点击显示 收起
1 From the Cardiovascular Nutrition Laboratory (AHL, NRM, SMJ, NAR, LMA) and the Lipid Metabolism Laboratory (EJS), Jean Mayer US Department of Agriculture Human Nutrition Research Center on Aging, Tufts University, Boston MA
2 Any opinions, findings, conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the US Department of Agriculture. 3 Supported by NIH grant HL 54727 and the US Department of Agriculture under agreement no. 58-1950-4-401. The experimental oils used in this study, with the exception of the partially hydrogenated soybean oil, were a gift from Solae Company, St Louis, MO. 4 Address reprint requests to AH Lichtenstein, Cardiovascular Nutrition Laboratory, Jean Mayer USDA Human Nutrition Center on Aging at Tufts University, 711 Washington Street, Boston, MA 02111. E-mail: alice.lichtenstein{at}tufts.edu.
ABSTRACT
Background: A variety of soybean oils were developed with improved oxidative stability and functional characteristics for use as alternatives to partially hydrogenated fat.
Objective: The objective was to assess the effect of selectively bred and genetically modified soybean oils with altered fatty acid profiles, relative to common soybean and partially hydrogenated soybean oils, on cardiovascular disease risk factors.
Design: Thirty subjects (16 women and 14 men) aged >50 y with LDL-cholesterol concentrations >130 mg/dL at screening consumed 5 experimental diets in random order for 35 d each. Diets contained the same foods and provided 30% of energy as fat, of which two-thirds was either soybean oil (SO), low–saturated fatty acid soybean oil (LoSFA-SO), high–oleic acid soybean oil (HiOleic-SO), low–-linolenic acid soybean oil (LoALA-SO), or partially hydrogenated soybean oil (Hydrog-SO).
Results: Plasma phospholipid patterns reflected the predominant fat in the diet. LDL-cholesterol concentrations were 3.66 ± 0.67b, 3.53 ± 0.77b, 3.70 ± 0.66b, 3.71 ± 0.64a,b, and 3.92 ± 0.70a mol/L; HDL-cholesterol concentrations were 1.32 ± 0.32a,b, 1.32 ± 0.35b, 1.36 ± 0.33a, 1.32 ± 0.33b, and 1.32 ± 0.32a,b mol/L for the SO, LoSFA-SO, HiOleic-SO, LoALA-SO, and Hydrog-SO diets, respectively (values with different superscript letters are significantly different, P < 0.05). No significant effects were observed on VLDL-cholesterol, triacylglycerol, lipoprotein(a), and C-reactive protein concentrations or on ratios of LDL cholesterol to apolipoprotein B (apo B) and HDL cholesterol to apo A-I. Total cholesterol:HDL cholesterol was lower after subjects consumed the unhydrogenated soybean oils than after they consumed the Hydrog-SO diet.
Conclusions: All varieties of soybean oils resulted in more favorable lipoprotein profiles than did the partially hydrogenated form. These soybean oils may provide a viable option for reformulation of products to reduce the content of trans fatty acids.
Key Words: Soybean oils • selective breeding • genetic modification • cardiovascular disease • CVD risk factors • trans fatty acids • LDL cholesterol • HDL cholesterol • triacylglycerol • C-reactive protein • fatty acids
INTRODUCTION
Notwithstanding the difficult goal of achieving and maintaining a healthy body weight, the cornerstone of dietary modification with the intent to prevent or decrease the risk of cardiovascular disease (CVD) is to minimize intakes of saturated and trans fatty acids and to replace them with unsaturated fatty acids (1, 2). The first approach to achieving this goal is to reduce intakes of animal (meat and dairy) and partially hydrogenated fats, which will result in a decreased intake of saturated and trans fatty acids, respectively. The second approach is to use vegetable oils, which are relatively high in cis-unsaturated fatty acids (1-4).
Modern plant husbandry, either through selective breeding or genetic modification, affords the opportunity to alter the fatty acid profile of plants. The former approach has been used since the adoption of modern agricultural cultivation practices. The result was the development of, for example, soybean plants, traditionally rich in polyunsaturated fatty acids, which are high in monounsaturated fatty acids, and rapeseed plants (canola), traditionally rich in monounsaturated fatty acids, which are high in polyunsaturated fatty acids (5, 6). Other modifications in the fatty acid profile of a variety of plants were likewise achieved (7). The primary intent of these modifications was to create trait-enhanced oils characterized by improved functional properties.
