Alzheimer's disease

• 1. Introduction

  ›

  In 1901, a German psychiatrist named Alois Alzheimer met a fifty-year-old woman he called Auguste D. He followed her case until she died in 1906, when he first reported publicly on what he had found inside her brain. Malformed protein deposits had gathered in her cerebral cortex. The condition that bears his name is now the most common form of dementia, accounting for around 60 to 70 percent of cases. As of 2020, roughly 50 million people worldwide were living with it. The first sign is usually small and easy to dismiss. People struggle to remember recent events, and families often mistake it for ordinary aging or stress. What follows is a long unraveling. How does a single misfolded protein begin a cascade that ends in a person unable to feed themselves? Why does it strike women more often than men? And why, after more than a century, can no treatment stop or reverse it?

• 2. Three Stages Of Decline

  ›

  The hippocampus, the brain region tied to memory, is the disease's first target, which is why short term memory loss appears before anything else. Detailed neuropsychological testing can reveal mild cognitive difficulties up to eight years before a person meets the clinical criteria for diagnosis. At this earliest point, people struggle to remember recently learned facts and cannot acquire new information. Apathy and depression can already be present, and apathy remains the most persistent symptom throughout the entire course of the disease.

  Older memories prove more durable than new ones in the early stage. Episodic memories of a person's life, facts they learned long ago, and implicit memory, such as how to use a fork or drink from a glass, are affected far less than fresh information. Language begins to thin out, marked by a shrinking vocabulary and decreased word fluency. At this point a person can still communicate basic ideas and perform many tasks independently, though they may need help with the most cognitively demanding activities.

  Progressive deterioration in the middle stage eventually makes independent living impossible. People begin substituting wrong words, called paraphasias, and may fail to recognize close relatives as long-term memory finally erodes. Roughly 30 percent develop delusional symptoms and misidentifications, and many lose insight into their own condition, a state called anosognosia. Wandering, irritability, and sundowning become common, alongside urinary incontinence.

  In the late stage there is complete dependence on caregivers. Speech collapses to single words and then to nothing, yet people can often still understand and return emotional signals. Muscle mass and mobility deteriorate until they are bedridden and unable to feed themselves. The cause of death is usually an external factor such as pneumonia or infection of pressure ulcers, not the disease itself. In some cases there is a paradoxical lucidity immediately before death, an unexpected recovery of mental clarity.

• 3. Plaques And Tangles

  ›

  Two abnormal proteins define the pathology: amyloid beta, which gathers outside neurons as amyloid plaques, and tau, which builds up inside neurons as neurofibrillary tangles. Both become dangerous the same way. They misfold into a shape rich in beta sheets, then spread through the brain by a prion-like mechanism of seeded protein aggregation, in which one misfolded molecule causes its neighbors to misfold too.

  The amyloid cascade hypothesis holds that the buildup of misfolded amyloid beta is the fundamental cause of the disease. In this view, abnormal amyloid beta leads to tauopathy and then to the complex degeneration of advanced Alzheimer's. People with trisomy 21, also known as Down syndrome, carry an extra copy of the gene for the amyloid precursor protein, and almost universally develop the symptoms and neuropathology by 40 years of age. A rare mutation that reduces amyloid beta production, by contrast, protects against the disease.

  Heiko Braak and colleagues found that tauopathy can be detected in certain neurons before amyloid plaques are visible, which supports the rival tau hypothesis. The disease appears to begin with hyperphosphorylation of tau in vulnerable populations such as the locus coeruleus. Tauopathy occurs in over 30 diseases beyond Alzheimer's. Current evidence still favors abnormal amyloid beta as the prime mover, but the two hypotheses are not mutually exclusive. Amyloid beta may start the disease while tauopathy is required for its full expression.

  Inside the neuron, tau normally stabilizes microtubules, the tracks that guide nutrients from the cell body to the ends of the axon. When tau becomes hyperphosphorylated, it pairs with other threads and disintegrates the transport system. Amyloid beta itself is a fragment, 39 to 43 amino acids long, cut from the larger precursor protein by the enzymes gamma secretase and beta secretase. It misfolds, self-assembles into fibrils, and clumps into the dense deposits seen outside neurons.

