Makes sense. Basically there are the pre-existing biases against things like red meat and processed meat, and in favor of things like vegetables. And then the plausibility and what seems interesting / sexy. And then the journals. All these different layers of what actually makes it to publication, and then there's the media coverage on top of that witch is sort of like another layer of the same biases. And in all of that making no mention of the inaccurate food frequency questionnaires and healthy user bias. It's a minefield out there, but pretty easy to see why you keep seeing articles about how terrible bacon is and how amazing blueberries are, for example. Most individual food ingredients are not that powerful one way or the other (though there are obviously some exceptions).
Interesting article. For the most part I agree with the conclusions. However, my takeaway is what I notice missing most from study design, is the lack of instructive distinctions in the host population. We are not all the same. Precision Medicine was supposed to address this by focusing more on the individual, but I think that has turned into window dressing and a way for corporate medicine to pretend they are actually following some of functional medicine principles efforts to find root causes, when in reality they are not. Perhaps it is NOT so much the input, but rather the health of the host which protects the body from various insults along the way. I'm not just referring to age, gender and ethnicity host variables, though those are important. I would like to see study distinctions such as microbial diversity in the gut microbiome, C section vs vaginal, breast-fed vs not breast-fed and for how long, vaccinated and unvaccinated, delayed vaccination, selective vaccination, age at first vaccination, autoimmunity or cancer in the immediate family as study variables.
I have seen a logical error occur repeatedly by all kinds of people, where the vaccinated are compared against covid, as if the two arms are somehow exclusive. Let me show a hypothetical example of what happens when this error is made.
Let us suppose that 5 people who are vaccinated suffer Harm, but 10 people who contract covid suffer harm. Comparing the vaccinated group with the covid group, we would conclude a relative harm of 0.5 for the vaccinated, group, which would be an actual relative benefit. However, in this situation, 5 of those in the covid arm are also vaccinated and the vaccinated are actually counted in each arm. If we compare the unvaccinated who got covid with the unvaccinated who got covid, then the relative harm is 1.0, which means that there is no benefit from vaccination as regards the Harm.
Here's data that would be helpful to examine for churnalism.
"According to data from Techniker Krankenkasse, the largest German medical insurance company, there were a total of 437,593 insurance claims billed under the four diagnostic codes for vaccine injury in 2021. To put those numbers in perspective, the total numbers billed for a vaccine injury code in the two preceding years was 13,777 and 15,044, respectively. As the Daily Skeptic notes, given that TK insures 11 million people, that means 1 in 23, or 4.3%, had a medical treatment billed for vaccine injury. And that assumes all 11 million were vaccinated. The background vaccination rate in Germany is 78%, although most of the unvaccinated are children, so the rate of injury per vaccinated person is likely even higher (5.1%)."
This also ties in with one of the comments by Vinu about vaccine injury. Hopefully there will be somebody willing and able to ask pertinent questions about the data and perhaps give a more nuanced perspective. If not, then maybe this is a smoking gun.
"what makes churnalism possible is shoddy research."
The following study makes HCQ early treatment for covid look like a parachute. Neither age nor the presence of comorbidities were the #1 factor correlating to mortality. The primary factor was failure to treat with HCQ within 72 hours of the onset of symptoms. Is the study churnalistic? Are there serious problems with it? Maybe there are problems with the statistics? If so, please discuss.
"Hydroxychloroquine / azithromycin in COVID-19: The association between time to treatment and case fatality rate"
"Results
A total of 1265 COVID-19 patients with an average age of 44.5 years were studied. Women represented 50.1% of patients, with an overall 5.9 symptom days, SpO2 97%, temperature of 37.3 °C, 41% with at least one comorbidity and 96.1% one symptom or sign. No patient treated within the first 72 h of illness died. The factors associated with higher case fatality rate were age (OR = 1.06; 95% CI 1.01–1.11, p = 0.021), SpO2 (OR = 0.87; 95% CI 0.79–0.96, p = 0.005) and treatment onset (OR = 1.16; 95% CI 1.06–1.27, p = 0.002), being the latter the only associated in the multivariate analysis (OR = 1.18; 95% CI 1.05–1.32, p = 0.005). 0.6% of our patients died."
Thank you. Just one small point. In an example you used the terms vegetarian and herbivore, as if the are interchangeable. As you know, vegetarian is a social construct which is an individual choice. Herbivore is a biological designation that is applied (species-wide) to a given species.
