BRAIN SCANS in series have helped British scientists uncover the very earliest structural changes associated with Alzheimer's disease that take place well before any symptoms appear.

We all forget some things sometimes. We lose our keys, miss appointments, stash important papers in spots so secret that we can't, after the fact, remember where they are. It's entirely normal. But for many middle-aged and elderly individuals, such episodes prompt intense anxiety over Alzheimer's disease. The devastating neurodegenerative disease slowly corrodes the brain until memories, moods and many other cognitive traits are permanently lost. And perhaps most frightening, it seems to creep up on its victims without much warning.

Researchers have over the years devised psychological screening tests to try to identify people in the early stages of Alzheimer's disease. Apart from memory loss, there are nine other warning signsamong them difficulty performing familiar tasks; problems with language, such as remembering words; disorientation in time and space; and changes in personality. Scientists have also noted that susceptible people appear to have a greater sensitivity to the drug tropicamide, which is ordinarily used to dilate the pupils in a routine eye exam. But for the most part, diagnosis has been a process of elimination.

Recently, however, a team of scientists from University College London and the Imperial College School of Medicine described in the journal The Lancet a new way to actually predict who will develop Alzheimer's as early as three years before symptoms appear. Nick Fox, Martin Rossor and their colleagues discovered distinctive patterns of progressive atrophy in the brains of pre-diagnosis Alzheimer's patients by taking series of MRI images. "It raises the hope that we might one day be able to intervene with therapy at a very early stage," the authors write, "before the devastating cognitive decline of the disease has already become established."

In fact, several early or preventive therapies have shown promise in animal studies. Researchers have devised vaccines that appear to prevent the formation of amyloid plaques associated with the disease. And scientists have removed existing plaques in mice by applying anti-plaque antibodies directly to the animals' brains. Using gene therapy, other experimentalists have successfully restored axonsthe vital fibers that relay messages between brain cellsin aged monkeys. Stalling or stopping the progress of Alzheimer's disease might become possible several decades out.

So to try to pinpoint the very earliest structural brain damage associated with Alzheimer's, Fox and colleagues chose to study the offspring of patients with the familial form of the disease. These people have a 50 percent chance of developing Alzheimer's at a fairly young age. They recruited four such symptom-free individuals, as well as 20 people having a probable diagnosis of familial or sporadic Alzheimer's disease and 20 of their spouses as controls.

All underwent annual serial MRI brain scansin addition to comprehensive physical examinationsover the course of five to eight years. The scientists very carefully matched the MRIs from individuals so that they could map volume changes over time, creating what are known as voxel compression maps. "Voxel compression mapping provides a valuable technique for in-vivo monitoring of the progression of Alzheimer's disease," Fox says.

As it turned out, the four at-risk subjects all progressed to Alzheimer's over the course of the study, providing the scientists with a clear view of the disease in its very beginnings. All of the subjects lost cerebral volume, but the rate of that loss was much greater in the people with mild to moderate Alzheimer's. In this respect, the four at-risk individuals more closely resembled the Alzheimer's patients at the study's start: their initial median rate of atrophy was significantly higher than that of the controls, well before they showed symptoms.

In addition to the different rates of loss, the scientists discovered that the various subject groups also exhibited different patterns of loss. In the Alzheimer's patients, widespread regions of gray and white matter were consistently wasted; only the primary motor and sensory cortices, brain stem and cerebellum were spared. The changes in the healthy controls were also somewhat diffuse. But in the four at-risk subjects, progressive atrophy was most pronounced in the posterior cingulate, the parietal lobe and, in particular, the medial temporal lobe.

The scientists warn that "some caution is needed in extrapolating results found in familial Alzheimer's disease to the more common sporadic Alzheimer's disease." Nevertheless, their findings accord well with other lines of study into the disease's histological changes (plaques, tangles, etc.), which probably precede any structural changes. "We have been able to show a presymptomatic phase of three years or more of increased rates of tissue loss," the authors explain. "The recognition of a presymptomatic phase, which extends beyond the medial temporal lobe, implies that structural changes might start earlier and are more widely distributed that previously appreciated."