mood disorders research

Chair's Welcome
History of Psychiatry at Yale
Giving Back
YPAA Executive Board
Alumni in the News
Alumni Webinars
Psychology Section
Assistant or Associate Professor – Psychiatric Emergency Services
Assistant Professor, Young Adult Service, Connecticut Mental Health Center
Bioinformatic Post-doctoral Associate, Xu Lab
Psychiatrist Specializing in Mood Disorders, Psychotic Disorders
Psychiatrist Specializing in Posttraumatic Stress Disorder
Yale Hispanic Psychiatry Fellowship Instructor
Full-Time Psychologist, Digestive Health Service
Department Leadership
Ladder Faculty
Research Faculty
Voluntary (Clinical) Faculty
Psychology Faculty
Adjunct Faculty
Emeritus Faculty
Junior Faculty Mentoring Program
Open Faculty Searches
Faculty Activity Map
Quick Links: Connecticut Mental Health Center
Quick Links: VA Connecticut Healthcare System
Quick Links: Yale New Haven Psychiatric Hospital
What is Recovery?
La Clinica Hispana
SATU—Substance use and Addiction Treatment Unit
West Haven Mental Health Clinic
Young Adult Service
Specialized Treatment in Early Psychosis (STEP)
OCD Research Clinic
Depression Research Program
Program for Recovery & Community Health
Services and Research
Leadership and Professional Staff
Graduate and Postgraduate Professional Training
CMHC Center for Digital Psychiatry
Peer Support at CMHC
Wellness Center
CMHC Cycles
Health & Wellness Resources
Ribicoff Research Facilities
If You Need Help
Spiritual Care Library
Community Services Network
CT Latino Behavioral Health System
CMHC News Archive
Connecticut WITS Training (Web Infrastructure for Treatment Services)
Talk to CMHC
Staff Directory
Clinical Programs
Research Programs
Educational and Training Programs
Yale New Haven Health
Yale New Haven Psychiatric Hospital
Psychiatric Emergency Service
Psychological Medicine Service
Success Stories
Publications
Juvenile Justice Mental Health
Therapeutic Junior and Senior High School
Outpatient Mental Health and Substance Abuse Treatment
Crisis Services
Employee Assistance Program (EAP)
Physician Staffing
Postgraduate Fellowship in Social Work
Behavioral Health Services at Hamden (BHSH)
Cedarhurst School
Connecticut Mental Health Center
Contracted Professional Clinical Staffing
Contracted Professional Physician Staffing
Employee Assistance Program
Employment Services
Mental Health Consultation in Juvenile Residential Settings
Peer Support Services
Psychological Assessments
Yale Medicine
Education & Training
Clinical Services
Olin Neuropsychiatry Research Center
The Steward House
News Archive
Research News
Photo Galleries
Faculty and Trainee Musicians
In Memoriam
Psychiatry@Yale Archive
Aging and Geriatric Psychiatry
Cognitive and Clinical Neuroscience
Ongoing Funded Research
Journal Club
Training Opportunities
Molecular Psychiatry
Public Psychiatry
Women's Behavioral Health Research
Get Involved
Sex & Gender & Alcohol Satellite Meeting
VA National Mental Health & Suicide Prevention ECHO
Laboratories
Faculty & Staff
Benefits Counseling
Assessing Money Mismanagement
ATM (Advisor Teller Money Manager)
Benefits Management/Payee Facilitation
GIFT (Gaining Immediate Financial Training)
Behavioral Pharmacology Laboratory
Center for Genes and Behavior
Center for the Translational Neuroscience of Alcoholism (CTNA)
Center for Wellbeing of Women and Mothers
Cocaine Research Clinic
Clinical Trials & Projects
How to Find Us
Implementation
Clinical Tools
Demonstration Videos
Neural Transplant and Neurobehavior Program
Participate in a Study

In the News

Opportunities.

PRIME Center
What is Clinical High Risk?
Our Services
Schedule Evaluation
Join a Research Study
Assessment Training
Program for Recovery and Community Health
Psychosis Program at Yale
Binge-Eating Disorder Studies
Bariatric Studies
Leadership and Staff
Presentations
Ongoing Research
News & Media
Support Our Research
Repetitive Transcranial Magnetic Stimulation (rTMS) Research Clinic
Schizophrenia Research Clinic
Smoking Cessation and Alcohol Treatment Clinic
VA Alcohol Research Center
Treatment for SAD
Risks of Light Treatment
How to Obtain a Light Box
Other Light Treatment Options
Collaborators
PD Ketamine Trial
Learning Based Recovery Center
Partnerships
Research and Collaborations
Clinical Trials
Funding Opportunity
Dissemination of Information
News & Events
Psychiatry & Behavioral Health Patient Experience
Biological Sciences Training Program (BSTP) in Psychiatry
Integrated Mentored Patient-Oriented Research Training
Research Training Fellowship in Substance Abuse
Research Training in Functional Disability Interventions
Resident/Intern Substance Use Research Education
Translational Research in Alcoholism Training Program
Active Addictive Behavior Clinical Trials
Active Mental Health Clinical Trials
Support Research
Department of Psychiatry Art Committee
2023 Grand Rounds
2022 Grand Rounds
2021 Grand Rounds
2020 Grand Rounds
2019 Grand Rounds
2018 Grand Rounds
First and Second Year Pre-Clinical Studies
Psychiatry Clerkship
Fourth Year Psychiatry Electives
PA Student Clerkship
Yale Medical Students Psychiatric Association
Why Train at Yale?
Clinical Curriculum
Research Training
Psychotherapy Training Program
Core Skills Training Program
Global Mental Health Program
The De-prescribing Elective
Individualized Education Plan
The Solnit Integrated Training Program
Training Program
Application
Research Clinics/Programs
Events & Resources
Seminar Series
Training Sites
Residency News
Leadership in Psychiatric Care Delivery
Neuroscience Research Training Program
Psychotherapy Distinction Track
Asian & Pacific Islander Resident Association
Cinema and Psychiatry
Climate Change and Mental Health Action
History, Humanities, and Health Interest Group
Interventional Psychiatry
Latinx/Hispanic Affinity Group
Psychodynamic Psychotherapy Group
Psychiatry Mental Health Policy Interest Group
Psychiatry Technology Group
Social Psychiatry
Solomon Carter Fuller Association
Women's Mental Health
Women in Psychiatry
Yale Global Mental Health Program
Year 1 - Taiwo
Year 2 - Ashley and Stephanie
Year 3 - Daniel
Year 4 - Tommy
Residency Roundtables
Alert & Critical Incident
Photo Album
Moonlighting
PRA Calendar
Resident Portal (password protected)
Residency Program Directors
Site Training Directors
Program Staff
Current Residents
Recent Graduates
Compensation & Benefits
Fellowship & Award Opportunities
PGY2-4 Positions
When You Visit
PGY2 and PGY4
Solnit Integrated
Forensic Psychiatry
Didactic Curriculum
Clinical Experience
Graduated Fellows
PPF Alumni Spotlight
Silver Hill Fellowship
Psychosocial Rehabilitation
Consultation-Liaison Psychiatry Fellowship
Research Fellowships
Post-MSW Fellowships
Program Directors
Faculty Mentors
Facilities and Resources
Social Justice and Health Equity Roster
Social Justice and Health Equity Publications and Presentations
GROW Core Curriculum
Research & Evaluation
Job Openings
2024 Speakers
2024 Sponsors and Exhibitors
2024 Posters
2023 Posters
Coordinators and Staff
Program, Presenters, and Parking
Video Archive
Past Award Recipients
RebPsych Conference
Submit Presentation
Travel to New Haven
Rebellious Lawyering
RebPsych 2020
RebPsych 2018
RebPsych 2017
2022 Webinars on Health Care of Refugees
Programme 2021
2021 Webinars on Health Care of Afghan Refugees in CT
Yale Stress and Resilience Town Halls
Kaye and Damisah
Enriquez-Geppert
Our Research
Legislative Advocacy Program (LEAD)
Meeting Minutes
Draft: Vision, Mission, and Values for the Yale Department of Psychiatry
Diversity, Equity, Inclusion, and Anti-Racism (DEIA) Resources
Steering and Subcommittee Grand Rounds and Newsletter Reports
Preparing a Grant (Pre-Award)
Managing Sponsored Awards (Post Award)
Transaction Team
Contacting Staff at NIH Institutes and Centers
Updates & Announcements
FY25 Faculty Compensation Information
Hiring Faculty and Postdocs
Search for the Residency Program Director for the Department of Psychiatry
Faculty Advisory Council
CV Part 2: Revised Structure and Descriptions
Junior Faculty Orientation
Project Synapse

INFORMATION FOR

Residents & Fellows
Researchers

Yale Mood Disorders Research Program (MDRP)

Our mission.

