SoylentNews is people
SoylentNews
Submitted via IRC for Bytram
No need to cut down red and processed meat, study says
Most people can continue to eat red and processed meat as they do now. A major study led by researchers at McMaster and Dalhousie universities has found cutting back has little impact on health.
A panel of international scientists systematically reviewed the evidence and have recommended that most adults should continue to eat their current levels of red and processed meat.
The researchers performed four systematic reviews focused on randomized controlled trials and observational studies looking at the impact of red meat and processed meat consumption on cardiometabolic and cancer outcomes.
In one review of 12 trials with 54,000 people, the researchers did not find statistically significant or an important association between meat consumption and the risk of heart disease, diabetes or cancer.
In three systematic reviews of cohort studies following millions of people, a very small reduction in risk among those who had three fewer servings of red or processed meat a week, but the association was uncertain.
The authors also did a fifth systematic review looking at people's attitudes and health-related values around eating red and processed meats. They found people eat meat because they see it as healthy, they like the taste and they are reluctant to change their diet.
The five systematic reviews, a recommendation and an editorial on the topic were published in the Annals of Internal Medicine today.
(Score: 5, Insightful) by theluggage on Wednesday October 02 2019, @11:30AM (5 children)
Thanks for taking the time to convert the extra-deaths-per-1000 into relative % risk - you've nicely demonstrated how to spin these sorts of statistics to maximise scaremongering.
What table actually says for "Overall cancer incidence" is:
Lifetime population risk: 185 per 1000 population.
Risk difference: -12 per 1000 population (95% confidence interval between -18 and -4)
So, to make that clear, around 185 people out of 100 are expected to get cancer, but everybody cutting down on red/processed meat might reduce that to between 167 and 181.
Oh, dear, that doesn't sound very scary so lets take the top end of that result (18 - it could just as easily be 4) and express it as a percentage of the 185 people (out of 1000) expected to get cancer anyway - so 9.7%... might as well round that up to 10% (even though you've already cherry-picked the top end of the confidence interval - a fairer figure would be 6% +/- 4% ) - so 10% increased cancer risk !!!!! - headline sorted!
So - to put it another way, you have a ~18.5% base risk of getting cancer, cutting red/processed meat might reduce that to ~16.7%-18.1% assuming that the original studies weren't just due to test subjects lying about jogging 10 miles a day and not smoking.
Unfortunately, kidnapping 1000 identical twins, locking them in a room for 10 years while feeding them healthy fruit and veg with either real or placebo bacon is neither ethical nor practical, so all of these "inference" studies should be taken with a huge pinch of
salt... er... maybe lemon juice or a hint of garlic?(Score: 0) by Anonymous Coward on Wednesday October 02 2019, @01:08PM
Is this one of those overunity probabilities they discovered in financial markets?
https://web.williams.edu/Mathematics/sjmiller/public_html/341Fa09/handouts/Burgin_ExtendedProbs.pdf [williams.edu]
(Score: 2) by JoeMerchant on Wednesday October 02 2019, @02:37PM
Yeah, that reincarnation-cloning technique is really cancer-prone...
🌻🌻 [google.com]
(Score: -1, Flamebait) by Anonymous Coward on Wednesday October 02 2019, @03:19PM (1 child)
So, quoting wrong numbers is your specialty?? It says, 185 and -18 (-26 to -11).
But whatever cherry picking you like so you can be outraged.
So sad that people don't understand statistics. Yeah, I guess you can eat all the fucking shit you want and we can defund cancer research. It's fucking useless anyway because "1% here and 1% there" ads up to .... 0% ??? So all these people that get cancer, well, fuck them.
(Score: 2) by theluggage on Wednesday October 02 2019, @05:54PM
Sorry - I misread the table- the -18 to -4 was from the next line. Doesn't change the principle of how quoting %changes in number of victims exaggerates the perceived level of risk.
(Score: 3, Interesting) by theluggage on Wednesday October 02 2019, @06:04PM
Whups - Muphrey's law was strong in this one.
The correct confidence interval was -26 to -11 (-18 to -4 came from the next, similarly titled, row) so 18 was the central value and I apologise for suggesting that the GP cherry=picked the extreme end of the range.
The point about how presenting risk as a % change in the number of victims rather than as the number of extra cases in the general population stands, though.
Oh, yes, and there's a missing zero...