Loading summary
Transcript35 lines
- [00:00]
A
Foreign.
- [00:14]
B
Hello, and welcome to another episode of Sigma Nutrition Radio. My name is Danny Lennon and with me is Dr. Alan Flanagan. Alan, how are you?
- [00:21]
A
I'm very well, thanks. I'm surprised we're back after episode 600,
- [00:25]
B
but there was enough of a fervor for us to return that we gave in to peer pressure from the and we have decided to return. Like all good retirements from sports, you always see them come back.
- [00:38]
A
It's the classic the Last Dance.
- [00:39]
B
Yeah. It's the boxer that's going to return and will stay returning until they get knocked out.
- [00:44]
A
We're punch drunk.
- [00:45]
B
Yeah, we just couldn't know how to give up. It's a classic one where they say, oh, he should have retired and should have stayed retired.
- [00:51]
A
We might be past our peak.
- [00:53]
B
Yes, I think certainly the deadpan nature of our last discussion maybe had caught a few out because I had to answer a number of messages and emails laying people's fears. The podcast was indeed gone, so maybe we need to signpost our deadpan delivery of those comments in future. But indeed, we are back and we are taking on a topic that continues to generate debate and also interest. In the past, we have talked about the topic of unprocessed red meat and chronic disease risk, broadly, mainly focusing on the epidemiology in the area. So we're not going to rehash all of that in today's episode, although it's important that we do touch on it. Recently on the podcast, I was talking with Dr. Gil Carvalho and Dr. Mario Kratz, specifically on the outcomes of type 2 diabetes, insulin resistance, and then some of the stuff around cardiovascular disease, specifically deciding not to touch on the outcome that we'll look at today, which is cancer. And I suppose the outcome of cancer is particularly important because it's perhaps the one where we have a lot of the research that gets debated and is perhaps one of the major concerns around consumption of unprocessed red meat. And again, with the same caveats that were applied to that episode today, we are going to keep our focus to unprocessed red meat. So not including processed red meat and so focusing in on everything that comes under that umbrella of red meat in an unprocessed sense, and looking specifically at the outcome of cancer. And as we'll discuss, that is typically relegated to one particular cancer type or a subtype. There's a few within them and we're going to walk our way through this. So maybe a good place to start here, Alan, is when most people hear about this topic around red meat and cancer, one of the things that gets referred to is that it is being classed as a carcinogen. And there's some differences here between processed and unprocessed. But even in relation to unprocessed meat, we have these designations by, for example, the International Agency for Research on Cancer as red meat having being probably carcinogenic to humans. The WCRF has a very similar label or being a probable cause of colorectal cancer. Can you maybe just touch on these designations of what these classifications mean, what they are built on, and anything else that people might have heard that we might need to add some nuance to?
- [03:23]
A
I think both are relatively similar in the conclusion that they arrive at from a causal standpoint, which is probabilistic language of unprocessed red meat as probably carcinogenic to humans. They slightly differ in the emphasis on evidence and the types of evidence that are emphasized. So with the IARC definition, that's something that prioritizes long term prospective cohort studies coupled with mechanistic research. And the test that they apply is that there are consistent associations in different populations that then mean that other factors like confounding or chance can be ruled out. And the wcrf, which also arrived at this determination of probable cause for red meat and colorectal cancer, is based largely, again similarly on identified mechanisms and those specifically in conjunction obviously with the long term epidemiology. And those specifically relate to the formation of what are known as heterocyclic amines and other compounds known as polycyclic aromatic hydrocarbons, and the potential for carcinogenesis to be induced by a group of compounds known as N nitroso compounds or N nitrosamines, and those compounds, the production of those compounds can be stimulated by heme iron, which is obviously highly concentrated in red meat. The main distinction, and where the controversy lay with the IARC and WCRF designations, was in relation to the distinction between processed meats and unprocessed meats. And processed meats were much more weighted towards a kind of causal determination in terms of the strength of the evidence and the consistency and the clear dose response relationship evident in prospective cohort studies, which would all satisfy a kind of probabilistic causal framework from epidemiological data. Processed meat, sorry, unprocessed meat, or what we might otherwise term fresh red meat, was the more controversial finding because some of the dose response analyses were not significant, for example, for unprocessed red meat per 100 gram a day increment increase. And so the conclusions and the use of probabilistic language was essentially in spite of a lack of significant association in some of the cohort studies with unprocessed meat intake specifically, and more of an emphasis on some of these mechanisms. And that was where most of the, shall we say, kind of issue was taken with certainly the WCRF classification. But I think, as we will hopefully highlight today, it wasn't a misguided evidence assessment. And while we do have issues that arise in the epidemiology of unprocessed meat in relation to dose thresholds at which risk may be observed, there is, when people tend to look at this evidence, a tendency to say, well, this is all based on epidemiology, so we'll never get further to reconciling that. And that's actually not the case. And there is a body of evidence that we'll discuss today that really goes back to the mid-1990s looking at, in humans, very controlled, elegant feeding studies looking at some of these intermediate mechanisms that may explain a long term relationship with colorectal cancer risk from unprocessed meat.
- [07:00]
B
And that will become important because it's often maybe left out of some of the conversations or debates that are commonly had and they end up devolving into whether someone wants to actually discuss the epidemiology and appraise that properly or they just want to dismiss it outright. But of course, that still serves as an important part of this evidence triangulation. We've talked about that previously on an episode more in depth. So if people really want the details of some of the best epidemiology in this area to answer some of these questions, that is worth revisiting. We also have an article on the website that might discuss that a bit more. And indeed some of the work. For example, Catherine Bradbury was on the podcast discussing their specific work in this area, answering these questions around red meat colorectal cancer risk and their epidemiology. They're based on the UK Biobank data, so I'll put those in the show notes if people want real details on that. But for now, given that we want to get into some of those real controlled studies looking at these controlled trials in humans, first let's touch on the epidemiology because it can't be left out. And it gives us a really important basis for seeing why we're even investigating this here. And there's probably a few of the most important pieces of evidence in this area that might do a good job of summarizing generally some of not only the direction of those effects, but some of the nuances within interpreting it, which is why there's maybe not one necessarily black or white answer. When we're looking at the EPI in this area, a lot depends on those cohorts, those exposures. We're looking at what specifically we're answering with these questions. So if we're to talk about some of the epidemiology in this area from this overview level, so we, without us getting all into the details of every single study, what are some of the most important points that we can touch on related to the epidemiology that you think lead us to be able to say some things with some degree of confidence that we can come away with the EPI with some clear signal relation to this question?
