BRAINBITES
WEEK OF SEP 11–17, 2026 · ISSUE 27
BRAINBITES — Your weekly health, fitness, and longevity digest. Five minutes that cover the highest-signal podcasts and science before they hit the mainstream.
This week: Michael Davidson told Peter Attia that the argument about cholesterol has been about the wrong variable, and brought his own brother as evidence. We paired it with a 1999 autopsy study that found the disease in people who died at nineteen. Plus Arthur Brooks on the three things happiness is actually made of, a Nature Medicine paper on why cancer is arriving earlier, and our first-ever Trend Barometer.
From: The Peter Attia Drive · JACC · JAMA · Nature Medicine · Huberman Lab
🔢 8 grams — the lifetime dose that appears to buy you heart disease
🎙️ Michael Davidson on two brothers, one drug graveyard, and the 30-year-old nobody treats
🔬 Deep Dive: what they found in the arteries of people who died at nineteen
🌡️ Trend Barometer: the cholesterol-lowering evidence ladder
📰 News Bits: happiness has three macronutrients · why cancer is arriving earlier · the fitness test you've never taken
GM.
Eight grams. That is roughly how much cholesterol has to pass through your arteries across a lifetime before the trouble starts — which makes the number on your last blood panel only half the story. The other half is how long it has been there.
⏱ ~5 min read
🧠 LEVEL × YEARS
Risk isn't a rate you carry. It's a total you accumulate.
Issue 26 made the case that metabolism isn't a speed. The same trick turns out to work one organ over. Cholesterol isn't a level either — it's a level multiplied by the years you've had it, and almost everything odd about heart disease prevention follows from that one piece of arithmetic.
🔢 THE VITAL SIGN · CARDIOLOGY / GENETICS
8 grams
The rough lifetime dose of LDL cholesterol that appears to buy you coronary artery disease.
The arithmetic is almost rude in its simplicity. An LDL of 200 mg/dL held for forty years gets you there. So does 100 held for eighty. So does 80 held for a hundred. Same total, different routes.
That isn't a figure from one trial — it's the arithmetic that falls out of the genetics. Brian Ference and colleagues pooled non-overlapping data on 312,321 people in JACC in 2012, looking at those who inherited gene variants giving them modestly lower LDL from birth. Every 1 mmol/L (about 39 mg/dL) of lifelong lower LDL came with a 54.5% lower risk of coronary heart disease. A statin started in middle age buys roughly a third of that per unit lowered.
Same molecule, same reduction, three times the effect — because the genetic version started at conception. The dose was never the number on the panel; it was that number multiplied by the years.
🎙️ THE EPISODE · CARDIOLOGY / LIPIDOLOGY
The hardest sell in medicine is a healthy 30-year-old.
Michael Davidson on the experiment his family ran by accident, a graveyard of expensive drugs, and why risk calculators tell young people to relax.
Michael Davidson, M.D. · The Peter Attia Drive #407 · 14 September 2026
Davidson runs the lipid clinic at the University of Chicago and heads a company with a cholesterol drug in late-stage trials. Both facts matter here.
Two brothers, and an experiment nobody meant to run.
Davidson's father died of a heart attack at 47. Michael was 16, his brother 14. An uncle checked both boys' lipids; both were equally bad. Michael started a statin the moment statins existed. His brother, also a doctor, went strict vegetarian and left the drugs alone for years. He had a bypass at 44. A sample of two, as Davidson says plainly — but the kind that stays with you.
A graveyard of drugs that raised the right number.
For twenty years the industry chased drugs that lifted HDL, the so-called good cholesterol. Torcetrapib, dalcetrapib, evacetrapib — one after another they raised HDL and did nothing useful, or worse. Davidson's read on the wreckage is that HDL was never the lever, and the survivors in that class are the ones that also lower LDL. His own company's candidate is the test of that reading, and that trial has not reported.
🎙️ Jump to ~27:15 (audio) for the graveyard — the best twenty minutes on how a field talks itself into something.
