Audio Transcript from February 10, 2026 R3 Seminar

This transcript is intended to provide an accurate, verbatim representation of the language used by the speakers during the R3 Seminar Long COVID Trajectories in Adults held on February 10, 2026. You can view a timestamped version of the transcript with the video on YouTube.

Watch the recording or read the summary on the R3 Recap page.

Disclaimer: This transcript has been generated using AI technology and lightly edited. Please note that some errors or omissions may have occurred due to the limitations of automated transcription.

Patrick Ahearn: 

So once again, welcome everyone to today’s RECOVER webinar on Long COVID trajectories in adults. My name is Patrick Ahearn with RTI, and I’ll be helping out with the virtual room today.

So just a few quick housekeeping notes before we get started. When you log in, your microphone will be automatically muted with your web camera turned off. But if you have any questions for our presenters today, please submit those into the Q&A window at any time. And if you run into any technical issues, please let me know in the Q&A window as well. Closed captions are available during today’s webinar. Just click on the show captions button on your main Zoom toolbar to turn those on.

So, at this time, I’ll go ahead and turn things over to Christine to kick off today’s session. Thank you.

Dr. Christine Bevc: 

All right. Thank you, Patrick. And welcome everyone to today’s R3 seminar. My name’s Christine Bevc, and I’ll be your moderator. Today’s session focuses on the recently published findings from the RECOVER adult observational study about clinical presentations of Long COVID and its progression over time. We’ll start with a presentation from the authors and then hear from our panel of discussants about what these findings mean for the diagnosis of Long COVID, patient care, and clinical trials. If you’re interested in reading the publication related to today’s session, we’re going to drop that link into the chat for your reference.

All right. And introducing our panel today, I am pleased to welcome Dr. Sarah Donohue and Dr. Tanayott Thaweethai. Dr. Sarah Donohue is director of research services, a research assistant professor in psychiatry and behavioral medicine, and the associate director of Population Health and Community Engagement in the Center for Health Research at the University of Illinois College of Medicine. She serves as principal investigator for the RECOVER ILLInet Hub.

And she’s joined by Dr. Tanayott Thaweethai, who is the lead biostatistician at the Data Resource Core for RECOVER. He’s an associate director at Massachusetts General Hospital Biostatistics and an assistant professor of medicine at Harvard Medical School. He’s also an assistant professor in the Department of Biostatistics at the Harvard T.H. Chan School of Public Health.

To help kick off our discussion of the findings, we also have with us today, Dr. Timothy Henrich and Dr. Mady Hornig. Dr. Henrich is an associate professor of medicine at the University of California, San Francisco, School of Medicine. His research group specializes in immune modulatory cytoreductive chemotherapeutic and stem cell transplantation approaches to HIV-1 cure.

Dr. Mady Hornig is a RECOVER Patient and Caregiver Representative. Since a SARS-CoV-2 infection in April of 2020, Dr. Hornig leverages her transdisciplinary scientific and lived experience perspectives in initiatives designed to help uncover the low-hanging fruit, especially watermelons for her, that hold promise of turning science into clinical traction. She’s a visiting scientist and professor at the Feinstein Institutes for Medical Research in New York and president of a new not-for-profit organization, the CORe Community Incorporated (COVID Recovery through Community), which is a 501(c)(3) that focuses on deciphering Long COVID and other infection-associated disorders with neuroimmune sequelae. She’s also a caregiver of a son with Long COVID.

All right, so as we jump in, if you’ve got questions about today’s session, as Patrick mentioned, please use that Q&A feature located in your Zoom menu bar. And thank you to those of you who submitted questions in advance. After today’s panel and discussion, our speakers will answer as many questions as possible. And please note, we cannot answer questions about individual cases, analyses, or proposals.

And with that, I think we’ll turn things over to Dr. Donohue to introduce us to this important work in the findings. Sarah?

Dr. Sarah Donohue:

Great. So, I’m happy to be here today to give you a little bit of background and context to this study. Next slide, please.

So, as you can see, this study came out in Nature Communications in November 2025, and we have a lot of authors on here, and I want to acknowledge all of their contributions. This was really an effort in team science where we got together people with a wide variety of expertise to come together to really better understand what these Long COVID trajectories are in the prospectively followed RECOVER adult US cohort. And I would also like to thank the RECOVER Observational Studies Consortium as a whole, all of the investigators, the study staff who worked tirelessly on this, the Patient and Community Representatives. And I would also like to thank our study participants, without whom none of this would’ve been possible. And finally, I would like to thank the National Institutes of Health [NIH] for their funding for this study. Next slide, please.

So, while I recognize a lot of people listening to this probably know what Long COVID is, I did want to start off with defining it. And the definition I’m using is from the National Academies of Sciences, Engineering, and Medicine. And this definition was basically the result of a consensus of over 1,300 individuals from patients to public health practitioners, researchers, and clinicians, all coming together to really define what Long COVID is. And they said it’s an infection-associated chronic condition that occurs after SARS-CoV-2 infection and is present for at least 3 months as a continuous relapsing and remitting or progressive disease state that affects one or more organ systems.

And with that, they also listed some common symptoms that people may experience with Long COVID, things related to fatigue; brain fog, like difficulty concentrating or memory changes; pulmonary symptoms such as shortness of breath and cough; gastrointestinal symptoms like bloating, constipation, diarrhea; cardiac symptoms, fast heart rate; and more. And these are just the common symptoms. This is by no means an exclusive list of the symptoms associated with Long COVID.

And just as the initial COVID infection presents differently in different people, with some people having certain symptoms and maybe a severe infection and other people having different symptoms and maybe a very mild infection, Long COVID also presents differently in different people with different constellations of symptoms and severity.

Now, one thing I want to point out about this definition is really the time course of it. So, when this was defined, people knew that it could be continuous or relapsing and remitting or progressive, but we didn’t really know exactly what the trajectories were specifically and how many people fell into which bucket. And so that’s really one of the things that this work we’re presenting set out to do. Next slide, please.

So, what exactly do I mean by trajectories? Well, the way I think about it, it’s basically the path of symptom burden or your condition over time. And so, what I’m showing here is a simplified version of this. But over time, you can imagine in some people, symptoms may stay stable. So, somebody may have a cough that just doesn’t go away and brain fog that doesn’t go away and fatigue that doesn’t go away. And those levels stay relatively stable over months or even years. Other people may find an increase in symptom burden over time. Some other people may find that symptoms decrease over time, and they’re starting to feel a lot better. Yet others may find kind of an increase or decrease, just this sort of oscillation of sometimes they’re feeling better and sometimes they’re feeling worse.

And while a lot of people who have Long COVID could probably tell you their own personal trajectory of this, how they’ve been feeling for the past weeks, months, years, et cetera, this really hasn’t been characterized fully. So, we don’t know exactly, or we didn’t know exactly, what these trajectories looked like and what proportion of people were falling into which category. Next slide, please.

So, there has been some previous work around the trajectories of Long COVID, and some of it has largely been symptom focused. So, there was a group out of Spain that followed a cohort of people who had been hospitalized pretty early on in the pandemic, and they followed them out for 8.4 months and then again at 18.3 months. And what they found is that for fatigue in particular, after 8.4 months, 57% of people who’d been hospitalized initially with COVID were still experiencing fatigue. And after 18.3 months, 43% of people were still experiencing fatigue. So, you can see that this trajectory is declining over time for this particular symptom, but this is of course just one symptom and not the whole picture of Long COVID.

