There is probably no subject in history that has received more print than “fat.” First, let me say that I’m not going to talk about “fat” à la “obesity”, or “fat” à la “how to get rid of it” via exercise/diet. What I am going to talk about is the science of dietary fat—what we think we may know about different fats and their role in the human diet. In my opinion, the science here is not “settled,” because the weight of evidence supporting one view or another is constantly shifting. Frankly, I’m confused.
In fact, when it comes to fat, there is a lot to be confused about. For one thing, the word “fat” is almost meaningless. The only way to understand what the word means is by the context in which it is used. For example, if someone refers to “fat” in their diet, you don’t know if they mean lard, oil, lipids, a specific fatty acid, a fatty acid supplement, or a mixture of some or all of these. If someone refers to “fat” in their blood, you don’t know if they mean triglycerides, cholesterol, or plaque build-up. I’d be just as happy if “fat” was eliminated from our vocabulary, as well as from our waists.
In sum, I’m more certain about the discovery of water on Mars than I am about the role of fat in the human diet.
And the subject is so complicated I think I’ll make it the topic of at least the next two blogs. Sorry. And, I can tell you right now my objective is to wade through studies to determine if there is an emerging consensus—OK, if you read no further, I’ll tell you: there isn’t one. At least I don’t see it. OR if there is a consensus it is very weak, so far.
For our purposes, the best way to think about fat is in terms of what it is made of: it consists of two kinds of molecules—fatty acids and glycerol. OK, we don’t care about glycerol (at least I think we don’t). What we REALLY care about are fatty acids.
What makes it confusing is that there are more than 20 different fatty acids. And they have frightening common names: oleic acid, palmitic acid, linoleic acid, linolenic acid, stearic acid, etc. Further, these fatty acids can be subdivided into “saturated” fatty acids and “unsaturated” fatty acids. Unsaturated fatty acids can be further divided into “polyunsaturated” fatty acids and “monounsaturated” fatty acids. And unsaturated fatty acids can also be described as “cis-“ and “trans-“ fatty acids.
Saturated fat. Trans-fat. Polyunsaturated fats. Peanut oil. Olive oil. Corn oil. Soybean oil. Lard. Margarine. All of these differ from each other by the types and amounts of fatty acids they contain.
By looking at the recommendations of the American Heart Association, the World Health Organization, the American Dietetic Association, the British Heart Foundation, the World Heart Federation, the Food and Drug Administration, and others, you would assume that there is a scientific consensus regarding appropriate amounts and types of dietary fat.
I’m sticking my neck out here, but after looking at the literature I am totally convinced that this is not the case. Let me say it in another way: in my opinion, there is no consensus by the scientific establishment on many of the recommendations concerning the consumption of fat that we all thought were gospel.
For example, we have all heard that we should eat less “saturated” fat and eat more “polyunsaturated” and “monounsaturated” fat. Following these recommendations, we have reduced our consumption of meat, cheese, butter, lard, pizzas, doughnuts, sausages, and cream (all high-saturated-fat foods), and increased our consumption of skim milk, nuts, chicken, fish, soybeans, margarine, and vegetables (all low-saturated-fat foods)—right? Isn’t that what we’ve been told to do for years now?
Clearly we need to take a look at the studies supporting these recommendations. But first, let’s define what we mean by “saturated” fat, “polyunsaturated” fat, and “monounsaturated” fat.
Simply, “saturated fat” means a fat that is composed of fatty acids that don’t have any double bonds between the carbon atoms. Simple as that. There are five common “saturated” fatty acids (butyric acid, lauric acid, myristic acid, palmitic acid, and stearic acid). For example, stearic acid has 18 carbon atoms, each of which is linked to another by a single chemical bond, and palmitic acid has 16 carbon atoms, also linked by single bonds. The fat found in ground beef, for example, is composed of 15% stearic acid, 26% palmitic acid, and 4% myristic acid. So 45% of the fat content of ground beef is saturated fat. Therefore, the remaining 65% of the fat in ground beef is polyunsaturated and monounsaturated fat.
