Brain scans may detect future Alzheimer's

Thursday April 14 2011

“Brain scans may be able to indicate potential Alzheimer's patients years before symptoms appear,” BBC News has reported. The BBC says that a small study has found that some parts of the brain may shrink up to a decade before outward signs of Alzheimer's disease appear.

This finding comes from a US study that looked at the thickness of nine regions of the brain (called AD-signature regions) in 65 cognitively normal elderly people and followed them for about a decade to see if they developed Alzheimer’s disease. It found that 55% of those with low thickness within the AD-signature regions developed the disease, compared with 20% of those with average thickness and none of those with high thickness. The results are of interest, but the small study size means the phenomenon will need to be confirmed in a larger sample. As yet, this method is not ready for use outside further research.

Knowing that a person is more likely to develop Alzheimer’s will only be helpful from a clinical perspective if treatments are available to slow or prevent the disease developing beyond this early stage. Currently, no such treatments are known, but the findings, if confirmed, may help researchers to study the very early stages of Alzheimer’s disease better, and potentially test treatments that may delay or prevent the disease’s progression.

Where did the story come from?

The study was carried out by researchers from the Massachusetts Alzheimer’s Disease Research Center in the US. It was funded by the US National Institutes of Health, the Alzheimer’s Association, the Mental Illness and Neuroscience Discovery Institute and the Illinois Department of Public Health. The study was published in the peer-reviewed medical journal Neurology.

The story was reported by the BBC News and the Daily Mail. Both sources report the research well and include caveats about the size of the study.

What kind of research was this?

This was a cohort study looking at whether the results of a brain scan could predict which individuals were more likely to develop Alzheimer’s disease in future. The brain changes that lead to Alzheimer’s are thought to start years before the symptoms of dementia appear and the researchers wanted to determine whether there was a non-invasive way of detecting these. In particular, they wanted to see whether a thinning of key areas of the cortex of the brain was associated with an increased risk of developing Alzheimer’s. The cortex is the outermost layer of the brain that contains various areas which control functions such as the senses, movement, and abstract thought.

This study design, where individuals are tested when they do not have symptoms and followed to see if they develop full-blown symptomatic disease, is the best way to answer this type of question.

What did the research involve?

The researchers assessed two separate samples of cognitively normal adults. They scanned their brains and measured the thickness of their cerebral cortex. They then followed them up over time to see who developed Alzheimer’s, and looked at whether those who did develop Alzheimer’s had lower cortex thicknesses than those who did not develop the disease.

The first sample included 33 community volunteers recruited at one hospital who were about 71 years old on average, and were followed up for an average of just over 11 years. The second sample included 32 community volunteers (average age about 76 years) recruited at another centre who were followed up for just over seven years on average. Individuals who had significant medical, neurological or psychiatric disease or major cardiovascular risk factors or disease were not allowed to participate. The study data suggest these elderly participants ranged in age from approximately 69 to 81 years of age, although it is not explicitly stated in the paper.

At the start of the study participants had a thorough assessment, including a clinical examination, neuropsychological tests and magnetic resonance imaging (MRI) brain scans. The brain scans were used to measure cortical thickness in nine areas of the cortex of the brain previously found to be affected in Alzheimer’s disease (called the AD-signature areas). The average thickness of these areas was calculated for each individual.

Participants also received annual clinical evaluations during the study. Only those who were cognitively normal at the start of the study, and for at least four years subsequently, were included in the current analysis. These follow-up evaluations identified those who developed mild cognitive impairment (MCI) or dementia. For the current study, only those with probable Alzheimer’s disease at their most recent assessment were included, not those with MCI or other forms of dementia.

For each sample, the researchers compared the average thickness of the cortex measurements of those who developed Alzheimer’s with those who did not. They also pooled the samples together and looked at what proportion of those who had low cortical thickness at the start of the study (one standard deviation below the average of the group or more) developed Alzheimer’s, compared with those with high  cortical thickness at the start of the study (one standard deviation above the average of the group or more), and those with average cortical thickness at the start of the study (i.e. not low or high).

What were the basic results?

During the study, eight out of the 33 people in the first sample developed Alzheimer’s, and seven out of the 32 people in the second sample.

In both sample groups the researchers found that, on average, those who developed Alzheimer’s had AD-signature areas that were 0.2mm thinner than those who did not develop the disease. Although this difference was small, it was statistically significant. The researchers then split the participants into groups based on their cortical thickness at the start of the study and looked at the prevalence of Alzheimer’s disease during follow-up:

  • 11 people had a low cortical thickness, of whom 55% went on to develop Alzheimer’s disease
  • 45 people had an average cortical thickness, 20% of whom went on to develop Alzheimer’s disease
  • 9 people had high cortical thickness, of whom none went on to develop Alzheimer’s disease

A reduction of one standard deviation in the thickness of the AD-signature areas of cortex was associated with a 3.4 time greater risk of developing Alzheimer’s during follow up.

How did the researchers interpret the results?

The researchers conclude that subtle but reliable changes in the areas of the brain affected by Alzheimer’s disease are detectable in cognitively normal individuals almost 10 years before onset of the disease. They say that these changes are a potentially important marker for early neurodegeneration.


This small study has suggested that measuring the thickness of certain areas of the brain may help to identify those who are at greater risk of developing Alzheimer’s. However, the study does have some limitations:

  • The number of people in the study was small (only 65 people). Ideally, these findings would be confirmed in a larger sample.
  • The individuals in this study were generally healthy, and may not be representative of the population as a whole.
  • The authors note that the two sample groups did have differing measurements, and the reasons for this were not clear. This needs further investigation.
  • A diagnosis of Alzheimer’s is difficult, and is only made once all other possibilities have been ruled out. Even then a diagnosis can only be finally confirmed by performing an autopsy. Ideally, the diagnoses of the individuals in this study would be confirmed in this way, to ensure they were correct.
  • Only just over half of those with thinner AD-signature areas developed Alzheimer’s disease over the 7-11-year follow-up period. Longer-term follow-up would be needed to determine what proportion of the remainder of individuals in this, and other cortical thickness groups, go on to develop the disease.

Knowing that a person is more likely to develop Alzheimer’s will only be helpful from a clinical perspective if treatments are available to slow or prevent the disease developing. Although there are some drugs available that can slow the progression of Alzheimer’s, they do not prevent or cure the disease. These drugs have also not been tested in individuals this early in the development of the disease, so their effects in this group would need to be assessed.

In addition, the cortical brain changes detected occur about 10 years before any symptoms, meaning any drugs used to slow the disease would need to be given for a long period before it could be established whether they had any effect on the disease.  Any potential benefits of any such treatment would have to be weighed up against any side effects, particularly if not all individuals with lower cortex thickness go  on to develop the disease. The fact that such individuals would not be expected to receive any benefit but will still be at risk of side effects would need to be considered.

At present, the findings do not have much direct clinical relevance but, if confirmed, will most likely help researchers to study the very early stages of Alzheimer’s disease better. This could potentially help with testing treatments that may slow or halt the progression of the disease.

Analysis by Bazian
Edited by NHS Choices