There was another study that looked at basically health-related quality of life. So, they had people followed after an initial COVID diagnosis out up to 18 months after this diagnosis. And what they found is that about 13% to 28% of people experienced a decline in their health-related quality of life 18 months after a COVID infection. And by health-related quality of life, this really measures domains of physical and mental, emotional, and social functioning. So, it’s really impacted by how well you’re feeling and how fit you are and how able you are to carry out your daily routine and activities.

And so, finding that a certain proportion of people were on this declining trajectory when it comes to health-related quality of life was quite interesting. And while they weren’t measuring Long COVID directly, so it’s likely that not everybody who experienced this decline was due to Long COVID. It is likely that a good proportion of these individuals did have this decline due to Long COVID. So, this provides sort of an initial picture of trajectories, but it hasn’t really characterized Long COVID trajectories as a whole. Next slide, please.

So, the goal of our current work was to characterize the trajectories of Long COVID in a group of acutely infected participants in the RECOVER adult observational cohort. Now, the RECOVER study as a whole has multiple scientific aims, and the objective of our work fits into parts of 2 of them. It fits into understanding the range of recovery from Long COVID and studying how Long COVID changes over time. Next slide, please.

So, the RECOVER adult cohort enrolled approximately 15,000 participants across 86 sites. You can see some of those on the map here. And this really allowed for basically diversity in terms of geography, in terms of demographic data as well. And participants were enrolled from both hospital-based and community-based sites for this study. Next slide, please.

So now, I’ll turn it over to Tony to talk about what we did and what we found.

Dr. Tanayott Thaweethai:

Thanks, Sarah. And thank you again to the organizers for inviting us to present. I’m really excited to jump into the methods, results, and conclusions from this paper that was published in November.

So, the analysis here focused on this cohort, which was defined by the following inclusion and exclusion criteria, and understanding the inclusion and exclusion criteria are really critical for understanding what the implications of our findings are and who they really apply to. So, this study focused specifically on RECOVER adult participants, so those aged 18 or older, who were followed prospectively from their very first SARS-CoV-2 infection. And there are 2 types of participants who fell into this category. First, acute infected participants. And second, crossover participants.

So acute infected participants are those who enrolled into RECOVER within 30 days of their very first infection with SARS-CoV-2 or their first experience with COVID. That means that individuals who enrolled into RECOVER after having been infected a year or even 2 years before enrolling were not included in this analysis. And that was because we really wanted to characterize the symptoms of individuals from the very beginning of when they experienced COVID. This meant that we were able to look at these symptoms in an unbiased way because we weren’t focusing on participants who maybe had been having symptoms for a long period of time. We were really focused on understanding the natural history of Long COVID as it progressed from the time of initial infection and then following them prospectively from that point.

So, we included these acute infected participants, but we also had an opportunity to include individuals who were classified as “crossover participants.” These are individuals who enrolled into RECOVER as uninfected, meaning they had never experienced COVID, but then they developed COVID over the course of the study. So, they kind of crossed over from being uninfected to becoming an infected participant. So, we were able to also follow these individuals from the time of their first infection with COVID.

So, this introduces a lot of strength to the study. We’re able to follow people from when they first had COVID, but it also means that given that RECOVER started enrolling patients near the end of 2021, that the study did not include participants who had COVID earlier than that because we were looking at people who, again, we were following them from the time of their first infection. So that’s just something to keep in mind when thinking about the inclusion criteria for the study.

So, we didn’t include participants who were enrolled more than 30 days after their first SARS-CoV-2 infection. In past R3 Seminars, we refer to these as post-acute participants, and these will be the focus of future analyses that are ongoing into understanding Long COVID trajectories. And then we also didn’t include participants who didn’t complete symptom surveys between 6 and 15 months after enrolling into the study for acutes or after their very first infection for crossover participants. Really, we included participants who had longitudinal data to contribute rather than just a single time point. Next slide.

So, the study focused from the time of first SARS-CoV-2 infection or first time that somebody had COVID. We followed them for up to 15 months, with study visits occurring every 3 months. Many of these visits were conducted remotely, which involved participants completing remote symptom surveys during which we were able to assess them for Long COVID in this analysis. So, they completed the symptom survey where they reported the presence and the nature of their Long COVID–related symptoms, and then we calculated the Long COVID Research Index at each visit, which I’ll describe on the next slide.

So, the Long COVID Research Index has been the subject of prior R3 Seminars, but I’ll give a brief refresher here. So the Long COVID Research Index, or the LCRI, was updated in 2024, and the citation for that is listed here, but essentially it’s a kind of scoring-based algorithm that can be used for research purposes to identify people who very likely have Long COVID or probable cases of Long COVID. And the way that it works is that you look at the reported symptoms that an individual reports at each given time point, and then you add up the points for each symptom that was reported to get a total score, or the total LCRI.

And then if the score is less than 11, that participant is referred to, for the purposes of this analysis, as “Long COVID unspecified,” and if they have a score of 11 or greater, that indicates likely or probable Long COVID. There’s a lot of important things to keep in mind here. Again, this Long COVID Research Index is only being used for research here as a way to characterize Long COVID symptoms at each visit and then over time.

And then importantly, those who have Long COVID research indices of below 11—we’re not necessarily saying that they don’t have Long COVID, but that they’re unspecified and that we’re not able to clearly state whether or not they have Long COVID, whereas those who have a score of 11 or greater, that’s highly indicative of having Long COVID. What we also showed in the papers that were published looking at the adult Long COVID Research Index, as well as the Long COVID research indices in pediatrics, which are not the subject of today’s presentation, is that individuals who had higher LCRI tended to have worse PROMIS [Patient-Reported Outcomes Measurement Information System] scores on various health metrics, including things like overall health, physical health, and in some cases where we looked at this, mental health as well. So, we found that this index did track well with other measures of overall health, and so it was the primary measure in which we characterized Long COVID symptoms over time in this study. Next slide.

So, for the statistical analysis, we have data for participants in this study at these 5 different time points between 3 and 15 months after first infection with COVID. And so, we used finite mixture modeling to group participants into distinct longitudinal profiles or trajectories, as Sarah showed on her earlier slide. There were numerous statistical issues we had to contend with, one of which was missing data. So, we used multiple imputations to address missingness and potential loss to follow-up during the study. We assumed that missing observations were missing at random, conditional on numerous variables that we included here. And then we used the averaged Bayesian information criterion, or BIC, across the imputed datasets in order to identify the optimal number of profiles, which I’ll show shortly.

And then once we identified these distinct longitudinal profiles or trajectories, we then did some additional subgroup analyses looking at the trajectories of participants who met criteria for Long COVID using the Long COVID Research Index at the 3-month visit. So, we looked at people who appeared to have Long COVID according to the index at 3 months and then looked at which groups they fell into for the rest of the time that they were in this analysis. Next slide.

So, let’s look at some results. Next.

So, this is a flow diagram showing the results of the inclusion and exclusion criteria for this analysis. In the purple box here, you can see that there were around 4,400 acute participants, so acute infected participants, so those who enrolled into RECOVER within 30 days of an infection. And then we also have a group of crossover participants in the green box, which are participants, again, who enrolled as uninfected but became infected during the course of their participation in RECOVER. Then we combine these 2 types of participants together to identify our group of participants who were prospectively followed from their first SARS-CoV-2 infection. After making some exclusions related to not having enough follow-up data, we then ended up with a cohort of 3,659 participants, most of whom were acute participants, but some were crossovers. Next slide.