What is monounsaturated fat? It is a fatty acid that has only one double bond. There are three common monounsaturated fatty acids: palmitoleic acid (with 16 carbon atoms), and vaccinic acid and oleic acid (each with 18 carbon atoms). Each of these fatty acids has only one place where the linkage between two carbon atoms consists of two bonds (a “double” bond). Olive oil is a good example of a fat that is high in monounsaturates (75%).
Finally, polyunsaturated fats. These are composed of fatty acids in which there is more than one double bond between the carbon atoms. For example, the famous “omega 3 fatty acid” that is found in fish oil and that we’ve all been advised to ingest more of, is actually a family of polyunsaturated fatty acids, the most famous being alpha-linolenic acid, which has 18 carbon atoms and 3 double bonds. The “omega 6 fatty acid” is another family of polyunsaturated fatty acids, of which the most famous is probably linoleic acid (not to be confused with the linolenic acid found in omega 3 above!). Linoleic acid is 18 carbon atoms long and has 2 double bonds.
Whew. I don’t know how anyone could keep this stuff in their heads for more than 5 seconds. So, for now just remember that fatty acids are either saturated, monounsaturated, or polyunsaturated.
Now let’s look at the recommendation we are most familiar with: “Thou shalt reduce the amount of saturated fat in your diet.” What I mean is, let’s look at the scientific underpinnings of this recommendation, which is practically gospel in most nutrition circles.
Well, since the 1960’s, dozens and dozens of studies have been performed to look at saturated fats. And, indeed, many studies show that reducing saturated fats results in a reduction in both mortality and cardiovascular disease. However, other studies show no relationship. So is there a consensus today?
One way at getting at the answer to this question is to look for what’s called a “meta-analysis.” A meta-analysis is a study that groups many studies together and then performs a statistical analysis of the grouped data. In this way you can go from an individual study of say 100 people (which I’m sure is too small to say anything definitive about dietary fat) to a single, comprehensive analysis of many grouped studies and thereby get your sample size up into the thousands. The goal is to have a study population that is large enough to say something definitive, that is, to do a statistical analysis that is powerful enough to identify statistically significant trends.
And for this reason I like meta-analyses. I mean, how else can one possibly answer the question, “What do the studies say?” without looking at each of the studies individually? A criticism of meta-analyses is that since they group many studies, they may lump sexes, races, ages, specific disease considerations, dietary interventions, and research methodologies together. Also, different studies have different numbers of people enrolled as test subjects—so a study based on 10 people should not have the same “weight” as a study with 10,000 people. But, some of these problems can be controlled by the scientists who are doing the meta-analysis: They can specify that they will only look at studies that meet specific criteria.
So, let’s look at a meta-analysis of studies that looked at the saturated fat question. I have found three recent such meta-analyses.
Meta-analysis #1*, published in 2011. This particular meta-analysis only included studies of adults 18 years or older in randomized experiments, with or without risk of cardiovascular disease, that included both males and females (pregnant or lactating women excluded). The authors settled on 60 studies with publication dates between 1965 and 2009, including 30 studies in North America, 26 in Europe, 3 in Australia/New Zealand, and one in the Middle East. There were approximately 70,000 total participants in the 60 studies combined.
For our purposes here, we can say that in one group of studies, saturated fat was reduced by approximately 30% of total calories. The test subjects made up for the resulting calorie deficit by eating more carbohydrates, fruits, vegetables, and/or protein.
The authors’ conclusion? The reduction in dietary saturated fat had no statistically discernable effect on mortality, stroke, heart attacks, “cardiovascular events,” cancer, or diabetes.
However, when polyunsaturated fats or monounsaturated fats were used in place of carbohydrates, vegetables, and fruits to bring the total calories back up to normal, there was a 14% reduction in “cardiovascular events”, but no specific reduction in strokes or heart attacks. The term “cardiovascular event” was not further defined.
Meta-analysis #2**, published in 2010. 21 studies were included in the analysis, with each study covering 5 to 23 years, and a combined total of 347,747 subjects. 11 studies included men only, 2 studies included women only, and 8 included both men and women. 12 studies were conducted in North America, 6 in Europe, 2 in Japan, and one in Israel.
There was wide variation in dietary modification or the lack thereof—some of the studies actually modified the subjects’ diets, and others were based on surveys of the subjects’ behavior without any attempt at dietary modification.
Conclusion? There was no evidence that reducing consumption of saturated fat was associated with a decrease in cardiovascular disease.
