Msleading title: the study didn't look at switching, just at correlations in the population.
this post was submitted on 29 Aug 2025
88 points (87.9% liked)
Public Health
1029 readers
294 users here now
For issues concerning:
- Public Health
- Global Health
- Health Systems & Policy
- Environmental Health
- Epidemiology
- etc.
🩺 This community has a broader scope so please feel free to discuss. When it may not be clear, leave a comment talking about why something is important.
Related Communities
- Medical Community Hub
- Medicine
- Medicine Canada
- Premed
- Premed Canada
- Public Health (📍)
See the pinned post in the Medical Community Hub for links and descriptions. link (!medicine@lemmy.world)
Rules
Given the inherent intersection that these topics have with politics, we encourage thoughtful discussions while also adhering to the mander.xyz instance guidelines.
Try to focus on the scientific aspects and refrain from making overly partisan or inflammatory content
Our aim is to foster a respectful environment where we can delve into the scientific foundations of these topics. Thank you!
founded 2 years ago
MODERATORS
And being vegan/vegetarian also correlates with wealth. And wealth correlates with lower rates of disease.
So yeah unless we get longitudinal studies I wouldn’t read much into these population wide correlations.
unless we get longitudinal studies
You can see my comment below on interpreting primary medical literature, so I'm not going to argue for the merits of specific claims about health based on this study. What I will address is what makes you think this study isn't longitudinal when it's called "Longitudinal associations between vegetarian dietary habits and site-specific cancers in the Adventist Health Study-2 North American cohort"? I'm also perplexed what you think "longitudinal study" means as it pertains to wealth.
There are dozens of meta-analyses broadly confirming a reduction in chronic disease with vegetarian and vegan diets if that's what you're interested in. There's so much that this particular study barely makes a dent except for the novel association between PBD and a reduced risk of lymphoma. (This still needs to be shown in meta-analyses, so it doesn't matter much.)
If the meta-analysis is based on studies with flawed methodology and/or assumptions, then the meta-analysis itself will be of little value. Is that not the case?
When a systematic review and meta-analysis is published (a meta-analysis is usually a component of the systematic review), methodologies are taken into account. It isn't just mindlessly throwing studies into a hydraulic press and fusing them together. For example, I can show you a 2017 systematic review (not a meta-analysis, but again, these usually come in one package) exploring possible links between managed bees and: 1) competition with native bees, 2) effect on fauna populations, and 3) pathogen transmission.
It's open-access if you want to read it, and over and over again the authors grill the methodology of the studies they're reviewing. They basically say "hey, the studies say this, but their collective methodologies are too weak, and so further, more robust studies need to be done." If they were just going by the studies ignoring methodology, they probably would've concluded managed bees are harmful to the ecosystem. But they couldn't because scientists care deeply about methodology. That's a major part of synthesizing scientific literature.
I've expressed some caveats to just citing the figure from this study before in another thread (hi, Sunshine lol), but overall yes, even without this study, there's robust evidence for double-digit reduced risk of CHD and overall cancer in plant-based diets compared to omnivorous ones. A recent meta-analysis for example shows a 15% reduction to overall cancer incidence risk, and even a substantial reduction in a vegetarian diet from an omnivorous one.
And to be clear, the evidence from this study is good. This isn't "study bad"; it's just primary medical literature that needs to later be interpreted by experts through things like meta-analyses.
Edit: Also, please don't take jet seriously below. All they do here is spread pseudoscientific nonsense about a diet with proven major health risks and no clinical evidence of benefits. This is akin to someone who believes vaccines are linked with autism evaluating a vaccine safety study. Neither you nor I nor especially they are qualified to interpret primary medical literature like this. Please just wait for the meta-analyses that include this study in them and leave it at that; we should want to be involved in science, but we should also acknowledge our own limitations and avoid the Dunning–Kruger effect. That's why I, a mod at /c/vegan, am not sitting here shitting my pants over this study that says vegans have a 24% lower cancer risk. It's cool, but I want to wait for people more qualified than I am to interpret it.
Please Try to focus on the scientific aspects and refrain from making overly partisan or inflammatory content. Our aim is to foster a respectful environment where we can delve into the scientific foundations of these topics.
Editing your comment to include a attack on someones character is inflammatory.