• 4. The Weight Of Genes

  ›

  Late-onset Alzheimer's is about 70 percent heritable, yet most cases are not familial and are termed sporadic. The strongest genetic risk factor for sporadic disease is APOEε4, one of four alleles of apolipoprotein E. Between 40 and 80 percent of people with Alzheimer's carry at least one copy. A single copy raises risk roughly threefold, and two copies raise it about fifteen times. That relationship is not universal. Nigerian Yoruba people do not show the link between APOEε4 dose and incidence seen in other populations.

  Only 1 to 2 percent of cases are inherited through autosomal dominant mutations, in what is called early-onset familial Alzheimer's disease. This form is rarer, tends to progress more rapidly, and is about 90 percent heritable. It can be traced to mutations in one of three genes: the amyloid-beta precursor protein and the presenilins PSEN1 and PSEN2. Most of these mutations increase production of amyloid beta 42, the main component of plaques.

  A Japanese pedigree revealed something stranger. A deletion mutation of codon 693 of the precursor protein gene, first reported in 2008 and known as the Osaka mutation, raises risk only in homozygotes. It accelerates amyloid beta oligomerization, but the proteins never form the fibrils that aggregate into plaques. That detail hints that the small toxic oligomers, rather than the plaques themselves, may be what damages the brain.

• 5. The Search For Causes

  ›

  Oskar Fischer, working in the early 20th century, likened amyloid plaques to small masses of a microbe called actinomyces, opening a line of inquiry that has never fully closed. Since then at least 15 different agents, including bacteria, viruses, fungi, and protozoa, have been proposed as causes. No definitive evidence shows that any single agent is necessary and sufficient. Human herpes viruses such as HSV-1, HHV-6, and HHV-7 have been linked to risk, and one large study of 6,245,282 patients reported increased risk following COVID-19 infection in cognitively normal individuals over 65.

  The cholinergic hypothesis points elsewhere, to the loss of neurons in the basal forebrain that produce the neurotransmitter acetylcholine. Those cells supply acetylcholine to the limbic system and cerebral cortex, and their death led directly to the development of drugs meant to raise acetylcholine in the brain. The efficacy of those drugs is limited, probably because many other neurotransmitter systems also degenerate.

  Sleep disturbance was once seen only as a consequence of the disease, but evidence now suggests the relationship runs both ways. One theory holds that the brain's clearance of toxic substances, including amyloid beta, is most active during sleep. Less sleep means more amyloid beta produced and less cleared, leading to accumulation. Smoking is described as a significant risk factor, and air pollution may contribute as well.

  One hypothesis reverses the order of development entirely. Retrogenesis proposes that just as a fetus builds the brain from neurulation through myelination, the Alzheimer's brain reverses the process, starting with demyelination and the death of white matter before reaching the gray. Some researchers have gone further still, proposing that Alzheimer's is a Type 3 diabetes because of correspondences with both Type 1 and Type 2.

• 6. Naming What You Cannot Cure

  ›

  Up to 23 percent of those clinically diagnosed with Alzheimer's may be misdiagnosed, carrying pathology that points to another condition entirely. For most of the 20th century the disease can only be definitively confirmed at autopsy, by examining brain tissue under a microscope for plaques and tangles. In the absence of autopsy, a clinical diagnosis is only ever possible or probable, built from medical history, observations from relatives, and behavioral changes. A caregiver's account is especially important, because a person with the disease is commonly unaware of their own deficits.

  Three sets of criteria guide clinical diagnosis: the DSM-5 from 2013, the NIA-AA definition revised in 2011, and the International Working Group criteria revised in 2010. Cognitive tests such as the mini-mental state examination, the Montreal Cognitive Assessment, and the Mini-Cog help measure impairment, though they can miss the earliest stages. Supplemental blood tests, thyroid checks, and scans rule out treatable causes such as tumors, strokes, vitamin B12 deficiency, and depression.

