,
17 tweets,
8 min read
Read on Twitter
Series of 6 papers published today in Ann Int Med from “NutriRECS” group.
They recommend people continue to consume unprocessed & processed red meat at current rates
I disagree, in USA
My perspective - in 17 tweets
#ReduceRedMeat in USA
n.pr/2mkxWN7
They recommend people continue to consume unprocessed & processed red meat at current rates
I disagree, in USA
My perspective - in 17 tweets
#ReduceRedMeat in USA
n.pr/2mkxWN7
3 of 17
The recommendations are stated to be “weak, with low-certainty evidence”.
I agree.
Beyond weak, they are reckless.
They will confuse the public & undermine scientific credibility.
Potential to harm public health & environment.
The recommendations are stated to be “weak, with low-certainty evidence”.
I agree.
Beyond weak, they are reckless.
They will confuse the public & undermine scientific credibility.
Potential to harm public health & environment.
4 of 17
3 of the 6 papers are reviews/meta-analysis of cohort studies. Different outcomes for each paper.
Authors restate what we know - observational studies show association, not causation.
Randomized trials provide stronger evidence
Do RCTs exist? Can they be done?
3 of the 6 papers are reviews/meta-analysis of cohort studies. Different outcomes for each paper.
Authors restate what we know - observational studies show association, not causation.
Randomized trials provide stronger evidence
Do RCTs exist? Can they be done?
5 of 17
12 trials identified
Only 1 trustworthy-WHI ?Not a meat reduction study?
Dismissed Lyon Heart?!
Omitted PREDIMED & DPP?! (others?)
How do you “pool” data from 1 study?
How is that a meta-analysis?
Better conclusion–not enuf RCTs for meta-analysis
12 trials identified
Only 1 trustworthy-WHI ?Not a meat reduction study?
Dismissed Lyon Heart?!
Omitted PREDIMED & DPP?! (others?)
How do you “pool” data from 1 study?
How is that a meta-analysis?
Better conclusion–not enuf RCTs for meta-analysis
6 of 17
More and better trials needed? Always.
Should be trustworthy…sure.
Trustworthy red meat RCT’s will require blinding?
WTF? BLINDING?
Assign to high vs low red meat, keep subjects blinded for 10-20 yrs? 🤔
Evidence grading criteria/method inappropriate
More and better trials needed? Always.
Should be trustworthy…sure.
Trustworthy red meat RCT’s will require blinding?
WTF? BLINDING?
Assign to high vs low red meat, keep subjects blinded for 10-20 yrs? 🤔
Evidence grading criteria/method inappropriate
7 of 17
GRADE method used to assess certainty of evidence
Grades of Rec Assess Dev’t & Eval
Designed for clinical practice guidelines
RCTs provide strongest evidence
But RCTs for clinical outcomes typically not available for lifestyle (diet, PA, sleep, etc)
Alternate method?
GRADE method used to assess certainty of evidence
Grades of Rec Assess Dev’t & Eval
Designed for clinical practice guidelines
RCTs provide strongest evidence
But RCTs for clinical outcomes typically not available for lifestyle (diet, PA, sleep, etc)
Alternate method?
8 of 17
HEALM
Alternate method of assessing certainty of evidence designed to address lack of RCTs for lifestyle medicine.
For red meat, need to consider obs. epi.
Interpret responsibly (association, not causation)
Put in proper perspective along continuum of causal inference
HEALM
Alternate method of assessing certainty of evidence designed to address lack of RCTs for lifestyle medicine.
For red meat, need to consider obs. epi.
Interpret responsibly (association, not causation)
Put in proper perspective along continuum of causal inference
9 of 17
Back to cohort studies
John Sievenpiper, MD, PhD is co-author on one of the 6 Ann Int Med papers that did met-analyses of cohort studies.
He is a conscientious objector of the conclusions stated in guidelines/recommendations paper, for which he was not a co-author.
Back to cohort studies
John Sievenpiper, MD, PhD is co-author on one of the 6 Ann Int Med papers that did met-analyses of cohort studies.
