I wanted to do a quick post about incidents of myocarditis (inflammation of the heart muscle) in young men after the Pfizer vaccine. This has been in the news for a bit, but I don't think the uncertainties involved in this issue have been properly explained. Ultimately, I'd like to be able to fit this risk into my risk analysis framework, but we have to deal with the uncertainties first, and in this case they make for a very difficult risk to assess.
Background Rate
Whenever I hear a report of a possible complication associated with one of the vaccines, the first thing I do is research to find out, what is the background rate of that complication? I then look at the rate at which this complication is being reported by vaccine takers, try to adjust based on how much I think this complication might be underreported, and then the difference between background rate and the adjusted rate for vaccine takers is the increase of risk or this complication that can be imputed to the vaccine.
With myocarditis, however, I found that there isn't a simple answer for determining what the background rate of myocarditis is, especially among young men which is the demographic of concern.
The most up-to-date summary on myocarditis incidence rates I know of is "Epidemiological Impact of Myocarditis", published just this past February. It describes the current state of knowledge as follows:
Myocarditis was considered a rare disease until intensified research efforts in recent decades revealed its true epidemiological importance. While it remains a challenge to determine the true prevalence of myocarditis, studies are underway to obtain better approximations of the proportions of this disease. Nowadays, the prevalence of myocarditis has been reported from 10.2 to 105.6 per 100,000 worldwide, and its annual occurrence is estimated at about 1.8 million cases. This wide range of reported cases reflects the uncertainty surrounding the true prevalence and a potential underdiagnosis of this disease.
"Between 10 to 105 cases per 100,000" is a very wide range. There are several reasons for this:
Variability of Severity and Symptoms
Myocarditis is a generic description of any sort of inflammation of the heart muscle. It can have a number of different causes, and it can be more or less serious, depending on the cause and the degree. The noticeable effects of myocarditis can range anywhere from "none" (as in, you have myocarditis but just don't notice anything) to "mild chest pains" (which I think is the most common manifestation) all the way up to "sudden death" (see Myocarditis in Clinical Practice).
It is this last part that has caused the medical world to try harder, in recent years, to get a handle on how often myocarditis is happening. What they have noticed was that in autopsies of people who died suddenly and unexpectedly from heart failure, myocarditis was frequently present even though the patients had not exhibited any of the classical observed symptoms (see Myocarditis and Sudden Death). If myocarditis is capable of being present with little or no symptoms and yet still causing or contributing to death, there must be many more people who have it at any given time and do not know it, nor suffer any obvious consequence from it. There would be no reason for these people to come to the attention of the medical world, so the conclusion has been that we don't really know how often myocarditis happens, but we know that it is very under-diagnosed.
One possible measure of how much it is under-diagnosed comes from a study of post-Covid athletes, Prevalence of Clinical and Subclinical Myocarditis in Competitive Athletes With Recent SARS-CoV-2 Infection. There, they evaluated patients just from symptoms first but later followed up with an MRI. When they did the MRIs, they identified 7.4 times as many incidences of myocarditis than they did by just symptoms alone. Now, this study was just for a very particular demographic in a very particular circumstance, so I'm not sure how well it generalizes to the population as a whole. However, it does mean that a difference between actual rates of myocarditis and reported rates of myocarditis all the way up to, say, 10x are within the realm of possibility.
Variability of Understanding
Given that the medical world has fairly recently recognized myocarditis as being under-diagnosed, increased effort has been put into understanding its spread. This has led to better tests and procedures developed to detect it in patients who come in with relevant symptoms. Indirectly, this also leads to a variety of answers to the question of "how prevalent is myocarditis?", depending on how old the specific paper is. Recently, I have seen this paper cited in support of an incidence rate of myocarditis in children at 1.13 per 100,000: The Incidence of Pediatric Cardiomyopathy in Two Regions of the United States (cited in this NPR article: Pfizer's COVID Vaccine In Teens And Myocarditis: What You Need To Know). However--crucially--this study was published in 2003. There are more recent studies that have reported some very different numbers, and I think that's important to recognize.
Variability by Age and Gender
One of the things we know for sure is that there is an age and a gender component to myocarditis: at least, reported myocarditis. Importantly for the current discussion, the demographic that is most susceptible to myocarditis is young men. This is something that has also been revealed more clearly in recent years. The 2003 study I mentioned above did see a slightly higher incidence rate of myocarditis in teenage boys compared to other children, but only by a bit less than 2x as much. A newer Finish study (2014), on the other hand, has identified a much higher difference--closer to 18x higher prevalence in teenage boys compared to other children (Occurrence and Features of Childhood Myocarditis: A Nationwide Study in Finland). In what might be considered a companion study of adults, the Finns discovered that the peak incidence occurred in mean aged 16-20 years, declining steadily by age after that (The effects of gender and age on occurrence of clinically suspected myocarditis in adulthood). Here's the key graph from the first study so that you can appreciate the dramatic difference that occurs for this age group specifically (unfortunately, I don't have access to the full data of the second study):
I don't believe it's completely understood why boys in particular have these higher rates of myocarditis, but I know testosterone levels are suspected.
