I need to tell you about the worst two days I had last year.

Not because I want your sympathy — but because what happened after those two days is exactly why I believe AI is one of the most important tools you can put in your hands right now, especially if you're over 50 and trying to make smart decisions about your health.

Here's what happened.

The headline that wrecked me

I'm 57. A few years ago, I did a 23andMe genetic test. Mostly out of curiosity — I'm a data person, I like knowing things.

What I found out is that I carry two copies of a gene variant called APOE4. One copy is common enough. Two copies — what's called homozygous APOE4 — is a different story. It's the highest known genetic risk factor for late-onset Alzheimer's disease.

I've known this for a few years. I've done my research, made some lifestyle adjustments, tried not to catastrophize. My mother has Alzheimer's. She lives in my guest house in Mexico. She knows I'm her daughter, but she can't remember my name. This is not abstract for me. I know what this disease looks like up close.

So last year, when a major study came out in Nature Medicine — one of the most respected medical journals in the world — and the headlines started hitting my feed, I paid attention.

"95% of APOE4/4 carriers will develop Alzheimer's."

"Virtually certain to develop Alzheimer's."

I read those words and something dropped out of my stomach.

I spent two days in a kind of low-grade terror. Not just "hmm, that's concerning." The real kind — where you do quiet math about how many good years you might have left. Where you look at your to-do list and wonder what the point is of planning five years out.

Two days. That's what a headline can do.

Then I went to the actual paper

After two days, I was done feeling helpless. I opened Claude and I said something like: "I need to find the actual Nature Medicine paper about APOE4 homozygosity that came out in 2024, and I need to understand what it actually found — not what the headlines said."

This is where AI changed everything.

Claude helped me find the paper. Then it helped me read it. Not skim the abstract — actually work through what the researchers measured, how they measured it, and what conclusions they drew versus what conclusions the headlines drew.

Here is what the paper actually said:

95%+ of people with two copies of APOE4 will have elevated amyloid levels in their cerebrospinal fluid by age 65.

Amyloid is a protein. When it accumulates in the brain, it's considered a biomarker — a measurable biological sign that something may be developing. The paper found that this biomarker pattern is nearly universal in APOE4/4 carriers. That's why the researchers reclassified homozygous APOE4 as a "distinct genetic form" — because of the near-universal biomarker picture.

Here's what the headlines never told you: being biomarker-positive is not the same as having Alzheimer's disease. Amyloid can accumulate for 10 to 20 years before any symptoms appear. And a meaningful percentage of people who accumulate amyloid never develop clinical dementia.

The study was about what happens in the brain at the molecular level. The headlines turned that into "you will get Alzheimer's." Those are not the same sentence. They are not even close to the same sentence.

What AI did that I couldn't do alone

Let me be specific, because this matters.

Could I have found the paper myself? Maybe — eventually — with a lot of frustrating searching through PubMed, which is the database of medical research. It's free and public. It's also dense and not designed for regular humans.

Could I have understood the technical language without help? Honestly, probably not without hours of additional research. "CSF amyloid biomarker status" is not a phrase most of us have floating around in our heads.

What AI did was compress what would have been a research afternoon into a conversation. It pulled up the paper. It explained what "cerebrospinal fluid biomarker" means in plain language. It helped me understand the distinction between preclinical changes in the brain and actual diagnosed dementia. It let me ask follow-up questions in real time.

"So what does this mean for what I should actually do differently?"

That's the question that changed everything for me. Not "am I going to get Alzheimer's" — which is a fear question that has no useful answer — but "what does knowing this mean for my choices?"

The answer to that question was a map, not a sentence. It led me to research on estrogen timing and brain health, lifestyle interventions with strong evidence behind them, clinical trials running specifically for people with my genotype, and blood tests that can monitor certain markers over time.

I went from two days of fear to a plan. That's what happens when you go to the actual data instead of the headline.

Why this matters for you, even if you're not APOE4/4

Here's the thing: this isn't really about my specific gene variant. This is about the gap between what research actually finds and what headlines are able to communicate in eight words.

That gap is everywhere. It affects how you think about every medication your doctor has mentioned. Every supplement you've considered. Every dietary approach you've tried or avoided. Every screening test you've declined or been scared into.

Health reporters are not bad people. But they're working in a system that rewards clicks, operates under tight deadlines, and requires them to translate highly technical findings into sentences that fit a subject line. Something always gets lost. Usually, it's the nuance that would actually help you make a decision.

AI is not a doctor. I want to be very clear about that. It's not a substitute for talking to your physician, especially about your specific situation.

But it is a tool that can help you do something that used to require a medical degree or a research librarian: go to the primary source, understand what the data actually says, and ask the right questions.

That's a skill. And it's one you can learn. And it might save you from spending two days terrified of a headline that wasn't even true.

One thing you can try this week

Next time a health headline scares you, don't just close the tab. Try this:

Open Claude (or whatever AI you use) and say: "I just read a headline that said [X]. Can you help me find the actual study this is based on and tell me what the researchers actually concluded?"

You might be surprised what you find. Sometimes the headline is accurate. Sometimes it's not. Either way, you deserve to know the difference.

That's what being not too old for AI actually looks like.

— Annette

P.S. I'm not sharing this story to scare you about Alzheimer's or APOE4 — that's not the point. But if you're curious about the research on lifestyle and brain health for people with genetic risk factors, I'll link to some solid resources in the comments. The evidence-based stuff is genuinely encouraging.