Here’s a summary of some very basic demographic characteristics of participants in the study. We see that our cohort—you can see the distribution of sex-assigned at birth, of age, and of race and ethnicity for this cohort. Our cohort was highly diverse. There were slightly more participants who were assigned female at birth than male participants. And then you can also see the age distribution of the cohort. In general, the acute infected cohort was not terribly different from the crossover cohort, but there were some differences here. For the full description of all of the demographic characteristics and clinical characteristics of the cohort, please see the actual published version of the study, which has the full information on the cohort.

One other thing I wanted to note is that a very small percentage of participants in RECOVER who were included in this study were hospitalized during their very first infection. And so, the findings from this paper may not generalize to participants who were hospitalized and experienced very severe critical illness during their first experience with COVID. So that’s just something to keep in mind about generalizability, that mostly these are people who recovered well in the acute phase and didn’t require going to the ICU [intensive care unit] but of course went on to experience lingering symptoms related to their Long COVID. Next slide.

One minor note about data completeness: Data completeness was actually quite good in this study when compared to many longitudinal cohort studies. The proportion of participants missing symptom data at visits 6 through 15 ranged only between 8% and 16%, which is quite good in terms of follow-up and retention. And in fact, 68% of the participants in this analysis had data measured at all 5 visits. So, we were able to really fully characterize their longitudinal trajectory and didn’t have to rely on any statistical strategies like multiple imputation. And in general, missingness was quite consistent across LCRI strata at the prior visit, except some participants who had higher LCRI were slightly more likely to miss the next visit. So, this is basically saying that participants who had a very high Long COVID Research Index were slightly less likely to attend the following visit, but we did our best to account for this type of missingness using our imputation-based approach. Next slide.

Okay. So here’s a figure, and I know it’s very busy, there’s a lot going on, so I’ll do my best to explain how to interpret this alluvial or sometimes called a Sankey plot. So here we have 5 different time points going from 3 all the way to 15 months, and we have grouped people in the study into their different categories based on the Long COVID Research Index. So there’s a group of 374 participants who met criteria for having Long COVID based on the LCRI at 3 months, and then a large group of participants who did not meet criteria for Long COVID, this LC unspecified group, as well as a very small number of people who had just been reinfected actually that quickly after their very first infection, a very small number of participants there.

So basically, these shaded areas that connect 3 to 6 months indicate the flow of participants between groups. And so, if the group is going from red to red, that means that they continue to meet the Long COVID Research Index criteria at 6 months and then 9 months and so on. But something that’s really interesting that we found is that there were participants who started out as having Long COVID and then moved into the unspecified group, which is divided into whether they had prior Long COVID or not. So, everybody who was in that red Long COVID group either continued to have Long COVID at 6 months or they went into this unspecified prior Long COVID. So, this is a group of people who at that time did not meet index criteria for having Long COVID but had it previously.

And then there’s this lighter blue, which corresponds to people who basically continued to never have Long COVID. There’s a lot of information here, but I think the main takeaway is that there is a lot of movement, that people are moving between categories with some frequency, but there also are some people who consistently are experiencing Long COVID–related symptoms all the way to 15 months and perhaps beyond that. So, there’s a lot of variety in what we’re seeing over time. It’s not that if people start out with Long COVID, they always have it, but also people are developing Long COVID over time. So, we sought to characterize this further in a more statistically principled way using the trajectory modeling approaches I referred to earlier.

So, if we go to the next slide, you can see the longitudinal profiles that were identified as parts of this study. So, in this slide, we identified 8 key longitudinal profiles, and here I’m showing Profile A, which comprised 5% of our study sample. And you can see in the light blue, the individual trajectories of the Long COVID Research Index over time where the X-axis is not currently showing, but it’s going from 3, 6, 9, 12, and 15 months, and then the Y axis is the actual Long COVID Research Index.

The dashed horizontal line is the “cutoff” for the LCRI of 11. And so, for this group, we’re seeing that they experienced that for these participants, their Long COVID Research Index was quite high for a long period for basically the duration of follow-up for this particular analysis, and then the solid line is the median. So, we would call this group a highly symptomatic and consistently symptomatic group. If we proceed, you can see Profile B, which includes people whose LCRI tended to oscillate a bit around the cutoff. So, these are folks who perhaps met the criteria for Long COVID at some visits, but then dipped below the criteria at other visits. So, it indicated that they did have symptom burden that was not quite as consistently high as in Profile A but still really oscillated. And then I think actually, rather than go one by one, if we just maybe advance 6 times, you can see all of the 8 different profiles together.

And then you can see the distribution of each of these. And then I think if you actually go one more slide ... Oh, I think, sorry, I should have organized this a little bit, but yeah, I think that’s right. If you just click all the way to the end of the following slide, it’s going to show the labels for all of them. Yes, perfect. Thank you so much. Sorry, that was confusing.

So here we have the labels for each of the 8 different profiles. The first was called persistently high, the second intermittently high. Profile C is improving or moderate. So here we’re seeing this trend where individuals were meeting, in some cases, were having slightly higher LCRIs early on at 3 months, but then there was a gradual decrease in the LCRI over time, indicating that some of the symptoms were actually resolving with time. And that’s noticed to an even greater extent in Profile D, where the symptoms were present at the beginning, but then really went away to being almost nonexistent towards the end of this 15-month follow-up period for participants here.

So I would say that Profile C and D were very encouraging for us to see that there were participants whose symptoms were improving over time, but of course, this is contrasted by participants who were in Profiles E and F, which are referred to as worsening, moderate, and delayed worsening. So, in Profile E, these are participants who had very few Long COVID–related symptoms at the beginning, indicated by their low LCRI, that gradually crept up over time and to the point of where they are at the end of the 15 months. And then in the delayed worsening group, Profile F, again, very few symptoms for basically a whole year after first COVID infection, and then there’s sort of a spike at the 15-month period. It’s only 6% to participants who are in this group, but it is something to be aware of, indicative that Long COVID symptoms don’t always show up early on, but they can have this kind of late onset appearance.

And then finally, for Profiles G and H, these are consistently low or consistent minimal, meaning that these individuals maybe had one visit over the course of their follow-up in this study where they had several, but in most cases, very few, or a couple of Long COVID–related symptoms, and then Profile H, which included, that was the largest group, had minimal to no symptoms. So, it’s a lot to digest, but these are the 8 profiles that we found. Next slide.

There were some specific findings that we mentioned in the paper that I wanted to highlight today. So in general, participants in Profile A who were in this persistent high symptom burden group, compared to those in Profile H who had consistent minimal to no symptom burden, these participants were slightly more often female, assigned female at birth, 77% versus 64%, and were slightly more likely to have been hospitalized during the acute phase of their very first SARS-CoV-2 infection. So, although hospitalized participants made up a very small percentage of the participants in the study, most of them were in Profile A, this highly symptomatic group.

Something that we were wondering also when thinking about Profiles E and F, which got gradually worse over the time, we wondered, were these individuals reinfected at higher rates than other groups? Is that explaining this late onset of symptoms that are related to Long COVID? What we found though is that the overall rate of reinfection in this cohort was 36%, and the reinfection rates in the worsening profiles were only marginally higher, 39% and 40%. So, reinfection did not seem to be a key driver of worsening symptoms in this analysis; there must have been something else that’s going on that’s explaining these worsening profiles, E and F.

