OUP user menu

Pronounced postprandial lipemia impairs endothelium-dependent dilation of the brachial artery in men

Hannes Gaenzer , Wolfgang Sturm , Guenther Neumayr , Rudolf Kirchmair , Christoph Ebenbichler , Andreas Ritsch , Bernhard Föger , Günter Weiss , Josef R Patsch
DOI: http://dx.doi.org/10.1016/S0008-6363(01)00427-8 509-516 First published online: 1 December 2001


Objective: Pronounced postprandial lipemia has been established as a risk factor for cardiovascular disease, but reports regarding its effect on endothelial function have been controversial. In the present study the influence of a standardized fatty meal with its ensuing postprandial lipemia of highly varying magnitude on endothelium-dependent dilation (EDD) was investigated. Methods: In 17 healthy, normolipidemic men EDD of the brachial artery was quantified in two series of three measurements each. In both series initial measurements were performed at 08:00 h after an overnight fast followed by measurements at 12:00 and 16:00 h, in the first series with continued fasting and in the second following the ingestion of a standardized fatty test meal 4 and 8 h postprandially. Results: Measurements of EDD in the fasting state revealed the recently appreciated diurnal variation with higher values in noon and afternoon hours compared with morning values (2.5±1.6% at 08:00, 7.5±2.7% at 12:00, and 7.0±2.1% at 16:00 h, P<0.001 by analysis of variance). Postprandial EDD values measured at 12:00 h were, at the average, lower than fasting EDD values measured at 12:00 h and correlated inversely with the magnitude of postprandial triglyceridemia (r=−0.81, P<0.001). In multivariate analysis, higher postprandial lipemia was associated with impaired postprandial EDD (P<0.001) independent of fasting triglycerides, low density lipoprotein (LDL) and high density lipoprotein (HDL) cholesterol, insulin, age and body mass index. Conclusion: We conclude that pronounced postprandial lipemia is associated with transient impairment of endothelial function. Our findings support the notion that impaired triglyceride metabolic capacity plays an important role in atherogenesis.

  • Arteries
  • Endothelial function
  • Lipid metabolism
  • Ultrasound
  • Vasoconstriction/dilation

Time for primary review 30 days.

1 Introduction

Fasting triglycerides, representing triglyceride metabolism under relaxation, have not generally been accepted as an independent risk factor for atherosclerosis including coronary artery disease (CAD) [1–3]. Individuals with normal fasting triglyceride levels exhibit highly varying postprandial triglyceride concentrations in the postprandial hours of a fatty test meal [4]. Therefore, postprandial lipemia (PPL), representing triglyceride metabolic capacity under challenge, is considered to be more informative for assessing the role of triglyceride metabolism in the development of atherosclerosis [5,6]. Consequently, a number of case–control studies showed impaired triglyceride metabolic capacity, defined as increased and prolonged postprandial hypertriglyceridemia, to be closely linked to the presence of CAD [6,7].

Endothelium-dependent dilation (EDD) of the brachial artery and intima-media thickness (IMT) of the common carotid artery represent surrogate functional and structural markers for early atherosclerotic disease. In studies of middle-aged healthy subjects both impairment of EDD [8–10] and increase of IMT [11] was strongly related to the extent of postprandial hypertriglyceridemia. However, other reports observed no correlation between endothelial dysfunction and magnitude of PPL [12–15]. Therefore, the object of our study was to test the hypothesis whether a transient endothelial dysfunction induced by a standardized fat load is dependent on the magnitude of lipemia and to uncover whether a threshold value of postprandial triglyceridemia exists which enables to distinguish between normal and impaired triglyceride metabolic capacity.

Recently, we [16] and another study group [17] found evidence for a diurnal rhythm of EDD with lower levels in the morning and higher ones during noon and afternoon. When evaluating endothelial function the influence of the diurnal rhythm on EDD should be considered for the full 8-h period of postprandial lipemia [4] which could explain some discrepancies of previous studies [12]. Considering this diurnal variation of EDD, we compared the effects of a standardized fatty meal and its ensuing PPL on endothelial function in healthy men with the vascular responses obtained in the fasting state.

