12 The Journalist Gary Taubes 12: Oil-Based Nutrition 2
Another diet trial that was premised on the therapeutic value of refined oils was the Los Angeles Veterans Study. Gary Taubes tells us that although the veterans who ate a diet dripping with polyunsaturated fats did experience a decline in their cholesterol which accompanied a lower rate of cardiovascular mortality, they also had a higher overall mortality rate than the control group, who consumed a lot of saturated fat. Gary Taubes quotes the lead author of the papers that came out of this trial, who wondered if there might be a noxious effect from consuming high levels of unsaturated fat. Perhaps that was a question that might have been asked before this trial began. Nevertheless, this study did provide us with some interesting observations. Before we look at it, just take note of Taubes’ reasoning here. Perhaps a diet high in polyunsaturated fats and low in saturated fats will lower mortality from heart disease, he seems to be saying. The price for that is an increased risk of other causes of death, especially cancer. This argument is composed of two separate questions which can be studied and for which there are answers. The first question is, will an oily diet like this lower the risk of mortality from heart disease compared to a diet heavy on saturated fats? The other is, do such diets cause cancer? If we are really searching for truth here, we need to address these questions separately. That must wait until later. Let’s look at this trial now instead.
There was a nice short paper published in The Lancet that summarized this trial nicely. As in Minnesota, the amount of fat in each diet was kept quite high. Also, the experimental diet was still very high in cholesterol. I doubt many on either side of the debate over saturated fat would consider this experimental diet to have been especially healthy. Notice what I have highlighted about their criteria for participation in this study. The participants could have had pre-existing heart disease. They were considered to be a participating member of the trial regardless of their possible absence from the hospital where the trial was run, and regardless of whether they actually cooperated in the trial. Using these loose standards, the adherence to the diet for the experimental group was only considered to be 49%. That means that the experimental group collectively received less than half of their meals from the staff of the trial. When we talk about the LA Veterans Study, we must remember that compliance to the oily experimental diet was quite poor.
Nevertheless, you can see clearly that the experimental group, here represented by the dotted lines, had a much lower incidence rate of cardiovascular events. Do you remember how I said there were two separate questions implied in the description Taubes gave of this study? This answers the first question. The diet with more saturated fat resulted in more adverse cardiovascular events. The confusionists raise the other question about overall mortality because they want to distract you from the clear answer to this question.
There are the numbers. The difference was quite clear.
This graph shows the overall death rate. I have circled the point at which the total death rate for the older people in the experimental group ticked up as the control group flattened. The cholesterol confusionists want you to focus on that one little segment of this graph, the part that focused on the older folks. Perhaps it takes a long time for the harm caused by vegetable oils to become apparent, they say. As you can see, that is a rather selective interpretation.
Here is Chris Masterjohn saying that this modest movement raises the question of what would have happened to total mortality had the study lasted longer. Well, Dr Masterjohn, you can wonder all you want about the data that isn’t there and what might have happened. Generally it is better to focus on real data in science rather than imaginary data. The Weston Price Foundation website is quite good if you like imaginary data and imaginative interpretations of the science.
This paper acknowledged the increased cancer and mortality in the experimental group. Their excess mortality amounted to only nine cases. These few deaths are the basis upon which the argument that these oils cause cancer is made.
This much longer paper gave the full accounting of the Los Angeles Veterans Study.
The detractors and supporters of diet-heart alike should focus on this study’s other more serious problems. On the left you see that it is implied that hydrogenated shortenings were used in the control diet. On the experimental side, a low-fat diet was never even considered. It was decided that such a diet would have required an unacceptable level of gastronomic sacrifice. For anyone doing a low-fat version of a whole food, plant-based diet, this reasoning must seem hopelessly outdated.
A criticism sometimes made of this trial was the unequal distribution of participants according to smoking behavior. I don’t see this as a particularly decisive factor. The control arm did have more total and heavy smokers, but they also had more light smokers and far fewer moderate smokers.
So what about that matter of excess cancer and other deaths in the polyunsaturated group? Here is the graph from the final paper showing that pattern change right at the end. Notice that the deaths in the control group flattened out. Masterjohn asks what would have happened had this trial continued. Maybe he thinks we can extrapolate this trend as well and conclude that after many years, saturated fat happens to become protective, just at the point when oils start killing people. That’s how imaginary data works, right?
Getting back to reality, it turns out, many of the cancer deaths were among those who were not adhering closely to the experimental diet. This makes any weak claim that this diet was causing cancer even weaker. If the people with cancer weren’t adhering to the diet, it’s a bit harder to blame the diet for their cancer. You’ll recall that overall adherence was very poor.
Moreover, any unusual pattern observed only in a very short stretch at the end was most likely a result of chance. Yes, Dr. Masterjohn, chance is always a factor to be considered. I don’t think you’ll make much out of that little change in mortality pattern in the one group at the very end unless you are trying to distract from the other results of this study, which did not appear to be caused by chance.
