The two best-selling brain supplements have the weakest evidence on this list. The strongest data belongs to a cheap fish-oil dose and a prescription you already know. 7 nootropics, ranked by how strong the HUMAN trial evidence is, not how loud the marketing is. 1. Omega-3 (DHA and EPA). The only one with moderate-grade trial evidence. Across 24 RCTs and 9,660 people, a real but small gain in executive function, showing up above about 500 mg a day. More isn't better: very high doses bring their own heart-rhythm risk. Caveat: the famous "30% lower Alzheimer's risk" is from population studies, not trials, and the big prevention trial in healthy elders found nothing. Best of a humble bunch. 2. Modafinil. A prescription-only wakefulness drug, Schedule IV controlled in the US, not something you can buy. The one compound here with a clean pooled positive in rested people, and it's tiny: average effect size 0.12 across 14 trials, mostly on hard executive tasks. It's on the list for evidence strength, not as something to go take. It may even dull creative thinking. 3. Caffeine plus L-theanine. A real, repeatable attention boost, but read the fine print: every trial is a single dose, in young adults, measured for about two hours. A review of 49 trials puts the attention effect at a modest 0.33. Cheap and reliable for the next two hours. Not a proven daily upgrade. 4. Creatine. A small memory signal, mostly in older adults. But Europe's food-safety regulator reviewed it in 2024 and rejected the cognition claim outright, and the largest "win" meta double-counts participants. In young healthy people the effect is zero. 5. Bacopa monnieri. Nine trials, 437 people, twelve weeks minimum. The effect is real but narrow: it speeds up attention, it doesn't boost memory. Expect stomach upset, the main reason people quit. 6. Lion's mane. One small win (30 older adults with mild impairment) that vanished four weeks after they stopped taking it. A 2025 trial in healthy young adults came back mixed and inconclusive, no clear overall benefit. Huge data in mice. Thin, messy data in people. 7. Ginkgo. The most thoroughly disproven supplement on this list. 3,069 people, six years, 120 mg twice a day: 277 got dementia on ginkgo, 246 on placebo. A second trial in 2,854 people agreed. Cochrane, 36 trials, calls the benefit "inconsistent and unreliable." Safe enough if you're not on blood thinners. It just doesn't work. Fish oil and modafinil have the best data. The two on the front shelf have the worst.

An elephant has about 100 times more cells than you and lives for decades. Every one of those cells is a lottery ticket for cancer. By the math, elephants should be cancer factories. They're not. About 4.81% of elephants die of cancer. For humans the figure is 11 to 25%. The clue is in a single gene. TP53 is the body's cancer alarm. When a cell's DNA gets damaged, TP53 decides whether to repair it or kill the cell before it turns dangerous. You carry one copy. African elephants carry at least 20. In lab dishes, elephant cells hit with DNA damage are far quicker to self-destruct than human cells. More alarms, faster shutdown, fewer chances for a tumor to start. Biologists call this Peto's Paradox. Across species, more cells does not mean more cancer. Evolution already solved the problem. We're just reading the answer.

Almost everyone thinks we’re losing to cancer. The age-adjusted death rate says otherwise: down 34% since 1991. Measured per person and adjusted for age, the US cancer death rate dropped by a third between 1991 and 2022. That adds up to roughly 4.5 million deaths that simply didn’t happen (about 3.0 million men and 1.4 million women). One caveat. That’s the death RATE per person, not the raw count. The total number of cancer deaths still rises each year because the country keeps getting bigger and older (about 618,000 deaths expected in 2025). Your odds of dying from cancer, though, keep falling. We bent the curve on the second-biggest killer without ever “curing” cancer.

The ADA meeting just ran five obesity drugs at once, and the weight numbers have never looked better. Here’s the board, and the question hiding behind all of it: 🥇Retatrutide (three hormones): 28% of body weight gone at 80 weeks. The most a drug has ever taken off in a major trial. 🥈Tirzepatide (Mounjaro): about 22%. Most people on a weight drug are already here. 🥉 CagriSema (adds the fullness hormone, amylin, on top): about 23%, from a peer-reviewed Phase 3 trial. 4️⃣ Eloralintide (that same fullness hormone, no Ozempic component): about 20%. 5️⃣ Orforglipron (a pill, no needle): about 11%. The first once-daily oral option in this class. Five drugs. Five ways to lose the fat. The fat is the part the field has figured out. None of those percentages tell you what happened to muscle. When you lose 28% of your body weight, about a quarter of the loss is lean mass. The careful body-scan data, from Ozempic’s own trial, shows that fraction is the same whether you used a drug or just dieted hard. The body sheds lean mass with every large weight loss. So the next race isn’t a bigger weight number. It’s keeping the muscle while the fat comes off. Bimagrumab already does it: paired with a weight drug in a trial this year, 92% of the loss was fat and lean mass barely moved. That’s why the pipeline looks nothing like this board.

Heart disease still kills about 680,000 Americans a year. For forty of those years the only fix was a pill you took every single morning, and plenty of people quit within a year or two. Then the shots arrived. Evolocumab, an antibody you inject every two weeks, drops LDL about 60%. Inclisiran went further, two shots a year, LDL roughly halved. Each one needed less of the patient than the last. VERVE-102 needs them once. One IV dose of a base editor switches off the liver's PCSK9 gene, the gene that keeps LDL high. In 35 patients with inherited high cholesterol, the top dose cut the PCSK9 protein a mean of 88% and LDL a mean of 62%, and in the 15 followed past a year, it held. The pill, the fortnightly shot, the twice-yearly shot, every one of them stops working the day you stop showing up. This is the first that doesn't ask you to. Phase 1b, no placebo, 35 people. @EliLillyandCo bought Verve to see what it does in 3,500. Would you take a one-time edit at 45, or wait for the outcomes data?