When you have this level of variability in a specific sub-group, this can lead to higher variability in end-results of a lot of studies, because it then becomes a question: how many of a particularly susceptible sub-group did you include within this study compared to that study? Especially when we don't know the exact mechanism which causes teenage boys to be more susceptible to myocarditis, this can make it difficult to get consistent results across studies.
Variability by Region and Viral Background
In the first study I mentioned, there was a reported rate as high as 105 cases per 100,000 people. This specific rate was reported for a particular region: Albania. Are Albanians particularly susceptible to myocarditis? Well, maybe--but on the other hand, it may have had more to do with the types of diseases that had been recently prevalent in Albania at and before the time of that regional study.
The majority of cases of myocarditis in the developed world are thought to be caused by viral infections: a virus attacks the heart muscle in some way, and then the immune system triggers inflammation of the damaged tissue. It's known that certain specific viruses are more prone to causing myocarditis than others, and some of the viruses that have this property are otherwise not very serious. The incidence rates of myocarditis, then, can fluctuate greatly depending on what specific diseases were recently going around in the region under study. This may well have been what happened to Albania here.
Here it is relevant to point out that Covid is one of those viruses that has been shown to cause myocarditis--at a rather high rate, as well (possibly even as high as 25%, depending on the study you look at). Further, we know that Covid has spread much more rapidly among young adults than was earlier appreciated (see here: Difference in Severe Acute Respiratory Syndrome Coronavirus 2 Attack Rate Between Children and Adults May Reflect Bias), and we know that it spreads in this demographic very often undetected. So we have at hand here in the United States a plausible, ready provider of large amounts of occult myocarditis in young adults.
Combined Variability
When you add up all these sources of variability, you get a lot of uncertainty, because each of the causes of uncertainty I've described so far are independent of each other. For example, the 105 cases per 100,000 were only from reported cases. Multiply that by 7.4 as a very plausible ratio between actual and reported, and you get a rate of 777 per 100,000. And if you were looking for the incidence rates for just the young men in that area, well, you could easily multiply that number by 2 or 3 as well because of the increased prevalence in that group. So we are looking at plausible ranges of background myocarditis at anywhere from 2 per 100,000 all the way up to *2300* per 100,000, depending on various things that we don't have a great way of knowing.
Rate of Myocarditis Among Vaccine Takers
Ok, so on that very unsatisfactory note, we now we have to move to the rate of myocarditis that we're actually seeing in young men taking the Pfizer vaccine. The best numbers I've been able to get for the States so far has been from this news report: CDC confirms 226 cases of myocarditis after COVID-19 vaccination in people 30 and under
The CDC is still going through cases they've found on VAERS (as of the time of that news article, they were about half way done), but what they've found so far is:
79 cases of myocarditis/pericarditis reported in teens ages 16 or 17 years after a second dose of vaccine, while the expected number was two to 19 cases, according to Dr. Shimabukuro. There were 196 cases in young adults ages 18-24 years, while eight to 83 were expected.
I know that the first number (79) is out of about 7 million teens who have so far been vaccinated over the course of about 1 month. If they were expecting 2-19 cases from 7 million people over one month, that means that they were expecting myocarditis at a rate of 0.3 - 3.4 per 100,000 per year. That corresponds OK with the older 2003 report I mentioned above, but it does not correlate well with the Finnish data. So I wonder if whoever is analyzing this for the CDC has good updated numbers on myocarditis prevalence, or whether they are using dated information. Then again, it could be argued that the 2003 study, while not as recent as the Finnish study, at least was a study done in the States, and as we have seen, there can be dramatic regional differences in background rates.
Meanwhile, Israel has reported that "between one in 3000 and one in 6000 men ages 16 to 24" who have been vaccinated have developed myocarditis (news story here: Israel reports link between rare cases of heart inflammation and COVID-19 vaccination in young men). That's between 17 and 34 per 100,000. The scientist (Mevorach) is quoted as saying that his is between 5 and 25 times the expected rate, which implies that he thinks the expected background rate is 1.4 to 3.4 per 100,000. That tells me, again, that he might be behind the times when it comes to background rates of myocarditis in young men, as those numbers don't agree with the Finnish study in the least. He might well be using the same numbers the CDC is using? That would seem less valid for Israeli data, though. Anyhow, compared to the Finnish study, the rates reported are more like 1x-2x the expected background rate. That's suggestive of a difference, but given the other possible variabilities here, hardly what I would call conclusive.
Europe has also reported some numbers, but as their vaccine program is heavily behind and still heavily targeting the elderly, their numbers are a lot less relevant.