- [09:05]
A
Yeah, for me, the most striking feature of this, and this relates to most of the epidemiology, or in fact, all of it, of unprocessed red meat and any disease outcome, including in this specific instance of our focus today, colorectal cancer, is regional variation in the presence of positive associations, that is higher risk associated with higher intake. And where we see a regional variation like that in any sort of exposure outcome relationship, that's a signal to look a little closer at the characteristics of the exposure in those different populations. And what we tend to see is a kind of general overview of this area is that in, for example, European or North American cohorts, we more consistently see higher risk associated with higher intake. And of course, as we've discussed on numerous podcasts that relate to nutritional epidemiology, the problem is this abstract use of the term high, quote, unquote. And so we really need to start to then define what we mean by high and what we mean by low, and what a range of actual intake, in terms of grams per day, for example, in absolute terms, is in that population, because a relative comparison of high and low, without further describing that, tells us very little about the characteristics between populations. And when we do that due diligence, we typically find that you will have certain populations, North America in particular, where the absolute actual intake is almost two to threefold higher than it would be, for example, if you took a cohort from the Asia Pacific region. And that gives us something to then start to work with. And generally speaking, a relatively consistent picture emerges where dose thresholds of at least around 100 gram a day or less of unprocessed meat intake don't really produce any sort of signal in the noise for any of these outcomes. And yet, if you go over, then in terms of Some of the North American, and indeed within Europe, some of the European cohorts that tend to have higher intakes of 160, 170 over grams a day. And that's factoring in the kind of well established underestimation that tends to result from dietary assessment methods in nutritional epidemiology, we're likely looking at higher again actual intakes on average terms. And so the regional variation is important because that regional variation speaks to differences in absolute intake. And that appears to, particularly for unprocessed meat, be a really crucial factor because the absolute intake, the intake in grams per day, is what is going to be important in trying to explain how unprocessed red meat could in fact drive some of these intermediate carcinogenic processes such as heme iron mediated nitroso compound formation or the production of polycyclic aromatic hydrocarbons or hydrocyclic amine, heterocyclic amines as well. And that for me, I think is probably the top line, most important consideration that we can have in this literature. And for colorectal cancer, I think a good example with the Chan and colleagues, well, it's now quite dated, but it's still a good meta analysis of that distinguished red and processed meat and colorectal cancer risk. In a meta analysis. And so you had 24 prospective cohort studies that spanned North America, Europe, which was 10 countries from the EPIC cohort, Asia Pacific region, including Japan, China, Singapore, and then you had Australia and Canada. And in that analysis, for example, red meat alone, unprocessed red meat per 100 gram a day increase was associated with a 17% higher risk of colorectal cancer. The same obviously for processed meat, that's typically a stronger association. But again, in terms of stratifying by geographic region, that was strongest in Europe and North America, whereas in Asia Pacific cohorts there was no significant association. And that's because you can't model a per 100 gram a day increase necessarily in a population that are consuming 70 grams or 60 or 50 grams a day on average. Right. You can model it, but again, you're not getting an association or a signal in the noise simply because the habitual intakes in those populations is relatively low and within this range at which we don't see risk. And I think that gives us a kind of general overview of the EPI in a way that allows us to then take some of those questions to stress test in some of the controlled feeding studies that we have.
- [13:56]
B
I think that's a particularly important point. That we can turn to these control trials because they allow us to maybe fill in gaps that are very difficult to do otherwise. So, for example, one thing people may be thinking now is that, as you mentioned, when we have these lower doses that we're looking at and maybe a comparison where everything is under a hundred grams and let's say we're comparing 70 as the quote unquote high in a certain position, and how we interpret that might be worth talking about because maybe there's a temptation to say, well, below these certain thresholds you've just mentioned, there is no risk, as opposed to saying below these certain thresholds we are unable to detect a risk from this particular epidemiology. And those are two, two different things. And you see, depending on how someone interprets that, they might say, well, that is true, we're not detecting effect, but based on all this direction, we would still assume there is risk even at these lower intakes. Someone else might say is, well, we can't say that. So at these levels of intakes, it seems that there's no increase in risk. And so therefore they're fine. And so those might be different conclusions to saying we're not detecting a risk. Can you maybe talk a bit about the interpretation of the epidemiology given this kind of gray area that we're referring to?
- [15:11]
A
Well, I think it obviously depends on the outcome that we're looking at, but you've outlined very well essentially the two main interpretations that were offered by an association like that. So what I think is always really important then is to really look at our risk estimate and the point estimate of that hazard ratio or the relative risk and the corresponding confidence intervals to see if there's any kind of direction that we can glean. Now, often let's take for example that Chan and colleagues cohort the relative risk by geographic area for the Asia Pacific analysis specifically was a relative risk of 0.94, so a 6% lower risk. But the confidence intervals went from 0.69 to 1.27. So that is spread entirely across. There's no direction of effect there evident in terms of that range. And so that's something that I think you would generally interpret more as a no association rather than perhaps interpreting it as not. And the potential there is that is something because of the confidence intervals that could, it's obviously no association in that analysis, but there is room open for it to be could not detect an association. Whereas with some of the cardiovascular studies, for example, when cardiovascular disease is an outcome in some of these East Asian cohorts, the entire point estimate and the confidence intervals range are actually in a direction of under 1.0 and in some cases statistically significant under 1.0. And that's often sex based those statistically significant differences in Asia Pacific or East Asian cohorts, specifically lower risk observed, that's statistically significant in women, but not men. And again, the paradox of heme iron in women in reproductive age, it may be because it's making a contribution to heme iron, who knows? They're all the speculative components. But I think in terms of how do we then say, is this just no association? But with the door open for potentially that particular analysis of that particular dose comparison range not being sufficient to detect an association, or whether we can say we're actually looking at more of a kind of a lack of risk. It depends really on what the effect or what the sorry risk estimate actually is and what we can glean from the overall and the confidence intervals of that risk estimate in particular.