Welcome! The Yale Mood Disorders Research Program (MDRP) is dedicated to understanding the causes of mood and related disorders, and suicide risk, across the lifespan. The MDRP brings together a multi-disciplinary group of scientists from across the Yale campus in a highly collaborative research effort. We use a wide variety of scientific methods to study how genetic and environmental factors affect the brain and lead to the development of mood disorders. Goals of the MDRP include the identification of biological markers for mood disorders and discovery of new detection and treatment strategies. We hope that these research efforts will lead to new and improved methods for early detection and treatment to reduce the suffering of mood disorders and suicide.

Examples of our Current Research Projects:

Studies on changes in the brain, mood symptoms and suicide risk with Brain Emotion Circuitry Targeted Self-Monitoring and Regulation Therapy (BE-SMART) in teens and young adults with bipolar and major depressive disorder, and at risk for bipolar disorder (e.g., parent or sibling with bipolar disorder)
Studies of symptom and behavioral changes with digital technologies such as actigraphy “watches” and smartphones
Multimodal magnetic resonance imaging (MRI) studies of brain circuitry differences in older adults with bipolar disorder, and how they change over time
Multimodal MRI study to identify brain markers of suicidal thoughts and behavior in older adults with mood and psychotic disorders
Study of novel mechanisms implicated in bipolar disorder from blood cells of stem cells and stem cell-derived brain cells
Study of brain changes related to genetic variations that are associated with bipolar disorder and suicide

Participating in a Yale Mood Disorders Research Program Study

Yale mood disorders research program participates in nami's first in-person walk since start of pandemic.

The Yale Mood Disorders Research Program (MDRP) celebrated Mental Health Awareness Month and the National Day of Hope in May by participating in the National Alliance for Mental Illness (NAMI) walk in Hartford's Bushnell Park on May 21. The event was especially important as it was NAMI's first in-person walk since 2019 and brought individuals, families, and loved ones together to support one another and walk to raise awareness of mental illness and to reduce the stigma. Pictured are Bernadette Lecza, LPC, left, and Erin Carrubba, LPC, right, study coordinators and therapists at MDRP who shared their team's mission to understand the science of mood disorders in hopes that their research efforts will lead to new and improved methods for early detection and treatment to prevent the suffering of mood disorders and suicide. Yale MDRP is directed by Hilary Blumberg, MD, John and Hope Furth Professor of Psychiatric Neuroscience and Professor of Psychiatry, and in the Child Study Center and of Radiology and Biomedical Imaging.

Transforming the understanding and treatment of mental illnesses.

Información en español

Celebrating 75 Years! Learn More >>

About the Acting NIMH Director
Advisory Boards and Groups
Strategic Plan
Offices and Divisions
Careers at NIMH
Staff Directories
Getting to NIMH

Mood Disorders Research Program

This program supports translational research on the etiology, core features, course, assessment, treatment, and prevention of mood disorders, including research aiming to improve understanding of their common and distinguishing features. It emphasizes a mechanistic, dimensional approach to psychopathology as outlined in the RDoC initiative and experimental therapeutics model. Ultimately, this program’s mission is to translate basic science and neuroscience into improved models, assessments, and treatments for prevention and intervention leading to significant public health impact.

Areas of Emphasis

Identification and validation of phenotypes and endophenotypes emerging from integrative research in psychopathology, neuroscience, and pathophysiology as clinical targets in mood disorders.
Advancing models of mood disorders based on emergent phenotypes and endophenotypes to improve precision of nosology and conceptualization.
Refining assessment of processes underlying psychopathology to improve definition and understanding of factors underlying mood disorders.
Developing novel psychosocial treatments targeting specific cognitive, behavioral, affective, or interpersonal components of psychopathology from prevention to intervention.

Alexander M. Talkovsky, Ph.D. Program Chief 6001 Executive Boulevard, Room 7131, MSC 9637 301-827-7614, [email protected]

Psychiatry and Behavioral Science Building on Quarry Road, Palo Alto

Pioneering Solutions for Depression and Bipolar Disorder

As many as 17 percent of Americans will experience some form of mood disorder in their lifetime—such as depression or bipolar disorder. Despite their prevalence, mood disorders remain one of the most widespread, misunderstood, and stigmatized health issues we face. Their impact reverberates far beyond an individual’s life. Families, friends, communities, economies—all are affected by these diseases. Depression now ranks first in rate of incidence among all illnesses worldwide; bipolar disorder ranks fifth. Tragically, suicide, often triggered by a mood disorder, takes more than one million lives worldwide every year.

Stanford Mood Disorders Center at a Glance

1,000 inpatients per year
8,000 total patients annually
55 clinical trials completed since 2003
1,000 undergraduates and 130 postdoctoral fellows trained between 2004 and 2017
250 medical students per year
200 leading-edge interdisciplinary research projects under way

Although the incidence and impact of mood disorders are undeniably on the rise, hope for solutions has never been higher. Through the Stanford Mood Disorders Center and Research Program, scientists and physicians are building on Stanford’s traditions of excellence, healing, and innovation. They are leveraging new knowledge of genetics and the brain’s molecular processes, and drawing on new techniques for imaging and healing the brain. Merging Stanford’s expertise across disciplines—psychiatry, biology, engineering, and myriad other fields—they are streamlining the process of translating laboratory discoveries into breakthrough treatments.

Through the research programs at the Stanford Mood Disorders Center, Stanford has led the quest for new knowledge and therapies for mood disorders. Today the center is expanding its reach and mobilizing Stanford’s diverse expertise toward a powerful shared mission: to overcome mood disorders through innovation and compassion.

Upcoming Event

Suicide-focused assessment & treatment webinar.

Wednesday, October 30, 2024 8:30 am – 2:30 pm PDT / 11:30 am – 5:30 pm EDT

Registrants will receive an extensive online resource that was updated in March 2024 and provides easily accessible references and guidelines to help clinicians approach the clinical challenge of a suicidal patient.

More Information & Registration

In the News

Dr. Alan Schatzberg describes the legal uses and potential issues related to Ketamine in this BBC interview. August 2024

Transformative Therapies & Technologies

Never before have we been so close to breakthroughs that will transform our approach to mood disorders, delivering advanced solutions for sufferers, their families, their friends, and their communities. Stanford is leading the way in understanding brain processes and transforming new knowledge, rapidly and efficiently, into new therapies and technologies.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Publications
Account settings
My Bibliography
Collections
Citation manager

Save citation to file

Email citation, add to collections.

Create a new collection
Add to an existing collection

Add to My Bibliography

Your saved search, create a file for external citation management software, your rss feed.

Search in PubMed
Search in NLM Catalog
Add to Search

Mood Disorders

PMID: 34881733
DOI: 10.1212/CON.0000000000001051

Purpose of review: This comprehensive review of mood disorders brings together the past and current literature on the diagnosis, evaluation, and treatment of the depressive and bipolar disorders. It highlights the primary mood disorders and secondary neurologic causes of mood disorders that are commonly encountered in a clinical setting. As the literature and our understanding evolve, recent additions to the current literature are important to bring forth to the readers.

Recent findings: Advancements in clinical medicine have strengthened our understanding of the associations of neurologic and psychiatric diseases. This article highlights the medications frequently used with newly identified mood disorders and the common side effects of these medications. A paradigm shift has moved toward newer treatment modalities, such as the use of ketamine, repetitive transcranial magnetic stimulation, and complementary and alternative medicine. The risks and benefits of such therapies, along with medications, are reviewed in this article.

Summary: Mood disorders are extraordinarily complex disorders with significant association with many neurologic disorders. Early identification of these mood disorders can prevent significant morbidity and mortality associated with them. With further expansion of pharmacologic options, more targeted therapy is possible in improving quality of life for patients.

PubMed Disclaimer

Publication types

Search in MeSH

Related information

Linkout - more resources, full text sources.

Ovid Technologies, Inc.
Wolters Kluwer
MedlinePlus Consumer Health Information
MedlinePlus Health Information

Research Materials

NCI CPTC Antibody Characterization Program

Miscellaneous

NCI CPTAC Assay Portal

Citation Manager

NCBI Literature Resources

MeSH PMC Bookshelf Disclaimer

The PubMed wordmark and PubMed logo are registered trademarks of the U.S. Department of Health and Human Services (HHS). Unauthorized use of these marks is strictly prohibited.

