brandonb

This study was really interesting to me since it showed that very low cholesterol levels--the average LDL was 43 mg/dL in the study--don't just slow the buildup of coronary plaque, but actually reverse it. About 4 in 5 patients in the study saw plaque volume go down, as measured by CCTA.

Every article I've seen about Galleri mentions false positives as the main objection.

So I ran the numbers. For adults age >50 (the recommended group), standard cancer screening generates 33 false positives per cancer case detected. Standard screening + Galleri detects 236% more cancers with just 8% more false-positives for this age group.

The numbers would be slightly worse for a younger cohort, but still better than I'd have thought.

Lp(a) is stable since it's mostly genetically determined.

ApoB is very responsive to diet and medication--you can lower it by about 85%.

A few reasons:

- If you made a dietary change (e.g., more fiber and less saturated fat), the tests tell you whether it actually worked.

- Statins are dosed based on the LDL or ApoB reduction you need. The most powerful statins cut cholesterol by 50%, but at higher doses the risk of side effects does go up.

- Other medications like PCSK9 inhibitors and ezetimibe exist, and whether to use them depends on the other numbers. For example, PCSK9 inhibitors have a secondary benefit of lowering Lipoprotein(a), and statins and GLP-1s have a secondary benefit of lowering inflammation.

- If blood pressure, HbA1c, etc is high, then there's a slightly different set of diet advice and a very different set of medications. For example, lowering blood pressure through diet is mostly about reducing sodium and increasing potassium.

Coronary CT calcium scans are getting more popular. But before getting one, I'd do these simpler tests first:

- ApoB - about 20% of people with normal cholesterol results will have abnormal ApoB, and be at risk of heart disease.

- Lp(a) - the strongest hereditary risk factor for heart disease.

- hs-CRP - inflammation roughly doubles your risk of heart disease

- HbA1c - insulin resistance is a risk factor for just about everything.

- eGFR - estimates the volume of liquid your kidneys can filter, and is an input to the latest heart disease risk models (PREVENT).

These are simple blood tests. You can get them for <$200 through my company or others:

https://empirical.health/product/comprehensive-health-panel

Imaging is great for detecting calcified plaque in your coronary arteries. But before you have calcified plaque, the above risk factors tell you about the buildup of soft plaque. And 4 out of 5 of them are modifiable through lifestyle, exercise, and medication.

Very fair caveat!

IMO, one of the most interesting things about the new hematology guidelines is that they're likely to get people treatment for symptoms like fatigue or low energy. A lot of doctors consider these "vague" complaints that are hard to diagnose. For some fraction of those cases, low ferritin or iron might be low hanging fruit.

Yep, LDL cholesterol is one of six biomarkers:

* ApoB or LDL cholesterol

* Blood pressure

* Inflammation (hs-CRP)

* Lipoprotein(a) (strongest hereditary risk factor)

* HbA1c (metabolic health)

* eGFR (kidneys)

Modern cardiovascular risk calculators like PREVENT take all of biomarkers factors and give you a 30-year forecast of your risk of heart disease.

Typo—thanks for pointing it out, I fixed the post.

The 55 mg/dL target in this study is somewhere between the 1st and 5th percentile.

70 mg/dL (the control) is around the 10th percentile.

IMO, surprising that reductions at the low end still seem to yield such large benefits. I'd expect it to saturate much earlier.

One more practical application: speech recognition sometimes uses a type of shortest paths search over a word or phoneme graph.

I'm curious-did you have any traditional risk factors (LDL/ApoB, blood pressure, HbA1c, hs-CRP, high PREVENT or ASCVD score)? Or put another way, did you ever find an underlying cause as to why you had a heart attack at age 29?

This is one of the things that makes me a little skeptical of their score. If it's accurate, there should be a positive correlation with known biomarkers. (I see you made a similar observation elsewhere.)

In their paper they actually find an inverse correlation with hs-CRP, which disappeared when removing two patients who were sick. So I wonder if the sample size being small is partly driving this.

Same thing happened with the Lp(a) drugs recently. Practically speaking, statins reduce LDL cholesterol (~50% at the strongest dose) and hs-CRP (~37%) and have cardiovascular outcomes data, so that's likely the route a patient would go based on the medications that are approved today.

This particular study looked at inflammation through imaging (CT scans) as part of Biograph's $17,500 program. You can get at the underlying signal fairly cheaply through a simple blood test (hs-CRP).

