In Ian Leslie's 'The Sugar Conspiracy', featured last week in the Guardian, we find a seductive narrative of the politics surrounding the process of setting the nation's dietary recommendations. Detailed within is the story of two researchers, Ancel Keys, a vocal advocate of lowering fat/saturated fat consumption, and John Yudkin, a humble researcher who warned of the dangers of sugar.
We are told in 'The Sugar Conspiracy' that, instead of being a matter of scientific differences, Ancel Keys, as part of the 'nutrition elite', squashed the dissenting voice of John Yudkin, who made the case that sugar was the real enemy.
This narrative is enticing; it, as the title notes, edges on conspiracy and addresses the deeply political perspectives on taking science and translating it into policy. The narrative crafted by Leslie, however, is dependent upon the idea that there weren't legitimate differences in the quality of evidence put forth by these researchers. There was nothing wrong with Yudkin's work; rather, politics overtook the science. This narrative is beautifully woven into dissenters ringing the alarm about the modern process of crafting dietary guidelines; history, according to 'The Sugar Conspiracy' is repeating itself.
For the skeptical reader of 'The Sugar Conspiracy', the question might arise: was Yudkin's research/perspective on the evidence so solid? The conspiracy would lose some of its allure if Yudkin's data was weak. For those interested, I have posted a blog thoroughly analyzing Yudkin's perspective; NutritionWonk has also posted a blog addressing some of Leslie's claims about Ancel Keys' data. In short, not only does Leslie misrepresent Keys' data, the power narrative he constructs ignore some of the major limitations of Yudkin's data/perspective.
To make sense of 'The Sugar Conspiracy' and my response, it's essential to consider a few things. The gold standard of evidence in the medical field is to conduct a placebo controlled, double-blind, randomized, controlled trial. Essentially, we take a well-defined population of people, randomly assign to them to two groups, one that gets the control diet and one that gets the intervention diet. As you can imagine, most diet literature is often unable to reach this gold standard level of evidence, because it'd be hard to 'blind' someone from the fact that they are in the intervention group receiving vegetable oil instead of butter; imagine trying to blind someone to the fact that you're not letting them eat any carbohydrates! #subtlebiases To do this trial and test the effects of the diet on cardiovascular disease, we'd need a huge group of people that were very adherent to their diets and we'd need to follow them for a really long time to see any differences in cardiovascular events.
Complicating this some, diet is super complex! People tend not to eat what researchers tell them to, and dietary interventions tend to have many 'moving parts'; that is , when you change one thing in the diet, like reduce saturated fat rich foods, you also remove the other nutrients that come with them. You can't just reduce saturated fats in a randomized controlled trial - the effects would be confounded by the removal of the calories; researchers must replace those calories, leading to ambiguous interpretation of the results: are saturated fats bad or was what they were replaced with good? All of the factors make designing nutrition research that reaches our best quality standards and that still retain some real world implications near impossible! Imagine trying to figure out how 1 nutrient affects disease outcomes that take decades to develop - prevention research is hard!
At the time of Keys' and Yudkin's research, we didn't have these diet randomized controlled trials with strong outcomes, like cardiovascular death/mortality. The research at the time was trying to understand what factors might be involved in the development of cardiovascular disease; before you can design a randomized controlled, you need a good idea about what risk factors to even target. There was a lot of animal research at the time looking at factors that affect cardiovascular disease development, but the relevance of how a 2 year old rabbit got cardiovascular disease lacks direct translation to humans. The limitations of animal models tend to lead us to study factors in large, population cohorts (epidemiological research) that are associated with cardiovascular disease.
Researchers at the time that the first dietary guidelines were developed had identified a few independent risk factors for cardiovascular disease at this point: high serum cholesterol, high blood pressure, and smoking. They certainly knew that we needed controlled trial evidence to affirm that therapies that lowered these risk factors were actually causal (and not confounded by some associated factor - e.g. if people who smoked also were of lower socioeconomic status and more likely to get infections, which might contribute to atherosclerosis via inflammation); even the demonized Ancel Keys was talking about the need for trial evidence to confirm the effects of lowering cholesterol on cardiovascular disease.
