What The Compounds Actually Do

Attia takes between five and seven compounds. He says the quiet part out loud at talks: most of what sits on the shelf in a pharmacy in Zurich or a gym in Los Angeles is expensive urine. He is not a purist about it. He is a filter, and the filter has three buckets. Strong evidence, do it. Mechanistically plausible, low risk, some data, maybe. No good evidence, avoid. That is the entire system, and it moves more money than most funds. I have spent sixteen years of field work in places where the evidence standard was not a preference but survival: rural Nigeria, post-conflict Serbia, Kazan, Fukushima. I deployed 210 energy systems across Africa and Asia. When a component fails in those conditions, nobody argues with the spec sheet. You read it. Either the number holds in the field, or the field eats the number. The supplement aisle does the reverse. It sells the mechanism and hides the outcome. Take the most successful category of the last decade: peptides. Tesamorelin has real randomised data, roughly 15 to 20 per cent visceral fat reduction, and it is approved for one narrow indication, HIV lipodystrophy. The fat returns after you stop. Epithalon rests on small Russian studies that were never replicated in the West. BPC-157 has convincing rodent work and zero human randomised trials, and the pharmaceutical-grade version people imagine does not exist. GHK-Cu has test-tube data and no measured outcomes in humans. Three of those four are sold to healthy men in their fifties at a premium. The fourth is sold as a subscription.

The sentence he repeats is a threshold, not a joke. Show me the randomised controlled trial, or do not bother. Apply it and the list collapses. Ashwagandha sells a cortisol story. The better trials show something like 15 to 25 per cent reduction, in small studies of mixed quality, and cortisol is not a villain. Acute elevation is how a body answers a load. Blunt it without a reason and you get thyroid shifts, occasional liver enzyme rises, sedation. Show me the outcome that matters, and I will care about the marker. Astaxanthin sells an internal sunscreen. The studies are small, short, and frequently funded by the people who sell it. Lutein and zeaxanthin do have a real trial behind them, AREDS2, and it applies to people with intermediate age-related macular degeneration. It does not apply to anyone who wants sharper eyes. The asymmetry is the story. A drug with a marketing budget gets to skip the trial, because nobody asks for the paperwork. A generic that works gets buried, because no sales force carries it into a clinic. The probability that the bottle in your hand has a positive randomised trial behind it is not zero for every category. For most of the shelf it is close, and almost nobody prices it that way. Ask what the trial endpoint was. Weight. Fracture. Infarction. Death. If the endpoint is a number nobody feels, you are buying a mechanism.

A mechanism is a story about a pathway. An outcome is a scar, a scan, or a death certificate. The distance between the two is where an entire industry lives. Phosphodiesterase type 5 inhibitors relax smooth muscle, raise cyclic GMP, and improve blood flow. The reasoning then runs: more flow to the brain, better cognition. Observational datasets do associate these drugs with lower dementia rates, and that finding is interesting. It is not definitive. Men taking them are also, on average, in a doctor’s office more often and drinking less. What is real is narrower. Men on tadalafil for prostate symptoms or erectile dysfunction report feeling sharper. The likeliest explanation is boring: they stop waking five times a night. Sleep improves. Endothelial function improves. The compound did something, it just did not do the thing on the label. I have been on the receiving end of that logic. I built companies across twelve countries, restructured a EUR 75 million industrial group, took a company through a Frankfurt IPO, and signed a EUR 20 million personal guarantee on a business that folded on Christmas Day 2008. In every one of those, the mechanism looked fine on paper. The outcome is the only thing that shipped. Medicine keeps learning that the slow way: stop grading the pathway, grade the endpoint.

The most useful thing in the file is not a compound. It is a man changing his mind in public. For years Attia ran strict time-restricted eating: sixteen hours of fasting, often a single meal a day. He credited it for weight control and mental clarity. Then he reversed the position, and said so, and the reasoning is the part worth taking. Muscle protein synthesis is dose limited. You need roughly 2.5 to 3 grams of leucine in one sitting, which lands around 30 to 40 grams of protein, and past that the excess is oxidised rather than used to build tissue. Stack 150 grams of protein into one evening meal and most of it never becomes muscle. Cramming the day into one sitting makes the target hard to reach in the first place. He lost muscle on a protocol he had defended for years, and he said that out loud too. Add the second effect. A long fast raises cortisol, and cortisol is catabolic. For an overweight person chasing calorie restriction, the trade can be worth it. For a lean, active man in his fifties, the same tool costs the tissue that keeps him out of a hospital at 78. His position now is three or four meals, protein spread evenly, fasting kept as an option rather than a rule. The lesson is not the schedule. A publicly reversed position is more credible than a consistent one, because it means the data was allowed to win.

