Association between influenza vaccination and risk of stroke in Alberta, Canada | Hacker News Reader

Association between influenza vaccination and risk of stroke in Alberta, Canada

ucalgary.ca

87 pointsgeox89 comments

The study: https://www.thelancet.com/journals/lanpub/article/PIIS2468-2...

mentions that the Hazad Ratio is reduced:

"Adjusted for demographics and comorbidities, recent influenza vaccination significantly reduced the hazard of stroke (hazard ratio 0·775 [95% CI 0·757–0·793])."

I had to look up what a Hazard Ratio is: https://statisticsbyjim.com/probability/hazard-ratio/

So many improvements seem to be quite meagre, I am not sure how to understand a HR but it looks to be a huge help.

One interpretation would be 1.000 - 0.775 = 22.5% decease risk of stroke after recent flu vax.

And of Alzheimers.

I wonder at what age flu shots provide overwhelming benefits. 40? 50? I'm still reasonably young and don't bother, but maybe I should start getting them.

https://pubmed.ncbi.nlm.nih.gov/35723106/

Also discussed here https://news.ycombinator.com/item?id=31899781

Now that the evidence seems pretty clear that it reduces your risk of dying from cardiovascular events by ~30%, you should probably just get it every year:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5256393/

It's like bike helmets. I didn't always wear them as a kid because they were uncomfortable, they weren't that effective, and we didn't know how bad head injuries were. But now that they're extremely comfortable, extremely effective, and we know how bad head injuries are, it doesn't really make sense to wear them any less than 100% of the time.

100% of the time? I don’t think I’ve ever thought about it, but you’re right. Everybody should be wearing bike helmets all the time to protect against head injuries.

Really, we should change the name “bike helmets” to something else if we want people to change their thinking and behavior. Perhaps “personal safety devices”?

We should mandate them!

If we just stopped one of the 1.5 million brain injuries each year — mandating permanently wearing helmets for the entire population is worth it!

https://www.ncbi.nlm.nih.gov/books/NBK459300/#_article-31648...

> If we just stopped one of the 1.5 million brain injuries each year — mandating permanently wearing helmets for the entire population is worth it!

I don't think most people would agree with you. If we passed a law in the US mandating no car can be capable of traveling faster than 25 mph, we would save 30,000 lives per year. But we don't do it because driving at higher speeds is worth the trade off in lives.

Yes — I was being sarcastic, by using the logic we heard justifying QALY negative public health measures during the pandemic.

I probably shouldn’t have used sarcasm online; it doesn’t translate well.

This claims flus cause heart attacks, so the flu vaccine reduces heart attacks by reducing flus. In my own case I wonder if there'd be any benefit, because I've never gotten the flu despite taking no precautions including shots.

Many influenza infections are asymptomatic. You might have been infected at some point without realizing it.

So?

So the greatgrandparent can at best say they have never had symptomatic flu. They cannot confidently say they have never gotten the flu. It is flawed reasoning to use no symptoms of flu = not gotten the flu = not getting vaccine.

This is not medical and vague argument.

Symptoms are what you "fight" against by medicine. Nonsymptimatic could be basically anything.

To stay with the helmet analogy, this sounds like "I won't wear a helmet because I've never been run over."

It depends a little on how you react to the vaccine, but just avoiding the flu once in a while is likely enough to be an overwhelming benefit.

just avoiding the flu once in a while is likely enough to be an overwhelming benefit

No one is getting the flu on purpose, if people could avoid it they would.

It depends a little on how you react to the vaccine

Statistically it doesn't and this is about statistics, not 'what if' scenarios. What is this supposed to mean?

There's people that claim they are impacted for days by vaccinations. I generally notice a little tiredness. I don't care to argue with them about whether they are actually impacted for days, but if they are it certainly changes the comparison vs being a little tired for part of a day.

And clearly people don't care to avoid the flu, vaccination rates are pretty low!

Note that they don't have a double bling randomized control group. We know this type of studies create a lot of false positive, for example there are many studies without a double blind randomized control group that "proved" that ivermectin was useful against covid-19.

