5 ms·

Aren't most studies contradictory ? For one study claiming A you can find another claiming the opposite. In the 80s/90s there was this constant fear about chol

Aren't most studies contradictory ? For one study claiming A you can find another claiming the opposite. In the 80s/90s there was this constant fear about cholesterol and eggs (perhaps promoted by the cereals lobby?) : did people's habits change? If yes, it was probably because of the monopoly of TV as the main source of "information". Nowadays with internet, every one follows his own little bubble of information.

em500 4y ago
Yup, according to studies, everything we eat both causes and prevents cancer[1]. AFAICT it's almost impossible to conduct convincing nutrition science. Even if the signs of the effect in observational studies are all in the same direction (usually they're not, see the reference), you still can't rule out many confounding effects. To demonstrate causality, you'd need to force a large number of people on specific diets for a long time in a proper randomized controlled trial, or show very specific bio-chemical reactions to particular food components, or (easier but still very hard) dream up some convincing "natural experiment" or "instrumental variable" to workaround the fact that people mostly choose their own diet. [1] https://www.sciencealert.com/everything-we-eat-both-causes-and-prevents-cancer https://www.sciencealert.com/everything-we-eat-both-causes-a...
- AlbertCory 4y ago
  There was a study on "gluten intolerance" (which unfortunately I deleted) where they did in fact control people's diets in a double-blind way, and check the validity of their symptoms. Maybe someone can find it, but even with all that rigor, it was somewhat inconclusive.
- phpisthebest 4y ago
  I think most nutrition science can be boiled down to 4 rules that most people should follow, and ignore the rest 1. Avoid processed Foods, Learn to cook your own meals 2. Eat a variety of foods 3. Control your portions (especially carbs) 4. Drink more water Everything else is mostly irrelevant to the normal person. Are eggs good or bad, is Red meat worse than Chicken, etc etc... Who Cares. Eat it all in moderation, while mixing in fruits and veggies all in proper portion I think it was Food Theory Channel or somewhere that had a video on the amount of influence various industries have had over the "recommendation" by "nutrition science" experts and / or the government, the end result is most "nutrition science" is really marketing not science
  - copperx 4y ago
    I would not even follow these 4 recommendations without a ton of evidence. Yes, they "make sense" because that's what we've been taught since we were children, but are these behaviors good compared to doing nothing?
    - phpisthebest 4y ago
      Well "doing nothing" would be not eating, and pretty sure "Breatharian" is right up there with the flat earth
- mrandish 4y ago
  Due to some fairly serious diet-related personal health issues (obesity and several related co-morbidities), 5 years ago I started closely following nutrition science. In this time, the clearest thing I've learned is that the vast majority of "Nutrition Science" is poor science - to the extent of hardly being science at all. * If the data is based on diary data, it's noisy to the extent of being nearly useless and no causation can be implied. * Doing rigorous controlled studies on humans is very difficult and astronomically expensive. Even then, it's incredibly easy for significant confounds and errors to slip in. * With nutrition studies in general, animal models are usually weak proxies for humans. A large number of significant nutrition results in animals fail to replicate in humans. * Humans metabolic responses to nutrition are highly variable and contingent on everything from genetic make-up and ethnicity to behavior and environment, etc. So, even in the rare cases statistical causation is demonstrated, it probably still depends on other variables. * There is a high degree of bias in nutrition science. This includes many funding sources having vested interests from commercial (food lobbies (meat, wheat, milk, etc)) and many researchers having ideological biases (anti-meat vegetarians, anti-meat climate concerns, anti-meat animal welfare, etc) * Accepting any result as significant requires skeptically examining the methods, statistics and history of the authors and funders.
  - copperx 4y ago
    I agree. And al of that leads to information fatigue. After reading many studies, I'm even worried about "common sense" nutrition knowledge, for example: * I'm worried about increasing my fruit and vegetable intake, because that will raise my pesticide burden. It is not clear whether the benefits of fruits and vegetables make up for the increased probability of cancer of pesticide consumption. Organic produce is not a solution, as the pesticide burden is even higher. * Increasing fish consumption of even "healthy" dishes will increase heavy metal consumption. Is it worth it? * Taking Omega 3 supplements greatly increases the risk of prostate cancer. Is it really worth it? You can't rely on anything to make dietary decisions for longevity. The best one can aim for is dietary advice to maintain a healthy weight. That, at least, can be quickly proven empirically.
    - mrandish 4y ago

