How the Brain Foresees Unpleasant Events
We just published a new collaborative study in PAIN investigating how the human brain learns to predict aversive (painful or unpleasant) events and avoid them. Led by Cleveland Clinic Research Scientist Dr. Raghavan Gopalakrishnana, our main goal was to pinpoint when and where the brain signals the surprise of pain. This signal is known as an “aversive prediction error” neural message, which is the difference between expected and actual pain, in a given context.
We tasked participants to avoid a mild painful stimulus in laboratory conditions, and discovered fast brain signals in a warning system of the brain that not only forecasts pain but also update behavior to steer clear of harm in the future.
We used magnetoencephalography (MEG) imaging to track millisecond brain activity. This level of time resolution is crucial because the brain’s prediction and surprise signals in this context are extremely rapid.
We found a specific brain circuit involved in predicting pain and registering surprise when outcomes aren’t what participants expected. This temporo-frontal circuit connects deep emotional and sensory regions to higher-level decision areas. Participants learned to choose an option less likely to yield a painful pressure on the thumb. As their expectations rose or fell, so did rhythmic activity in this brain network. In fact, moments before an outcome, certain regions in the frontal lobe (notably the lateral orbitofrontal cortex) and in the deep brain (midbrain/diencephalon areas) showed changes in an alpha active (~10 Hz) that reflected how much pain the person expected.
Higher expected pain caused more disruption of these alpha brain waves, as if the circuit was tensing up in anticipation. About 150 milliseconds after the outcome — pain or no pain — this circuit reacted strongly if it was a surprise, i.e., related to a prediction error. Notably, the deep midbrain region and the orbitofrontal cortex are wired together in reward and punishment pathways. Therefore, our study confirms that this midbrain–frontal connection is a key hub for aversive learning, helping us adjust our choices and avoid future pain.
We also observed how the brain uses this information to drive decisions. After a painful outcome, participants often switched their choices to avoid pain next time – and during those decisions, frontal areas like the anterior cingulate cortex (ACC) lit up in the alpha-band as well. The ACC and a neighboring region (dorsomedial prefrontal cortex) are thought to evaluate conflict and choices, and here they were more active when people had to rethink their strategy after a surprise shock. This finding suggests that the brain’s pain-prediction circuit doesn’t work alone; it connects with decision-making centers (ACC/dmPFC), which then help decide possible changes in behavior.
Together, we believe our findings paint a picture of an integrated system: first, predict (temporo-frontal circuit sets an expectation), then feel the outcome and compare (generate a prediction error if it’s unexpected), and finally adjust (ACC engages to modify behavior). Our data reveal a sophisticated loop that underlies the way we learn from painful mistakes and try to prevent them in the future.
Looking forward, our findings not only advance scientific understanding of pain and learning, but they also bring hope that we can better address chronic pain and anxiety disorders in the future, where the brain’s prediction abilities are often go awry. The brain’s crystal ball isn’t mystical at all; it’s grounded in fast electrical rhythms and circuits, and studies like this are helping us read it more clearly than ever.
—
The study is available in open-access from the journal PAIN