Taking the lead on long COVID studies

How are researchers tackling long COVID?
18 October 2022

Interview with 

Leora Horwitz


A stylised coronavirus particle next to a woman wearing a facemask.


At the moment, it looks like about 1 person in 20 will get some kind of longer term syndrome in the aftermath of a dose of Covid-19, and 15% of the time, that might persist indefinitely. Women look like they’re more susceptible than men. So how are researchers trying to pursue this? Speaking with Chris Smith, Leora Horwitz is the person helping to lead the charge in the US…

Leora - My name's Leora Horwitz. I'm a general internist. I'm a professor of population health and medicine at NYU Langone Health in New York City. And, I am helping to lead the clinical science core for the NIHS Recover program, which is a program seeking to understand long COVID.

Chris - And have I read this right, that you've scored hundreds of millions in funding to look at this? Everyone wants to be your friend now!

Leora - Yes, but we do pass it all on to others! Congress has appropriated, a billion or so dollars to study long COVID and that includes doing observational studies, electronic health record studies and clinical trials. A portion of that has come to us here at NYU to distribute to hundreds of sites around the country that are doing observational studies of long COVID.

Chris - Do we agree yet, Leora, what long COVID actually is?

Leora - No, it's actually our first goal for the observational work and our foundational work. If we can't define long COVID, we can't do trials to see if we can make long COVID better.

Chris - So how are we defining it at the moment when we're trying to explore this entity? What are we actually thinking or saying we're studying?

Leora - Well, it's actually extremely difficult. So the WHO have made some stabs at definitions like having new symptoms or problems that started after COVID and that last a few months. And then others have taken much narrower definitions with specific symptoms. We are doing something in between, which is we are asking people about over 50 symptoms and whether they are new or different since COVID and using those, we will be able to come up with probably not one definition, but multiple definitions of long COVID.

Chris - So is it likely then that what we're dealing with is an entity that is an umbrella term, but it's united by the fact that everyone who's got these complex symptoms has been infected with COVID at some point. But what they've probably got is a range of different syndromes that they arrive at through different roots and possibly with totally different mechanisms and pathologies going on.

Leora - That's exactly what we think. Yes.

Chris - So how on earth are you gonna get a handle on this?

Leora - Well the Recovery study is studying adults, children and also people who have died to come up with some answers to the exactly those questions. What we do is we prospectively enroll people some at the time that they have their first infection and some after they've had an infection in the past. Every three months we ask them about these 50 plus symptoms so we can understand how those evolve over time. We ask them if they've had them before their infection and we also enroll people who've never had infection at all and ask them about the same symptoms because these are symptoms that ordinary people have in the course of their lives, even if they haven't had COVID.

Chris - Because Terence Stephenson, who did a similar sort of thing a bit on a much smaller scale and looking at younger people in the UK, found that while there was evidence that young people who had had coronavirus infection did have a certain complex of symptoms, more often he got a very similar number when he asked people who hadn't had coronavirus infections. So there must be a lot of noise in this system. It's quite hard to disentangle what's really COVID and what's background just because of what we've all been through in the last two years.

Leora - Yeah, there's actually many reasons. This is extremely hard and one of them is precisely that. One is that symptoms like fatigue, exhaustion, brain fog and joint pain and so on and so forth, any particular symptom you could come up with does exist in the general population. That's one problem. But there's other problems too. So one is that the virus itself has changed over time and it could well be that the kinds of symptoms people have long term from the omicron variant are different from the kind that people would've gotten from the original variant or delta or something other. Another challenge is that the treatments for COVID infection have changed dramatically over time as well. They may be better or worse able to prevent long-term symptoms. Another issue is that people are now vaccinated. So it could be that people have a different frequency or type or severity of long COVID because of that. And then finally the world around us has changed too. And living now is different from living in March of 2020, when the world was sort of falling apart and in 2021 and so on. So the other influences on people's symptoms and experiences have also changed over time.

Chris - How do you also get around the question of what epidemiologists and health statisticians call recall bias? If you go up to somebody who's got a problem, they're more likely to be an accurate historian or to remember things relevant to that than someone who never thought about it before.

Leora - Yeah, it's a real problem. We try to get around this in two ways. One is we enroll people who have not had COVID and we ask them about the same things in the same time periods. So hoping that, we're specifically asking them about things and it will jog their memory. The second way we do this is that we've reserved half of the slots in our study for people enrolled at the time of an acute infection, meaning they don't know yet if they have long-term consequences or not. And so we can then prospectively ask them, that means going forward in time every three months, ask them about their symptoms so that there's not bias there. And then I will finally say that we don't just ask questions. We collect large amounts of blood and other specimens.

Leora - We do a huge number of blood tests and other sorts of tests on people. We do x-rays, we do MRIs, we do at physical examinations, we do cognitive evaluations and so on. So we have as well objective data, which is a little bit less subject to bias. We also are doing tests to try to understand causes of long COVID. Here, we look at a variety of hypotheses. We are looking right now for evidence that there's still virus present in the body. We look for evidence of immune dysregulation. We look for evidence of other viruses being reactivated like EBV (Epstein-Barr virus). So we are looking for all of those different types of potential causes. And it's quite likely that some people will have some and others will have others.


Add a comment