The Memory Genes Don't Explain

Not "nothing" as in the study failed. "Nothing" as in the thing everyone was certain would be there wasn't there, and the researchers wrote that down and published it anyway — which is a small act of nerve in a field that runs on causes.

Here's the setup. SuperAgers are people past eighty whose episodic memory performs like someone decades younger. Not "sharp for their age." Genuinely indistinguishable from a person twenty or thirty years down the timeline. Emily Rogalski and her colleagues at Northwestern put the category on the map in 2012, and it has been quietly embarrassing gerontology ever since, because these people are not supposed to exist in the numbers they do. The standard story of the aging brain is a slow subtraction problem. SuperAgers simply decline to do the arithmetic.

So the obvious question: what's protecting them?

The obvious answer, in 2026, is genetics. It's always genetics. Genetics is the modern explanation of first resort — the place we file anything we can't otherwise account for, because a gene feels like a reason. Somewhere in there is a variant, a lucky allele, a spelling difference in the source code. Find it, bottle it, sell it.

A team at the University of Chicago's HAARC Center, working with the Translational Genomics Research Institute, went looking. They compared 142 SuperAgers against 89 cognitively average older adults, recruited across five sites in the U.S. and Canada. They checked APOE — the single most famous risk gene in Alzheimer's research. They checked three separate polygenic risk scores, which aggregate thousands of small genetic contributions into a single number.

And they could not tell the two groups apart.

The paper — Piras and Capuano as co-first authors, Matt Huentelman and Rogalski as senior authors — landed in Alzheimer's Research & Therapy on July 22nd, and what it reports is a hole. A precisely measured, well-controlled, 231-person hole where the explanation was supposed to be.

Now, be careful about how big a hole it is. A polygenic risk score for Alzheimer's is a score for Alzheimer's risk, not for memory superiority — so this tells you SuperAging isn't hiding in the Alzheimer's risk architecture, and says nothing about the rest of the genome. It could still be boringly genetic by way of variants nobody has thought to score. Two hundred thirty-one people is also a small null, and SuperAgers are volunteers who reached eighty-plus intact and chose to enroll, which is a filter that can hide a great deal.

So the honest claim isn't the answer isn't in the blueprint. It's narrower and stranger: the answer isn't in the part of the blueprint we built instruments to read.

Which raises the question worth the whole study. Why did we build the instrument pointed there?

Not because genes are the likeliest answer. Because a genome is legible and ownable. A polygenic risk score compresses thousands of variants into one number attached to one body — fundable, patentable, billable, locatable inside a single customer. Meanwhile the candidates Rogalski's team is chasing next, in a list that runs through imaging and biomarkers and immune profiling and sleep, end at "social and environmental measures." Community. Who you talk to. Those are distributed, relational, and unownable. Nobody can bottle having friends.

So we built a magnificent apparatus for reading the individual, almost nothing for reading the between, and then act surprised when the answer isn't where the light is. A null result from an instrument optimized for individual, monetizable causes isn't a mystery. It's the instrument telling you where it was aimed.

And I should admit the obvious about my own delight here. I like the relational answer partly because it's the one that lets me end on go do something. That's a motivated preference running the same direction as the genetic determinism I just criticized, only with better manners. The bias doesn't make the argument wrong. It does mean the between deserves an instrument, not a cheer.

Consider what's actually on the table. You are a roughly 86-billion-neuron structure, assembled by chemistry, running on about twenty watts, that somehow maintains a stable representation of your ninth birthday for eighty years. That's already an outrageous claim about what matter can do. And some of these structures hold that capacity deep into their ninth decade for reasons we cannot currently locate anywhere we know how to look.

An honestly held unknown is not an absence. It's a map with a coastline drawn around territory nobody has walked — infinitely more useful than a map with a confident guess printed over the blank. And this particular blank has an address: it's in the places our knowledge apparatus was never built to see, which are reliably the places where the answers are collective.

Some eighty-year-olds remember everything, and we have no idea why. Isn't that fantastic? There's still something in there. In you, presumably, since you're made of the same stuff — and possibly, if the hole is where it looks like it is, in the people around you.

Go find out.