This is why it was the opinion of the authors that the increase in mortality for the experimental group did not represent a toxic effect from the diet, despite the fact that they had previously raised this question. Gary Taubes repeated their question about this noxious effect, but he didn’t tell you the answer they gave. He didn’t like the answer they gave so he decided to leave the impression that their question was just left hanging, only to one day be answered by smart and sciency low carbers like him.
This was a trial using people who were advanced in years, people who had probably been consuming a lot of animal foods all their lives. I think there is a more reasonable interpretation of all this that you won’t hear from the cholesterol confusionists. Even though the men in the experimental group had been given a terrible diet laden with nutritionally empty refined oils, they still managed to have similar total mortality and much lower cardiovascular mortality than the saturated fat-based group. That’s how badly the saturated fat-based diet performed. If I were a defender of saturated fats, I’d point to the line about hydrogenated oils and blame this dismal performance on that.
Recall that there was that separate question of whether oily cholesterol-lowering diets cause an increase in cancer incidence. The answer is that even among all the terrible oily diets that have been used this way in trials, they have not been observed to contribute to cancer.
Before he described the trial in Los Angeles, Mr Taubes told us about the Anti-Coronary Club, a large group of men in New York who adopted the high-meat, high-oil approach to heart health. They had issues with excess mortality as well.
Mr Taubes tells us that “the mortality problem was the bane of Keys’s dietary-fat hypothesis, bedeviling every trial that tried to assess the effects of a low-fat diet on death as well as disease.” He then starts in about Los Angeles. By now the blatant inaccuracy of this assertion should be obvious. These diets were not low-fat in any sense. Instead, they were purposely equally loaded with fat, just fat of a different kind. They did not represent any sort of contradiction of Keys because he was promoting a diet lower in fat. He explicitly expressed his disapproval of such oily diets. This is yet another case of the journalist Gary Taubes making stuff up. Low carbers prefer to debate straw men and dead men.
Keys said in 1974 that the promotion of such oily diets was “propaganda”. Taubes needs a straw man, though, and he isn’t burdened by the constraints of honesty and integrity.
Taubes could not claim there was a worse effect on mortality during the Helsinki trial, which did demonstrate improved longevity and heart disease risk.
Unfortunately, that trial had problems, too. Ramsden and colleagues were concerned about the unequal use of the drug thioridazine between the two arms of the trial.
And you can see their concerns are justified.
For my purposes, I just want you to know that Helsinki was merely another oil-based intervention, as these researchers like the others felt that only fatty foods reminiscent of animal-sourced foods would be acceptable to their participants. Diet trials are usually exercises in low standards. That's why you don't see many whole food vegan diet trials.
Consequently, both the control and experimental diets in Helsinki were quite high in fat. The intervention diet relied on refined, non-nutritional vegetable oils, so it would have been lower in nutrition for this reason alone. Vegetable fats were used to replace dairy fats. Other components of the diets were not all that different. The study was designed simply to demonstrate the effects of an exchange of different fats. It was not designed to investigate the benefits of a cardioprotective and generally healthful diet.
Just so you are familiar with it if you haven’t seen it, here is the abstract. You can see that the diet intervention was declared to be successful. While this trial had its problems, it was definitely not a win for the saturated fat proponents like Taubes.
But getting back to the Anti-Coronary Club, that diet was intended to be one third fat, so it was not a low-fat diet trial. Taubes wants you to mistakenly think it was low-fat anyway.
On the bottom, you see that they sought a balance of saturated, polyunsaturated, and monounsaturated fats, and they served plenty of meat along the way. This diet was also not low in cholesterol.
Gary Taubes tells us that in this trial, the experimental group lost more members to heart attack than the control group. He tells us that one of the researchers wrote that this was “somewhat unusual” before he talked more about what he wanted to talk about, which was an improvement in risk factors. Taubes is giving us the impression that this diet caused more deaths than the control diet, and they tried to gloss over this embarrassing fact. Please pause and read this carefully. “Eight members of the club died from heart attacks, but none of the controls.” This was called “somewhat unusual.” This is pretty clear, right?
Now read the paper for yourself. “Of the eight subjects with new coronary disease events in the fully participating experimental group, three died of coronary heart disease, one died of other causes, and the other four were still alive at the end of the observation period.” This says there were eight new coronary cases in the experimental group.
Taubes did not mention that the saturated fat-based control group racked up 12 new coronary cases, did he? Yes, they all were still alive at the end, and here you can see that this was indeed called unusual. But then keep reading. The experimental group had worse risk factors from the start. The groups were not precisely matched. You can see at the bottom that these were very big differences. 11% more were obese at entry. 15% more had hypertension. More participants had multiple risk factors. 17% fewer had no risk factors. The experimental group had the odds stacked against them from the beginning by design because they started with less-healthy people, but Taubes didn’t tell you that. Yet he had to have seen this to use that quote. How is that journalism? Why do you low carbers trust him so much?
In the first paragraph, you can see that they said they expected more heart disease events in the experimental group for all these reasons.
