| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letter |
Nutrition Research Division, Food Directorate, Health Products and Food Branch, Health Canada, Ottawa, Ontario, CANADA
PROTECTIVE EFFECTS OF FLAX MEAL AGAINST HYPERCHOLESTEROLEMIA AND HYPERTRIGLYCERIDEMIA IN RATS
The paper by Bhathena et al. [1] in a recent issue of the Journal concluded that flaxseed meal may have important therapeutic implications in patients with hypertriglyceridemia and hypercholesterolemia and deserve further study with humans with these disorders. We certainly find this conclusion very interesting, but before pursuing human studies it is imperative to repeat the animal study, because as indicated below, the study by Bhathena et al. [1] had several serious fundamental flaws in experimental design.
1. The study used growing rats, and the American Institute of Nutrition (AIN) has recommended that the diet of growing rats should contain 7% fat as soybean oil. Soybean oil is a good source of both linoleic and 
-linolenic acids, which are essential fatty acids for growth and proper function of cell membranes. AIN estimated that 12 g of linoleic acid and 2 g/kg of -linolenic acid were the minimal requirements for rats [2]. Contrary to AIN recommendations, this study used corn oil as the source of fat and at a level of 4% of total diet. Corn oil contains no -linolenic acid; therefore, the three diets used in this study were deficient in -linolenic acid. It is not known to what extent this deficiency influenced the results of this study.
2. It is known that commercially-prepared flaxseed meal does contain some residual fat. Unfortunately, this study did not characterize the test protein sources including flax meal and soy protein concentrate, as well as the final diets for fat content, fatty acid and amino acid composition. This is a fundamental deficiency of this study, because even a small amount of residual -linolenic acid (which is possible in flaxseed meal, because of the extremely high content of -linolenic acid in flaxseed) could influence serum triglyceride levels. It is known that -linolenic acid and other long-chain n-3 fatty acids can effectively lower serum TG. Thus it is probable that the lower serum TG levels seen in animals fed flaxseed meal diet might be partly due to the presence of -linolenic acid.
3. The authors stated that the experimental diets were prepared according to the specifications of AIN-93 and that the three protein sources were added as the sole source of protein to provide 20% calories from protein. However, it was not shown whether the protein composition of AIN-93G (20% protein) or AIN-93M diet (14% protein) was simulated in the present study. In the absence of this information and the lack of data on proximate composition of the test protein sources, it is difficult to figure out the amount of protein sources, especially flax meal, in the experimental diets.
4. Flaxseed meal protein is known to be limiting in lysine for rat growth [3]. Although the differences in growth among the three dietary groups were apparently small, total protein in rats fed the flax meal diet was significantly lower than in those fed the other two protein sources, i.e. casein and soy protein (Table 2), suggesting lower protein status in rats fed the flax diet. Moreover, flax protein contains considerably higher cystine than does casein (2.0% vs. 0.5% of protein). Addition of 0.18% supplemental cystine to all the experimental diets, as done in the present study, would further increase the cystine content of the flax diet. Therefore, the higher cystine content of the flax meal diet would have been partly responsible for its hypocholesterolemic effect, as a significant (p < 0.05) negative correlation between dietary cystine and blood cholesterol has been reported in rats fed different animal and vegetable proteins [4].
5. The authors concluded that the substitution of flaxseed meal or soy protein for casein in the diet significantly reduced fat accumulation in the liver of obese SHR/-cp rats. We disagree with this conclusion, because it was based on an assessment of liver fat by oil red O stained sections. Since fat is not uniformly distributed in the liver, assessment based on oil red O does not reveal the real amount of fat in the liver. The actual amount of fat can be determined accurately only by chemical analysis.
6. The authors should explain why there was no correlation between body weight and liver weight in rats fed the flax meal diet (Table 2). Usually liver weight increases with increased body weight as noted in the case of casein and soy protein fed animals. The lower liver weight in animals fed the flax meal diet may perhaps suggest a deleterious effect.
7. The authors noted that flax meal had a significant cholesterol lowering effect compared to casein control. However, they did not mention that the HDL/total cholesterol ratio in the flax meal group (0.20) was lower than in the soy (0.25) or the casein group (0.23).
8. Flax meal is also known to contain a vitamin B-6 antagonist called linatine [5]. Normally, in animal studies, diets containing high levels of flax meal are supplemented with extra vitamin B-6, which apparently was not done in the present investigation. Moreover, the effect of the presence of a large amount of mucilage in the flax meal diet was not addressed.
