) as substrate. To keep the blank at minimum, the labeled triolein was repurified by thin-layer chromatography according to the manufacturer's recommendations. This emulsion was mixed with an incubation medium containing 12% bovine albumin (SIGMA A 6003; Sigma-Aldrich, St Louis) and 0.02% (wt:vol) 193 U sodium heparin/mg (Novo 3091007; Novo Nordisk A/S, Bagsvaerd, Denmark) and inactivated rat serum containing apo C-II, an activator of lipoprotein lipase (provided by the Department of Experimental Medicine, Panum Institute, Copenhagen University). All determinations were done in triplicate.
We were aware that measurement of low lipase activities (without prior heparin injection) requires a sensitive assay system. All samples for any one individual were analyzed in the same batch on the same day. Preheparin lipoprotein lipase controls were included at the beginning and at the end of series together with several blanks. The intraassay CV was 5.1%. We used fasting pre- and postheparin plasma, which was drawn after injection of heparin (75 IU/kg body wt) on another occasion, as controls (intra- and interassay CVs: 4.3% and 7.9%, respectively) and a postheparin sample provided by the Department of Medical Biochemistry and Biophysics (Umeå University, Umeå, Sweden). Samples and blanks were counted in a liquid scintillation counter. Lipoprotein lipase activity is expressed in U/L, which corresponds to 1 nmol fatty acid released/min.
Plasma CETP activity was determined according to the method of Albers et al (36) as modified by Tato et al (37), which measures CETP activity as the percentage of total [3H]cholesterol ester transferred from HDL3 (donor lipoprotein) to LDL (acceptor lipoprotein) in the presence of a small volume of plasma. Donor and acceptor lipoproteins were obtained from fasting normolipemic volunteers and isolated and prepared as described previously (37). LDL was diluted to a final cholesterol concentration of 5 mmol/L and HDL3-containing [3H]cholesterol ester to a final cholesterol concentration of 1 mmol/L. For the assay, 50 µL [3H]HDL3 was mixed with 200 µL LDL and 20 µL test plasma was diluted 1:3 (by vol) with tris buffer. Blanks and quality controls were included in triplicate in each assay, whereas plasma samples were determined in duplicate. All assay samples were incubated for 16 h at 37°C and the reaction stopped by placing the tubes on ice for 15 min. Assay samples were adjusted to d = 1.063 kg/L and the HDL (donor) and LDL (acceptor) fractions were separated by ultracentrifugation at 245000 x g for 18 h at 4°C; 1 mL of the HDL fraction was counted for 20 min in a liquid scintillation counter. The interassay CV was 13.8%. The plasma concentration of C-reactive protein (analyzed by an immunoturbidimetric method) was determined to rule out the presence of any infectious diseases in the subjects at the time of blood collection. Values were in the normal range (<5 mg/L).
Statistical analysis
Repeated-measures analysis of variance (SPSS Inc, Chicago) with Huynh-Feldt adjustment of df was used to assess the effect of time, differences in the effect of the experimental fats, and the interaction between effects of time and type of fat during the 0–8-h period of the day. A significant interaction between the effects of time and type of fat means that the mean difference between the 2 fats varied with time.
RESULTS
Plasma total, chylomicron-, VLDL-, and HDL-triacylglycerol concentrations peaked 4 h after intake of all test fats and there were no significant differences between values at 4 h (Figure 1). LDL-triacylglycerol concentrations increased from 2 to 8 h after intake of all test fats except stearic acid, which peaked at 6 h. Fatty acids decreased initially from 0 to 2 h after all test fats were consumed and increased thereafter up to 8 h. The test fats containing the LCSFAs stearic and palmitic acids generally resulted in less of an increase in plasma total and chylomicron triacylglycerols after 4 h than did the other test fats and the return to postabsorptive values was slower. The differences in the response of fatty acids to the test fats mirrored to some extent the differences in chylomicron triacylglycerols: fatty acid concentrations tended to be lower between 2 and 6 h after stearic acid intake than after intake of the other test fats, followed by an increase between 6 and 8 h after intake of both stearic and palmitic acids. However, compared with changes in chylomicron triacylglycerol, changes in postprandial plasma fatty acids were small.
FIGURE 1. . Mean plasma total, chylomicron (chylo), VLDL, LDL, and HDL triacylglycerol (TG) and fatty acid (FA) concentrations before (0 h; fasting) and 2, 4, 6, and 8 h after intake of the test meals (1 g fat/kg body wt) high in palmitic + myristic acid (M, ; n = 8), palmitic acid (P, ; n = 16), stearic acid (S, *; n = 16), oleic acid (O, •; n = 16), trans 18:1 (T, no symbol; n = 15), and linoleic acid (L, ; n = 16). Bars depict mean (±SEM) values 4 h after intake of the test meals. Note that the y axes differ between panels.