So since I just showed you that only three died in the experimental group from heart disease, how does Taubes figure that eight members of the club died?
He added in five deaths from another group in the trial, an inactive experimental group.
Here you can see that these men were called inactive because they were barely in the program. Two thirds got an annual physical and some lab work. The rest only were in contact through the phone or the mail.
Now remember, the journalist Gary Taubes wants you to think the experimental group had a mortality problem.
He wants you to believe that rather than this. Here you can see with your own eyes the differences in coronary disease events. This is the truth he would not permit you to know. If we compare using the active experimental group, we can see that the controls eating more saturated fat had adverse events at almost three times the frequency as the men eating less saturated fat. And remember, the control group had less obesity and high blood pressure going into this. This is the extent to which the journalist Gary Taubes is willing to twist the facts. I am not asking much, Mr Taubes.
TAUBES: Just, stay honest, Dean. That’s all I’m asking, ok?
Here it is spelled out. “Three times as high.” Moreover, the performance of the experimental diet seemed to be improving over time. This raises questions about what might have happened had the trial been extended. Isn't that right, Chris Masterjohn?
The mortality issue is a red herring to distract you from this powerful evidence. Over time, it has become clear that there was no increased mortality compared to saturated fat-based controls during these trials that emphasized vegetable oils. Once again, I reject the use of these oils, but I am here to report the science.
Beyond that sort of research, there is a lot of other research on low cholesterol itself and its effect on mortality. This important study by Strandberg measured mortality over a 39-year stretch. Low cholesterol in mid-life predicted better survival.
Please read this slide. “There was a graded association between mortality risk and baseline serum cholesterol.” “Differences were highly statistically significant.”
Look over this slide and you’ll see why this is such a good study. The authors explained that comorbities tend to lower cholesterol later in life. They understood the need for long time frames to see the effects of high cholesterol. They were aware of statistical traps like regression to the mean.
Additionally, there has been research into the effects of genetic polymorphisms that lower cholesterol. This study found that people with genetically lowered LDL had a much lower risk of heart disease. They also found that genetically lowered LDL had no association at all with increased cancer risk.
Mr Taubes, the science is in on this. Your two implied questions are answered. Lower cholesterol levels result in fewer cardiac events. Lower cholesterol levels do not increase mortality. Why don’t you publicly state that your attempts to scare people away from trying to lower their cholesterol levels have been wrong and counterproductive and harmful to the health of your readers?
Before I leave the topic of diet-heart intervention trials, I would be remiss without mentioning the Oslo Diet and Antismoking Trial. This was an important study that you won’t read about in any Gary Taubes book. In this one, the results on both coronary disease mortality and overall mortality clearly supported efforts to lower saturated fat. Separate cohorts were created using men with high cholesterol and a high risk of suffering a first manifestation of heart disease. The intervention group was told to lower their intake of cholesterol and saturated fat and to cut back on their smoking.
After five years, the guys in the intervention group did lower their consumption of saturated fat, and as a consequence, their cholesterol levels were 10 percent lower than the control group's.
Even though they were asked to quit smoking, the intervention group’s smoking behavior didn’t differ that much from the control group’s. The intervention group had a huge 47 percent reduction in events related to coronary disease. They also had a 33 percent reduced risk of mortality.
The researchers who ran this trial found that the explanation for this success was a drop in blood cholesterol levels, not smoking behavior. In these slides you are looking at a later paper from these investigators which followed up on the participants after eight to nine years. Even after the end of the trial, the intervention group was still doing better than the control group.
If we look at the paper that was published after the five year trial, we can see that Oslo stands apart from the oil-based studies. While this was far from an ideal diet intervention, the experimental diet really was lower in fat than the control diet, and that was mostly due to a reduction in saturated fat. They didn’t get crazy with the vegetable oils like the other researchers did.
While cholesterol wasn’t lowered all that much, remaining far too high in the intervention group, the difference was enough that the researchers could clearly separate the effects of their diet from the effects of their reduction in smoking behavior. The dominant factor in the reduction in heart disease events was clearly the dietary change.
The Oslo trial was a pretty good demonstration of the validity of diet-heart even though their intervention diet shouldn’t be a model for anyone. This study is harder for the deniers like Taubes to pick apart. Even if there were a reasonable critique to be made of it, eventually, a fair-minded person should notice that between this trial and all the others, saturated fat has certainly never been shown to be protective against heart disease in a trial. If saturated fat were truly healthy and innocent and if polyunsaturated fats and lower cholesterol levels were so bad, you would think that there would be at least one study out there that has shown us that. I don’t know of any such trial. The Minnesota study seems to be the best the deniers have, and as you saw, it wasn’t worth much.
In the next video, I will approach the misreporting of Gary Taubes in another way. Taubes is very selective in his use of quotations. He thinks that as long as he is quoting accurately, he is reporting honestly. Honest reporting doesn’t work like that. I’ll show you that only the quotes he likes make it onto the pages of his books, even if they misrepresent the ideas of the person he is quoting. The Taubes Filter is next.