Alzheimer's Disease & Dementia Causes
Alzheimer's Disease & Dementia Causes
Alzheimer's disease is defined by two protein changes that begin years before any symptom and spread through the brain in a broadly predictable order. Around that core is a set of contributors — vascular, genetic, and environmental — that change when the disease becomes visible rather than whether it is present. This page sets out what happens in the brain, what genes can and cannot tell you, which risk factors are established and which are speculated, and the distinction that matters most to families reading it: nothing here was caused by anything the person did.
What happens in the brain
Two proteins are central, and they fail in different places and on different timetables.
| Process | What goes wrong | Why the timing matters |
|---|---|---|
| Amyloid accumulation | A fragment of a larger protein is not cleared as fast as it is produced, and it aggregates outside the cell into plaques | It begins first, and it plateaus relatively early — which is why amyloid burden on a scan correlates poorly with how impaired someone is |
| Tau spread | Tau, which stabilises the internal skeleton of the neuron, becomes abnormally phosphorylated, detaches and clumps inside the cell | It follows later and it tracks symptoms closely. Where tau is, is approximately what the person has lost |
| Synapse loss and atrophy | Connections are lost and the tissue shrinks, first in the regions that lay down new memories | This is the change that correlates with function, and it is what the MRI shows |
| Neuroinflammation | The brain's immune cells change their behaviour, with effects that are protective early and damaging later | One of the active research areas. It is a real part of the process and it is not yet a treatment target with proven benefit |
| Vascular and clearance failure | Small vessel disease, impaired perivascular drainage and reduced clearance of waste from the brain | It lowers the threshold at which the pathology becomes symptomatic, and it is partly modifiable |
| Network disconnection | The regions that work together in a network stop doing so, partly through the loss of the cells that carry the signal between them | It explains why the deficit follows a functional system — memory, language, visuospatial — rather than an arbitrary one |
The single most important consequence of that timetable is the silent phase. Amyloid begins accumulating, in most estimates, more than a decade before anything is noticeable, and for much of that time the brain compensates. By the time a person or a family notices, substantial pathology is already present. That is why the disease cannot currently be prevented by acting at the point of symptoms — and why the research effort has moved towards the pre-symptomatic phase.
It is also why the biology and the symptoms are not the same thing. A person can carry considerable plaque burden and think clearly, because what makes the difference is where tau has reached and how much network has been lost.
Genes: what they do and do not tell you
Genes matter in Alzheimer's disease in two entirely different ways, and conflating them is what makes the subject frightening. They need to be kept separate.
| Category | What it is | What it means for an individual |
|---|---|---|
| APOE ε4, common | A risk allele carried by a substantial minority of the population; one copy raises risk, two raise it considerably more | It raises probability and lowers the average age of onset. It is not a diagnosis, many carriers never develop the disease, and most people with the disease are not carriers of two copies |
| Other common variants | A large number of common variants identified in genome-wide studies, each with a small individual effect | They contribute to a polygenic picture that is not currently useful for individual prediction, and they are not part of clinical testing |
| APP, PSEN1, PSEN2, deterministic | Mutations that cause rare early-onset familial Alzheimer's disease | They cause the disease with very high penetrance. They account for a small fraction of all cases, and the family pattern is usually unmistakable across generations |
| Down syndrome | An extra copy of chromosome 21, which carries the gene for amyloid precursor protein | Alzheimer's pathology develops at a markedly younger age, and cognitive change in this group should be assessed rather than attributed |
Two practical rules follow. Genetic testing is preceded by counselling, not followed by it, because a result has consequences for relatives who did not ask for it and, in some jurisdictions, for insurance and employment. And a negative result changes nothing about management in the common form — the plan is the same whether or not a risk allele is present. Testing is most useful in early-onset disease, in a clear family pattern, or where the result would change a decision.