The problem is that it's very difficult to compensate for all the other variables. They even get more strokes in the vaccinated group than in the unvaccinated group. This is not so surprising, because old people and ill people are more encouraged to get the flu vaccine. After correcting for the age, they even get less strokes in the vaccinated group than in the unvaccinated group.

It's difficult to get all the variables that are important to be sure they are not forgetting to make more adjustments. For example, people without access to a good medical system (because they live far away from a city or are too poor to take a day to get t the doctor) are probably less vaccinated, less checked and have probably more health problems.

Correcting for all the unknown unknowns is too difficult. So the trick is to use a double bling randomized control group, to try to avoid other factors that select who is vaccinated od who not.

"...less checked and have probably more health problems."

More checks = more possibilities to find some kind of problem. So statistically, in wealth countries diagnoses are higher in measured numbers.

Why not link study directly?

https://www.thelancet.com/journals/lanpub/article/PIIS2468-2...

I may be reading it wrong, but when I look at chart "Table Characteristics stratified by having ever received any influenza vaccination" the Never received influenza vaccine group has much less diseases - Atrial fibrillation 0.67% vs vaxxed 2.12%, Chronic obstructive pulmonary disease 1.52% vs vaxxed 4.1% and so on and then we reach "Any stroke event" 0.52% vs vaxxed 1.25% while not that much hiugher average age 48.34 vs 38.98 (unvaxxed).

So even if they suposedly lower the risks within the vaxxed group, by joining the vaxxed group your risk actually grows compared to staying in unvaxxed group and I don't really see odds for both comparable groups.

> Due to the nature of this research using administrative data across a population, participants of this study did not agree for their data to be shared publicly, so supporting data are not available.

I would take any study coming from authoritarian Canada, which froze bank accounts of people opposing COVID lockdowns, with big grain of salt.

Also even according CDC flu vaccine efficacy is only 40 to 60% and that's assuming it works at all, if they don't match the correct variant you are somewhere around 0% efficacy, might as well take placebo with same effect.

https://www.cdc.gov/flu/vaccines-work/vaccineeffect.htm

Right under the chart you reference, the study states that these figures are confounded:

> [...] this crude observation was confounded. Adjusted for age, sex, comorbid illness, and socioeconomic status, recent vaccination (within 182 days) was associated with a reduced hazard of stroke

It's scary how many people don't know the difference between raw data statistics (which are almost always worthless) and a multivariate regression with control variables (which can actually tell you something about the data). Everyone can read a paper and look at numbers, but to understand what is really happening you do need some formal education that most people simply seem to lack. Even people with this kind of education sometimes step into pitfalls, but normal people drawing conclusions from these things is like running through a minefield while blindfolded.

> from authoritarian Canada, which froze bank accounts of people opposing COVID lockdowns

This hyperbole discredits the rest of your comment.

I have no idea how an they "adjust" for a fact people who had vaccines perhaps in general are more familiar with the heath are system, because they might have more dealings with it in the past. Other vaccines they took alongside this one, other illnesses. Or the opposite way, perhaps people who took those vaccines also take better care of what they eat, do more exercise etc.

I wouldn't read too much from this one study. It really takes a lot more to have some level of certainty.

Flu vaccine increases risk of shingles: https://onlinelibrary.wiley.com/doi/full/10.1111/irv.13055

“There was a slight increase in risk of HZ in people receiving influenza vaccine in the first 1–15 days after vaccination.” That’s kind of overselling it, isn’t it? The OP study says that stroke chances are lowered for 6 months. I think many people would take that trade off.

Figure 4D in the stroke study shows how small the hazard is for younger people, and how small the difference is. The shingles study shows the shingles IRR is higher for people under 65. As another comment says "I wonder at what age flu shots provide overwhelming benefits.", that is the reason why I posted it.

I actually got shingles 2 days after flu vaccination in 2017. I had chicken pox twice as a kid and the common idea (old wives' tale?) at that time was if you got chicken pox you wouldn't get shingles. I don't know that the vaccine contributed or not, but it's an odd coincidence. Luckily it was a mild case handled by amoxicillin.

It’s actually the other way around. Going through chicken pox as a kid increases the chance to have a shingles as it is caused by the same virus being dormant in your body.

When infected the first time you can have chicken pox as an adult.