The calculators are answering a different question.
Standard risk tools forecast ten years ahead, so a healthy 30-year-old scores near zero and is told to come back later. Davidson's objection: nobody is trying to prevent a heart attack at 40. They are trying to prevent one at 75, and the plaque is being laid down now. His counter-tools are unglamorous — family history, Lp(a) measured once, a calcium scan when the answer would change something.
💡 DEEP DIVE · WHEN THE CLOCK STARTS
It's already there
The cumulative-dose argument is usually made from genetics and drug trials. There is a blunter way to make it, and it comes from a morgue.
Between 1987 and 1994, pathologists across a network of American medical centres examined the arteries of 2,876 people aged 15 to 34 who had died of external causes. Not patients — people who were, as far as anyone knew, fine. Jack Strong and the PDAY group published the findings in JAMA in 1999.
Raised fatty streaks — the stage where a smear of cholesterol starts to become a structure — were in the right coronary artery of about one in ten of the 15-to-19-year-olds. By ages 30 to 34, about one in three. The disease was not waiting for middle age. It was in progress, quietly, in people who had never had a symptom or a reason to be tested.
Two honest limits. People who die violently are not a random sample of the living, and a fatty streak is not a heart attack — the flat early ones can regress. But the raised ones do not, and they were already accumulating at an age when every risk calculator in the building would have said there was nothing to discuss.
📈 TREND BAROMETER · SORTED BY EVIDENCE, NOT BY MARKETING
Plenty of things lower cholesterol. Far fewer have been shown to lower your odds of a heart attack. The ladder:
Proven on events. Statins, and on top of them ezetimibe and the PCSK9 inhibitors. Large randomised trials with heart attacks and deaths as the endpoint, not blood panels.
Lowers the number; verdict pending. Obicetrapib, from this week's episode. It lowers LDL, apoB and Lp(a). Whether that means fewer events is what the outcomes trial exists to answer.
Genuinely contested. Omega-3s. The trials disagree with each other, and Davidson — who ran some of the earliest — is far more careful about them than the supplement aisle is.
A statin wearing a hat. Red yeast rice. Its active ingredient, monacolin K, is the same molecule as lovastatin, and one survey of commercial products found the dose varying more than sixtyfold between brands. People who avoid statins buy it by the jar.
📰 NEWS BITS
Arthur Brooks: happiness has three macronutrients, and you are probably eating one.
Brooks splits happiness into enjoyment, satisfaction and meaning, and argues most unhappiness is a deficiency rather than a disorder — plenty of one, none of the others. The sharper half is his list of four things people substitute when the real ones are hard to get: money, power, pleasure, fame. Each works briefly. That is the problem.
— Huberman Lab, Dr. Arthur Brooks, 14 September 2026
Cancer is arriving earlier. One suspect is the clock itself.
Yin Cao's group at WashU compared biological age with calendar age across more than 154,000 young adults in the UK Biobank and over 10,000 in the US All of Us programme. Later-born cohorts looked biologically older at the same calendar age, and the fastest-aging third carried a 15% higher risk of early-onset solid cancers. Immune aging tracked with lung cancer, fat-tissue aging with colorectal. Observational, and biological-age scores are estimates, not readings.
— Nature Medicine, Tian, Cao et al., 14 September 2026
Andy Galpin would like to know your grip strength.
Galpin breaks fitness into nine separate components and points out the obvious thing nobody does: test them before training them. Grip strength, push-ups, movement quality, endurance — most people have measured none of it, and so have no idea which of the nine is holding them back. Untested training is guessing.
— Huberman Lab Essentials, Dr. Andy Galpin, 17 September 2026
What did you think of today's BrainBite?
🧠🧠🧠🧠🧠 Nobody ever mentioned that the years were part of the dose.
🧠🧠🧠 The barometer was the useful bit. Glad it finally ran.
🧠 I'm thirty. Ask me again in fifty years.
We listen to every health podcast so you don't have to. Weekly. No noise.
Level times years. That was the whole formula. See you next week.
