Substack cometh, and lo it is good. (Pricing)

The role of obesity in the COVID-19 crisis


There has been a fair amount of anecdotal and a bit of statistical evidence that obesity is somehow associated with individuals who have worse progression of COVID-19. The data out of China I saw wasn’t significant statistically speaking. The problem? There didn’t seem to be enough obese people in their samples. Then anecdotes and some data came out of Europe implicating obesity as a risk factor. And, doctors started reporting a disproportionate number of obese patients in the ICU.

Now we have really good evidence, Factors associated with hospitalization and critical illness among 4,103 patients with COVID-19 disease in New York City:

We conducted a cross-sectional analysis of all patients with laboratory-confirmed Covid-19 treated at a single academic health system in New York City between March 1, 2020 and April 2, 2020, with follow up through April 7, 2020. Primary outcomes were hospitalization and critical illness (intensive care, mechanical ventilation, hospice and/or death). We conducted multivariable logistic regression to identify risk factors for adverse outcomes, and maximum information gain decision tree classifications to identify key splitters….Strongest hospitalization risks were age ≥75 years (OR 66.8, 95% CI, 44.7-102.6), age 65-74 (OR 10.9, 95% CI, 8.35-14.34), BMI>40 (OR 6.2, 95% CI, 4.2-9.3), and heart failure (OR 4.3 95% CI, 1.9-11.2)…In the decision tree for admission, the most important features were age >65 and obesity; for critical illness, the most important was SpO2<88, followed by procalcitonin >0.5, troponin <0.1 (protective), age >64 and CRP>200. Conclusions: Age and comorbidities are powerful predictors of hospitalization; however, admission oxygen impairment and markers of inflammation are most strongly associated with critical illness.

click to enlarge

I’ve reformated table 3 of the regression below. It’s important to note here that the whole population is infected. The table is assessing the risk out of the infected sample that someone is going to go critical (which means a host of things, but entails hospitalization). Remember that a lot of the comorbidities associated with obesity are in the table. That means the risk of obesity is viewed as an independent variable. One can make some mechanistic arguments about the inflammatory effects of lipids, etc. That’s neither here nor there.

When assessing the risk of various nations is that 3% of Japanese are obese, while 40% of Americans are obese.

Table 3: Multivariable regression results, hospitalization of patients with COVID-19
     
CharacteristicNOdds ratio95% intervalp
Age
0-18534.96(2.75-8.98)<0.001
19-441501Reference
45-546942.57(2.06-3.2)<0.001
55-647674.17(3.35-5.2)<0.001
65-7457710.91(8.35-14.34)<0.001
≥7551166.79(44.73-102.62)<0.001
Cancer1851.24(0.81-1.93)0.329
Chronic kidney2153.07(1.78-5.52)<0.001
Coronary artery2350.88(0.57-1.4)0.59
Diabetes6142.81(2.14-3.72)<0.001
Male20722.8(2.38-3.3)<0.001
Heart failure1314.29(1.89-11.18)0.001
Hyperlipidemia7380.67(0.51-0.87)0.003
Hypertension9831.23(0.97-1.57)0.094
Obesity
BMI <303003Reference
BMI 30-409154.26(3.5-5.2)<0.001
BMI >401856.2(4.21-9.25)<0.001
Pulmonary disease3121.33(0.96-1.84)0.087
Race
White1812Reference
African American6550.88(0.69-1.11)0.28
Asian2841.44(1.04-1.98)0.026
Other/Multiracial9271.99(1.62-2.45)<0.001
Unknown4250.9(0.67-1.21)0.501
Tobacco (current/former)8780.71(0.57-0.87)0.001

16 thoughts on “The role of obesity in the COVID-19 crisis

  1. I’m not certain what I’m looking at it here. Does that mean if you have a BMI >40, you are . . . roughly six times more likely to be hospitalized?

  2. Does that mean if you have a BMI >40, you are . . . roughly six times more likely to be hospitalized?

    “all other variables controlled” basically. so it’s not a simple interpretation. you might want to look at the description tables which show proportions hospitalized vs. not based on BMI etc.

  3. How do they address the multicollinearity issue here? Obesity and other comorbities are related to each other so we might not be getting reliable estimates

  4. I doubt it is lipids; it is probably the increased expression of ACE-2 on epithelial cells due to obesity. But as you said you’re not interested in biology.

    The low tobacco risk is surprising to me; I would not be surprised if vaping/marijuana smoking are also major risk factors. Other studies are showing asthma is not a risk factor, although COPD is. I didn’t see that here.

    As I said earlier the NHAMES study seems to be best source on this. It would be nice if they could break down the weight by age range as well.