Noting from the chart, sometimes coffee is good, sometimes bad:
In the 90's/00's, I saw 2 conflicting studies. One said coffee was good, the other bad. Then a third study that noticed that bad coffee was unfiltered, like espresso or percolator. "Good" coffee was paper filtered, like typical drip. So the 3rd confirmed the results of the first two.
Even years ago, me being just an engineer who took some statistics as an undergrad, I learned that studies must be studied, not just headlined.
After an admittedly scanning read of this article, I must suggests that similar kinds of invalid studies have utterly degraded the entire field of "climate change" . People believe what they want to believe regardless of actual good scientific data.
Vinay Prasad, you are a treasured expert in deconstructing studies, meticulously exposing the bad apples. Why would you put your name to this article reinforcing a "preponderance of the evidence" approach to all studies regardless of individual quality? There is considerable concern that industry is flooding the sphere with questionable studies in order to overwhelm good studies critical of the respective product with noise
On the side of negative BS studies, we have seen studies looking at the impact of vitamin D supplementation on healthy, vitamin D-sufficient populations, showing no benefit.
Over the last 70+ years of my life, food fads have come and gone. Studies have made claims that certain foods are bad or good for you, only to decide that it wasn’t bad or good after all. Eggs, butter, margarine, meats... are just a few. Why was the question. With enough research money and motive you can prove or disprove most studies. Critical thinking is seriously needed to navigate the maze.
I normally shudder inwardly at the mention of the word statistics, thank you for making it easy to follow. The chart is very good.
I think people generally have a strong feeling that all the reports can't be right but they don't have the time or the skills to do what you are doing and try to find which are based on solid evidence.
Even if we do find it, propaganda, advertising and marketing in supermarkets is hard to overcome.
I am optimistic that the last two years will have opened up the discussion around food/diet to a far wider audience.
Makes sense. Basically there are the pre-existing biases against things like red meat and processed meat, and in favor of things like vegetables. And then the plausibility and what seems interesting / sexy. And then the journals. All these different layers of what actually makes it to publication, and then there's the media coverage on top of that witch is sort of like another layer of the same biases. And in all of that making no mention of the inaccurate food frequency questionnaires and healthy user bias. It's a minefield out there, but pretty easy to see why you keep seeing articles about how terrible bacon is and how amazing blueberries are, for example. Most individual food ingredients are not that powerful one way or the other (though there are obviously some exceptions).
Interesting article. For the most part I agree with the conclusions. However, my takeaway is what I notice missing most from study design, is the lack of instructive distinctions in the host population. We are not all the same. Precision Medicine was supposed to address this by focusing more on the individual, but I think that has turned into window dressing and a way for corporate medicine to pretend they are actually following some of functional medicine principles efforts to find root causes, when in reality they are not. Perhaps it is NOT so much the input, but rather the health of the host which protects the body from various insults along the way. I'm not just referring to age, gender and ethnicity host variables, though those are important. I would like to see study distinctions such as microbial diversity in the gut microbiome, C section vs vaginal, breast-fed vs not breast-fed and for how long, vaccinated and unvaccinated, delayed vaccination, selective vaccination, age at first vaccination, autoimmunity or cancer in the immediate family as study variables.
I have seen a logical error occur repeatedly by all kinds of people, where the vaccinated are compared against covid, as if the two arms are somehow exclusive. Let me show a hypothetical example of what happens when this error is made.
Let us suppose that 5 people who are vaccinated suffer Harm, but 10 people who contract covid suffer harm. Comparing the vaccinated group with the covid group, we would conclude a relative harm of 0.5 for the vaccinated, group, which would be an actual relative benefit. However, in this situation, 5 of those in the covid arm are also vaccinated and the vaccinated are actually counted in each arm. If we compare the unvaccinated who got covid with the unvaccinated who got covid, then the relative harm is 1.0, which means that there is no benefit from vaccination as regards the Harm.
Here's data that would be helpful to examine for churnalism.
"According to data from Techniker Krankenkasse, the largest German medical insurance company, there were a total of 437,593 insurance claims billed under the four diagnostic codes for vaccine injury in 2021. To put those numbers in perspective, the total numbers billed for a vaccine injury code in the two preceding years was 13,777 and 15,044, respectively. As the Daily Skeptic notes, given that TK insures 11 million people, that means 1 in 23, or 4.3%, had a medical treatment billed for vaccine injury. And that assumes all 11 million were vaccinated. The background vaccination rate in Germany is 78%, although most of the unvaccinated are children, so the rate of injury per vaccinated person is likely even higher (5.1%)."
https://www.conservativereview.com/horowitz-german-insurance-claims-vaccine-injury-2657863726.html
This also ties in with one of the comments by Vinu about vaccine injury. Hopefully there will be somebody willing and able to ask pertinent questions about the data and perhaps give a more nuanced perspective. If not, then maybe this is a smoking gun.
