点击显示 收起
1 From the Second Department of Internal Medicine (MS, IC, NH, and NT), the Department of Clinical Pharmacology and Therapeutics (KY and SU), and the Faculty of Medicine, University of the Ryukyus, Okinawa, Japan
2 Address reprint requests and correspondence to M Shimabukuro, Second Department of Internal Medicine, Faculty of Medicine, University of the Ryukyus, 207 Uehara, Nishihara, Okinawa 903-0215, Japan. E-mail: mshimabukuro-ur{at}umin.ac.jp or me447945{at}members.interq.or.jp.
ABSTRACT
Background: Abnormalities during the postprandial state contribute to the development of atherosclerosis. Reportedly, postprandial hyperglycemia, hypertriglyceridemia, and hyperlipacidemia independently cause postprandial cytokine activation. However, it is not clear which dietary composition preferentially affects postprandial endothelial function in healthy subjects.
Objective: We aimed to examine the associations of dietary composition and postprandial endothelial function in healthy subjects.
Design: The effects of a single ingestion of a high-carbohydrate meal (300 kcal, 100% carbohydrate), a high-fat meal (30 g fat/m2, 35% fat), or a standard test meal (478 kcal; 16.4% protein, 32.7% fat, 50.4% carbohydrate) on postprandial plasma concentrations of adiponectin and forearm blood flow (FBF) during reactive hyperemia were studied in healthy subjects.
Results: The peak FBF response and the total reactive hyperemic flow (flow debt repayment; FDR), indexes of resistance artery endothelial function, were unchanged after ingestion of a high-carbohydrate and standard test meal but decreased 120 and 240 min after a high-fat meal. After a high-fat meal, decreases in peak FBF and FDR were well correlated with an increase in plasma free fatty acid (FFA) concentrations but not with the other biochemical variables, including triacylglycerol, insulin, glucose, total cholesterol, HDL cholesterol, and adiponectin.
Conclusions: Postprandial endothelial function was impaired only after the high-fat diet and not after the high-carbohydrate or standard test meal in healthy subjects. Because such endothelial dysfunction after a high-fat meal was closely correlated with FFA concentrations, postprandial state could be hazardous, mostly through acute hyperlipacidemia in healthy subjects.
Key Words: Free fatty acids • endothelial function • adiponectin • postprandial state • hyperlipidemia • hyperglycemia
INTRODUCTION
The abnormalities of the postprandial state are important contributing factors to the development of atherosclerosis. There is evidence that postprandial hypertriglyceridemia is a risk factor for cardiovascular diseases (CVD) (1), whereas in diabetic subjects, postprandial hyperglycemia has been proposed as an independent risk factor for CVD (2).
The risk of coronary artery disease is increased by consumption of a diet rich in saturated fatty acids (3). In both healthy subjects and in diabetic patients, a single high-fat meal induces endothelial activation, which is associated with increased inflammatory cytokine production (4). Meanwhile, postprandial hypertriglyceridemia (4–6) and hyperglycemia (7) can elicit endothelial dysfunction via an independent and a cumulative increase in oxidative stress. Free fatty acids (FFAs) elevated by intravenous lipid or heparin infusion directly impair the vasodilatory response to acetylcholine in healthy humans, which may be pathophysiologically relevant to the development of postprandial endothelial dysfunction in patients with obesity and insulin resistance (8–10).
Adiponectin is an adipocyte-derived plasma protein (adipocytokine) that accumulates in injured arteries and has potential antiatherogenic properties. Maintenance of adiponectin concentrations is closely associated with normal endothelial function in humans (11, 12); therefore, postprandial changes in plasma adiponectin concentrations (13–15) could be related to postprandial endothelial dysfunction. However, such a relation has not been evaluated. We investigated effects of diet composition on postprandial plasma concentrations of adiponectin and endothelial function in healthy subjects.
SUBJECTS AND METHODS
Subjects
Twelve healthy subjects (6 men and 6 women) aged 30–42 y (
On 3 separate mornings 7 d apart, participants ingested a high-carbohydrate meal (300 kcal, 100% carbohydrate) (16), a high-fat meal (30 g fat/m2, 35% fat, 342 kcal/100 g), or a standard test meal (478 kcal; 16.4% protein, 32.7%fat, 50.4% carbohydrate; proposed by a working group of the Japan Diabetes Society) after fasting overnight (17). The order for meal ingestion was randomized in a crossover design. Endothelial function was measured by using forearm blood flow (FBF) before and 120 and 240 min after the ingestion of each meal. Blood samples were obtained at 0, 30, 60, 120, and 240 min. FBF was measured with the use of a mercury-filled silastic strain-gauge plethysmograph (EC-5R; DE Hokanson Inc, Issaquah, WA) as described (Figure 1) (11, 16).
FIGURE 1.. Representative tracing of forearm blood flow (FBF). FBF was calculated from gradient of tracing (mV/s) with a mercury-filled strain-gauge plethysmograph at baseline and after 5 min of upper arm cuff occlusion.
