Quick read is great if you are debunking something. IF you are trying to Teach us something more detail might be needed so we understand the issues instead of just trusting your opinions (which I normally do trust)
I appreciate your highlighting important aspects of health and nutrition while needing to be diplomatic about certain topics in order to maintain credibility across the board.
I enjoy the timeliness that helps zero in the flaws/nuances while I’m being flooded with rehashes and clickbait. Still, the in depth is also good as allows a deeper understanding of issues that are not black and white. Cheers
Hello Nick, I've written a response to your debunk. This response includes some suggestions for improving your criticisms.
Response to claim 1:
Nick’s focus on how investigators define Unprocessed Red Meat suggests he believes it’s a mistake to group mixed dishes, like sandwiches
with standalone items like steak. However, it’s unclear why this is necessarily an error. If Nick’s concern is about accurately capturing red meat consumption, he can refer to reference 22 (https://pubmed.ncbi.nlm.nih.gov/37552965/). If he’s worried about confounding factors, such as the bun in a burger, Nick should propose the causal theory for these being confounders. To help with this, he might consider studies on populations outside the U.S., where pairing red meat with bread is less common. If Nick insists on the confounding claim, a sensitivity analysis would be necessary to assess its impact. Unless the confounding effect is substantial compared to the treatment effect, the investigators can reasonably dismiss Nick’s “debunk.”
Claim 2:
Now, it’s not clear to me if Nick is using healy user bias in the same way I understand it to mean. Nevertheless, I think my response will probably be the same. Let’s suppose there was indeed selection bias present despite the investigators best efforts at dealing with it. It is up to Nick to tell us how large the magnitude of the bias is. Because if the magnitude (be it positive or negative) is not larger than the effect of treatment, this can safely be ignored.
Claim 3 and 4:
Sadly to me, I couldn't parse them enough to really provide a response. It's not super clear how this would be part of the "dubunk".
My preference would probably be for an "executive summary" of your thoughts on these types of things and then a bit more of a deep dive, ideally with something like an outline or perhaps sections with a bold subject line where we could skim to portions that most interest us to drill down into the details if we are so inclined. But I realize that your time is very valuable to you. On the other hand, the extra time you spend on making your valuable information available to others is multiplied exponentially because of the many readers/listeners... and those to whom they communicate with.
Thank you so very much for what you do. I think your style and what you are doing is cutting edge and should (hopefully) lead to better, more fruitful and more accurate ways of studying, interpreting and disseminating the data to help people live healthier, happier lives.
"My preference would probably be for an "executive summary" of your thoughts on these types of things and then a bit more of a deep dive" -- your wish is granted! Did you see my 8 carnivore myths newsletter from the day after this newsletter was posted?
It is the type of scientific inquiry that is so sorely lacking right now & I really appreciate it. I was amazed to see how crazy our medical reporting/scientific community(?) had become probably a decade or more ago. I was reading the articles in the MSM regarding health and medicine & thought I was informed... until I started gaining weight in my 40s. I had never had that problem before, but following conventional wisdom I cut out as much fat from my diet as I could, but that didn't help. Then I went almost vegan for a bit. Eating bowls of mashed potatoes, rice, bread, pasta, even candy. I had been told it was all calories in, calories out & carbs had less than half the calories per gram as meat. Even sugar was ok per some articles, but their drawback was they were "empty calories"... no vitamins. So I ate candy for snacks & took vitamins. I gained even more weight and rapidly. Luckily, I saw an article about a book signing by Gary Taubes. His book was "Why get fat and what to do about it." I got the book from the library and was astounded that this top science writer was saying to go to a high fat, low carb diet. It was the OPPOSITE of what "all of the medical experts" were saying. But I was at the low end of obese by this time & what I was doing wasn't working, and he had a lot of science and reason behind his recommendations so I reluctantly tried it, with some trepidation.
Again I was astounded to see the pounds melting away. I was eating to contentment but was consistently losing ~ .2 lbs per day.
I was still worried about the health consequences, it went against the recommendations of the American Heart Association, the NIH, the National Cancer Institution and virtually every other high profile medical organization. But the more that I've read since then makes me really question their systems. You have exposed a few egregious errors even recently. Today I just watched your video ( https://youtu.be/RlksFU_lx_E ) about red dye... or really the basic logic that absence of evidence is evidence of absence. It is a logical fallacy that I've seen multiple times in the reporting on medical issues and even assertions or denials by medical experts and organizations that should certainly know better.
Thank you again from the bottom of my heart for the important public service that you are doing!
