Big Biobanks Are Coming. Why You Should Own Yours.

I've been working on something

I published ten articles in “Thinking About Your Own Health” between February and September of last year, and then I went quiet. Here is what happened: I was working on an idea.

Over the past year I’d been writing to you, my friends and readers, about the personal health record. To make it and own it yourself, imagine a health notebook — a well-organized record of the data, curated by you, that describes your health. Not just what’s in your medical chart, but everything that matters to you: what you eat, how you move, what your biomarkers say, what you’re exposed to, how you sleep. Your vaccine record. Your family history. The data your wearable gadget collects while you’re not thinking about it. Because most health decisions and health results don’t happen in a doctor’s office — they happen in your kitchen, your grocery store, your gym, your bathroom. The health notebook could, in theory, capture all of it.

As a cardiologist, I’ve been tracking my own biomarkers since long before companies were selling that as a service. As a competitive masters athlete, I’ve tracked my physical performance on some very well standardized tests — like the 100-meter dash (fun!) and the 400 meters (painful and not fun, but a hell of a test that makes it hard to deny your age.) I’ve written about self-experimentation, about evaluating supplement claims, about why uncertainty is a superpower, not a weakness.

People ask me great health questions

Should I worry about seed oils? What about gluten — is it really bad for me, or is that just a fad? If the mother eats spicy food during pregnancy, will the baby tolerate more kinds of food? My blood pressure was really high but I was stressed, so it’s OK, right? Should I worry about microplastics? Should I be a vegan, or a carnivore?

You might have one friend who worries about xenoestrogenic pesticides, thermal printer receipt tape, and WiFi. Right, or wrong? You might have another who doesn’t worry about mixing salad with her bare hands after handling raw chicken. Right, or wrong? Well, wrong!

But my honest answer, more often than not, is: I’m not sure.

Adam Grant’s *Think Again* makes the case that this kind of uncertainty is a superpower — that the willingness to say “I don’t know” and to change your mind based on new evidence is more valuable than confident certainty. I agree with Adam Grant. So if “seed oils are bad” is part of your personal political and tribal identity, then you won’t be likely to take in evidence to the contrary. And if “WiFi and 5G cause brain fog” is a core part of your personal belief system, then defending that belief becomes stronger than a counter-argument. In “Uncertainty” I wrote about having breakfast with the double Nobel laureate Linus Pauling, who had rather strong beliefs about vitamin C. Should I take vitamin C every day like Pauling did? I’m not sure. I’ve run lots of clinical trials. As a clinical investigator, I know exactly what study we would need to do to find out — and it has never been done.

So that’s where I hit a wall and had to stop to think for a few months. One person’s health notebook can show trends, correlations, and early warnings. It has real personal value. But even my quite detailed notebook cannot establish that taking vitamin C caused my low back pain to feel better. In a made-up example in my Biomarkers article, Mrs. Duffer can see from her notebook’s years of bloodwork data that her fasting glucose has been creeping up. It might mean something. But when she changes her diet and the next glucose reading a month later goes down, she cannot know for certain whether the diet was the reason, or whether the data would have gone back towards average (“regressed to the mean”) without the new diet. I can watch my VO₂max declining over the decades but it’s impossible for one person to separate the impact of aging from the impact of reduced training volume from the impact of shifting metabolism. There’s a good reason that the gold standard of clinical trials requires testing a single intervention against a comparison population. In clinical trial design, “N” is number of people in the whole study. One person’s data in a health notebook, often called “N-of-1,” is still a powerful tool, but it’s lonely.

So during these quiet months, I’ve been asking: what would it take to actually get answers? Well, if a million highly detailed health notebooks were assembled in one data set, that would be a very powerful tool. If it contained genetic information, behaviors and habits, and medical outcomes, it would be called a biobank.

Your health data is valuable. A million peoples’ health notebooks would be really valuable. Owning your health data is the way forward.

A biobank that we own, together

What if a million health notebooks weren’t just personal? What if people who actually care about their data — my friends the masters athletes; self-experimenters; people tracking their biomarkers; people asking exactly the questions my friends ask — what if they pooled their notebooks. You’d better believe that big tech is already working on it. That insurance companies are in the hunt.

Your health notebook has personal value. But it also has tremendous economic value — the aggregated data from a million notebooks could transform health science, leading to better guidance, better treatments, more precise drug and supplement selection -- real impact.

I am arguing that you should own the value in your health data -- one of your most personal assets. I came to believe that we should not let this value be extracted from us — we should own it outright in the coming AI age. What if our million health notebooks were collected into a structure owned and controlled by its users? Not handing it to 23andMe or Apple or a university. Not giving away our information to the tech aggregators, like we’ve done with every other kind of click, like, purchase, and view in our digital lives, but instead governing it ourselves, kind of like a credit union for health data. With enough notebooks in the pool, you’d finally have the statistical power to approach real answers: Does that supplement actually work — and for whom? Is your statin suppressing your athletic performance? What does your genome say about your health risk from air pollution — and can diet modify it?

That’s the idea I was building, and the idea leads to an organization. In the next articles, I’ll show you specifically what this cooperative could answer — questions no existing study can touch.

The cooperative is called Co-Op Health.AI. If you’d be interested in learning more, sign up for the early list at https://coophealth.ai/.