Mental health

What mental disorders do you study?

We estimate the burden due to 12 mental disorders in the GBD study. These follow definitions proposed by the Diagnostic and Statistical Manual of Mental Disorders and the International Classification of Diseases. They include:

• Depressive disorders
• Anxiety disorders
• Schizophrenia
• Autism spectrum disorders
• Bipolar disorder
• Conduct disorder
• Idiopathic developmental intellectual disability 
• Eating disorders
• Attention-deficit hyperactivity disorder (ADHD)

We estimate disability and total cases for all the mental disorders in GBD, but only estimate mortality for eating disorders. You can use the GBD Compare tool to find these measures, including years lived with disability, disability-adjusted life years, and more, across locations and time.

For example, explore the number of years lived with disability (YLDs) caused by depressive disorders for each country in 2021.

Open GBD Compare
 

What are the risk factors for mental disorders?

We currently investigate 3 risk factors for mental disorders in GBD: 

• Childhood sexual abuse
• Intimate partner violence
• Bullying victimization

We plan to continue investigating how other risk factors might also contribute to the burden, how they vary across different demographic groups, and how we can work to prevent them at a policy level.  

To include a risk factor in GBD, we need evidence of elevated risk from high-quality longitudinal cohort studies showing a significant elevation in risk of a mental disorder diagnosis following exposure to the risk factor. This is to ensure the causal direction of the risk. For example, people who are bullied may be more likely to develop a depressive disorder, but also people with a depressive disorder may be more likely to be bullied. The longitudinal cohort design allows for inference of the causal direction – that the risk factor came before the mental disorder and not the other way around.  

These studies are not very common as they are long and costly, representing a major challenge in adding new risk factors to GBD. 

Why are there so few years of life lost (YLLs) attributed to mental disorders in the GBD?

It is well established that individuals with mental disorders are at an increased risk of mortality. However, this is not well captured within GBD estimates where deaths and fatal burden can only be estimated for eating disorders.  

Individuals with mental disorders are at an increased risk of mortality from other diseases like ischemic heart disease, cancer, and diabetes, or from injuries – for example, death by suicide. These deaths are assigned directly to these causes in the GBD estimation of years of life lost (YLLs) rather than to the underlying mental disorder. Where there is sufficient evidence available, it is possible for us to use comparative risk assessment – a method used as part of GBD’s analysis of burden from risk factors – to estimate the contribution of mental disorders to premature mortality.  

For instance, in a separate piece of work conducted after GBD 2010, we found that the inclusion of burden from suicide occurring because of an underlying mental and substance use disorder would have increased their global ranking from fifth- to third-leading cause of disability-adjusted life years (DALYs) globally.  

Additional research to quantify the causal pathway between mental disorders and other health outcomes is required to allow us to expand on this work for other fatal outcomes. 

What are the areas of method development and innovation for mental disorder burden analyses in upcoming GBD studies?

We are actively exploring the inclusion of additional mental disorders that are not currently represented in GBD, for example personality disorders and binge eating disorder. We are also exploring how burden and severity of a disorder can change depending on access to treatment.

There is limited epidemiological data available for mental disorders in many parts of the world. With each iteration of GBD studies, we look to expand on the underlying epidemiological datasets informing burden of disease estimates through systematic literature reviews and working directly with researchers undertaking new mental health surveys.

The GBD mental disorders team welcomes any opportunity to discuss the addition of new epidemiological data sources within our GBD datasets.  

• We include data on the prevalence, incidence, remission/duration, excess mortality, or severity of selected mental disorders collected from 1980 onward.  
• Estimates disaggregated by age and sex are preferred, but those spanning broad age groups or males and females combined are also accepted.  
• Estimates with past-year recall or less are preferred to limit recall bias, but for disorders with little to no remission such as bipolar disorder and schizophrenia, lifetime recall is also accepted.  
• Mental disorders must be identified using survey instruments adhering to definitions in the Diagnostic and Statistical Manual of Mental Disorders or the International Classification of Diseases.  
• Study samples must be representative of the general population. Study samples where participants may be at lower or higher risk of mental disorders (e.g., pharmacological treatment samples, case studies, veterans, or refugee samples) are excluded.  
• No limitation is set on the language of publication.

Want to collaborate on our mental health research?

To get involved, please email [email protected].