2 Methods

2.1 Subjects

Seventeen healthy, normolipidemic, non-smoking men aged 29–43 years, participated in the study. All subjects were free of diabetes, thyroid disease, dyslipemia, hypertension or family history of premature vascular disease, and none was taking cardiovascular medication or antioxidant agents. The subjects were recruited from University staff members after giving written informed consent. The investigation conformed with the principles outlined in the Declaration of Helsinki.

2.2 Study protocol

For each subject vascular reactivity of the brachial artery was determined in two series of studies. In the first series in which the study subjects had to continue abstaining from food intake for additional 8 h after an overnight fast, vascular measurements were taken at 08:00, 12:00 and 16:00 h. One week later, a second series of measurements was performed. Here, the first measurement was carried out at 08:00 h also following an overnight fast of 12 h. Immediately thereafter, a standardized oral fat load [4] was ingested, and vascular measurements repeated 4 and 8 h after the oral fat load at 12:00 and 16:00 h, respectively. Plasma levels of triglycerides, cholesterol, LDL cholesterol, HDL cholesterol, insulin and glucose were determined at each time point when vascular measurements were performed.

2.3 Brachial artery study

After having taken medical history and having measured supine resting pulse and blood pressure the ultrasound procedures for assessing endothelium-dependent and endothelium-independent arterial dilation were performed as described by others [18]. The brachial artery diameter was measured on B-mode ultrasound images by the use of a 13.0-MHz linear-array transducer and a standard Acuson Sequoia 512 system (Acuson, Mountain View, CA). In all studies, scans were obtained with the subject at rest, during reactive hyperemia and after sublingual administration of 400 μg glyceryl trinitrate. The subjects lay quietly for at least 10 min before the first scan was obtained in a temperature-controlled room (22°C). The brachial artery was scanned in a longitudinal section 2–15 cm above the elbow.

When a satisfactory transducer position was found the skin was marked, and the arm kept in the same position throughout the study. A resting scan was obtained, and the velocity of arterial flow was measured with a pulsed-Doppler signal at a 70° angle to the vessel. Increased flow was then induced by the inflation of a pneumatic torniquet placed around the forearm (distal to the scanned part of the brachial artery) to a pressure of 250 mmHg for 4.5 min, followed by pressure release. A second scan was performed at 45–90 s after deflation of the cuff, including a repeated recording of flow velocity for the first 15 s after pressure release. Thereafter, a period of 10 min was allowed for recovery of the vessel, after which an additional resting scan was performed. Then glyceryl trinitrate (400 μg) was sublingually administered to induce endothelium-independent dilation. A final scan was performed 3–4 min later.

Photographic images of five consecutive cardiac cycles were taken R-wave triggered at end-diastole. The diameter of the vessel was measured in a blinded manner with respect to daytime of examination and magnitude of postprandial lipemia by the same investigator to avoid interobserver variability. Reproducibility of EDD measurements performed at 08:00 h in both series, showed a coefficient of variation <3%. EDD and glyceryl trinitrate-induced vasodilation were determined as the percentage change in diameter relative to the mean value of the corresponding five baseline measurements. Volume flow was calculated by multiplying the velocity–time integral of the Doppler flow signal (corrected for angle) by the heart rate and the cross-sectional area of the vessel (πr2). Reactive hyperemia was calculated as the maximal flow in the first 15 s after cuff deflation divided by baseline flow.

2.4 Oral fat load

In the second series a standardized liquid fatty meal described and used in all our studies on postprandial lipemia [4,6,19] was given. Briefly, the test meal contained per square meter body surface 730 kcal and 65 g of fat from heavy whipping cream (41 g saturated fatty acids, 20 g monounsaturated fatty acids and 4 g polyunsaturated fatty acids) with a ratio of polyunsaturated to saturated fat of 0.06, and with 24 g carbohydrates, 4.75 g protein and 240 mg cholesterol.