- [17:45]
B
So that turns to why it might be important for us to look at these other areas that could fill in that gaps to suggest, well, these areas where either we're unable to detect an effect in those situations, that could be down to, first of all, not there not being an effect or negative effect, let's say at those levels of intake. Or it could be, as we've said, it could be that there is some degree of effect, but just not that we're able to detect it here. And so to be able to try and answer which one of those it is, we can try and look to some of this other work. And as you've alluded to, these are really nicely done studies, but maybe are oftentimes left out this conversation. We spend all the time looking at this EPI and debating what we can take from it or not. And this might be useful to, to walk through. You've already discussed those primary mechanisms that have been typically hypothesized as to why red meat could be particularly could be playing this role in the development of colorectal cancer, specifically, some of those relate to one one another, for example, the heme iron, potentially setting the stage for some of these other compounds that are having a problem. So if we start walking through some of these and referencing some of the studies that you've highlighted as particularly useful and particularly interesting to touch on these topics, that might be a way to give people an idea of what we actually know in this area. In these controlled trials in humans, one of the first and one of the ones you've highlighted that talks to some of this issue around these N nitroso compounds that we've referenced was from Bingham and colleagues in 1996. Can you walk us through why this particular study stands out to you, why this is a useful one for us to take into account?
- [19:28]
A
Yeah, so there's, like I said, a body of work in very nicely controlled trials that we can draw on. And this I think is quite a seminal paper in this area published in 1996. And it was a metabolic ward study where eight participants, male participants, age 24 to 32, so young, lived in a controlled metabolic facility and consumed diets that were fully controlled, prepared under metabolic kind of kitchen conditions. And they had two different protocols that the participants were exposed to with different setups for their diet. In the first, they had a low red meat diet with 60 grams a day of red meat. They also had a high red meat diet with 600 grams a day of red meat. And then they had a high red meat diet that was supplemented with additional 20 grams of wheat bran to add a fiber based supplement to the high meat diets. The same 600 gram a day red meat dose with that diets were matched for energy intake. Their fat content was the same across all diets. And it used a crossover design for the participants to go through the first two diets, either the low or high red meat, and then cross over to the brand supplemented high red meat diet. And then they also had a second protocol with a low red meat, same sorry, low white meat, 60 grams a day, high white meat, 600 grams a day, turkey, chicken and fish primarily, and then the high red meat again 600 gram a day diet. And they were primarily interested as their primary outcome in fecal excretion levels of total N nitroso compounds. So these compounds are these alkylating agen that can promote potentially mutagenic and carcinogenic processes in the colon. And they then also looked at faecal nitrite, nitrate, iron levels and other outcomes. And for the primary outcome of the faecal nitroso compound formation, those compounds, the generation of those increased significantly on the high red meat diet, as we might expect with a dose maybe of 600 grams a day, given kind of hypothesis in this area. And then interestingly, on the high red meat, with the addition of the bran, the nitroso compound formation was still significantly higher than the low red meat 60 gram a day condition, but not significantly different from the high red meat diet. So effectively the increased red meat consumption led to this significant and quite rapid increase in endogenous faecal nitroso compound formation and the level of nitroso compounds that was produced from the 600 gram a day red meat diet was comparable to the carcinogenic exposure that you might get from tobacco smoke. The white fish condition, that second protocol, the white meat or the white fish condition, didn't increase faecal nitroso compound formation or nitrite, which again would be consistent with the epidemiological evidence that we have, which you see no real association for white meat and you see benefits in substitution analysis. That model replacing, for example, 100 grams of red meat with 100 grams of white meat and the bran supplement, and this is important, did not reduce the N nitroso compound formation. And so at that high dose of 600 grams a day, the addition of that bran supplement made no difference to the N nitroso compound. So now would the same effect be seen with an intermediate dose of something in the region of 250 grams or 300 grams? That's the limitation of this study. It was deliberately in its setup creating a very wide exposure contrast. But it would also, in terms of ecological validity, a dose of 600 grams a day probably not reflect, unless we're talking full carnivore, the way that regular meat eaters might actually consume. But from a mechanistic standpoint, it's given us something to go with one, is there some plausibility to the lack of association we see in East Asian cohorts at doses of under 60 grams a day or 100? Well, in some way, yes, because this study used a dose of 60 grams a day, didn't see that significant increase. And then we do see this increase with the high red meat intake that's unprocessed, driving a nitroso compound production endogenously to very high levels. But there's the artificiality of the dose that was used in this specific study.
- [24:14]
B
There's some really nice components that they've thought of here. Obviously, as you mentioned, the limitations being this wide exposure, which was done because we're trying to tease out what's going on here. But that aside, we have this really nice comparison of not only the low level of red meat to the high, but then taking into account this additional fiber, which is another classic line, we hear that once you have enough fiber in your diet, then you don't need to worry about any red meat consumption here. Even in those situations, we're seeing this almost looks like a threefold increase in some of these formation of these compounds. And then when we look at then the white meat or fish. You mentioned that there was no increase there in the. So straight away we're seeing that we can isolate this increase to red meat driving that and it's not offset by at least a supplementation of wheat bran in this case. Now you mentioned there that obviously this is comparing that 60 grams per day to 600 grams per day, as opposed to some intermediary numbers that might give us a bit more information. There was another trial that came after that. This was the Hughes and colleagues study that you mentioned that did indeed look at different ranges from 0, 60, 240 and 420. Can you maybe walk us through the important parts of this trial and maybe what this adds on top of what we've just discussed?
