I wonder if it's BMI per se, or the lack of exercise? While the two are correlated, they are not the same - you can be fat and muscular at the same time. And if it's just BMI, then through what mechanism it increases the risk.
"fat and muscular" still means you're fat. Carrying excess fat is indicative of having metabolic syndrome and the many issues associated with that (chronic inflammation, insulin resistance, etc).
(I'm ignoring the fact literal BMI is a poor measure and BF% should be used instead to properly account for the vast variety of body shapes and sizes which have healthy BF% but don't align with the very out-of-date BMI charts & studies.)
BF% is hard to measure. Different manufacturers use different models to map resistivity data to BF%, so for the same person one can get rather different answers.
Resistance based measurements are pretty poor especially with only 2-4 points of contact, the accurate ones are displacement based like bodypod and X-ray based.
I’m actually surprised that this isn’t something that people have thrown ML at yet, it seems like a problem that an IR camera with 360 degree body imaging and a good model might actually be able to solve.
At the minimum, you need a friend to use calipers on you, and take those measurements over time to see trending and averages.
Even better is a dunk tank or Dexascan to get actual numbers from direct readings, most cities/universities have such facilities but not as accessible to everyone. In a perfect world it would be part of your annual physical, but they still use BMI charts :(
Muscle mass alone isn't enough. Obese people have far less aerobic headroom (lower VO2_max per kg), so viral pneumonia that decreases their lung efficiency even further is harder to survive.
There are many factors:
"Fat in the abdomen pushes up on the diaphragm, causing that large muscle, which lies below the chest cavity, to impinge on the lungs and restrict airflow."
"the blood of people with obesity has an increased tendency to clot—an especially grave risk during an infection that, when severe, independently peppers the small vessels of the lungs with clots"
"Immunity also weakens in people with obesity, in part because fat cells infiltrate the organs where immune cells are produced and stored"
"people with obesity also suffer from chronic, low-grade inflammation... Those effects may compound the runaway cytokine activity that characterizes severe COVID-19"
There are some exceptions like professional bodybuilders (often with the aid of steroids), but we're talking about reasons why mortality rates are linked to BMI, not declaring absolutes.
I agree with you, just want to say I think the reason skywhooper is bringing it up is "you get what you measure" and that hard rules about BMI risks harming and disincentivizing the body building community.
You’d have to design a prospective study to find out because most patients are going to lack good data on exercise frequency and type, whereas BMI is directly available from almost every medical record. Therefore you’d need to be asking for the information you want.
This is a better question than it may seem. Research suggesting lack of muscle may matter more than presence of fat for insulin resistance, among other things:
As noted by others, BMI isn't strictly synonymous with fat. The military tape tests body builders who can't meet weight to prove they aren't actually fat. When my husband was a military recruiter, every year some civilian organization told them they were "overweight." They all had to pass PT tests and meet weight to remain in the military. Muscle weighs more than fat.
"The coronavirus disease 2019 (COVID-19) pandemic presents an unprecedented health crisis to the entire world."
"Unprecedented"?
Have you heard of the 1918-19 flupandemic? Or even the 1957 or 1968 influenza pandemics? I'm not saying they're exactly the same, but the last two are in the same range, and actually the first one was way worse.
"Unusual", sure. "Worst in half-a-century", maybe, although some might argue HIV did more damage (albeit over a longer time period). "Unprecedented" means, "Never before seen, done, or experienced; without precedent". While every pandemic is different in some way, this is not 'unprecedented'.
Unfortunately this is also what the current administration in the US has been saying. That it would have been impossible to predict that pandemic and that no one could have seen it coming. Despite many warnings and the establishment of a pandemic response team because this was seen coming. (Of course, not specifically COVID-19 at the specific time it did, but there were plenty of warning signs.)
Probably because we already had an influenza vaccine which was developed in the 1940s. We have never had an approved vaccine for a corona virus. Thankfully there has been a lot of work on SARS and new vaccine technology for Ebola and Zika etc that could be applied to SARS-CoV-2 so we didn’t have to start from nothing this time.
Were they? I haven't seen any history on that one way or the other [checks wikipedia]...oh, they were! My guess is that it was because we had an established infrastructure for making influenza vaccines in particular. Even then, the vaccine didn't appear in any quantity until almost the end of 1957, and the influenza was first seen in 1956.
"unprecedented" in that it was rapidly distributed by modern travel before quarantine could be established and it spread across the EU, Asia, Oceania simultaneously.
I know my country, Sweden has a chance to create a quarantine, but chose not to. Because the medical leader was foolishly confident in contact tracing, which fell apart almost immediately.
This is somewhat confounded by the fact that it was suggested early on that people with BMI>30 would be de-prioritised for treatment/ICU beds if hospitals became overwhelmed.