We also wanted to point out that some of these late-onset symptoms in these profiles where people tended to get worse over time were symptoms like post-exertional malaise, chronic cough, and gastrointestinal issues. And it seemed that these types of symptoms were driving the late onset in contrast to symptoms like loss of smell, which is not something that tended to appear a year after initial infection, but symptoms like the ones that are listed here did for people in these groups. Next slide.

And then finally, in terms of results, we did this subgroup analysis where we looked at participants who met symptom criteria for Long COVID at 3 months. It was around 10% of participants who met criteria at 3 months, and then we looked at the distribution of their trajectories after that. And we found that slightly under half of these participants were in this highly symptomatic group, and 35% of them were in this sort of moderate group, but fortunately, 18% of these participants were in the group that got better over time.

And so here, again, we’re focusing really on the participants who appeared to have Long COVID early on, and then we can see the distribution of their trajectory profiles after that, as opposed to the other slide, which looked at the entire cohort as a whole, many of whom never developed Long COVID symptoms. Here, we’re really focused on those who had Long COVID symptoms early on. Okay, next slide.

All right. So, I’ll wrap up with just a few discussion slides to kick things off for the rest of the panel. So just as a summary, we conducted this large prospective longitudinal analysis of adult participants who all had a history of SARS-CoV-2 infection, and we identified 8 distinct post-infection trajectories. Five percent of them persistently met the threshold for Long COVID and were highly symptomatic, and then 12% had intermittently high Long COVID symptom burden. These were in participant group or Profile B.

And then just to summarize that last slide, among the 10% of participants who met the criteria for Long COVID ... I’m sorry, this should say 3 months and not 10 months. That’s my mistake. So, among participants who met Long COVID criteria at 3 months, less than half of them were in the persistent Long COVID group, and you can see the distribution of the other groups as well.

Finally, there was unfortunately 14% of participants who did not appear to have many symptoms early on, but then had symptoms that worsened over time or developed over time, and those were Profiles E and F, and it did not seem that reinfection was explaining this phenomenon entirely, that again, it’s possible that symptoms have a later onset or perhaps something else was happening with these participants. Next slide.

There are some strengths and limitations I wanted to mention about our study. Our study is very large and diverse in terms of the cohort, and we had frequent, very detailed serial assessments of Long COVID from the time of initial SARS-CoV-2 infection, thus avoiding selection bias of including only participants who maybe were already highly symptomatic at the beginning. And our frequent detailed serial assessments, I think, distinguishes it from other studies—for example, ones that Sarah mentioned earlier that had less frequent follow-up. So, our very high frequency of follow-up, I think, was one of the strengths of the study. And then also with a huge testament to the participation of participants in RECOVER and their continued follow-up is that we really had minimal loss to follow-up in this study, which really enhanced our ability to characterize these longitudinal trajectories.

Some limitations to note. As I mentioned at the beginning, we only included participants who were recruited around the time of their infection, and so participants who had very early strains of COVID would not be included in this analysis. So again, something to keep in mind about generalizability. Also, though great efforts were taken to sample and recruit participants from community-based settings and hospitals, this isn’t strictly a population-based sample, so there’s not a lot that we can say about the population as a whole. We wouldn’t say that this applies to everybody in the country or the world but that our cohort was highly diverse and includes many people from rural settings and other settings. So, we think that it’s very valuable to have characterized Long COVID trajectories in this highly diverse cohort, even though it’s not necessarily exactly representative of the country or the world.

Finally, though we do have some findings that we were able to report about individual symptoms, we were able to bring together many symptoms simultaneously by looking at the Long COVID Research Index. And so, our analysis didn’t focus on the trajectory of individual symptoms, but instead of Long COVID as a whole, which can be thought of both as a strength and a limitation. Next slide.

And so finally, just to wrap up, we feel that the variability across individual-level trajectories will really enable future studies to evaluate risk factors and biomarkers from biological samples that could predict future outcomes and explain differences in time to recovery. So really, this should set the stage for future analyses to understand why is it that people fall into these different trajectories. Future analyses will also include time-varying exposures that can be incorporated, including things like vaccinations, a more detailed look at reinfections and other medications that people might be using to treat their Long COVID. And then finally, these profiles may serve as useful targets for enrollment into future studies of pathophysiologic mechanisms, preventive strategies, and treatments.

So, with that, I think just the next slide ... Yes. Finally, to echo Sarah’s comments from earlier, we just wanted to thank all the participants enrolled in RECOVER, without whom, of course, none of this would be possible, as well as the National Community Engagement Group, NCEG, as well as all Patient, Caregiver, and Community Representatives who contributed to this paper and to all the papers that are part of the RECOVER Initiative. And here we’ve listed the grants that are also funding this work. For a full list of grants and disclosures, please refer to the manuscript itself. So, I’ll end there and then hand it back to the rest of the panel.

Oh, Christine, I think you’re—

Dr. Christine Bevc: 

Oh, I thought I pushed the button to unmute myself. All right. Thank you, Tony and Sarah, for that exciting set of results there and kicking us off. If anyone has questions for our authors or our panel, please use the Q&A in the Zoom menu. We’re getting a lot of questions in there, which is exciting. Some of them have answers, some of them are still to be answered, but let’s go ahead and bring in our discussants to help us sort through what all this means for diagnosis, patient care, and the future of Long COVID. Mady, let’s start with you in this. And for our audience, can you tell them a little bit about your personal experience with Long COVID as a patient and a caregiver?

Dr. Mady Hornig:

Yeah, of course. I think that this is a real opportunity for me to speak about that dual experience. But I’ve been sick since April 2020, which came at a time when I was actually writing a paper, a white paper for a group called SolveME to bring to the Hill about the potential persistence of SARS-CoV-2 infections, and that actually helped in the RECOVER funding in May of 2020. I have had over $400,000 in medical billings over this period of time. Fortunately, I have good insurance that’s covered most of that, but I don’t meet criteria for the Long COVID Research Index. I would be in the indeterminate group. And I think that what this research really helps us to start to delineate is what my experience, and I think many of those people in the indeterminate or “possible group” have experienced that the severity of illness may be, and the pattern of illness may be a little bit different.

I’ve had cardiovascular symptoms, tachycardia that came out of nowhere that was not associated with standing up or position like a postural orthostatic tachycardia. I was just sitting there, and my pulse would shoot up to 160. I had COVID toes at the beginning of my infection that led to swelling of my entire leg, and I had to be ... There was a risk of deep vein thrombosis and pulmonary embolism. I’ve had anemia—iron deficiency that required infusions of iron. I’ve had crisis-level hypertension, but yet at any one time, I didn’t meet all of those symptoms that would’ve put me in the bin of having Long COVID Research Index. And so I think that what is really nice here in this research is that we’re seeing that the symptom burden patterns are really relevant for us to understand how to predict perhaps who’s going to be ill for how long, and also potentially when we put it together with biomarkers to see more about these other possibilities for prediction.

I have also had gastrointestinal symptoms. I’ve been hospitalized numerous times. Again, that intersection and that frustration, which many of those who’ve had brushes or who are still having symptoms consistent with Long COVID. I’m a physician, and I still would be in the position where I’d have to be telling my care team what the issues were. And I think maybe we can talk in a few moments about medical gaslighting, but I wanted to talk about my own experience as a caregiver. My younger son who came home when I could not actually sit up because I had an inflamed muscle that was so painful, he came home to live with me and then he got COVID, and he got a big plug, mucus plug in the large part of his bronchus. And he’s in the living room, his oxygen saturation is plummeting, and I pull out my stethoscope, and I was so gaslit myself, I thought I had forgotten how to use a stethoscope because there were no breath sounds in the bottom of his right lung.