- [25:35]
A
Yeah. So this was also a metabolic ward, tightly controlled study where participants completed, in a kind of crossover randomized order design, four different 10 day dietary periods. In one, they consumed zero grams a day of meat, of red meat. In the other, in the second condition, 60 grams a day. So that same dose from the previous study and then 240 grams a day and then 420 grams a day. So we're into slightly more ecologically valid absolute thresholds of intake. And we still have a very high dose with the 420 gram a day dose. So this was provided primarily as roast beef and other meat preparations. And because it was a fully controlled study, again in a metabolic facility, protein intake, total fat and fiber levels in the diets were all matched between these diets, differing only in their dose of red meat intake. And the study was designed specifically to look at dose response relationships. Then across that range from 0 to 420, primary outcome was still N nitroso compounds in feces. And then they had similar secondary outcomes to the previous study. So for this, the generation of fecal and nitroso compounds from 0 to 60 grams a day, there was no significant increase. So at 60 grams a day, the difference from zero there was no difference. It was 53 micrograms per day in the 0 grams per day meat group. And in the 60 grams per day meat group it was 52 micrograms per day. So there was no difference whatsoever. When you then went to the 240 gram a day meat dose, that jumped up to 159 micrograms a day. And when you went to the 420 gram a day meat group, that faecal nitroso compound formation jumped up to 198 grams per day. So there was evidence of significant increased levels of N nitroso compounds. Once they went from the 60 to 140 to 60 to 240 gram a day dose and then again an increase from 240 up to 420. And that was obviously then evidence of a kind of dose dependent increase in nitroso compound formation, with those increases evident at a more moderate dose of 240 grams a day. Which again in terms of when categorical analyses in epidemiology model, particularly in North American cohorts and some of the European cohorts model their very high group, that is again often in this region of over certainly 180 grams a day, possibly into this kind of range, that's reflected more by that 240 gram dose. So at that moderate level you were still seeing quite a significant increase in nitroso compound formation. There was a lot of inter individual variation, which is potentially obviously a factor always in why some people see outcomes related to these mechanisms and potentially some don't. And for an interesting aspect of this study, they maintained a subgroup of four of the participants voluntarily obviously on the 420 gram a day red meat diet for 40 days. And in that faecal nitroso compound formation showed no sort of adaptation effect. The levels just remained high over time and did not decline at all. So there was no kind of adaptation. There was no. And this is outside of the kind of control of the metabolic world where they're just sustaining the meat intake and other aspects of diet have kind of lost their control in this. So again, prolonged intake of a high red meat diet is not going to. We hear this a lot, right? And pushback and sometimes in some of the kind of ancestral myths around diet is humans are adapted to that word. Adaptation is often invoked as humans are designed to eat this way. And this clearly showed that there was no source of adaptation against the production of these carcinogenic compounds. So I think this particular study is possibly more persuasive than the previous study in linking back to the epidemiology. What we have is an even more now plausible mechanistic support for stating that when we don't see risk at these doses of under 100 grams a day, in and around this average dose of 60 grams a day, that could very well be genuinely, because there is no risk, we didn't see any difference in nitroso compound formation at that dose of 60 compared to 0 compared to a no meat diet, a no red meat diet. So I think that lends to more of the plausibility for making that statement in relation to what we see in the epidemiology. And we now have a degree of biological plausibility to associations that are observed with high red meat intake in populations at more moderate red meat doses that could be reflective of more consistent intake. Knowing that the production of these compounds is effectively immediate and knowing that production is essentially sustained as long as the meat intake is consistent in that diet.
- [31:19]
B
There's a couple of things you've just said that I think are really important and worth restating and emphasizing for people. One is this connection to the epidemiology and helping us explain maybe what some of those observations are. Now, of course there's a number of candidate mechanisms that relate red meat to this colorectal cancer risk. So we can't say with certainty just based on what we've said. But indeed, if we are to accept this formation of N nitroso compounds as important, or maybe one of those driving factors, or if that is indeed the mechanism, then we can see based on this work that we've just mentioned, that up to that 60 grams per day level, we're not seeing any change in the production of those compounds. We're not seeing an increase which then would then fit in and at least make plausible sense as to why maybe we aren't detecting those effects at that level of intake in the epidemiology. So marries up nicely there, once we start pushing it beyond that to these ranges we used in this study, 240 or 420 grams per day, we are seeing this much elevated level of these compounds. And again that might map on to why we would be seeing more risk being detected in that epidemiology. And I think that's a really crucial point that helps us connect some of these dots. The other thing that was nice about the study that you outlined was this maintenance of that intake of the high intake over that 40 day period after the controlled feeding study. And again, this gives us again room to at least for the moment say we can call into question some of those claims people have of oh, this is just a transient response you're seeing over a couple of days. After that there's going to be an adaptation, there's nothing to worry about. Whereas we see here that's not the case. If you continue to eat these higher levels of red meat that was described here, you would continue to see that high level of and nitroso compound formation. So really nice useful work there. Based on what we've discussed so far, we're starting to see that with red meat we, we've been able to specifically Say it's red meat as opposed to, let's say white meat or fish. We can also say it's specifically red meat. Even in the context of this added brand that we mentioned earlier here, we're now starting to see that maybe there's a dose threshold to pass. The next step is then to say, well, if we are detecting this effect, what component of that red meat actually is it? What is the thing that's driving some of the formation of these N nitroso compounds? What aspect of red meat leads to that? And this is maybe where we can get into the cross study because this is really useful and perhaps one of the most important mechanistic trials people should be aware of in this area and helps us try and answer some of that question of what are the components of concern that might be driving these effects. Can you maybe speak to the setup of this cross study and again, what this adds to our understanding?