    There is a lot of stuff floating around that we are being too aggressive on the ventilators; doctors are seeing very low levels of peripheral oxygen but at the same time C02 is being cleared. Think altitude sickness, not ARDS.

    The British and european press are reporting large death tolls in nursing homes. Again we’ve know since January who are the people at risk, but someone we’ve convinced young families they have the most to lose.

    RXs are down almost 25%; people are not taking drugs, going to the doctor for checkups, heart pains, and manageable diseases. This is driving the bumps in deaths across the country — not viral deaths. Just as a reminder, about 7000 people die every day in the US.

  5. There was also another study in the last week that highlighted chronic kidney problems as a major risk factor. Cant find the link right now.

    only 878 current/former smokers in the entire cohort? that seems very low. Or 878 among those hospitalized?

    Would be good if they could break out hypertension vs people being treated with drugs; there is some interesting stuff there but again data limitations may be the problem.

  6. BMI is not the way to measure obesity. Body fat percentage is. If you’re a guy, you want your body fat percentage to be less than 15%. If you are a lady, you want it to be less than 22%. Body fat percentage is easily measured at any fitness gym.

  7. “We then fitted multivariable logistic regression models with admission and with critical illness as the outcomes to identify factors associated with those outcomes. We included all selected predictors based on a priori clinical significance after testing for collinearity using the variance inflation factor (VIF) and ensuring none had VIF>2”

    you can’t ever get rid of all colinearity i think. so take the results as important but don’t fixate on the exact number.

    an ER doc friend says that he is seeing young obese patients in ICU without any other official comorbities…. (inflammation due to obesity?)

  8. BMI is not the way to measure obesity

    i was going to add the addendum. this is true on the individual level. but obesity is fine to assess populations. you’ll get some false positives (muscular people) and false negatives (skinny fat). but that’s ok

  9. Since women in the U.S. are more likely to be obese (BMI >30): 35.0% of men and 40.4% of women, I find it difficult to understand the strong gender splits going the other way.

    My intuition is that higher BMI is much worse for men?

  10. Hmm… If obesity matters, one consequence of that is that any IFR / hospitalization estimated from Germany / Italy is probably going to be higher than we see in the developing Asia, just as an IFR calculated on their age structure would be high for the developing world.

    We’d need to adjust down from whatever the true IFR is identified in Germany/Italy by removing effect of obesity.

    (E.g. Germany+Italy about 22-20%, USA 36%, China 6.2%, Thailand 10%, India 3.9% – https://ourworldindata.org/obesity).

    From what I can remember, Waist-to-height ratio is probably superior to BMI (because BMI has a relationship where it tends to overstate at taller heights), and because of some other things to do with ethnic body composition, and probably WHtR with some fat measurement would be better still… but I don’t think you’ll go *that* far wrong with BMI.

  11. Diabetes and obesity for heavily comorbid, but obesity as a continuous trait is likely to catch severity of a diabetes cases that are not captured by a “yes” or “no” diabetes variable.

    Since they are so highly comorbid, without examining Type I diabetes cases which are probably too rare to be statistically significant, I’m not sure it is meaningful to say that one or the other is the “real cause” distinct from the other.

  12. @JoeQ States with high obesity in the U.S. also generally have fewer efforts and less compliance with efforts to reduce R for the spread of the virus (in both cases for cultural reasons). But, they also have (1) lower population density, on average, which helps, (2) warmer temperatures with higher humidities which may or may not help, and (3) weaker connections to the global economy through in bound physical contact, which helps.

  13. Nice little article on testing who socially distanced in the USA – https://medicalxpress.com/news/2020-04-people-didnt-social-distance.html. Young people comply less, older people comply more (and older people tend to be more relaxed and accepting of it).

    There’s also another finding that – “Residents in Republican counties are less likely to completely stay at home after a state order has been implemented relative to those in Democratic counties. We also find that Democrats are less likely to respond to a state-level order when it is issued by a Republican governor relative to one issued by a Democratic governor.”https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3569098

    So although you find less orders on social distancing in “red states”, the actual behaviour of not socially distancing may be that it’s being more driven by younger “blue voters”, within those states.

    Large, blue cities with relatively youthful, sociable populations, in red states seem like they’d then be expected to have a lot of prevalence, and then if they also have enough old you’d start seeing higher deaths. It seems like that’s also what you find with New Orleans as an example?

  14. Purely anecdotally, I am not seeing an obesity effect in the patients I am caring for. Being in the upper midwest, just about all my patients seem to be at least a bit overweight, so maybe that’s why. My sample is small too, so whatever.

Comments are closed.