Pairwise comparisons with significantly different responses to the test fats (diet x time interaction) are shown in Table 2. (Changes between baseline and values at the different time points were analyzed for triacylglycerol in lipoprotein fractions, whereas absolute values 2, 4, 6, and 8 h after the meals began were analyzed for plasma fatty acids.) In general, the effect of the LCSFAs differed from that of the other test fats; the response to stearic acid was significantly different from that of all other test fats, except palmitic acid, and the response to palmitic acid differed significantly from that of trans 18:1 and linoleic acid. Intake of cis and trans MUFAs did not affect postprandial lipemia differently. The response of chylomicron and VLDL triacylglycerols paralleled that of total triacylglycerols (Table 3), whereas there was no association between TRLs and fatty acids and lipoprotein lipase. There was a weak association between lipoprotein lipase and fatty acids.
View this table:
TABLE 2. Statistical comparison between test fats with regard to postprandial responses in plasma triacylglycerols (TG) in lipoprotein fractions and in fatty acids (FA) and preheparin lipoprotein lipase (LPL) activity1
View this table:
TABLE 3. Correlation matrix of postprandial lipemia, fatty acids (FA), preheparin lipoprotein lipase (LPL) activity, cholesterol in lipoprotein fractions, and cholesterol ester transfer protein (CETP) activity in healthy young men 4 h after intake of fatty meals1
There was an overall increase in chylomicron and VLDL cholesterol after intake of all test fats, which peaked at 4 h (Figure 2). LDL, HDL, and HDL3 cholesterol decreased initially after all test fats; the lowest values were at 4 h, after which time concentrations began to increase. Pairwise comparisons of responses to the test fats that were significantly different (diet x time interaction: changes between baseline and values at the different time points) are shown in Table 4. The lipid with the most significant differences between responses to the different test fats was VLDL cholesterol. Specifically, the response to the LCSFAs stearic and palmitic acids differed significantly from that to trans 18:1 and linoleic acid. Although there were fewer significant differences in VLDL cholesterol than in TRLs, the effect of dietary fatty acids on VLDL cholesterol mirrored the effect on VLDL triacylglycerol, which was confirmed by the strong correlation coefficient (Table 3).
FIGURE 2. . Mean plasma chylomicron, VLDL, LDL, HDL, HDL2, and HDL3 cholesterol (C) before (0 h; fasting) and 2, 4, 6, and 8 h after intake of the test meals (1 g fat/kg body wt) high in palmitic + myristic acid (M, ; n = 8), palmitic acid (P, ; n = 16), stearic acid (S, *; n = 16), oleic acid (O, •; n = 16), trans 18:1 (T, no symbol; n = 15), and linoleic acid (L, ; n = 16). Bars depict mean (±SEM) values 4 h after intake of the test meals. Note that the y axes differ between panels.
View this table:
TABLE 4. Statistical comparison between test fats with regard to postprandial responses to test fats in plasma cholesterol in lipoprotein fractions and cholesterol ester transfer protein (CETP) activity1
Concentrations of apo B increased after intake of palmitic acid, peaking at 4 h. However, concentrations decreased between 0 and 6 h after intakes of palmitic + myristic acid, trans 18:1, and linoleic acid, followed by an increase between 6 and 8 h (Figure 3).The only significant differences were between the responses to palmitic acid and trans 18:1 and between the responses to palmitic and palmitic + myristic acids (diet x time interaction). No significant differences were observed for apo A-I.
FIGURE 3. . Mean apolipoprotein (apo) B and apo A-I before (0 h; fasting) and 2, 4, 6, and 8 h after intake of the test meals (1 g fat/kg body wt) high in palmitic + myristic acid (M, ; n = 8), palmitic acid (P, ; n = 16), stearic acid (S, *; n = 16), oleic acid (O, •; n = 16), trans 18:1 (T, no symbol; n = 15), and linoleic acid (L, ; n = 16). Bars depict mean (±SEM) values 4 h after intake of the test meals. Note that the y axes differ between panels.
CETP activity decreased between 0 and 4 h after intakes of stearic and linoleic acids and increased between 4 and 6 h after intakes of all test fats, except oleic acid; intake of oleic acid resulted in virtually no changes in CETP activity over the test period (Figure 4). Pairwise comparisons of the effects of the test fats on CETP activity (ie, differences between baseline and the 2-, 4-, 6-, and 8-h time points) indicated significant differences between the effects of oleic acid and those of stearic acid, palmitic + myristic acid, and trans 18:1 and between the effects of palmitic + myristic and trans 18:1 (Table 4). There was a strong correlation between CETP activities at baseline and those 4 h after intake of the test meals (r = 0.86, P < 0.001; data not shown), but no correlation between changes in CETP activities at baseline and those between 4 and 6 h (data not shown). Four hours after intake of the fatty meals, CETP activities correlated negatively with HDL cholesterol and positively with VLDL cholesterol, LDL triacylglycerol, and LDL cholesterol (Table 3). As a result of CETP activity, postprandial HDL-triacylglycerol and VLDL-cholesterol concentrations correlated positively with postprandial triacylglycerol concentrations (Table 3).