And so now it’s been a challenge sort of addressing that. Fortunately, we’ve both had the opportunity to be on an agent called Inspiritol, which is nebulized, not yet FDA [U.S. Food and Drug Administration] approved, but that has been enormously helpful for both of us, but it’s not approved, and we really hope that that’s going to be something that we’ll be able to get out there.

Dr. Christine Bevc: 

So, turning to that more challenging topic, and you alluded to that, and we hear this from other participants in this study, is the medical gaslighting. And how can what Sarah and Tony have just shared potentially help to address this issue?

Dr. Mady Hornig: 

Well, I think that the more education that we can have of clinicians, that the patterns over time may vary. For me, one specific organ domain was prominent at one time, but maybe not altogether. And so, understanding how symptoms in different individuals may fluctuate over time may be really important. So, I use this little metaphor of checking the windshield wipers. So, they’re working, they’re not working, they’re working, they’re not working. So, you need time, you need to look at the issue over time, and our clinicians are conditioned to get an elevator pitch from their patients. You got 2, 3 minutes to tell, what’s your problem? What are you here for? We don’t have integrative teams. We don’t have teams that are working together, and that’s been a really enormous frustration in my mind.

And in fact, as somebody who’s a researcher long-term in infection-associated chronic complex illnesses, including Long COVID, but also myalgic encephalomyelitis or chronic fatigue syndrome [ME/CFS], it is very frustrating to have this continued sort of defiance on clinicians, but I can’t really necessarily blame all clinicians also because the system is organized around them having 15 minutes at most to address a problem and then you’re out the door.

So, I think that that is something that education can really help and research that helps to address some of the persistent myths amongst the clinical community can be really valuable. And the more that we put in the biomarker studies that are ongoing in RECOVER like the systems biology work, as well as all of the pathobiology studies, and the more we understand about how and if, and in whom viral persistence is an issue, in whom are vascular problems an issue and cardiovascular sequelae, dementias or neurodegenerative issues—are we putting people at accelerated risk on a neurocognitive decline? The more we have research like what RECOVER has the capacity to do, as well as the other researchers around the world that are working on these issues, I think we need to take the lessons we learn and bring that into clinicians.

Dr. Christine Bevc: 

All right. So, again, expanding on if you had one takeaway of implications of these findings for patient care, what do you think it would be?

Dr. Mady Hornig: 

Believe your patients.

Dr. Christine Bevc: 

Well, thank you for sharing your experience with us today. We’re going to bring you back for the Q&A because your insights are invaluable. But I want to turn quickly to Tim Henrich, a clinical researcher. He provides care for individuals with Long COVID and other infection-associated chronic conditions. Tim, can you tell us a little bit about your experience with Long COVID in a clinical role? When did you start to see patients with Long COVID? Thinking about the trajectory that Tony shared, at what point are your patients on that journey?

Dr. Timothy Henrich:

Yeah, thank you. Most of my interactions now are on a clinical research or translational research scale. So, through our UCSF [University of California, San Francisco] LIINC [Long-term Impact of Infection with Novel Coronavirus] cohort, we see a lot of people with and without Long COVID that we enroll both in our prospective clinical trials, but also our PET [positron emission tomography] imaging studies, our cohort studies, our tissue biopsy program. But what’s interesting is that we actually started to see Long COVID before Long COVID was really a recognized thing. So back in early 2020, we had put together the UCSF LIINC cohort to really study the immunologic outcomes of SARS-CoV-2. It was very clear that this was not your run-of-the-mill respiratory virus; this was not just flu, that there was going to be long-term implications, but we had no idea what those were going to be yet. But just clearly from the severity of initial illness, how it was spreading and how people were presenting during those initial strains, we really felt that we needed to start looking and collecting stories and samples from people that were infected very early on during that cohort.

And again, initially we were interested in what the immune system was going to be like. Are people going to get reinfected, were they going to be protected? But what we started to notice is that 3 to 4 months after we started enrolling people back in March, April actually of 2020, so before the RECOVER Initiative, that we started to see people just weren’t getting better. That was obviously a minority of individuals with infection, but there were some folks that really just had profound changes in activities of daily life, symptomology that was not improving, the smell and taste phenomenon. And it was about the same time that this was also being recognized in the lay press, and it was being talked about more and more. And so, we were struggling to figure out what was going on, but also just really have never had seen something on this scale before.

Now, I had experiences with ME/CFS and infection-related chronic illnesses about 15 years ago in Boston on the clinical and research side, and in my outpatient clinic there in Boston as well. But I think that just the scale of this was unprecedented. And I think that we rapidly learned that we needed to look at this. This is something that we had to look at systematically to really understand what was going on, what was driving this, how to make people better, et cetera. But to get back, I think that the paper with Tony and colleagues is really quite representative of what we’ve been seeing, especially in the more modern era. I think there are differences between kind of pre- and post-Omicron, pre-vaccine, pre-reinfection era and now. But certainly, we are seeing people that ... I think the good news is that a lot of people do fully recover after COVID-19, but some don’t.

And those that don’t do have these different trajectories; some just have very persistent symptoms over time. Some can wax and wane as shown, some kind of come up and down, and some people do gradually get better, although for some people, that can be very prolonged. And I think this is talking to what Mady was saying too about medical gaslighting and just the perceptions on the clinician side, is that we have a very heterogeneous disease that we don’t really have a full biological understanding yet. We have a lot of clues, but really not a holistic, full biological understanding at this point. And so, I think this has caused a lot of frustration and also a lot of just lack of knowledge. And I think unfortunately, when clinicians are in a position of not understanding something and not seeing something, that this can become a gaslighting situation and certainly has been the case, I think, for many people.

And so, yeah. I think this certainly does mirror what we’re seeing but also presents some of the challenges of what we’re seeing. This is not a single homogeneous syndrome that everybody experiences the same. And we really need to understand that patients and our participants and people really have different experiences through this process.

Dr. Christine Bevc: 

So on the point of the medical gaslighting that Mady brought up, how can the findings that we heard about today, how can those potentially help to address the challenge?

Dr. Timothy Henrich:

Yeah. Well, I think Mady made a great point earlier that understanding Long COVID is absolutely critical for clinicians, that anything like this paper that actually really tells us a little bit about what we would expect, what types of patients are we to expect to see in the clinic, what are their trajectories? This is actually very valuable information. And I think that this does go ... Information like this is important to impart to people, especially those who are not working with people with Long COVID on a full-time basis, et cetera, et cetera. But it also may lay the framework, as I was saying, for more comprehensive care. If we understand that this is going to be a prolonged waxing and waning heterogeneous process, then certainly we’re going to need to build and find those resources and infrastructure to really provide that comprehensive care. And I think also to understand the trajectory to try to get people into clinical trials, into research studies where we can better understand what’s happening, and that we can actually do those studies where we find evidence-based therapeutics that will work across many individuals.

And I think that obviously also with that lacking can be frustrating on the clinician’s end. But yeah, I think we need information like this. We really have to understand and educate our fellow clinicians on what to expect, what to see, and how to approach and engage with this major problem.

Dr. Christine Bevc: 

Is there anything that you heard today or from the article that surprised you about these study findings?