So what's this got to do with my perspective deviating from Leslie's? Whereas Leslie would have us believe that the dietary recommendations at the time were made because Ancel Keys and the nutrition elite had some undue influence over the policy recommendations, I think there's a lot more to this story. For starters, the limitations of the available evidence at the time was evident and frequently noted by the scientific establishment - a convenient fact left out by Leslie's piece. Multiple reports from the American Heart Association at the time, including the 1961 report that included Ancel Keys as an author, had discussed the limitations of the evidence and the fact that the role of diet had not yet been proven; to quote the report when discussing diet: "it must be emphasized that there is as yet no final proof that heart attacks or strokes will be prevented by such measures". A lot of the data was associative (non-causal) at the time, opening it up to different biases and potential confounding.
Apart from associative data, there were experiments that tested the way that diet effected the established risk factors at the time (again: serum cholesterol and blood pressure). While Yudkin had some data showing that high sugar diets (some of Yudkin's experiments had people eating over 400g of sugar) affected one component of serum cholesterol, the triglycerides, the human evidence was quite weak at the time. These studies from Yudkin and others at the time were quite short-term feeding studies, using doses significantly higher than the average sugar consumption; even with these designs, they could only show a slight increase in triglycerides/serum cholesterol. A major limiting factor of this research was that it had been shown at the time that this rise in triglycerides was transient - it took several months for humans to adapt and for serum cholesterol to be lowered. The fact that Yudkin and other researchers at the time were using such high doses in short term feeding studies was an area of critique by other researchers at the time, including Ancel Keys, who noted that animal/human experiments using mega-doses of sugar to produce slight increase in serum cholesterol weren't very relevant to human disease development. Consider these studies using 100s of grams of sugar per day to increase serum cholesterol a bit with the more sizable impact that saturated fat intake has on serum cholesterol, using very real world doses and the reduction in serum cholesterol you could get by consuming polyunsaturated fats. The argument, based on how diet affects established risk factors, for the recommendation to change fat intake and the type of fats was stronger than that to reduce sugar.
Yudkin also used some weaker case-control study designs to suggest that individuals with cardiovascular disease consumed more sugar than his controls; however, not only were these studies not able to be repeated by other researchers at the time, but they didn't assess sugar intake before the heart attack; this is pretty important, because you should be able to show that something that causes/affects the development of the disease occurs before the disease, not after - what if people who had heart attacks started eating a lot of sugar afterwards, or the trauma of their event biased their ability to recall their food intake months/years before? One prospective cohort study at the time, assessing sugar intake in individuals and then following them to see who had a heart attack, did not find a significant effect of sugar intake on cardiovascular events; even if it had, there would still be the concern that this association was confounded - it was known at the time that high sugar consumption was associated with being a smoker, a major risk factor for CVD. If you're interested in all of the data that Yudkin put forth and a detailed analysis of it, see my former post!
If you notice in 'The Sugar Conspiracy', Leslie refers to a quote by a researcher named Reiser; Leslie cites a quote that Reiser allegedly told Gary Taubes about Keys' behavior in addressing Yudkin's research. I bring up Reiser because of a review that he wrote in 1975 summarizing the data on sugar and disease (keep in mind that 1975 was just before the McGovern report and the first dietary guidelines were released so this review is the evidence available to the authors of these reports at that time). Reiser concludes in this review that sucrose alone has not been shown to be a significant risk factor in the development of diabetes and cardiovascular disease; he merely suggested that, based on some small trials at the time, there might be some individuals who are more sensitive to the effects of sugar on blood cholesterol and they might benefit from a reduction. It was noted by other researchers at the time that those who hyperlipemic (high starting triglycerides) were more likely to see a greater rise in triglycerides from consuming high sugar diets than normolipemic people - this kind of research would inform a medical professional, like a registered dietitian, when working one on one with a patient with abnormal lipids, but it's not the type of research that we include in general, population wide recommendations.
