Vegetable Oil Studies That Need a Closer Look

Vegetable Oil Studies That Need a Closer Look

Producer Potts 0:12

Since publishing your in depth article on vegetable oils and busting the myth that they are toxic and inflammatory, I have seen a lot of pushback and quite a few people really wanting to debate you on this topic. Those types of conversations are not always productive. However, I wanted to give you the floor today to take us through these old studies that keep popping up as proof that these vegetable oils are unhealthy. Can we please start with the Sydney Diet Heart Study?

Dr. Sarah 0:47

Absolutely. I think there’s three studies that are quoted over and over and over again from the 60s and 70s. As like See, here’s here’s the proof that vegetable oils are actually toxic and inflammatory. And I think it’s really important to kind of go through these studies and look at the methodological flaws and show that while there’s some interesting things that come out of these older studies, they typically followed study participants for a lot longer than a lot of the modern studies do, which is really cool. There there’s other problems with these studies that mean that more recent studies are a higher level of scientific evidence than these older ones. So yeah, let’s start with this Sydney Heart Study, and I’m actually going to pull up some notes so that I can go through these methodically. Yeah, let’s, let’s dive in almost. This is almost like a non debate, debate. So dive in and and show us what’s what. So the Sydney Diet Heart Study ran from 1966 to 1973 and it included 458 men aged 30 to 59 and they had all recently suffered a coronary event. So even though the original study was published in 1978 an updated analysis was published in 2013 that kind of reopened the conversation about this study and implicating that vegetable oils resulted in a higher risk of death from all causes, coronary heart disease and cardiovascular disease. So one of the strengths of this study was the use of safflower oil, because it is almost exclusively linoleic acid, so it prevents some potential like confounding effects from other vegetable oils like canola oil, that also have high omega three intake. Although canola oil didn’t exist yet in the 60s and 70s, have not yet been invented, which also fun fact. So in the re-analysis and looking at like both of these publications together, the claim is that the vegetable oil group, so these were study participants swapping out saturated fat for vegetable oil. It showed that vegetable oils increased total mortality by 62% coronary heart disease mortality by 70% and cardiovascular disease mortality by 74% which is very, very scary. Those are huge, huge percentages. So here’s here’s the flaw. Here is the reason why we can’t put any stock into those results the group that was given safflower oil, they were they so they were swapped out saturated fats in their diet, like butter for safflower oil. Were given safflower oil in standard liquid form, like you, you buy the bottle of safflower oil from the grocery store, as well as a soft margarine a brand called Miracle, that should get little, little alarm bells going, because in the 60s and 70s, margarines had very high contents of trans fatty acids. Miracle specifically was 15% trans fats. So these study participants did swap saturated fats for linoleic acid, but they also dramatically increased their trans fat intake. So which, at the time we had we didn’t know how bad trans fats were, right? Like we had no idea.

And it’s worthwhile emphasizing here that trans fats have been phased out of the food supply. They cannot be added to foods anymore. There can be small amounts formed in processing, but there’s limits on how much a food can contain. I think it’s like point 5% or something like that. So margarine nowadays does not have trans fats. Trans fats came from partially hydrogenated oils. Nowadays, margarines contain fully hydrogenated oils, which is turning the Omega six fats in the vegetable oil into a saturated fat, not turning it into a trans fat. So just being very clear, there’s trans fats and margarines. So just as a point of clarification, because I this is mind blowing, I did not know this. So essentially, they’re not even using the same product as what we have now, basically Correct, yeah, so we can’t even, like, redo this study. Well, first of all, it wouldn’t be ethical to give people trans fats when you know it’s going to make them sicker, right? Like, can’t we can’t do that, but, but, yeah, if you were to redo this study today, I would say I don’t even know that giving like safflower oil and a safflower oil based margarine, I don’t think we would even do that. It would literally be just safflower oil, and you would basically say, like, yeah, don’t, don’t butter or margarine, your toast, right? Like, that’s, that’s why these study participants were given a solid form. It was to replace butter. So butter for cooking. Well, it’s easy to pour some oil into a pan instead of a pad of butter into a pan, but as a, you know, on a bread, or, you know, I don’t know, a muffin or something like that, that’s where the solid fat would have been used. And I think nowadays we’d say, you know, hydrogenated safflower oil, or in those days, partially hydrogenated safflower oil, that’s no longer the same fat. So that doesn’t actually help us understand the linoleic acid. That’s a variable that we would now control, and it wasn’t controlled in the 60s and 70s. That’s fascinating. Okay, so that’s, that’s, that’s the big we don’t know how much of those effects was trans fat people.

