Author + information
- Received August 2, 2000
- Revision received February 5, 2001
- Accepted February 15, 2001
- Published online June 1, 2001.
- Paul J Nestel, MDa,* (, )
- Hideki Shige, MDa,
- Sylvia Pomeroy, BSa,
- Marja Cehuna and
- Jaye Chin-Dusting, PhDa
- ↵*Reprint requests and correspondence: Dr. P. J. Nestel, Baker Medical Research Institute, P.O. Box 6492, St. Kilda Rd. Central, Melbourne, Australia
We sought to examine the effects of plasma lipids, especially in remnants after a fat meal, on systemic arterial compliance (SAC), a newly recognized cardiovascular risk factor.
Post-prandial remnants correlate with coronary heart disease events through mechanisms that may include vascular dysfunction, although the effect on SAC has not been studied.
Systemic arterial compliance was measured non-invasively over 6 h after a fat meal in 16 subjects with varying plasma triglyceride levels. Changes were related to rises in plasma lipids and remnant lipids. Systemic arterial compliance was measured in 20 subjects after a control low-fat meal.
The fat meal induced increments in plasma triglyceride and remnant cholesterol and triglyceride (respectively +54%, 50% and 290% at 3 h, analysis of variance <0.001). Systemic arterial compliance fell at 3 h and 6 h by 25% and 27% (analysis of variance <0.001). Baseline SAC correlated significantly with all lipid concentrations at 0, 3 h and 6 h, but only with triglyceride on stepwise regression analysis. The SAC response to the low-fat meal was very small and not significant.
This is the first demonstration of SAC becoming impaired after a fat meal. Remnant lipids and plasma total triglyceride appeared to contribute to the fall in SAC.
Hypertriglyceridemia is generally recognized to be an independent predictor of cardiovascular events, possibly more so in women than in men (1). The uncertainty that had delayed acceptance of such a relationship was due in part to the heterogeneity of triglyceride-rich lipoproteins (TRL), which leads to inhomogeneity in the risk for any given hypertriglyceridemic population (2). In prospective studies, the progression in the severity of arterial atherosclerotic lesions and the incidence of new events has been strongly associated with higher levels of remnants of TRL, such as intermediate density lipoproteins and small, partly catabolized very-low-density lipoproteins (VLDL) (3–5). With the development of better techniques for isolating and quantifying the TRL moiety that putatively confers the greatest risk, such as the cholesterol-enriched remnants of partly catabolized chylomicrons and VLDL, the link to coronary risk has become clearer.
Of several techniques, one that identifies remnants through immunoprecipitation of other lipoproteins (6)has recently provided compelling evidence. Kugiyama et al. (7)examined prospectively the occurrence of further clinical coronary events in 135 patients with established disease and observed that higher levels of circulating remnants independently predicted future events. Several cross-sectional comparisons of subjects with and without clinical coronary artery disease confirmed that TRL remnant concentrations were significantly higher in the patients with disease (8–10).
Of the several mechanisms through which elevated remnant levels may give rise to clinical disease, adverse effects on vascular functions may be particularly relevant (11–13). Hypercholesterolemia clearly impairs the capacity of resistance vessels and of muscular arteries to respond to vasodilatory agonists (14,15). Similar endothelial dysfunction has also been reported in hypertriglyceridemic subjects, but less consistently (16–18). Kugiyama et al. (13)showed that even in subjects with apparently normal coronary arteries, raised levels of remnants attenuated the vasodilatory response to acetylcholine infused into the coronary circulation. The same group of investigators found that brachial artery dilation in response to hyperemia was diminished at higher concentrations of remnants (12).
Less well-defined is the endothelium-dependent response to acute elevation of plasma triglyceride or of remnants after a fat meal. De Man et al. (19), who reported impaired dilation in the microcirculation of the forearm of hypertriglyceridemic individuals, failed to observe impairment after infusion of a synthetic triglyceride-rich emulsion. By contrast, Lundman et al. (20), who infused a similar emulsion into healthy people, did find transient impairment of flow-mediated dilation of the brachial artery. Vogel et al. (21)were the first to report endothelial dysfunction, as measured by flow-mediated dilation, over several hours after a fat meal. Since then, their findings have been confirmed in some (22,23)laboratories but not in others (24).
