BD² Unlocks Largest Psychiatric Dataset to Decode Bipolar Disorder

WASHINGTON, D.C. – March 30, 2026 – In a move aimed at transforming a historically fragmented field, the nonprofit BD²: Breakthrough Discoveries for thriving with Bipolar Disorder today announced the public release of the largest and most complex dataset ever assembled for a psychiatric condition. The initial data from the BD² Integrated Network Longitudinal Cohort Study (LCS) offers an unprecedented, multi-layered view into the lives of 615 individuals with bipolar disorder, marking a potential turning point in the quest for personalized treatments.

This release provides qualified researchers worldwide with access to a meticulously curated collection of clinical assessments, high-resolution brain scans, blood-based biomarkers, and continuous data from wearable sensors. The initiative seeks to dismantle the data silos that have long hindered progress and build a unified foundation for discovery.

“For decades, bipolar disorder research has been severely underfunded and deeply fragmented,” said Cara Altimus, PhD, CEO of BD², in the official announcement. “Without shared standards or shared data, meaningful progress was nearly impossible. BD² was created to fundamentally change that ecosystem so that we can make a positive change in the lives of those living with bipolar disorder.”

A New Foundation for a Fragmented Field

The challenge in bipolar disorder research has been its complexity and variability. The condition manifests differently in each person, making it difficult to develop one-size-fits-all treatments. The BD² Integrated Network aims to address this by following participants for five years, tracking biological changes and life experiences before, during, and after mood episodes to identify predictive patterns.

Central to this strategy is the study’s integration within a Learning Health Network (LHN). This model creates a continuous feedback loop between laboratories, clinics, and patients. As researchers uncover insights from the data, the LHN framework is designed to rapidly translate those findings into improved clinical practices. Simultaneously, real-world observations from doctors and patients can shape the next wave of research questions, ensuring the science remains grounded in patient needs.

“By bridging the gap between biological discovery and clinical practice, we can finally address long-standing care gaps and identify the precise factors that allow individuals with bipolar disorder to truly thrive,” noted Mark Frye, MD, a Scientific Director for the network and a professor at Mayo Clinic.

The Power of Multi-Modal Data

What makes the BD² dataset revolutionary is not just its size, but its depth and diversity. By integrating four distinct types of data, it offers a holistic picture of the condition.

• Clinical Assessments: Comprehensive reports from clinicians and patients cover everything from mood and cognition to social determinants of health, capturing the immense clinical variety of the disorder.

• High-Resolution Brain Imaging (MRI): Structural and functional MRI scans provide a window into the brain, allowing researchers to investigate how its wiring and activity relate to clinical outcomes and potentially identify objective biomarkers for the illness.

• Wearable Sensor Data (Fitbit): Continuous, real-world data on sleep, activity, and circadian rhythms offers a powerful tool for understanding the daily disruptions that are hallmarks of bipolar disorder. This passive monitoring can reveal subtle changes that may predict the onset of a mood episode.

• Biological Samples: Standardized blood tests explore the roles of inflammation, stress, and metabolic function. These biomarkers are emerging as critical tools for differentiating bipolar disorder from other conditions and tailoring medication choices.

“For the first time, we have the scale and breadth of data required to identify modifiable treatment targets and generate findings that can be directly translated into better care and improved lives,” emphasized Katherine E. Burdick, PhD, the network's Scientific Director and a professor at Columbia University.

Contextualizing the “Largest Dataset” Claim

While BD² bills its release as the “largest integrated dataset in the history of psychiatry,” it enters a landscape that includes colossal health data projects. The UK Biobank, for instance, involves over 500,000 participants and includes extensive genetic and imaging data. However, its focus is on broad population health, not the deep, longitudinal phenotyping of a single psychiatric condition.

The uniqueness of the BD² dataset lies in its focused mission and multi-modal integration specifically for bipolar disorder. While other datasets may be larger in participant numbers or focus on a single data type like genetics, none have combined clinical, imaging, wearable, and biological data with this level of rigor and specificity for a psychiatric illness. This specialized depth is what enables researchers to move beyond isolated observations and begin asking the complex questions necessary to unravel the condition.

Open Science, Big Challenges

The project is built on a foundation of open science, a principle intended to maximize its impact. “With open science as a priority and uniting multimodal data, clinical practice, and the lived experience of those with bipolar disorder, we are building the foundation required to move from scientific insight to sustained, real-world impact,” said Ekemini A.U. Riley, PhD, a BD² Program Board Member.

However, an initiative of this scale is not without significant hurdles. The foremost challenge is protecting the privacy of participants. Mental health data is exceptionally sensitive, and the organization has stressed that all identifiable information has been removed before making the dataset available. Maintaining this security and the trust of participants is paramount.

Furthermore, the generalizability of the findings will depend on the diversity of the initial cohort. For the research to be truly transformative, it must be applicable to the wide spectrum of people affected by bipolar disorder across different racial, ethnic, and socioeconomic backgrounds.

Finally, the technical complexity of harmonizing such disparate data types—from the gigabytes of an MRI scan to the daily logs of a Fitbit—is immense. BD² states that dedicated working groups have rigorously curated and standardized the data to reduce this burden, a crucial step in making the resource immediately useful to the global research community.

Despite these challenges, the launch of the BD² dataset represents a landmark investment in mental health. It signals a paradigm shift away from siloed research and toward a collaborative, data-driven approach. By providing an unprecedented, unified view into the biological and lived realities of the condition, BD² has laid the groundwork for a new era of discovery in the quest to help individuals with bipolar disorder not just survive, but thrive.