A woman came to see me last year because her mother and her aunt had both died with Alzheimer's, and she had decided she was next. She was sixty-one, sharp, still running a business, and terrified. Her internist had ordered the usual things. Thyroid, B12, a basic metabolic panel, a structural MRI that read as normal for her age. Everything came back reassuring, and none of it touched the question she actually walked in with, which was whether the disease was already underway inside her and whether anyone could see it before it announced itself.
For most of my career the honest answer was that we could not see it early, and that even if we could, there was little to do. A study published in Nature on August 6, 2025, is the strongest reason I have to think both halves of that answer are changing. It points to a deficiency most physicians have never measured, in a metal most people associate only with psychiatric medication, and it may sit closer to the origin of Alzheimer's than anything we have found so far.
What the Harvard team actually found
The work came out of the lab of Bruce Yankner at Harvard Medical School, a group that has spent decades on the biology of the aging brain. They did something that sounds simple and is technically hard. Using mass spectrometry sensitive enough to detect metals in trace amounts, they measured 27 different metals in the brains of people who had died at three different stages: cognitively healthy, mild cognitive impairment, and advanced Alzheimer's. The brain tissue came from the Rush Memory and Aging Project, one of the best-characterized aging cohorts in the world.
Of all 27 metals, one stood out. Lithium was markedly lower in the prefrontal cortex of people with mild cognitive impairment and Alzheimer's than in people who stayed sharp. Nothing else moved that way. And the timing is the part I want you to hold onto. Lithium was already depleted at the mild cognitive impairment stage, which is the earliest point at which anyone notices a problem, often years before a diagnosis. This was not a late consequence of a brain already destroyed. It was present at the beginning.
When the team looked closer, they found where the lithium had gone. It was concentrated inside the amyloid plaques themselves, and the amount trapped there climbed as the disease advanced. The plaques were acting like a sponge, pulling lithium out of the surrounding tissue and locking it away where neurons could not use it. So the brain was not simply low on lithium. It was being actively robbed of it by the very pathology lithium seems to hold back.
From correlation to cause
A finding in dead tissue tells you what is associated with the disease, not what drives it. The next set of experiments is why this paper matters. The researchers took mice and cut the lithium in their diet by half, down to a level that mirrors the depletion seen in human Alzheimer's brains. Those mice developed more amyloid, more tau tangles, activated microglia, the brain's inflammatory cells, and measurable memory loss.
Here is the finding that reframes the whole disease for me. When they fed a low-lithium diet to ordinary aging mice that had no Alzheimer's genetics at all, those animals still built up more of the plaque-forming amyloid protein and still lost memory. Lithium deficiency alone, without any inherited risk, was enough to push a normal aging brain toward the pathology. That moves lithium from a bystander to something that looks like a lever.
Why the type of lithium matters
If you know anything about lithium, you know it as a psychiatric drug given at doses that require blood monitoring because the margin between helpful and toxic is narrow. That is lithium carbonate, and the doses used for bipolar disorder are roughly a thousand times higher than what we are talking about here. The Yankner team tried lithium carbonate in their Alzheimer's mice and it barely worked, because carbonate is drawn to the same negatively charged plaques that were sequestering the brain's natural lithium in the first place. The drug got trapped before it could do its job.
So they screened 16 different lithium salts looking for one that could slip past the plaques. They landed on lithium orotate. At tiny doses, in the micromolar range, lithium orotate cut amyloid and tau, restored the synapses that had been lost, and reversed memory deficits in the mice. Given long-term to aging animals, it prevented synapse loss and cognitive decline with no detectable toxicity. The dose that worked was a fraction of a psychiatric dose, closer to what your brain would get from a mineral than from a prescription.
What this does and does not mean for you
This is exactly the kind of finding that gets turned into a supplement fad by the weekend, so the caveats carry weight. The work is in mice and in postmortem human tissue. Yankner himself has said plainly that no one should start taking lithium on their own, that the safe and effective dose in living people is not yet established, and that his group is planning a human trial. Buying lithium orotate off a shelf and dosing yourself is not the lesson here, and it could do harm.
The real lesson is about measurement, and it fits everything I believe about how brain medicine should work. For decades we have treated Alzheimer's as a black box that opens only when symptoms are undeniable and the damage is done. This study is one more piece of evidence that the disease writes its signature in the brain's chemistry long before that, in trace metals, in blood biomarkers like p-tau217, in patterns a standard workup never goes looking for. The woman in my office did not need reassurance that her MRI looked normal. She needed someone to actually look, with the tools that can see the beginning of a disease rather than its end. That is the whole design of our Intensive Brain Health Program, which builds a real baseline of your brain's chemistry, structure, and function instead of waiting for the floor to give way.
Cognition is the asset that everything else in your life is built on. Alzheimer's does not begin the day you forget a name. It begins years earlier, quietly, and the science is finally catching up to the point where we can look for it and act. The question worth answering is not whether your brain feels fine today. It is what your brain's chemistry actually shows right now, while there is still room to change the outcome. That is a question you can get answered.