Relatively unexplored is the effect of these modifications on indicators of CVD risk traditionally associated with dietary fat. Although the effects of changes in the fatty acid profile of the diet can be estimated with the use of predictive equations (8-10), only actual feeding studies can confirm these estimates. The intent of this work was to assess the efficacy of novel soybean oils with modified fatty acid profiles, relative to soybean and partially hydrogenated soybean oils, on CVD risk factors in middle-aged and older moderately hypercholesterolemic and postmenopausal women and men.
SUBJECTS AND METHODS
Subjects
Thirty subjects (16 women and 14 men) aged >50 y with LDL-cholesterol concentrations >130 mg/dL at the time of screening were recruited for this study from the greater Boston area. All subjects fulfilled the following criteria: had normal kidney, liver, thyroid, and cardiac functions; had normal fasting glucose concentrations; were not taking medications known to affect blood lipid concentrations; were nonsmokers; and were postmenopausal for all women. Subjects using other medications were requested to continue on the same regimen throughout the study period. Subjects were counseled to maintain habitual levels of physical activity. Characteristics of the study subjects at the time of screening are shown in Table 1. Twelve subjects were initially recruited who did not complete the study, 9 of whom terminated during phase 1 and 10 of whom were replaced. Their data were not included in the statistical analysis reported. The reasons were as follows: 3 for time constraints, 4 for noncompliance, 2 for change in medical status (1 subject during phase 4), 1 for loss of medical insurance (phase 2), 1 for move out of state (phase 3), and 1 for dislike of food. Of these subjects 75% terminated participation within the first 2 wk of their start date. The Human Investigation Review Committee of New England Medical Center and Tufts University approved the protocol.
View this table:
TABLE 1. Characteristics of the subjects at the time of screening1
Experimental design
Study subjects were provided with each of 5 experimental diets in random order varying in the main source of fat for periods of 35 d per diet phase. The subjects, investigators, and laboratory personnel were blinded to the order and identification of the diet phases. Subjects reported to our metabolic research unit 3 times per week, had their body weight and blood pressure measured at each visit, and consumed one meal on site. All other food and drink were provided to the subjects in containers appropriate for either microwave or conventional ovens with the intent of obviating the need to transfer food before consumption so as to minimize potential losses. Subjects were required to consume all that was provided and not supplement with any additional food or drink with the exception of water and noncaloric beverages. Initial caloric requirements were estimated by using the Harris-Benedict formula and were adjusted, when necessary, to maintain body weight. The mean (±SD) caloric intake was 2259 ± 291 for the female subjects and 2783 ± 482 for the male subjects. Three times after day 28 of each diet phase, fasting blood samples were obtained for biochemical determinations. The mean value of the 3 time points is reported and was used for statistical analysis.
Diets
The diets were designed to provide 30% of energy as fat with two-thirds of the fat contributed by the experimental oils. This was accomplished by first formulating a diet providing 10% of energy as fat and then for each of the individual diets adding the experimental oil to various food mixtures to achieve the final fat target of 30% of energy. These foods were incorporated into breakfast, lunch, and dinner menus. The experimental oils were provided by Solae Company (St Louis, MO) and were soybean oil (SO), low–saturated fatty acid soybean oil (LoSFA-SO) developed by selective breeding, high–oleic acid soybean oil (HiOleic-SO) developed by genetic modification, and low–-linolenic acid soybean oil (LoALA-SO) developed by selective breeding. Partially hydrogenated soybean oil (Hydrog-SO) was a commercially available product (Whirl; Proctor and Gamble Company, Cincinnati, OH). Nutrient analysis was determined by Covance Laboratories America Inc (Madison, WI).