We have studied the effect of fat meals on another parameter of arterial function, systemic arterial compliance (SAC), which reflects predominantly the distensibility of large elastic arteries. In 16 subjects over a range of plasma triglyceride concentrations, we have related changes in SAC over 3 h and 6 h to the fasting and post-prandial concentrations of remnant lipids.
Thirty-two subjects were recruited for the study. Twenty participated in a low-fat control study that examined the fluctuations in SAC over the 6-h duration of the test study, which was fat-supplemented. Of those 20, four also took part in the high-fat study. Sixteen subjects, recruited partly from a lipid clinic, participated in the test study that examined the effects on SAC of a 50-g fat meal over 6 h.
The subjects tested with the high-fat meals were selected to provide a range of plasma triglyceride concentrations: four in the normal 1 to 2 mmol/l range, five in the 2 to 3 mmol/l range, five in the 3 to 5 mmol/l range and two in the 5 to 7 mmol/l range. Seven of the hypertriglyceridemic subjects (defined as >2 mmol/l) also had plasma cholesterol levels >5.5 mmol/l and were therefore classified as having combined hyperlipoproteinemia. Their mean ages, body weights and gender are shown in Table 1. None had been receiving lipid-lowering or antihypertensive medication, and they were asked to cease vitamin and other supplements for a month beforehand; only one woman was being treated with hormone replacement. None smoked. Before the day of the test, they drank no alcohol and did not undertake unusual exercise. They fasted and drank only water during the preceding 12 h. Table 1shows that the 20 subjects who ate the control low-fat meal were of similar age and weight but had significantly lower diastolic blood pressure than did the 16 test subjects. The control subjects, so termed in the tables because they consumed the control low-fat meals, included men and women whose plasma triglyceride concentrations ranged from 0.65 to 3.4 mmol/l with a mean of 2.33 ± 0.66 mmol/l. Fifteen of these subjects had fasting triglyceride values >2 mmol/l, and their mean and median plasma triglyceride concentrations were not significantly different from those in the group that ate the high-fat test meal.
The study involved a 6-h post-prandial determination of SAC and plasma lipids measured before, 3 h and 6 h after the test meals.
Sixteen subjects ate a mixed meal immediately after the first set of measurements and fat-free, low-energy fluids as requested. The meal consisted of a ham and cheese sandwich, buttered bread, a drink of whole milk, and ice cream. It provided about 670 kilocalories and 50 g fat (of which two thirds was saturated fat). The 20 subjects who ate the control low-fat meal obtained a similar energy value, mainly from carbohydrate and protein from breakfast cereal, bread and fruit and 6 g of fat from low-fat milk and spreads.
During the 6-h study period, the subjects were free to move around but did not undertake strenuous activity.
Measurement of plasma lipids
Samples were stored at −80°C and assayed in batches. Plasma total cholesterol and triglyceride measurements were carried out by standard enzymatic procedures; high-density lipoprotein cholesterol (HDL-C) was assayed after precipitating apolipoprotein B lipoproteins.
Remnant lipoprotein cholesterol and triglyceride concentrations were determined by the assay produced by JIMRO (Takasaki, Japan). Remnant lipoproteins are separated from serum on immuno-affinity gels that contain monoclonal antibodies to apolipoprotein A-1 and most apoliporotein B-100. Thus HDL, most chylomicrons (which contain apolipoprotein A-1), low-density lipoproteins (LDL) and most VLDL are adsorbed. The supernatant contains lipoproteins not recognized by the antibodies, including those apolipoprotein E, apolipoprotein B-48 and triglyceride-rich lipoproteins (TRL) that constitute remnants of chylomicrons. The method is fully described by Nakajima et al. (6), and the limitations of the assay, including minor contamination by nonremnant TRL, are discussed elsewhere (25). The triglyceride and cholesterol concentrations in the supernatant were measured in a Cobas autonalyzer (Roche, Basel, Switzerland). The apparatus for the assay is manufactured by Otsuka Pharmaceuticals (Rockville, Maryland). Standards and plasma samples containing the range of lipid concentrations in the test samples were assayed repeatedly to establish inter- and intra-assay variation. Interassay coefficients of variation were 5% for remnant cholesterol and 7% for remnant triglyceride. Linearity of the assay over three values was established with each batch of measurements.