Evidence pointing at sugar as a major driver of cardiovascular disease just wasn't there at the time. Of note, given the fact that Leslie cites multiple low carbohydrate diet folks and notes that randomized controlled trials for low carbohydrate diets show consistent evidence (though it's only for weight loss and modifying the same risk factors that he casts doubt on), Yudkin's evidence all pointed to replacing sucrose with starch; while he suggested a low carbohydrate diet was great for weight loss, one of the reasons he quoted at the time was that low carbohydrate diets also tended to cut fat! Additionally, while some folks cited by Leslie cast doubt on the 'conventional wisdom of calories' and paint larger pictures about the role of insulin in fat loss, Yudkin can be found in the literature ranting about the 'anti-calorie' folks.
Given these points about the limitations of Yudkin's data (and the majority of the data at that time), it's a shame that Leslie didn't write a more genuine article asking the real question - why were any dietary guidelines given before strong trial evidence? There are some reasonable explanations for this, but they wouldn't really write an article called 'The Sugar Conspiracy'. For starters, many would argue that waiting for big, randomized controlled trials with meaningful outcomes, like a heart attack, would take years to do. Think about how many more people would have died from smoking related diseases if we had waited until we enrolled 10,000 patients into a trial and told half to stop smoking, and see how many develop lung cancer, a disease that takes decades to develop. Actually, the fact that it took so long to get the evidence sufficient to act on reducing smoking incidence/prevalence in the population led to the creation of a set of criteria to understand better how to decide when its a good time to make policy/recommendations when we only have observational evidence. At the time that the dietary guidelines were set, we didn't have those criteria. There's also an ethical question that arises: if we think that having high cholesterol is a big risk factor for a disease, is it ethically okay to keep telling people to eat diets with foods that raise cholesterol? This is obviously open to a lot of interpretation and debate, but it's one that could be invoked.
The politics at the time of the setting of the first dietary guidelines are important to consider as we think about why dietary recommendations were set without strong trial evidence. The article I think Leslie should have written was actually already written in 1998 - it's titled 'Science, Policy and Controversy in the Cholesterol Arena', penned by Karin Garrety. While it mentions Ancel Keys' passion in one line of the publication, it details a much more complex history. One where the media like Times Magazine, political influences, diet books, industry, and public enthusiasm that believed a 'healthy diet' would prevent disease led to the perfect storm for recommendations that were, to some degree, premature. It's not a story of some 'nutrition elite' but rather, one of multiple sources of pressure to address a diseases on the rise, despite lacking large trial evidence.
Ultimately, the McGovern Report and the Dietary Guidelines were published and, conveniently not mentioned by Leslie's article, recommended consuming a low sugar diet! They simply didn't advocate consuming low sugar diets for the purpose of reducing cardiovascular risk, nor should they have, because the evidence didn't support doing so at the time.
Fast forward to present day, and it's still clear that saturated fats raise cholesterol, specifically the low density lipoproteins. The science doesn't doubt that at all - what the science actually doubts is the assumption that any intervention which lowers LDL will be beneficial. The evidence implicating LDL in cardiovascular disease is stronger than ever - with the advent of large genetic datasets, we've identified a number of genetic variants that affect the levels of LDL. Because genetic variants are generally randomly distributed at birth, this mimics the randomization process in clinical trials. Individuals with genetic variants that lower LDL, reflecting a lifetime of low LDL (we have zero lifetime randomized controlled trials) have a lower risk of developing cardiovascular disease, relatively proportionate to the effect size of this LDL lowering. In addition to these natural human genetic studies, we have multiple lines of evidence from animal genetic experiments and pharmacological experiments that further support the effectiveness of lowering LDL/concerns about having high LDL. When Leslie cites the Zoe Harcombe, another low carbohydrate diet enthusiast, analysis of LDL levels, he doesn't mention that this is extremely low quality evidence - it is a simple correlation between total cholesterol (no subfractions like LDL or HDL) and cardiovascular disease deaths; it doesn't take into account any other factors that might augment this relationship (many wasting diseases that occur at the end of life can lower cholesterol; it doesn't even take into account whether or not those with low cholesterol were on cholesterol-lowering drugs, indicating that they were already sick!). It's extremely dangerous, and lacking in integrity, for a journalist to not check these sources and balance them with harder science, conveying misinformation to the public.