Producer Potts 7:00

When people bring up that study, they don’t, they don’t mention that there were trans fats. Conveniently Interesting. Okay, just a point of like, I see you. Okay. So the second study that came up a lot was called the Minnesota coronary experiment. So can you walk us through that one?

Dr. Sarah 7:20

Yeah, for sure, and again, checking my notes to make sure I have this exactly right. So the Minnesota coronary experiment was conducted from 1968 to 1973 it included 9000 patients that were given either a control diet, which was 18% saturated fat and 5% polyunsaturated fat, or the intervention diet, which is 9% saturated fat and 15% polyunsaturated fat. It’s worthwhile taking a step back to actually talk about why these experiments were done in the first place. They were done to test the Diet Heart hypothesis, which was the idea that saturated fat was increasing cardiovascular disease risk, which there is a fairly good body of scientific literature showing, although it’s complex and it’s being re evaluated now, but that’s that’s why they were comparing such a high saturated fat diet to a high polyunsaturated fat diet, because that was, that was, that was the question that we’re trying to answer at the time. So similar to the the Sydney Diet Heart Study, this one was first published in 1989 and then there was a re analysis published in 2016 so the original analysis basically showed that the vegetable oil reduced study participants cholesterol oil or cholesterol levels, which is good, right? News, right? We’re like, yay. Lower Cholesterol, good. So that was what was published in 1989 but what was published in 2016 was that the people with the lower cholesterol levels had an increased risk of death, which is weird, and it only applied to study participants over the age of 65 and it was literally For every 30 milligrams per deciliter reduction in total cholesterol, risk of death increased by 22% which, again, that’s a very strange thing. That’s not what all of our modern science tells us. So here’s here’s the flaw, here’s where that weird result comes from. 83% of the study participants were lost to follow up in this study. That means they started the study and they weren’t there at the end. That is a crazy attrition rate that is not seen in Contemporary Studies unless you’re evaluating something that’s like, really hard. So like some of the, you know, really extreme diet studies will have the the group that has the really, really hard diet will have really, really high dropout rates. But when you like, you’ve, they’ve lost them. It’s not just that this, the participant dropped out. It’s like they couldn’t get in touch with them again. They have no idea what happened to them. So.

Producer Potts 10:21

So let me just ask a quick question here. So, in today, how we do science now, and how we we would that even be published today? No, see, that’s fascinating. So it was just, I mean, it’s just like, how did they get away with that? Though, that seems like such a I mean, I don’t even have a good comprehension of how scientific studies are run, but to me, that’s like a very basic thing of like, you didn’t have the people at the end. So like, how, how did Wow, that’s mind blowing.

Dr Sarah 10:59

It was a little bit easier to lose people in the 60s and 70s than it is now. So like, now, we have so many different ways of contacting people, right? We have the internet like we have we have different ways of keeping track of of people. Whereas back then, like, if you moved and didn’t let the study know, they would go, that’s a good point. Send a letter to your address and you just don’t live there anymore, right? So, okay, that does make sense. Then I was thinking, like, how would it, how would that even Yeah, so that makes sense. That makes sense. It wasn’t necessarily the big alarm bells that you would think now, like, that’s 83% of all of the study participants, not just the vegetable oil group. It’s not like all the vegetable oil people went away. It’s of all of the study participants. So they just like they they lost track. They started with with 9000 people and ended up with whatever 83% loss of 9000 is, yeah, 1000 and change so, so, yeah, that’s, that’s, you know, at the time, it was like a little bit the difficulty, but then you don’t know what extra factors you’re introducing into your data set at this point, so it causes, potentially, something called attrition bias, which they wouldn’t have known at the time. Although the 2016 reevaluation should have, should have been, that should have been a major consideration. But what it means is, like you don’t know, are you losing more people from one group than the other, or are you losing people for specific reason that might skew the data, right? So the people that you lost you lost, not because they moved away, but because they died. And we don’t have the same kind of records to be able to look that up now that like nowadays, we could just look that up, but in the 60s and 70s, little bit harder to track down that information, so that was, and that is, basically, that is the spoiler. That is the flaw of this study was attrition bias, so selective dropout of participants who have important differences than those who stay in the study so they become lost, not because they moved away or because they found it really annoying to participate in the study, but because they became ill or because they died. So looking at the data more specifically, there were 21 excess deaths in the vegetable oil group, exclusively among patients over the age of 65 in the control group, 15 more patients in this age range were lost to follow up without explanation, compared to the vegetable oil group. So the re-analysis goes, oh, a vegetable oil group had 21 more deaths, but they’re not accounting for 15 extra patients lost in the control. So they’re comparing against, you know, control. But what we don’t know is if those 15 people in the control group actually died, in which case, it’s not 21 excess deaths, it’s six. And that’s not meaningful with this big of a data set, right? So that is like, that is the question, and I’ve got more evidence for why that’s likely. What’s going to happen the association between greater reductions in cholesterol and higher risk of mortality wasn’t necessarily due to diet.

