Anxiety, heart attack, or stroke?

One of my friends who has regular anxiety attacks asked me to help her disambiguate between anxiety attacks and strokes. I wrote up a brief email, and then figured it might be worth sharing with the broader world. I’m not an expert on this, this is just the result of a little searching, so anything you hear from a genuine trusted expert who knows your particular situation should probably override this, but I expect it’s better than no info at all. I answered the question for her particular situation but it should be easy to modify as needed.

Anxiety attack or stroke?

Assuming symptoms are identical

Anxiety attacks happen to my friend at least every 4th day. Strokes happen to 0.7% of women aged 20-39 ever (she’s in this category), and that's over a 20 year range. So the average woman 20-39 has a stroke once every 7,000 / 0.007 = 1 million days. That means that on any given day, my friend has a 1 in 4 chance of an anxiety attack, and a 1 in a million chance of a stroke. Therefore, if she’s having symptoms characteristic of a stroke or anxiety attack, there’s a 1 in 250,000 chance that it's a stroke.

Using symptoms to disambiguate

After some more investigation, I found a website (archive here) with some concrete info on how to tell the difference between strokes and anxiety attacks. Strokes are sudden and more typically involve asymmetrical paralysis, while anxiety attacks fade in, and more typically involve shaking or a racing heart. My best guess is that asymmetrical paralysis is a relatively strong indicator of a stroke.

Anxiety attack or heart attack?

Assuming symptoms are identical

About 0.6% of women aged 20-39 have a heart attack, roughly the same base rate as for strokes, so the numbers are about the same - if anxiety attacks and heart attacks had the same symptoms, it should turn out to be a heart attack about once every quarter of a million times.

Using symptoms to disambiguate

Unfortunately, heart attacks are famously similar to anxiety attacks, so we don’t get much additional info from this. However, it looks like anxiety involves whole-body effects, tingling in the extremities, or a sense of tightening in the neck more often than heart attacks do, so if she experiences either of those, she should somewhat lower her estimate of the chances of a heart attack, and if she experiences neither, she should slightly raise it. Overall, this review suggests that it’s pretty hard to tell and you should pretty much go on base rates:

When looking at specific characteristics of chest pain, panic disorder has frequently been reported in those having atypical angina or atypical chest pain. This review supports these findings. However, panic disorder has also been seen in patients with typical angina (4% to 65%). Further complicating the angina-atypical angina link to panic disorder is that 10% of patients with ischemic chest pain have panic disorder. Conversely, only 64% of patients with heart disease have chest pain, atypical chest pain is seen in coronary artery disease, and only 79% of patients with significant coronary artery disease have angina. In emergency department patients with acute chest pain, angina was most common in those patients with both panic disorder and acute ischemia. However, there are studies that have either failed to find an association between panic disorder and angina or have found no difference in the prevalence of angina and atypical angina (19%) in patients with panic disorder. Thus, although certain atypical features may suggest panic disorder, many of the characteristics classically associated with CAD are common in patients with panic disorder or anxiety.

[...] Even when panic disorder and [coronary artery disease] coexist, the distress perceived by patients with chest pain is typically caused by the panic disorder.

How accurate is this?

Probably not very. Probably within an order of magnitude. This is mainly just a demonstration of what you can do with some simple statistics, put up on the internet because it’s cheap to do so. Please let me know if I am making any substantive errors!

It’s pretty difficult to find useful medical information online on some things. For example, advice about disambiguating emergencies is pretty terrible. Much of it includes useless disclaimers like, “if you need additional help, go to an emergency room / call 911”. These are useless because they don’t actually give you diagnostic help to figure out whether to take that step. Hopefully this helps fill that gap a tiny bit by collecting a bit of the info out there with decision-relevant content.

Leave a Reply

Your email address will not be published.