Researchers from the Virginia Polytechnic Institute and State University (Virginia Tech) have identified distinct brain chemistry patterns that can differentiate Parkinson’s disease from essential tremor.
Their findings, published in Nature Communications, concern differences in dopamine and serotonin signaling during decision-making tasks, revealing a possible new direction for understanding these two common movement disorders.
“From adapting behavioral game theory into something that functions like a medical test to refining machine learning models that can see brain chemistry in real time, we’ve always aimed to translate insights into something clinically useful,” said co-senior author Read Montague, director of Virginia Tech’s Center for Human Neuroscience Research. “This study takes a clear step in that direction.”
Distinct neurochemical patterns emerge
The study analyzed data collected during deep brain stimulation (DBS) surgeries conducted in 2017 and 2018, where patients with either Parkinson’s disease or essential tremor completed decision-making tasks while playing a game involving monetary offers.
While patients made choices between fair and unfair offers, researchers used a machine learning-enhanced electrochemical technique to measure rapid fluctuations in dopamine and serotonin levels in a brain region involved in decision-making and reward processing – the caudate of the striatum.
In 2018, the researchers published early findings from this experiment, revealing the first-ever recordings of sub-second fluctuations in dopamine and serotonin during active decision-making in a conscious human participant.
In this latest study, researchers used computational tools to model how participants formed and updated their expectations during the task.
“The data had been collected as long as eight years ago by Montague and his team at the research institute and their collaborators at Wake Forest University, but we came back at it with better tools and a fresh perspective — and finally saw what was there all along,” said co-senior author William “Matt” Howe, an assistant professor in the School of Neuroscience at Virginia Tech.
Serotonin emerges as a key differentiator
The team found distinct neurochemical signatures tied to each disorder while playing the game. In patients with essential tremor, unfair offers that violated their expectations led to a “seesaw” chemical pattern: dopamine levels rose while serotonin dropped. This pattern was absent in patients with Parkinson’s disease. This kind of oppositional response, where one neurotransmitter rises while the other falls, has been seen in other studies of brain activity during decision making.
However, this neurochemical signaling pattern was found to be absent in patients with Parkinson’s disease.
The loss of dopamine-producing neurons is a well-known feature of Parkinson’s disease. However, when the researchers looked closer, they found that serotonin best distinguished the two conditions.
“What surprised us was how much serotonin stood out,” Howe said. “It wasn’t just that dopamine was disrupted, which was expected. It was that the normal back-and-forth between dopamine and serotonin was gone. There’s neither the serotonin dip nor the dopamine rise. It’s not just one system being disrupted — it’s the lack of that dynamic interaction that turned out to be the clearest difference between Parkinson’s and essential tremor.”
Serotonin has historically not been a very prominent figure in theories of Parkinson’s disease; these findings may shine powerful new light onto the disease.
Applying models from reinforcement learning
To uncover these differences, the team applied a form of machine learning known as “reinforcement learning”, which gradually improves its ability to detect patterns as more data is processed by receiving rewards or penalties for its actions. By reframing the task using an “ideal observer model” – which simulates how an optimal agent would perceive, interpret and respond to behavioral data – the team were able to extract new insights from human patient decision-making behavior.
“What they added was a computational model of what the subjects expected would happen,” Howe said. “When we reframed the data that way, we were able to reveal a difference in how the brain responded in these two patient groups.”
The researchers found that mismatches between what participants expected and what they received – a concept known as prediction error – were closely tied to serotonin activity and could be used to differentiate which disease the patient had.
“It’s very powerful to link moment-to-moment changes in internal beliefs — here what a person expects from others — to measurable chemical signals in the brain,” said study author Dan Bang, an associate professor at the Center of Functionally Integrative Neuroscience at Aarhus University in Denmark, and adjunct associate professor at Virginia Tech’s Fralin Biomedical Research Institute. “This opens a new window into how deeply human cognitive processes, like social evaluation, are shaped by disease.”
Revisiting older data with improved tools
Although the recordings used in this study were made several years ago, improved analytical tools allowed the team to uncover new insights.
“These models improve over time as they’re trained on more data,” said study author Seth Batten, a senior research associate in the Montague Lab. “The version we used in this study was far more refined than what we had early on. But just as important was the collaborative approach — bringing in new people with different expertise allowed us to see patterns we hadn’t recognized before.”
“It’s exciting to see that effort applied in a way that might help diagnose or stratify real clinical populations” added Montague.
The study highlights how combining computational neuroscience with intraoperative recordings can reveal subtle differences in brain signaling. It also underscores the potential of serotonin as a clinically relevant biomarker in differentiating Parkinson’s disease from essential tremor.
Reference: Hartle AE, Kishida KT, Sands LP, et al. Caudate serotonin signaling during social exchange distinguishes essential tremor and Parkinson’s disease patients. Nat Commun. 2025;16(1):7958. doi: 10.1038/s41467-025-63079-w
This article has been republished from the following materials. Note: material may have been edited for length and content. For further information, please contact the cited source. Our press release publishing policy can be accessed here.
This content includes text that has been generated with the assistance of AI. Technology Networks’ AI policy can be found here.