Everything You Know About Dementia Might Be Based on the Wrong Category
Dementia gets treated like a single disease. It's a label we slapped on a dozen different things happening in the brain and body, and the confusion is costing us.
Your mother gets a diagnosis: "dementia." You hear one word, and your brain fills in the rest: memory loss, decline, nothing to be done.
But that single word? It's doing enormous work to represent what is actually a collection of distinct biological processes that happen to share one functional outcome.
And because we keep treating the label like it's the condition, we keep getting the response wrong.
The word "dementia" works like the word "fever"
Dementia is a syndrome, not a diagnosis. It tells you that cognitive changes are severe enough to disrupt daily function, but it tells you almost nothing about why.
Think of it this way: a doctor who says "you have a fever" hasn't told you whether it's malaria, a urinary tract infection, or your body fighting off a cold. Fever describes a pattern and so does dementia.
Alzheimer's disease, vascular dementia, Lewy body dementia, frontotemporal lobar degeneration, Parkinson's-related dementia: these have different cellular mechanisms, different genetic profiles, different progression patterns, and different symptom presentations.
Frontotemporal dementia often starts with personality changes and disinhibition, while memory stays intact. Lewy body dementia frequently begins with visual hallucinations and motor symptoms. They got grouped together because they share a functional outcome: cognitive impairment severe enough to affect independence.
That's clinical convenience. It's not biological unity.
Why this matters more than it sounds
The category problem is probably the single biggest reason pharmaceutical trials keep producing null results. When you run a drug trial on "dementia patients," you're testing a molecule on a heterogeneous population and hoping it works, despite the fact that the participants may have completely different things going wrong at the cellular level.
The amyloid cascade hypothesis is the clearest example: for 30 years, Alzheimer's research poured resources into the idea that clearing amyloid plaques from the brain was the path to treatment. Over $40 billion invested in amyloid-targeting drug development since the mid-1990s, with limited clinical benefit.
The foundational 2006 paper in Science, supporting a specific amyloid subtype as causal, was retracted in 2022 due to image manipulation. That doesn't demolish the entire hypothesis (amyloid is clearly involved in the biology), but it should give us pause.
Many cognitively normal older adults carry significant amyloid burden. Many Alzheimer's patients don't show the expected amyloid profile. Amyloid is a marker and possibly a contributor but treating it as the cause has been expensive.
And that's just within Alzheimer's. Extend the same single-molecule logic to the full category of "dementia" and you can see why the hit rate has been so low.
Some of what looks like dementia is treatable right now
This is where the category problem gets personal.
A meaningful subset of presentations that look like dementia are caused by B12 deficiency, untreated hypothyroidism, normal pressure hydrocephalus, medication toxicity, depression, or subdural hematoma. All of these are treatable and some are fully reversible.
But the assumption that dementia equals irreversible, has historically led clinicians to stop investigating once the pattern matched.
If you already believe the trajectory is one-way, you don't go looking for the thyroid panel or the medication interaction that might actually explain what you're seeing.
Some underlying mechanisms genuinely are irreversible, but others are not.
Whether the whole presentation can be reversed depends heavily on which mechanisms are driving it in this specific person. You can't know that if you stop at the label.
The question that changes everything
The traditional question is: "Does this person have dementia?"
The better question, rebuilt from what the evidence actually supports: "What is driving the decline in this specific person, and which of those drivers are modifiable?". That reframe changes the entire posture.
Instead of matching someone to a category and offering a prognosis, you're looking at the individual biology. What's the cardiovascular health? Sleep quality? Systemic inflammation? Metabolic function? Medication load?
Since the FINGER trial in 2015 and subsequent lifestyle intervention studies, showed meaningful effects on cognitive trajectory, through interventions that had nothing to do with the brain in isolation: exercise, nutrition, cognitive training, vascular risk management. The effect sizes are real.
Treating dementia strictly as a brain problem forecloses most of the modifiable terrain.
What this means if someone you love has been diagnosed
The concept of dementia, rebuilt from what probably survives scrutiny, is less tidy than the inherited version. But it's more useful.
It means asking the diagnosing physician exactly which mechanisms they've investigated and which they've ruled out. It means not assuming that a diagnosis of "dementia" is a complete clinical answer. It means understanding that some of the factors affecting trajectory (sleep, physical activity, social engagement, systemic health) are things you can actually influence.
It won't change the biology. But it might change how much of the biology gets addressed.
If you want to understand more about what's actually happening in the brain during dementia and how to respond to it, explore the Decoding the Dementia Brain workshop or book a discovery call with AEF-CNP.
Amelia Enginco-Figueroa is a Swiss-educated Cognitive Neuroscientist specializing in applied brain science for caregivers, educators, and institutions. Learn more at aef-cnp.com.