- [34:12]
A
Yeah, like you say, I think this is probably one of the most important of the studies that kind of came out from this group in this period from kind of 96 to the early. And this was a 2003 paper published in Cancer Research. And there were two protocols, again, randomized crossover dietary interventions in healthy male volunteers that were living in a metabolic ward, consistent with the kind of rigor and control that we've described for the previous studies. Each diet period in this study lasted 15 days with no, sorry, with a washout period, basically for the first 10 days leading into each study, arm each protocol, the first protocol had three diets over those three 15 day periods where you had a low red meat, again the 60 gram a day dose, high red meat, this 420 gram a day dose, and a vegetarian diet matched for protein content to the high meat diet. And then the second, and this is the real kind of genius, so to speak, of this study, is the three period crossover again with the 60 gram a day red meat diet alone or 60 gram a day diet supplemented with heme iron at around 8 milligrams per day, or the same diet supplemented with inorganic iron, which was 35 milligrams a day of a ferrous gluconate tablet. Again, nitroso compound formation was the primary driver or primary outcome. But what they were specifically interested in, particularly with this protocol too, was heme iron and the type of heme iron, a potential driver of that N nitroso compound formation. And what was brilliant about the design with the full control and the rigor that they had over the study was that between the vegetarian diet matched for protein content and the red meat high and low diets, they actually controlled for nitrate intake, so that actual exogenous nitrate intake through the diet was matched between those diets. So for the outcomes for the first protocol, if you Remember, compared the 60 grams and 420 grams a day of red meat with the vegetarian diet with the fecal N nitroso compound formation, the low red meat diet, the high red meat diet was just under 1300. As in 1300 micrograms. The way they expressed it was micrograms per kilogram of body weight. What that translated to was about 190 micrograms per day in an individual. In contrast, the low red meat diet was 42 micrograms per day. That's consistent with what we saw in their earlier work, where the N. Nitrosa compound formation at their 60 gram a day diet in these otherwise healthy participants was around 52 grams micrograms per day. And the vegetarian diet, then in terms of micrograms per day was around 63. So there was no significant difference between the low red meat 60 gram a day diet and the vegetarian diet, despite the fact that the vegetarian diet was protein matched to the high red meat diet. But the high red meat diet was a kind of fourfold increase in nitroso compound formation relative to both the low red meat diet and the vegetarian diet. And then when we then look to the second protocol for the addition of heme iron, you've got the low red meat 60 gram a day diet, which was similar in this range. This time it was 77 micrograms per day of N. Nitrosa compound formation on the low red meat diet. But when they added the 8 milligram a day heme iron supplement to that, that jumped up to 156 micrograms a day. When they added the inorganic ferrous iron supplement, there was no significant difference between the low red meat diet that had no iron added whatsoever. So the increase in N nitroso compound formation was driven by the addition of heme iron, even in the context of a low total red meat diet, whereas inorganic iron had no significant effect. And so this, I think, is where we build on what we've seen in some of those previous studies in that. Because sometimes you can see associations or suggestions that protein content is a kind of relevant factor in red meat and in unprocessed red meat in some of these associations. But it wasn't the protein content that was driving the nitroso compound formation, it was heme iron that was responsible for that increase associated with red meat consumption. And so what this indicated was that heme iron itself catalyzes the formation of potentially carcinogenic and nitroso compounds in the colon. And that provides, if we think back to the epidemiology and we can look to specific cohorts in epidemiology that have modeled, for example, or measured heme iron or nitrates or nitrates, and what you will see in several analyses, including analysis of, for example, the National Institutes of Health retired Person cohort in the US or the UK Biobank cohort, is in mediation analyses in epidemiology. Heme iron explains substantial proportion of the association between unprocessed meat and colorectal cancer risk. So we now have a plausible biological mechanism evidenced in a controlled intervention with a really elegant design that specifically isolated independent effects of the addition of heme iron as a supplement. That clearly showed that the increase in nitroso compound formation observed when you give people 420 grams a day of red meat was achieved when you give people 60 grams a day, but add 8 milligrams of heme iron to that dose. And I think this is an incredibly powerful kind of line of research, building on some of their previous work with the dose response relationships that ties together. Yes, you've got these endpoints of N nitroso compounds that are potentially mutagenic and carcinogenic, but now you actually have an explanation for why unprocessed red meat, which is low in preformed nitrates, as opposed to processed meat, which which people typically are happy with, kind of to conclude that there is that association with processed meat. And I think this study really gives us the causal chain that would link back to heme iron rather than the protein content or other component of unprocessed red meat.
- [41:17]
B
It's a really beautifully done study and well thought out study some of these components that help us isolate those. As you say, we have this initial comparison of the low red meat at 60 grams per day to the 420 to the vegetarian diet, Matra protein, which gives us not only that, we see this increased N nitroso compound formation with the high red meat compared to low. But even then, if we match the vegetarian diet for protein, we don't see an increase, thus allowing us to conclude, well, it's not the protein that's driving it, which was maybe one of those hypotheses we can then look at. Even in the low red meat intake, that 60 grams per day, if we give the addition of a heme iron supplement, we see these compounds Jump to the same types of levels we're seeing with the high red meat intake. Again, just really nice work that allows us to isolate down that what we're seeing is not a function of the protein load. It's also not a function of iron per se, because we're comparing that heme iron to the inorganic ferrous iron that you mentioned. So we can specifically say it's this heme iron that seems to be driving it. And we have this nice clear causal pathways. So some really nice work with that. Then we turn into another area and there's a collection of studies here, there's two or three at least, that you've highlighted that speak to this topic of genotoxicity. And essentially we're thinking about the colonic environment and where maybe that can become genotoxic. Can you maybe give us an overview of some of these studies, what they might be referring to, and again, what this collection of studies might add to our understanding?