em500 4y ago

Yup, according to studies, everything we eat both causes and prevents cancer[1]. AFAICT it's almost impossible to conduct convincing nutrition science. Even if the signs of the effect in observational studies are all in the same direction (usually they're not, see the reference), you still can't rule out many confounding effects. To demonstrate causality, you'd need to force a large number of people on specific diets for a long time in a proper randomized controlled trial, or show very specific bio-chemical reactions to particular food components, or (easier but still very hard) dream up some convincing "natural experiment" or "instrumental variable" to workaround the fact that people mostly choose their own diet. [1] https://www.sciencealert.com/everything-we-eat-both-causes-and-prevents-cancer https://www.sciencealert.com/everything-we-eat-both-causes-a...

AlbertCory 4y ago
There was a study on "gluten intolerance" (which unfortunately I deleted) where they did in fact control people's diets in a double-blind way, and check the validity of their symptoms. Maybe someone can find it, but even with all that rigor, it was somewhat inconclusive.
phpisthebest 4y ago
I think most nutrition science can be boiled down to 4 rules that most people should follow, and ignore the rest 1. Avoid processed Foods, Learn to cook your own meals 2. Eat a variety of foods 3. Control your portions (especially carbs) 4. Drink more water Everything else is mostly irrelevant to the normal person. Are eggs good or bad, is Red meat worse than Chicken, etc etc... Who Cares. Eat it all in moderation, while mixing in fruits and veggies all in proper portion I think it was Food Theory Channel or somewhere that had a video on the amount of influence various industries have had over the "recommendation" by "nutrition science" experts and / or the government, the end result is most "nutrition science" is really marketing not science
- copperx 4y ago
  I would not even follow these 4 recommendations without a ton of evidence. Yes, they "make sense" because that's what we've been taught since we were children, but are these behaviors good compared to doing nothing?
  - phpisthebest 4y ago
    Well "doing nothing" would be not eating, and pretty sure "Breatharian" is right up there with the flat earth
mrandish 4y ago
Due to some fairly serious diet-related personal health issues (obesity and several related co-morbidities), 5 years ago I started closely following nutrition science. In this time, the clearest thing I've learned is that the vast majority of "Nutrition Science" is poor science - to the extent of hardly being science at all. * If the data is based on diary data, it's noisy to the extent of being nearly useless and no causation can be implied. * Doing rigorous controlled studies on humans is very difficult and astronomically expensive. Even then, it's incredibly easy for significant confounds and errors to slip in. * With nutrition studies in general, animal models are usually weak proxies for humans. A large number of significant nutrition results in animals fail to replicate in humans. * Humans metabolic responses to nutrition are highly variable and contingent on everything from genetic make-up and ethnicity to behavior and environment, etc. So, even in the rare cases statistical causation is demonstrated, it probably still depends on other variables. * There is a high degree of bias in nutrition science. This includes many funding sources having vested interests from commercial (food lobbies (meat, wheat, milk, etc)) and many researchers having ideological biases (anti-meat vegetarians, anti-meat climate concerns, anti-meat animal welfare, etc) * Accepting any result as significant requires skeptically examining the methods, statistics and history of the authors and funders.
- copperx 4y ago
  I agree. And al of that leads to information fatigue. After reading many studies, I'm even worried about "common sense" nutrition knowledge, for example: * I'm worried about increasing my fruit and vegetable intake, because that will raise my pesticide burden. It is not clear whether the benefits of fruits and vegetables make up for the increased probability of cancer of pesticide consumption. Organic produce is not a solution, as the pesticide burden is even higher. * Increasing fish consumption of even "healthy" dishes will increase heavy metal consumption. Is it worth it? * Taking Omega 3 supplements greatly increases the risk of prostate cancer. Is it really worth it? You can't rely on anything to make dietary decisions for longevity. The best one can aim for is dietary advice to maintain a healthy weight. That, at least, can be quickly proven empirically.
  - mrandish 4y ago