点击显示 收起
1 From the Institute of Experimental and Clinical Pharmacology and Toxicology (TKR, KR, ES, JT-A, RHB, and RM) and the Department of Cardiology and Angiology, Heart Center (TKR), University Hospital Hamburg-Eppendorf, Hamburg, Germany, and the Institute of Pharmacy, University of Hamburg, Hamburg, Germany (UR)
2 Supported by the Institute of Experimental and Clinical Pharmacology and Toxicology of the University Hospital Hamburg-Eppendorf. 3 Reprints not available. Address correspondence to TK Rudolph, Department of Cardiology and Angiology, Heart Center, University Hospital Hamburg-Eppendorf, Martinistrasse 52, 20246 Hamburg, Germany. E-mail: t.rudolph{at}uke.uni-hamburg.de.
ABSTRACT
Background: High-fat meals have negative effects on endothelial function, but vitamin-rich side orders may prevent these negative effects.
Objective: The acute effects of conventional and alternative fast-food meals on vascular function and various cardiovascular biomarkers were investigated.
Design: In a crossover study, flow-mediated endothelium-dependent dilatation (FMD) and cardiovascular disease risk markers were investigated in 24 healthy volunteers before and 2 and 4 h after 3 fast-food meals: a conventional beef burger with French fries, ketchup, and carbonated lemon-flavored soda (meal 1); a vegetarian burger with French fries, ketchup, and carbonated lemon-flavored soda (meal 2); and a vegetarian burger with salad, fruit, yogurt, and orange juice (meal 3).
Results: FMD decreased after all 3 fast-food meals: the values were 9.7 ± 2.5%, 7.5 ± 3.5%, and 6.2 ± 3.3% for meal 1; 9.2 ± 3.4%, 7.1 ± 3.4%, and 6.3 ± 4.0% for meal 2; and 8.8 ± 3.3%, 6.2 ± 4.0%, and 6.8 ± 4.3% for meal 3 at baseline, 2 h, and 4 h, respectively. There were significant intraindividual differences for time (P < 0.001) but not for type of meal (P = 0.677). A postprandial increase in baseline diameter of the brachial artery was significant for time (P < 0.001) but not for type of meal (P = 0.148).
Conclusions: Against common expectations, a conventional beef burger meal and presumably healthier alternatives with or without vitamin-rich side orders did not differ significantly in their acute effects on vascular reactivity. The frequently reported postprandial decline in FMD may be attributed in part to a postprandial increase in baseline arterial diameter.
Key Words: Fast food • healthy volunteers • endothelial function • ratio of L-arginine to asymmetric dimethylarginine • ADMA • oxidative stress
INTRODUCTION
Worldwide consumption of fast food is steadily increasing. In the United States alone, sales of fast food reached a value of $161 billion in 2004, of which burgers accounted for 53% (1). Today in the United States, up to 37% of adults and up to 42% of children regularly consume fast food. These adults and children have higher intakes of energy, fat, saturated fat, sodium, and carbonated soft drinks and lower intakes of vitamins A and C, milk, fruit, and vegetables than do adults and children who do not consume fast food regularly (2-4). Epidemiologic, clinical, and in vitro evidence indicates that frequent consumption of fast food may have an unfavorable effect on the cardiovascular disease (CVD) risk profile in general, because fast food has been shown to promote weight gain and insulin resistance (5). Several studies suggested that these adverse long-term effects are already caused by the acute impairment of endothelial function and the increase in oxidative stress observed after the intake of high-fat meals that are low in vitamins (6). Endothelial dysfunction and oxidative stress are considered important factors in the development and progression of atherosclerosis (7-9). The presence of endothelial dysfunction and oxidative stress has repeatedly been shown to be associated with a greater risk of future cardiovascular events and death (10). In contrast to fast food, some diets, eg, the vegetarian or the Mediterranean diet, that contain high amounts of vitamins and unsaturated fat may have protective effects (11-20). For example, the administration of high doses of antioxidant vitamins such as vitamin C and E prevented diet-induced impairment of endothelial function (21). In response to growing public awareness, several large fast-food chains began to adapt their supply to a new demand for healthier fast food. Vegetarian burgers and vitamin-rich side orders and fruit juices are increasingly offered as a "healthy alternative" or add-on to conventional beef burger meals and are promoted as having supposedly less negative effects on the vascular risk profile.
Plausible as they are, these claims have never been investigated for complete meals as they are sold and consumed by the public. Therefore, the present study was set up to address 3 questions. First, does commercially sold fast food have an effect on vascular function and CVD risk markers? Second, does it make a difference, with respect to acute effects on vascular function and CVD risk markers, to consume a conventional (beef burger) or a vegetarian fast-food meal? Third, what is the effect of presumed vitamin- and antioxidant-rich side orders on vascular function CVD risk markers?
