When the painkiller ibuprofen was given to rats in the laboratory injected with a substance to cause Alzheimer's it kept the disease at bay for 90 days.

One characteristic of Alzheimer's is that plaques - made of a protein called beta-amyloid - are formed in the brain.

It had been thought that these were what caused the degeneration of brain tissue associated with the disease, and that they were thus responsible for the impaired function and dementia it causes.

However, researchers at the University of Ulster have found that more damage is done to the brain when particles produced by the body's immune system came into contact with the plaques.

They said the two react to produce toxins that destroy brain tissue, while the amyloid plaques on their own probably do little damage.

Rat model

The team, led by Dr Eugene O'Hare, conducted their experiments on rats, injecting them to stimulate the growth of plaques and then monitored them.

Nothing happened for 35 days, and then the animals started to show signs of impaired function. The researchers realised that was when the immune system became aware of the plaques and started to fight them off.

They found that when immune system chemicals came into contact with the plaques, they produced inflammatory toxins that washed over surrounding brain cells and damaged them.

"Most probably, the amyloid itself has a minimal effect on the immune system on the neurons, especially in the early stages of the disease - in fact, it's the immune system's efforts to help that are causing a bigger problem," Dr O'Hare said.

Slow progress

The team combined their findings with those of another study, which noted that arthritis patients who took anti-inflammatory drugs had a lower incidence of Alzheimer's than the general population.

They gave ibuprofen - an anti-inflammatory drug - to another set of rats and found that it slowed down progress of the disease, with the rats showing no impairments 90 days later.

"The anti-inflammatory drug protected them from the memory and learning deficits that occur normally when they are untreated," Dr O'Hare said.

However, he stressed that this was only the case in their rat model.

A spokeswoman for the Alzheimer's Disease Society said it was too early to comment on the significance of the findings.