Brain Shrinkage Does Not Follow a Pattern in Alzheimer's

Brain shrinkage in people who develop Alzheimer’s disease follows no specific or uniform pattern, according to a new study from University College London (UCL) and Radboud University.

The study, published in Alzheimer’s & Dementia, is the first to compare individual patterns of brain shrinkage over time in people with mild memory problems or Alzheimer’s disease against healthy controls.

Understanding brain shrinkage in Alzheimer's

Alzheimer’s disease is the most common cause of dementia, affecting nearly 7 million Americans. Understanding the development and progression of the disease is crucial to finding new ways to treat or prevent it.

Excess brain shrinkage in Alzheimer’s patients can be measured using MRI scans, but these shrinkage patterns vary hugely between individuals and can therefore easily be missed.

These patterns could provide valuable information on how individual patients’ thinking and memory abilities – or cognitive performance – can change over time. To capture these differences, UCL researchers used a normative modeling approach to gain insights into individual variability between Alzheimer’s patients.

Changes linked to disease progression, but no single pattern

Assessing brain scans for the “fingerprints” of disease, researchers found that, among people with mild memory issues, those who develop a greater number of “outlier” changes more quickly are more likely to develop Alzheimer’s. An outlier is classified as a specific brain area that, when adjusted for age and sex, has shrunk more than normal.

To develop these brain maps, the researchers used data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), comparing 3,233 MRI brain scans from 1,181 people with Alzheimer’s disease or mild memory issues to brain scan data collected from over 58,000 healthy people. The MRI scans were typically taken at one-year intervals, with most participants having two or three scans.    

After processing to assess the thickness and volume of the brain, the resulting individualized brain maps of Alzheimer’s patients could then be charted over time against healthy maps, serving as a “benchmark” for comparison.

Alzheimer’s patients had 15–20 outlier brain regions on average at the start of the study, rising to ~30 after 3 years. Patients with mild memory issues started with around 5–10 outliers but accumulated on average only 2–3 additional outliers over the same period. In both groups, a higher number of outlier regions was associated with poorer memory.

Among people with mild memory issues, those who went on to develop dementia within three years accumulated four new outliers every year. In contrast, people who remained with mild memory issues developed less than one new outlier per year on average.

However, though there were trends with the number of outlier changes, researchers also found that, despite some overlap, there was no uniform pattern to the way the brain shrank in those who developed Alzheimer’s. Though most participants starting out with similar-sized brains, different patterns of brain shrinkage were seen between individuals over time – such as differences in the progression or brain regions affected.

“If we look at which areas of the brain have shrunk most in different people with Alzheimer’s, there’s no single pattern,” explained James Cole, the study’s senior author and professor of neuroimage computing at UCL.

“The approach taken in our study means we can get a better sense of individual variability in Alzheimer’s disease progression,” Cole said. “By making these brain maps, which are unique ‘fingerprints’ of a patient’s brain health, we can spot if separate brain regions are changing and how rapidly.”

Dr. Serena Verdi, the study’s first author and research fellow in UCL’s Department of Computer Science, said: “While it’s true that some regions of the brain, such as the hippocampus, are particularly important in Alzheimer’s disease, we wanted to avoid focusing on specific regions in this study. Our results confirm that everyone is different, the regions affected by disease in one person may not be the same in the next.

“I think we need to pivot towards a new way of thinking to get away from the idea that ‘this area is important, this area isn’t’. The big picture and the individual variability contained within it, is what counts.”

Some of this individual variability may stem from the fact that many people with Alzheimer’s have more than one cause of cognitive illness, such as vascular dementia or fronto-temporal dementia, the authors said. Genetic and environmental factors, such as brain injuries, alcohol consumption or smoking are also thought to play a part.

Potential for far-reaching implications

Researchers say this new understanding could eventually help to predict how an individual’s illness will progress based on the early brain changes identified in scans. However, further research is needed to pinpoint which brain changes are predictive of specific future symptoms.

“We know that Alzheimer’s affects everyone differently,” said study co-author Jonathan Schott, a professor of neurology at UCL. “Understanding and quantifying this variability has important implications for the design and interpretation of clinical trials, and potentially in due course for counselling patients and developing personalized approaches to treatment.”

Reference: Verdi S, Rutherford S, Fraza C, et al. Personalizing progressive changes to brain structure in Alzheimer’s disease using normative modeling. Alzheimer’s & Dementia. 2024. doi: 10.1002/alz.14174

This article is a rework of a press release issued by UCL. Material has been edited for length and content.