No. Triage is supposed to prioritize those for whom treatment is most likely to make the difference between surviving and not surviving, which is not at all the same as prioritize those most likely to survive.
If you say "I won't treat fat people (the same) because they are less likely to survive",
and then afterward say "Fat people were less likely to survive (so this earlier decision was justified)", then that's severely faulty and dangerous logic.
The reason fat people had worse outcomes is, just as likely, because they had worse treatment.
Ah, I see what you are saying. Possibly, though early on the people that had the worst outcomes were those that were aggressively treated with ventilators.
It was found that there was a linear relationship
(Pnon-linearity = 0.242) between BMI and the risk
of critical COVID-19
They seem to mean that the log odds ratio is a linear function of BMI. This just can't be true. Sufficiently low BMI has to be bad, not good.
Figures 7 and 8 in the paper actually show a near-constant risk from BMI of 18 to BMI of about 27. The good results for BMI of 18, which is pretty definitely underweight, actually seem rather dubious, on the basis of past studies (see my blog post at https://radfordneal.wordpress.com/2020/04/06/body-mass-and-r...). But in any case, there is clearly no reason to think that risk goes down linearly as BMI gets smaller than about 27, and there's little reason to think it goes up linearly above that either, considering the very small number of data points.
Headline is wrong here. The paper only asserts an “association” and the abstract mentions a correlation. This headline incorrectly asserts a direct cause-effect relationship.
39 comments
[ 3.2 ms ] story [ 97.2 ms ] thread(I'm ignoring the fact literal BMI is a poor measure and BF% should be used instead to properly account for the vast variety of body shapes and sizes which have healthy BF% but don't align with the very out-of-date BMI charts & studies.)
I’m actually surprised that this isn’t something that people have thrown ML at yet, it seems like a problem that an IR camera with 360 degree body imaging and a good model might actually be able to solve.
At the minimum, you need a friend to use calipers on you, and take those measurements over time to see trending and averages.
Even better is a dunk tank or Dexascan to get actual numbers from direct readings, most cities/universities have such facilities but not as accessible to everyone. In a perfect world it would be part of your annual physical, but they still use BMI charts :(
There are many factors:
"Fat in the abdomen pushes up on the diaphragm, causing that large muscle, which lies below the chest cavity, to impinge on the lungs and restrict airflow."
"the blood of people with obesity has an increased tendency to clot—an especially grave risk during an infection that, when severe, independently peppers the small vessels of the lungs with clots"
"Immunity also weakens in people with obesity, in part because fat cells infiltrate the organs where immune cells are produced and stored"
"people with obesity also suffer from chronic, low-grade inflammation... Those effects may compound the runaway cytokine activity that characterizes severe COVID-19"
https://www.sciencemag.org/news/2020/09/why-covid-19-more-de...
There are some exceptions like professional bodybuilders (often with the aid of steroids), but we're talking about reasons why mortality rates are linked to BMI, not declaring absolutes.
https://news.ycombinator.com/item?id=8748147
As noted by others, BMI isn't strictly synonymous with fat. The military tape tests body builders who can't meet weight to prove they aren't actually fat. When my husband was a military recruiter, every year some civilian organization told them they were "overweight." They all had to pass PT tests and meet weight to remain in the military. Muscle weighs more than fat.
"Unprecedented"? Have you heard of the 1918-19 flupandemic? Or even the 1957 or 1968 influenza pandemics? I'm not saying they're exactly the same, but the last two are in the same range, and actually the first one was way worse.
"Unusual", sure. "Worst in half-a-century", maybe, although some might argue HIV did more damage (albeit over a longer time period). "Unprecedented" means, "Never before seen, done, or experienced; without precedent". While every pandemic is different in some way, this is not 'unprecedented'.
Yes we did Donald. Yes we did.
The spanish flu was even more severe since its lethality followed a W-shape instead of predominately affecting old people.
https://en.wikipedia.org/wiki/Triage
If you say "I won't treat fat people (the same) because they are less likely to survive",
and then afterward say "Fat people were less likely to survive (so this earlier decision was justified)", then that's severely faulty and dangerous logic.
The reason fat people had worse outcomes is, just as likely, because they had worse treatment.
Figures 7 and 8 in the paper actually show a near-constant risk from BMI of 18 to BMI of about 27. The good results for BMI of 18, which is pretty definitely underweight, actually seem rather dubious, on the basis of past studies (see my blog post at https://radfordneal.wordpress.com/2020/04/06/body-mass-and-r...). But in any case, there is clearly no reason to think that risk goes down linearly as BMI gets smaller than about 27, and there's little reason to think it goes up linearly above that either, considering the very small number of data points.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7283831/
The numbers (n,m) were small in both tables.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7430889/