Amoxicillin is an antibiotic and can be given for the risk of opportunistic infections but it doesn’t treat a viral disease by itself.

Good to know! I suspected that was the case since it is a variant of herpes, dormant in most cases but with the potential to reassert itself when the immune system is struggling.

The anti-Vaxer threads were flagged and are dead, presumably because someone considered them dangerous.

I fear that doing so mostly just emboldens them and helps their cause. Censoring like that makes me, and probably others, suspicious.

The better thing to do would just be to help them understand the statistics and provide links + explanations for how they are misinterpreting the data.

Convincing suspicious vaccine-skeptics of the value of vaccines is not the goal here. We're not a public health service; we're a forum for curious conversation. Tedious rehashes of antivax arguments aren't curious; they're just tedious.

I would consider discussing statistics from a study to be pretty relevant for this site, especially since you can just ignore or collapse a thread if you don’t want to participate.

I’m not even saying I’m suspicious of anything related to vaccine efficacy. I’m saying that it doesn’t look good to shut down people completely like that and that it would be better for society if we tried to correct misinfo instead of censoring people.

Again: the goals you're referring to here aren't those of Hacker News. There's a whole wide Internet to prosecute these kinds of arguments on; I hear Twitter is newly receptive to them. Here, our concerns about how most compellingly to correct misinformation must give way to the prime directive, which is curious conversation.

But there are a lot of curious things about vaccines, and plenty of those curious things are reasons why somebody might not be excited at the prospect of getting one.

If they're the same "curious" arguments over and over again, the conversations aren't "curious" in the sense the guidelines refer to.

I for one find constant links about the benefits without any freedom to actually discuss the trade offs incredibly tedious.

I’m a scientific individual, but I don’t appreciate getting nagged.

Edited to add:

I find the repetition of silly debates tedious, too. Like open source licenses, android, iOS, etc. But that doesn’t mean I don’t learn things from some of those debates.

Also, I want to know all the information and know that I know it. I currently know that attitudes like yours are keeping me from learning all there is to know about various vacccines. Because any information that could be interpreted negatively is verboten.

Edited to add one more thing: I was speaking about interesting scientific articles that reveal and discuss the downsides. You implied that I was talking about arguments under those articles being interesting. You do see the difference between those things, right? Banning information because some people react badly to that information is what I’m objecting to.

Banning information because some people react badly to that information

This is done on HN all the time or it wouldn't work at all. It's not a meaningful argument about how to treat particular topics.

Additionally, disproving some bullshit claim takes a lot of effort - while coming up with further claims is rather easy. It's a very, very uneven battle and one not worth starting. Furthermore, some of the antivax are genuinely misinformed and (to some degree) expressing rational doubts, but there are also a lot of trolls who just enjoy wasting your efforts. They have a lot of time on their hands, and there will always be more trolls tomorrow. Don't bother.

Most HN threads are tedious and just rehashes of the same opinions people in the SWE world have had for the past decades. If we truly flagged tedious stuff regardless of personal preference you'd see far fewer posts and comments on this website

I flag tedious technical arguments all the time; best case in point: language war comments on threads about Go or Rust. I'm just responding to the parent comment, which suggests we have a duty to entertain stupid arguments to better society. No, we don't!

Certainly, a thoughtful comment expressing skepticism about a study has a home here; for example:

https://news.ycombinator.com/item?id=34177238

> Tedious rehashes of antivax arguments aren't curious; they're just tedious.

I find authoritarians shushing curiosity to be tedious.

I do appreciate the variety to which people like yourself justify bad behavior, though:

You’re not an anti-intellectual silencing debate on a matter of public health — you’re stopping “tedious” conversation that “bores” you.

I think this is a great example of the banality of evil — and has certainly made me lose a lot of respect for you.

How dare you.

Yep — I get it.

You’re going to treat it as a joke that during an emergency people like yourself were authoritarians who silenced public debate — leading us to enacting QALY negative public health measures, killing hundreds of thousands.

You do that because you can’t admit what that reality implies about who you are — so you’ll cling to your pretense. Like a Chinese citizen saying it’s “tedious” to discuss the plight of Uyghurs. And you’ll walk away high on your own feigned moral virtue.