SUBJECTS AND METHODS
Subjects and study protocol
The study was conducted at the Institute of Experimental and Clinical Pharmacology of the University Hospital Hamburg-Eppendorf. Twenty-four healthy volunteers aged 18–65 y were enrolled in this observer-blinded, randomized, 3-way crossover trial. Major exclusion criteria were CVD, diabetes, dysfunction of coagulation, acute infection, anemia, and intolerance of any components of the different meals as assessed by clinical examination and interview 1 wk before the first study day.
Written informed consent was obtained from all volunteers. The study was conducted in accordance with the principles outlined in the Declaration of Helsinki and was approved by the local Ethics Committee.
On 3 study days separated by 1 wk, each volunteer presented after a 12-h overnight fast for baseline blood sampling and vascular function testing. Two hours after baseline testing, each volunteer received in random order a conventional beef burger meal consisting of a 211-g beef burger (Big Mac; McDonald's Corp, Oak Brook, IL) with 152 g French fries, 20 mL ketchup, and 500 mL carbonated lemon-flavored soda (Sprite; Coca-Cola Co, Atlanta, GA) (meal 1); a vegetarian burger meal with standard side orders consisting of a 203-g vegetarian burger (GemüseMac; McDonald's Corp, Munich, Germany) with 152 g French fries, 20 mL ketchup, and 500 mL carbonated lemon-flavored soda (meal 2); and a vegetarian burger meal with vitamin-rich side orders consisting of a 203-g vegetarian burger, 90 g salad, 30 mL balsamic dressing, 306 g yogurt with fruit, and 500 mL orange juice with 200 mg vitamin C [Minute Maid; Minute Maid (a division of the Coca-Cola Co), Houston, TX] (meal 3).
The dietary composition of the fast-food meals, as derived from publicly available dietary information provided by McDonald's Corp on the Internet (22) and as cross-checked against an independent source (23), is given in Table 1. The percentage of calories from carbohydrate, protein, and fat protein was 52.3%, 11.0%, and 36.7%, respectively, for meal 1; 56.6%, 5.7%, and 37.6%, respectively, for meal 2; and 63.0%, 9.4%, and 27.6.3%, respectively, for meal 3. Meal 1 corresponds of one of the world's most frequently consumed fast-food meals, meal 2 was chosen to allow comparison to an isocaloric vegetarian fast-food component (to address the hypothesis that pure vegetarian meals have a less negative effect on vascular function), and meal 3 was chosen to allow further comparison to assess the effect of a lower fat and higher vitamin content.
View this table:
TABLE 1. Nutrition facts for the 3 test meals1
All components were anonymously obtained at McDonald's restaurants. At baseline and 2 and 4 h after the meal, vascular function tests and blood sampling were performed. In addition, urine was collected in the 2 h before and the 4 h after ingestion of the fast-food meal.
Vascular function testing
Methods of assessing endothelium-dependent vasodilation followed the principles set by the International Brachial Artery Reactivity Task Force (24). Endothelial function was assessed in each volunteer's right arm in a quiet, temperature-controlled room (22 °C) by using high-resolution ultrasound (Siena 12-MHz linear array transducer; Siemens, Erlangen, Germany). Longitudinal scans of the brachial artery were obtained 5 cm proximal to the antecubital fossa. The transmit focus zone was set at the depth of the anterior wall. Anatomical landmarks and snapshot images were used to assess flow-mediated endothelium-dependent dilatation (FMD) in the exact same vessel section on each study day and at each time point. A view of a 5-cm longitudinal section of the brachial artery (BA) was recorded for periods of 30 s at baseline and before and during peak (ie, 1 min after cuff release) reactive hyperemia (after deflation of a blood pressure cuff previously inflated to 50 mm Hg above the volunteers systolic blood pressure around the forearm for 5 min). Each 30-s recording was digitalized (Vascular Imager 4.1.3; Medical Imaging Applications LLC, Iowa City, IA) at a rate of 10 high-resolution frames/s (= 300 frames per recording), by using specialized software (BRACHIAL ANALYZER, version 4.1.3; Medical Imaging Applications LLC). This allowed the averaging of maximal, minimal, and mean vessel diameters over 25–40 heart beats. Results obtained for maximal, minimal, and mean vessel diameter were very similar, and thus data based on mean vessel diameters are presented, because they gave the best reproducibility. FMD was calculated as the percentage change in diameter from the baseline value before cuff release to the value 1 min after cuff release. Ultrasound studies and image analysis were performed separately by independent investigators in an observer-blinded fashion. The mean intraindividual CV of the baseline diameter of the preprandial brachial artery of the baseline measurements obtained on the 3 separate study days (representing the cumulative intraindividual day-to-day variation of the vessel diameter and the intraobserver and interobserver variation in measurement) was 4.6%. For the baseline measurements of FMD in the present study, we observed a mean (intraindividual) day-to-day difference in FMD of 2.68 ± 1.58% after a mean baseline vasodilation of 9.26 ± 2.47%.
