We've known since the 1980s that some proteins behave like pathogens if they aren't folded properly.
These self-replicating molecular troublemakers called prions can cause brain tissue to deteriorate, as seen in Mad Cow disease and its human counterpart, Creutzfeldt-Jakob disease.
Over the past few decades, some intriguing studies have suggested that something similar could be happening in dementia.
The sticky clumps of amyloid beta protein found in the brains of people with Alzheimer's disease might behave like prions and may even spread to other people under very rare circumstances involving medical procedures.
Prion diseases do not transmit through direct contact between humans. But prions can jump from cows to humans through the consumption of contaminated meat, and kuru – a neurodegenerative prion disease found in Papua New Guinea – once spread through the ceremonial practice of eating deceased relatives' brains.
Similarly, if Alzheimer's disease can be transmitted via a misfolded protein, it won't be possible to catch it just by spending time with someone with the disease, researchers say.
But a small number of historical cases and a few animal studies have seeded concerns that unsuspecting patients could be infected with dangerous Alzheimer's proteins during neurosurgery if the same tools are used on multiple patients.
Patients could also be exposed when receiving a tissue transplant from a person with misfolded amyloid beta. In both cases, the effect wouldn't be seen until decades later as the protein takes a long time to propagate.
The first whisperings of this idea were heard in 2006, when researchers implanted transgenic mice with brain tissue from people who had died from dementia. These mice developed the classic amyloid beta plaque seen in people with Alzheimer's disease, whereas the control group did not.
In this experiment, the rate of plaque development was proportional to the amount of beta amyloid in the brain tissue, and the time taken to incubate. These are "patterns you would expect to see if the extracts caused the plaques," the researchers wrote.
Synthetic amyloid beta injected into the brains of mice had a similar, although less potent, effect.
However, it was not until 2015 that the concept of 'contagious' dementia started to make headlines.
In a small study published in Nature, researchers examined the brain tissue of eight young adults who had died from Creutzfeldt-Jakob disease.
Around 30–40 years earlier, as children, they were injected with growth hormones from the pituitary gland of human cadavers to treat their short stature – and had unwittingly been infected with the misfolded protein that causes Creutzfeldt-Jakob disease.
Four of these adults also had a substantial buildup of amyloid beta in their brains.
This kind of plaque is usually only seen in older people with moderate-to-severe Alzheimer's disease, and was a surprising finding for people who had died so young.
It raised the possibility that the cadaver injections had seeded people with amyloid-beta proteins, which had snowballed into larger plaque deposits later in life.
The use of cadaver-derived growth-hormone injections was halted in 1985, when researchers realized that a small proportion of children had received contaminated samples and were developing Creutzfeldt-Jakob disease.
Three decades later, researchers hunted down archival samples of the original injections and confirmed that they contained amyloid beta.
They injected these old samples into the brains of young mice, and found that they developed amyloid plaques and a related brain-bleed condition called cerebral amyloid angiopathy.
Another study published in 2018 found that eight people who developed cerebral amyloid angiopathy under the age of 60 had undergone brain surgery as children or teenagers.
"These findings raise the possibility that amyloid-beta pathology may be transmissible, as prion disease is, through neurosurgical procedures," the researchers concluded.
Earlier this year, in September, a large population study from Denmark and Sweden found that people who had received blood transfusions from people who later developed brain bleeds were more likely to develop brain bleeds themselves. The study "may suggest a transfusion-transmissible agent," the researchers reported.
This line of inquiry raises the specter of transmissible dementia that can pass from older to younger people if the same surgical tools are used.
This is concerning because common sterilization techniques, such as boiling, dunking in formaldehyde or drying, don't seem to affect amyloid beta, the sticky protein seen clogging the brains of people with Alzheimer's disease.
That said, neurosurgery for children is usually done at children's hospitals, where the tools used have probably never been near a patient with Alzheimer's.
The more we understand Alzheimer's the better we can treat it, but we've been burned before by amyloid-beta research that looked promising but failed to pay off. And the evidence is nowhere near conclusive yet.