A recent article in Trends in Cognitive Sciences suggests a novel direction for the research of stress and its memory effects.
Stress is usually conceptualized as the sum of the physiological and psychological response to the disruption of homeostasis (the internal balance of the organism – e.g., blood sugar concentration, body temperature) or psychological well-being. Stress is a popular topic in neuroscience and psychology, but the many researchers interested in it use several distinct approaches and their interpretations of the data can be quite disparate, thus the current state of this area is a bit confused.
However, the idea that stressful events are well remembered, is supported by a wealth of observations and experiments. According to the authors, this effect is due to the prediction errors elicited by stressful events. Prediction error is a central notion in several influential models of nervous system function. These state that the brain is constantly trying to predict its inputs (rewards, punishments or other sensations) based on previous experiences. A prediction error is generated when the brain fails to accurately predict these variables and this signals a need to learn, to improve the predictions. The authors emphasize that stressful events are usually unexpected, unpredicted and thus they may generate a significant prediction error.
The most successful predictive coding framework is that of reinforcement learning. It showed that the learning of reward associations is dependent on dopaminergic neurotransmission. This neurotransmitter is currently not extensively studied in the context of stress, so studies focusing on its role may yield interesting findings. The research tradition of reinforcement learning may also be of great benefit to stress research. This approach will hopefully inspire new and valuable studies on stress and its effects on cognition.