Systemic arterial compliance (SAC)
Systemic arterial compliance is a function of aortic flow and pressure that defines the distensibility or, conversely, the stiffness of the large arterial system, predominantly that of the aorta. Flows and pressures are determined by ultrasound probes placed over the ascending aorta (suprasternal notch) and the right carotid artery, respectively. Ascending aortic flow rate is measured by a hand-held Doppler velocimeter (MD1, Multi-Doplex, Huntleigh Technology, Cardiff, United Kingdom) and aortic root driving pressure at the proximal part of the right carotid artery by applanation tonometry using a Millar Mikro-Tip pressure transducer (Millar Instruments, Houston, Texas). The area under the diastolic blood pressure decay curve and the calculation of SAC as described by Liu et al. (26)are computed by software designed by Dr. J. Cameron (Laboratory Software, Melbourne, Australia). Although the procedure is carried out manually, the computation of SAC is computerized, with observer input limited to selecting 10 satisfactory pressure and flow waves with a predetermined low standard error. The procedure is performed at least twice for each measurement of compliance to ensure optimal assays of flows and pressures. The entire procedure takes up to 20 min, and subjects were first made familiar with the technique on a previous day. The methodology is fully described elsewhere (27)and provides a parameter of compliance. We chose not to measure the aortic root diameter (by echocardiography), which provides a value for the absolute volume of systolic flow, because it was extremely unlikely that aortic root diameter would change over several hours. Aortic root diameter was found not to vary over 6 h following either a high-fat or low-fat meal. In five individuals, other than those in the study, aortic root diameter was similar before and after a high-fat and also a low-fat meal, the minor variation being repeatability of an echocardiographic procedure. The coefficient of variation on repeat measurement of SAC carried out in a large number of subjects is 9%.
Liang et al. (27), using identical methodology including the software, have estimated the number of subjects required to establish a 15% to 20% difference in SAC through an intervention. Because the number was between 11 and 18, we recruited 16 subjects to the study, anticipating on the evidence of Vogel et al. (21)that, if fat meals affect SAC (as they appear to affect flow-mediated dilation of the brachial artery), then a reduction of at least 15% might be found.
The distributions of the concentrations of remnant lipids and of total triglyceride were not normal, as is often the case with TRL. Statistical calculations were therefore carried out by repeated measures of analysis of variance on ranks; if significant, pairwise multiple comparisons were carried out by the method of Dutton.
Systemic arterial compliance was subjected to the same analysis. Results in the tables therefore show medians with the 25% and 75% values, as well as means ± standard deviations. A ttest or Mann-Whitney Rank Sum Test was used to compare values between the two groups. Because four subjects participated in both meal studies, membership of the group was added as covariant.
Plasma triglyceride levels did rise slightly but not significantly in the 20 subjects who ate only 6 g fat in their meal. Among them, plasma total triglyceride concentrations rose from 2.33 ± 0.66 to 2.55 ± 1.20 at 3 h and to 2.44 ± 0.91 at 6 h, changes that were not significant. Remnant lipoproteins were therefore not measured in this group.
In the 16 subjects who ate the 50-g fat meals, plasma triglycerides, remnant cholesterol and remnant triglyceride rose after 3 h and 6 h, with one exception in whom the 6 h values were below the fasting levels (Table 2). The increments at 3 h and 6 h compared with baseline were significant for all three lipids (repeated measures analysis of variance on ranks p < 0.001, subsequent pairwise multiple comparisons p < 0.05). Neither the mean nor median values at 3 h and 6 h were significantly different; the concentrations of all three lipids failed to rise consistently further between 3 h and 6 h.
Plasma total cholesterol, LDL cholesterol and HDL cholesterol were not on average altered by the consumption of fat.