The authors of this meta-analysis point out that several individual studies have shown that reducing saturated fat and increasing polyunsaturated fat in the diet is beneficial to coronary heart disease, and that more benefits are seen as the amount of polyunsaturated fat increases and the amount of saturated fat decreases. The authors’ own meta-analysis did not reach this conclusion, however.
So, there is a suggestion in both of these studies that reducing saturated fats and increasing polyunsaturated fats MAY have a beneficial effect on cardiovascular disease.
Which brings me to the next study.
Meta-analysis #3***, published in 2010. 8 studies, 13,614 participants. The average length of the studies was 4.25 years. In these studies, the test subjects REPLACED saturated fat with polyunsaturated fat in their diets. On average, the “polyunsaturated” fat group got 15% of their energy from polyunsaturated fats, whereas on average the “saturated” fat group got only 5% of their energy from polyunsaturated fats.
The authors concluded that there was a 19% reduction in heart attacks or deaths due to cardiovascular events among the “polyunsaturated” fat group in comparison with the “saturated” fat group.
So, two of these three meta-analyses indicate that reducing saturated fats and increasing polyunsaturated fats may be beneficial.
Now for something controversial. If no one responds to this, I’ll know that nobody has read this blog! Here is the deal: These three studies also reported on summary results for the individual studies, and SOME of the individual studies showed that THE REDUCED SATURATED-FAT DIETS WERE CORRELATED WITH AN INCREASE IN DEATHS/CARDIOVASCULAR EVENTS. Let me say that a different way: Some (many) of the individual studies showed that decreasing saturated fats in the diet was actually HARMFUL.
What is interesting is that none of the authors discussed this. Is that because it would have been too controversial? Is that because no one wants to advocate a diet high in saturated fat, when we have been told for decades that saturated fat is evil? I do not know.
Further, an analysis was done in the second meta-analysis to test for reporting bias, and the authors concluded that there may be some. That is, studies showing negative results may be under-reported. In this case, a finding that decreasing your consumption of saturated fat is bad for you would be considered a negative result.
What do I think? I don’t know for sure, but I suspect that too few subjects have been analyzed thus far and that the dietary interventions used in these studies were not detailed enough. For example, suppose that in order to lower intake of saturated fat, you ask the test subjects to reduce their consumption of red meat. The problem is that red meat is about 50% saturated fat and 50% monounsaturated fat. So when you analyze the results, are you seeing the effects of reduced saturated fats or reduced monounsaturated fats? Further, there are many different types of polyunsaturated fats—some may be “good,” and some may be “bad.” If a test subject decreases consumption of saturated fat and increases consumption of a “bad” polyunsaturated fat (such as, perhaps, the omega 6 fatty acids found in safflower oil), then the two may cancel each other out. Who knows?
Do I think there is something counter-intuitive is going on here with regard to consumption of fatty acids? There is certainly that suggestion. Do I think you should conclude that increasing saturated fat in your diet is good for you? No, the data does not say that, and neither am I. Am I saying that consumption of saturated fat is neutral (neither good nor bad)? Don’t know that either. Do I think increasing SOME TYPES OF polyunsaturated fats may be beneficial? Some of the data suggest that.
And so, after looking at the conclusions reached in all these studies and struggling to read the itty-bitty fine print in those charts, am I still confused? Yes. Do I think you should get yearly blood tests for cholesterol, triglycerides, and all the other usual suspects? Definitely. Do I think you should lose weight, get more exercise, and decrease overall calories and probably carbohydrates as well? Absolutely. Do I have a lot of data to support these assertions? Not yet!