- Google Chrome

Intended for healthcare professionals

My email alerts
BMA member login
Username * Password * Forgot your log in details? Need to activate BMA Member Log In Log in via OpenAthens Log in via your institution

Search form

Advanced search
Search responses
Search blogs
Diagnosis and...

Diagnosis and management of bipolar disorders

Related content
Peer review
1 Precision Medicine Center of Excellence in Mood Disorders, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
2 Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, USA
Correspondence to: F S Goes fgoes1{at}jhmi.edu

Bipolar disorders (BDs) are recurrent and sometimes chronic disorders of mood that affect around 2% of the world’s population and encompass a spectrum between severe elevated and excitable mood states (mania) to the dysphoria, low energy, and despondency of depressive episodes. The illness commonly starts in young adults and is a leading cause of disability and premature mortality. The clinical manifestations of bipolar disorder can be markedly varied between and within individuals across their lifespan. Early diagnosis is challenging and misdiagnoses are frequent, potentially resulting in missed early intervention and increasing the risk of iatrogenic harm. Over 15 approved treatments exist for the various phases of bipolar disorder, but outcomes are often suboptimal owing to insufficient efficacy, side effects, or lack of availability. Lithium, the first approved treatment for bipolar disorder, continues to be the most effective drug overall, although full remission is only seen in a subset of patients. Newer atypical antipsychotics are increasingly being found to be effective in the treatment of bipolar depression; however, their long term tolerability and safety are uncertain. For many with bipolar disorder, combination therapy and adjunctive psychotherapy might be necessary to treat symptoms across different phases of illness. Several classes of medications exist for treating bipolar disorder but predicting which medication is likely to be most effective or tolerable is not yet possible. As pathophysiological insights into the causes of bipolar disorders are revealed, a new era of targeted treatments aimed at causal mechanisms, be they pharmacological or psychosocial, will hopefully be developed. For the time being, however, clinical judgment, shared decision making, and empirical follow-up remain essential elements of clinical care. This review provides an overview of the clinical features, diagnostic subtypes, and major treatment modalities available to treat people with bipolar disorder, highlighting recent advances and ongoing therapeutic challenges.

Introduction

Abnormal states of mood, ranging from excesses of despondency, psychic slowness, diminished motivation, and impaired cognitive functioning on the one hand, and exhilaration, heightened energy, and increased cognitive and motoric activity on the other, have been described since antiquity. 1 However, the syndrome in which both these pathological states occur in a single individual was first described in the medical literature in 1854, 2 although its fullest description was made by the German psychiatrist Emil Kraepelin at the turn of the 19th century. 3 Kraepelin emphasized the periodicity of the illness and proposed an underlying trivariate model of mood, thought (cognition), and volition (activity) to account for the classic forms of mania and depression and the various admixed presentations subsequently know as mixed states. 3 These initial descriptions of manic depressive illness encompassed most recurrent mood syndromes with relapsing remitting course, minimal interepisode morbidity, and a wide spectrum of “colorings of mood” that pass “without a sharp boundary” from the “rudiment of more severe disorders…into the domain of personal predisposition.” 3 Although Kraepelin’s clinical description of bipolar disorder (BD) remains the cornerstone of today’s clinical description, more modern conceptions of bipolar disorder have differentiated manic depressive illness from recurrent depression, 4 partly based on differences in family history and the relative specificity of lithium carbonate and mood stabilizing anticonvulsants as anti-manic and prophylactic agents in bipolar disorder. While the boundaries of bipolar disorder remain a matter of controversy, 5 this review will focus on modern clinical conceptions of bipolar disorder, highlighting what is known about its causes, prognosis, and treatments, while also exploring novel areas of inquiry.

Sources and selection criteria

PubMed and Embase were searched for articles published from January 2000 to February 2023 using the search terms “bipolar disorder”, “bipolar type I”, “bipolar type II”, and “bipolar spectrum”, each with an additional search term related to each major section of the review article (“definition”, “diagnosis”, “nosology”, “prevalence”, “epidemiology”, “comorbid”, “precursor”, “prodrome”, “treatment”, “screening”, “disparity/ies”, “outcome”, “course”, “genetics”, “imaging”, “treatment”, “pharmacotherapy”, “psychotherapy”, “neurostimulation”, “convulsive therapy”, “transmagnetic”, “direct current stimulation”, “suicide/suicidal”, and “precision”). Searches were prioritized for systematic reviews and meta-analyses, followed by randomized controlled trials. For topics where randomized trials were not relevant, searches also included narrative reviews and key observational studies. Case reports and small observations studies or randomized controlled trials of fewer than 50 patients were excluded.

Modern definitions of bipolar disorder

In the 1970s, the International Classification of Diseases and the Diagnostic and Statistical Manual of Mental Disorders reflected the prototypes of mania initially described by Kraepelin, following the “neo-Kraepelinian” model in psychiatric nosology. To meet the primary requirement for a manic episode, an individual must experience elevated or excessively irritable mood for at least a week, accompanied by at least three other typical syndromic features of mania, such as increased activity, increased speed of thoughts, rapid speech, changes in esteem, decreased need for sleep, or excessive engagement in impulsive or pleasurable activities. Psychotic symptoms and admission to hospital can be part of the diagnostic picture but are not essential to the diagnosis. In 1994, Diagnostic and Statistical Manual of Mental Disorders , fourth edition (DSM-IV) carved out bipolar disorder type II (BD-II) as a separate diagnosis comprising milder presentations of mania called hypomania. The diagnostic criteria for BD-II are similar to those for bipolar disorder type I (BD-I), except for a shorter minimal duration of symptoms (four days) and the lack of need for significant role impairment during hypomania, which might be associated with enhanced functioning in some individuals. While the duration criteria for hypomania remain controversial, BD-II has been widely accepted and shown to be as common as (if not more common than) BD-I. 6 The ICD-11 (international classification of diseases, 11th revision) included BD-II as a diagnostic category in 2019, allowing greater flexibility in its requirement of hypomania needing to last several days.

The other significant difference between the two major diagnostic systems has been their consideration of mixed symptoms. Mixed states, initially described by Kraepelin as many potential concurrent combinations of manic and depressive symptoms, were more strictly defined by DSM as a week or more with full syndromic criteria for both manic and depressive episodes. In DSM-5, this highly restrictive criterion was changed to encompass a broader conception of subsyndromal mixed symptoms (consisting of at least three contrapolar symptoms) in either manic, hypomanic, or depressive episodes. In ICD-11, mixed symptoms are still considered to be an episode, with the requirement of several prominent symptoms of the countervailing mood state, a less stringent requirement that more closely aligns with Kraepelin's broader conception of mixed states. 7

Epidemiology

Using DSM-IV criteria, the National Comorbidity Study replication 6 found similar lifetime prevalence rates for BD-I (1.0%) and BD-II (1.1%) among men and women. Subthreshold symptoms of hypomania (bipolar spectrum disorder) were more common, with prevalence rate estimates of 2.4%. 6 Incidence rates, which largely focus on BD-I, have been estimated at approximately 6.1 per 100 000 person years (95% confidence interval 4.7 to 8.1). 8 Estimates of the incidence and lifetime prevalence of bipolar disorder show moderate variations according to the method of diagnosis (performed by lay interviewers in a research context v clinically trained interviews) and the racial, ethnic, and demographic context. 9 Higher income, westernized countries have slightly higher rates of bipolar disorder, 10 which might reflect a combination of westernized centricity in the specific idioms used to understand and elicit symptoms, as well as a greater knowledge, acceptance, and conceptualization of emotional symptoms as psychiatric disorders.

Causes of bipolar disorder

Like other common psychiatric disorders, bipolar disorder is likely caused by a complex interplay of multiple factors, both at the population level and within individuals, 11 which can be best conceptualized at various levels of analysis, including genetics, brain networks, psychological functioning, social support, and other biological and environmental factors. Because knowledge about the causes of bipolar disorder remains in its infancy, for pragmatic purposes, most research has followed a reductionistic model that will ultimately need to be synthesized for a more coherent view of the pathophysiology that underlies the condition.