FIGURE 4. . Mean changes in cholesteryl ester transfer protein (CETP) activity as the percentage of total [3H]cholesterol ester transferred from HDL3 (donor lipoprotein) to LDL after intake of the test meals (1 g fat/kg body wt) high in palmitic + myristic acid (M, ; n = 8), palmitic acid (P, ; n = 16), stearic acid (S, *; n = 16), oleic acid (O, •; n = 16), trans 18:1 (T, no symbol; n = 15), and linoleic acid (L, ; n = 16). Bars depict mean (±SEM) values 4 h after intake of the test meals.
Lipoprotein lipase activities decreased between 0 and 2 h after intakes of stearic, linoleic, and oleic acids and increased between 0 and 2 h after intakes of palmitic acid, trans 18:1, and palmitic + myristic acid (Figure 5). Activities increased between 2 and 4 h after intakes of all test fats. Between 6 and 8 h, lipoprotein lipase activity decreased after intakes of palmitic + myristic acid, oleic acid, trans 18:1, and linoleic acid, whereas activities increased slightly after intakes of palmitic and stearic acids. The responses to stearic and palmitic acids were significantly different from those to both trans 18:1 and linoleic acid, and the responses to oleic acid were significantly different from those to trans 18:1 (diet x time interaction); absolute values 2, 4, 6, and 8 h after intake of the meals are compared in Table 2. No other significant differences between test fats were observed.
FIGURE 5. . Mean preheparin lipoprotein lipase (LPL) activity before (0 h; fasting) and 2, 4, 6, and 8 h after intake of the test meals (1 g fat/kg body wt) high in palmitic + myristic acid (M, ; n = 8), palmitic acid (P, ; n = 16), stearic acid (S, *; n = 16), oleic acid (O, •; n = 16), trans 18:1 (T, no symbol; n = 15), and linoleic acid (L, ; n = 16). Bars depict mean (±SEM) values 4 h after intake of the test meals.
DISCUSSION
Consumption of the high-fat test meals resulted, as expected, in pronounced postprandial lipemia. The response of chylomicron and VLDL triacylglycerols paralleled that of total triacylglycerols; maximum increases occurred 4 h after intake of the fatty meals, a finding that was also observed by others (5, 38, 39). The tendency of a relatively higher degree of lipemia 4 h after consumption of MUFAs (17) and PUFAs than after LCSFAs was observed previously (40, 41). We observed a higher increase in TRLs after consumption of the test fat containing myristic acid than after those containing LCSFAs. This finding agrees with the results of other studies, which showed that butter fat (which contains short-chain and medium-chain SFAs), but not fats containing LCSFAs, resulted in higher postprandial lipemic responses than did consumption of MUFAs and PUFAs (14, 18). A slower return to postabsorptive values after intake of the LCSFAs than of PUFAs was also shown previously (14).
The markedly lower lipemia after stearic acid intake, and to some extent after palmitic acid intake, than after intakes of the other test fats was probably due to either slower absorption or to increased lipolysis. Changes in plasma lipoprotein lipase activity and plasma fatty acid concentrations do not explain the increased lipolysis. Thus, low postprandial lipemia is more likely to reflect slower or less-efficient absorption of LCSFAs, as was shown by others (17, 42) and supported by our analyses of the fatty acid composition of chylomicrons (data not shown).
The similar postprandial responses of VLDL and chylomicron cholesterol and VLDL triacylglycerol to the test fats was also observed by others (43, 44). The increase in LDL cholesterol between 4 and 8 h after intakes of the test fats was probably due to short-term regulation of hepatic receptor activity (17). The decrease in HDL cholesterol resulted predominantly from a decrease in the HDL3 fraction, the concentration of which was lowest 4 h after intake of the fatty meals, which agrees with earlier findings (6, 17, 40), suggesting stimulation of CETP activity (45).
