Researchers at the Indian Institute of Technology Kanpur (IIT Kanpur), in collaboration with Ganesh Shankar Vidyarthi Memorial (GSVM) Medical College, Kanpur, have found that a combination of brain and gastric electrical signals, along with clinical symptoms, can help predict antidepressant treatment outcomes within 7–10 days of starting treatment. The findings could enable earlier assessment of treatment response, compared with the conventional 4–6 week period typically needed to assess treatment response.
Depression affects an estimated 5% of adults worldwide and around 4.5% of India’s population. More than half of patients may not respond adequately to their first antidepressant, often requiring weeks of trial and error before an effective treatment is identified. The ability to identify likely treatment response quickly could help address a significant challenge in depression care.
“Our study shows that objective non-invasive brain and gut electrophysiological signals collected in about the first week of treatment already contain valuable information about treatment response to precisely guide the intervention,” said Dr. Pragathi Priyadharsini Balasubramani, Assistant Professor, Department of Cognitive Science, IIT Kanpur, and corresponding author of the study.
“We found that different symptom profiles were associated with distinct patterns of brain and gut physiology linked to treatment outcomes. Recognizing these biological subtypes helps explain why patients respond differently to the same medication and facilitates personalized treatment strategies,” said Amal Jude Ashwin Francis, PHD Scholar, Department of Cognitive Science, IIT Kanpur, and the first author of the study.
The study examined electrical activity in the brain and stomach using electroencephalography (EEG) and electrogastrography (EGG), respectively, together with clinical symptom data. The researchers found that these signals, recorded within the first 7–10 days of treatment, could help identify patients who were unlikely to respond to an antidepressant.
The study included 206 participants, including 144 treatment-naive patients with depression. Researchers recorded EEG and EGG signals at the start of treatment and again approximately one week later, and examined whether these early biological signals, together with clinical information, could predict treatment outcomes assessed 4–6 weeks after antidepressant therapy began.
The predictive model identified patients unlikely to respond to treatment with 84% sensitivity and 78% specificity during model evaluation. When tested on an independent patient cohort, the model achieved 77.3% overall accuracy, with 80% specificity and 71.4% sensitivity for identifying nonresponders.
The tools used in this study are currently accessible through Neuroclinical Innovative Solutions (NCIS) Private Limited, which received financial support from the Biotechnology Industry Research Assistance Council (BIRAC) for this translational project. This support has helped advance the research towards real-world clinical applications, enabling the development of technologies that can facilitate earlier and more personalized treatment decisions for patients with depression.
The research opens up possibilities for developing more timely and personalized approaches to depression treatment. At the same time, further studies across larger and more diverse patient groups will be needed to validate the approach.