Brand new subscriber, so maybe shouldn't weigh in, but... I like this post yet would've appreciated something longer, because I didn't see a food lumping issue (beef in lasagna is still "unprocessed" meat), and the researchers did control for differences in smoking etc. I've written about other flaws in this study related to measurement and coding but still find the results credible, even if it's unclear exactly how much processed red meat intake would increase risk of dementia.
They didn't control for it. The made 'statistical adjustments' to observed healthy user bias elements that (i) shouldn't assume to be complete (ii) don't capture the full scope of healthy user bias elements. If you agree with their general conclusions, I'd direct attention to the fact that the biological plausibility statements are mostly not specific to processed meat. So why wasn't an effect observed for unprocessed meat?
Incorrect. Statistical adjustments Nick, does mean they controlled for x in their model. Your comment about biological plausibility doesn't make much sense to me here so perhaps you can explain a bit more of what you left our in your response. Another correction, a non-zero point estimate was provided. It just wasn't statistically significant. There could be a number of reasons why they failed to reject, failed model assumptions could be one.
I agree they didn't capture the full scope of healthy user bias variables (maybe no study looking at a broad health outcome like dementia could?). I was more concerned about the absence of dose-response analyses. Feels like a researcher degrees of freedom problem that they trichotomized intake into low, medium, and high but gave a squishy explanation of how the cuts were made.
I appreciate here and in your post the attention to biological plausibility, given that researchers so often prioritize significant effects over plausibility. But the measure of intake is so crude - self-report summaries of consumption over the past 2-4 years - I'd be surprised to see perfect alignment with any plausibility-based predictions. (Also, maybe give them credit for not HARKing?)
So we are in agreement that the claim red meat is bad for the brain that's circulating in the media in not truly supported by the "squishy" data, and that the biological plausibility argument are out of sync with the findings?
I'd rather say the data suggests that processed red meat is bad for the brain, but the measurement and analytic approach reveal nothing about dose-response relationships or how much people have to eat before they start experiencing greater risk, and these shortcomings make it challenging to find signals in the noise that are consistently biologically plausible.
I'm struck by the fact that their high-consumption group has no upper limit with respect to intake. Perhaps the effects are driven by a relatively small number of folks whose red-meat consumption is extreme. No way of knowing.
I have to acknowledge that what you wrote earlier about healthy user bias weakens my argument. I thought the authors did a good job in the choice of covariate to adjust for, but we can't measure everything.....
I agree that they didn't capture the full scope of healthy user bias variables (though maybe no study looking at a broad health outcome like dementia could?). I was more concerned about the absence of dose-response analyses. Feels like a researcher degrees of freedom problem that they trichotomized intake into low, medium, and high but gave a surprisingly squishy explanation of how the cuts were made.
I appreciate here and in your post the attention to biological plausibility, given that researchers so often prioritize significant effects over plausibility. But the measure of intake is so crude - self-report summaries of consumption over the past 2-4 hear - I'd be surprised to see perfect alignment with any plausibility-based predictions. (Also, maybe give them credit for not HARKing?)
Thanks for your thoughtful responses. I'll clarify I think it's plausible that processed red meat has a negative effect on the brain. However, I think a lot of things are 'plausible." It's seems like we're both open to the possibility, but also feel these data do not make a strong claim to that effect and - furthermore & more importantly - make no claim as to whether unprocessed red meat is bad for the brain. In fact, they perhaps argue in favor of the null hypothesis. As for credit for nor HARKing, my intention is not to evaluate the intentions of the authors but the value (or lack thereof) of the data. Fair?
Quick read was great but I also like the longer reads too!!! Not helpful I know!
I was hoping to hear this :)
I agree with Penny.
Concur! Happy to read Nick's long deep dives, but equally content with short and punchy.
Thanks Amy!
All your content is greatly appreciated!
Thanks!
If the information is correct (it is nearly 100% of the time) then both formats are good ... obviously.
Seems to be the consensus. Love it!
Quick read is great if you are debunking something. IF you are trying to Teach us something more detail might be needed so we understand the issues instead of just trusting your opinions (which I normally do trust)
Good response!
I like the snippets just as much as the long form. Bite sized for those times when I don’t have time to digest your full serving of science.
I'll dish up the shrimp cocktail and the 16 oz ribeye then ;)
Mmmm. Reminds me: shrimp and steak are on the menu for dinner today!! 😀
Nice!
Anything that you post is useful to someone Nic
I prefer shorter posts but if it takes a few more words then do it 👍⛓️
Thans Ernie :).
All your content is awesome, short or long.
