long term covid effects: commentary on yelin et al. (2020)

On this blog, we’ve had a debate over my opinions on COVID policy. To summarize, here is my view:

  1. Most of the damage from COVID is disproportionately shouldered by the elderly.
  2. Harm from COVID is relatively rare among non-elderly adults and young people.
  3. It is possible to reduce transmission to elderly people with relatively simple policies like age screening, increased testng for care givers and encouraging the elderly to reduce contacts with younger people who have been out in the community.
  4. Widespread lockdowns and closure have some very damaging effects such as unemployment, mental health problems, and erosion of the community.

Points #1 and #2 are very much the consensus. Most people seem to accept #4. #3 is more contended – but I am puzzled about why focusing on at risk people is seen as such a misguided policy. It’s what we do for other illnesses (e.g., if your family doesn’t have a history of sickle cell anemia, we won’t monitor you for it). The conclusion , however, is pretty simple: a relatively “open” policy – relase the young, protect the elderly. There are a growing number of epidemiologists who are moving toward this position. Even the WHO is backing off on lockdowns as a default method for COVID mitigation.

In the comments of my blog posts on the “open” position, people told me I am severly misunderstanding things or misrepresenting things. For example, people will link to a news article reporting that a young person died from COVID. This evidence is not what is needed to reverse my position. I have never said that young people are immune from COVID or that they don’t get infected. Rather, my point – which is the consensus – is that the damage from COVID is much, much more rare among the young than the elderly. Thus, if you point to an article about a single college student succumbing to COVID, that doesn’t undermine the major point. Policy is about larger trends, not rare cases. Also, news about infection rates should not change the position because infections by themselves are not the issue. Hospitalization and death is the issue. I also note here that I focus on death rates because infection rates are tough to interpret – an increase in public funding for test may increase *reporteed* cases, not actual cases.

Now, I want to turn to an issue that people brought up. People claimed that my “open” position is untenable because of the possibility of “long haul” COVID effects. In other words, COVID may have lingering and highly damaging effects on people and I don’t factor that in. Remember, when dealing with public policy, you are usually trying to deal with larger trends and allocating resources to people who need it. A small handful of cases will not change the argument. Thus, my argument about the “open” position for COVID stands if the people who get long term effects are the same people who were initially hospitalized or died – the elderly, mainly. If people who do not initially show symptoms, but are infected, still end up getting long terms effects, then that’s a chance to revise the police.

In arguing for a more “closed” position because of “long haul” effects, one commenter linked to this article in Lancet: Infectious Disease by Yelin et al., as a way of criticizing my point. So remember: if long haul effects are mainly concentrated among people who showed symptoms early on, then that reinforces my point becuase long haul effects would be an issue affecting a very specific group that we can protect (the elderly). If long haul effects appear in asymptomatic people (mainly younger people), then it would undermine my point because the assumptions behind the trade-off involved in an “open” position are incorrect.

So here is what the article says. Let’s start with the first sentence:

Weeks and months after the onset of acute COVID-19, people continue to suffer.

Right there, this article is about people who showed acute symptoms from the start. It’s about comorbidity conditioned on breathing problems and other severe symptoms. How can we tell? Here is the data that motivates the article: “78 of 100 patients in an observational cohort study who had recovered from COVID-19 had abnormal findings on cardiovascular MRI (median of 71 days after diagnosis) and 36 of those reported dyspnoea and unusual fatigue.” Severe breathing problems. Who were these 100 patients? A recent JAMA article (Pentmann et al. 2020) on *hospitalized* COVID patients. How sick were they? “This is a prospective observational cohort study of 100 patients diagnosed with severe acute respiratory syndrome coronavirus 2 by reverse transcription–polymerase chain reaction on swab test of the upper respiratory tract who fulfilled inclusion criteria for this CMR investigation.”

Let’s look at Table 1. in Pentmann et al. (2020). This table describes the 100 people who had severe COVID induced breathing problems. Look at the 1st column. What do we see? Average age 49. 18% had diabetes. 22% had hypertension. BMI? 25 vs. 23 in the control group. Blood pressure? 129 v. 122 for the control group. Pentmann et al. also created a group of 57 people who were risk matched, in order to compare equally ill people who did and did not have *acute* COVID.

Not suprisingly, the group of more obese, more diabetic and older acute COVID patients have more cardiac problems post-COVID than people who did not have acute COVID. Also, not surprisingly, COVID victims compared to “risk matched” non-COVID individuals did worse. But Yelin at al. (2020) put this in the proper perspective. They do not estimate the population level frequency of long haul effects among acute COVID victims, nor do they say anything about their appearance in otherwise low risk population.