Adjusting the Reported Rate
So, given the bare rates reported above, how would we need to adjust them to get closer to the true number of cases of myocarditis after vaccination? Unfortunately, this is going to be very difficult to do.
Since VAERS is a self-reporting database, there are likely to be other cases that happen that are unreported. However, no one really has a great handle on what this ratio between reported and unreported incidents is.
We have done research on VAERS in the past that makes it quite clear that the more serious the side-effects are, the more likely they are to be reported in VAERS. About 76% reporting for the more serious side effects was the highest that particular report found (see here: The reporting sensitivity of the Vaccine Adverse Event Reporting System (VAERS) for anaphylaxis and for Guillain-Barré syndrome )
But that high rate of reporting was during the swine flu epidemic, and the researchers rightly pointed out that the high levels of media reporting on the pandemic may have increased reporting percentage in that case. So what are we to expect in the age of Covid, in which media reporting and public concern about possible side effects of a vaccine under an unprecedented Emergency Use Authorization rollout? Especially given the fact that the official guidelines for reporting Covid vaccine related side-effects are much more stringent and inclusive than they ever have been for any other vaccine (see the special "Covid-vaccine-only" official guidelines here: https://vaers.hhs.gov/faq.html )
I think it's very reasonable to expect that side-effect reporting for the Covid vaccines will be much higher than normal, and probably higher than they have ever been. But in reality, no one has actually studied or measured this, and we can only go from intelligent guesses. Personally, I think the rate is likely to be all the way up to 90% or higher--but I don't have much to back that number up aside from feeling. Lower numbers are plausible--for the most trivial of side-effects and for other times, reporting rates as low as 1% have been seen.
I think it's very reasonable to expect that side-effect reporting for the Covid vaccines will be much higher than normal, and probably higher than they have ever been. But in reality, no one has actually studied or measured this, and we can only go from intelligent guesses. Personally, I think the rate is likely to be all the way up to 90% or higher--but I don't have much to back that number up aside from feeling. Lower numbers are plausible--for the most trivial of side-effects and for other times, reporting rates as low as 1% have been seen.
So I could see plausible ranges of reporting ratios here all the way from 10% to 90%, given that the symptoms of mild myocarditis are frequently very mild and could be dismissed as no big deal. So instead of 17 to 34 cases per 100,000, maybe we are actually seeing more like 170 to 340 cases per 100,000.
Bottom Line: What can we say about vaccines causing myocarditis?
Unfortunately, I think the honest answer to this question is, "not a lot". Given the numbers that I am seeing, it is entirely plausible that the vaccines have no causal link to myocarditis at all. It is also entirely plausible that the vaccines are causing myocarditis in young men, maybe even at a rate as high as about 1 in 3000. The problem is that there is too wide a range of plausible answers for us to say for certain what is actually the case.
There can be other clues aside from mere numerical prevalence: for example, there is some coincidence of timing, where apparently there are clusters of cases around 4 days after the second dose. However, I have seen things like this appear chimerically in data on an initial analysis, only to vanish after longer vigilant analysis, so I don't trust this datapoint yet.
I, personally, am very suspicious that this link between vaccination and myocarditis didn't become evident until after the vaccine was approved for 12-16 year olds--precisely the age group that the latest studies are showing to be much more prone to myocarditis than anyone else. I think it's quite likely that this is a false signal based on an underappreciation of the true background rate of myocarditis in young men. However, other more expert people than myself have held that there is *probably* something here, so I admit my opinion is not a consensus. Also, I think that the contrary to my opinion is very possible as well.
So what should we do, in practice? Well, the first thing not to do is avoid taking the vaccine. At a very worst case, the vaccine would be causing myocarditis at a rate of something like 1 in 300--that's multiplying the Israeli number by 10x for all of the very mild, asymptomatic cases we assume might be happening without being caught. If you also account for the same sort of thing on the Covid side, however, you get a rate of myocarditis at 1 in 4. In other words, if you get Covid, you are at least about 100x as likely to get myocarditis than if you got the vaccine, and probably more like 200x-500x more likely. You would have to be *really* certain you were not going to get Covid in order to make those odds work out in your favor.
Another reason to get the vaccine is that the cases of myocarditis we have seen from the vaccine have definitely been on the mild side, and again, that is less the case with what we've seen from the virus. So the incidence numbers alone don't tell the whole story.
Another reason to get the vaccine is that the cases of myocarditis we have seen from the vaccine have definitely been on the mild side, and again, that is less the case with what we've seen from the virus. So the incidence numbers alone don't tell the whole story.
The one practical thing that I believe we can take away from this is a single recommendation: for teens and young men who get the vaccine, I would personally recommend avoiding strenuous exercise for the week following the second dose, just in case. Strenuous exercise is the single most common circumstance that is seen in conjunction with myocarditis that leads to a serious attack. In case there is any increased risk of myocarditis with the vaccines, abstaining from strenuous exercise for the likely critical period seems to me to be a prudent decision.
