What is your suggestion, on what is beneficial and what is not? And how we distinguish between them?
EDITED TO ADD VIDEO LINK AT THE BOTTOM OF WELL FORMULATED STUDIES (from around 10 to 23 mins). IT IS BREATHTAKING TO SEE SOME OF THE IGNORED EVIDENCE FROM 1976, NOTE THE TOP RIGHT REFERENCE AT 15 mins 21 seconds (THIS IS WHY SOME OF US THINK THERE IS A CONSPIRACY)
Where to start with this, as for example science creates unbelievably tasty treats based on the bliss principal, is this processed food good for us? I would say only in moderation, which is almost impossible.
A Typical Bad Study
If you are determined to prove meat causes cancer then do the following:
- Start with mice of whom 80% die of cancer no matter what
- Extract an element of meat
- Give the mice an agent that speeds up cancer
- Over feed the mice the extract
- Turn cancer indicators on and off via the extract
- Claim this proves meat causes cancer
Ignore all populations that eat meat mainly in a "cleaner" fashion, as if they don't exist and take in millions of believers. (by cleaner I mean minimally processed diet).
I think science should start with "clean hands", and a willingness to accept you might be wrong. You are meant to start with a hypothesis and then do your best to disprove it, along with others. There really is no point in the sugar industry investigating the benefits or otherwise of sugar. Alternative is full independence is a must.
I would say do not have multiple variables which could all be the cause of your hypothesis and then cherry pick what suits your case. Alternative is to measure real candidates who for example truly really eat what is of concern.
Stop retesting old epidemiology studies with new algorithms (torturing data) to effectively. I believe this is too compromised. Alternative for example cancer in meat, ask the carnivores on this site, and elsewhere to test their cancer (and other bio markers), as well as first world tribes.
Stop the ridiculous meta-analysis of bad datasets. Just because you averaged out 10 studies is completely meaningless if the 10 studies were junk in the first place. Alternative, don't do it.
Never knowingly mislead. Alternative, do not make public anything that does not pass the hazard ratio (2), and if you are going to discuss only cite the absolute risk (relative should be banned for obvious reasons).
Upfront ahead of the title declare all interests including financial backers.
We distinguish between good and bad studies by not having obvious contradictory observational evidence, again using meat, explain the current longest lived being the highest meat consuming in Hong Kong. By having a massive risk differential such as with lung cancer and smoking. By being able to prove something closer to conclusively, such as high LDL vs CAC score.
I am not a scientist but I have read enough to know the bare faced, let's call it what it is "lies" that some of these studies put out. I am excited for the day of reckoning.
I can feel my answer becoming too long so will end here.