We measured baseline and postprandial preheparin lipoprotein lipase activity (without prior injection of heparin) because preheparin lipoprotein lipase activity is suggested to be a better marker of lipoprotein lipase activity in the lipemic phase than is postheparin lipoprotein lipase activity (46). In addition, injected heparin is known to interfere with lipid metabolism. The general increase in lipoprotein lipase activity between 2 and 4 h and visual parallelism between lipoprotein lipase activity and plasma triacylglycerol is consistent with the findings of others (47). The positive association between the increase in lipoprotein lipase activity and of plasma fatty acids 4 h after intake of the meal agrees with previous findings (47, 48) and indicates a fatty acid feedback control of lipoprotein lipase, as suggested by others (24, 48). Despite the low preheparin lipoprotein lipase activity observed in the present study, we observed some differences in the response to the test fats. Our lipoprotein lipase results should be interpreted with caution because the measured in vitro lipoprotein lipase activity may not entirely mirror the interaction of chylomicrons (which have different fatty acid contents) with vessel wall components. However, consumption of the test fats high in the LCSFAs stearic and palmitic acids resulted in less of an increase in lipoprotein lipase activity during the day than did consumption of the other test fats. In contrast, the other test fats resulted in a more pronounced increase between 2 and 4 h and in most cases a decrease between 6 and 8 h. The more moderate increase in lipoprotein lipase activity after consumption of SFAs than after consumption of unsaturated fatty acids agrees with results for postheparin lipoprotein lipase in animal studies (49, 50), in adipose tissue (51, 52), and in vitro studies (21, 22), but not in one study in humans (53). Taken together, intake of fatty meals with a markedly different fatty acid composition only caused small differences in preheparin lipoprotein lipase activity. Because of the corresponding pattern of lipoprotein lipase activity with plasma fatty acids, as discussed above, we suggest that these differences in lipoprotein lipase activity mainly reflect the total plasma triacylglycerol concentration of TRLs, which was reported to be an important determinant of lipoprotein lipase activity (20).
The postprandial increase in CETP activity agrees with observations by others (54, 55). The strong correlation between total plasma triacylglycerol and HDL triacylglycerol and VLDL cholesterol agrees with the fact that when TRLs are present in excess in plasma, as after the consumption of fatty meals, CETP promotes the triacylglycerol enrichment of HDL with concomitant cholesteryl ester accumulation in VLDLs (56, 57). In addition, CETP activity correlated negatively, as expected, with HDL cholesterol and positively with VLDL and LDL triacylglycerol and LDL cholesterol, as shown by others (54, 58). As mentioned above, we observed some differences in the response of CETP activity to the individual test fats. Although other studies showed an effect of dietary fatty acid consumption on CETP activity in fasting plasma (30, 59–63), postprandial effects have rarely been investigated. A single study reported that the greater extent of lipemia after PUFA than after SFA consumption caused a greater response in CETP activity (32). This finding disagrees with the finding of another study (64), which showed no significant differences in the response of CETP activity to SFA and PUFA consumption. The finding that CETP activity did not increase after oleic acid intake agrees with the results from other studies in which consumption of a high-MUFA diet resulted in relatively low CETP activity in the fasting state (29). The observed difference between the effects of oleic acid and trans 18:1 suggests that cis-trans isomerization may affect CETP activity. Because the test fat high in oleic acid did not result in an increased transfer of cholesterol from HDL to LDL, we speculate that this is the reason why oleic acid intake did not result in a decrease in HDL cholesterol. The somewhat lower CETP activity after intake of stearic acid than after trans 18:1 agrees with the results of the in vitro study mentioned previously (59).
The observed lipemic peak was considerably less after intake of LCSFAs than after intake of the other fatty acids, including trans fatty acids, which did not result in a lipemic response pattern different from that elicited by oleic acid. Low postprandial lipemia, besides being less procoagulant (65–67), may result in a lower production of atherogenic TRLs (68). In comparison with the effects of the intake of unsaturated fatty acids, the intake of LCSFAs resulted in a slightly slower return to the postabsorptive state, which could be associated with an increased risk of CVD (4). However, because the slightly slower return to the postabsorptive state after intake of LCSFAs than after intake of unsaturated fatty acids did not result in a higher CETP activity, the intake of LCSFAs was probably more beneficial postprandially than was the intake of MUFAs and PUFAs by healthy young men. More subtle differences between responses to test meals may have been uncovered in elderly or hypertriglyceridemic persons.
In conclusion, dietary fatty acid composition influenced postprandial lipemia. LCSFAs caused a relatively lower lipemia and a later return to the postabsorptive state than did the other test fats, probably because LCSFAs comparatively are absorbed less and at a lower rate gastrointestinally. Fatty acid chain length and degree of saturation may determine the amount of fat absorbed, affecting the extent and duration of lipemia, which may affect hepatic VLDL output indirectly. These effects seem to not be mediated via effects on lipoprotein lipase and CETP activities.
ACKNOWLEDGMENTS
We thank Gunilla Olivecrona (Department of Medical Biochemistry and Biophysics, Umeå University, Umeå, Sweden) for assistance with the lipoprotein lipase measurements and for the generous provision of the substrate and antibodies; Gloria Vega (Southwestern Medical Center, University of Dallas) for assistance with the analysis of CETP; our technicians, especially Karen Rasmussen, for outstanding contributions; and dietitians Hanne Jensen and Berit Christensen and other staff of the metabolic kitchen.
REFERENCES