This is not about your character. It's about your beliefs being incompatible with modern medical science. You mod this public health community but have pseudoscientific beliefs which are diametrically opposed to a modern understanding of public health. Attacking your character would be to say you're doing this out of malice or baselessly accusing you of shilling. I believe you believe this, and I believe that if you believe a Western pattern diet is killing people (it is; this is the main thing we agree on), it shows strong ethical character to try to warn people about it. What's problematic is that your solution to it is so divorced from science that people who just read comments and nod along because they sound plausible need to be aware of the lens you're interpreting this study through.
You are bringing in external bias into a discussion of one study, which is what this post is about.
We disagree on diet decisions, but nothing I've said in this post has been about my personal choices. You are escalating a grudge you have had historically with me into a unrelated discussion. That is a definitional attack on character.
Please stay on topic, which is the study in the post, if you want to attack my character as human or literate abilities you may do so in a different community, but not here.
The actual study Longitudinal associations between vegetarian dietary habits and site-specific cancers in the Adventist Health Study-2 North American cohort
abstract
Background
Associations between vegetarian diets and risk of common cancers are somewhat understood, but such data on medium-frequency cancers are scarce and often imprecise.
Objectives The objectives of this study was to describe multivariable-adjusted associations between different types of vegetarian diets (compared with nonvegetarians) and risk of cancers at different bodily sites.
Methods The Adventist Health Study is a cohort of 95,863 North American Seventh-day Adventists established between 2002 and 2007. These analyses used 79,468 participants initially free of cancer. Baseline dietary data were obtained using a food frequency questionnaire and incident cancers by matching with state and Canadian provincial cancer registries. Hazard ratios (HRs) were estimated using proportional hazard regression. Small amounts of missing data were filled using multiple imputations.
Results Overall cancers, all vegetarians combined compared with nonvegetarians, had HR: 0.88; 95% confidence interval (CI): 0.83,0.93; P < 0.001, and for medium frequency cancers, HR: 0.82; 95% CI: 0.76, 0.89; P < 0.001. Of specific cancers, colorectal (HR: 0.79; 95% CI: 0.66, 0.95; P = 0.011), stomach (HR: 0.55; 95% CI: 0.32, 0.93; P = 0.025), and lymphoproliferative (HR: 0.75; 95% CI: 0.60,0.93; P = 0.010) cancers, were significantly less frequent among vegetarians. A joint test that HR = 1.0 for all vegetarian subtypes compared with nonvegetarians was rejected for cancers of the breast (P = 0.012), lymphoma (P = 0.031), all lymphoproliferative cancers (P = 0.004), prostate cancer (P = 0.030), colorectal cancers (P = 0.023), medium frequency cancers (P < 0.001), and for all cancers combined (P < 0.001).
Conclusions These data indicate a lower risk in vegetarians for all cancers combined, as well as for medium-frequency cancers as a group. Specific cancers with evidence of lower risk are breast, colorectal, prostate, stomach, and lymphoproliferative subtypes. Risk at some other sites may also differ in vegetarians, but statistical power was limited.
I've ranted about my Standards for Nutritional Evidence before, but here are the major issues
- Observational
- Weak Hazard Ratios
- Absolute risk not calculated or published
- Compares only against the Standard American Diet [SAD]
- Cannot inform on cause and effect
- proportional hazard regression is just a fancy way of saying they tried to model out confounders by guessing
- Food Frequency Questionnaire administered asking about 1 year of food administered once, in a study that has been running since 2002 - 23 years! 21 years since a FFQ
TLDR: It's a interesting study, it is hypothesis generating and should be used for further science or to design a interventional study, it should not be used to inform on personal health choices, it simply does not have the power or the rigor
Don't you think calling a proportional hazards model "guessing" is doing a disservice to Lemmy posters here? Not everyone has a background in statistics, and honestly quite a few people here struggle with highschool math. It's not much different than linear regression. Its fancy linear regression, not fancy guessing.
Does their model not fit or something?
At its most essential level - when you have a epidemiology dataset you don't know the relationships until you analyze the data, in order to make controls for some factors in the data you have to assume some relationship for that factor. It's typically assumed to be some linear relationship. If you knew the relationship between factors with certainty, you wouldn't need a epidemiological dataset in the first place, but since we are trying to control for a confounder by definition we don't know the relationship. It is a guess in colloquial terms, a educated guess to be sure, but still a guess.