_____________________
* Hooper L, Summerbell CD, Thompson R, Sills D, Roberts FG, Moore H, Smith GD (July 2011). "Reduced or modified dietary fat for preventing cardiovascular disease". The Cochrane Library (7): CD002137
**Siri-Tarino et. al. 2010. Meta-Analysis of Prospective Cohort Studies Evaluating the Association of Saturated Fat with Cardiovascular Disease. The American Journal of Clinical Nutrition 91 (3): 535–46
*** Mozaffarian D, Micha R, Wallace S (March 2010). Katan, Martijn B. ed. "Effects on Coronary Heart Disease of Increasing Polyunsaturated Fat in Place of Saturated Fat: A Systematic Review and Meta-Analysis of Randomized Controlled Trials". PLoS Medicine 7 (3): 1–10
In fact, when it comes to fat, there is a lot to be confused about. For one thing, the word “fat” is almost meaningless. The only way to understand what the word means is by the context in which it is used. For example, if someone refers to “fat” in their diet, you don’t know if they mean lard, oil, lipids, a specific fatty acid, a fatty acid supplement, or a mixture of some or all of these. If someone refers to “fat” in their blood, you don’t know if they mean triglycerides, cholesterol, or plaque build-up. I’d be just as happy if “fat” was eliminated from our vocabulary, as well as from our waists.
In sum, I’m more certain about the discovery of water on Mars than I am about the role of fat in the human diet.
And the subject is so complicated I think I’ll make it the topic of at least the next two blogs. Sorry. And, I can tell you right now my objective is to wade through studies to determine if there is an emerging consensus—OK, if you read no further, I’ll tell you: there isn’t one. At least I don’t see it. OR if there is a consensus it is very weak, so far.
For our purposes, the best way to think about fat is in terms of what it is made of: it consists of two kinds of molecules—fatty acids and glycerol. OK, we don’t care about glycerol (at least I think we don’t). What we REALLY care about are fatty acids.
What makes it confusing is that there are more than 20 different fatty acids. And they have frightening common names: oleic acid, palmitic acid, linoleic acid, linolenic acid, stearic acid, etc. Further, these fatty acids can be subdivided into “saturated” fatty acids and “unsaturated” fatty acids. Unsaturated fatty acids can be further divided into “polyunsaturated” fatty acids and “monounsaturated” fatty acids. And unsaturated fatty acids can also be described as “cis-“ and “trans-“ fatty acids.
Saturated fat. Trans-fat. Polyunsaturated fats. Peanut oil. Olive oil. Corn oil. Soybean oil. Lard. Margarine. All of these differ from each other by the types and amounts of fatty acids they contain.
By looking at the recommendations of the American Heart Association, the World Health Organization, the American Dietetic Association, the British Heart Foundation, the World Heart Federation, the Food and Drug Administration, and others, you would assume that there is a scientific consensus regarding appropriate amounts and types of dietary fat.
I’m sticking my neck out here, but after looking at the literature I am totally convinced that this is not the case. Let me say it in another way: in my opinion, there is no consensus by the scientific establishment on many of the recommendations concerning the consumption of fat that we all thought were gospel.
For example, we have all heard that we should eat less “saturated” fat and eat more “polyunsaturated” and “monounsaturated” fat. Following these recommendations, we have reduced our consumption of meat, cheese, butter, lard, pizzas, doughnuts, sausages, and cream (all high-saturated-fat foods), and increased our consumption of skim milk, nuts, chicken, fish, soybeans, margarine, and vegetables (all low-saturated-fat foods)—right? Isn’t that what we’ve been told to do for years now?
Clearly we need to take a look at the studies supporting these recommendations. But first, let’s define what we mean by “saturated” fat, “polyunsaturated” fat, and “monounsaturated” fat.
Simply, “saturated fat” means a fat that is composed of fatty acids that don’t have any double bonds between the carbon atoms. Simple as that. There are five common “saturated” fatty acids (butyric acid, lauric acid, myristic acid, palmitic acid, and stearic acid). For example, stearic acid has 18 carbon atoms, each of which is linked to another by a single chemical bond, and palmitic acid has 16 carbon atoms, also linked by single bonds. The fat found in ground beef, for example, is composed of 15% stearic acid, 26% palmitic acid, and 4% myristic acid. So 45% of the fat content of ground beef is saturated fat. Therefore, the remaining 65% of the fat in ground beef is polyunsaturated and monounsaturated fat.
What is monounsaturated fat? It is a fatty acid that has only one double bond. There are three common monounsaturated fatty acids: palmitoleic acid (with 16 carbon atoms), and vaccinic acid and oleic acid (each with 18 carbon atoms). Each of these fatty acids has only one place where the linkage between two carbon atoms consists of two bonds (a “double” bond). Olive oil is a good example of a fat that is high in monounsaturates (75%).
