Obsessive-compulsive disorder (OCD) is an anxiety-related mental condition defined by intrusive and recurrent thoughts or obsessions, specific repetitive behaviors, or compulsions. In extreme circumstances, OCD patients are unable to lead normal lives.

To comprehend the origins of OCD, researchers applied the reinforcement learning theory. They showed how the cycle between obsession and compulsion can be exacerbated when learning parameters for reinforcement and punishment are significantly out of balance. In this paradigm, delays and both positive and negative prediction mistakes must be taken into account. The result of a certain decision is often available after a specific wait. As a result, recent decisions made within a specific time period should receive both praise and punishment. Credit assignment is what reinforcement learning theory refers to as and it is implemented as a memory trace.

The researchers noted that agents subconsciously learn obsessive-compulsive behaviour when the trace decay factor for memory traces of prior behaviours associated with negative prediction mistakes (n-) is significantly smaller than that connected to positive prediction errors (n).  168 healthy control participants and 45 OCD patients played a computer game with monetary prizes and punishments. As expected, given the computational properties of OCD, patients with OCD had substantially lower n- when compared to n.

Thus this computational model demonstrates that, despite our belief that we always make sensible choices, we occasionally unintentionally promote nonadaptive behaviours. According to this computational model, patients with severely unbalanced trace decay factors could not benefit from just behavioural therapy. (Hindustan Times, Sep 06, 2022)