This is a good overview of cause and effect in inferential statistics, and confounders (start at the 5 minute mark) https://www.youtube.com/watch?v=n4YV7tEtg3I
If you prefer something written with more rigor: https://pmc.ncbi.nlm.nih.gov/articles/PMC4017459/
the researchers should notice that wrong assumptions about the form of the relationship between confounder and disease can lead to wrong conclusions about exposure effects too.
This is a critical weakness of epidemiology when inferences are made about something not directly measured.
while all true, I'm taking issue with you calling it guessing, not that it's a perfect method.
also, we use epidemiological data because it's kind of hard to do a double blind study where you tell some group of people to eat meat for 20 years, and another group of people to not eat meat for 20 years, and then have them live exactly identical lives for that 20 years.
you're kind of not mentioning that. it's kind of dishonest when the audience (Lemmy) is full of layman who are definitely not reading your linked citations, I certainly don't have time to. I'm not defending this study at all because I haven't read it, I'm just taking issue with how you are presenting these (useful) techniques
These are useful techniques to generate hypothesis to test, absolutely!
The results from epidemiology, especially weak hazard ratios, and poor confounders, really have no business being publicized to lay people to get them to change any aspect of their life.
also, we use epidemiological data because it’s kind of hard to do a double blind study where you tell some group of people to eat meat for 20 years, and another group of people to not eat meat for 20 years, and then have them live exactly identical lives for that 20 years.
Sure, but that isn't science. Science is a falsifiable hypothesis that can be tested, if we say we can't test these things then we are not in the realm of empiricism but of theology. That is fine, but we should be clear that the message isn't backed by science.
Good lord there should be a confirmation for the delete button.
Anyway,
The results from epidemiology, especially weak hazard ratios, and poor confounders, really have no business being publicized to lay people to get them to change any aspect of their life.
This is certainly a problem with science reporting.
if we say we can’t test these things then we are not in the realm of empiricism but of theology
I would like to know how you think we've established the link between smoking and cancer. Or air quality, etc. It's just a tool, not something perfect.
theology
This is the key of my issue with your statements here. I am no vegetarian. When you are being hyperbolic like this, it makes everything else you say suspect.
I would like to know how you think we’ve established the link between smoking and cancer. Or air quality, etc. It’s just a tool, not something perfect.
Ah, Good question! I do cover this in my evidence standards post (i know, I know, no time to read, but I'll quote the bits here) https://discuss.online/post/25820268
What about smoking? Smoking causes cancer and that was all observational epidemiology.
That epidemiology had hazard ratios of 6000 (far greater then 4), was consistent across different reputable studies, demonstrated in animal interventions… and most importantly there is no medical benefit to smoking… Giving up smoking is all upside, no real tradeoff. That being said… we actually don’t know that smoking causes cancer in all contexts - the health of the subject, their diet, their lifestyle, their genetics… there are smokers who die without lung cancer.
theology
This is the key of my issue with your statements here. I am no vegetarian. When you are being hyperbolic like this, it makes everything else you say suspect.
I'm not being hyperbolic, if the response to feedback about the rigor of something is that the thing is untestable, that is no longer science.
Depending on your lemmy interface there should be a undelete button too.
That epidemiology had hazard ratios of 6000
Yes, fine, this is what I am saying: Take issue with the findings of the model, not epidemiological data (edit: as a technique that is akin to theology). Focus on that.
I’m not being hyperbolic
It was theology before, but now that hazard ratio is fine, because the number is big? There's big numbers in the bible too, friend. This is what I would call hyperbole. Either it's theology or it's not.
Yes, fine, this is what I am saying: Take issue with the findings of the model, not epidemiological data. Focus on that.
I totally agree with you, actually.
Under what circumstances would I personally look at a observational epidemiology study and consider it to modify my behavior?
- Hazard Ratios greater then 4 (far greater honestly, but 4 is the floor)
- Absolute Risk reported in the paper (not relative)
- Clear signal across different studies
However, this is so rare, that it is exceptional.
It was theology before, but now that hazard ratio is fine, because the number is big? There’s big numbers in the bible too, friend. This is what I would call hyperbole. Either it’s theology or it’s not.