The strain gauge was attached to the upper arm, held above the right atrium, and connected to a plethysmographic device. At baseline and during 180 s after release of 5-min occlusion of the upper arm at 200 mm Hg (reactive hyperemia), FBF was measured as follows. First, the wrist cuff was inflated to a pressure of 200 mm Hg to exclude hand circulation, and then the upper arm cuff was repetitively inflated to 40 mm Hg to occlude venous outflow from the arm (7 s) and deflated (8 s) in each 15-s cycle with a rapid cuff inflator (EC-20; DE Hokanson Inc). The FBF output signal was transmitted to a recorder (U-228, Advance Co, Nagoya, Japan). FBF was calculated from gradient of tracing (mV/s) with a mercury-filled strain-gauge plethysmograph at baseline and after 5 min of upper arm cuff occlusion and was expressed as milliliters per minute per 100 mL of forearm tissue. A representative tracing of an FBF curve is shown in Figure 1.
Calculations of blood flow debt incurred during arterial occlusion, reactive hyperemic flow, and blood flow debt repayments (FDR) were made as described (11) below:
RESULTS
The meals were well tolerated by all patients, and no adverse events were observed during the study. Systemic hemodynamic and metabolic variables before test meal ingestion were comparable on the 3 study mornings.
The effects of the high-carbohydrate, high-fat, and standard test meals on plasma glucose, plasma insulin, serum lipid, and serum adiponectin concentrations are shown in Figure 2. After the high-carbohydrate and standard test meals, plasma glucose concentrations increased from baseline by 5 mmol/L to a peak of 7.2 mmol/L at 60 min and returned to baseline at 240 min (Figure 2). Plasma glucose concentrations did not change after the high-fat meal. After the high-carbohydrate and standard test meals, plasma insulin concentrations increased from baseline by 50 pmol/L to a peak of 540–590 pmol/L at 60 min. The plasma insulin concentration did not change after the high-fat meal. Serum FFA concentrations decreased during the 120 min after the high-carbohydrate and standard test meals and returned to baseline at 240 min. Serum FFA concentrations increased to 0.87 ± 0.06 mmol/L at 240 min after the high-fat meal. Serum triacylglycerol concentrations increased to 2.77 ± 0.83 mmol/L at 240 min after the high-fat meal, but did not change after the high-carbohydrate and standard test meals. Serum concentrations of total and HDL cholesterol and adiponectin did not change during 240 min after either test meal.
FIGURE 2.. Mean (±SEM) blood concentrations of glucose, insulin, free fatty acid (FFA), triacylglycerols, total and HDL cholesterol, and adiponectin after loading of a high-carbohydrate, high-fat, and standard test meal in 12 healthy subjects (6 men and 6 women). Multigroup comparisons of time course curves were analyzed by 2-factor repeated-measures ANOVA. If the multigroup difference was significant, intragroup comparisons were made by one-factor repeated-measures ANOVA followed by Tukey's post hoc test. *Significantly different from 0 min, P < 0.01 (Tukey's post hoc test).
The effects of the high-carbohydrate, high-fat, and standard test meals on FBF are shown in Figure 3, A, B, and C. The peak FBF was unchanged before and after the high-carbohydrate and standard test meals, but decreased significantly 120 and 240 min after the high-fat meal (Figure 3, A and B). The total reactive hyperemic flows (FDR), indexes of resistance artery endothelial function, also decreased 240 min after the high-fat meal (Figure 3, C).
FIGURE 3.. Effects of dietary composition (high-carbohydrate, high-fat, or standard test meal) on forearm blood flow (FBF) and total reactive hyperemic flow (flow debt repayment; FDR) in 12 healthy subjects (6 men and 6 women). A: Measurements were taken at baseline and during reactive hyperemia 120 and 240 min after the meals. B: Peak FBF was measured during reactive hyperemia before (0 min) and 120 and 240 min after the meals. C: Total reactive hyperemic flow (FDR) was measured during reactive hyperemia. FDR (%) = (reactive hyperemic flow/blood flow debt) x 100. Data are presented as means ± SEMs. One-factor ANOVA with Tukey's post hoc test was used to compare intragroup means. *Significantly different from 0 min, P < 0.01.
After the high-fat meal, changes from baseline in peak FBF (peak FBF) and FDR (FDR) were inversely well correlated with the change in plasma FFA concentration (FFA) (Figure 4) but not with the other biochemical variables, including triacylglycerol, insulin, glucose, total cholesterol, HDL cholesterol, and adiponectin (data not shown).
FIGURE 4.. Changes from baseline in peak forearm blood flow (peak FBF) and flow debt repayment (FDR) and changes in serum free fatty acids (FFA) and triacylglycerols 120 and 240 min after high-fat meal loading in 12 healthy subjects. After adjustment by ANCOVA with the use of time point after loading as covariance, Pearson's correlation coefficients (r) and P values were calculated.