Insights from genetics

From its earliest descriptions, bipolar disorder has been observed to run in families. Indeed, family history is the strongest individual risk factor for developing the disorder, with first degree relatives having an approximately eightfold higher risk of developing bipolar disorder compared with the baseline population rates of ~1%. 12 While family studies cannot separate the effects of genetics from behavioral or cultural transmission, twin and adoption studies have been used to confirm that the majority of the familial risk is genetic in origin, with heritability estimates of approximately 60-80%. 13 14 There have been fewer studies of BD-II, but its heritability has been found to be smaller (~46%) 15 and closer to that of more common disorders such as major depressive disorder or generalized anxiety. 15 16 Nevertheless, significant heritability does not necessarily imply the presence of genes of large effect, since the genetic risk for bipolar disorder appears likely to be spread across many common variants of small effect sizes. 16 17 Ongoing studies of rare variations have found preliminary evidence for variants of slightly higher effect sizes, with initial evidence of convergence with common variations in genes associated with the synapse and the postsynaptic density. 18 19

While the likelihood that the testing of single variants or genes will be useful for diagnostic purposes is low, analyses known as polygenic risk studies can sum across all the risk loci and have some ability to discriminate cases from controls, albeit at the group level rather than the individual level. 20 These polygenic risk scores can also be used to identify shared genetic risk factors across other medical and psychiatric disorders. Bipolar disorder has strong evidence for common variant based coheritability with schizophrenia (genetic correlation (r g ) 0.69) and major depressive disorder (r g 0.48). BD-I has stronger coheritability with schizophrenia compared with BD-II, which is more strongly genetically correlated with major depressive disorder (r g 0.66). 16 Lower coheritability was observed with attention deficit hyperactivity disorder (r g 0.21), anorexia nervosa (0.20), and autism spectrum disorder (r g 0.21). 16 These correlations provide evidence for shared genetic risk factors between bipolar disorder and other major psychiatric syndromes, a pattern also corroborated by recent nationwide registry based family studies. 12 14 Nevertheless, despite their potential usefulness, polygenic risk scores must currently be interpreted with caution given their lack of populational representation and lingering concerns of residual confounds such as gene-environment correlations. 21

Insights from neuroimaging

Similarly to the early genetic studies, small initial studies had limited replication, leading to the formation of large worldwide consortiums such as ENIGMA (enhancing neuroimaging genetics through meta-analysis) which led to substantially larger sample sizes and improved reproducibility. In its volumetric analyses of subcortical structures from MRI (magnetic resonance imaging) of patients with bipolar disorder, the ENIGMA consortium found modest decreases in the volume of the thalamus (Cohen’s d −0.15), the hippocampus (−0.23), and the amygdala (−0.11), with an increased volume seen only in the lateral ventricles (+0.26). 22 Meta-analyses of cortical regions similarly found small reductions in cortical thickness broadly across the parietal, temporal, and frontal cortices (Cohen’s d −0.11 to −0.29) but no changes in cortical surface area. 23 In more recent meta-analyses of white matter tracts using diffuse tension imaging, widespread but modest decreases in white matter integrity were found throughout the brain in bipolar disorder, most notably in the corpus callosum and bilateral cinguli (Cohen’s d −0.39 to −0.46). 24 While these findings are likely to be highly replicable, they do not, as yet, have clinical application. This is because they reflect differences at a group level rather than an individual level, 25 and because many of these patterns are also seen across other psychiatric disorders 26 and could be either shared risk factors or the effects of confounding factors such as medical comorbidities, medications, co-occurring substance misuse, or the consequences (rather than causes) of living with mental illness. 27 Efforts to collate and meta-analyze large samples utilizing longitudinal designs 28 task based, resting state functional MRI measurents, 29 as well as other measures of molecular imaging (magnetic resonance spectroscopy and positron emission tomography) are ongoing but not as yet synthesized in large scale meta-analyses.

Environmental risk factors

Because of the difficulty in measuring and studying the relevant and often common environmental risk factors for a complex illness like bipolar disorder, there has been less research on how environmental risk factors could cause or modify bipolar disorder. Evidence for intrauterine risk factors is mixed and less compelling than such evidence in disorders like schizophrenia. 30 Preliminary evidence suggests that prominent seasonal changes in solar radiation, potentially through its effects on circadian rhythm, can be associated with an earlier onset of bipolar disorder 31 and a higher likelihood of experiencing a depressive episode at onset. 31 However, the major focus of environmental studies in bipolar disorder has been on traumatic and stressful life events in early childhood 32 and in adulthood. 33 The effects of such adverse events are complex, but on a broad level have been associated with earlier onset of bipolar disorder, a worse illness course, greater prevalence of psychotic symptoms, 34 substance misuse and psychiatric comorbidities, and a higher risk of suicide attempts. 32 35 Perhaps uniquely in bipolar disorder, evidence also indicates that positive life events associated with goal attainment can also increase the risk of developing elevated states. 36

Comorbidity

Bipolar disorder rarely manifests in isolation, with comorbidity rates indicating elevated lifetime risk of several co-occurring symptoms and comorbid disorders, particularly anxiety, attentional disorders, substance misuse disorders, and personality disorders. 37 38 The causes of such comorbidity can be varied and complex: they could reflect a mixed presentation artifactually separated by current diagnostic criteria; they might also reflect independent illnesses; or they might represent the downstream effects of one disorder increasing the risk of developing another disorder. 39 Anxiety disorders tend to occur before the frank onset of manic or hypomanic symptoms, suggesting that they could in part reflect prodromal symptoms that manifest early in the lifespan. 37 Similarly, subthreshold and syndromic symptoms of attention deficit/hyperactivity disorder are also observed across the lifespan of people with bipolar disorder, but particularly in early onset bipolar disorder. 40 On the other hand, alcohol and substance misuse disorders occur more evenly before and after the onset of bipolar disorder, consistent with a more bidirectional causal association. 41

The association between bipolar disorder and comorbid personality disorders is similarly complex. Milder manifestations of persistent mood instability (cyclothymia) or low mood (dysthymia) have previously been considered to be temperamental variants of bipolar disorder, 42 but are now classified as related but separate disorders. In people with persistent emotional dysregulation, making the diagnosis of bipolar disorder can be particularly challenging, 43 since the boundaries between longstanding mood instability and phasic changes in mood state can be difficult to distinguish. While symptom overlap can lead to artificially inflated prevalence rates of personality disorders in bipolar disorder, 44 the elevated rates of most personality disorders in bipolar disorder, particularly those related to emotional instability, are likely reflective of an important clinical phenomenon that is understudied, particularly with regard to treatment implications. 45 In general, people with comorbidities tend to have greater symptom burden and functional impairment and have lower response rates to treatment. 46 47 Data on approaches to treat specific comorbid disorders in bipolar disorder are limited, 48 49 and clinicians are often left to rely on their clinical judgment. The most parsimonious approach is to treat primary illness as fully as possible before considering additional treatment options for remaining comorbid symptoms. For certain comorbidities, such as anxiety symptoms and disorders of attention, first line pharmacological treatment—namely, antidepressants and stimulants, should be used with caution, since they might increase the long term risks of mood switching or overall mood instability. 50 51

Like other major mental illnesses, bipolar disorder is also associated with an increased prevalence of common medical disorders such as obesity, hyperlipidemia, coronary artery disease, chronic obstructive pulmonary disease, and thyroid dysfunction. 52 These have been attributed to increase risk factors such as physical inactivity, poor nutrition, smoking, and increased use of addictive substances, 53 but some could also be consequences of specific treatments, such as the atypical antipsychotics and mood stabilizers. 54 Along with poor access to care, this medical burden likely accounts for much of the increased standardized mortality (approximately 2.6 times higher) in people with bipolar disorder, 55 highlighting the need to utilize treatments with better long term side effect profiles, and the need for better integration with medical care.

Precursors and prodromes: who develops bipolar disorder?

While more widespread screening and better accessibility to mental health providers should in principle shorten the time to diagnosis and treatment, early manifestation of symptoms in those who ultimately go on to be diagnosed with bipolar disorder is generally non-specific. 56 In particular, high risk offspring studies of adolescents with a parent with bipolar disorder have found symptoms of anxiety and attentional/disruptive disorders to be frequent in early adolescence, followed by higher rates of depression and sleep disturbance in later teenage years. 56 57 Subthreshold symptoms of mania, such as prolonged increases in energy, elated mood, racing thoughts, and mood lability are also more commonly found in children with prodromal symptoms (meta-analytic prevalence estimates ranging from 30-50%). 58 59 Still, when considered individually, none of these symptoms or disorders are sensitive or specific enough to accurately identify individuals who will transition to bipolar disorder. Ongoing approaches to consider these clinical factors together to improve accuracy have a promising but modest ability to identify people who will develop bipolar disorder, 60 emphasizing the need for further studies before implementation.