Biochemical analysis
Blood samples were collected after a 12-h fast. Serum was separated by centrifugation at 1100 x g at 4 °C for 20 min and assayed for total cholesterol, HDL cholesterol, LDL cholesterol, triacylglycerol, and high-sensitivity C-reactive protein (CRP) with the use of Roche Diagnostics reagents (Indianapolis, IN) and apolipoprotein A-I (apo A-I), apo B, and lipoprotein(a) [Lp(a)] with the use of Wako Diagnostics reagents (Richmond, VA) with a Hitachi 911 automated analyzer (Ingelheim, Germany). VLDL cholesterol was calculated as
RESULTS
The experimental fats were distinguished from each other by their fatty acid profiles (Figure 1). Relative to the SO diet, the LoSFA-SO diet had 50% the content of saturated fatty acids, primarily attributable to 16:0. The HiOleic-SO diet had 4 times the amount of oleic acid, representing 85% of total fatty acids, primarily at the expense of polyunsaturated fatty acids and to a lesser extent saturated fatty acids. The -linolenic acid content of the LoALA-SO diet was reduced to 54% of the SO diet. Note, this concentration of -linolenic acid was similar to that in the HiOleic-SO and Hydrog-SO diets. The fatty acid profile of the Hydrog-SO diet was distinguished by the relatively high concentration of trans fatty acids, 13% of the total fatty acids, and a shift in the fatty acid profile of the fat from polyunsaturated to monounsaturated fatty acids and a lesser extent saturated fatty acids.
FIGURE 1.. Fatty acid composition of the experimental fats as determined by chemical analysis. SFA, saturated fatty acid; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; FA, fatty acid; SO, soybean oil; LoSFA-SO, low–saturated fatty acid soybean oil; HiOleic-SO, high–oleic acid soybean oil; LoALA-SO, low–-linolenic acid soybean oil; Hydrog-SO, partially hydrogenated soybean oil. *Excludes 18:3n–3.
By design, the total fat content of the diet was 30%, two-thirds of which was represented by the experimental fat (Table 2). Likewise, the contents of protein, carbohydrate, cholesterol, and fiber were similar among diets. The fatty acid profiles of the diet reflected those of the experimental fats, albeit to a less dramatic extent because of the influence of the other components of the diet. The exception was the LoALA-SO–enriched diet in which the difference in the -linolenic acid content attributable to the oil itself was obscured.
View this table:
TABLE 2. Nutrient composition of the experimental diets as determined by chemical analysis of food1
The phospholipid fraction of plasma isolated at the end of each diet phase reflected, to a limited extent, the predominate fat in the diet and provided one indicator of dietary compliance (Table 3). Little effect of the diets was observed on the relative proportion of saturated fatty acids, likely attributable to endogenous synthesis. In contrast, after the subjects consumed the diet enriched with the HiOleic-SO, cis-monounsaturated fatty acids represented 17% of the total fatty acids, which was 40% higher than after the subjects consumed the SO, LoSFA-SO, and LoALA-SO diets. cis-Monounsaturated fatty acids were intermediate after subjects consumed the Hydrog-SO diet, relative to the HiOleic-SO diet and the other 3 oils, indicative of the shift in the fatty acid profile of soybean oil after partial hydrogenation. The pattern observed with monounsaturated fatty acids was complementary to that observed for cis-polyunsaturated fatty acids. Concentrations were lowest after subjects consumed the HiOleic-SO diet, intermediate after they consumed the Hydrog-SO diet, and highest after they consumed the SO-, LoSFA-SO–, and LoALA-SO–enriched diets. Plasma phospholipid trans fatty acids were 3.5 times higher when subjects consumed the Hydrog-SO–enriched diet relative to the other diets. These patterns were similar for women and men.
View this table:
TABLE 3. Phospholipid fatty acid profiles at the end of each diet phase1
The effect of the soybean oils modified in fatty acid profile on plasma lipid and lipoprotein concentrations was modest, with few exceptions. LDL-cholesterol concentrations were highest after the subjects consumed the Hydrog-SO diet and LoALA-SO diet than after the other diets (Table 4). LDL-cholesterol concentrations while consuming the LoSFA-SO and HiOleic-SO diets were lower than that observed relative to the Hydrog-SO diet but not significantly different relative to the SO diet. Relative to the SO diet, the percentage of difference in LDL-cholesterol concentrations were –3.2% for the LoSFA-SO diet, 1.4% for the HiOleic-SO diet, 0.8% for the LoALA-SO diet, and 5.6% for the Hydrog-SO diet (Figure 2). This pattern of response to the different diets was reflected in plasma apo B concentrations. In contrast, HDL-cholesterol concentrations were not significantly different between the diets of modified fats and either the SO diet or the Hydrog-SO diet, although for men, HDL-cholesterol concentrations on the HiOleic-SO diet were significantly greater than with the SO diet.