Dr. Timothy Henrich:

Yeah. I mean, again, having worked with this for the last 5 years or so, 5.5 years, we’re always surprised with things that come up sometimes in a great way, sometimes in a not-so-great way. Yeah. I think for me, the major take-home message is that there is hope that actually many people do get better. It can be a slow-going process for some, but unfortunately, there is a small percentage of people after COVID that just really have these persistent long-term types of symptoms. And I think that really kind of reiterating that that is the case and looking for cases, I also thought it was very interesting to show that there can be symptoms that develop longer term after time. So, you can have this period of time without symptomology and then it comes up. And I think it really, again, adds to that complexity of what we’re seeing and also adds to the complexity of ... Speaking of gaslighting as well, people may say, “Well, how do we know that’s COVID related? Could this be something else that happened interim, et cetera?”

So, I think it just makes it more complex for both the research and the clinical community, but it’s a challenge that we need to take on.

Dr. Christine Bevc: 

So, this is all beginning to feel like a really pivotal time in the research, and I want to open it up to our full panel. So, bringing Sarah and Tony back into this, about what you’d like to see as the next steps in research. And Sarah, why don’t you start us off?

Dr. Sarah Donohue: 

Yeah, so I think I’ve seen this come up a little bit in the Q&A, and I think one of the key things is really looking at these trajectories for people who are pre-Omicron, people who were infected early on in the pandemic and understanding really how that’s going for different classes of people, and if that looks the same as what we’re seeing for Omicron, how it varies, and then going more into the pathobiology behind it to really understand why some people end up in group A versus group B, et cetera.

Dr. Tanayott Thaweethai:

Yeah, those are great points. Yeah, so I think it echoes this trade-off that I was referring to earlier that we were able to really closely follow people from the time of their first infection, but we could only do that for people who had essentially Omicron or later variants. And so, it’s a strength that we were able to follow them so closely from the time of their first infection but also a huge limitation of the study. And I’m excited for future analyses that will incorporate data from participants who were infected earlier on, as many of these people, I mean by definition, have been suffering from Long COVID for even longer. And so, it’s even more critical that we better understand their symptoms. And then, I think going in the other direction; this was limited to just 15 months of follow-up. I’m very curious to see that, as people continue to participate in RECOVER, we have more and more data over longer follow-ups to understand and characterize.

So, we’re kind of going in both directions. We’re going to look at earlier infections as well as later symptoms of Long COVID. I think finally, RECOVER is not just a study of adults. It’s also a study of children and understanding how these trajectories are potentially different in different ages, especially as children go through different developmental stages and have different presentations of Long COVID, I think will be really important to characterize. And so, I’m very interested in understanding these trajectories in children for whom Long COVID is even more of an overlooked public health problem than adults in some cases.

Dr. Mady Hornig: 

I would love to just chime in on—I think those are just really tremendously important points for the future. I think also what we saw in the paper was that there was a signal relating to reinfections, not always, but there weren’t enough for us to really know how that is going to play out, how there weren’t enough reinfections. So, I think I’m particularly interested in knowing about that as we go along. What should we be doing when somebody has a reinfection? I recently had a reinfection myself—how does it help if you acutely have Paxlovid or other agents in the acute period?

And I was able to get back on track with a combination of $814 5-day Paxlovid course, which was pretty astoundingly expensive. But also, I was able to sit up after those 5 days to continue taking this medication, which you nebulize. So, I lost some ground, took me some time to get back again, but I think that those data are going to be really important. I was the “OG” variant, but then obviously it was whatever variant it is, but docs are not asking people to come in, to check. We’re not paying attention to reinfection. Some people don’t even bother to test. So, I think that the more we understand about at least the identified reinfections and their impact on their trajectories, I think, and RECOVER, I think it is poised to potentially help with those insights.

Dr. Timothy Henrich:

And maybe I can just add briefly on some of the very early infections, because I know RECOVER took a little while to get off the ground. I mean, we certainly have studied and interacted with and enrolled people that were infected very early during the pandemic. And I actually saw that there were some Q&A in the questions as well. And I think we actually, and just anecdotally, have certainly seen that these people tend to have more of the persistent severe phenotypes over time. We’re wondering if this has to do with initial viral seeding and reservoirs, and different tissues maybe have been more during that era. It is a different virus in the way that it engages the ACE2 receptor, et cetera, and different tissues. And so, yeah, I think that they are discrete. I mean, there’s obviously a continuum there. This is all COVID-19 related, but certainly there are some discrete instances.

And we still have people that were infected in April of 2020 that still have a lot of symptoms, and we’re enrolling in clinical trials. And I think we see less of that just anecdotally in our post-Omicron era, and not to minimize at all, but certainly we see a very high burden still, I think, in that initial cohort.

Dr. Christine Bevc: 

All right. Well, now let’s turn to some of the questions that we’ve been asked by our audience today. For our audience, we’ve received over 100 question submissions today live coming in. So, we most likely are not going to be able to answer all of them within this time period, but we will be sharing a summary of today’s seminar and answering those questions that’ll be shared on the website later. So, if we run out and we’re not able to answer it, responses will be provided on the website. So, let’s start with a couple questions that we haven’t had a chance to cover. And Tim, you actually alluded to this in your statement there. And to Tony, for the numbers here, what percentage or number of long-haulers go beyond the 3 years with Long COVID symptoms? We’ve gotten a couple questions about this from people who’ve had Long COVID for 3+ years.

Dr. Tanayott Thaweethai:

This is a hard question to answer. I think it’s beyond the scope, I think, of the analysis that we conducted, which did just look at people for up to 15 months after their infection. So, I can’t comment on specific rates of continuing to experience persistent symptoms. I think we can speak to people, from a high-level point of view, from the data that we’ve seen in RECOVER that there are many people still reporting many symptoms years out from their initial infection. Me saying that I don’t know the number doesn’t mean that it’s not high or that it’s not something to be very concerned about. And I think it’s certainly going to be the subject of future analyses for sure.

Dr. Christine Bevc: 

Oh, that’s exciting. We’re looking forward to that. Can you just address, and well, I’ll open this up to Tim and Mady also in terms of relapses of Long COVID symptoms and how that’s captured in the data, and then I’ll turn it to Tim for a clinical perspective.

Dr. Tanayott Thaweethai:

Sure. Mady, go ahead.

Dr. Mady Hornig: 

Oh, sure. So, in my own personal experience, my recrudescence, if you will, of symptoms always came with some other new infection. While recently I had my first reinfection with SARS-CoV-2, I had a common cold type of coronavirus 229E back in ‘22 that put me in the hospital. I had bilateral pneumonia and reactivation, kind of like shingles, but without the rash, only with itching. So, herpes zoster—I’ve had that 4 times. And each time I’ve had those additional issues, it’s been sort of a recrudescence of various symptoms, but which organ system is involved has varied. So, I think the more we’re able to understand what that’s about, and if I have viral persistence with SARS-CoV-2, how is that? Is it because SARS-CoV-2 is becoming reactivated in my GI [gastrointestinal] tract that I end up with GI symptoms, or is it really the overall impact on my immune system? And maybe it’s just a random thing, which organ it ends up impacting. I’ve had some vascular things, cardiovascular things and so forth. So, I think this is really why this research and continued funding for this research is so, so critical.