Screening for bipolar disorder

Manic episodes can vary from easily identifiable prototypical presentations to milder or less typical symptoms that can be challenging to diagnose. Ideally, a full diagnostic evaluation with access to close informants is performed on patients presenting to clinical care; however, evaluations can be hurried in routine clinical care, and the ability to recall previous episodes might be limited. In this context, the use of screening scales can be a helpful addition to clinical care, although screening scales must be regarded as an impetus for a confirmatory clinical interview rather than a diagnostic instrument by themselves. The two most widely used and openly available screening scales are the mood disorders questionnaire (based on the DSM-IV criteria for hypomania) 61 and the hypomania check list (HCL-32), 62 that represent a broader overview of symptoms proposed to be part of a broader bipolar spectrum.

Racial/ethnic disparities

Although community surveys using structured or semi-structured diagnostic instruments, have provided little evidence for variation across ethnic groups, 63 64 observational studies based on clinical diagnoses in healthcare settings have found a disproportionately higher rate of diagnosis of schizophrenia relative to bipolar disorder in black people. 65 Consistent with similar disparities seen across medicine, these differences in clinical diagnoses are likely influenced by a complex mix of varying clinical presentations, differing rates of comorbid conditions, poorer access to care, greater social and economic burden, as well as the potential effect of subtle biases of healthcare professionals. 65 While further research is necessary to identify driving factors responsible for diagnostic disparities, clinicians should be wary of making a rudimentary diagnosis in patients from marginalized backgrounds, ensuring comprehensive data gathering and a careful diagnostic formulation that incorporates shared decision making between patient and provider.

Bipolar disorder is a recurrent illness, but its longitudinal course is heterogeneous and difficult to predict. 46 66 The few available long term studies of BD-I and BD-II have found a consistent average rate of recurrence of 0.40 mood episodes per year in historical studies 67 and 0.44 mood episodes per year in more recent studies. 68 The median time to relapse is estimated to be 1.44 years, with higher relapse rates seen in BD-I (0.81 years) than in BD-II (1.63 years) and no differences observed with respect to age or sex. 1 2 In addition to focusing on episodes, an important development in research and clinical care of bipolar disorder has been the recognition of the burden of subsyndromal symptoms. Although milder in severity, these symptoms can be long lasting, functionally impairing, and can themselves be a risk factor for episode relapse. 69 Recent cohort studies have also found that a substantial proportion of patients with bipolar disorder (20-30%) continue to have poor outcomes even after receiving guideline based care. 46 70 Risk factors that contribute to this poor outcome include transdiagnostic indicators of adversity such as substance misuse, low educational attainment, socioeconomic hardship, and comorbid disorders. As expected, those with more severe past illness activity, including those with rapid cycling, were also more likely to remain symptomatically and psychosocially impaired. 46 71 72

The primary focus of treating bipolar disorder has been to manage the manic, mixed, or depressive episodes that present to clinical care and to subsequently prevent recurrence of future episodes. Owing to the relapse remitting nature of the illness, randomized controlled trials are essential to determine treatment efficacy, as the observation of clinical improvement could just represent the ebbs and flows of the natural history of the illness. In the United States, the FDA (Food and Drug Administration) requires at least two large scale placebo controlled trials (phase 3) to show significant evidence of efficacy before approving a treatment. Phase 3 studies of bipolar disorder are generally separated into short term studies of mania (3-4 weeks), short term studies for bipolar depression (4-6 weeks), and longer term maintenance studies to evaluate prophylactic activity against future mood episodes (usually lasting one year). Although the most rigorous evaluation of phase 3 studies would be to require two broadly representative and independent randomized controlled trials, the FDA permits consideration of so called enriched design trials that follow participants after an initial response and tolerability has been shown to an investigational drug. Because of this initial selection, such trials can be biased against comparator agents, and could be less generalizable to patients seen in clinical practice.

A summary of the agents approved by the FDA for treatment of bipolar disorder is in table 1 , which references the key clinical trials demonstrating efficacy. Figure 1 and supplementary table 1 are a comparison of treatments for mania, depression, and maintenance. Effect sizes reflect the odds ratios or relative risks of obtaining response (defined as ≥50% improvement from baseline) in cases versus controls and were extracted from meta-analyses of randomized controlled trials for bipolar depression 86 and maintenance, 94 as well as a network meta-analysis of randomized controlled trials in bipolar mania. 73 Effect sizes are likely to be comparable for each phase of treatment, but not across the different phases, since methodological differences exist between the three meta-analytic studies.

FDA approved medications for bipolar disorder

View inline

Summary of treatment response rates (defined as ≥50% improvement from baseline) of modern clinical trials for acute mania, acute bipolar depression, and long term recurrence. Meta-analytic estimates were extracted from recent meta-analyses or network meta-analyses of acute mania, 73 acute bipolar depression, 86 and bipolar maintenance studies 94

Download figure
Open in new tab
Download powerpoint

Acute treatment of mania

As mania is characterized by impaired judgment, individuals can be at risk for engaging in high risk, potentially dangerous behaviors that can have substantial personal, occupational, and financial consequences. Therefore, treatment of mania is often considered a psychiatric emergency and is, when possible, best performed in the safety of an inpatient unit. While the primary treatment for mania is pharmacological, diminished insight can impede patients' willingness to accept treatment, emphasizing the significance of a balanced therapeutic approach that incorporates shared decision making frameworks as much as possible to promote treatment adherence.

The three main classes of anti-manic treatments are lithium, mood stabilizing anticonvulsants (divalproate and carbamazepine), and antipsychotic medications. Almost all antipsychotics are effective in treating mania, with the more potent dopamine D2 receptor antagonists such as risperidone and haloperidol demonstrating slightly higher efficacy ( fig 1 ). 73 In the United States, the FDA has approved the use of all second generation antipsychotics for treating mania except for lurasidone and brexpriprazole. Compared with mood stabilizing medications, second generation antipsychotics have a faster onset of action, making them a first line treatment for more severe manic symptoms that require rapid treatment. 99 The choice of which specific second generation antipsychotic to use depends on a balance of efficacy, tolerability concerns, and cost considerations (see table 1 ). Notably, the FDA has placed a black box warning on all antipsychotics for increasing the risk of cerebral vascular accidents in the elderly. 100 While this was primarily focused on the use of antipsychotics in dementia, this likely class effect should be taken into account when considering the use of antipsychotics in the elderly.

Traditional mood stabilizers, such as lithium, divalproate, and carbamazepine are also effective in the treatment of active mania ( fig 1 ). Since lithium also has a robust prophylactic effect (see section on prevention of mood episodes below) it is often recommended as first line treatment and can be considered as monotherapy when rapid symptom reduction is not clinically indicated. On the other hand, other anticonvulsants such as lamotrigine, gabapentin, topiramate, and oxcarbazepine have not been found to be effective for the treatment of mania or mixed episodes. 101 Although the empirical evidence for polypharmacy is limited, 102 combination treatment in acute mania, usually consisting of a mood stabilizer and a second generation antipsychotic, is commonly used in clinical practice despite the higher burden of side effects. Following resolution of an acute mania, consideration should be given to transitioning to monotherapy with an agent with proven prophylactic activity.

Pharmacological approaches to bipolar depression

Depressed episodes are usually more common than mania or hypomania, 103 104 and often represent the primary reason for individuals with bipolar disorder to seek treatment. Nevertheless, because early antidepressant randomized controlled trials did not distinguish between unipolar and bipolar depressive episodes, it has only been in the past two decades that large scale randomized controlled trials have been conducted specifically for bipolar depression. As such trials are almost exclusively funded by pharmaceutical companies, they have focused on the second generation antipsychotics and newer anticonvulsants still under patent. These trials have shown moderate but robust effects for most recent second generation antipsychotics, five of which have received FDA approval for treating bipolar depression ( table 1 ). No head-to-head trials have been conducted among these agents, so the choice of medication depends on expected side effects and cost considerations. For example, quetiapine has robust antidepressant efficacy data but is associated with sedation, weight gain, and adverse cardiovascular outcomes. 105 Other recently approved medications such as lurasidone, cariprazine, and lumateperone have better side effect profiles but show more modest antidepressant activity. 106

Among the mood stabilizing anticonvulsants, lamotrigine has limited evidence for acute antidepressant activity, 107 possibly owing to the need for an 8 week titration to reach the full dose of 200 mg. However, as discussed below, lamotrigine can still be considered for mild to moderate acute symptoms owing to its generally tolerable side effect profile and proven effectiveness in preventing the recurrence of depressive episodes. Divalproate and carbamazepine have some evidence of being effective antidepressants in small studies, but as there has been no large scale confirmatory study, they should be considered second or third line options. 86 Lithium has been studied for the treatment of bipolar depression as a comparator to quetiapine and was not found to have a significant acute antidepressant effect. 88