View this table:
TABLE 4. Lipid and lipoprotein concentrations at the end of each experimental diet phase1
FIGURE 2.. Mean (±SE) percentage change in plasma lipid and lipoprotein concentrations relative to the soybean oil (SO) diet. Changes were analyzed by ANOVA with main effect of diet and subject as repeated measures followed by Tukey's t test for multiple comparisons. Error bars with different lowercase letters are significantly different, P < 0.05.
Relative to the SO diet, these differences were modest, ranging from –0.6% to 4.1%. The difference in HDL-cholesterol concentrations was attributable primarily to the HDL3 subfraction. Apo A-I concentrations mirrored those of HDL cholesterol. As a result of the differential effects of the dietary fats on total and HDL-cholesterol concentrations, the ratio of total cholesterol to HDL cholesterol was least favorable (higher) after the subjects consumed the Hydrog-SO and LoALA-SO diets (4.0% and –0.1% relative to the SO diet, respectively). The ratios of LDL cholesterol to apo B and HDL cholesterol to apo A-I were similar at the end of the diet phases, suggesting that there was little change in the composition of the lipoprotein particles. No significant effect of the dietary perturbations was observed on VLDL-cholesterol, triacylglycerol, Lp(a), or CRP concentrations. The responses of the female and male subjects were similar to the dietary perturbations.
DISCUSSION
Over the years there has been and continues to be a shift in the fatty acid profile of plant oils (6, 15-18). The primary intent is to improve the functionality of the fats, such as physical characteristics or chemical stability. One of the approaches to achieve this end was to hydrogenate liquid oil. The result was decreased fluidity (liquid to semisolid or solid) and reduced susceptibility to oxidation. These changes were attributable to an increase in the concentration of trans fatty acids and a shift in the proportion of fatty acids from polyunsaturated to monounsaturated and saturated. In the 1990s concern was raised about the effects of trans fatty acids on lipoprotein profiles and CVD risk factors (19-22). This resulted in an increased interest in identifying an alternate fat that had physical characteristics similar to those of partially hydrogenated fat without the undesirable biological effects.
In addition to minimizing trans fatty acids there are several other reasons to modify the fatty acid profile of vegetable oils. Saturated fat intake is positively associated with LDL-cholesterol concentrations (1). LDL-cholesterol concentrations are positively associated with risk of developing CVD (1). Decreasing the relative proportion of saturated fatty acids would make an oil or food product made thereof more desirable from a "heart health" perspective. Increasing chemical stability and altering the physical characteristics of fats are other reasons to modify the fatty acid profile of oils. Highly unsaturated fatty acids increase the susceptibility of the fat to oxidation (23). Oxidized fats used for food preparation (ie, frying fat) or formed in packaged foods with increased storage time impart undesirable flavors and odors (24). Two approaches to optimize fats from the perspective of chemical stability are to decrease the proportion of highly unsaturated fatty acids by either shifting the relative proportion of fatty acids from polyunsaturated to monounsaturated and saturated by partial hydrogenation or to cultivate plants with a reduced content of fatty acids most susceptible to oxidation. As early as the mid-1960s vegetable fats with modified fatty acid profiles began to appear (25).
Concomitant with the development of new varieties of oil-yielding plants is a dearth of information about the effect of these oils on health indicators commonly associated with the fat component of the diet. The results of the current investigation, which focused on CVD risk factors, suggest that the altered fatty acid profile of soybean oils currently available when fed at relatively high amounts, with the exception of partial hydrogenation, had no or a modest effect on the indicators assessed, either positive or negative. Additional data are limited on this issue. Wardlaw and Snook (26) compared the effect of butter with corn and high-oleic sunflower oils with respect to lipoprotein concentrations. The vegetable oils had similar effects, and both resulted in significantly lower total and LDL-cholesterol concentrations than did butter. In a direct comparison of a standard and selectively bred low linolenic acid soybean oil on serum lipid concentrations, Lu et al (27) reported small differences and no significant effect on the ratio of LDL to HDL. More recently, Allman-Farinelli et al (28) compared high oleic acid sunflower oil with saturated fat and observed results comparable to what would be expected from a conventional oil rich in monounsaturated fatty acids.