2.5 Laboratory methods

Plasma levels of triglycerides, cholesterol, LDL and HDL cholesterol were determined using enzymatic methods for cholesterol (Cholesterol PAP, MA-kit 100, Roche) and triglycerides (Triglyceride PAP, Uni-Kit III, Roche). HDL cholesterol was quantified using a precipitation procedure with dextran sulphate and magnesium chloride [20]. For all these procedures a Cobas Mira autoanalyzer (Roche, Basle, Switzerland) was used. LDL cholesterol was calculated using the Friedewald formula [21]. The magnitude of PPL was quantified by calculating the area under curve (AUC) of postprandial triglycerides normalized for the fasting level [4]. In terms of atherosclerosis a meaningful classification of postprandial response is achieved by two postprandial blood collections [11] when one late discriminating time point is included [6]. Plasma insulin was measured by radioimmunoassay (Insulin RIA, Pharmacia).

2.6 Statistical methods

Analyses were performed by using the statistical software package SYSTAT 7,01 (SPSS, Chicago, IL). Data are presented as mean values±standard deviation (S.D.). All variables examined showed normal distribution and homogeneity of variances with Kolmorogov–Smirnov statistics and thus, were used untransformed for analysis of variance. Metabolic and vascular variables obtained by serial measurements at 08:00, 12:00 and 16:00 h were evaluated for trends over time by repeated measures analysis of variance with Hotelling’s T values for the time effect indicated. Associations between metabolic and vascular variables were sought by calculating univariate Pearson correlation coefficients. To analyze the combined effects of demographic and lipid data on postprandial EDD a multivariate regression analysis was performed, using the stepwise procedure. In this model triglyceride AUC, fasting triglycerides, LDL and HDL cholesterol, insulin, age and BMI were included as independent and postprandial EDD measured at 12:00 h as dependent variables, respectively.

Inspection of the regression between magnitude of PPL and endothelial function (Fig. 2) made us wonder as to whether there exist two subgroups because no data points could be found in the zone of 800–950 mg/dl·8 h triglyceride AUC. These subgroups refer to a stratification according to the triglyceride tolerance. Trends over time for metabolic and vascular variables were evaluated in both subgroups by repeated measures analysis of variance.

Fig. 2

Relationship between endothelium-dependent dilation (EDD) measured at 12:00 h in the postprandial state and magnitude of postprandial lipemia in 17 healthy male subjects.

3 Results

3.1 Baseline characteristics and lipid data

Baseline characteristics of the participants are summarized in Table 1 and lipid levels are illustrated in Table 2. After ingestion of the fatty meal the average level of triglycerides increased from 95±35 mg/dl at 08:00 h to 223±135 mg/dl at 12:00 h with widely varying individual values between 69 and 419 mg/dl and dropped to 134±73 mg/dl at 16:00 h (Hotelling’s T value 1.25, P=0.002). In the postprandial period no significant changes in the average levels of total cholesterol, LDL cholesterol, HDL cholesterol, insulin and glucose were observed.

View this table:
Table 2

Fasting and postprandial plasma lipid, insulin and glucose levels in 17 healthy men

View this table:
Table 1

Baseline characteristics in 17 healthy men

3.2 Vascular study results

All results of vascular reactivity measurements performed in the fasting and postprandial state are summarized in Table 3. Serial measurements of EDD in the fasting state exhibited significant changes over the 08:00–16:00-h period averaging 2.5±1.6% at 08:00 h to 7.5±2.7% at 12:00 h and to 7.0±2.1% at 16:00 h (Hotelling’s T value 5.41, P<0.001) with higher values at 12:00 and 16:00 compared to 08:00 h (Table 3 and Fig. 1). Measurements of EDD in the postprandial state exhibited values of 2.3±2.0% at 08:00 h, 5.0±2.6% at 12:00 h and 6.0±2.7% at 16:00 h (Hotelling’s T value 2.96, P<0.001) again showing higher values at 12:00 and 16:00 h than at 08:00 h (Table 3 and Fig. 1). Using a general linear model to compare the time course of fasting versus postprandial EDD values we observed significantly higher values of EDD in the fasting state as in the postprandial state with a Hotelling’s T value of 0.26; P=0.027 (Fig. 1). None of the other vascular parameters, including glyceryl trinitrate-induced dilation (16.4±3.5% at 08:00, 14.2±3.9% at 12:00, 14.1±3.6% at 16:00 h, P=NS) revealed any changes over time neither in the fasting nor in the postprandial state.