The above-noted points must be addressed to enable other researchers to duplicate this study and to assess the validity of the results and the study’s potential application to human nutrition.
REFERENCES
Beltsville Human Nutrition Research Center, Agricultural Research Service, U.S. Department of Agriculture, Beltsville, Maryland (S.J.B.)
Department of Food Science, Ain Shams University, Cairo, Egypt (A.A.A.)
The Jerome Holand Laboratory, American Red Cross, Rockville, Maryland (C.H.)
Department of Medicine, The George Washington University Medical Center, Washington, DC (P.L., T.R., M.T.V.)
Virginia State University, Petersburg, VA (A.I.M.)
National Institutes of Health, Animal Genetic Resource, Bethesda, Maryland (C.H.)
We appreciate the critique of our paper by Drs. Ratnayake and Gilani. However, we strongly disagree that our study has several serious fundamental flaws in experimental design.
1. Our study was designed to compare soy protein concentrate containing isoflavones and flaxseed meal containing lignans with casein. In order to study the differences in the effects of protein sources (and phytonutrients), we purposely used corn oil. Use of soybean oil may complicate the interpretation as to whether the effects observed are due to n-3 fatty acids. It is important to note that most studies before 1993 (when recommendation for use of soybean oil came out) and several studies since then used oils other than soybean oil in rat diets. Corn oil was the recommended source of oil for rat diets recommended in AIN-76 and AIN-76A diets. Corn oil does contain a small amount of -linolenic acid.
Ratnayake and Gilani go on to state that "it is not known to what extent this deficiency (of -linolenic acid) influenced the results of this study." However, the majority of studies where soybean oil has not been used have not reported any deleterious effects. Several papers published by Ratnayake and colleagues since 1993 report the use of oils other than soybean in control diets for experiments comparing the effects of various oils and other dietary ingredients on metabolic and other parameters [1,2]. These studies found no significant differences of these diets on body weight. Our manuscript quoted another paper by Ratnayake et al. (Ref. 15) [3] in which corn oil and lard (not soybean oil) were used for control diets fed to rats to study the comparative effects of various levels of flaxseed in the diet. Their own data did not show any significant effect of these control diets on food intake, body weight and organ weights (liver, kidney and heart). In addition, their data showed no significant difference in serum lipids between rats fed corn oil (with only a small amount of -linolenic acid) and those fed 10% flaxseed containing -linolenic acid and no difference in blood glucose levels between control rats fed corn oil and those fed 10% to 40% flaxseed. Thus, low levels of -linolenic acid do not appear to have an effect on lowering cholesterol or triglyceride levels in rats. Based on the data of Ratnayake et al. and others, it seems unlikely to us that our findings in these experiments are much influenced by a deficiency of -linolenic acid in the corn oil.
2. Ratnayake and Gilani also criticize our study for failing to analyze our diets for fat content, fatty acid and amino acid composition. They state that "... it is probable that the lower serum TG levels seen in animals fed flaxseed meal diet might be partly due to the presence of -linolenic acid." We cannot rule out the possibility since the design of our study did not include an analysis of the effects of different amino acids or fatty acids on lipid profile. However, we think that a significant effect of linolenic acid is unlikely, since other studies by Ratnayake et al. [3] have shown no difference in serum triglyceride between rats fed corn oil (with no added -linolenic acid) and those fed 10% flaxseed containing a significant amount of -linolenic acid. On the other hand, Ratnayake et al. [3] also failed to provide an analysis of amino acids in the flaxseed diet used in their study, so any specific contributory effect of linolenic acid to the effects of flax has not actually been determined.
3. The next point raised by Drs. Ratnayake and Gilani is confusing. They raise concern about the composition of the protein sources in the diets used in our study and question whether the protein compositions were simulated. We have compared the effects of proteins from three sources at 20% of energy level, namely casein (AIN-93), soy protein concentrate and flaxseed meal. The protein content of the diets was adjusted by the percentage of energy content. The amino acid composition is expected to be different. The nature of the question is unclear.
4. The probable deficiency of lysine in flaxseed meal in our study had no significant effect on growth as assessed by weight gain. The weight gain of rats fed flaxseed meal was similar to that in rats fed either casein or soy protein concentrate in both lean and obese rats. In fact, in obese rats, weight gain was higher in rats fed flaxseed meal than in those fed casein or soy protein concentrate. There were no significant differences in plasma protein in lean rats fed three different dietary proteins (Table 1). Similarly, in obese rats no significant difference was observed in plasma total proteins between rats fed flaxseed meal and those fed soy protein concentrate. It is important to note that lean and obese animals respond quantitatively differently to different diets. In the study by Sarwar et al., quoted by Ratnayake and Gilani, no apparent difference was observed in weight gain or food intake between rats fed casein and flax (lacking in lysine). Since no standard error or statistical analysis was provided, it is difficult to see whether differences in weight gain (10.4 g vs. 9.3 g) were significantly different.