Finally, polyunsaturated fats. These are composed of fatty acids in which there is more than one double bond between the carbon atoms. For example, the famous “omega 3 fatty acid” that is found in fish oil and that we’ve all been advised to ingest more of, is actually a family of polyunsaturated fatty acids, the most famous being alpha-linolenic acid, which has 18 carbon atoms and 3 double bonds. The “omega 6 fatty acid” is another family of polyunsaturated fatty acids, of which the most famous is probably linoleic acid (not to be confused with the linolenic acid found in omega 3 above!). Linoleic acid is 18 carbon atoms long and has 2 double bonds.
Whew. I don’t know how anyone could keep this stuff in their heads for more than 5 seconds. So, for now just remember that fatty acids are either saturated, monounsaturated, or polyunsaturated.
Now let’s look at the recommendation we are most familiar with: “Thou shalt reduce the amount of saturated fat in your diet.” What I mean is, let’s look at the scientific underpinnings of this recommendation, which is practically gospel in most nutrition circles.
Well, since the 1960’s, dozens and dozens of studies have been performed to look at saturated fats. And, indeed, many studies show that reducing saturated fats results in a reduction in both mortality and cardiovascular disease. However, other studies show no relationship. So is there a consensus today?
One way at getting at the answer to this question is to look for what’s called a “meta-analysis.” A meta-analysis is a study that groups many studies together and then performs a statistical analysis of the grouped data. In this way you can go from an individual study of say 100 people (which I’m sure is too small to say anything definitive about dietary fat) to a single, comprehensive analysis of many grouped studies and thereby get your sample size up into the thousands. The goal is to have a study population that is large enough to say something definitive, that is, to do a statistical analysis that is powerful enough to identify statistically significant trends.
And for this reason I like meta-analyses. I mean, how else can one possibly answer the question, “What do the studies say?” without looking at each of the studies individually? A criticism of meta-analyses is that since they group many studies, they may lump sexes, races, ages, specific disease considerations, dietary interventions, and research methodologies together. Also, different studies have different numbers of people enrolled as test subjects—so a study based on 10 people should not have the same “weight” as a study with 10,000 people. But, some of these problems can be controlled by the scientists who are doing the meta-analysis: They can specify that they will only look at studies that meet specific criteria.
So, let’s look at a meta-analysis of studies that looked at the saturated fat question. I have found three recent such meta-analyses.
Meta-analysis #1*, published in 2011. This particular meta-analysis only included studies of adults 18 years or older in randomized experiments, with or without risk of cardiovascular disease, that included both males and females (pregnant or lactating women excluded). The authors settled on 60 studies with publication dates between 1965 and 2009, including 30 studies in North America, 26 in Europe, 3 in Australia/New Zealand, and one in the Middle East. There were approximately 70,000 total participants in the 60 studies combined.
For our purposes here, we can say that in one group of studies, saturated fat was reduced by approximately 30% of total calories. The test subjects made up for the resulting calorie deficit by eating more carbohydrates, fruits, vegetables, and/or protein.
The authors’ conclusion? The reduction in dietary saturated fat had no statistically discernable effect on mortality, stroke, heart attacks, “cardiovascular events,” cancer, or diabetes.
However, when polyunsaturated fats or monounsaturated fats were used in place of carbohydrates, vegetables, and fruits to bring the total calories back up to normal, there was a 14% reduction in “cardiovascular events”, but no specific reduction in strokes or heart attacks. The term “cardiovascular event” was not further defined.
Meta-analysis #2**, published in 2010. 21 studies were included in the analysis, with each study covering 5 to 23 years, and a combined total of 347,747 subjects. 11 studies included men only, 2 studies included women only, and 8 included both men and women. 12 studies were conducted in North America, 6 in Europe, 2 in Japan, and one in Israel.
There was wide variation in dietary modification or the lack thereof—some of the studies actually modified the subjects’ diets, and others were based on surveys of the subjects’ behavior without any attempt at dietary modification.
Conclusion? There was no evidence that reducing consumption of saturated fat was associated with a decrease in cardiovascular disease.
The authors of this meta-analysis point out that several individual studies have shown that reducing saturated fat and increasing polyunsaturated fat in the diet is beneficial to coronary heart disease, and that more benefits are seen as the amount of polyunsaturated fat increases and the amount of saturated fat decreases. The authors’ own meta-analysis did not reach this conclusion, however.