Awe! Thanks!
On point as usual. I expect no less from you.
I appreciate your highlighting important aspects of health and nutrition while needing to be diplomatic about certain topics in order to maintain credibility across the board.
Thank you Alessandro. You get it!
Thanks Nick! Shorter is better for me. (Time crunch)
Also I like to share your articles on Bluesky, and shorter more likely to be picked up and reposted.
BTW, I'd luv to see you and other great docs I used to follow on X on Bluesky #medsky .
I appreciate what you do!
Cheers, Barry Davidson
Thanks Barry. I'll consider it depending on whether Bluesky starts to take off. Lots of socials to manage at the moment and I don't have my own team.
Many thanks for doing everything that you do. Appreciate it.
Thank you Raghu :)
I join the (so far) consensus view - both formats are useful, but (a), it depends on the topic, and (b), neither option should dominate.
Does indeed seem to be the consensus. Registered.
I enjoy the timeliness that helps zero in the flaws/nuances while I’m being flooded with rehashes and clickbait. Still, the in depth is also good as allows a deeper understanding of issues that are not black and white. Cheers
Thanks George. Consensus seems to be that a mix would nice nice. Love to hear it.
Hello Nick, I've written a response to your debunk. This response includes some suggestions for improving your criticisms.
Response to claim 1:
Nick’s focus on how investigators define Unprocessed Red Meat suggests he believes it’s a mistake to group mixed dishes, like sandwiches
with standalone items like steak. However, it’s unclear why this is necessarily an error. If Nick’s concern is about accurately capturing red meat consumption, he can refer to reference 22 (https://pubmed.ncbi.nlm.nih.gov/37552965/). If he’s worried about confounding factors, such as the bun in a burger, Nick should propose the causal theory for these being confounders. To help with this, he might consider studies on populations outside the U.S., where pairing red meat with bread is less common. If Nick insists on the confounding claim, a sensitivity analysis would be necessary to assess its impact. Unless the confounding effect is substantial compared to the treatment effect, the investigators can reasonably dismiss Nick’s “debunk.”
Claim 2:
Now, it’s not clear to me if Nick is using healy user bias in the same way I understand it to mean. Nevertheless, I think my response will probably be the same. Let’s suppose there was indeed selection bias present despite the investigators best efforts at dealing with it. It is up to Nick to tell us how large the magnitude of the bias is. Because if the magnitude (be it positive or negative) is not larger than the effect of treatment, this can safely be ignored.
Claim 3 and 4:
Sadly to me, I couldn't parse them enough to really provide a response. It's not super clear how this would be part of the "dubunk".
Yes, I like this style.
My preference would probably be for an "executive summary" of your thoughts on these types of things and then a bit more of a deep dive, ideally with something like an outline or perhaps sections with a bold subject line where we could skim to portions that most interest us to drill down into the details if we are so inclined. But I realize that your time is very valuable to you. On the other hand, the extra time you spend on making your valuable information available to others is multiplied exponentially because of the many readers/listeners... and those to whom they communicate with.
Thank you so very much for what you do. I think your style and what you are doing is cutting edge and should (hopefully) lead to better, more fruitful and more accurate ways of studying, interpreting and disseminating the data to help people live healthier, happier lives.
"My preference would probably be for an "executive summary" of your thoughts on these types of things and then a bit more of a deep dive" -- your wish is granted! Did you see my 8 carnivore myths newsletter from the day after this newsletter was posted?
No, I didn't, but I just found it and loved it!
It is the type of scientific inquiry that is so sorely lacking right now & I really appreciate it. I was amazed to see how crazy our medical reporting/scientific community(?) had become probably a decade or more ago. I was reading the articles in the MSM regarding health and medicine & thought I was informed... until I started gaining weight in my 40s. I had never had that problem before, but following conventional wisdom I cut out as much fat from my diet as I could, but that didn't help. Then I went almost vegan for a bit. Eating bowls of mashed potatoes, rice, bread, pasta, even candy. I had been told it was all calories in, calories out & carbs had less than half the calories per gram as meat. Even sugar was ok per some articles, but their drawback was they were "empty calories"... no vitamins. So I ate candy for snacks & took vitamins. I gained even more weight and rapidly. Luckily, I saw an article about a book signing by Gary Taubes. His book was "Why get fat and what to do about it." I got the book from the library and was astounded that this top science writer was saying to go to a high fat, low carb diet. It was the OPPOSITE of what "all of the medical experts" were saying. But I was at the low end of obese by this time & what I was doing wasn't working, and he had a lot of science and reason behind his recommendations so I reluctantly tried it, with some trepidation.