Fig. 1

Line plots of endothelium-dependent dilation (EDD) trends over the time period of 08:00 to 12:00 and 16:00 h in the fasting (○) and postprandial (●) state in 17 healthy male subjects. Evaluation of trends over time by repeated measures analysis of variance showed P<0.001 for the fasting and postprandial state indicating diurnal variation of EDD for each state of absorption. Comparison of the two line plots was performed by using a general linear model and gave a P value of 0.027. Data are expressed as means±S.D.

View this table:
Table 3

Vascular reactivity studies over time periods of 8 h in the fasting and postprandial state in 17 healthy men

In univariate linear regression analysis postprandial EDD values of 12:00 and 16:00 h were correlated inversely with the magnitude of PPL (r=−0.81, P<0.001 for 12:00 h and r=−0.57, P=0.017 for 16:00 h values) (Fig. 2). No correlation was found between postprandial EDD and fasting triglyceride levels (r=−0.409, P=1.0).

In a multivariate linear regression analysis, including cholesterol, LDL and HDL cholesterol, insulin, age and body mass index, respectively, only the magnitude of PPL remained statistically significant for the postprandial value of EDD measured at 12:00 h (Table 4).

View this table:
Table 4

Multiple stepwise regression analysis for determinants of endothelium-dependent dilation (EDD) at 12:00 h in the postprandial state in 17 healthy men

3.3 Post-hoc subgroup analysis

Scrutinizing the strong relationship between magnitude of PPL and 12:00 h postprandial EDD values (Fig. 2) data points appeared to be grouped below 800 and beyond 950 mg/dl·8 h triglyceride AUC. We, therefore, divided the 17 study subjects into eight subjects with high PPL (<950 mg/dl·8 h triglyceride AUC) and nine subjects with low PPL (<800 mg/dl·8 h triglyceride AUC). Clinical characteristics and fasting lipid levels did not reveal any differences between the two groups. In the low PPL group the diurnal variation of EDD values was unaffected by PPL with very similar values for fasting and postprandial 12:00 h EDD (7.0±2.5% fasting vs. 6.9±1.8% postprandial, P=0.67) (Fig. 3a). In contrast, the postprandial EDD values were severely affected in the high PPL group (Fig. 3b). In this subgroup postprandial EDD values were significantly lower at 12:00 h (2.9±1.5%) compared to fasting values at the same time (8.2±2.9%, P=0.003).

Fig. 3

Line plots of endothelium-dependent dilation (EDD) trends over the time period of 08:00 to 16:00 h in the fasting (○) and postprandial (●) state in the low postprandial lipemia (n=9) (a) and high postprandial lipemia group (n=8) (b). Evaluation of trends over time by repeated measures analysis of variance showed P<0.001 for the fasting and postprandial state in the low postprandial lipemia group (a) and P<0.001 for the fasting and P=0.012 for the postprandial state in the high postprandial lipemia group (b). Comparison of the line plots in each subgroup was performed by using a general linear model and resulted in a P value of 0.67 in the low postprandial lipemia group and a P value of <0.001 in the high postprandial lipemia group. Data are expressed as means±S.D.

4 Discussion

The main finding of our study is a transient impairment of endothelial function in the postprandial phase following ingestion of a standardized fatty meal which strongly depends on the magnitude of PPL. In contrast, endothelium-independent dilation was unaffected by PPL. A lipemia threshold of 800–950 mg/dl·8 h triglyceride AUC appears to define a degree of PPL above which endothelial function is clearly impaired.

This finding completes and confirms the results of a prior case control study using the identical fat load methodology [6]. In this study 101 patients with angiographically demonstrated CAD exhibited PPL values averaging 1006 mg/dl·8 h triglyceride AUC, whereas healthy controls showed 711 mg/dl·8 h triglyceride AUC revealing a discriminating lipemia value of 800–950 mg/dl·8 h triglyceride AUC. Therefore, the extent of PPL seems to allow segregation of individuals with normal postprandial triglyceride metabolism from those with proven CAD [6] and even from those with a preceding condition of atherosclerosis, i.e., impaired endothelial function. Since all subjects of our study had normal fasting triglyceride levels according to recent definitions [22] but showed impairment of endothelial function depending on the magnitude of PPL, we consider PPL metabolism to play a crucial role in atherogenesis [5,6,23]. This assumption is also supported by the study of Sharrett et al. [11] who found pronounced PPL to be associated with structural evidence of atherosclerosis, i.e., IMT of the carotid artery.