  Neuroimaging has slowly made diagnosis possible in living patients. On MRI or CT the disease usually shows cortical atrophy and shrinkage of the hippocampus. PET imaging can detect protein deposits, and the FDA has approved radiopharmaceutical agents florbetapir in 2012, flutemetamol in 2013, florbetaben in 2014, and flortaucipir in 2020. Because many insurance companies in the United States do not cover the procedure, its use is largely limited to clinical trials.

  In May 2025 the FDA approved a blood test, the Lumipulse G pTau217 amyloid ratio device from Fujirebio Diagnostics, for early detection of amyloid plaques in adults aged 55 and older who show signs of the disease. It marks a shift from the autopsy table toward a simple sample of blood.

• 7. Treatment And Its Limits

  ›

  Four acetylcholinesterase inhibitors, tacrine, rivastigmine, galantamine, and donepezil, treat the cognitive symptoms of mild to severe Alzheimer's by slowing the breakdown of acetylcholine. The benefit is small, and it does not delay symptom onset. Memantine, an NMDA receptor antagonist first used as an anti-influenza agent, blocks overstimulation by glutamate and offers a small benefit in moderate to severe disease. The combination of memantine and donepezil has been described as of statistically significant but clinically marginal effectiveness.

  Two monoclonal antibodies, donanemab and lecanemab, have been approved to target amyloid beta, yet their role remains uncertain. Lecanemab carries a boxed warning about amyloid-related imaging abnormalities. A 2026 meta-analysis found that anti-amyloid antibodies have no effect, and a 2026 Cochrane review found no effect or at best a small one. The minor cognitive effect of antibodies that clear amyloid beta has prompted some researchers to reconsider the amyloid cascade hypothesis itself.

  When drugs reach their limits, the work falls to people. Caregiving is essentially the treatment, since the disease gradually renders people incapable of tending to their own needs. Simplified routines, safety locks, and labeling household items can reduce risk in the earlier stages. The use of physical restraints is discouraged, and the VIPS framework has been shown to reduce hours per day of restraint. Music therapy is effective at reducing behavioral and psychological symptoms, while emotion-oriented approaches such as reminiscence therapy show inconsistent results.

  The burden lands hard on families. In the United States, informal family care is estimated to make up nearly three-fourths of all caregiving, at a cost of 234 billion dollars per year and roughly 18.5 billion hours. Worldwide, the cost of caring for people with the disease is projected to increase nearly tenfold and reach about 9.1 trillion dollars by 2050.

• 8. A Growing Shadow

  ›

  By 2050 the number of people with Alzheimer's is estimated to triple, reaching 152 million worldwide, up from the 50 million of 2020. Advancing age is the primary driver. Every five years after the age of 65, the risk of acquiring the disease roughly doubles. In the United States in 2020, prevalence was estimated at 5.3 percent in the 60 to 74 age group, rising to 13.8 percent in the 74 to 84 group and to 34.6 percent in those over 85.

  Women are affected more often than men, a gap long attributed to their longer lifespans. The picture is more complicated than that. The Framingham study found women with almost twice the lifetime risk of men. Tau protein accumulates faster in women, APOE4 raises risk more in women, and for the same amount of pathology a woman shows greater cognitive decline. As of 2025 the reason women are more commonly affected is still unknown.

  The disease has left its mark on art and film. It was portrayed in Still Alice in 2014, about a Columbia University professor with early onset disease, featuring Julianne Moore in the title role, and in Iris in 2001, based on John Bayley's memoir of his wife Iris Murdoch. The American artist William Utermohlen drew self-portraits from 1995 to 2000 as an experiment in showing his own disease through art, watching his hand and eye change on the canvas.

  The research effort is vast. The US National Plan to Address Alzheimer's Disease holds a budget of 3.98 billion dollars for fiscal year 2026, and as of 2025, 182 clinical trials were testing 138 drugs against multiple targets. More than a century after Auguste D., the central question of why certain neurons are destroyed while others are spared, what researchers call selective vulnerability, remains unanswered.