A Preventive Cardiologist on the HRV Question.

It is a real signal with a very honest measurement problem. Heart rate variability is a readout of autonomic tone, and autonomic tone is meaningful. Lower variability, under most circumstances, is a system working harder than it should be to hold a resting state. Higher variability, under most circumstances, is a system with room. That part is real. The problem is that the measurement on a wrist is noisy, and that the one-morning reading on the app is often misread by the member. The app has trained a generation of otherwise-analytical people to read a single data point as a verdict, and that is not what a single data point is.

As a trend over fourteen to twenty-eight days, compared to a member’s own baseline. Not against population norms. Not against last Tuesday. Against their own average over a month. A sustained drift of fifteen to twenty percent below their baseline is a real signal. A single morning outside their range is almost always noise — dehydration, poor sleep, a glass of wine, a head cold. The pattern matters. The pixel does not.

The wearables miss ventricular function, coronary plaque burden, lipoprotein particles, blood pressure trajectory, and structural heart disease. In other words, almost everything that actually kills people from cardiovascular causes. The devices are very good at heart rate. They are decent at heart-rate variability. They are pretty good at sleep staging. They are poor at blood pressure on the wrist, and they cannot see inside the chest at all. The most important conversation I have with a new patient is usually the one where I tell them the device has been telling them the wrong story for two years — not because the device is lying, but because it is measuring the wrong thing, confidently.

ApoB. Coronary artery calcium score. And systolic blood pressure measured correctly, meaning at home, seated, at rest, multiple mornings, with a validated upper-arm cuff. Those three, taken together, predict cardiovascular mortality over the next two decades better than the wearable data of a lifetime. The first is a single blood test and it is cheap. The second is a single imaging study between ages forty and fifty-five for most patients. The third is a twenty-dollar cuff and a habit. None of them are on any Apple Watch I have ever seen.

Because apoB counts the particles. LDL cholesterol measures the cargo inside the particles. In a patient with small dense LDL — very common in the metabolic-drift population — you can have a reassuring LDL number and a frankly elevated particle count, and it is the particle count that deposits in the arterial wall. ApoB is a better read-through. It is available, it is inexpensive, and the cardiology literature has been clearer about it for a decade than the popular guidelines reflect. I order it on almost every patient. I would like to see it become standard of care. It is not yet.

Because it requires a referral, a CT scanner, and about a hundred dollars, and none of those are in the consumer-wellness supply chain. The test takes ten minutes. It gives you a picture of what twenty-five years of cardiovascular behavior have actually done to the arteries of the heart. A zero score at fifty is very reassuring. A score above a hundred at fifty is a conversation that changes what I prescribe and how aggressively. The test has been validated across hundreds of thousands of patients in the MESA and CAC Consortium studies. It should be more routine than it is. I write the referral for almost every appropriate patient. About a third of them go. I keep writing it.

Charitably: a desire for agency in a system that is not set up to give them enough of it. Less charitably: a very effective marketing funnel that has outpaced the evidence. There are a few supplements I think have a reasonable case — creatine, for example, in older adults for strength and cognition, and fish oil at the right dose in patients with elevated triglycerides. There is a much longer list where the case is weaker than the enthusiasm. I spend a lot of my consults simplifying a patient’s supplement list. We take things away. The patient usually feels the same or better. That tells us most of what we need to know.

Time. I have twelve minutes with a patient. Dr. Chaudhari has fifty. That is not a difference of quantity, it is a difference of medicine. In twelve minutes I can rule out an emergency, I can review a recent test, I can adjust a prescription. I cannot ask what time their bedroom light turns off. I cannot ask about the third cup of coffee. I cannot walk them through the five years their numbers have been drifting before the threshold was crossed. The hour is where the preventive conversation actually happens, and the healthcare system is not built to pay for the hour. Private facilities that are doing this carefully — and I mean carefully, not as a supplement-pushing front — are doing something medicine should be doing more of. That is an uncomfortable thing for an academic to say. It is still true.

Continuous glucose monitors in non-diabetic adults. The data is interesting and almost none of it has been shown to improve cardiovascular outcomes in the metabolically healthy. I expect a correction in the next three years, after a wave of members who chased every spike on the graph and made themselves anxious without making themselves healthier. The signal in a CGM is real in the metabolic-drift population and in diabetics. It is noisy-to-meaningless in a forty-year-old with a normal fasting insulin. Use the right tool on the right patient. That is most of clinical reasoning.

The move from mouse models to human recommendations. The mouse data is often beautiful. The translation to humans, at human doses, over human lifespans, is almost always more modest than the podcast length allows. Rapamycin is an example. Metformin in non-diabetics is another. There are careful voices on these questions — Attia is more cautious than his audience, to his credit. But the audience takes the enthusiasm and loses the caveats. I spend a lot of clinical time returning the caveats.

Measure their blood pressure at home, sitting, at rest, five mornings in a row, with a validated upper-arm cuff. Then average the readings from the second through fifth. If the average is above one-twenty over eighty, call their physician and have a conversation. That is the single highest-yield thing a forty-to-sixty-year-old can do this week for their cardiovascular future. The cuff costs twenty dollars. The five minutes cost nothing. The conversation is the point.