Our finding of a tight association of increased PPL and impairment in EDD is in good agreement with previous studies [8–10], but also contradicts the results of four recent studies demonstrating no correlation between magnitude of PPL and EDD [12–15]. From numerous factors we regard differences in the study methodology and patient selection to be mainly responsible for this controversy.

The dose of fat load used in some studies was considerably lower than the amount of fat used in our study [8,9,14] resulting in a lower PPL probably not sufficient to appropriately challenge triglyceride metabolic capacity [4]. As all previous studies about the influence of different doses of fatty meals consistently demonstrated no impact of low-fat meals on EDD compared to high-fat meals [8,9,13], we decided not to investigate the effects of a low-fat meal in our study. The kind of PPL quantification is another crucial issue in this context. Quantification by using single postprandial triglyceride values, particularly at different timepoints, appears to be inadequate for the comparison of various studies [8,9,14]. Differences in the magnitude of PPL can best be detected by calculating the AUC of postprandial triglycerides over an 8-h period after the meal. In addition, a prolongation of postprandial hypertriglyceridemia has been shown to reflect impaired triglyceride metabolic capacity with a close association to CAD [6,7] and even familial risk for CAD [24].

Differences in the content of an experimental meal represent another important determinant for inducing endothelial dysfunction. In recent studies oxidative stress has been accused to represent the underlying mechanism responsible for the impaired endothelial function observed during PPL [8,9,13]. Plotnick et al. [8] have reported that pre-treatment with antioxidant vitamins C and E blocks this decrease in endothelial function, a finding which was confirmed by Vogel et al. [25] who demonstrated that olive oil meals containing antioxidant vitamins or foods did not reduce EDD. Williams et al. [13] speculated that heat-modified fat vulnerable to oxidative degradation contributes to impaired EDD following a high-fat meal commonly served in fast food restaurants. In contrast, concomitant intake of red wine containing antioxidant polyphenols did not prevent endothelial dysfunction after a high fat meal [12].

Marchesi et al. [10] and Anderson et al. [15] used the same standardized fat load [4] and PPL quantification as we did. The application of an identical study design allows a direct comparison of results. Our group and Marchesi et al. [10] found an inverse correlation between PPL and EDD, whereas Anderson et al. [15] could not confirm this finding. We consider differences in gender of the study participants to be responsible for this contradictory results because in the latter study five of 12 subjects were female, whereas in a prior study [10] and in our study only male subjects were investigated. Recently, Schillaci et al. [26] had clearly proved a gender difference concerning postprandial endothelial function. In women the magnitude of PPL after a high-fat meal was much lower than in men and probably not high enough to alter EDD. Aging is another confounding factor associated with delayed and impaired clearance of PPL [27] and impaired endothelial dysfunction again influenced by gender [28]. The mean age of our study population was 36 years, lying below that one in the study of Anderson et al. [15] (43 years) but above that one in the study of Marchesi et al. [10] (23 years).

Recently, two independent reports have demonstrated a diurnal variation of EDD [16,17]. In contrast to other confounding factors, e.g., occlusion cuff position and occlusion time to induce hyperemia, the influence of diurnal variation was not been appreciated in studies, so far. Our study is the first that takes this confounding factor into account comparing endothelial function in the fasting and postprandial state. The EDD value of 2.5±1.6% found at 08:00 h in our study may seem low when compared to previous reports [8–10,12,29]. In these studies baseline EDD varied widely from 4.2 [14] to 10.6 [29] most probably due to the different morning time points of measurements which in our opinion is the only variable not well controlled for in these previous studies. In one study investigating young healthy individuals, three of 16 subjects were even excluded because their baseline EDD values of less than 5% were deemed too low [12].