Laboratory measurements
All laboratory measurements were performed in a blinded fashion. We analyzed 8-iso-prostaglandin F2 (8-iso-PGF2) in urine samples by using gas chromatography–mass spectrometry (GC-MS) as previously described (25). Plasma L-arginine and asymmetric dimethylarginine (ADMA) were measured by using a liquid chromatography–tandem mass spectrometry (LC-MS/MS)–based method with one derivatization step and sample pretreatment consisting of protein precipitation as recently detailed (26). Plasma insulin was measured by using an enzyme-linked immunosorbent assay (Mercodia SA, Sweden). All other laboratory variables, eg, plasma lipids, lipoproteins, glucose, uric acid, and blood cells, were obtained immediately by using certified routine clinical chemistry procedures.
Statistical analysis
We aimed to detect a minimal difference of 30% in the primary target variable of the mean diet-induced change in FMD between the 2 meals with a two-sided type I error protection of 0.05 and a power of 0.80. On the basis of previous trials and pilot data, we estimated the SD of the difference between the 2 values in the same participant to range from 20–50% (in relative terms), which gave formal sample size requirements of 10–24 subjects.
All data were tested for normal distribution by using the Kolmogorov-Smirnov test. Continuous variables were expressed as arithmetic means ± SDs if normally distributed or, otherwise, as medians and 25th–75th percentiles. Parametric (t test) and nonparametric (Mann-Whitney U (2-sided) and Kruskal-Wallis H) tests were used for comparisons, as appropriate. Unless indicated otherwise, descriptive data are presented as means ± SDs or medians and 25th–75th percentiles. Categorical variables were compared by using a chi-square test or exact Fishers test as appropriate. Correlations were assessed by Pearsons's or Spearmans test, as appropriate.
Effects of meals were assessed by using general linear models (GLM) for repeated measures with type of meal (3 levels) and time of measurement (2–3 levels) as fixed factors. Comparisons are presented for differences between meals, time points, and meal x time point interaction (the Huynh-Feldt correction was applied when the sphericity assumption was not met). When the analyses indicated a significant effect of diet (P < 0.05), pairwise comparison of the diets was conducted. Between-diet comparisons were corrected by using the Bonferroni correction as adequate. Effects of individual meals were assessed by using a GLM for repeated measures with type of meal (3 states) and time of measurement (2–3 states) as a fixed factor and posttest Bonferroni correction. For the GLM, variables with a highly skewed distribution (ie, 8-iso-PGF2, triacylglycerols, and insulin) were log transformed.
No further correction of the type I error was made for the number of global GLM models used. A value of P < 0.05 (2-tailed) was considered to be significant. All calculations were performed with SPSS software (version 12.0; SPSS, Chicago, IL).
RESULTS
Baseline characteristics
The study group consisted of 24 healthy volunteers (14 F, 10 M; aged 32 ± 11 y). Fasting serum concentrations of cholesterol, LDL cholesterol, and triacylglycerols were 175.5 ± 33.7, 95.9 ± 25.3, and 72.0 mg/dL, respectively (range: 52.5–112.8 mg/dL). The mean body mass index (BMI; in kg/m2) was 24 ± 5. The volunteers were following no particular diet and were not taking additional vitamins. Women had a smaller brachial vessel diameter at baseline than did men (3.50 ± 0.57 and 4.06 ± 0.44 mm, respectively; P = 0.017), but the mean FMD at baseline did not differ significantly (8.65 ± 3.02% and 9.81 ± 3.29%, respectively; P = 0.388).
Effect of fast-food meals on flow-mediated endothelium-dependent dilatation and hemodynamics
Preprandial and postprandial hemodynamic factors and laboratory values are shown in Table 2. There were no significant differences among the 3 baseline values for each variable (P > 0.1 for all).