It does not prove causation, there is no downside to giving up smoking, so why not? Does smoking cause cancer in all circumstances, no. So, give up smoking, sure why not. Does smoking cause cancer? It hasn't been proven.
There is more nuance here, in some contexts smoking is correlated with cancer. I have my own personal theories on the incidence of cancer increasing even though smoking has existed throughout documented history, but that is neither here nor there.
there is no downside to giving up smoking, so why not?
To you there is no downside. People actually do take up smoking for reasons. For example, I have worked shitty jobs where smokers get extra breaks, or get extra time to bullshit with the boss. They also might do it because they feel it looks cool. These are not valid reasons for me (being that it is unhealthy, expensive, and messy). It sure seems like I'm being nit picky here, but this statement just isn't true! It's also pretty hard to quit if you've started, why bother doing it? The money may be less important than the downsides of withdrawals there. It's why it's important to point out that smoking is bad for you, and epistemological studies is one of the tools we have for that.
Similarly, people give up meat for reasons that do not make sense to you: It can be expensive, it can contain pathogens, industrial farming is a blight, etc etc etc. For them, the benefits do not outweigh the negatives. I'm not litigating this. I'm just pointing it out. I eat meat. This isn't part of my identity, it is the force of gravity for me. Eating meat is easy.
There is more nuance here
Incredible that you're speaking about nuance when you've just called epistemology theology. I mean I totally agree with you, the devil is in the details, but.... damn dude. :')
epistemology theology.
Ah, I see our disconnect. I don't think of epidemiology as theology at all. I think of the abandonment of science throwing up all enquiry on a subject because its hard to test, but still using weak epidemiology to inform public policy, guidelines, and even lifestyle... that is theology.
Epidemiology is a tool that can be used in science, it is hypothesis generating after all, but by itself it is not science, it is a part of science, not the end of science.
Weak epidemiology can be engineered for any result you want... Paper - Grilling the data: application of specification curve analysis to red meat and all-cause mortality
Yes, there does seem to be a disconnect here.
Using weak epidemiology to inform public policy, etc, is bad.
Calling epidemiology guessing, or saying that it's use is "not in the realm of empiricism but of theology" is hyperbole. If you're going to critique a paper because it's being presented to a layman audience, you should probably avoid that (that being: exaggeration. Don't do that.).
This has, more or less been my point for this entire comment chain. Your exaggeration is harmful to your overall argument. Especially because people take up a sports-team sort of ideological following for eating meat vs not eating meat. I'd be especially avoidant of exaggeration for that reason.
I didn't say epidemiology was guessing
I said the statistical controls for confounding variables are guesses. And that is true
I didn't say epidemiology was theology.
The abandonment of science, falling back onto week epidemiology is theology
I don't know how to express this more clearly
Thanks for the leg work of getting to the actual study instead of "science journalist" often incorrect or misunderstood understanding of things.
“science journalist” often incorrect or misunderstood understanding of things.
I think they understand perfectly, the study is just a talking point for some agenda. If the journalist was interested in science they would have mentioned some of the things I did.
- Food Frequency Questionnaire administered asking about 1 year of food administered once
I don't think you read that whole part of the study. The direct follow-up to that reads: "Validation of the dietary habits was primarily by comparison with 6 24-h dietary recalls administered by telephone in a representative validation subgroup (1100 participants)." (Per the methodology section.)
From there, I'm not going to discuss the other points because I know neither you nor I are qualified to interpret primary medical literature; we should leave that to the experts, and I'm not going to fan the flames of this by doing something I'm not qualified to do. But I can objectively point out that your last bullet point is highly selective and misleading. I can also point out for readers at home that "Compares only against the standard American diet*" isn't necessarily true? You're assuming omnivorous Seventh day Adventists generally follow that diet similar to the general population, and moreover, you subscribe to a pseudoscientific diet which is notoriously even less healthy somehow. There is no reason people should trust you to interpret primary medical literature when your qualifications are that you follow and advocate a diet which the medical community considers deeply unscientific and very unhealthy.
* I'm not sure where you got the term "SAD" from, but it's generally called the "Western pattern diet".