Again I was astounded to see the pounds melting away. I was eating to contentment but was consistently losing ~ .2 lbs per day.
I was still worried about the health consequences, it went against the recommendations of the American Heart Association, the NIH, the National Cancer Institution and virtually every other high profile medical organization. But the more that I've read since then makes me really question their systems. You have exposed a few egregious errors even recently. Today I just watched your video ( https://youtu.be/RlksFU_lx_E ) about red dye... or really the basic logic that absence of evidence is evidence of absence. It is a logical fallacy that I've seen multiple times in the reporting on medical issues and even assertions or denials by medical experts and organizations that should certainly know better.
Thank you again from the bottom of my heart for the important public service that you are doing!
Thanks for the comprehensive comment and words of appreciation. Have a lovely evening Jeff :)
Brand new subscriber, so maybe shouldn't weigh in, but... I like this post yet would've appreciated something longer, because I didn't see a food lumping issue (beef in lasagna is still "unprocessed" meat), and the researchers did control for differences in smoking etc. I've written about other flaws in this study related to measurement and coding but still find the results credible, even if it's unclear exactly how much processed red meat intake would increase risk of dementia.
They didn't control for it. The made 'statistical adjustments' to observed healthy user bias elements that (i) shouldn't assume to be complete (ii) don't capture the full scope of healthy user bias elements. If you agree with their general conclusions, I'd direct attention to the fact that the biological plausibility statements are mostly not specific to processed meat. So why wasn't an effect observed for unprocessed meat?
Incorrect. Statistical adjustments Nick, does mean they controlled for x in their model. Your comment about biological plausibility doesn't make much sense to me here so perhaps you can explain a bit more of what you left our in your response. Another correction, a non-zero point estimate was provided. It just wasn't statistically significant. There could be a number of reasons why they failed to reject, failed model assumptions could be one.
I agree they didn't capture the full scope of healthy user bias variables (maybe no study looking at a broad health outcome like dementia could?). I was more concerned about the absence of dose-response analyses. Feels like a researcher degrees of freedom problem that they trichotomized intake into low, medium, and high but gave a squishy explanation of how the cuts were made.
I appreciate here and in your post the attention to biological plausibility, given that researchers so often prioritize significant effects over plausibility. But the measure of intake is so crude - self-report summaries of consumption over the past 2-4 years - I'd be surprised to see perfect alignment with any plausibility-based predictions. (Also, maybe give them credit for not HARKing?)
So we are in agreement that the claim red meat is bad for the brain that's circulating in the media in not truly supported by the "squishy" data, and that the biological plausibility argument are out of sync with the findings?
I'd rather say the data suggests that processed red meat is bad for the brain, but the measurement and analytic approach reveal nothing about dose-response relationships or how much people have to eat before they start experiencing greater risk, and these shortcomings make it challenging to find signals in the noise that are consistently biologically plausible.
I'm struck by the fact that their high-consumption group has no upper limit with respect to intake. Perhaps the effects are driven by a relatively small number of folks whose red-meat consumption is extreme. No way of knowing.
I have to acknowledge that what you wrote earlier about healthy user bias weakens my argument. I thought the authors did a good job in the choice of covariate to adjust for, but we can't measure everything.....
I agree that they didn't capture the full scope of healthy user bias variables (though maybe no study looking at a broad health outcome like dementia could?). I was more concerned about the absence of dose-response analyses. Feels like a researcher degrees of freedom problem that they trichotomized intake into low, medium, and high but gave a surprisingly squishy explanation of how the cuts were made.
I appreciate here and in your post the attention to biological plausibility, given that researchers so often prioritize significant effects over plausibility. But the measure of intake is so crude - self-report summaries of consumption over the past 2-4 hear - I'd be surprised to see perfect alignment with any plausibility-based predictions. (Also, maybe give them credit for not HARKing?)
Thanks for your thoughtful responses. I'll clarify I think it's plausible that processed red meat has a negative effect on the brain. However, I think a lot of things are 'plausible." It's seems like we're both open to the possibility, but also feel these data do not make a strong claim to that effect and - furthermore & more importantly - make no claim as to whether unprocessed red meat is bad for the brain. In fact, they perhaps argue in favor of the null hypothesis. As for credit for nor HARKing, my intention is not to evaluate the intentions of the authors but the value (or lack thereof) of the data. Fair?
Fair. I think that nicely summarizes our common ground.
Quick summaries are good… so much “academic” research is structured incomprehensiblity…