It's fairly well-known that inflammation is an independent risk factor for heart disease, alongside LDL cholesterol (or ApoB) and blood pressure: https://www.empirical.health/blog/inflammation-and-heart-hea...

The American College of Cardiology looked at a bunch of inflammation biomarkers and standardized on hs-CRP: https://www.jacc.org/doi/10.1016/j.jacc.2025.08.047

The 4th generation heart rate sensor is quite interesting -- Apple Watch used to sample every ~5 minutes, and now it's every 5 seconds (basically continuous). I think that'll enable a lot of interesting health & wellness use cases.

UCSF puts on a "Mini Med School" series of lectures for the public. If you don't live near San Francisco, some of the video are online. One by Dr. Scheinman is here: https://www.youtube.com/watch?v=Fm3SwORdBS4

Good thing is that all of these conditions have screening and treatments, so you can do something about it.

This is a good addition. I'll update the article. While tadalafil is not FDA-approved for this indication, it looks like current American Urological Association guidelines do recommend it as an off-label use.

(Replied to a comment that was since deleted)

IMO 34.6% is a surprisingly high number, especially in that age range. Most men in their 30's would have a CAC of 0, so being able to observe soft plaque through these newer imaging modalities is telling us something genuinely new.

Whether that's good or bad isn't fully known. We'd expect it's bad: soft plaque is more likely to rupture than hard plaque.

But the study itself measured plaque in a cross-section of participants, at a moment in time. To get a definitive answer, we'd need to follow participants for several decades, observe a large number events (heart attacks, strokes, etc) and then see if soft plaque is predictive above and beyond conventional risk scores.

One thing that struck me about this particular study is that they used two different imaging techniques (CCTA and ultrasound) to measure soft plaque. (This is necessary to capture both coronary and peripheral arteries.)

Lepodisiran is a drug in clinical trials which reduces Lp(a), the strongest hereditary risk factor for heart disease.

This study analyzed the trial results and found it also reduced oxidized lipids, namely oxidized apoB. These are LDL cholesterol particles which are believed to be especially harmful.

Related and worth reading: The Third Culture by the former executive editor of Wired (when Wired was good): https://www.science.org/doi/10.1126/science.279.5353.992

This is roughly how GPUs for neural networks got started: after Andrew Ng left Google Brain, he no longer had access to a 10,000-CPU cluster used to train the original DistBelief system. But his Stanford students could buy a GPU...

Agreed that we can't do as much as we'd like. There are some general health things that are believed to prevent cognitive decline, such as managing cholesterol, HbA1c, blood pressure, sleep apnea, etc. One phrase I've heard is that cholesterol at age 50 predicts cognitive health at age 70.

But hopefully we'll start to have more options and wider approvals for things like anti-amyloid drugs.

This Alzheimer's blood test, PrecivityAD2, is based on the p-tau217 biomarker.

In one recent study, people very high p-tau217 had a 38% chance of progressing to cognitive impairment within 5 years vs 12% for those with low levels. The current tests cost about $200-300, so they're not unreasonable as a screening test. PrecivityAD2 looks to be priced around $1,400-$1,500 so at that price, this specific test likely only makes sense for people with established disease.

Curious-is there a particular diagnosis they should be compensating for in these algorithms?

(OP here) I worked on one of the first studies on detecting diabetes from wearables in 2018. It was kind of fun to review how the field has evolved. The basic approach is the same (pretrain a self-supervised foundation model on sensor data, fine-tune for a particular application) but the scale is several orders of magnitude larger.

Uncompensated care attributable from uninsured patients is ~2–3% of operating revenue/costs. So it's there but not massive.

Those are two real effects, but together they wouldn't compensate for the rate cuts.

About 8% of the population is uninsured. The uninsured population skews younger, with less healthcare utilization (Medicare already covers everyone 65 and older).

Another comment in this thread estimated billing overhead at 8.5%. Medicare for All would eliminate some, but not all of this, since Medicare is still a claims-based system. You would remove a lot of overhead around prior auths, which I agree is a good thing, but could be achieved with more focused legislation.

There's a sketch of part of the proof in "the Mellin ansatz." I'd definitely recommend reading the sections before it, though, since they attempt to establish the more foundational connection between sphere packing and harmonic analysis (which is necessary to understand OpenAI's proof).