Antidepressants

Owing to the limited options of FDA approved medications for bipolar depression and concerns of metabolic side effects from long term second generation antipsychotic use, clinicians often resort to the use of traditional antidepressants for the treatment of bipolar depression 108 despite the lack of FDA approval for such agents. Indeed, recent randomized clinical trials of antidepressants in bipolar depression have not shown an effect for paroxetine, 89 109 bupropion, 109 or agomelatine. 110 Beyond the question of efficacy, another concern regarding antidepressants in bipolar disorder is their potential to worsen the course of illness by either promoting mixed or manic symptoms or inducing more subtle degrees of mood instability and cycle acceleration. 111 However, the risk of switching to full mania while being treated with mood stabilizers appears to be modest, with a meta-analysis of randomized clinical trials and clinical cohort studies showing the rates of mood switching over an average follow-up of five months to be approximately 15.3% in people with bipolar disorder treated on antidepressants compared with 13.8% in those without antidepressant treatment. 111 The risk of switching appears to be higher in the first 1-2 years of treatment in people with BD-I, and in those treated with a tricyclic antidepressant 112 or the dual reuptake inhibitor venlafaxine. 113 Overall, while the available data have methodological limitations, most guidelines do not recommend the use of antidepressants in bipolar disorder, or recommend them only after agents with more robust evidence have been tried. That they remain so widely used despite the equivocal evidence base reflects the unmet need for treatment of depression, concerns about the long term side effects of second generation antipsychotics, and the challenges of changing longstanding prescribing patterns.

Pharmacological approaches to prevention of recurrent episodes

Following treatment of the acute depressive or manic syndrome, the major focus of treatment is to prevent future episodes and minimize interepisodic subsyndromal symptoms. Most often, the medication that has been helpful in controlling the acute episode can be continued for prevention, particularly if clinical trial evidence exists for a maintenance effect. To show efficacy for prevention, studies must be sufficiently long to allow the accumulation of future episodes to occur and be potentially prevented by a therapeutic intervention. However, few long term treatment studies exist and most have utilized enriched designs that likely favor the drug seeking regulatory approval. As shown in figure 1 , meta-analyses 94 show prophylactic effect for most (olanzapine, risperidone, quetiapine, aripiprazole, asenapine) but not all (lurasidone, paliperidone) recently approved second generation antipsychotics. The effect sizes are generally comparable with monotherapy (odds ratio 0.42, 95% confidence interval 0.34 to 0.5) or as adjunctive therapy (odds ratio 0.37, 95% confidence interval 0.25 to 0.55). 94 Recent studies of lithium, which have generally used it as a (non-enriched) comparator drug, show a comparable protective effect (odds ratio 0.46, 95% confidence interval 0.28 to 0.75). 94 Among the mood stabilizing anticonvulsant drugs, a prophylactic effect has also been found for both divalproate and lamotrigine ( fig 1 and supplementary table 1), although only the latter has been granted regulatory approval for maintenance treatment. While there are subtle differences in effect sizes in drugs approved for maintenance ( fig 1 and table 1 ), the overlapping confidence intervals and methodological differences between studies prevent a strict comparison of the effect measures.

Guidelines often recommend lithium as a first line agent given its consistent evidence of prophylaxis, even when tested as the disadvantaged comparator drug in enriched drug designs. Like other medications, lithium has a unique set of side effects and ultimately the decision about which drug to use among those which are efficacious should be a decision carefully weighed and shared between patient and provider. The decision might be re-evaluated after substantial experience with the medication or at different stages in the long term treatment of bipolar disorder (see table 1 ).

Psychotherapeutic approaches

The frequent presence of residual symptoms, often associated with psychosocial and occupational dysfunction, has led to renewed interest in psychotherapeutic and psychosocial approaches to bipolar disorder. Given the impairment of judgment seen in mania, psychotherapy has more of a supportive and educational role in the treatment of mania, whereas it can be more of a primary focus in the treatment of depressive states. On a broad level, psychotherapeutic approaches effective for acute depression, such as cognitive behavioral therapy, interpersonal therapy, behavioral activation, and mindfulness based strategies, can also be recommended for acute depressive states in individuals with bipolar disorder. 114 Evidence for more targeted psychotherapy trials for bipolar disorder is more limited, but meta-analyses have found evidence for decreased recurrence (odds ratio 0.56; 95% confidence interval 0.43 to 0.74) 115 and improvement of subthreshold interepisodic depressive and manic symptoms with cognitive behavioral therapy, family based therapy, interpersonal and social rhythm therapy, and psychoeducation. 115 Recent investigations have also focused on targeted forms of psychotherapy to improve cognition 116 117 118 as well as psychosocial and occupational functioning. 119 120 Although these studies show evidence of a moderate effect, they remain preliminary, methodologically diverse, and require replication on a larger scale. 121

The implementation of evidence based psychotherapy as a treatment faces several challenges, including clinical training, fidelity monitoring, and adequate reimbursement. Novel approaches, leveraging the greater tractability of digital tools 122 and allied healthcare workers, 123 are promising means of lessening the implementation gap; however, these approaches require validation and evidence of clinical utility similar to traditional methods.

Neurostimulation approaches

For individuals with bipolar disorder who cannot tolerate or do not respond well to standard pharmacotherapy or psychotherapeutic approaches, neurostimulation techniques such as repetitive transcranial magnetic stimulation or electric convulsive therapy should be considered as second or third line treatments. Electric convulsive therapy has shown response rates of approximately 60-80% in severe acute depressions 124 125 and 50-60% in cases with treatment resistant depression. 126 These response rates compare favorably with those of pharmacological treatment, which are likely to be closer to ~50% and ~30% in subjects with moderate to severe depression and treatment resistant depression, respectively. 127 Although the safety of electric convulsive therapy is well established, relatively few medical centers have it available, and its acceptability is limited by cognitive side effects, which are usually short term, but which can be more significant with longer courses and with bilateral electrode placement. 128 While there have been fewer studies of electric convulsive therapy for bipolar depression compared with major depressive disorder, it appears to be similarly effective and might show earlier response. 129 Anecdotal evidence also suggests electric convulsive therapy that is useful in refractory mania. 130

Compared with electric convulsive therapy, repetitive transcranial magnetic stimulation has no cognitive side effects and is generally well tolerated. Repetitive transcranial magnetic stimulation acts by generating a magnetic field to depolarize local neural tissue and induce excitatory or inhibitory effects depending on the frequency of stimulation. The most studied FDA approved form of repetitive transcranial magnetic stimulation applies high frequency (10 Hz) excitatory pulses to the left prefrontal cortex for 30-40 minutes a day for six weeks. 131 Like electric convulsive therapy, repetitive transcranial magnetic stimulation has been primarily studied in treatment resistant depression and has been found to have moderate effect, with about one third of patients having a significant treatment response compared with those treated with pharmacotherapy. 131 Recent innovations in transcranial magnetic stimulation have included the use of a novel, larger coil to stimulate a larger degree of the prefrontal cortex (deep transcranial magnetic stimulation), 132 and a shortened (three minutes), higher frequency intermittent means of stimulation known as theta burst stimulation that appears to be comparable to conventional (10 Hz) repetitive transcranial magnetic stimulation. 133 A preliminary trial has recently assessed a new accelerated protocol of theta burst stimulation marked by 10 sessions a day for five days. It found that theta burst stimulation had a greater effect on people with treatment resistant depression compared with treatment as usual, although larger studies are needed to confirm these findings. 134

Conventional repetitive transcranial magnetic stimulation (10 Hz) studies in bipolar disorder have been limited by small sample sizes but have generally shown similar effects compared with major depressive disorder. 135 However, a proof of concept study of single session theta burst stimulation did not show efficacy in bipolar depression, 136 reiterating the need for specific trials for bipolar depression. Given the lack of such trials in bipolar disorder, repetitive transcranial magnetic stimulation should be considered a potentially promising but as yet unproven treatment for bipolar depression.

