A team led by scientists at the University of Cambridge describe using a combination of imaging techniques to examine how patterns of tau relate to the wiring of the brain in 17 patients with Alzheimer's disease, compared to controls. The study was published in the journal Brain

One hypothesis is that harmful tau starts in one place and then spreads to other regions, setting off a chain reaction. This idea – known as 'transneuronal spread' – is supported by studies in mice.

When a mouse is injected with abnormal human tau, the protein spreads rapidly throughout the brain; however, this evidence is controversial as the amount of tau injected is much higher relative to brain size compared to levels of tau observed in human brains, and the protein spreads rapidly throughout a mouse's brain whereas it spreads slowly throughout a human brain.

There are also two other competing hypotheses. The 'metabolic vulnerability' hypothesis says that tau is made locally in nerve cells, but that some regions have higher metabolic demands and hence are more vulnerable to the protein. In these cases tau is a marker of distress in cells.

The third hypothesis, 'trophic support', also suggests that some brain regions are more vulnerable than others, but that this is less to do with metabolic demand and more to do with a lack of nutrition to the region or with gene expression patterns.

Dr Cope and colleagues looked at the functional connections within the brains of the Alzheimer's patients – in other words, how their brains were wired up – and compared this against levels of tau. Their findings supported the idea of transneuronal spread, that tau starts in one place and spreads, but were counter to predictions from the other two hypotheses.

Confirmation of the transneuronal spread hypothesis is important because it suggests that we might slow down or halt the progression of Alzheimer's disease by developing drugs to stop tau from moving along neurons.

The same team also looked at 17 patients affected by another form of dementia, known as progressive supranuclear palsy (PSP) In PSP patients, tau tends to be found at the base of the brain rather than throughout.

The researchers found that the pattern of tau build-up in these patients supported the second two hypotheses, metabolic vulnerability and trophic support, but not the idea that tau spreads across the brain.

In Alzheimer's patients, they showed that as tau builds up and damages networks, the connections become more random, possibly explaining the confusion and muddled memories typical of such patients.

In PSP, the 'highways' that carry most information in healthy individuals receives the most damage, meaning that information needs to travel around the brain along a more indirect route. This may explain why, when asked a question, PSP patients may be slow to respond but will eventually arrive at the correct answer.