The real area of scientific debate is the best way to lower cholesterol via diet. Since the time of the first dietary guidelines, there have been a number of diet trials that have tried to deduce the effects of lowering saturated fat; as you might guess, nearly all are plagued by some limitation, whether it is some co-intervention (e.g. not just lowering saturated fat), a lack of adherence, confounding (a lot of these trials were done in sick populations), short follow up time (the recent headlines about the Minnesota Coronary Experiment highlight these last two points), inappropriate/failure to randomize, lack of blinding, etc. The most recent examination of this evidence, performed by the Cochrane group, generally considered a source for high quality meta analyses and systematic reviews, concluded that there was moderate evidence that replacing saturated fats with different polyunsaturated fats (no specific recommendations about omega3 or 6) reduces cardiovascular events, but not cardiovascular or total mortality. Based on this evidence, and other epidemiological evidence modeling replacements of saturated fats with polyunsaturated fats, the current dietary guidelines continue to recommend replacing saturated fats with polyunsaturated fats. For those who find it interesting, this recommendation is reinforced by studies of nonhuman primates, our closest relatives.
Is the human evidence perfect? Absolutely not. Will it ever be perfect? Likely not. Consider that we have no human randomized controlled trials, not even terribly confounded ones, showing that lowering sugar, or carbohydrates for that matter, affects cardiovascular events/mortality, a fact not put forth by the string of low-carbohydrate diet folks quoted by Leslie. Maybe that's the true 'Sugar Conspiracy'.
We are told in 'The Sugar Conspiracy' that, instead of being a matter of scientific differences, Ancel Keys, as part of the 'nutrition elite', squashed the dissenting voice of John Yudkin, who made the case that sugar was the real enemy.
This narrative is enticing; it, as the title notes, edges on conspiracy and addresses the deeply political perspectives on taking science and translating it into policy. The narrative crafted by Leslie, however, is dependent upon the idea that there weren't legitimate differences in the quality of evidence put forth by these researchers. There was nothing wrong with Yudkin's work; rather, politics overtook the science. This narrative is beautifully woven into dissenters ringing the alarm about the modern process of crafting dietary guidelines; history, according to 'The Sugar Conspiracy' is repeating itself.
For the skeptical reader of 'The Sugar Conspiracy', the question might arise: was Yudkin's research/perspective on the evidence so solid? The conspiracy would lose some of its allure if Yudkin's data was weak. For those interested, I have posted a blog thoroughly analyzing Yudkin's perspective; NutritionWonk has also posted a blog addressing some of Leslie's claims about Ancel Keys' data. In short, not only does Leslie misrepresent Keys' data, the power narrative he constructs ignore some of the major limitations of Yudkin's data/perspective.
 |
| stats.statsexchange.com |
To make sense of 'The Sugar Conspiracy' and my response, it's essential to consider a few things. The gold standard of evidence in the medical field is to conduct a placebo controlled, double-blind, randomized, controlled trial. Essentially, we take a well-defined population of people, randomly assign to them to two groups, one that gets the control diet and one that gets the intervention diet. As you can imagine, most diet literature is often unable to reach this gold standard level of evidence, because it'd be hard to 'blind' someone from the fact that they are in the intervention group receiving vegetable oil instead of butter; imagine trying to blind someone to the fact that you're not letting them eat any carbohydrates! #subtlebiases To do this trial and test the effects of the diet on cardiovascular disease, we'd need a huge group of people that were very adherent to their diets and we'd need to follow them for a really long time to see any differences in cardiovascular events.