Rather, they correctly, note that many infectious diseases can have catastrophic long terms effects: “Rare long-term sequelae can result after other viral infections—eg, infectious mononucleosis, measles, and hepatitis B.” They also note that since COVID is novel, we’ll need to conduct more research on what such long term effects are:

  1. “The number of people affected by COVID-19 is unprecedented. We owe good answers on the long-term consequences of the disease to our patients and health-care providers. The obvious answer is in research.”
  2. “Careful recording of symptoms and patient examination should allow understanding of which part of the sequelae is common to all severe infections, which symptoms might be explained by the anxiety caused by a new disease and by the isolation,9 and which symptoms are secondary to a complicated form of COVID-19 (eg, pulmonary involvement during the acute disease).”

Yelin et al. 2020 have a simple and important message: if you have acute COVID, you may have lingering side effects, just like other infectious diseases. Also, we need to figure out which problems are due to infectious disease generally, anxiety, and isolation. Notice, that nowhere in this article do Yelin et al. fundamentally take issue with the disease burden. In fact, the big study they describe (and others in the references) confirms what we know about COVID – it’s harms are way, way more focused on the elderly, and that’s the group we should help.

50+ chapters of grad skool advice goodness: Grad Skool Rulz ($4.44 – cheap!!!!)
Intro to sociology for just $1 per chapter – INSANE BARGAIN!!!!!
A theory book you can understand!!! Theory for the Working Sociologist (discount code: ROJAS – 30% off!!)
The rise of Black Studies:  From Black Power to Black Studies 
Did Obama tank the antiwar movement? Party in the Street
Read Contexts Magazine– It’s Awesome!!!!


Written by fabiorojas

October 19, 2020 at 8:13 pm

Posted in uncategorized

15 Responses

Subscribe to comments with RSS.

  1. OMG, Did you actually endorse the Great Barringnton Declaraton?! The authors of that document have been widely critiqued and in many cases discredited. It’s less epidemiological consensus and more of a libertarian economic policy promotion grounded on the flimsiest of expertise. Here is a “counter-declaration” from people who actually know stuff:

    I don’t even know how you can keep pushing for your position, since we see real world evidence that it is impossible to isolate the old and the vulnerable. And I have know idea what’s with the straw-man argument about opposition wanting widespread lockdowns. Most people who actually know stuff are pushing for ramped up testing and tracing that allows for careful and “dynamic” closing and reopening. You are pretty much making the same arguments as Trump, who claims that his opponents want widespread lockdowns. You must be very proud to have arrived there.

    Liked by 2 people


    October 19, 2020 at 8:37 pm

  2. There are other studies of Long COVID that focus on how brain fog and other symptoms are hurting young healthy people who had mild cases of COVID, and how the brain damage among previously healthy people might be permanent.

    There was even a NYT article about some of it. But there has since been other research. You should do a better lit review before writing about this publicly.



    October 19, 2020 at 9:39 pm

  3. Catallus:

    1. The NY Times article documents that brain fog happens to some people. I have never disputed side effects. For example, my aunt – hospitalized with COVID – has this condition. Nor do I dispute that it happens to some young people. The issue is how often. Does the epidemiology of brain fog look different than acute COVID? The NY Times article does not offer an answer to that question. If you can find an article that shows wide spread brain fog in a large population of young people with asymptomatic COVID or very mild COVID, please post it.

    2. A previous commenter claimed the Lancet ID article undermined my basic point about the age burden of COVID. It does not. My blog post focused on that claim. Also, check out the references for Yelin et al 2020. Article. Lots of citations to long term effect research and I didn’t see any that clearly challenges my point about the epidemiology of COVID. Feel free to link to research that shows that 1. Acute COVID is not strongly associated with age or 2. There is a separate epidemic of long term COVID effects among non elderly populations with mild or asymptomatic covid. Happy to discuss it.



    October 19, 2020 at 10:06 pm

  4. Fabio, you continue to show you are completely out of your element each time you opine about COVID. It should be embarrassing for you. But you’ve consistently shown Trumpian-like limited capacity for self reflection. Look no further than your linking to the Great Barrington Declaration by claiming that “There are a growing number of epidemiologists who are moving toward this position,” but failing to recognize that no more than a few people who’ve signed that travesty actually have a background in epidemiology. What’s next, are you going to point to Eric Feigl-DIng as the foremost expert on COVID-19?



    October 19, 2020 at 10:26 pm

  5. Here is a nice visual for you: Who knew that a defender of open borders would eventually end up agreeing with Scott Atlas….



    October 20, 2020 at 8:16 pm

  6. Puzzled: I am puzzled right back at you! My claim has always been about relative risk, not absolute numbers. If a disease is very frequent in one population, but infrequent in another, then targeted interventions make sense. The size of the problem does not shift that basic argument.