The mechanism underlying the morning attenuation of EDD is unclear but may reflect diurnal rhythm of many pathophysiological factors, including sympathetic activity, blood pressure, heart rate, basal vascular tone, vasoactive hormones, prothrombotic activity, platelet activity and blood viscosity [30] whose time-dependent variation can be correlated with an increased risk of cardiovascular events occurring in the morning [31]. It was hypothesized that either nitric oxide production is attenuated or nitric oxide degradation is augmented at this time of day [17].

The wide range of PPL observed in our study in healthy, middle-aged men arises the question which underlying mechanisms may lead to a pronounced PPL. The determinants responsible for the extent of PPL are numerous, including environmental and genetic factors. Beside gender and age [26,28], physical activity [4,32], diet [33], heavy alcohol consumption [34] and smoking [35] have been shown to affect the magnitude of PPL. Gender and age have already been mentioned. A large number of genetic factors whose products control or modulate postprandial lipoprotein metabolism are also associated with impaired triglyceride metabolic capacity leading to pronounced PPL. Lipoprotein lipase is the key enzyme in the first step of triglyceride-rich lipoprotein catabolism and more than 80 mutations of the lipoprotein lipase gene have been described [36]. In a study by Miesenböck et al. [19] heterozygous carriers with a mutation at codon 188 of the lipoprotein lipase gene exhibited pronounced PPL, even though fasting triglycerides were in the normal range. Polymorphisms in the lipoprotein lipase gene [37], the apoE genotype [38] and a common polymorphism of a signal peptide of apoB [39] have been reported to influence PPL. We have determined plasma lipids, glucose and insulin only and therefore cannot contribute to the question why some of our study subjects demonstrated a pronounced PPL despite normal fasting lipid levels. Similar levels of fasting insulin and glucose obtained in our study in the high and low PPL group enable us, however, to exclude insulin resistance as a responsible factor for pronounced PPL in our subjects. Moreover, we did not find an association of postprandial insulin increase with impaired EDD which contributes to the finding that insulin enhances endothelial function by release of nitric oxide [40].

In the postprandial state triglyceride-rich lipoproteins accumulate [6,41] and lead to an extraction of cholesteryl esters from HDL and LDL. Since these postprandial metabolic effects are strongly dependent on the magnitude of PPL, the resulting structural alteration of LDL to small dense LDL and the depression of HDL will be exaggerated in the case of pronounced PPL [4,6]. Small dense LDL, which are more easily oxidized [42], and depressed HDL, with diminished antioxidant potential [43], have been proposed to represent important mechanisms for endothelial dysfunction and atherogenesis [6,42–45]. Recent data confirm that triglyceride-rich lipoproteins and oxidative stress are intimately linked [46]. In this way an impaired triglyceride metabolic capacity defined by an increase of triglyceride AUC above a level of 800–950 mg/dl·8 h might represent the central underlying mechanism for impaired EDD.

The number of subjects investigated was relatively small. However, it is comparable with those of other studies using similar time-consuming and demanding methodologies [8–10,12–15] most probably sufficient to uncover a true biological phenomenon. Of course, our observations need further confirmation by other studies particularly when a meaningful threshold value has to be worked out for normal and pathological lipemic response to a standardized fatty meal. A clear quantitative definition (e.g., >800–950 mg/dl·8 h triglyceride AUC) of the lipemic response will be essential, if triglyceride metabolic capacity is ever to be incorporated into risk assessment procedures and, hopefully, intervention efforts of cardiovascular diseases.

The present study shows that pronounced postprandial hypertriglyceridemia impairs endothelial function transiently. This was demonstrated by a significant alteration of diurnal variation of EDD, attenuated at noon and afternoon, during an exaggerated PPL. With a standardized oral fatty meal it appears possible to identify a PPL value that enables to distinguish between individuals with impaired triglyceride metabolic capacity resulting in endothelial dysfunction from subjects with normal PPL. Our data suggest that impaired triglyceride metabolism may play an important role in atherogenesis.


This work was supported by the grant S07106 — MED of the Austrian Fond zur Förderung der Wissenschaftlichen Forschung to JRP.


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
View Abstract