A supplement is a drug with a marketing budget and no trial behind it.

There is a phenotype that ends lean, active men while the check-up says the numbers are fine. Thin outside, fat inside. Normal body mass index, sometimes a visible abdomen, and a visceral fat depot that behaves like an organ with an agenda. Fat around the organs is metabolically active. It secretes interleukin 6 and tumour necrosis factor, it drives insulin resistance, and it tracks with cardiovascular disease, cancer, and all-cause mortality. The scans make the point better than any argument. One man presents at a body mass index of 22 with 4 pounds of visceral fat and is the sicker of the two people in the room. The second man sits at 28 with 1 pound and no metabolic markers. The index printed on a form is close to useless for the question that actually matters. The threshold in the file is about 1.5 to 2 pounds on a DEXA scan. Above that, address it. Lean men arrive there through the same roads as everyone else: genetics, chronic stress, broken sleep, alcohol, ultra-processed food driving liver fat production. The body runs out of room for subcutaneous fat and starts overflowing into the liver, the pancreas, and the muscle. This is the one place in the whole category where I would spend money before I spent attention. Get the scan. Then argue about the shelf.

Two things from the food aisle and one thing from a hot room came through. Small fatty fish come first. Sardines, anchovies, mackerel, salmon: dense in EPA and DHA, dense in protein and calcium, low in mercury because they sit low on the food chain. The target is a red blood cell omega-3 index of 8 to 12 per cent. Most people in the United States sit at 4 to 5 per cent, and food alone rarely closes the gap. Flaxseed does not close it either, because conversion from plant ALA runs at roughly 5 to 10 per cent. Fish oil is one of the few supplements the man takes, at 2 to 4 grams of EPA and DHA a day, and the reason is that the endpoint shows up in a blood test. Broccoli sprouts come second. They carry 20 to 50 times the glucoraphanin of mature broccoli. Glucoraphanin becomes sulforaphane, which switches on the NRF2 pathway: antioxidant response, glutathione, phase two detoxification. The human trial literature is thinner than the marketing, and the preclinical work on cancer prevention is genuinely interesting. The catch is myrosinase. Heat above 70 degrees Celsius destroys the enzyme that makes the conversion, which is why most capsules underdeliver, and why the file grows them on a sill and eats them three to four times a week. Heat comes third, with a caveat. Finnish cohort data link four to seven sauna sessions a week to lower cardiovascular, dementia, and all-cause mortality. Cohort data carries a healthy-user problem, and nobody should pretend otherwise. The pattern is twenty minutes at 175 to 185 degrees Fahrenheit, three to four times a week. It is not a substitute for training, and it is not detoxification. It is a hormetic stressor with unusually good observational cover.

I lost savings in the 2013 Cyprus bail-in. I was in the building at Lehman Brothers for the 2003 blackout. Neither of those is why I care about this. The reason is arithmetic. The chronic diseases that end men in their fifties and sixties are mostly slow, mostly measurable, and mostly visible years before they become events. Visceral fat shows up on a scan. Protein intake shows up in a grip strength test. Omega-3 status shows up in a blood draw. Sleep shows up in all of it. The supplement market is built on the opposite instinct: buy the feeling of having done something. It is a subscription to the appearance of care. The compounds that work are unglamorous, cheap, and boring. Fish oil that tastes of fish. Broccoli sprouts on a window sill. A barbell. A scan you pay for yourself, because the standard check-up will not order it. The book was published on 8 April 2026, 393 pages, and the argument in it is the same argument running here: the map you were handed is wrong, and the correction does not come from a bottle with a story printed on it. It comes from the endpoint. So the question does not change, for the compounds that matter or the ones that do not. Show me the trial. Then show me the outcome, never the mechanism. What I will not hand you is a protocol. Not a dose, not a stack, not the number from my own chart. Find a doctor who reads the same papers, and bring them the trial.