- [42:56]
A
Yeah, so there's a couple of studies that I think we'll go through now for listeners. But before we do, I think it will be just helpful to describe the outcome that they use in these studies. And the outcome used in a couple of the papers that we'll discuss now is called faecal water genotoxicity. And what this is, it's a kind of ex vivo measure of DNA damage in a colon cancer cell line. So how this is a model that has been used for years in colorectal cancer research. And what it does is you feed your participants your dietary intervention or exposure. You then take fecal samples and you incubate the participants fecal samples with a specific colon cancer cell line. And the faecal water looks then when it's incubated with these particular cell lines, you are then looking at the level of DNA damage that the faecal water from your sample causes when it interacts with these colon cancer cell lines. And so it's a marker. And yes, it's not sticking something into something into a participant's colon to directly measure this as it's happening in vivo. But it's quite a kind of good substitute methodologically to try and assess DNA damage and genotoxicity on human colon cancer cell lines using this method. So when we're talking about faecal water genotoxicity, which sounds a bit strange, that's essentially what it is. So you're looking at the potential DNA damage in a colon cancer cell line caused by taking a sample of faecal water from a participant and incubating it with that particular cell line. And an early paper to do this that I think is quite helpful to look AT was a 1999 paper published by Rieger and colleagues in the journal Carcinogenesis. And it compared two diets that were consumed sequentially for 12 days. There was a one week washout period in between where they just consumed their normal diet. The first diet was 50% of total energy from fat was high in meat and it was also high in sugar, but low in vegetables and dietary fiber and free of wholemeal products, roughly getting at a kind of crappy western diet type of approach. And the second diet was high in vegetables and whole meal products, but lower in total fat, which was now just 19% of total energy and low in meat. And again, the primary outcome was, as I just described, this fecal water genotoxicity. And when they looked at DNA damage as this outcome after incubation with fecal water after the high meat and fat diet, that was 28%. And whereas the second diet that was high in fiber and vegetables was 17%. So the DNA damage from the fecal water incubated with these cancer cell lines following diet one was basically twice that of the second diet. And there was in terms of some of the other kind of outcomes, they looked at oxidative DNA damage as well as significant increase in oxidative DNA damage compared to control. And this, I think was an analysis then that basically you conclude that the diet high in meat and fat but low in fiber and vegetables increases the genotoxic potential of fecal water. That's reflected in increased DNA damage evident and oxidative DNA damage in these human colon cancer cells. And again, that could be something that contributes to an elevated risk of colon cancer with these kinds of diets. Now, the caveat of this study, as you probably already inferred, is this wasn't really conducted with the isolated control of some of the previous research. We were really looking at dietary pattern type of interventions where there's multiple component parts and that are not matched between those diets. So we're not really left with, as I said describing the outsets, diet 1 was high in meat and fat, but it was also high in sugar. So we're left with a lack of ability to more specifically isolate the independent effects of red meat in this particular analysis.
- [47:36]
B
So we have this support for the broader idea that diet itself and certain types of dietary changes can alter this genotoxic potential for the in the colon. But given the limitation of this study, we can't isolate this effect to red meat because we have all these various different things changing within the diet from fiber to fat content to the amount of vegetables they're consuming. But it does show us this potential and one of the things that maybe other work can allude to that we might see these changes within the colon relative to these types of differences between diet.
- [48:12]
A
Yeah, so there's, I think there's two papers that we could look at to, to look at some of the fecal water genotoxicity outcomes. There's a Dino and colleagues paper which was actually more recent, was published in December of 2025 and there's a Hebbels and colleagues paper which was published a decade earlier, which was 2012 and just over a decade. The Hebbles colleagues paper basically had individuals who had a baseline average red meat intake of 88 grams on average per day. So again, not dissimilar to some of the kind of low intake, low were intakes that we've seen in some of the Amanda Jane Cross and Bingham and colleagues papers. They then went on to a 7 day high red meat diet with around 300 grams a day consisting only of beef products that were pan fried. In terms of their, the cooking method and their specific outcome of interest was this faecal water genotoxicity and also nitroso compound formation with the fecal water genotoxicity that increased significantly with the addition obviously of this 300 gram a day meat diet compared to the, compared to the kind of baseline lower red meat intake, it didn't show an increase interestingly in faecal nitroso compounds formation following the 300 gram a day red meat intake. And so this particular study is the first that you don't see that increase in faecal nitroso compound formation at a dose of 300 grams a day, which is actually somewhat intermediate relative to some of the doses used in the previous research. But the previous research I think was better controlled. So I would still run with the conclusions we've seen with some of that kind of metabolic ward studies. But this basically added to that previous study by looking at this kind of faecal water genotoxicity outcome and seeing a significant increase in that in terms of a seven day sustained diet that was high in the, that was high in red meat. Now I think the problem with this study potentially is that you also had participants with moderate grade inflammatory bowel disease. So I think that somewhat limits in terms of obviously making inferences based on intestinal health. Six of them had diagnosed inflammatory bowel disease and another six also had Irritable bowel syndrome. And while that is something that we wouldn't expect an underlying pathology in terms of gastrointestinal function, I still think the kind of presence of half the subjects with IBD probably limits some of the kind of inferences that we can make, particularly in relation to the nitroso compound formation, whatever may be going on in that regard, or overlaps with inflammation. And in the previous studies we were looking at otherwise healthy individuals. So with that, I think the more recent paper, the Dinu and colleagues paper is actually probably more instructive, at least in my opinion, because it was a really nice design that thought through some of the potential moderating or modifying factors that you can see in the literature. So they had three exposure diets, intervention diets, it was a crossover design or, sorry, a parallel group design where they were either consuming a 900 gram per week red meat intake, high meat intake, and that was a mix of unprocessed red meat, processed meat, and they also had one 150 gram serving of poultry. So I think important to bear in mind that we're looking at weekly doses of intake. Now in this study