Dr. Timothy Henrich:

Yeah, I agree that we really need to understand this better. We need this research to happen. I can always speak anecdotally. I’m a bit biased, because in our clinical studies, we often have people that are on the sicker end or more persistent end, where we do see that people who persist are unfortunately the ones who tend to persist. And so, once you get to a couple of years, those are usually the folks that we see go much longer. And anecdotally, again, we have people 5+ years that have been impacted that we are now enrolling in some of these studies. But I think having further research and further understanding of why some people do fully recover, some people take a long time, but recover. And then in terms of the relapse as well, this is certainly, again, anecdotally something that we see, especially when we do our questionnaires for different studies.

It seems like there is some type of insult to the immune system or inflammation that happens, whether it’s another illness, another infection, potentially reinfection, that can certainly trigger these. And so, that’s certainly just anecdotally something that we do see as well, of course, in the population. And I think for me, really what is frustrating is that we would love to have therapeutics that work for a large number of people that we can roll out and either stop these or do that progression. Again, we need more research in order to do this and to have the biggest impact on everyone.

Dr. Christine Bevc: 

And Tony, you alluded to this a little bit at the end of your presentation as one of the follow-up directions, but can you just clarify whether you’re able to differentiate between the various strains, recognizing that the early strains were quite different from the latter strains and what ... I think you mentioned this in the limitations there, but could you just reiterate that for us?

Dr. Tanayott Thaweethai:

Sure. Yeah. So I think just related to first, I think the relapse or late-onset symptoms just in general, I think it’s just to echo Mady and Tim’s thoughts, I think that it’s so important for us to understand why it is that these, what’s going on there, if and how it’s being triggered by other infections or other pathophysiologic mechanisms, it can be very difficult for us to assess this in a rigorous statistical and epidemiological framework. There are so many possibilities. And the classic issue is that if you test a million things, you’re going to find some false positives, so to speak. And so, it’s important for us to be very thoughtful and rigorous when we’re analyzing these, so that we’re not tricked by spurious findings.

And so, I think it becomes very complicated when analyses are longitudinal in this sense, because of a phenomenon known as time-varying confounding by which there are very ... What’s happening to try to characterize someone’s entire natural history is very complicated, because you’re trying to say, “Well, what would’ve happened if this person received this treatment? Would their symptoms improve afterwards, or would they have gotten worse?”

It’s very tricky and there’s a lot of things you have to account for. So, I think that’s, again, the subject of future work so that we can better understand really what the triggers are. And I think a lot of the work into the pathophysiology of PASC [post-acute sequelae of SARS-CoV-2] through ROAs [Research Opportunity Announcements] and other ancillary studies and announcements that RECOVER is putting out to invite future research is to really look at these, what’s special about people who are in the worsening groups as well as the improving groups, and that’s going to be really critical going forward. To answer your question specifically, Christine, about the ... I of course could just go on and on, but about the strains.

So in this analysis, knowing that our cohort was already restricted to people who essentially were Omicron era or later, we didn’t formally adjust for any strain, because there were not people included in the analysis who had earlier strains, but I think that once we get into analyses that incorporates, again, to use RECOVER lingo, the “post-acutes,” meaning those who are enrolled in RECOVER more than 30 days after their first infection, so these are going to be folks who were infected as early as 2020 and earlier on in 2021, that this is something that we definitely want to account for in analyses of those, so that we can better understand whether there are maybe specific interactions between the type of strain that someone has and then the things that happen to them over the course of their follow-up, to understand if certain findings are applicable, perhaps to people who had a particular strain. So, it’s an excellent question, and it’s really how you address it. It relates to which cohort you’re looking at, and I think I agree it’s quite important to factor that in.

Dr. Mady Hornig: 

One piece of that, again, these are such great points. One point that I don’t think I emphasized quite as clearly as I might have wanted to, which is that one of the long-term things that I think RECOVER has in the more recent years focused more intently on some of the symptoms that might tell us about the development of ME/CFS, myalgic encephalomyelitis/chronic fatigue syndrome. And I think that the more we do that, I think it will be really very helpful. We should be bringing in all sorts of other comparator groups into these studies, people who have infectious mono, which is usually caused by Epstein-Barr virus, doing follow-ups of those individuals, whether they’re pre-pandemic or in the current era.

I think looking at all of the infectious stimuli that can cause these similar types of patterns of symptoms, yes, SARS-CoV-2 has its own unique pattern. Epstein-Barr virus has also its own unique patterns and HIV and all sorts of—Middle East respiratory syndrome, et cetera, that I think we need to think especially about a comparative biology or comparative virology and the immunology that goes along with that as well as the clinical features so that we can ... Yes, there are some commonalities, and we’ve done a great job at trying to pull out those common threads. And the more that we do that, the more we will learn, but we also need to see maybe they’re different colored threads or have different sizes, et cetera. Those differences are also going to be important. And especially for disorders like ME/CFS, which is largely a disease that’s initiated by an infectious and viral-like syndrome. The more that we’re going to be able to understand both about SARS-CoV-2 and Long COVID, but also pre-pandemic or other infectious agents that might lead to these disorders, and also SARS-CoV-2-associated ME/CFS.

Dr. Christine Bevc: 

So I want to shift a little bit and start, Tim, with just some of your anecdotal experience around this, around how vaccines play into this and how individuals who have received their vaccine after a Long COVID onset, how that plays into some of their symptom trajectories. And so Tim and then Tony, if that’s again, future research piece there, and then Mady. So Tim, from your experience in the clinic.

Dr. Timothy Henrich:

Yeah, I mean, again, I’m mainly interacting with people through research on this case. I think it’s mixed. I think first of all, there is actually data showing that having the vaccine before infection actually can reduce the chances of having Long COVID. So I just want to throw that out there, that is evidence, and it makes sense logically as well from an immunological standpoint that if you can reduce the severity and/or control viral replication during initial infection, then that you’re probably going to have ... It’s going to be beneficial for not developing Long COVID down the line. And I think that’s starting to play out now in the data. I think, again, just from questionnaires, it’s highly variable depending on what people see. Some people actually get better, some people feel worse, many people don’t feel different. And so it really is a personal experience. And I don’t have a lot of data otherwise. Other folks I think are looking at this question specifically. I am not personally, so I don’t have much more than just the anecdotal stuff, but it’s variable depending on the person.

Dr. Christine Bevc: 

Right. Let’s throw this over to Tony since he has some data in there, and not that we have the results from it.

Dr. Tanayott Thaweethai:

Yeah. So just wanted to ... There’s been a lot of evidence in support of what Tim is saying that being vaccinated prior to first infection can be protective against Long COVID. We published this as part of RECOVER pediatrics, finding that in adolescents, this paper came out late last year in demonstrating that adolescents who were vaccinated prior to infection were less likely to develop Long COVID later on. I think it doesn’t answer the question, I think, of what happens to people who already have Long COVID and what happens to them after they get vaccinated. What I can say here is that all of the 8 profiles that I showed, all of those groups are mixtures of people who did and didn’t get vaccinated, got vaccinated after they developed Long COVID symptoms. And so, it was not something that we had the capacity to do a rigorous job of evaluating for this paper.

There was already so much that we were trying to understand, but it is something that we’re thinking about. And I think future analyses that are ongoing are looking at let’s find, for example, somebody who has Long COVID, maybe has been experiencing Long COVID symptoms for months or years, and then let’s find somebody else in RECOVER who has a similar symptom profile longitudinally and their similar symptom history who didn’t get the booster in a given season or something, and then let’s compare them to each other to see what happens to them later on. And so these kind of advanced epidemiological techniques to do sophisticated matching can help us try and better understand what happens to them later, but it involves very careful comparisons because you want to make sure that you’re just an unadjusted comparison of people who get vaccinated versus not may lead you down a very wrong path because of all of the things that may be leading people to get vaccinated. I think that there are people with various conditions that mean that their doctor tells them not to get vaccinated.

