back
136 comments
Hey HN!

This project was inspired by a personal experience I had--I put my numbers into a spreadsheet provided by the American Heart Association and found I had a 50% lifetime risk of a heart attack. Yikes!

I've gone to regular check-ups all my life and always been a little outside of the cholesterol guidelines, but I never knew how seriously to take that. I didn't know you could translate those numbers into percentage chance of getting a heart attack, whether I was "high" or "low," or how much I could improve by making lifestyle changes.

I figured there must be other people out there who are similarly ignorant about their own health, so I built this. (Please don't take this as a substitute for going to the doctor; there are all sorts of things your doctor can tell you that a web page can't. This is just for education!)

Please let me know if you have questions!

I've never really understood what life time risk means...

Does 50% mean I'll get a heart attack in my 40s, or mean I'll get one in my 80s instead of getting cancer, etc.

What I'm getting at is that if nothing else kills you, eventually your heart will fail. Shouldn't it be possible to decrease my heart attack risk by increasing my cancer risk?

For those purposes, do you know how smoking marijuana maps to smoking cigarette? E.g. is 1 "dose" of pot = 1 cigarette? 10 cigarettes? Or does it not map at all?

What about electronic cigarettes?

Cool tool, I like the simplicity. Just a bug, if I scroll back up and update the numbers/recalculate, the text changes but the graph does not.
Just wondering how you protect the data? I don't super care, and I almost care more about my email than I do about the numbers.
Is there anything like this that uses ABSI or some other BMI alternative?
This is really just a spreadsheet still with a graph at the end. Would be more valuable if you actually show the ways to raise your HDL, and how to lower blood pressure, etc. Perhaps grab some data from some reputable health websites.
And none of these numbers can be expressed with proper measurement units most of the world uses... Instead we have lbs feet and inches.
I came here to say exactly the same. Maybe you could add a radio button, or better, try to get geographical information from the IP and display the units accordingly (but don't forget about the other two countries that still use the imperial system, the USA aren't yet alone ;) ).

In any case I'm glad that I'm at relatively low risk despite being overweight (but a little less each day). I guess that low blood pressure and zero smoking are that good.

My caffeine intake is also zero (Ok, I think that cocoa powder has some, so not literal zero). I wonder if that has any effect on my risk, maybe something indirect through blood pressure.

Hey! Which units would you prefer?

For height/weight, I'd assume meters and kg?

When you get your cholesterol measurements, do use mg/dL, mmol/L, or something else? What about blood pressure?

And none of this website can be expressed with a proper language most of the world uses... Instead we have English.
Unfortunately, until we figure out an easy way to convert between unit systems, our only recourse is to complain on hacker news.
Based on following the research, I don't believe all these risk factors have been established scientifically. What i've gleaned over the years:

Total cholesterol: probably doesn't matter much

HDL: matters

Blood Pressure: I don't know

Smoking: matters

If you're curious about the risk modeling research literature, here's a good example of models built for 30-year risk prediction: http://circ.ahajournals.org/content/119/24/3078.abstract

The models tend to be close to equivalent in their inputs: either total or LDL cholesterol (small-particle LDL is the one that drives atherosclerosis, but if the model includes HDL cholesterol, then it doesn't make a big difference in accuracy whether the other variable is total or LDL cholesterol), HDL cholesterol (seems to cause a protective effect, although randomized trials of HDL-boosting drugs haven't shown improvements, so may not be causal), blood pressure, and smoking. People also sometimes include newer biomarkers like triglycerides or hs-CRP, but they don't consistently show accuracy gains (LDL, triglycerides, and hs-CRP are highly correlated, so each one individually is a good predictor but having all three doesn't necessarily give you a more accurate model).

There's still a lot of active research to find out which of these variables are causal and which are merely associated with heart disease. For example, LDL-lowering drugs like statins do reduce the risk of heart attacks in randomized trials.

TC is less predictive than HDL:Triglyceride ratio by a significant margin. people with "high" TC of 200-220 and a good HDL:Trig ratio outlive those with "excellent" TC levels.

There are cases where BP matters less as well. For example the cochrane review on sodium consumption indicated that while lowering sodium intake aggressively did successfully lower BP, it resulted in worse mortality.