View this table:
TABLE 2. Hemodynamic and laboratory variables before and 2 h and 4 h after the 3 fast-food meals1
First, we compared FMD at all time points. The intraindividual comparison of preprandial and postprandial (2 h and 4 h) FMD of all meals in a repeated-measures GLM indicated that time point but not type of meal was associated with differences in FMD (P < 0.001 for time and P = 0.677 for type of meal).
There was a small but significant postprandial increase in diameter of the brachial artery (P < 0.001 for time of measurement). The comparison of all 3 meals found no significant difference in the effect of the 3 meals on the vessel diameter (P = 0.148).
Effects on plasma markers and markers of oxidative stress
Serum triacylglycerol concentrations increased significantly after each meal (P < 0.001 for time of measurement), but without significant difference between the 3 meals (P = 0.828). Serum HDL declined after the meals (P = 0.003 for time of measurement), but without significant difference between the 3 meals (P = 0.403). Serum total cholesterol did not change from baseline after any of the 3 meals. Two hours after each type of meal, plasma insulin was significantly elevated, and it remained high at 4 h—again, without a significant intraindividual difference between the 3 meals (P = 0.895). Postprandially, plasma glucose increased significantly (P = 0.002 for time point), but there was no significant effect of the meals on postprandial glucose (P = 0.504).
Plasma concentrations of the endogenous nitric oxide (NO) synthase inhibitor ADMA remained unchanged after all 3 meals (Table 2). However, the 3 meals differed in their effects on plasma L-arginine (P < 0.001) and the corresponding ratio of L-arginine to ADMA (L-arginine:ADMA) (P = 0.008). We observed significant increases (Table 2) after the 2 meals with the highest protein content (conventional beef burger meal (1) and vegetarian burger meal with vitamin-rich side orders (3).
The effects of the 3 meals on urinary excretion of 8-iso-PGF2 did not differ significantly (P = 0.226). Although the underlying postprandial urinary excretion of 8-iso-PGF2 was numerically higher than baseline, it did not differ significantly from baseline (P = 0.067 for time of measurement).
Correlation of vessel diameter, flow-mediated endothelium-dependent dilatation, and other markers
To elucidate possible mechanisms contributing to the postprandial decline in FMD, we correlated the postprandial decline in FMD or changes in baseline diameters with changes in factors associated with vascular function (ie, triacylglycerols, glucose, L-arginine, ADMA, insulin, and urinary 8-iso-PGF2 excretion). We did not observe a clinically relevant correlation in any of the groups (r < 0.3 for all, P > 0.05 for all).
DISCUSSION
This study is the first clinical trial simultaneously comparing the acute effects of one of the world's most frequently consumed fast-food meals and vegetarian alternatives (with and without vitamin enrichment) on endothelial function, oxidative stress, and a wide array of other vascular markers. The most important findings are that we observed a decline in FMD after all 3 investigated fast-food meals and that this decline did not differ significantly between the meals. This result is somewhat unexpected, given numerous previous reports that various dietary components or vitamin supplements could prevent acute diet-induced changes in vascular function (21). Lack of sensitivity of the FMD determination is an unlikely explanation of this negative finding, because the modest decline in FMD observed after all 3 meals reached significance. The changes in FMD can most likely be attributed to diet effects, because, in the few studies conducted to examine circadian rhythms, the circadian changes in FMD were rather moderate and not significant (27). Moreover, in studies with designs similar to the design of the present study, some meals induced changes in FMD, but others did not. If circadian rhythm were a significant confounder, it would have blunted results in all of the studies.