    The graphic you linked to correctly notes that we have about 300k excess deaths. If we use the 57 and 68 flus as a guide, this is horrible, but not unexpected. The graphic also notes that we have increases in mortality in young people and minorities, which is also horrible, but expected if we have a new contagion.

    However, the graphic does not provide information on fatality rates by age group, which is the core of the argument. It only says that it increased within one group compared to 2019. Let me quote it for you: “From late January to October 2020, the U.S. had 299,000 more deaths than the typical number in previous years… The largest percentage increases were among people who were Hispanic or Latino and adults aged 24-55.”

    Nowhere, does the infrographic say that fatality after infection for people aged 24-55 is the same, or even the same magnitude, as for 65+. Because COVID has increased mortality among young people, it does not imply that mortality conditional on infection (the “IFR” rate) is not still way, way higher for old people than younger people. It is.

    Please go here ( and check out Table 1, row 1. This is not a media infrographic, it’s the CDC report on COVID mortality. Now, please compute the ratio of death for elderly vs. kids (.028 / .00002) = 1,000 (!!!), elderly vs. adults (.028 / .00007 ) = 400. elderly vs. middle age (50-69) = (.028/ .0025) = 11.2. Also, let me note the qualitative difference in IFR. Soldiers in Iraq had a death rate was 335/100,000 = .003 (see this article in Military Medicine – page 222 Table I). That is actually lower than the IFR for the elderly!!

    Let me respond to a chartiable version of your criticism. It may be the case that the absolute size of COVID requires an intervention different than what I suggested (i.e., age graded mitigation). It’s just so horrible that you have to ignore the hetereogeneous impact of the problem. That’s fair, but then you need to follow up: What is the cost of the intervention? Would the risk level of the problem be acceptable without intervention?

    In response to the first, we know the answer: mass unemployment, social isolation and dpression. Second, people would not accept a fatality rate that elderly COVID patients have, but people seem tolerate other reasonably dangerous events and activities, such as the 57 and 68 flus, auto fatlities,fatlities and disease due to alcohol consumptins and the tens of thousands of people who die every year from hospital acquired infections. Thus, while COVID among young people is horrible, shutting down society also has very big side effects, which many might find acceptable given what we know about how people tolerate risk in other circumstances.



    October 20, 2020 at 9:39 pm

  7. What I find fascinating and puzzling about your ongoing crusade for protecting the old and releasing the young, which only the widely discredited Barrington Declaration seems to support while you have not presented any experts who supposedly argue for “shutting down society”. You are arguing against straw-man. No one is advocating for the point you are trying to refute. Most experts are suggesting 1) sophisticated test-trace-isolate operations and 2) targeted and short-term shutdowns of high-risk places. So the mainstream experts are actually trying to avoid the large-scale societal shutdowns you seem to disparage.

    It is very hard for you to sustain your position when you point to GBD as evidence that: ” There are a growing number of epidemiologists who are moving toward this position.” Their main accomplishment is having gotten the ear of Atlas and Trump, but I don’t think any serious epidemiologists would acknowledge that crew as evidence of mainstream thinking on disease control.

    So your posts on open borders, on org theory on mentoring Ph.D. students have been insightful. But with this crusade against a position that no one takes, and by making arguments that are only supported most cynically by people who have no serious interest in getting the pandemic under control, you are not helping anyone.


    Still extremely puzzled

    October 20, 2020 at 10:10 pm

  8. Still Extremely Puzzled: If I got you on 2/3 of my issues, I’ll happily take it!! But I do not think I am straw manning. For example, literally today, Washtenaw County in Michigan (the county for Univ of Michigan), issued a stay at 2 week home order. Many colleges (not yours?) have moved to remote instruction, as have many K-12 schools. In the education sector, “stay at home” or “remote learning” is the norm in many places, not “come back, but sheild the elderly.” I am fortunate in that my school district is allowing a wise mix – most kids came back to school but some are remote if they have elderly folks at home… and after two months, minimal health issues.

    But there are lots of places where more stringent measures are being instituted. For example, in Europe, some nations are rolling out lockdowns again. Since we are tossing around NY Times articles, here is an article on how Ireland is re-imposing lockdowns. Go ahead – click on it: . I may be wrong in arguing that “help the elderly, release the young” is a good policy, but we have seen lockdowns in the Spring and now they are appearing in Europe again. This is not a fake issue, it happened and it’s happening again in some places.



    October 20, 2020 at 10:22 pm

  9. Note that for Ireland: “The government of Ireland announced a six-week lockdown beginning Wednesday night”. Six weeks. Not nothing but not long-term either. Because they know that having very low case numbers is key to resuming any sense of normalcy for the society – both young and old. This is what we see in China now as well as other countries that have done well, like New Zealand, parts of Canada (Atlantic Provinces, definitely not Quebec or Ontario), Taiwan, South Korea. Definitely not Sweden, whose approach is closer to what you are advocating, which has suffered as much economic harm as its neighbors with 7 times the deaths of its neighbors.