Here’s another factor. So there’s some illnesses where like leading up to the illness, part of the illness is a drop in cholesterol level, and that is another possible explanation for why lower cholesterol was associated with increased risk of death. It could be that it was a like symptom of the illness, not a cause of the illness. Does that does that make sense? Yes, that makes sense. Yes, totally. They also had the trans fat problem that was also in the study. So they also. Consumed some of their like vegetable oils in the form of margarine. That was additional problem here again. Okay, fill in in here, but here’s where we go. There’s, there’s something going on. Either there was more, you know, people who were were sick, and that’s why their cholesterol is lower and or we’ve got some kind of attrition bias effect where the people who dropped out dropped out for like, a reason that skews the data, because this study also showed that smoking higher body mass index and higher diastolic blood pressure were all associated with lower mortality risk. So these other three things, right? High blood pressure, high body mass index and smoking are all associated typically with increased risk of all cause mortality, and this study showed the opposite. So this study had all kinds of results that are not any sense with, yeah, you’re not in line with like the vast, huge, giant body of scientific evidence. If this study is showing that smoking reduces mortality risk, we really need to call into question the fact that it showed that vegetable oils increased it right? Something wackadoo is happening, and it’s almost certainly because of that 83% of study participants lost to follow up.

Producer Potts 16:33

Gotcha, wow. Okay, so that study had a lot of issues. Okay? Problematic. Bye, bye, Minnesota. Okay, the last one is the LA Veterans Administration study. So now this is, I almost feel like I’m getting the inside tea. This is actually fascinating to me. And I think this is so interesting, okay, I think the the it’s like the drama that must have been happening behind these scenes when these studies are like, going on and being published.

Dr. Sarah 17:06

I mean, hang on, because let’s, let’s actually, like, jump ahead before we even talk about the like LA Veterans Administration study. Because often these three studies are presented as, right? These older studies that were so they were done for so long, and they were, you know, like they use safflower oil, which is all linoleic acid, right? All of the strengths of the studies are presented, and they’re also presented as, like the three big studies from the 60s and 70s that show the vegetable oils are bad. How, how did we get here with all these vegetable oils? It must be insert conspiracy theory, right? But these were not the only studies from that era looking at this question, and there’s a bunch of other big studies with all of these same strengths and, you know, some of the same limitations, right, like unknown amounts of trans fats, but that all showed that vegetable oils were beneficial. So these were like even cherry picking these studies. First of all, you know, putting higher weight on these flawed studies from the 60s and 70s than all of the way more rigorous, well designed, well controlled studies from the 90s, the knots, the 2010s and all of the meta analyzes that look, you know, pool together all of those studies. Like first of all, when you’re using these studies as your argument, you are ignoring way better design studies, but you’re also ignoring the other studies from the same era that don’t support your argument. So you’re already cherry picking studies from the 60s and 70s to support your argument, ignoring all of the important science that came after. That’s fascinating. Okay, well, that’s major tea okay, so I like it.