We do not think that addition of 0.18% cystine had any significant effect on plasma cholesterol levels. In the study by Sautier et al. quoted by Ratnayake and Gilani, no difference in plasma total cholesterol or HDL-cholesterol (Table 3) was observed in rats fed casein (low in cystine) and those fed lactoglobulin, ovalbumin, alfalfa and gluten (containing three to five times more cystine than casein) (Table 1), clearly showing no effect of cystine level. To our knowledge, no study has added lysine to flaxseed, including the one by Ratnayake et al. [3]. Our study supports the findings of several other studies in rats that suggest that the cholesterol-lowering effects of flaxseed meal and soy protein concentrate are more likely due to lignan SDG in the case of flaxseed and to the type of protein and isoflavones in the case of soy protein.
5. The next criticism concerns the use of histological assessment as a means of quantifying steatosis (fat) on liver biopsy. This means of assessing fat is a standard approach for determining the extent of fatty change in hepatocytes. In general, fatty change is a diffuse process in the liver. Although it is possible for a needle biopsy or small sampling of liver to misrepresent the extent of fatty change in liver tissue, it is unusual in the absence of localized effectors or significant underlying liver disease such as cirrhosis. The protocol used in our study assessed two generous wedge biopsies of liver tissue from each animal. One sample of liver was formalin-fixed, paraffin-embedded and stained with H&E, while the second biopsy underwent frozen-section and staining with oil red O. Both specimens were evaluated for extent of fatty change. There was good relative agreement between the two methods of assessment, although oil red O seemed to be more sensitive and easier to read than the H&E sections. The fat seen in these samples was uniform in distribution across several lobules. It seems to us unlikely that these samplings were not representative of the extent of relative fatty change in each liver.
6. The critique also questions why there was no correlation between body weight and liver weight in rats fed the flaxseed meal diet. It is interesting that the loss of fat in the liver did not always correlate with a loss in body weight as seen in two recent studies. In human subjects with non-alcoholic steatohepatitis (NASH) treated with pioglitazone, there was a reduction in liver fat but an increase in body weight [4]. In an experimental animal model, Zucker rats given powdered green tea, there was an increase in the concentration of total lipids in liver, but a decrease in liver weights [5]. Further studies are needed to better understand the meaning of these observations, but dissociations between liver fat (weight) and whole body fat do occur. It is well established, however, that substantial changes or improvement in the metabolic parameters in obesity may occur with only minimal changes in total body weight or fat, which may be related to differences in responsiveness in hepatic or visceral fat and subcutaneous and extra abdominal fat.
In our experiments, we found the discordance between body weight and liver weight to be most remarkable in the obese rats (Table 2), which have significant fat deposits in various fat depots other than liver. In the lean rats (Table 1), there were no significant differences between body weight or liver weight. The lower liver weight in obese rats fed flaxseed meal is associated with a decrease in liver fat content, as stated above. A mere reduction in liver weight induced by dietary modification does not necessarily reflect a deleterious effect of the diet. In fact, the lower liver weight in rats fed flaxseed meal may be a beneficial effect, along with other beneficial effects that have been reported in the settings of obesity, diabetes and renal disease, summarized by us in previous publications [6,7].
7. Drs. Ratnayake and Gilani correctly note that the HDL/cholesterol ratio was lower in rats fed the flax meal diet than the other diets. As Table 2 demonstrates, total cholesterol and all cholesterol fractions are lower in rats fed flaxseed meal than those fed casein diet. No significant differences were observed in the HDL/total cholesterol in lean rats (Table 1). We also observed significant decreases in HDL cholesterol, a finding different from that in many other studies. As stated above, obese animals respond quantitatively differently from lean phenotypes.
8. Drs. Ratnayake and Gilani refer to a paper from 1967 which describes the presence of a vitamin B6 antagonist, linatine, in flaxseed. Linatine has been shown to have a deleterious effect in chicks, but not in rats and mice. To our knowledge, no study using flaxseed has supplemented the diet with vitamin B6, including the study by Ratnayake et al. [3].
We believe that our results on the effects of flaxseed clearly support the findings observed in previous animal and human studies, as referred to and discussed in our paper (Refs. 15–19, 31, 34–36, 46 and 47).
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