Complicating this some, diet is super complex! People tend not to eat what researchers tell them to, and dietary interventions tend to have many 'moving parts'; that is , when you change one thing in the diet, like reduce saturated fat rich foods, you also remove the other nutrients that come with them. You can't just reduce saturated fats in a randomized controlled trial - the effects would be confounded by the removal of the calories; researchers must replace those calories, leading to ambiguous interpretation of the results: are saturated fats bad or was what they were replaced with good? All of the factors make designing nutrition research that reaches our best quality standards and that still retain some real world implications near impossible! Imagine trying to figure out how 1 nutrient affects disease outcomes that take decades to develop - prevention research is hard!
At the time of Keys' and Yudkin's research, we didn't have these diet randomized controlled trials with strong outcomes, like cardiovascular death/mortality. The research at the time was trying to understand what factors might be involved in the development of cardiovascular disease; before you can design a randomized controlled, you need a good idea about what risk factors to even target. There was a lot of animal research at the time looking at factors that affect cardiovascular disease development, but the relevance of how a 2 year old rabbit got cardiovascular disease lacks direct translation to humans. The limitations of animal models tend to lead us to study factors in large, population cohorts (epidemiological research) that are associated with cardiovascular disease.
Researchers at the time that the first dietary guidelines were developed had identified a few independent risk factors for cardiovascular disease at this point: high serum cholesterol, high blood pressure, and smoking. They certainly knew that we needed controlled trial evidence to affirm that therapies that lowered these risk factors were actually causal (and not confounded by some associated factor - e.g. if people who smoked also were of lower socioeconomic status and more likely to get infections, which might contribute to atherosclerosis via inflammation); even the demonized Ancel Keys was talking about the need for trial evidence to confirm the effects of lowering cholesterol on cardiovascular disease.
So what's this got to do with my perspective deviating from Leslie's? Whereas Leslie would have us believe that the dietary recommendations at the time were made because Ancel Keys and the nutrition elite had some undue influence over the policy recommendations, I think there's a lot more to this story. For starters, the limitations of the available evidence at the time was evident and frequently noted by the scientific establishment - a convenient fact left out by Leslie's piece. Multiple reports from the American Heart Association at the time, including the 1961 report that included Ancel Keys as an author, had discussed the limitations of the evidence and the fact that the role of diet had not yet been proven; to quote the report when discussing diet: "it must be emphasized that there is as yet no final proof that heart attacks or strokes will be prevented by such measures". A lot of the data was associative (non-causal) at the time, opening it up to different biases and potential confounding.
Apart from associative data, there were experiments that tested the way that diet effected the established risk factors at the time (again: serum cholesterol and blood pressure). While Yudkin had some data showing that high sugar diets (some of Yudkin's experiments had people eating over 400g of sugar) affected one component of serum cholesterol, the triglycerides, the human evidence was quite weak at the time. These studies from Yudkin and others at the time were quite short-term feeding studies, using doses significantly higher than the average sugar consumption; even with these designs, they could only show a slight increase in triglycerides/serum cholesterol. A major limiting factor of this research was that it had been shown at the time that this rise in triglycerides was transient - it took several months for humans to adapt and for serum cholesterol to be lowered. The fact that Yudkin and other researchers at the time were using such high doses in short term feeding studies was an area of critique by other researchers at the time, including Ancel Keys, who noted that animal/human experiments using mega-doses of sugar to produce slight increase in serum cholesterol weren't very relevant to human disease development. Consider these studies using 100s of grams of sugar per day to increase serum cholesterol a bit with the more sizable impact that saturated fat intake has on serum cholesterol, using very real world doses and the reduction in serum cholesterol you could get by consuming polyunsaturated fats. The argument, based on how diet affects established risk factors, for the recommendation to change fat intake and the type of fats was stronger than that to reduce sugar.