rather than high daily doses. Then they had the same diet. So three, sorry, what they were asked for 150 gram servings of unprocessed red meat, three 50 gram servings of processed meat and one 150 gram serving of poultry. That is same high red meat diet but supplemented with 100 milligrams a day of alpha tocopherol. And then there was a Pesco vegetarian diet where participants excluded all meat and poultry and included three 150 gram servings of fish per week. They were prescribed a similar macronutrient intake. And this was also looking at this outcome of faecal water genotoxicity as its primary outcome. And for fecal water genotoxicity, DNA damage increased on the meat based on the high meat based diet by 12% compared to the Pesco vegetarian diet and interestingly by a similar amount compared to the meat based diet plus alpha tocopherol. There was no significant difference in in the generation of DNA damage measured through this faecal water genotoxicity assay between the meat based diet with added alpha tocopherol and the Pesco vegetarian diet. And when you looked at the within group difference, it was the high meat diet that basically showed this significant within group increase from their before diet by 16% increase in DNA damage compared to baseline. Whereas there was no significant difference in either the meat based diet, same diet plus alpha tocopherol or the pesca vegetarian diet. And there were some other interesting changes in terms of markers of lipid peroxidation. Although some of the more robust measures of lipid peroxidation that you can look at, there were no significant differences between groups. And this trial for me is a really good example of the potential for moderating factors. In the previous study we discussed, there was the addition of this kind of high fiber wheat bran to a high meat condition, and that didn't abolish the generation of N nitroso compound formation at a dose of 600 grams a day of red meat intake. This is a weekly dose and I think we have to bear that in mind distributed across the week. But it's still very interesting that the addition of the primary fat soluble antioxidant, which is alpha tocopherol to the meat based diet, effectively abolished the increase in fecal water genotoxicity observed on the high meat based diet. So that I think speaks to some of the complexity that we can sometimes see in some of the epidemiology and the potential for moderating factors to influence associations. And some of these are obviously not necessarily always captured in prospective cohort studies. And it does suggest at least some degree of moderation by fat soluble alpha tocopherol in at least this specific outcome. What it would say for other potential mediating effects and what it might mean if the study focused exclusively on unprocessed red meat, whereas this actually included white meat as well, although much less of a dose of that. All of those are open questions the study doesn't particularly answer, but it is consistent with suggesting overall that the diets obviously lower in unprocessed red meat have a lower colorectal cancer risk profile. But there are other factors within a diet that may moderate the relationship, at least with some of these kind of intermediate outcomes.
- [55:34]
B
And so when we're looking at some of these, this mechanistic work to try and piece together, is there this causal effect? One of the other studies maybe worth touching on before we start pulling all of this together was 2021 paper Gurjao and colleagues. And this was interesting because this specifically looked at human colorectal cancer tumors and found this association with the higher red meat intake to these, this kind of mutational signature on these tumors. So essentially these signatures that they're able to trace back to see what might have been involved here in a causal way and connecting it to some of what we're seeing with these components of red meat, is there any, anything within this particular Study that's maybe worth highlighting that might be of interest that could
- [56:23]
A
add something else, I think, for this with the. So this was an analysis of the us, the kind of two long standing US cohorts, the Nurses Health Study and the Health Professionals Follow Up Study. And it was specifically looking at, like you said, this alkylating mutational signature in colorectal tumors. And this analysis, I think is probably. So they looked at. They identified seven different mutational signatures that could be identified in the colorectal cancer tumours, one of which was this particular alkylating signature that they highlighted was undescribed previously in colorectal cancer. It was enriched in tumours that promoted methylation as a kind of process by which cancer cell proliferation can occur. And this was associated with a high intake not just of processed red meat, but also unprocessed red meat and overall red meat. And no association was found for either white meat and chicken or fish. So, yeah, I think this is an interesting additional analysis from these cohorts that gives us again some degree of linkage between some of what we saw in the kind of short term mechanistic studies and some of what we might see in the longer term prospect of studies. These are studies of hundreds of thousands of men and women in the us. It is very limited demographically and does not represent the wider US population. So I think that has to be said as well. But in terms of these signature, kind of these kind of metabolomics essentially based alkalizing signatures associated with unprocessed meat, but not with chicken or fish, I think that's relatively consistent with what we've seen with the outcomes of nitroso compound formation generated by heme iron in the controlled interventions. So it provides, I think, another link in the evidential picture of there are mechanistic carcinogenic processes in the colonial that are influenced by high. And these are cohorts that do have high intakes of unprocessed red meat. And I think that's important then when we're thinking about the dose response, I mean, typically in the NHS or the HPFs, the high quintile is often in this range of over 180 grams a day. So, yeah, I think it's a nice analysis that kind of adds to the picture of causal pathways that could explain a relationship between this exposure of unprocessed red meat and colorectal cancer as an outcome.
- [59:10]
B
Yeah, it's another one that kind of fits in and aligns with some of that. And again, still with open questions, I think Here we do have within that red meat consumption. It was unprocessed and processed here, but we still see that kind of 10 those the 10% of highest red meat intake had the highest amount of that alkaline signature. And that intake ends up corresponding to those types of figures we mentioned. I think about 150 grams or above. So, so fits in there, but just maybe adds another little piece. So I suppose to start pulling this together and to get to what people really want to hear from you on a topic that as we've noted is got some lots of interesting questions that maybe still we can say certain things with confidence, but maybe when we come to clear conclusions, that's maybe another thing. So if we were to pull this together from everything we've said, Alan, I suppose what people want to get to is first from purely talking about the evidence we have in this area, what are some things you think we can conclude with some degree of confidence that are fair conclusions to say about the current state of evidence related specifically to unprocessed red meat and this outcome of colorectal cancer risk? What would you have? Yeah, a fair degree of confidence in saying this is what we can state based on the current evidence base?
