A shock to the complex dynamic system

Abstract

Psychological complaints are traditionally classified into diagnostic categories, such as depression or PTSD. However, individuals with the same diagnosis often experience vastly different symptoms. Moreover, many fall through the cracks or receive multiple diagnoses. Therefore, a new approach to psychopathology is needed, focusing on the unique situation of the individual and the interconnectedness of symptoms.

In this dissertation, we examined how symptoms develop over time and influence each other as parts of a complex, dynamic system. We studied a cohort of over 55,000 Japanese military personnel deployed after the 2011 earthquake, tsunami, and nuclear disaster, followed until 2018.

First, we mapped the course of PTSD symptoms, identifying five common symptom trajectories. Risk factors such as personal experience of the disaster, specific tasks, and overwork predicted a more severe course.

Next, we investigated symptom dynamics using an innovative technique: Dynamic Time Warping (DTW). This allowed us to determine which symptoms are related and which change first when PTSD increases or decreases. DTW also reflected connections in the network more reliably than traditional network models in many situations. These findings offer valuable tools for monitoring and preventing PTSD symptoms, and in the future, these techniques could aid in developing more personalized treatments.