Comment by ravenstine | Hacker News Reader

Comment by ravenstine | Hacker News Reader

Parent Full thread

The standard of proof should be conducting actual science, not epidemiology or surveys. Epidemiology is good for some things, but as someone with an ongoing interest in health and especially metabolism, I don't believe epidemiology is very good at all for assessing what's optimal for human health unless the datapoints are overwhelming (ex. smokers have tar in their lungs that is rarely found in non smokers). Such evidence doesn't present itself against meat eating, nor dietary cholesterol or saturated fat. To assume that the science is settled for any one of those things in isolation is a display of ignorance of the highest order. Though they can be debated, as can anything.

But we're unlikely to see better scientific research because there's ethical problems around keeping people behind forcing people to eat specific foods and keeping them behind lock and key for observation, and so they don't eat a single gram of anything outside the boundaries of the study. Unless more of such research can be carried out, too much of what we do is uncontrolled to make meaningful observations about what's unhealthy and what's not.

What meat eating has that other diets don't have is not just the phenotype of our digestive systems (being like that of other meat eaters), or archaeological evidence of our ancestors using sharp tools and arrows, but stable isotope testing of fossil evidence linking our recent ancestors to a substantial amount of meat eating for around 4 million years. If you compare that to the amount of time we've had agriculture, ~30k years, the time in which we've been subsisting not only on more plants but plants selectively bred to be more nutritious than they were in the wild is a flash in the pan.

My opinion is that we should reconsider the ethics around researching nutrition, because it's been an overall failure, and has probably harmed more people than would scientific studies. This isn't to say that ethics in science should be abolished, but relying on the mouse model and epidemiology doesn't appear to be doing the most good.

> What meat eating has that other diets don't have is not just the phenotype of our digestive systems (being like that of other meat eaters), or archaeological evidence of our ancestors using sharp tools and arrows, but stable isotope testing of fossil evidence linking our recent ancestors to a substantial amount of meat eating for around 4 million years.

How would that invalidate research showing, for example, a link between certain cancers and the consumption of red meat? That some behaviour produces a competitive advantage at one time does not mean it comes without risk, nor that newer behaviours aren't better in some or even all ways, nor that it would remain advantageous for all time.

It wouldn't necessarily invalidate any research around a link to cancer. What it shows is that any link to cancer we currently have in regards to meat eating cannot be conclusive evidence that it's deleterious to such an extent that it is an unacceptable risk to the individual or the population as a whole. If humans developed to eat meat, that doesn't support the idea that red meat causing cancer (hypothetically speaking) is reason enough to avoid it. When people come from the perspective that humans are actually supposed to eat mostly plants, or that plants are generally good for us, or that meat is empty calories, the conclusion that follows from such studies is commonly "we should eat less meat." That is wrong. The idea that plants are better for humans and that meat is both devoid of nutrition and dangerous is by no means an uncommon viewpoint.

You are absolutely right about the cost-benefit balance in the case of cancer and red meat (or really in the consumption of anything). That's why it's important for people to understand our evolutionary past, even if it might seem peripheral to the cancer link itself. If we go into subjects like this with a flawed mindset, we'll get flawed results. If it was more commonly understood that meat isn't just tasty, non-nutritious, artery clogging filler food that greedy humans only recently started doing because slaughtering animals is fun, jumping to conclusions about red meat and cancer wouldn't be nearly as popular.

It’s well understood that meat was a calorie dense food that helped human beings not starve in the past. This was its primary evolutionary advantage.

In the world where food is abundant and chronic disease caused by overconsumption of calorie dense foods is extremely prevalent, it’s worth re-examining meat as a staple food. This is not touching on our polluted world and meat’s role as bioamplifier of toxins in our foood chain, along with a bunch of other good reasons to re-examine our relationship to meat as a society

It's hard to clearly go from "a link between certain cancers and the consumption of red meat" to a causal statement.

Like, it's plausible that the red meat is causing these cancers. It's also plausible to have a different cause - e.g. if people's lifestyles lead to a correlation between consumption of red meat and smoking (for the sake of argument - I have no idea if there actually is one) then we would observe a link between lung cancer and the consumption of red meat without there being any causal relation.

It's also quite plausible that certain ways of preparing food is causing these cancers (e.g. nitrates in bacon, and known carcinogens from high temperature frying on pan or open flame), and we're seeing the effect because we're currently often preparing red meat in this way, but replacing the red meat with fish or vegetarian equivalent in the same fried dish would still have the same carcinogens, and thus the intervention would not get the expected results.

So, a correlation between certain cancers and the consumption of red meat is a good probable cause for an investigation of the causal relationships, but I'd be wary of using it to argue for major dietary changes - the last decades had far too many counterproductive dietary recommendations (e.g. regarding cholesterol) because of non-causal evidence leading into wrong assumptions.

Ignoring that I haven't made any case for major dietary changes (not here, at least), and though I haven't read the GBD I doubt they will have said any more than a reduction in an individual's red meat consumption is probably a good thing - ignoring all that, I'm willing to bet they've taken into account an obvious factor like smoking. Additionally, the claim in the GBD, according to the Times, is regarding unprocessed red meat.

And indeed, correlations aren't causation, but the claim is “The study said it had identified causal relationships between eating red meat and coronary heart disease, breast cancer and strokes, on top of links already established with diabetes and colon cancer” and the "serious errors" are that “the study had failed to take into account the nutritional benefits of red meat as a source of vitamins, iron and protein, that should be weighed against any risks”. I find that a misleading use of the word "error". Don't you?

> My opinion is that we should reconsider the ethics around researching nutrition, because it's been an overall failure

That just isn’t true. Advances in nutritional science have had a major impact. Vitamin C made transatlantic shipping empires far more viable. The discovery of folate eliminated a whole class of birth defects. Fluoride improved the oral health of nations. The whole reason that nutritional science has any credulity today is that it had an unequivocal positive effect on mankind’s quality of life at barely any cost.

Reply on news.ycombinator.com