Slow down and be critical before using early warning signals in psychopathology

Early warning signals are considered to be generic indicators of a system’s accumulating instability and ‘critical slowing down’ prior to substantial and abrupt transitions between stable states. In clinical psychology, these signals have been proposed to enable personalized predictions of the impending onset, recurrence and remission of mental health problems before changes in symptoms occur, thereby facilitating timely therapeutic interventions. In this Perspective, we question the idea that early warning signals in a person’s emotion time series can predict changes in mental health symptoms. Using the empirical findings to date and the theoretical and methodological limitations inherent in their application, we argue that there is little support for the use of early warning signals based on critical slowing down in clinical psychology. Deepening our knowledge of the theoretical foundations of these predictors and improving their measurement are key to clarifying the potential and boundaries for their use in psychopathology. It is necessary to build on the insights gained from early warning signal studies and to improve and evaluate alternative methods, keeping in mind that clinical applications require prospective, real-time predictions that not only indicate whether, but also when, a specific person is likely to experience changes in their mental health.