AS Review

The Centre for Urban Mental Health at the University of Amsterdam and Amsterdam UMC is conducting an extensive systematic review and meta-analysis (more than 300.000 studies) to find out what factors and interaction of factors contribute to the onset, maintenance, and relapse of depressive-, anxiety-, and substance use disorders.


Common mental disorders, such as anxiety, substance use and depression, are highly prevalent and disabling conditions, affecting an estimated 20% of people globally each year. Prevalence rates are even higher when including the harmful use of substances such as tobacco and alcohol.

For a proper understanding of common mental disorders' etiology, maintenance and relapse, it is of utmost importance to have a complete overview of the evidence for all possible predictive and hypothesized factors that may contribute to their onset, relapse and maintenance.

Although thousands of such articles have been published on risk factors and mechanisms for common mental disorders, a clear and exhaustive overview of all possible risk- and preceding factors, and their potential interactions, still need to be included. We, therefore, investigated the current state of affairs on evidence in all published data on these topics.

The main aim of the current project is to use a systematic search with the help of machine learning to create a database that makes all potentially relevant studies on these topics available.

Studies Included

To search the published studies on common mental disorders, ASReview was used—a supervised machine learning software developed for reviewing academic research.

After searching the Medline, PsycInfo, Embase, and Scopus databases, initially, 2.739.753 records were identified. Due to this large number of hits, adjustments to the search terms were necessary. After adjustments of the search terms (reported on DANS, below), 70.065 papers were found for anxiety disorders, 83.371 for substance use disorders, and 161.760 for depressive disorders—a total of 315.196 studies.

For the procedure, search terms, search queries, selected key papers and de-duplication process, click here to see the search protocol document on OSF.

Screening Protocol

The screening protocol repository contains the procedure for screening the records identified in the search protocol. Click here to access it.


The PROSPERO pre-registration, which you can access by clicking here, was created for the MegaMeta project. The pre-registration describes the review question, the types of studies to be included, the population, interventions, exposure(s) and more.

Click here to access it.

All scripts, data files, and output files of the software are available via Zenodo (for Github code), the Open Science Framework (for protocols, output), and DANS (for the datasets) and are listed in the following sections.

Data Base

The final database will include all articles and meta-data on all possible factors involved in the onset, maintenance, and relapse/recurrence of depression, anxiety, and addiction.

The data and output are available in repositories for research data. Click on the bolded terms to access.


Server Setup

To allow for continuation of the screening project between different screeners, a server was set up. Details on how this server was set up can be found in the GitHub repository or in the Zenodo publication.

Hyperparameter Tuning

This repository stores the scripts and plugins used to create the final data files for screening titles and abstracts in a second round. These final project files were created using a classifier based on a convolutional neural network optimized using Optuna. Click on the linked words to access the Hyperparameter Tuning on GitHub or on Zenodo.


This repository stores the scripts to post-process the data after the title and abstract screening phase. Click on the linked words to access the Post-Processing on GitHub or on Zenodo.


Principal Investigator

Prof. Claudi Bockting

Research and Technical Team




This project is funded by a grant from the Centre for Urban Mental Health, University of Amsterdam, The Netherlands.