DISCUSSION
In this study, we investigated the effects of diet composition on postprandial plasma concentrations of adiponectin and endothelial function in healthy subjects. Effects of the high-carbohydrate, high-fat, and standard test meals on plasma glucose, insulin, and serum lipid concentrations were different, and postprandial forearm endothelial function was impaired after high-fat diet but not after the high-carbohydrate and standard test meals.
Postprandial hyperlipidemia (4) and postprandial hyperglycemia independently produce endothelial dysfunction and both are now largely recognized as potential underlying mechanisms of macrovascular events in subjects with normal or impaired glucose tolerance (1, 2). Postprandial state is a complex dynamic phase. Hyperlipidemia, hyperglycemia, or changes in the bloodstream of various humoral factors are simultaneously present in the postprandial phase. Thus, the specific molecule mostly responsible for endothelial dysfunction has not been determined. We therefore examined the effects of diet composition on postprandial biochemical variables and endothelial function. The study was done on 3 separate mornings, and the order in which the 3 types of meals were ingested was randomized in a crossover design to minimize the other confounding factors.
Plasma glucose and insulin concentrations increased at 60 min and returned to baseline at 240 min after the high-carbohydrate and standard test meals but did not change after the high-fat meal. Serum FFA and triacylglycerol concentrations increased only at 120 and 240 min after the high-fat meal. Plasma adiponectin concentrations did not change during the 240 min after either test meal. Some (13, 14), but not all (15), studies reported that adiponectin concentrations changed postprandially. Esposito et al (13) reported that adiponectin concentrations decreased from baseline after the high-fat meal, but not after the high-carbohydrate meal. Although the reason for the discrepancy between their study and ours is unknown, the difference in the percentage of energy as fat of the diets could be one reason (60% compared with 32.9%).
We reported that the high-carbohydrate and standard test meals did not produce endothelial dysfunction in healthy subjects with normal glucose tolerance (16, 17). However, it has been reported that the high-fat meal could impair endothelial function, even in healthy subjects (4–7). Our results confirm the previous findings that endothelial function is impaired after ingestion of a high-fat meal but not after ingestion of a high-carbohydrate or standard test meal.
As shown in Figure 4, decreases in peak FBF and FDR after a high-fat meal were well correlated with increases in serum FFA concentrations, but not with the other biochemical variables, including triacylglycerol, insulin, glucose, total cholesterol, HDL cholesterol, or adiponectin. Because adiponectin concentrations did not change after the meals, adiponectin oscillation is unrelated to postprandial changes in acute vascular reactivity. Our previous study showed that plasma adiponectin concentrations were well correlated with baseline (fasting) endothelial function in nondiabetic subjects (11). We postulated that there are 2 possible mechanisms by which hypoadiponectinemia decreases endothelial function: 1) hypoadiponectinemia can cause endothelial dysfunction by decreasing insulin sensitivity, and 2) hypoadiponectinemia may be directly linked to early atherosclerotic vascular damage and subsequent endothelial dysfunction.
In the present study, we showed that the FFA concentration, rather than the plasma adiponectin concentration, is mainly associated with postprandial endothelial dysfunction in healthy subjects. Taken together, hypoadiponectinemia could be linked to chronic vascular injury but not to acute postprandial vascular injury. It has been shown that the FBF response to the intraarterial infusion of acetylcholine was impaired acutely after systemic infusion of lipid and heparin in healthy subjects (9, 10). Previous findings support the notion that acute hyperlipacidemia directly impairs endothelial function, which was observed in this study.
The mechanism for such FFA-related endothelial dysfunction could not be drawn from this study. One can assume that hypertriglyceridemia, a cumulative effect of endothelial activation, or both may be associated with increased serum concentrations of inflammatory cytokines such as tumor necrosis factor-, interleukin-6, intercellular adhesion molecule-1, and vascular cell adhesion molecule-1 (13).
Another possible explanation is the production of reactive oxygen species (ROS) by FFAs. Ceriello (7) reported that increased production of ROS, determined by nitrotyrosine, was closely associated with endothelial dysfunction after consumption of a high-fat meal. We confirmed that FFA-induced endothelial dysfunction could be protected by co-infusion of antioxidant, vitamin C, in healthy humans (10). Experimentally, we observed that FFA directly inhibited vascular response to acetylcholine in stripped aorta isolated form normal animal models (19). Also, we found that FFA directly enhanced the production of vascular ROS via up-regulation of NADPH oxidase subunit mRNA, and the inhibition of ROS production by N-acetyl-cysteine (a general antioxidant), diphenyleneiodonium, and apocynin (NADPH oxidase inhibitors) can prevent FFA-induced endothelial dysfunction (19).
In summary, we investigated the effects of diet composition on postprandial plasma concentrations of adiponectin and endothelial function in healthy subjects. Postprandial endothelial function was impaired only after ingestion of the high-fat diet and not after consumption of the high-carbohydrate and standard test meals. Because such endothelial dysfunction after a high-fat meal was closely correlated with FFA concentrations, the postprandial state could be hazardous, mostly through acute hyperlipacidemia in healthy subjects.
ACKNOWLEDGMENTS
None of the authors had a conflict of interest.
REFERENCES