Systemic arterial compliance (SAC)
Basal SAC values were not statistically significant between the groups who ate the low-fat and the high-fat meals. In the 20 subjects who ate the control low-fat meal (6 g fat), mean SAC did not change, the mean values at 0 h, 3 h and 6 h being 0.220 ± 0.060, 0.232 ± 0.059 and 0.226 ± 0.057 units, respectively (Fig. 1).
By contrast, in the 16 subjects who ate the 50-g fat meal, there was an average fall in SAC at 3 h (25%) and 6 h (27%), both being significant (Table 3, Fig. 1). The respective values for SAC at 0 h, 3 h and 6 h were 0.189 ± 0.066, 0.148 ± 0.043 and 0.131 ± 0.051 units, respectively. The falls at 3 h and 6 h, compared with baseline, were statistically significant by repeated measures analysis of variance on ranks (p < 0.001) followed by pairwise multiple comparisons (p < 0.05 for both baseline vs. 3 h and baseline vs. 6 h). The difference between 3 h and 6 h was not significant, reflecting the similar trend in lipid values. Systemic arterial compliance values at 3 h and 6 h were significantly lower after the high-fat meal than after the low-fat meal (p < 0.01 by analysis of variance and p < 0.05 by subsequent paired comparison). The significance of the differences between the high-fat meal subjects and the low-fat meal subjects was not altered when the common membership of four individuals was added as a covariate.
Arterial pressures and heart rates did not change significantly during the 6 h: respective 0 h and 6 h values were 121 ± 27 and 117 ± 23 for systolic pressures; 78 ± 14 and 75 ± 13 diastolic (mm Hg); 65 ± 8 and 62 ± 7 beats per min for heart rate.
Among the 16 subjects who ate high-fat meals, univariate analysis showed that baseline fasting concentrations of plasma triglycerides, remnant triglyceride and remnant cholesterol were significantly and inversely related to baseline SAC (r = −0.61, −0.61 and −0.58 respectively; p < 0.01 for each). The initial SAC values maintained these inverse significant correlations with the three lipid parameters at 3 h and 6 h (r = −0.51 to −0.66 for the six correlations; p < 0.05 for each). The final 6-h SAC values correlated significantly and inversely with both 3-h remnant lipids (r = −0.49 for cholesterol and −0.52 for triglyceride; p < 0.05), as well as with the plasma triglyceride concentration over the 6-h period (r = −0.56; p < 0.05). Cholesterol in LDL and HDL did not correlate with SAC.
Stepwise regression analysis revealed that plasma triglyceride was the best predictor of basal SAC, but there were high degrees of intercorrelation among the three lipids (data not shown).
Changes in systemic arterial compliance
The present study is the first to demonstrate an adverse effect of a fat meal on SAC, which fell by 25% after 3 h and by 27% after 6 h. Furthermore, SAC in the fasting state was significantly and inversely correlated with plasma triglyceride and remnant lipid concentrations. The lower 6-h SAC, as well as the change in SAC over the 6-h study period, also correlated with plasma triglyceride and remnant lipids at baseline and at 3 h and with the 6-h area under the triglyceride curve. However, on stepwise regression analysis, plasma triglyceride emerged as the strongest correlate with baseline and 6-h SAC. On the other hand, neither LDL cholesterol nor HDL cholesterol was related to SAC. Remnant lipoproteins are present in fasting plasma even in normolipidemic individuals and rise in concentration in hypertriglyceridemia (28). The present observations clearly reflect the operation of post-prandial events that influenced SAC even after 12 h to 14 h of fasting. The clinical relevance of the findings pertains to the significance of impaired arterial compliance and diminished remnant clearance.
Systemic arterial compliance as measured in this study represents the elasticity or, conversely, the stiffness of large arteries such as the aorta. The walls of such arteries are complex, comprising elastic and collagenous matrix, muscular layers and endothelium. The predominant factors that determine compliance in such a vessel are also complex and less well-established than is the endothelial component in conduit, largely muscular arteries or in the microcirculation. It is likely that SAC in the current study involves modification in the tonicity of the smooth muscle cell-containing layer, the elastic response to exerted load, and the responsiveness of the endothelium (29). Endothelial dysfunction can be reversed rapidly with a single LDL apheresis (30), and rapid changes can also be induced in muscular conduit arteries. It was therefore not unrealistic to anticipate changes in SAC over several hours.
