In this monologue, the speaker argues that randomized controlled trials (RCTs), the current gold standard of evidence‑based medicine, are fundamentally inadequate for delivering personalized health care. By aggregating heterogeneous populations, RCTs mask individual variability, leading to drugs that help only a small fraction of patients despite achieving statistical significance. The speaker cites that the top‑selling drugs benefit roughly 4%–25% of users and uses a drug‑placebo delta example to illustrate the disconnect between population‑level efficacy and patient relevance. To overcome this, the speaker proposes a shift to N‑of‑1 medicine, which relies on extensive personal data—particularly longitudinal multi‑omics (genome, transcriptome, proteome, metabolome, microbiome) combined with AI/ML analysis—to identify root‑cause metabolic dysfunction and tailor interventions. The talk highlights pioneering efforts such as Stanford’s Snyder Lab and the company N‑of‑One, which aim to empower individuals with their own health data. The speaker calls for an intellectual and cultural revolution in medicine, moving away from heterogeneous RCTs toward individualized, data‑driven care, and invites viewers to discuss the provocative ideas in the comments.
Chapters
The randomized controlled trial is dead. Okay, what do I mean by this? Well, let me present you with an information gap. Here's where we are currently. Fewer than 10% of people are metabolically healthy, and the majority are confused and frustrated about how to manage their own unique health concerns. But here is where we are going to be in 25 years. Decreasing rates of obesity and metabolic disease, with people having clarity and taking targeted action to improve their own metabolic health. But to get from where we are now to that future, we are going to need to kill the randomized controlled trial in favor of n=1 medicine. Welcome to my channel. Stay curious. Here's the problem. The way things work right now, in order to test a specific intervention, like a new drug, you ideally want to take a large population of people and then randomize them to the drug, intervention, or an alternative treatment, like a placebo. And this is called a randomized controlled trial, and it's considered the gold standard in modern medical research. It is the basis for evidence-based medicine. But by virtue of taking a large enough population of people to get statistically significant results, you are inevitably lumping together a heterogeneous mass of different human beings. And what are the results? Well, the top 10 highest-grossing drugs only help between 1 in 25 and 1 in 4 people who take them. Now, that statistic is from 2015, but I'm not sure much has changed because our approach to medicine has not changed. So while you can drive up the number of people in a trial to get statistically significant results, what is lost is specificity, individuality. There will almost always be responders and non-responders, strong responders and weak responders, and the non-responders and the weak responders might be the majority, but To sell a drug, all that matters is the population average and statistical significance. If you can drive that p-value below 0.05. But for the patient, what matters is something different. By way of example, if a drug provides a 40% improvement in the drug group and only a 34% improvement in the placebo group, that delta, that 6%, might be statistically significant. It might be enough to sell the drug. But if you're the patient and you're not in those 6%, do you care? I'd argue no. No. So what we need is something better, something that takes into account individuality, and that is n=1 medicine. So how do we get to n=1 medicine? Well, we need data. Data, lots of data, your data. We are moving towards a future that can collect big datasets on an individual. In the most extreme examples, and the sexiest examples in my opinion, this is called longitudinal multi-omics. Multi-omics referring to a compilation and an integration of all the omes. The microbiome, the genome, the transcriptome, the proteome, the metabolome, all the ohms that make you unique. And then an integration of these massive datasets paired with AI and machine learning to parse what's going on with you, what's going on right, and what's going on wrong with your individual metabolism, your individual physiology. And then targeting the problems with root cause targeted protocols aimed at the underlying physiology, the root cause, not the symptoms. And there are already labs, research teams, and companies that are pushing us in this direction. The future is now. So for example, the Snyder Lab at Stanford is pioneering research in longitudinal multiomics for individual health. And the company, say, N-of-One— N-of-One, get it? Who? They're trying to put patients' data in their own hands, empowering those patients, you, me, to do what the current healthcare system can't: develop bio-individual treatment protocols based on root cause physiology. I personally think this is the way forward. This is the way we move into that future where everybody has access to metabolic health. But to do this really will require an intellectual revolution and a cultural shift in medicine moving away from randomized controlled trials based on heterogeneous groups of humans towards gold standard N=1 medicine. It's a really provocative thought. It's a bold statement. But given where things are now with the metabolic health of this country, this world, don't you think we need provocative thoughts and bold statements? I'll leave you with that. Let me know your thoughts. In the comments below, and let's have a discussion.
