Read more

February 25, 2021
2 min read
Save

Brain age serves as sensitive marker for information processing speed in MS

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

Brain age represented a valuable, volume-based principal component of MRI findings in patients with MS that correlated with information processing speed, according to results presented at the ACTRIMS virtual meeting.

“Aging is a complex process,” Stijn Denissen, a PhD student in the AIMS lab at the Center for Neurosciences of Vrije University in Belgium, said during his presentation. "Brain age is how old the brain looks. The difference with chronological age is the Brain-Predictive Age Difference."

Brain scan
Source: Adobe Stock
Stijn Denissen

That difference was found to be “predictive” in patients with MS as well as dementia, according to Denissen, who also noted that, in dementia, Brain-Predicted Age Difference correlated with cognitive function. In the current study, the researchers aimed to determine the value of brain age in MS based on MRI volumetry, particularly regarding information processing speed, an area of cognition that is “commonly affected” in MS.

In the first phase of the study, Denissen and colleagues trained a machine learning tool to predict chronological age based on brain volumes in a sample of 1,743 healthy controls from various publicly available collections of T1-weighted MRI. Ages in that sample ranged from 8 to 94 years, according to the study findings.

In the second phase of the study, the researchers used the algorithm to determine brain age based on 133 T1-weighted MRI from patients with MS in two clinical centers. Ages in that sample ranged from 14 to 75 years, according to Denissen and colleagues.

The researchers fitted a principal component analysis on healthy brain volumes, including whole brain, white matter, (cortical) grey matter and lateral ventricles. Denissen and colleagues then applied that model to brain volumes of patients with MS, pulling the first principal component that explained the greatest variance. They also evaluated information processing speed in the MS sample with the Symbol Digit Modalities Test.

"The essence of a principal component is that it is actually a 'summary' of a set of features, in this case volumetric features of the brain MRI," Denissen told Healio Neurology. "The first principal component is the one that provides the best summary of all principal components that we can extract from the volumetric features."

Denissen and colleagues found that brain age, chronological age and the first principal component “correlated strongly” with the Symbol Digit Modalities Test (r = –0.57, r = –0.43 and r = –0.56 respectively; P < .001), indicating that brain age showed the strongest relationship. Additionally, the researchers found that the first principal component and brain age appeared to carry “almost identical information,” according to the study results (r = 0.98; P < .001).

"The main point of the poster presentation is that this 'best summary of brain volumetry' approximates the 'brain age' that was also calculated from the volumetric features," Denissen told Healio Neurology.

In his presentation, Denissen noted that brain age represented "a sensitive marker for information processing speed in MS."

“It approximates our first principal component of the feature space," he said. "It could therefore be considered an explainable principal component of brain volumetric data.”

Editor's Note: This article was updated on March 1, 2021, to include original commentary from Stijn Denissen.