"what makes churnalism possible is shoddy research."
The following study makes HCQ early treatment for covid look like a parachute. Neither age nor the presence of comorbidities were the #1 factor correlating to mortality. The primary factor was failure to treat with HCQ within 72 hours of the onset of symptoms. Is the study churnalistic? Are there serious problems with it? Maybe there are problems with the statistics? If so, please discuss.
"Hydroxychloroquine / azithromycin in COVID-19: The association between time to treatment and case fatality rate"
"Results
A total of 1265 COVID-19 patients with an average age of 44.5 years were studied. Women represented 50.1% of patients, with an overall 5.9 symptom days, SpO2 97%, temperature of 37.3 °C, 41% with at least one comorbidity and 96.1% one symptom or sign. No patient treated within the first 72 h of illness died. The factors associated with higher case fatality rate were age (OR = 1.06; 95% CI 1.01–1.11, p = 0.021), SpO2 (OR = 0.87; 95% CI 0.79–0.96, p = 0.005) and treatment onset (OR = 1.16; 95% CI 1.06–1.27, p = 0.002), being the latter the only associated in the multivariate analysis (OR = 1.18; 95% CI 1.05–1.32, p = 0.005). 0.6% of our patients died."
https://www.sciencedirect.com/science/article/pii/S1477893921002040
Thank you. Just one small point. In an example you used the terms vegetarian and herbivore, as if the are interchangeable. As you know, vegetarian is a social construct which is an individual choice. Herbivore is a biological designation that is applied (species-wide) to a given species.
Here's my favorite dodgy study:
"Adverse events following mRNA SARS-CoV-2 vaccination among U.S. nursing home residents"
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8162901/
How did it get published?
Noting from the chart, sometimes coffee is good, sometimes bad:
In the 90's/00's, I saw 2 conflicting studies. One said coffee was good, the other bad. Then a third study that noticed that bad coffee was unfiltered, like espresso or percolator. "Good" coffee was paper filtered, like typical drip. So the 3rd confirmed the results of the first two.
Even years ago, me being just an engineer who took some statistics as an undergrad, I learned that studies must be studied, not just headlined.
After an admittedly scanning read of this article, I must suggests that similar kinds of invalid studies have utterly degraded the entire field of "climate change" . People believe what they want to believe regardless of actual good scientific data.
Vinay Prasad, you are a treasured expert in deconstructing studies, meticulously exposing the bad apples. Why would you put your name to this article reinforcing a "preponderance of the evidence" approach to all studies regardless of individual quality? There is considerable concern that industry is flooding the sphere with questionable studies in order to overwhelm good studies critical of the respective product with noise
On the side of negative BS studies, we have seen studies looking at the impact of vitamin D supplementation on healthy, vitamin D-sufficient populations, showing no benefit.
Sebastian Rushworth analyzed VD supplementation at https://sebastianrushworth.com/2020/08/03/do-vitamin-d-supplements-protect-against-respiratory-infections/
I thought the secondhand smoke thing has since been found to be wildly overstated?
Unrelated, but I forgot to mention you in my Twitter thread, and I think that everybody could be interested:
https://twitter.com/daniel_corcos/status/1559493023217717249
....which is why we need more journals like the Journal of Articles in Support of the Null Hypothesis.
Pre-registration of studies also helps.
Over the last 70+ years of my life, food fads have come and gone. Studies have made claims that certain foods are bad or good for you, only to decide that it wasn’t bad or good after all. Eggs, butter, margarine, meats... are just a few. Why was the question. With enough research money and motive you can prove or disprove most studies. Critical thinking is seriously needed to navigate the maze.
I normally shudder inwardly at the mention of the word statistics, thank you for making it easy to follow. The chart is very good.
I think people generally have a strong feeling that all the reports can't be right but they don't have the time or the skills to do what you are doing and try to find which are based on solid evidence.
Even if we do find it, propaganda, advertising and marketing in supermarkets is hard to overcome.
I am optimistic that the last two years will have opened up the discussion around food/diet to a far wider audience.