Just be aware that the AI may be hallucinating. I’d go with Bempedoic Acid and Ezetemibe with your CAC and LDL. However, we all have our own preferences. Either way, I wish you a long and healthy life!

5 Likes

It depends on your prompt, if you ask it to reduce your risk of MACE over 10, or 20 year period, or longer and consider and prioritize correctly causal factors, and independent risk factors, it might answer differently. Best practice is to work on creating the prompt with it that you’ll feed it back to it + your data and when it should reason and about what + search. The prompt can be really long and detailed.

3 Likes

MACE can put someone in the retirement home and end healthspan. Independent + causal risk factors is important for this and to weigh factors correctly. Time period is also important (is it 10, 20, 30 year risk, or longer?).

1 Like

You might also try ChatGPT Research AI version. Please let us know if you see much variance in the recommendations.

1 Like

I believe this is true, as far as ASCVD goes. This is the analogy of somebody being tied down to the train tracks and simply getting them off the tracks extends their life. It seems to be simple enough that of you bottom out ApoB (and others like Lp(a)) then you basically won’t develop ASCVD in your expected lifetime.

However, other variables like inflammation, blood pressure etc matter for overall cardiovascular health. For example, high blood pressure alone can cause ventricular wall thickening etc. Infections can cause myocarditis. There’s plenty of CVD outside of ASCVD. But you’re right - I think confusing comes from the fact that atherosclerosis is so common and predominant that ASCVD becomes equated with all CVD.

There honestly isn’t. The science is pretty damn settled, particularly at the population level. What makes a difference is the individual level. Some people, for whatever reason, are very resistant to atherosclerosis. They can have mega high LDL-C/Apo-B and zero atherosclerosis. And some people can generate plaque with LDL-C of 70, which is below most current therapeutic targets. However, the overall science of how lipids are involved in atherosclerosis is not really subject to back and forth - it gets more solidified each year.

I remember in your case that you are 70-something, have high LDL, but a relatively low calcium score for your age, plus a healthy lifestyle. So in your specific case there is less argument in favour of statins, since you seem to not accumulate much plaque. However, there are users here in their 40s with positive calcium scores, and they’re hopefully expecting to live another 40, 50 or 60 years. For them, statins (or BA, PCSK9i, Ezetimibe etc) should be a no-brainer.

In your case, if I were you, I’d still look to bring down that LDL-C level through some easily-available pharmacology. A CAC of 108 is not zero, and CAC only tells us about advanced disease. ASCVD is responsible for ~50% of all deaths, and you already have it to some extent. Thus, I’d look to minimizing that risk as much as possible, since you’re presumably hoping to be alive 20 years from now.

4 Likes

@relaxedmeatball what do you think about ATTR prevention in the scenario of already preventative treatment (e.g preventing atherosclerosis with low apoB, optimal BP, obesity, T2D, avoiding CKD, etc)? I found only one study that 70% of supercentenarians die from it, and prevalence is 25% for those 80+ or so. If a graph is drawn from that age to 110+ it seems quite prevalent with time. Seems possible to stop the misfolding with e.g drugs and maybe this will make a lot more people live to 80, 90, 100, etc, especially if already doing a lot of other things right. But I don’t know much about it. One drug seemed to have low side effects.

3 Likes

I’m sorry I don’t know enough about it say anything with confidence. But I have two colleagues who are cardiologists, and I’ll ask them and try to get back to you. Next March I’ll be at an aging/neuroscience conference so I’ll ask around there too.

2 Likes

Did this per your suggestion - no difference in recommendation

1 Like

The science is pretty damn settled, particularly at the population level. What makes a difference is the individual level.
This is the key. My reading of the science is that all (except perhaps the LMHR study) studies of statins have cohorts of metabolically unhealthy subjects.
I happen to be quite fit for a 72 yo and metabolically healthy - based on various functional metrics have the biological age of a 35 yo and live at high altitude to boot. BP 104/68, BMI 18, VO2 max 50, insulin of 3, ESR of 2. I also take tadalfil at low dose, 3 gms of omega 3 and 10,000 fu of natto which help.
Statin kill me - tried a 20 mg dose for 10 days earlier this year and knocked my fitness score off 30% (I have a Carol bike which measures this)

Honestly, every physician has a bias based on “standard of care”. I find these latest AI’s to be the best Dr I have ever visited

4 Likes

I find the medical AI’s to be smarter than my doctor. They are certainly useful as a second opinion. One place they are useful is determining which is the best medication.

My medications are based on AI suggestions, which I then talked over with my doctor and got the suggested medication.

One example was replacing losartan with telemisartan. My doctor readily agreed it was the better choice. I didn’t pursue the question, “Then why the hell did you prescribe losartan instead of telelmisartan in the first place?”

8 Likes

There are incredible time pressures due to the corporate overlay and metrics. They have 10-15 minutes for an appointment, including their documentation time. It has to be a quick and focused time and initiating a change in medication requires the mental energy and knowledge, along with having the time to explain why a change is sensible.

I see a lot of people on medications that have just been continued, and no one actually looks at them and makes changes as new options are available or more data comes forward.

When I review patients we go through this and make recommendations for changes. Many people have been on the same medication for 20 years. It is unlikely that something given 20 years ago is still (or ever was) the best choice.

But doing this and starting to take a comprehensive approach to healthcare takes at least 45 minutes in review and charting, plus 60-90 minutes in a visit. Even then, we have things to revisit and dive into more detail on. In that same time, there would have been 8-9 visits in standard primary care.

Considering all this, it’s not unexpected that patients need to research and be pro-active, because the system is reactionary, not pro-active, and goals for “normal” not “optimal.” The average patient I see for a heart attack had “normal” blood work the day before. Anyone who assures their patients that everything is good by basic blood work (and many doctors do) is not taking a scientific approach.

14 Likes

I so wish you were available via telemedicine to people in New York. But I know NY has vast red tape making it close to impossible. I have to move to another state some day.

1 Like

Intermediate-risk participants with CAC score greater than 0 and less than 400 were randomized to usual or CAC score–informed care (which included 40 mg of atorvastatin). The CAC score–informed group showed a sustained reduction in total and low-density lipoprotein cholesterol and less progression of total, noncalcified, and fibrofatty and necrotic core plaque volumes at 3 years. The association of CAC score guidance with these plaque volume changes was independent of baseline plaque volume and risk factors.

Effects of Combining Coronary Calcium Score With Treatment on Plaque Progression in Familial Coronary Artery Disease A Randomized Clinical Trial 2025

7 Likes

This is interesting, and seems to be more of a psychology study rather than a biological study.

I interpret this as: showing patients (and their doctors) a positive CAC score scares the crap out of them and they actually take their medication and care about their health more.

2 Likes

Agreed. The frontier LLMs, especially with the search/research capability are so powerful. Just in the last month, they have gotten so much better.

2 Likes

That’s not how I understood the study. I thought it was: in one group we do the usual (so nothing unless you have a stratospheric LDL) and in the other one we use the CAC to start a statin earlier. But that’s not very clear and you might be right.

1 Like

Liz Parrish has done the PCSK9 gene therapy. I think about 5 years ago.

Get a mail box in a state where your fav Dr can do Telemedicine and a burner phone from that state :slight_smile: but don’t tell anyone :slight_smile:

3 Likes

Which AI tool did you use for that? I want to try it for my list of medications.

1 Like

https://www.openevidence.com/, https://www.droracle.ai/ Also, the Chinese Deep Seek seems to be in agreement. https://chat.deepseek.com/

5 Likes