Stanford’s scientists used Machine Learning to elucidate the relationship between stroke and depression by identifying a biomarker in stroke survivors. The findings may provide new hope for post-stroke patients struggling with mood disorders.

Stroke, a major cause of disability globally, often accompanies a complex web of challenges, including the common and persistent complication of depression. The prevalence of depression after stroke is striking, affecting up to 50% of survivors within the first year. Even after two years, around 20% of survivors continue to grapple with this emotional burden. In light of its adverse effects on quality of life and the potential for increased mortality, it is important to understand the underlying mechanisms of post-stroke depression.

We can now look at a stroke survivor’s
blood and predict their mood.

Author: Marion Buckwalter

Post-Stroke Depression and Its Unique Challenges

Irrespective of its widespread prevalence and significant impact, the molecular basis of post-stroke depression remains largely unknown. Its intricate interplay with the aftermath of a stroke distinguishes it from depression in other contexts. Factors like prior history of depression, gender, age, and lesion location have been implicated, but a comprehensive understanding has remained elusive. Moreover, if chronic post-stroke depression lasts for more than six months, it brings new challenges such as cognitive decline, exhaustion, and loss of independence, showing the possibility that the stroke itself is the cause of this biological process.

One promising avenue of exploration is inflammation, which has been shown to persist in stroke scar tissue for extended periods. Chronic peripheral inflammation, in turn, can lead to neuroinflammation. The inflammatory response has been linked to major depression, as it can cause neurochemical changes that lead to depressive symptoms. While studies in major depression have reported inconsistent findings regarding specific inflammatory markers, a comprehensive examination of chronic post-stroke depression has been lacking.

A Proteomic Approach

To bridge this gap, a proteomic approach was employed, scrutinizing 1,196 plasma proteins in samples from 85 chronic ischemic stroke survivors. The utilization of Olink technology, known for its precision and reproducibility in small sample volumes, provided a robust platform for analysis. This comprehensive protein array aims to capture both established and novel pathways associated with mood.

Connecting the Dots

The study employed a machine learning algorithm to discern whether plasma protein levels alone or in combination with age and time since stroke could predict mood scores. The integration of proteomics data into multivariable regression models, alongside relevant clinical features, yielded promising results. Notably, incorporating age and time since the stroke significantly enhanced the accuracy of mood predictions.

Individual Proteins: A Multifaceted Puzzle

At the individual protein level, no single entity emerged as a definitive predictor of mood. However, through univariate analyses, proteins strongly correlated with mood were identified. Remarkably, this list contained proteins previously implicated in major depression, establishing a potential link between the two conditions.

The Immune System’s Role

An intriguing finding was the overactivity of immune proteins in individuals with worse moods. This observation suggests a broader immune system involvement in chronic post-stroke depression. When the immune proteins get more active, it sets off a series of changes in how serotonin works and our brain’s flexibility. This all adds up and can lead to feeling depressed.


This groundbreaking study sheds light on the intricate relationship between plasma proteomic signatures and depression in chronic stroke survivors. The study utilized machine learning algorithms, proteomics, and patient information to explore the link between inflammation, body defense, and post-stroke depression, offering insights for improved treatment and outcomes.

Story Source: Reference Paper | Reference Article

Learn More:

 | Website

Dr. Tamanna Anwar is a Scientist and Co-founder of the Centre of Bioinformatics Research and Technology (CBIRT). She is a passionate bioinformatics scientist and a visionary entrepreneur. Dr. Tamanna has worked as a Young Scientist at Jawaharlal Nehru University, New Delhi. She has also worked as a Postdoctoral Fellow at the University of Saskatchewan, Canada. She has several scientific research publications in high-impact research journals. Her latest endeavor is the development of a platform that acts as a one-stop solution for all bioinformatics related information as well as developing a bioinformatics news portal to report cutting-edge bioinformatics breakthroughs.


Please enter your comment!
Please enter your name here