He is a conscientious objector of the conclusions stated in guidelines/recommendations paper, for which he was not a co-author.
10 of 17
NutriRECS fails to appreciate & address a fundamental issue – in obs epi, for those who eat less meat, what do they eat instead?
If it is more pizza w/ refined flour & processed cheese, no health benefit expected.
Critical to model: what would/could replace red meat?
NutriRECS fails to appreciate & address a fundamental issue – in obs epi, for those who eat less meat, what do they eat instead?
If it is more pizza w/ refined flour & processed cheese, no health benefit expected.
Critical to model: what would/could replace red meat?
11 of 17
⬇️Red meat alone wouldn’t impact health much if replaced w/ ultraprocessed foods (refined grain/added sugar).
⬆️Quality of diets requires ⬆️veg/beans/fruit/fiber.
Focus there
What would they replace?
US meat consumption is highest in world - lots of room to decrease
⬇️Red meat alone wouldn’t impact health much if replaced w/ ultraprocessed foods (refined grain/added sugar).
⬆️Quality of diets requires ⬆️veg/beans/fruit/fiber.
Focus there
What would they replace?
US meat consumption is highest in world - lots of room to decrease
12 of 17
But with less meat how would we get enuf protein?
(didja see the last tweet – the rest of the world eats less, in many cases much less)
Got that covered….here - doi.org/10.1093/nutrit…
But with less meat how would we get enuf protein?
(didja see the last tweet – the rest of the world eats less, in many cases much less)
Got that covered….here - doi.org/10.1093/nutrit…
13 of 17
NutriRECS authors note that they do not address ethical or environmental issues.
Thanx for acknowledging
Thanx to @aaronecarroll and @DrTiff_ for concluding with this mention in editorial.
NutriRECS authors note that they do not address ethical or environmental issues.
Thanx for acknowledging
Thanx to @aaronecarroll and @DrTiff_ for concluding with this mention in editorial.
14 of 17
Notably missing was more extensive assessment of intermediate risk factors
Blood lipids & blood pressure listed among surrogate measures assessed
Diet can change these factors in 6-8 weeks
But >6 month eligibility likely problematic for these – many trials excluded
Notably missing was more extensive assessment of intermediate risk factors
Blood lipids & blood pressure listed among surrogate measures assessed
Diet can change these factors in 6-8 weeks
But >6 month eligibility likely problematic for these – many trials excluded
15 of 17
Circling back
Red meat RCTs don’t/won’t exist for mortality/morbidity
Risk factor RCTs likely to be <6 months – excluded here
Cohort data DO support benefit from decreased red meat – but these are dismissed
Self-fulfilling “weak rec’s w/ low-certainty of evidence”
Circling back
Red meat RCTs don’t/won’t exist for mortality/morbidity
Risk factor RCTs likely to be <6 months – excluded here
Cohort data DO support benefit from decreased red meat – but these are dismissed
Self-fulfilling “weak rec’s w/ low-certainty of evidence”
16 of 17
Singling out red meat not helpful.
Focus on food patterns, finding more agreement, making it less confusing to public.
Given:
Current meat intake > required/recommended
& veg/fruit/bean intake < recommended
Try:
*More (whole) veg/beans/fruits
*Less meat/better meat
Singling out red meat not helpful.
Focus on food patterns, finding more agreement, making it less confusing to public.
Given:
Current meat intake > required/recommended
& veg/fruit/bean intake < recommended
Try:
*More (whole) veg/beans/fruits
*Less meat/better meat
17 of 17
Has self-appointed NutriRECS group successfully challenged & refuted previous rec's for reducing red meat established & published by:
-Am Heart Assoc.
-Dietary Guidelines for Americans
-World Health Organization
-Eat-Lancet Commission
+ more
?
No
#ReduceRedMeat in USA
Has self-appointed NutriRECS group successfully challenged & refuted previous rec's for reducing red meat established & published by:
-Am Heart Assoc.
-Dietary Guidelines for Americans
-World Health Organization
-Eat-Lancet Commission
+ more
?
No
#ReduceRedMeat in USA