That’s the banality of evil — as you’ve so wonderfully demonstrated.

Can you please stop posting flamewar comments? This kind of thing isn't allowed here for what should be obvious reasons if you'll read https://news.ycombinator.com/newsguidelines.html.

I realize it's frustrating to hold a minority/contrarian view but if accounts break the rules this badly, we end up having no choice but to ban them.

The anti vaxer threads were flagged and are dead because they were scare spam with no substantive facts or references.

I'm not a medical expert but isn't this obvious? Severe infections are hugely stressful for the body. Influenza vaccination leads to fewer severe infections which leads to lower risk of a multitude of diseases.

It’s still nice to have evidence to back up obvious things, because sometimes we find counter-intuitive results.

I agree. My comment wasn't meant to criticize their work but the anti-vax critics.

> The crude incidence of stroke was higher among individuals who had ever received an influenza vaccination (1·25%) compared with those who had not (0·52%; table). However, this crude observation was confounded. Adjusted for age, sex, comorbid illness, and socioeconomic status, recent vaccination (within 182 days) was associated with a reduced hazard of stroke (hazard ratio [HR] 0·775 [95% CI 0·757–0·793]).

I doubt that the result of this paper is valid.

The crude results have more strokes for people who vaccinated, but they used statistics to demonstrate that vaccination helps. Well there are a lot of choices to make when correcting for factors, and it’s quite hard to evaluate truth that way.

The thing which sticks out particularly is correcting for cormorbid conditions. If you do that, you could very well just be pushing out sick people. Two kinds of people will vaccinate more: people who are sick and get encouraged by doctors they see more often, and people who pay more attention to their health. If you “correct for” people with comorbid conditions you might have a larger influence of healthy people who pay attention to their health.

In other words, it is plausible that the actual results of the study are “being the sort of person who vaccinates is associated with lower incidence of stroke” not because the vaccine has any effect but because general things people who try to take care of their health do tend to prevent a few strokes (like going to a doctor who notices high blood pressure and getting treatment)

Yes, this is quite a big deal additional studies to check various aspects would be warranted to confirm this.

Plus, it’s possible that the group that has comorbidities is the larger group. The “correction” could then hide the fact that encouraging people to get the vaccine will result in more people having strokes overall.

> I doubt that the result of this paper is valid.

Agreed. Or at least highly suspect.

Correcting for confounding factors is not an exact science. There is a large degree of "art" to it.

When the raw data says "inverse effect" then you "adjust for confounding factors" and it says the opposite, that should be viewed with suspicion.

I agree that stuff can get washed out during the comorbidity control process, but I thought that the comparison of vaccination status + each comorbidity to be reasonably persuasive (Figure 4 and 5). Certainly not bullet proof, but reasonable.

Since they said that effect disappears after adjusting for things including age, I would guess it’s actually just more likely to be because older people are both far more likely to have strokes, and also more likely to have ever had a flu vaccine…

Not trying to be funny, but it's important that in a study like this, you take into account that other causes of mortality might rise, therefore reducing the number of eligible people to die for a stroke, or killing someone faster than an eventual stroke.

100%

Quite a few comorbities were controlled for. In statistics controlled for means "no, this isn't the reason you're seeing this number." Rather than making vague and unsubstantiated claims. Why don't you tell us what mortality might have risen that you believe confounded the results? And back it up with a study. These are the factors that were chosen a-priori: age, sex, anticoagulant use, atrial fibrillation, chronic obstructive pulmonary disease, diabetes, hypertension, income quintile, and rural or urban home location.

So unless you have some magical mystery comorbitiy, it was controlled for and therefore your comment is vague confusion peddling.

> We know that upper respiratory infections often precede heart attacks and strokes. Preventing or reducing the severity of influenza provides a protective factor particularly for stroke

So it is really just a secondary benefit to protecting against the flu. One could say that wearing a mask reduces stroke risk for the same reason, or washing hands thoroughly. Anything that reduces the chance of getting the flu.

You can reduce your risk of stroke by increasing the risk of dying by other means.

This strengthens the hypothesis that a lot of conditions with unknown origins actually stem from dormant viral infections or their leftovers, like how EBV causes MS.