Right now QRISK2 is considered the most accurate when it comes to predicting morbidity and mortality of cardiovascular events. This calculator is great, but missing one important factor if either of your first degree parents had a heart attack before 60 then you are at greater risk. This mostly ties into things like homocystine levels, hrcp, and LDL particle size, but its a good ballpark for the genetic variation.

My advice get genetically screened if you are worried and make changes. If you carry alterations on the 9p21 chromosome chances are you are at higher risk. If so you can simply eat more fruit and vegetables this has been shown in multiple studies to dramatically lower rates independent of other factors.

What a great way to get detailed demographic data of the HN crowd!
One critical factor missing is immediately family history of early cardiac disease, such as a parent or sibling having a heart attack before age 45
Slick. I liked the additional prompts when entering information.

The graph animation was slow in my Firefox 33.

If you don't enter height or weight into "debug my life" it says: "Your current BMI is 23, which is in the healthy range" instead of "you didn't give us what we need to calculate your BMI".
The current risk in the graph doesn't update when I change a value and click Calculate My Risk again; e.g., if I calculate with blood pressure blank then recalculate with it set to 130.

I'm using Firefox 33 and Chrome 38-ish for Mac.

Just wonder, do people know about kilograms, meters, international system of units?
This is great, brandonb, and I love to see people making tools that raise awareness. If anyone is looking for this in iOS app form, I made something a few months ago: https://itunes.apple.com/us/app/heartgram-physicians-cvd-ris... .
Interesting that family history isn't included. I'd have thought that was a pretty important factor.
Total cholesterol, really (which goes up with HDL, which has no upper limit)? And no triglycerides?
Interesting, but not exactly new. This has been promoted in various ways in the past (http://www.cvriskcalculator.com/ - which is updated with recent BP management data as well)
There was also that thing in the news about some effects of smoking being mitigated if one quit by about age ~32. Still I wonder (as someone who previously smoked) how there couldn't be some kind of damage after several years of smoking
This has got to be off.

"I had a 50% lifetime risk of a heart attack. Yikes!"

not for nothing, but I entered in my numbers and got 68% and I smoke about a pack of day. If that's the case, then I'm not quitting any time soon.

A good metric would be the hours spent per week in doing cardio exercises.
The sensitivity analysis being displayed automatically is a nice feature, as it immediately shows the effects of changes. That's a clever bit of UI that most similar apps don't have.
Minor bug: the heart icon gets clipped if the risk factor comes out to 12% in the latest version of Chrome.

http://imgur.com/l27HcXA

This seems pretty neat. Are you considering turning this into a startup, then? X for Y pitch of "machine learning for personal health"? I say go for it. :D
Great app - thank you for sharing. Unfortunately the answer to this seems similar to most health advice sent my way, "Lose some weight!"
Cool, but it doesn't take into account your drinking habits. 1-2 drinks per day reduces the chances of coronary heart disease by up to 40%.
So far there are only two links to published research in this thread.

That's suboptimal when discussing something where the science is moving.

I don't know if you submitted your app to the Million Hearts challenge, but it would have made a nice fit.

http://www.health2con.com/devchallenge/one-in-a-million-hear...

Are you using the Archimedes Indigo back end? https://archimedesmodel.com/indigo-api

How does BMI affect numbers when I do weight lifting? I.e. I assume I have more muscles than average Joe?
Shouldn't alcohol intake be a major factor as well along with smoking?
These numbers seem slightly off, or the way they are presented is a bit confusing:

"Your lifetime risk of heart disease is 56% as of today. You can reduce that to 38% if you ... That would make your risk about 34% lower. Not too shabby!"

How might I increase my HDL (good cholesterol)?
> Your risk of heart disease is 25%

But the graph shows 26%

Does age play any part into one's risk?
All I can say is big thank you :-)
Nice try, insurance company.
Great domain name.
Please add support for SI units (mmol/L) for cholesterol. Some of us live in more enlightened countries...
There's an interesting controversy around the lipid hypothesis: http://en.wikipedia.org/wiki/Lipid_hypothesis#Minority_disag...
So if I have relatively more muscle mass compared to my height, that means that I am at a higher chance for getting a heart attack? Well, that could be the case for all I know.