Yudkin also used some weaker case-control study designs to suggest that individuals with cardiovascular disease consumed more sugar than his controls; however, not only were these studies not able to be repeated by other researchers at the time, but they didn't assess sugar intake before the heart attack; this is pretty important, because you should be able to show that something that causes/affects the development of the disease occurs before the disease, not after - what if people who had heart attacks started eating a lot of sugar afterwards, or the trauma of their event biased their ability to recall their food intake months/years before? One prospective cohort study at the time, assessing sugar intake in individuals and then following them to see who had a heart attack, did not find a significant effect of sugar intake on cardiovascular events; even if it had, there would still be the concern that this association was confounded - it was known at the time that high sugar consumption was associated with being a smoker, a major risk factor for CVD. If you're interested in all of the data that Yudkin put forth and a detailed analysis of it, see my former post!
If you notice in 'The Sugar Conspiracy', Leslie refers to a quote by a researcher named Reiser; Leslie cites a quote that Reiser allegedly told Gary Taubes about Keys' behavior in addressing Yudkin's research. I bring up Reiser because of a review that he wrote in 1975 summarizing the data on sugar and disease (keep in mind that 1975 was just before the McGovern report and the first dietary guidelines were released so this review is the evidence available to the authors of these reports at that time). Reiser concludes in this review that sucrose alone has not been shown to be a significant risk factor in the development of diabetes and cardiovascular disease; he merely suggested that, based on some small trials at the time, there might be some individuals who are more sensitive to the effects of sugar on blood cholesterol and they might benefit from a reduction. It was noted by other researchers at the time that those who hyperlipemic (high starting triglycerides) were more likely to see a greater rise in triglycerides from consuming high sugar diets than normolipemic people - this kind of research would inform a medical professional, like a registered dietitian, when working one on one with a patient with abnormal lipids, but it's not the type of research that we include in general, population wide recommendations.
Evidence pointing at sugar as a major driver of cardiovascular disease just wasn't there at the time. Of note, given the fact that Leslie cites multiple low carbohydrate diet folks and notes that randomized controlled trials for low carbohydrate diets show consistent evidence (though it's only for weight loss and modifying the same risk factors that he casts doubt on), Yudkin's evidence all pointed to replacing sucrose with starch; while he suggested a low carbohydrate diet was great for weight loss, one of the reasons he quoted at the time was that low carbohydrate diets also tended to cut fat! Additionally, while some folks cited by Leslie cast doubt on the 'conventional wisdom of calories' and paint larger pictures about the role of insulin in fat loss, Yudkin can be found in the literature ranting about the 'anti-calorie' folks.
Given these points about the limitations of Yudkin's data (and the majority of the data at that time), it's a shame that Leslie didn't write a more genuine article asking the real question - why were any dietary guidelines given before strong trial evidence? There are some reasonable explanations for this, but they wouldn't really write an article called 'The Sugar Conspiracy'. For starters, many would argue that waiting for big, randomized controlled trials with meaningful outcomes, like a heart attack, would take years to do. Think about how many more people would have died from smoking related diseases if we had waited until we enrolled 10,000 patients into a trial and told half to stop smoking, and see how many develop lung cancer, a disease that takes decades to develop. Actually, the fact that it took so long to get the evidence sufficient to act on reducing smoking incidence/prevalence in the population led to the creation of a set of criteria to understand better how to decide when its a good time to make policy/recommendations when we only have observational evidence. At the time that the dietary guidelines were set, we didn't have those criteria. There's also an ethical question that arises: if we think that having high cholesterol is a big risk factor for a disease, is it ethically okay to keep telling people to eat diets with foods that raise cholesterol? This is obviously open to a lot of interpretation and debate, but it's one that could be invoked.