Therefore, other mechanisms, such as the small but significant postprandial increase in baseline diameter, must be considered, whereas the absolute vasodilation after cuff occlusion remained rather constant. As a net effect, the postprandial increase may have translated into the apparent decline in FMD. This observation is consistent with some (28, 29) but not all (30, 31) previous studies concerning diet-induced changes in vascular function. Unfortunately, baseline diameters are reported in only a fraction of studies. It should be noted, however, that the postprandial change in baseline diameter was small in the present study, which makes it rather unlikely that the postprandial decline in FMD can be attributed solely to an increase in baseline diameter. A postprandial attenuation of vascular function has frequently been attributed to dietary triacylglycerols (32). The observation that triacylglycerols increased not only after meal 1 and meal 2 but also after meal 3 can most likely be attributed to the fact that carbohydrates from fruit and fruit juices can also increase triacylglycerols (33). Nevertheless, postprandial changes in triacylglycerols did not correlate with changes in baseline diameter or FMD. This observation is in line with several studies in patients with comparably low baseline triacylglycerol concentrations (6, 21, 28), whereas other studies, mainly performed in subjects with higher baseline triacylglycerol concentrations, more frequently reported an inverse correlation of the changes in triacylglycerol concentrations and FMD (32, 34, 35). Another frequently considered explanation of the postprandial decline in FMD is diet-induced oxidative stress. Evidence for diet-induced oxidative stress is usually obtained rather indirectly through the attenuation of diet-induced decline in FMD by supplementation of antioxidants such as vitamin C and E (21, 36-38). Whereas we observed a considerable interindividual variation in urinary 8-iso-PGF2 excretion, the crossover design showed a remarkable intraindividual stability of baseline urinary 8-iso-PGF2 excretion. In comparing the 3 meals by GLM analysis, we did not find any significant differences in postprandial changes in urinary 8-iso-PGF2 excretion. Thus, in contrast to pharmacologic (high) doses of vitamins, which were often given intravenously in positive studies (39), a vitamin-rich side order of mixed salad, orange juice (containing 200 mg vitamin C), and yogurt did not significantly alter urinary excretion of 8-iso-PGF2 and could not prevent impairment of FMD. Possibly, a single vitamin-rich meal is not sufficient to reduce oxidative stress measured by 8-iso-PGF2 (40). Engler et al (41) showed that an antioxidant vitamin therapy significantly improved FMD without an effect on biomarkers for oxidative stress, but this too was a long-term study. Unfortunately, there are no data regarding the precise time point at which (after exposure to oxidative stress) 8-iso-PGF2 is increased in the urine. Therefore, it is possible that the complete postprandial extent of oxidative stress has not been identified in our study.
In a study in patients with diabetes (42), a high-fat diet induced an acute increase in plasma ADMA concentration, which has repeatedly been associated with impairment of endothelial function (43-47). In our healthy volunteers, we observed no postprandial change in plasma ADMA concentrations. Furthermore, we observed no correlation of plasma ADMA and FMD or baseline diameter. This is somewhat in contrast to expectations from previous studies. It may in part be explained by the fact that plasma ADMA concentrations in our healthy volunteers were low and in a narrow range. It is, however, in accordance with a previous study in men with stable angina (48). On the other hand, after ingestion of meal 1 and after ingestion of meal 3, we observed both a small but consistent increase in the baseline diameter and an increase in the plasma L-arginine, whereas, after meal 2 (which had the lowest protein/L-arginine content), the plasma L-arginine did not change significantly. However, it is of importance to note that the acute increase in plasma L-arginine obviously did not prevent the postprandial decline in endothelial function. Several studies suggest that L-arginine:ADMA determines NO synthesis and vascular function (49, 50). The postprandial increase of plasma L-arginine was associated with an increase in L-arginine:ADMA after meal 1 and meal 3. Furthermore, we found a moderate but consistent correlation between L-arginine:ADMA and the baseline diameter of the brachial artery. This finding points toward a possible association of L-arginine:ADMA and baseline vascular tone. It is surprising that the improved L-arginine:ADMA seen after meals did not translate into improved FMD. This observation is in line with a previous study showing no association between L-arginine:ADMA or changes therein and endothelial dysfunction in patients with stable angina who were treated with L-arginine (41). Two conclusions, not necessarily exclusive, can be drawn from this observation: first, changes in L-arginine:ADMA in the physiologic range may increase basal NO generation and vascular tone but are not sufficient to increase stimulated NO generation, and, second, factors present in the meals prevent the generation of NO or attenuate the vascular response to NO.
Limitations
This study was designed to address short-term effects of different fast-food meals based on surrogate endpoints. Long-term effects of chronic ingestion of fast food cannot be inferred from the present data, and they remain to be addressed. The present study used publicly sold fast food rather than experimental diets, and it did not attempt to dissect the effects of all single nutrients in the meals or to determine the healthiest mixture of foods for the endothelium.
Conclusions
Against common expectations, a conventional beef burger meal and presumably healthier alternatives such as vegetarian burgers with or without vitamin-rich side orders did not differ significantly in their acute effects on vascular reactivity. The frequently reported postprandial decline in FMD may be attributed at least in part to an increase in the baseline diameter.
ACKNOWLEDGMENTS
The authors thank Mariola Kastner and Elisabeth Silberhorn for their excellent technical assistance.
The responsibilities of the authors were as follows—RM, TKR, and RHB: the conception and design of the study; TKR, KR, and RM: medical assessment of the volunteers, vascular function testing, integrated analysis of all data, and drafting of the manuscript (TKR and RM contributed equally to the manuscript); UR and JT-A: the biochemical analyses and drafting of the Methods section; ES: the final quality control of all biochemical data; and RHB and ES: revision of the manuscript. None of the authors had a personal or financial conflict of interest.
REFERENCES