Hey no fair updating your post with insults after I responded nicely to you.
can also point out for readers at home that
Your just grand standing, you don't want to talk to me about what I wrote at all.... That is a bit rude.
There is no reason people should trust you to interpret primary medical literature when your qualifications are that you follow and advocate a diet which the medical community considers deeply unscientific and very unhealthy.
That is just an appeal to consensus fallacy, anyone who wants to learn more is welcome to discuss, ask question, etc at !carnivore@discuss.online but that has nothing to do with this post at all. My analysis of the posted study does not rely on my personal dietary choices.
The direct follow-up to that reads: “Validation of the dietary habits was primarily by comparison with 6 24-h dietary recalls administered by telephone in a representative validation subgroup (1100 participants).”
Yup! But they were validating the globally administered FFQ, so that doesn't change the rigor of the data, just the confidence in the FFQ.
I know neither you nor I are qualified to interpret primary medical literature; we should leave that to the experts, and I’m not going to fan the flames of this by doing something I’m not qualified to do. But I can objectively point out that your last bullet point is highly selective.
I am trained in the maths, so I can speak to the weakness of the signal here.
Did you happen to look at my post on standards of evidence? I welcome feedback, and I know you do care about data and science even if we disagree on diets, so I would appreciate your thoughts. https://discuss.online/post/25820268
Jet, I can't honestly take you seriously when you follow a carnivore diet. I'm sorry. It's like someone who believes in autistic enterocolitis trying to debunk a study about vaccine safety. That's the heart of the matter: your interpretation of scientific literature is so asinine and filled with disinformation that I honestly believe you believe this but also have to acknowledge that talking with someone on a carnivore diet about nutrition science is pigeon chess. "It's not me who's wrong! It's the entirety of medical academia and all of the health institutes who are wrong!"
your interpretation of scientific literature is so asinine and filled with disinformation that I honestly believe you believe this but also have to acknowledge that talking with someone on a carnivore diet about nutrition science is pigeon chess.
Shouldn't our divergent views be manifest in our standards of evidence? We might be stuck in a loop of pre-selecting a conclusion before examining the data, hence our inability to be empathetic to the others conclusions.
I don't need you to agree with me, but if you can speak to my consistency of evidentiary standards we could have a productive discussion.
“It’s not me who’s wrong! It’s the entirety of medical academia and all of the health institutes who are wrong!”
Not the entirety! https://thesmhp.org/
Jet, linking me to the SMHP this would be like me linking you to the PCRM in a reality where the broader scientific community supports an animal-based diet. "Not all doctors believe that! Check out the PCRM!" in a world where an exclusive carnivore diet reduces the rates of major chronic diseases by double digits, and you'd tell me I'm cherry-picking like the far-right cherrypicks climate scientists who don't believe in man-made climate change.
Closing the loop on soliciting feedback on my evidentiary standards process: You can't look at it because you know you disagree with my conclusions. That seems rather circular
A document which answers many of the critiques you brought up in your post edits above (wrt SAD).
I'm simply providing counter evidence to the blanket statement that "the entirety of medical academia and all of the health institutions that are wrong"
I'm reading what you wrote, and responding to it as written, I'm taking you seriously, which I hope you do for me as well.
you’d tell me I’m cherry-picking like the far-right cherrypicks climate scientists who don’t believe in man-made climate change
Now your just putting words in my mouth.
Okay, to be fair, that's what I'm hoping you would say to me in that reality. Even with animal ethics at stake in that alternate reality, I would hope you'd call me out. When I say the entire biology community believes in evolution, the entire climatology community believes in man-made climate change, and the entire medical community believes vaccines do not cause autism, that isn't a literal mathematical universal qualification; it's saying that the support is so overwhelming that any dissent is absolutely negligible and not even worth considering.
it’s saying that the support is so overwhelming that any dissent is absolutely negligible and not even worth considering.
If that is your worldview on things you believe no wonder you always end up attacking me and my character.
I think it was based on AHS-2 - https://adventisthealthstudy.org/studies/AHS-2
90k people, only one FFQ administered at the 2-3 year mark, no FFQ followup. The study started in 2002-2007; So we are 23ish years into the study and the last FFQ was done 21 years ago.
It is a confounder that really does need to be highlighted when examining the publications.
I just see this as american meat being a bad thing. glad I don't/won't eat it