The politics at the time of the setting of the first dietary guidelines are important to consider as we think about why dietary recommendations were set without strong trial evidence. The article I think Leslie should have written was actually already written in 1998 - it's titled 'Science, Policy and Controversy in the Cholesterol Arena', penned by Karin Garrety. While it mentions Ancel Keys' passion in one line of the publication, it details a much more complex history. One where the media like Times Magazine, political influences, diet books, industry, and public enthusiasm that believed a 'healthy diet' would prevent disease led to the perfect storm for recommendations that were, to some degree, premature. It's not a story of some 'nutrition elite' but rather, one of multiple sources of pressure to address a diseases on the rise, despite lacking large trial evidence.
Ultimately, the McGovern Report and the Dietary Guidelines were published and, conveniently not mentioned by Leslie's article, recommended consuming a low sugar diet! They simply didn't advocate consuming low sugar diets for the purpose of reducing cardiovascular risk, nor should they have, because the evidence didn't support doing so at the time.
Fast forward to present day, and it's still clear that saturated fats raise cholesterol, specifically the low density lipoproteins. The science doesn't doubt that at all - what the science actually doubts is the assumption that any intervention which lowers LDL will be beneficial. The evidence implicating LDL in cardiovascular disease is stronger than ever - with the advent of large genetic datasets, we've identified a number of genetic variants that affect the levels of LDL. Because genetic variants are generally randomly distributed at birth, this mimics the randomization process in clinical trials. Individuals with genetic variants that lower LDL, reflecting a lifetime of low LDL (we have zero lifetime randomized controlled trials) have a lower risk of developing cardiovascular disease, relatively proportionate to the effect size of this LDL lowering. In addition to these natural human genetic studies, we have multiple lines of evidence from animal genetic experiments and pharmacological experiments that further support the effectiveness of lowering LDL/concerns about having high LDL. When Leslie cites the Zoe Harcombe, another low carbohydrate diet enthusiast, analysis of LDL levels, he doesn't mention that this is extremely low quality evidence - it is a simple correlation between total cholesterol (no subfractions like LDL or HDL) and cardiovascular disease deaths; it doesn't take into account any other factors that might augment this relationship (many wasting diseases that occur at the end of life can lower cholesterol; it doesn't even take into account whether or not those with low cholesterol were on cholesterol-lowering drugs, indicating that they were already sick!). It's extremely dangerous, and lacking in integrity, for a journalist to not check these sources and balance them with harder science, conveying misinformation to the public.
The real area of scientific debate is the best way to lower cholesterol via diet. Since the time of the first dietary guidelines, there have been a number of diet trials that have tried to deduce the effects of lowering saturated fat; as you might guess, nearly all are plagued by some limitation, whether it is some co-intervention (e.g. not just lowering saturated fat), a lack of adherence, confounding (a lot of these trials were done in sick populations), short follow up time (the recent headlines about the Minnesota Coronary Experiment highlight these last two points), inappropriate/failure to randomize, lack of blinding, etc. The most recent examination of this evidence, performed by the Cochrane group, generally considered a source for high quality meta analyses and systematic reviews, concluded that there was moderate evidence that replacing saturated fats with different polyunsaturated fats (no specific recommendations about omega3 or 6) reduces cardiovascular events, but not cardiovascular or total mortality. Based on this evidence, and other epidemiological evidence modeling replacements of saturated fats with polyunsaturated fats, the current dietary guidelines continue to recommend replacing saturated fats with polyunsaturated fats. For those who find it interesting, this recommendation is reinforced by studies of nonhuman primates, our closest relatives.
Is the human evidence perfect? Absolutely not. Will it ever be perfect? Likely not. Consider that we have no human randomized controlled trials, not even terribly confounded ones, showing that lowering sugar, or carbohydrates for that matter, affects cardiovascular events/mortality, a fact not put forth by the string of low-carbohydrate diet folks quoted by Leslie. Maybe that's the true 'Sugar Conspiracy'.
Comments
Post a Comment