Rutgers research explains why brains think at different speeds

Every moment, the brain balances signals that unfold at different speeds. Some arrive in milliseconds, such as a sudden sound or movement. Others build slowly, such as understanding context, meaning, or intent. A new study from Rutgers Health, published in Nature Communications, explains how the brain merges these fast and slow streams into coherent thought and behavior.

The research was led by Linden Parkes, an assistant professor of psychiatry at Rutgers Health and a core member of the Rutgers Brain Health Institute and the Center for Advanced Human Brain Imaging Research. Working with colleagues from Rutgers and Cornell University, the team studied how the brain’s wiring supports timing differences across regions.

“To affect our environment through action, our brains must combine information processed over different timescales,” Parkes said. “The brain achieves this by leveraging its white matter connectivity to share information across regions, and this integration is crucial for human behavior.”

Different brain areas specialize in holding information for different lengths of time. Scientists call these durations intrinsic neural timescales. Sensory regions respond quickly and let activity fade fast. Regions involved in memory, planning, and attention hold signals longer.

These timing differences do not appear randomly. They track with gene activity, cell types, and local structure. They also follow a consistent pattern across species, from humans to mice, suggesting deep biological roots.

To explore how timing and wiring interact, the Rutgers team analyzed brain imaging data from 960 people. The data came from the Human Connectome Project, which maps how brain regions connect through white matter pathways.

For each participant, researchers built a detailed map of brain connections, known as a connectome. They then used mathematical tools to model how activity flows across these networks over time.

“Our work probes the mechanisms underlying this process in humans by directly modeling regions’ intrinsic neural timescales from their connectivity,” Parkes said. “This draws a direct link between how brain regions process information locally and how that processing is shared across the brain to produce behavior.”

To understand how the brain moves between activity patterns, scientists often use network control theory. In this framework, brain structure shapes how easily activity can shift from one state to another.

A key concept is control energy. This measures how much effort the brain needs to change its overall activity pattern. Lower energy means the brain’s wiring naturally supports that transition. Higher energy suggests resistance.

"Traditional models assume every brain region behaves the same way internally. Each area is treated as losing activity at the same rate. This simplification helps math models stay stable, but it ignores decades of evidence showing that brain regions operate on very different timescales," Parkes explained to The Brighter Side of News.

"Our team addressed this gap by allowing each region to have its own internal timing. We introduced region-specific decay rates that reflect how quickly activity fades in each area. Instead of fixing these values, the model learned them directly from real brain data," he continued.

Allowing regions to follow their own timing patterns produced immediate gains. Transitions between brain states required less control energy across every tested condition. This drop happened even though the model did not aim to reduce energy directly.

The reduction occurred because the model better matched how the brain truly behaves. When internal dynamics became more realistic, the brain’s structure supported function more efficiently.

Tests confirmed that this effect depended on real biology. When researchers scrambled the brain’s wiring, energy increased. When they scrambled activity patterns, energy dropped unrealistically. Together, these tests showed that real structure and real activity must align for efficient brain control.

The study also addressed a practical problem. Many models assume that every brain region can receive its own control signal. Real brain stimulation methods cannot do this.

With region-specific timing, successful transitions required far fewer control points. In some cases, only about one-sixth of brain regions needed direct input. Standard models required nearly twice as many.

These transitions also became more stable. Individual regions influenced only certain patterns of brain activity, rather than activating everything at once. This selectivity made control more efficient.

The team tested whether the model’s timing estimates matched real biology. Regions with slower modeled dynamics also showed slower activity decay in real brain recordings.

These regions aligned with higher expression of genes linked to somatostatin-related inhibitory neurons. Faster regions aligned with parvalbumin-related markers. These patterns match known differences between association areas and sensory cortex.

The model also tracked intracortical myelin. Heavily myelinated regions, which transmit signals quickly, showed faster timescales. This further grounded the findings in physical brain features.

To test robustness, researchers applied the model to another human dataset and to a mouse brain connectome. In both cases, region-specific timing reduced control energy and matched known biology.

The mouse results were especially striking. Despite major differences in brain size and layout, the same principles held. This consistency suggests a fundamental rule of mammalian brain organization.

The final step focused on individual variation. The team analyzed nearly 1,000 individual brains rather than group averages.

People whose brain timing matched their optimized models more closely showed more flexible transitions between brain states. Control energy from the new model also predicted cognitive performance better than standard approaches.

“We found that differences in how the brain processes information at different speeds help explain why people vary in their cognitive abilities,” Parkes said.

The study involved collaboration with Avram Holmes, an associate professor at Rutgers Health, along with Ahmad Beyh, Amber Howell, and Jason Z. Kim of Cornell University.

These findings offer a clearer framework for understanding how brain structure supports thought and behavior. By linking timing, wiring, and biology, the work moves brain modeling closer to real-world conditions. This approach could improve how researchers study cognition and mental illness.

The results also point toward more precise brain stimulation strategies. Knowing which regions operate on slower or faster timescales could help guide targeted therapies for conditions such as schizophrenia, bipolar disorder, and depression. The team is already extending the model to study these disorders.

Over time, this work may help clinicians tailor interventions to individual brain dynamics. It also deepens understanding of how timing differences allow the brain to act efficiently, flexibly, and with purpose.

Research findings are available online in the journal Nature Communications.

Like these kind of feel good stories? Get The Brighter Side of News' newsletter.