Okay, so the LA Veterans Administration study included 846 male veterans. They substituted vegetable oils for animal fats, and looked at cardiovascular disease risk. Both groups included or had diets that were 40% calories from fat. In the control group, 10% was from linoleic acid, and in the intervention group, 38% was from linoleic acid. So basically swapping out like three, nope, more than that, swapping out most of the saturated fat right for for vegetable oil. And the results were originally published in 1969 after an eight year follow up period. So what this study showed that when they pooled cardiovascular events together, so that included heart attacks, sudden death and stroke. The vegetable oil group saw significantly fewer events than the control group, so the instance rate was 31% compared to nearly 48% so. And the benefits of vegetable oil consumption were seen in study participants under 65 and a half, as well as for everyone who had like higher cholesterol at baseline. So cholesterol over 233 for total. But there was so that that all falls in line with vegetable oil is good like that. That’s not, that’s not the thing that the anti vegetable oil crowd are quoting about the study. What they are quoting about this study is this surprising result that was not related to cardiovascular disease, which is that the vegetable oil group experienced an excess of cancer deaths compared to the control group. So the vegetable oil group had 31 deaths due to cancer compared to 17 in the control group. And with, you know, only 850 people in this study, that is a meaningful difference. So then what the quote of vegetable oil people will say is there were more heavy smokers in the vegetable oil group, 70 versus 45 who were smoking at least a pack a day. So because of that, you would expect the control group to be even higher risk of cancer to begin with. So that must mean this vegetable oil effect causing cancer is even worse than what this data shows. So that is, that is the the narrative from this surprising result from this study. But of course, there’s more. There’s always more. There’s always there’s more. I’m waiting. The single most important thing is that the difference in deaths, while it’s it’s a it’s a big number, right? Like, what did I say? 31 versus 17? Like, that’s a big difference in a study of the size. It still wasn’t statistically significant. So when you actually do the statistics and you test to see what is the likelihood that this difference is due to random chance, or is a real difference, it failed that statistical test, which means there’s a higher chance that this is just random, that this is not an actual, real thing. That’s why the statistics is done, and those statistics were done in the 60s, so, like it was known way back then, that there’s a high chance that this is completely chance. But the other thing is, the heavy smokers statistic is very misleading, because the vegetable oil group actually had more total smokers. So if you just, if you just take heavy smokers, yes, the control group had more. But if you look at all smokers, and smoking was much more common in those days, there were more total smokers in the vegetable oil group, which we know you don’t have to be a heavy smoker for that to increase risk of cancer, right? That even secondhand smoke increases risk of cancer. So more total smokers is actually the more important measurement. It’s just that in the 60s, they didn’t know that yet either, right? So they were doing the analysis to try to control for heavy smokers, not trying to control for total smokers. Have that data, and when you look at that data, you go, Aha. You know, there was more total smokers in the vegetable oil group, and that almost certainly accounts for that difference in cancer incidents that plus, you know, it’s, it’s noisy, it’s nice, there’s, there’s enough question about whether or not that’s even a real difference in the first place. We can’t, we can’t put like, we can’t draw conclusions on that. And when you compare that to this huge body of scientific literature that has come since that shows that vegetable oil actually doesn’t have a meaningful impact on cancer incidents like it’s pretty neutral. When it comes to cancer, it reduces risk of cardiovascular disease and cardiovascular mortality. It reduces risk of all cause mortality, and it does reduce risk of cancer mortality. But other than breast cancer, where vegetable oils do look protective, more data needed. Overall. There doesn’t seem to be an impact of vegetable oil consumption on cancer, and that’s that’s okay, like, not everything we eat needs to reduce risk of every bad thing that could possibly happen to us. Vegetable Oils reduce risk of so many different bad things. But yeah, there’s, there’s tons of data that has come since this study that has helped to to clarify the role of vegetable oils in cancer incidence. And this was pretty, pretty unreliable data to start with, interesting.

Producer Potts 24:28

Okay, well, I actually found this really fascinating. Like I said, I felt like I got the behind the scenes, like a tea of these things is really interesting, like the details of the study and how they’re done, really interesting, and especially the fact that I have been with you today, and we went through the video that talked about, you know, busting the myth and all the studies that were in that, and then the health benefits and all the studies that were in that, by the way, we’ll make sure all the. Videos are all in the playlist for you. Sarah has a detailed article. I’ll make sure that’s here. But I think it was really also good to take a look at these studies that just get keep people keep bringing up, and to know, like, how they were done, the amount of people, what was wrong with them, and, you know, the way that they’re brought up makes them sound like they’re this, like, end all, be all. And actually, they’re really nothing that’s

Dr. Sarah 25:29

That’s correct. If you put this video side by side with the video where we looked at all the health benefits of vegetable oil, and we’re talking about, you know, sample sizes of 40,000 300,800 100,000 800,000 you know, right since versus 850 right, like, right 9000 but we lost 83% of them. Right? Like, it’s just the thing, the quality of those data sets are not comparable.

Producer Potts 25:59

Gotcha totally, totally. And it was good to have the comparison. And like I said, yeah, those two videos side by side. Definitely just again, you’ve given us the sledgehammer to break through this wall of like, vegetable oil, fear mongering, myths, all this, all the mess, all the mess. Okay, well, thank you for this. We have a couple more videos to do on this topic. So again, they’ll all be in a playlist. There’s an article. If you have a question, leave it below. We might do some more follow up videos on this and just keep your comments kind and full of curiosity is what we ask. Thanks. Dr Sarah, thank you.