Diminished compliance or increased arterial stiffness is closely related to age, blood pressure (31)and diabetes (32). In general, LDL cholesterol has not been found to influence SAC, which we have recently confirmed to be the case even with substantial reduction in LDL (33).
The strong association of diminished SAC with rising remnant levels, although not previously reported, resembles the observations made by Vogel (21)and by Plotnick et al. (34), that flow-mediated dilation of the brachial artery becomes significantly impaired over several hours after eating fat. Because taking vitamin E with the meal appeared to overcome the effect of fat (34), the possibility of oxidized fatty acids impairing endothelial function has been raised (11,12). Consonant with that possibility is the observation by Doi et al. (35)that incubating endothelial cells with remnant lipoproteins increases intracellular oxidant levels that may interfere with nitric oxide mediated functions. Furthermore, Williams et al. (22)had demonstrated impaired endothelial function in people who had eaten oils that contained oxidized lipid from frequent usage.
Nevertheless, the issue of post-prandial impairment of flow-mediated dilation is not resolved. Apart from the original observations by Vogel et al. (21)and by Plotnick et al. (34), infusion of a synthetic emulsion of triglycerides has also led to impaired flow-mediated dilation (20), and Wilmink et al. (23)have also reported on the adverse effects of a fatty meal. Conversely, Raitakari et al. (24)showed an absence of changed flow-mediated dilation and increased forearm blood flow in response to hyperemia that was attributed to hemodynamic responses to eating. No obvious factors account for the inconsistent results, although it is being recognized that key biological regulators that may affect vascular function can be influenced rapidly after a fatty meal. Blanco-Colio et al. (36)have shown activation of mononuclear NF-κB several hours after a fat-rich breakfast.
Kugiyama et al. (12)showed that flow-mediated dilation was diminished in proportion to the circulating levels of remnant cholesterol, suggesting that the endothelial dysfunction induced by eating fat probably resulted from the accumulation of remnants of chylomicrons. Endothelial dysfunction in the coronary circulation has also been related to remnant cholesterol concentration (13).
Although our present data suggest that changes in plasma triglyceride levels predicted changing SAC at least as well as did remnant concentrations, this may be a consequence of the high intercorrelation between the three lipid parameters measured. It does not negate the likelihood that remnant lipids exert a biological effect on vasculature that is as important as that of triglyceride, especially because measures of total triglyceride include a highly heterogeneous group of lipoproteins. Several other reports deal with the effects of plasma total triglyceride levels on vascular function. Although they are not consistent, some studies indicate that endothelium-dependent dilation of resistant vessels is likely to be impaired in hypertriglyceridemic subjects (16–18). The relationship between endothelium-dependent vascular responses to raised triglyceride levels is clearly not as consistent as with hypercholesterolemia.
The 20 studies that were carried out with control meals contained about one-tenth the amount of fat in the test meals. Because SAC was not impaired, our observations suggest that the response was related to the amount of fat, although it is recognized that removing one nutrient leads, in an isocaloric situation, to an increase in another nutrient. In this context, it should be noted that there have been reports of improved endothelial function when carbohydrate replaced fat (37,38). Although eating leads to a variety of hemodynamic responses (39), these are more likely to lead to vasodilation than to factors that resulted in diminished SAC.
Significance of post-prandial remnants
Although post-prandial lipemia is closely correlated to the fasting plasma triglyceride concentration, it is now clear that delayed clearance of alimentary particles occurs in patients with coronary heart disease (CHD), independently of fasting hypertriglyceridemia (40). The post-prandial response after a fat meal also appears to predict future CHD events (41).
Raised concentrations of remnant lipoproteins are found in fasting samples of plasma in a variety of disorders that also cause delayed clearance of alimentary particles. Remnant cholesterol rises with age, after menopause in women, in subjects with diabetes (25)and in patients with chronic renal disease (42). Lower concentrations occur in healthy subjects than in those with established CHD, as recently reviewed by Cohn et al. (25).