What raises the risk, and what is speculated
Age is the dominant risk factor, and it is not close. Beyond that, the honest position is a list of associations of varying strength, with the strongest and most actionable ones being cardiovascular and acting in mid-life rather than late.
| Factor | Direction | How it should be read |
|---|---|---|
| Age | Increases steeply after sixty-five | The dominant factor. Everything else modifies a probability that age has already set |
| Family history and APOE ε4 | Increase | Real and non-deterministic. A raised probability, not a prediction |
| Mid-life hypertension | Increases | One of the better-established modifiable associations, and one of the few with a treatment that changes it |
| Diabetes and insulin resistance | Increases | Established association; whether tight control changes dementia risk specifically is less clear than that it changes vascular risk generally |
| Smoking | Increases | Established, and the single most reversible item on the list at any age |
| Physical inactivity | Increases | Consistent association across studies. Causality in the specific direction is not established, and the general health case does not depend on it |
| Untreated hearing loss | Increases | Among the larger modifiable associations reported, and uniquely relevant because treatment is available now and the effect on daily life is immediate |
| Social isolation and low education | Increase apparent risk | Both partly reflect cognitive reserve: more reserve means the same pathology becomes visible later, not that it is absent |
| Traumatic brain injury | Probable increase, particularly with repeated injury | The association is consistently reported and the mechanism is not settled. It is an argument for head protection rather than a cause to be assigned after the fact |
| Aluminium, single foods, dental amalgam, stress | Not established as causes | Investigated and not supported. Stress can unmask decline already underway, which is why diagnoses often follow life events |
The final row is the one families need most. This is not caused by anything the person did, and the recurrence of that question in clinic is a measure of how much guilt the illness generates rather than of any real causation.
What makes the symptoms worse right now
Separate from risk — which is about getting the disease — is what makes an existing impairment worse today. This list is more useful than the risk list, because every item on it is actionable and the effect is measured in weeks.
| Factor | How it shows up | What is done |
|---|---|---|
| Anticholinergic and sedative medicines | Increased confusion, dry mouth, constipation, urinary retention, drowsiness | A burden calculation and a substitution plan, reviewed weeks later |
| Infection, even a mild one | A step down over a day or two, often with reduced appetite and increased sleep | Looked for in any sudden change and treated. The threshold for investigating is low |
| Pain and constipation | Agitation and apparent confusion in someone who cannot report either | Asked about directly and treated; prevention rather than rescue in the case of constipation |
| Sleep disruption and apnoea | Daytime sleepiness, inattention, sundowning, mood change | Sleep assessment, treatment of apnoea, and a consistent daily rhythm |
| Uncorrected hearing and vision loss | Apparent confusion, withdrawal, and answering questions that were not asked | Testing and correction, and checking that the aids are worn and fit |
| Unmanaged vascular risk | Additional small vessel injury on top of the Alzheimer's pathology | Blood pressure, lipids, diabetes, smoking and atrial fibrillation treated as active targets |
| An unfamiliar environment | A marked step down in hospital, on holiday, or after a move | Familiar objects and a familiar person present, a consistent routine, and short stays where a stay is unavoidable |
That last row is worth a line to itself. A hospital admission is the commonest setting for a step down that never fully recovers, and the measures that prevent it are cheap: a familiar person present, glasses and hearing aids in place, a clock and a calendar in view, and the shortest stay that is safe.
What is established, and what is not
Being explicit about the boundary is more useful to a family than a confident summary, because most of the confusion they will encounter comes from sources that do not observe it.
| Question | What can honestly be said |
|---|---|
| Is it a single disease? | The pathology is defined and reproducible. Whether it is one disease or several that share a final pathway is genuinely debated |
| Do amyloid and tau cause it? | They are the defining pathology and the strongest evidence points to them as central. Whether amyloid is the driver or a participant is contested, and the modest clinical effect of removing it is part of that debate |
| Can it be prevented? | Not guaranteed. Reducing modifiable risk is associated with lower incidence and later onset at a population level, and it is worth doing for reasons that do not depend on dementia |
| Can it be reversed? | No. No treatment currently restores lost function, and any claim that one does should be treated as a claim rather than as medicine |
| Do the new antibodies change the outlook? | They produce a small average slowing over the studied period in a narrow group. Whether that translates into a meaningful difference in years of independence is not yet established |
| Is anything about this the person's fault? | No. Not aluminium, not a food, not a supplement not taken, not a character flaw, and not a failure to try harder |
Where there is a real question about cause in an individual — an unusually young onset, an unusually fast tempo, or an atypical phenotype — the answer is usually found in the exclusion work rather than in the biology. The diagnosis page sets out what that involves, and the living-with-dementia page sets out what is worth doing regardless of the answer.