Can network analysis be used to inform better treatment plans in anxiety and depression randomised trials? Assessing underlying assumptions and predicting treatment efficacy.

Anxiety disorders are currently the most common mental health disorders, affecting 301 million (4%) of the world’s population, as of 2019. People with an anxiety disorder are up to six times more likely to be admitted to hospital for psychiatric care than those who do not suffer from such disorders. Similarly, the World Health Organisation has estimated that depression accounts for 10% of non-fatal disease burden worldwide. Depending on age, gender and country, depression is ranked in the top 15 most impactful conditions, and the World Health Organisation projects it to be the leading cause of disability by 2030.

Symptoms caused by conditions such as generalised anxiety disorder and depression can vary from person to person and there are a number of different ways these can be assessed. Among the most common is the use of surveys, such as clinician or patient reported outcome measures. These types of measures allow clinicians and researchers to measure the severity of common depression symptoms, such as low mood, low energy or insomnia.

Network analysis (NA) is a modern method to explore relationships and patterns in complex data, such as the outcome measures used to assess anxiety/depression. It helps to identify the most important or ‘central’ symptoms from a group (network) of related symptoms. This information that could guide the development of better patient treatment plans. For example, if depressed mood is a key or ‘central’ symptom in a depression network, focusing on treating depressed mood may improve the overall outcomes more effectively than targeting other, less central, symptoms. However, for NA to be reliable, the overall network structure needs to remains stable, and similar among different subgroups, such as age or gender. Our project aims to explore whether this is the case.

We aim to determine whether NA is a reliable method for use in depression studies by examining the stability over time and among different groups. To do this, data will be compiled from thousands of participants who took part in previously conducted multinational clinical trials of both anxiety and depression. Network models of each will be developed for both before treatment and 8 weeks after treatment starts. Then, we will use advanced analysis to assess whether both anxiety and depression symptoms can be reasonably compared across different groups, and determine which symptoms are most strongly related to treatment efficacy.