The other major form of neurostimulation studied in both unipolar and bipolar depression is transcranial direct current stimulation, an easily implemented method of delivering a low amplitude electrical current to the prefrontal area of the brain that could lead to local changes in neuronal excitability. 137 Like repetitive transcranial magnetic stimulation, transcranial direct current stimulation is well tolerated and has been mostly studied in unipolar depression, but has not yet generated sufficient evidence to be approved by a regulatory agency. 138 Small studies have been performed in bipolar depression, but the results have been mixed and require further research before use in clinical settings. 137 138 139 Finally, the evidence for more invasive neurostimulation studies such as vagal nerve stimulation and deep brain stimulation remains extremely limited and is currently insufficient for clinical use. 140 141

Treatment resistance in bipolar disorder

As in major depressive disorder, the use of term treatment resistance in bipolar disorder is controversial since differentiating whether persistent symptoms are caused by low treatment adherence, poor tolerability, the presence of comorbid disorders, or are the result of true treatment resistance, is an essential but often challenging clinical task. Treatment resistance should only be considered after two or three trials of evidence based monotherapy, adjunctive therapy, or both. 142 In difficult-to-treat mania, two or more medications from different mechanistic classes are typically used, with electric convulsive therapy 143 and clozapine 144 being considered if more conventional anti-manic treatments fail. In bipolar depression, it is common to combine antidepressants with anti-manic agents, despite limited evidence for efficacy. 145 Adjunctive therapies such as bright light therapy, 146 the dopamine D2/3 receptor agonist pramipexole, 147 and ketamine 148 149 have shown promising results in small open label trials that require further study.

Treatment considerations to reduce suicide in bipolar disorder

The risk of completed suicide is high across the subtypes of bipolar disorder, with estimated rates of 10-15% across the lifespan. 150 151 152 Lifetime rates of suicide attempts are much higher, with almost half of all individuals with bipolar disorder reporting at least one attempt. 153 Across a population and, often within individuals, the causes of suicide attempts and completed suicides are likely to be multifactorial, 154 affected by various risk factors, such as symptomatic illness, environmental stressors, comorbidities (particularly substance misuse), trait impulsivity, interpersonal conflict, loneliness, or socioeconomic distress. 155 156 Risk is highest in depressive and dysphoric/mixed episodes 157 158 and is particularly high in the transitional period following an acute admission to hospital. 159 Among the available treatments, lithium has potential antisuicidal properties. 160 However, since suicide is a rare event, with very few to zero suicides within a typical clinical trial, moderate evidence for this effect emerges only in the setting of meta-analyses of clinical trials. 160 Several observational studies have shown lower mortality in patients on lithium treatment, 161 but such associations might not be causal, since lithium is potentially fatal in overdose and is often avoided by clinicians in patients at high risk of suicide.

The challenge of studying scarce events has led most studies to focus on the reduction of the more common phenomena of suicidal ideation and behavior as a proxy for actual suicides. A recent such multisite study of the Veterans Affairs medical system included a mixture of unipolar and bipolar disorder and was stopped prematurely for futility, indicating no overall effect of moderate dose lithium. 162 Appropriate limitations of this study have been noted, 163 164 including difficulties in recruitment, few patients with bipolar disorder (rather than major depressive disorder), low levels of compliance with lithium therapy, high rates of comorbidity, and a follow-up of only one year. Nevertheless, while the body of evidence suggests that lithium has a modest antisuicidal effect, its degree of protection and utility in complex patients with comorbidities and multiple risk factors remain matters for further study. Treatment of specific suicidal risk in patients with bipolar disorder must therefore also incorporate broader interventions based on the individual’s specific risk factors. 165 Such an approach would include societal interventions like means restriction 166 and a number of empirically tested suicide focused psychotherapy treatments. 167 168 Unfortunately, the availability of appropriate training, expertise, and care models for such treatments remains limited, even in higher income countries. 169

More scalable solutions, such as the deployment of shortened interventions via digital means could help to overcome this implementation gap; however, the effectiveness of such approaches cannot be assumed and requires empirical testing. For example, a recent large scale randomized controlled trial of an abbreviated online dialectical behavioral therapy skills training program was paradoxically associated with slightly increased risk of self-harm. 170

Treatment consideration in BD-II and bipolar spectrum conditions

Because people with BD-II primarily experience depressive symptoms and appear less likely to switch mood states compared with individuals with BD-I, 50 171 there has been a greater acceptance of the use of antidepressants in BD-II depression, including as monotherapy. 172 However, caution should be exercised when considering the use of antidepressants without a mood stabilizer in patients with BD-II who might also experience high rates of mood instability and rapid cycling. Such individuals can instead respond better to newer second generation antipsychotic agents such as quetiapine 173 and lumateperone, 93 which are supported by post hoc analyses of these more recent clinical trials with more BD-II patients. In addition, despite the absence of randomized controlled trials, open label studies have suggested that lithium and other mood stabilizers can have similar efficacy in BD-II, especially in the case of lamotrigine. 174

Psychotherapeutic approaches such as psychoeducation, cognitive behavioral therapy, and interpersonal and social rhythm therapy have been found to be helpful 115 and can be considered as the primary form of treatment for BD-II in some patients, although in most clinical scenarios BD-II is likely to occur in conjunction with psychopharmacology. While it can be tempting to consider BD-II a milder variant of BD-I, high rates of comorbid disorders, rapid cycling, and adverse consequences such as suicide attempts 175 176 highlight the need for clinical caution and the provision of multimodal treatment, focusing on mood improvement, emotional regulation, and better psychosocial functioning.

Precision medicine: can it be applied to improve the care of bipolar disorder?

The recent focus on precision medicine approaches to psychiatric disorders seeks to identify clinically relevant heterogeneity and identify characteristics at the level of the individual or subgroup that can be leveraged to identify and target more efficacious treatments. 1 177 178

The utility of such an approach was originally shown in oncology, where a subset of tumors had gene expression or DNA mutation signatures that could predict response to treatments specifically designed to target the aberrant molecular pathway. 179 While much of the emphasis of precision medicine has been on the eventual identification of biomarkers utilizing high throughput approaches (genetics and other “omics” based measurements), the concept of precision medicine is arguably much broader, encompassing improvements in measurement, potentially through the deployment of digital tools, as well as better conceptualization of contextual, cultural, and socioeconomic mechanisms associated with psychopathology. 180 181 Ultimately, the goal of precision psychiatry is to identify and target driving mechanisms, be they molecular, physiological, or psychosocial in nature. As such, precision psychiatry seeks what researchers and clinicians have often sought: to identify clinically relevant heterogeneity to improve prediction of outcomes and increase the likelihood of therapeutic success. The novelty being not so much the goals of the overarching approach, but the increasing availability of large samples, novel digital tools, analytical advances, and an increasing armamentarium of biological measurements that can be deployed at scale. 177

Although not unique to bipolar disorder, several clinical decision points along the life course of bipolar disorder would benefit from a precision medicine approach. For example, making an early diagnosis is often not possible based on clinical symptoms alone, since such symptoms are usually non-specific. A precision medicine approach could also be particularly relevant in helping to identify subsets of patients for whom the use of antidepressants could be beneficial or harmful. Admittedly, precision medicine approaches to bipolar disorder are still in their infancy, and larger, clinically relevant, longitudinal, and reliable phenotypes are needed to provide the infrastructure for precision medicine approaches. Such data remain challenging to obtain at scale, leading to renewed efforts to utilize the extant clinical infrastructure and electronic medical records to help emulate traditional longitudinal analyses. Electronic medical records can help provide such data, but challenges such as missingness, limited quality control, and potential biases in care 182 need to be resolved with carefully considered analytical designs. 183

Emerging treatments

Two novel atypical antipsychotics, amilsupride and bifeprunox, are currently being tested in phase 3 trials ( NCT05169710 and NCT00134459 ) and could gain approval for bipolar depression in the near future if these pivotal trials show a significant antidepressant effect. These drugs could offer advantages such as greater antidepressant effects, fewer side effects, and better long term tolerability, but these assumptions must be tested empirically. Other near term possibilities include novel rapid antidepressant treatments, such as (es)ketamine that putatively targets the glutamatergic system, and has been recently approved for treatment resistant depression, but which have not yet been tested in phase 3 studies in bipolar depression. Small studies have shown comparable effects of intravenous ketamine, 149 184 in bipolar depression with no short term evidence of increased mood switching or mood instability. Larger phase 2 studies ( NCT05004896 ) are being conducted which will need to be followed by larger phase 3 studies. Other therapies targeting the glutamatergic system have generally failed phase 3 trials in treatment resistant depression, making them unlikely to be tested in bipolar depression. One exception could be the combination of dextromethorphan and its pharmacokinetic (CYP2D6) inhibitor bupropion, which was recently approved for treatment resistant depression but has yet to be tested in bipolar depression. Similarly, the novel GABAergic compound zuranolone is currently being evaluated by the FDA for the treatment of major depressive disorder and could also be subsequently studied in bipolar depression.