Dietary fat in any typical day comes from a wide variety of foods. For the current study the system was exaggerated by providing 20% of energy from each of the experimental fats. This was done to make it more likely that, if there was a difference in response among the test fats, it would be of a sufficient magnitude to be observed. It was not done to actually mimic the effect of some of the newer varieties of vegetable oils as currently present in the food supply. When any of these oils were used in place of Hydrog-SO, a positive effect was observed on the total and LDL-cholesterol concentrations. The magnitude of difference compares favorably with those previously reported when SO or corn oil were compared with their partially hydrogenated counterparts (20, 29). Effects on HDL-cholesterol concentrations were small. It was previously reported that trans fatty acids lower HDL-cholesterol concentrations (19, 20, 29). This effect of trans fatty acids on HDL-cholesterol concentrations was predominately reported relative to saturated not unsaturated fatty acids, as were the comparison fats in this study. As recently shown from lipoprotein kinetic studies, the differences seen in response of HDL cholesterol to saturated and trans fatty acids is more likely a lack of the ability of trans fatty acids to raise HDL-cholesterol concentrations as do saturated fatty acids than their actual ability to lower HDL-cholesterol concentrations (30). Nevertheless, the higher total cholesterol concentrations resulting from the Hydrog-SO diet resulted in the highest mean total cholesterol:HDL cholesterol at the end of that diet phase, hence, was the least favorable option.
Modifying the fatty acid profile of soybean oils by selective breeding, genetic modification, or partial hydrogenation had no significant effect on CRP concentrations. In general, dietary fat has little effect on CRP concentrations, with the exception of very long chain n–3 fatty acids (eicosapentaenoic and docosahexaenoic acids) which have been reported to decrease CRP concentrations (31, 32). A similar effect was not observed with a plant-derived n–3 fatty acid, -linolenic acid.
-Linolenic acid (18:3n–3) is the most common and quantitatively important highly unsaturated fatty acid present in the main vegetable oils used for food preparation in the United States (soybean and canola oils). -Linolenic acid is a member of the n–3 fatty acid series and is an essential fatty acid. The exact requirement has yet to be determined, in part because -linolenic acid is one of several essential fatty acids, some of which serve as precursors for others. -Linolenic acid can be converted, albeit at a low rate, to longer chain n–3 fatty acids, eicosapentaenoic acid and docosahexaenoic acid (33). Although considerable effort has been directed at determining the effect of -linolenic acid intakes on a range of health outcomes, for the most part the data are equivocal (33-37). The results of this intervention confirm this observation; essentially decreasing -linolenic acid content of the soybean oil by half had no significant effect on the measured indicators.
There are several limitations of this study. To isolate the potential effect of each of the unique soybean oils, diets were designed to maximize potential differences by exclusively using a single oil as the predominant fat and keeping the other components of the diet constant. Likewise, the study was not designed to assess the effect of a 1:1 substitution of one fatty acid for another but rather the effect of displacing one type of fat with another within the context of a heart healthy diet. Subjects represented a relatively narrow range of the US population. Specifically recruited were older moderately hypercholesterolemic subjects. This specific type of subject was selected because these subjects are the ones frequently targeted for dietary intervention and most likely to be affected were an effect to be observed.
The effect of novel soybean oils with altered fatty acid profiles resulted in, for the most part, plasma lipid, lipoprotein, apolipoprotein, Lp(a), and CRP concentrations that were similar to those of soybean oil. Some of the modifications, such as the reduction in the saturated fatty acid content, resulted in nonsignificant trends toward lowering LDL-cholesterol concentrations relative to the other nonhydrogenated soybean oils. However, the magnitude of the difference was small in the oil tested. All resulted in more favorable lipid and lipoprotein concentrations than did partially hydrogenated fat and hence are viable alternatives because the food industry is poised to phase out trans fatty acids in their products.
ACKNOWLEDGMENTS
We thank the staff of the Metabolic Research Unit for the expert care provided to the study subjects and gratefully acknowledge the cooperation of the study subjects, without whom this investigation would not have been possible.
AHL was the principal investigator for this study and wrote the initial draft of the manuscript. All other authors contributed to critically reviewing the manuscript. NRM was responsible for the fatty acid analysis. NAR provided technical assistance. SMJ was involved in all aspects of the biochemical analysis. EJS was involved in the design of the study. and LMA was responsible for the statistical analysis. None of the authors had a conflict of interest.
REFERENCES