    Shutdowns, like in Michigan or Ireland are not the option that experts encourage. It is the thing that happens when the normal test-trace-isolate approach is not pursued (e.g., the US) or does not succeed.

    So yes, I think you are straw-manning it. No one advocates for the widespread shutdowns. It is what happens when the consensus view is not followed, and the society decides that maybe killing millions of people is not a good idea after all. So pursuing your strategy or the more callous herd immunity without vaccine is what gets us shutdowns. Shutdowns are not what experts want. They are an acknowledgement that whatever we are doing is not working and we are out of ideas.


    Still Puzzled

    October 20, 2020 at 10:44 pm

  10. I honestly don’t think I am straw manning if (a) people are actually calling for shut downs in actual countries – and six weeks is not trivial!, (b) data on the stringency for COVID measures shows that many regions and nations have maintained strict measures (see the Oxford data – on their scale, the US peaked at 8/10 in stringency but has declined to a mere 6/10 and some nations are increasing – and (c) I haven’t seen a lot of age graded mitigation. If on a standard scale of stringency, your country is hovering between 6/10 and 8/10, I would say that leans towards my point. If you don’t think so, we are at an impasse and stuck at definitions. Still, I thank you for your debate and discussion.



    October 20, 2020 at 10:59 pm

  11. Very many of those who have long term effects from COVID were never hospitalized. Some don’t even know about the long-term effects–I’ve seen studies of athletes who had few symptoms but ended up with myocarditis, and an analysis suggesting that scuba divers who get COVID may be symptom-free but never able to dive again due to the small but significant decline in lung function. Your refusal to acknowledge that anything other than hospitalization and death matter–your refusal to believe that long term disability is a real thing–is either remarkably clueless or willfully obtuse.

    My mother has a friend who lost nine close relatives. A friend of mine’s husband in decent if not perfect health has not been able to work in 5 months (and lost his job and health insurance coverage). My city councilor lost a cousin who was young, healthy, and pregnant. Quantitative data is not telling the whole story here. You need to listen to the human cost. And yes, a closed economy also has costs–but we could have made different choices, keeping bars and colleges closed while opening elementary schools and getting more people back to work. We made the wrong choices. People are continuing to die and be permanently disabled.

    And, quite frankly, I am tired of hearing healthy people say it’s okay to sacrifice those who do not have perfect health. Lives are not less valuable because they are lived in overweight bodies or with diabetes or asthma. Solving the problem we face does not occur by opening everything up and watching our loved ones die. It occurs by investing in universal basic income, protecting people from eviction, investing in public education at the K-12 and college levels, ensuring that everyone has access to health care, providing bridge funding for small businesses, I could go on. I want to live in a world in which we care about people and provide them with what they need to thrive, not one in which people like you say we should sacrifice the overweight because it would make your life more convenient.



    October 21, 2020 at 3:31 am

  12. Also this: (on excess mortality among young people)



    October 21, 2020 at 4:17 pm

  13. Fabio, you continue to embarrass yourself. As others point out, you advocate a position that basically no one holds, except in the fevered imagination of MAGA-land.

    Protecting the vulnerable and letting the less vulnerable get on with their lives is the goal everyone shares. It is impossible to do so without a robust public health approach of testing-tracing-quarantine which does not exist in this country. It is also impossible to do when the vulnerable and less vulnerable, by reality and necessity, mix frequently in multi-generational households and communities. The only realistic solution is a version of what New Zealand and others have tried: go hard and early. In other words, targeted shutdowns when cases rise.

    Your views are irresponsible. If adopted, they would kill many needlessly. The fact that you can’t see that impugns your judgement more broadly. I am embarrassed that you are the incoming chair of the Political Sociology Section. Perhaps that should not be the case.

    Liked by 1 person


    October 22, 2020 at 3:58 am

  14. Trump tonight, about 19.30 minutes in: “…take a look at North Carolina. They’re having spikes. And they’ve been closed. And they’re getting killed financially. We can’t let that happen, Joe. You can’t let that happen. We have to open up, and we understand the disease. We have to protect our seniors, we have to protect our elderly. We have to protect especially our seniors with heart problems and diabetes problems and we will protect them. We have the best testing in the world by far. That’s why we have so many cases.”

    Congrats, Fabio. This is the company you keep.



    October 23, 2020 at 3:26 am

  15. I remember that the author of this post also declared that the US was in a post-racist society probably when this blog just started. So why are you all surprised :)?



    October 27, 2020 at 5:20 pm

Comments are closed.

%d bloggers like this: