Quote:
Originally Posted by S=kBlogW
Millions of Americans have had covid, only a small percentage (5%?) of them get really sick, go to hospital and/or die. Of this cohort, 78% are overweight or obese.
|
Something like 70% of Americans are overweight or obese. If someone blindly picked 148,000 people they are likely to find that around 70% are overweight or obese. "Random" is never really random so the percentages would not be exactly the same. The question is whether 78% is higher in a statistically significant sense.
Looking at it from the other direction:
There are millions of Americans. Perhaps (I can't find the data) 30% of them fall into the the category of not overweight or obese. A study of about 148,000 patients found that only 22% of them fell into the category of not overweight or obese. Either the wording on the article was poor (distinct possibility) or the study found that the demographics of the hospitalized patients closely matched the demographics of Americans as if they were randomly picked off the street with no regard to weight.
Note that one of the limitations of the study was it could only consider patients where height and weight were listed. The data could be biased towards overweight or obese as the hospitals look for comorbidities. In other words, the hospital *might* list height and weight for overweight and obese patients more often than for patients who had a more "healthy" weight.