So, there is a suggestion in both of these studies that reducing saturated fats and increasing polyunsaturated fats MAY have a beneficial effect on cardiovascular disease.
Which brings me to the next study.
Meta-analysis #3***, published in 2010. 8 studies, 13,614 participants. The average length of the studies was 4.25 years. In these studies, the test subjects REPLACED saturated fat with polyunsaturated fat in their diets. On average, the “polyunsaturated” fat group got 15% of their energy from polyunsaturated fats, whereas on average the “saturated” fat group got only 5% of their energy from polyunsaturated fats.
The authors concluded that there was a 19% reduction in heart attacks or deaths due to cardiovascular events among the “polyunsaturated” fat group in comparison with the “saturated” fat group.
So, two of these three meta-analyses indicate that reducing saturated fats and increasing polyunsaturated fats may be beneficial.
Now for something controversial. If no one responds to this, I’ll know that nobody has read this blog! Here is the deal: These three studies also reported on summary results for the individual studies, and SOME of the individual studies showed that THE REDUCED SATURATED-FAT DIETS WERE CORRELATED WITH AN INCREASE IN DEATHS/CARDIOVASCULAR EVENTS. Let me say that a different way: Some (many) of the individual studies showed that decreasing saturated fats in the diet was actually HARMFUL.
What is interesting is that none of the authors discussed this. Is that because it would have been too controversial? Is that because no one wants to advocate a diet high in saturated fat, when we have been told for decades that saturated fat is evil? I do not know.
Further, an analysis was done in the second meta-analysis to test for reporting bias, and the authors concluded that there may be some. That is, studies showing negative results may be under-reported. In this case, a finding that decreasing your consumption of saturated fat is bad for you would be considered a negative result.
What do I think? I don’t know for sure, but I suspect that too few subjects have been analyzed thus far and that the dietary interventions used in these studies were not detailed enough. For example, suppose that in order to lower intake of saturated fat, you ask the test subjects to reduce their consumption of red meat. The problem is that red meat is about 50% saturated fat and 50% monounsaturated fat. So when you analyze the results, are you seeing the effects of reduced saturated fats or reduced monounsaturated fats? Further, there are many different types of polyunsaturated fats—some may be “good,” and some may be “bad.” If a test subject decreases consumption of saturated fat and increases consumption of a “bad” polyunsaturated fat (such as, perhaps, the omega 6 fatty acids found in safflower oil), then the two may cancel each other out. Who knows?
Do I think there is something counter-intuitive is going on here with regard to consumption of fatty acids? There is certainly that suggestion. Do I think you should conclude that increasing saturated fat in your diet is good for you? No, the data does not say that, and neither am I. Am I saying that consumption of saturated fat is neutral (neither good nor bad)? Don’t know that either. Do I think increasing SOME TYPES OF polyunsaturated fats may be beneficial? Some of the data suggest that.
And so, after looking at the conclusions reached in all these studies and struggling to read the itty-bitty fine print in those charts, am I still confused? Yes. Do I think you should get yearly blood tests for cholesterol, triglycerides, and all the other usual suspects? Definitely. Do I think you should lose weight, get more exercise, and decrease overall calories and probably carbohydrates as well? Absolutely. Do I have a lot of data to support these assertions? Not yet!
_____________________
* Hooper L, Summerbell CD, Thompson R, Sills D, Roberts FG, Moore H, Smith GD (July 2011). "Reduced or modified dietary fat for preventing cardiovascular disease". The Cochrane Library (7): CD002137
**Siri-Tarino et. al. 2010. Meta-Analysis of Prospective Cohort Studies Evaluating the Association of Saturated Fat with Cardiovascular Disease. The American Journal of Clinical Nutrition 91 (3): 535–46
*** Mozaffarian D, Micha R, Wallace S (March 2010). Katan, Martijn B. ed. "Effects on Coronary Heart Disease of Increasing Polyunsaturated Fat in Place of Saturated Fat: A Systematic Review and Meta-Analysis of Randomized Controlled Trials". PLoS Medicine 7 (3): 1–10
No comments:
Post a Comment