- [60:28]
A
Yeah, I think we've got based on evidence from the controlled feeding studies that we were discussing. If we're trying to understand is the relationship between unprocessed red meat as an exposure and colorectal cancer as an outcome that we observe in epidemiology, is that a true effect or is that something that people can just dismiss as well? That's just correlations, not causation and all that kind of jazz. I think we have for unprocessed red meat. We have clear evidence of an increase in carcinogenic compound production. We've clear evidence that is primarily mediated by heme iron and we have clear evidence of a dose response relationship that this effect is primarily observed at doses of at least based on the evidence we can say over 240 grams a day. Now is there likely to be similar effects at something like 200 grams a day? Arguably, when we do look at the total body of these controlled feeding studies, those increases in carcinogenic compound production are relatively consistent at these higher doses of unprocessed red meat intake. That is generally consistent with the type of threshold effects that we tend to see in the epidemiology. So I do think that it is fair to say that there is a degree of biological plausibility to the effect of a high unprocessed meat diet. Sustained over time in terms of long term colorectal cancer risk. But I also think that there's an equal plausibility based on these same studies for the lack of association that we see with intakes of less than 100 grams a day. And I think that is not simply an absence of being able to detect an effect which we could potentially be left with sometimes wondering in the epidemiology because of the fact that so many of these trials had control or comparative exposures of 60 grams a day, where we didn't really see any sort of difference in the production of these same carcinogenic compounds that we see at higher doses. And I think when you have a dietary pattern with red meat in that average range, that also typically would be concomitantly higher in either other white meats or fish, or indeed higher in fiber rich whole grain foods, legumes and vegetables and fruits typical of dietary patterns that are lower in red meat. I find it very difficult to conclude on that basis that unprocessed red meat is still quote, unquote, causal in that kind of context. And so I think that what is offered by the epidemiology that there is some sort of dose threshold in this relationship between unprocessed meat, I believe is supported by the evidence that we have from these controlled feeding studies.
- [63:20]
B
And again, there's a lot to discuss, maybe for another time. I know we have a plan to revisit this topic of substitutions comparisons. We have a previous episode that I encourage people to go back and listen to. I was a 300 and something called Compared to what? We'll be revisiting that topic in a upcoming episode. But there's also. That is quite pertinent to some of the things that you just discussed, given that, as people have noted, not only through the studies we've discussed today, but also the epidemiology we see in the trials, maybe different comparators that are being compared to red meat in each of these situations. And then with epidemiology, depending on what type of substitution analysis we're doing, we can have it being compared to not only different things, but typically the background diet that is filling in the gaps. And there's some nuance for us to discuss, but again, that's maybe for another point. So given that those conclusions that you just put in place to maybe put you on the spot even further as we finish here, there's people that maybe are thinking, okay, going beyond that as an individual, that they might want to make some decisions based on this evidence and thinking about the question of to what level of intake should I feel safe consuming, let's say, unprocessed red meat, given that there is this zone where maybe we're not seeing an effect, pause would be that we're not. There is no an effect at these certain levels that there's some evidence that, as you said, could support that. But given that we're seeing that for depending on what we're comparing it to and what's maybe replacing that red meat or not in the diet, there could be differences here. What would be an intake that someone should say, okay, if I am going to reduce my red meat intake because currently it's now high, where should that go? And would I foreseeably continue to see a benefit if I continue to reduce that down to, let's say, ultimately zero, depending on what then replaces that? How do they work through that at an individual level given the current gaps we might have in the evidence?
- [65:24]
A
Yeah, I think given those gaps, the only thing that you could really conclude is that under 100 grams a day is a desirable target for someone that currently consumes a high red meat diet. I think it's pretty clear that the substitution that would be desirable for that reduction would be primarily if the diet, if they are eating a high red meat diet and it's low concomitantly in other plant protein sources like legumes and stuff, that that would be the primary substitution that we want to make. And secondarily, if they also eat an omnivorous diet, then in terms of getting that red meat consumption down to those intake levels, then obviously then white meat or fish probably would be preferable even to white meat, I think, for various reasons. And that would probably be the kind of order of hierarchy by which they would prioritize changing their diet. But I don't think that we can, based on the evidence and the gaps that do exist, fill in the gaps with anything more than the kind of dose threshold that we've identified. And I think that really the only way that we could say that, and obviously this then depends on individual risk, but I think the only reason really threshold level or range that we could say with any degree of confidence that this is somewhere where if you were determined, if someone was saying, I'm not going to cut meat out of my diet, then I think that compromise would be, well, then this is an intake range where you're at least at lower risk, particularly if substitutions are made with legumes and whole grains and the kind of plant component of diet, and make sure that vitamin E is in there somewhere.
- [67:10]
B
So that's really useful. We have a situation that if someone currently consumes high intakes of unprocessed red meat, maybe because their position has been, well, it's unprocessed, it's not necessarily negative, or the rest of my diet is. Okay, the implication is here, well, if you're consuming a high level, it's probably worth limiting, eliminating that where we can have confidence, where you would see that risk reduction is getting it to below that, say 100 gram per day level. If someone is below that level, we can confidently say there's a lower risk there compared to higher intakes. And then where it becomes a bit more tricky then is saying going beyond that point. If someone did continue to reduce that further, would there be continued risk reduction and how far that would extend? Maybe one then thing that could fit in with that is if someone did want to, let's say theoretically reduce all possible risks that red meat unprocessed metal rate might contribute, of course then the option is to reduce that to zero if they wish. But again, that is different from concluding that maybe based on current evidence that we can say that would be the only safe level for someone without reducing risk. We're a bit away from that. But again, at an individual level, someone could do that and say, I want to reduce the possibility of it contributing to that to absolutely zero. And so there, that is maybe as much as we can conclude given the these kind of numbers we're seeing.
- [68:38]
A
Yeah, yeah, I think so. I think so.
- [68:41]
B
Excellent. I think this has been a really useful episode and hopefully a lot of you listening have enjoyed it. Alan and I will be back with another episode coming soon. We have some what we think are really interesting topics that we're going to be discussing in the not too near future. So hopefully you join us for that. For those of you who are Sigma Nutrition premium subscribers, you can get a set of detailed study notes that goes through everything that we have discussed in today's episode, as well as an edited transcript that will be linked up in the description box. And if you're listening on the public feed of the podcast, then click into the description box for more information about premium subscriptions or just anything related to today's episode. So that is that Alan and I will be back another episode soon. Please join us for that and in the meantime, have a great week. Stay safe and take care.
- [69:34]
A
Sam.