The present technique of measuring remnant lipids is one of several introduced in recent years (25). Although the technique measures minor amounts of VLDL, most of the lipid is derived from partially catabolized chylomicrons and VLDL and therefore represents true remnants (6). The same assay has been used in several publications referred to earlier (7,12,13).
Clinical significance of diminished arterial compliance
Age-related effects on the microstructure of the walls of elastic arteries that give rise to arterial stiffness include fracture of elastic lamellae and accumulation of glycosaminoglycans (29). Arterial stiffness in diabetics is likely to have a similar etiology (32), whereas that associated with hypertension includes an adaptation of arterial wall material to distending pressure (43). The adverse prognostic outcomes on carotid stenosis, stroke and all-cause mortality from increasing pulse pressure, itself a result of arterial stiffening, have been documented in the Systolic Hypertension in the Elderly Program (SHEP) publications (44).
Limitations of the study
Whether episodic reductions in arterial compliance after the consumption of a fat meal, as reported in this study, may cumulatively lead to adaptive or structural changes of a permanent nature cannot be answered. The clinical importance of our finding rests on that possibility, but at least our findings reinforce current advice for a reduction in fat consumption.
We thank Dr. K. Nakajima and Dr. Tao Wang of Otsuka America Pharmaceutical Inc. and the National Heart Foundation of Australia for their support of this study.
☆ Supported in part by grants-in-aid from Otsuka America Pharmaceutical, Inc., and The National Heart Foundation of Australia.
- high-density lipoprotein
- low-density lipoprotein
- systemic arterial compliance
- triglyceride-rich lipoprotein
- very-low-density lipoprotein
- Received August 2, 2000.
- Revision received February 5, 2001.
- Accepted February 15, 2001.
- American College of Cardiology
- Austin M.A
- Nestel P.J
- Phillips N.R,
- Waters D,
- Havel R.J
- Hodis H.N
- Nakajima K,
- Okazaki M,
- Tanaka A,
- et al.
- Kugiyama K,
- Doi H,
- Takazoe K,
- et al.
- Reardon M.F,
- Nestel P.J,
- Craig I.H,
- Harper R.W
- Steiner G,
- Schwartz L,
- Shumak S,
- Poapst M
- Doi H,
- Kugiyama K,
- Ohgushi M,
- et al.
- Kugiyama K,
- Motoyama T,
- Doi H,
- et al.
- Yokoyama I,
- Ohtake T,
- Momomura S
- Lewis T,
- Dart A.M,
- Chin-Dusting J.P.F
- Chowienczyk P.J,
- Watts G.F,
- Wierzbicki A.S,
- et al.
- Schnell G.B,
- Robertson A,
- Houston D,
- et al.
- De Man F.H,
- Weverling-Rijnsburger A.W.E,
- van der Larse A,
- et al.
- Lundman P,
- Eriksson M.D,
- Schenck-Gustaffson K,
- et al.
- Williams M.J.A,
- Sutherland W.H.F,
- McCormick M.P,
- et al.
- Wilmink H.W,
- Banga J.D,
- Hijmering M,
- et al.
- Raitakari O.T,
- Lai N,
- Griffiths K,
- et al.
- Cohn J.S,
- Marcoux C,
- Davignon J
- Liu Z,
- Brin K.P,
- Yin F.C
- O’Rourke M,
- Kelly R.P,
- Avolio A.P
- Tamai O,
- Matsuoka H,
- Itabe H
- Devereux R.B,
- Roman M.J,
- Paranicas B.A
- ↵Shige H, Dart AM, Nestel PJ. Simvastatin improves arterial compliance in the lower limb but not in the aorta. Atherosclerosis 2001;155:245–50.
- Doi H,
- Kugiyama K,
- Oka H,
- et al.
- Blanco-Colio L.M,
- Valderrama M,
- Alvarez-Sala A,
- et al.
- Groot P.H.E,
- Von Stiphout W.A.H.J,
- Krauss X.M,
- et al.
- Weintraub M.S,
- Grosskopf I,
- Rassin T,
- et al.