And so there’s so many factors that you have to take into account, and it’s a complicated question, but I think that one that data from RECOVER is well poised to answer because we collect such detailed data on people’s symptom histories as well as their immunologic conditions that they may be experiencing that can help better answer this question.

Dr. Christine Bevc: 

Mady, did you want to add anything on that one or my next questions for you?

Dr. Mady Hornig: 

Yeah, no, we can move on.

Dr. Christine Bevc: 

All right. So, looking, one of the things you mentioned, Tony, is about the percentage of adults with Long COVID that recover, but not all of them do, and there’s this new normal. But, Mady, what does baseline look like from your perspective? Pre-COVID baseline might not be what folks might expect, but from your perspective on a good day?

Dr. Mady Hornig: 

Well, I tend to look at it in percentages. I think of myself as getting back again to where I was before this reinfection a month ago, and I was about 85%, 90% of my pre-pandemic self, but I came into my first COVID infection with 7 autoimmune diseases. So, my baseline may have been a little bit different than other people, but I wasn’t sleeping 16 hours a day, which I was for a lot of the period with Long COVID. I wouldn’t even call it unrefreshing sleep. I just needed a whole lot more of it. And so I think that one of the key things is really being able to be functional in terms of, and getting back to one’s former level of function, even if that was at a lower level than you might’ve wanted it to be because of other preexisting diseases. So, I think that it is really important when I go to the doctor and I’m looking for assistance with ... I tend to do a lot of deciding on my own whatever I have the competence to try to decide.

But when I’m asking my docs, I want to be able to think about the things that are really, really interfering most with my function. And for me, it’s not only the percentage of pre-pandemic self, but having enough hours in a day where you’re functional. So, I save up certain tasks where because it’s going to take me another 2 hours than I should need to just get going in the morning with my matcha and the double-strength green tea and ginger and all sorts of other things on top of the Inspiritol agent that I’m taking. So, function, being able to do things at a level that preexisting, not having to change the size of the cup. There’s this whole thing about, is the cup half full or is it half empty?

Sometimes I was probably thinking about, “Well, I’ll just shrink the cup and then I’ll be overflowing.” I was like, “Have a little tiny thimble at the end.” But you shouldn’t have to adjust your hope. If there are answers that, as I was saying, that the low-hanging fruit, the watermelons that are on the ground, there may be some things that we can do that help now that clinicians could know about that we could perhaps be tapping into. So I think there’s a lot of data in RECOVER about things that people are doing, supplements and meds, et cetera, a lot of data that we may be able to mine in order to get some of those answers while we’re waiting for really robust curative answers. And so, when I think of my ... I think all of us with Long COVID and or any other chronic illnesses deserve the opportunity to have hope that there’s an opportunity to get back to the before times, but keeping in mind also that it may not be possible yet, but there may be some things that we can do that can help.

And also getting vaccinated, wearing a mask when you’re in crowded places, campaigning for clean air and HEPA [high-efficiency particulate air]-filtered air in indoor spaces to avoid reinfection, those are also—even with vaccination and reinfection can still hit some people hard.

Dr. Christine Bevc: 

All right. Well, we’ve got just a few more minutes left, so I want to close out with one last question to the panel, and Mady already we’ve touched on this in terms of hope and what you’re hopeful for from this study and this work that’s ongoing. And Sarah, I know we’ve had you over there, and you’ve been listening, but from all this and the work that you’ve been doing, what does the future hold for this?

Dr. Sarah Donohue: 

I mean, I really see this as kind of just the beginning, and I think we have so much data from RECOVER thanks to all of these participants. And we can go in so many different directions to study different aspects of this. And I think that’s what’s really helpful is that once we have the capacity to study all of these different aspects, then we can actually start understanding these and understanding what treatments may be effective in them and roll out clinical trials. So, I see this as basically the first step in a series to get us to where we need to be.

Dr. Christine Bevc: 

Tim, where do we go from here?

Dr. Timothy Henrich:

Well, I think if we think of Long COVID as a giant 1,000-piece puzzle, we’ve now flipped all their puzzle pieces over to the appropriate area. We’ve started to work on the edges. So, I think that there’s been a huge amount done in terms of understanding the clinical trajectories, obviously, as Sarah and Tony have really highlighted today. But on the pathophysiologic side as well, I think there’s been a huge amount of work. There’s a lot more understanding now about how other infection-related chronic illnesses play a role in ME/CFS, et cetera. I think there’s a lot more engagement from the scientific physician community as well. And I think this is all very promising, and we just need to put those pieces in their final position. They may be different for each person, but I think we have those pieces, and I would agree with Dr. Donohue that this is the start, but I feel confident that we’re in the right direction.

Dr. Christine Bevc: 

Tony, Mady, anything to add to the puzzle?

Dr. Mady Hornig: 

No. I think both Tim and Sarah handled that extremely well. I am only focused on trying to ensure that we have the infrastructure and the continued funding to be able to do this sort of work. I think that’s going to be really, really critical in order to have the answers that we need and to know what Long COVID is and also understand what it’s not, how it differs from other disorders. Where can we borrow treatments that have been developed for other disorders, and where do we need to discover and develop specific treatments for these long-term SARS-CoV-2 sequelae?

Dr. Christine Bevc: 

Tony.

Dr. Tanayott Thaweethai:

Yeah, I really like Tim’s analogy of the puzzle because I think it’s also like we’re at the edges, and we’re not looking for just one puzzle piece to bring it all together. I think it’s like we’ve seen ... I think that the theme that comes up over and over in Long COVID is how, and it’s come up today, it’s like how heterogeneous it is, that maybe it is the case that there is one explanation for everything, but it seems highly unlikely. And so I think that it’s the collective work of people with all different types of backgrounds. Today, the group that’s assembled here, we all have such different backgrounds working together on this problem to find all the pieces that fit together, I guess.

Dr. Christine Bevc: 

On that note, we want to thank you. Thank you so much to our panel for sharing your expertise, your knowledge, your experience, and thank you to our audience for joining us today and engaging in today’s sessions and all of the questions that we’ve seen there. For interested researchers, and several of you posted this into the Q&A, we do have an active opportunity that’s open. It is our current Research Opportunity Announcement and Notice for Ancillary Studies from NIH. Kate just dropped in the link to this. This is an opportunity for ancillary studies that continue to open. We’re accepting applications on a rolling basis. These are some of the ways that researchers both within and outside of RECOVER can apply to use the participant data, so the data that Tony presented today, as well as some of the emerging data that’s going to be coming from future or ongoing studies at this point.

So, you can propose your own research and request access to the data. So, thank you, Kate, for dropping that in there. You can also, there was the pop-up survey that came up there. We’d appreciate your feedback on today’s seminar as well as any future topics that you’d like to see. Recording of this along with the Q&A will be available on recovercovid.org in about a week. We’ll be posting that summary along with responses to the questions that were submitted today.

Thank you so much for joining us. We’re going to be taking a short spring break, but we’ll hope to see you in a future session, so be sure to check out the R3 Seminars webpage for updates and registration for future events. Thank you and have a great rest of your day.