The annual flu shot is a guess. Due to manufacturing lead time and distribution the prediction for which strain will be the dominant strain in the future is made (afaik) approx a year earlier.

The point being, its possible any give year's jab is not the ideal jab given the strain that ended up the actual dominant.

Point being, the study should account for that, or at least explain the significance (or insignificance) of the match / mismatch. That is, is the jab creating an immune response that mitigates stroke. Or is the jab creating an immune response that mitigates flu and less flu means less stroke.

Otherwise, this feels like correlation.

Another variable that causes issues in the results is the effectiveness of the vaccine and the evolution of the virus over years with retained immunity.

For example:

A flu shot from years ago can protect against a flu from this year or in the future. So even if the flu vaccine isn’t effective in its current year, the immunity lasts for many years to those antigens. The strain of flu that’s currently dominant can fluctuate to newer or older variants. Controlling for this is even more confusing

A guess this year could help in future years

Maybe. Reference link(s), please.

Perhaps people who get the flu shot are generally in better health or exercise more?

It's the opposite. The absolute occurence of stroke was higher in people who got the vaccine, but other risk factors were also higher; adjusting for those, they found a "less-higher than expected" amount of stroke, which they attribute to the flu vaccine.

If you have a lot of faith in these sorts of statistical adjustments and the assumptions behind them, then you may find this result compelling.

Funding: None.

Those are good people I guess.

"However, vaccination was associated with a reduced risk of stroke overall across all ages and risk profiles with the exception of individuals without hypertension."

https://www.thelancet.com/journals/lanpub/article/PIIS2468-2...

ROFL So, if you are normal (un)healthy person and you DON'T HAVE HYPERTENSION then this amazing flu vaccine won't lower your risk of stroke at all, this study applies only to people with hypertension.

Yeah, it's actually kind of nuts how much higher your baseline risk of stroke is with hypertension that even though only ~10% of their study population has hypertension that the effect (vaccination correlated with reduced risk of stroke) just shows up on a the global analysis.

This would be a better dunk if hypertension wasn't the dominant risk factor in all strokes, but it is, so you haven't actually made a point here.

I've made point, the headline reads "Flu vaccine lowers risk of stroke", which is very different from "Flu vaccine lowers risk of stroke for people with hypertensia".

The linked article does not mention this hypertensia fact at all, there is not a single mention of hypertensia in it and it is not clearly communicated anywhere in study besides one small fine print, that their findings actually affect only very narow group of people.

You've managed to add "in the cohort of pts likely to have strokes" to the headline, which is not an interesting wrinkle. That's the population you'd most hope to see the effect in.

And if 47% of adults in the US didn't have hypertension.

Only 20% of adult Canadians.

"Seven and a half million people in Canada, about 1 in 5 adults, live with hypertension."

https://hypertension.ca/about-us/

“Without adjustment for other covariates, influenza vaccination was not protective in patients either with or without hypertension. After adjusting for comorbidities, vaccination was strongly associated with a reduced risk of stroke in patients with hypertension; however, there was a non-intuitive interaction, contrasting with previous studies, such that vaccination in patients without hypertension was associated with a marginally higher risk of stroke compared with no vaccination. We observed that the non-hypertensive vaccinated population was much more likely to be female, and that the young average age of the non-hypertensive never-vaccinated stratum was associated with a small number of outcomes. A possible explanation for this finding is that the increase in stroke risk in vaccinated non-hypertensive individuals might represent vaccine-seeking behaviour due to other non-vascular risk factors for poor health outcomes, such as a history of cancer or autoimmune disease, conditions for which we could not adjust. In this situation, vaccination is acting as a latent variable, identifying these individuals at risk of stroke rather than being causative in increasing the risk of stroke. Another possibility is the misclassification of hypertension status by way of undiagnosed hypertension. We also do not have blood pressure measurement or medication data and therefore cannot classify the severity of hypertension or compliance with hypertension treatment, which might influence our findings. Alternatively, this finding might simply be due to chance.”

Note that if you separate a population by enough factors, at least one will show an effect different from the aggregate, by pure probability.

Reply on news.ycombinator.com