Unfortunately, given the general efficacy for most patients of available treatments, few scientific and financial incentives exist to perform large scale studies of novel treatment in mania. Encouraging results have been seen in small studies of mania with the selective estrogen receptor modulator 185 tamoxifen and its active metabolite endoxifen, both of which are hypothesized to inhibit protein kinase C, a potential mechanistic target of lithium treatment. These studies remain small, however, and anti-estrogenic side effects have potentially dulled interest in performing larger studies.

Finally, several compounds targeting alternative pathophysiological mechanisms implicated in bipolar disorder have been trialed in phase 2 academic studies. The most studied has been N -acetylcysteine, a putative mitochondrial modulator, which initially showed promising results only to be followed by null findings in larger more recent studies. 186 Similarly, although small initial studies of anti-inflammatory agents provided impetus for further study, subsequent phase 2 studies of the non-steroidal agent celecoxib, 187 the anti-inflammatory antibiotic minocycline, 187 and the antibody infliximab (a tumor necrosis factor antagonist) 188 have not shown efficacy for bipolar depression. Secondary analyses have suggested that specific anti-inflammatory agents might be effective only for a subset of patients, such as those with elevated markers of inflammation or a history of childhood adversity 189 ; however, such hypotheses must be confirmed in adequately powered independent studies.

Several international guidelines for the treatment of bipolar disorder have been published in the past decade, 102 190 191 192 providing a list of recommended treatments with efficacy in at least one large randomized controlled trial. Since effect sizes tend to be moderate and broadly comparable across classes, all guidelines allow for significant choice among first line agents, acknowledging that clinical characteristics, such as history of response or tolerability, severity of symptoms, presence of mixed features, or rapid cycling can sometimes over-ride guideline recommendations. For acute mania requiring rapid treatment, all guidelines prioritize the use of second generation antipsychotics such as aripiprazole, quetiapine, risperidone, asenapine, and cariprazine. 102 192 193 Combination treatment is considered based on symptom severity, tolerability, and patient choice, with most guidelines recommending lithium or divalproate along with a second generation antipsychotic for mania with psychosis, severe agitation, or prominent mixed symptoms. While effective, haloperidol is usually considered a second choice option owing to its propensity to cause extrapyramidal symptoms. 102 192 193 Uniformly, all guidelines agree on the need to taper antidepressants in manic or mixed episodes.

For maintenance treatment, guidelines are generally consistent in recommending lithium if tolerated and without relative contraindications, such as baseline renal disease. 194 The second most recommended maintenance treatment is quetiapine, followed by aripiprazole for patients with prominent manic episodes and lamotrigine for patients with predominant depressive episodes. 194 Most guidelines recommend considering prophylactic properties when initially choosing treatment for acute manic episodes, although others suggests that acute maintenance treatments can be cross tapered with maintenance medications after several months of full reponse. 193

For bipolar depression, recent guidelines recommend specific second generation antipsychotics such as quetiapine, lurasidone, and cariprazine 102 192 193 For more moderate symptoms, consideration is given to first using lamotrigine and lithium. Guidelines remain cautious about the use of antidepressants (selective serotonin reuptake inhibitors, venlafaxine, or bupropion) in patients with BP-I, restricting them to second or third line treatments and always in the context of an anti-manic agent. However, for patients with BP-II and no rapid cycling, several guidelines allow for the use of carefully monitored antidepressant monotherapy.

Bipolar disorder is a highly recognizable syndrome with many effective treatment options, including the longstanding gold standard therapy lithium. However, a significant proportion of patients do not respond well to current treatments, leading to negative consequences, poor quality of life, and potentially shortened lifespan. Several novel treatments are being developed but limited knowledge of the biology of bipolar disorder remains a major challenge for novel drug discovery. Hope remains that the insights of genetics, neuroimaging, and other investigative modalities could soon be able to inform the development of rational treatments aimed to mitigate the underlying pathophysiology associated with bipolar disorder. At the same time, however, efforts are needed to bridge the implementation gap and provide truly innovative and integrative care for patients with bipolar disorder. 195 Owing to the complexity of bipolar disorder, few patients can be said to be receiving optimized care across the various domains of mental health that are affected in those with bipolar disorder. Fortunately, the need for improvement is now well documented, 196 and concerted efforts at the scale necessary to be truly innovative and integrative are now on the horizon.

Questions for future research

Among adolescents and young adults who manifest common mental disorders such as anxiety or depressive or attentional disorders, who will be at high risk for developing bipolar disorder?

Can we predict the outcomes for patients following a first manic or hypomanic episode? This will help to inform who will require lifelong treatment and who can be tapered off medications after sustained recovery.

Are there reliable clinical features and biomarkers that can sufficiently predict response to specific medications or classes of medication?

What are the long term consequences of lifelong treatments with the major classes of medications used in bipolar disorder? Can we predict and prevent medical morbidity caused by medications?

Can we understand in a mechanistic manner the pathophysiological processes that lead to abnormal mood states in bipolar disorder?

Series explanation: State of the Art Reviews are commissioned on the basis of their relevance to academics and specialists in the US and internationally. For this reason they are written predominantly by US authors

Contributors: FSG performed the planning, conduct, and reporting of the work described in the article. FSG accepts full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish.

Competing interests: I have read and understood the BMJ policy on declaration of interests and declare no conflicts of interest.

Patient involvement: FSG discussed of the manuscript, its main points, and potential missing points with three patients in his practice who have lived with longstanding bipolar disorder. These additional viewpoints were incorporated during the drafting of the manuscript.

Provenance and peer review: Commissioned; externally peer reviewed.

↵ . Falret’s discovery: the origin of the concept of bipolar affective illness. Translated by M. J. Sedler and Eric C. Dessain. Am J Psychiatry 1983;140:1127-33. doi: 10.1176/ajp.140.9.1127 OpenUrl CrossRef PubMed Web of Science
↵ Kraepelin E. Manic-depressive Insanity and Paranoia. Translated by R. Mary Barclay from the Eighth German. Edition of the ‘Textbook of Psychiatry.’ 1921.
Merikangas KR ,
Akiskal HS ,
Koukopoulos A ,
Jongsma HE ,
Kirkbride JB ,
Rowland TA ,
Kessler RC ,
Kazdin AE ,
Aguilar-Gaxiola S ,
WHO World Mental Health Survey collaborators
Bergen SE ,
Kuja-Halkola R ,
Larsson H ,
Lichtenstein P
Smoller JW ,
Lichtenstein P ,
Sjölander A ,
Mullins N ,
Forstner AJ ,
O’Connell KS ,
Palmer DS ,
Howrigan DP ,
Chapman SB ,
Pirooznia M ,
Murray GK ,
McGrath JJ ,
Hickie IB ,
↵ Mostafavi H, Harpak A, Agarwal I, Conley D, Pritchard JK, Przeworski M. Variable prediction accuracy of polygenic scores within an ancestry group. Loos R, Eisen MB, O’Reilly P, eds. eLife 2020;9:e48376. doi: 10.7554/eLife.48376 OpenUrl CrossRef PubMed
Westlye LT ,
van Erp TGM ,
Costa Rica/Colombia Consortium for Genetic Investigation of Bipolar Endophenotypes
Pauling M ,
ENIGMA Bipolar Disorder Working Group
Schnack HG ,
Ching CRK ,
ENIGMA Bipolar Disorders Working Group
Goltermann J ,
Hermesdorf M ,
Dannlowski U
Gurholt TP ,
Suckling J ,
Lennox BR ,
Bullmore ET
Marangoni C ,
Hernandez M ,
Achtyes ED ,
Agnew-Blais J ,
Gilman SE ,
Upthegrove R ,
↵ Etain B, Aas M. Childhood Maltreatment in Bipolar Disorders. In: Young AH, Juruena MF, eds. Bipolar Disorder: From Neuroscience to Treatment . Vol 48. Current Topics in Behavioral Neurosciences. Springer International Publishing; 2020:277-301. doi: 10.1007/7854_2020_149
Johnson SL ,
Weinberg BZS
Stinson FS ,
Costello CG
Klein & Riso LP DN
Sandstrom A ,
Perroud N ,
de Jonge P ,
Bunting B ,
Nierenberg AA
Hantouche E ,
Vannucchi G
Zimmerman M ,
Ruggero CJ ,
Chelminski I ,
Leverich GS ,
McElroy S ,
Mignogna KM ,
Balling C ,
Dalrymple K
Kappelmann N ,
Stokes PRA ,
Jokinen T ,
Baldessarini RJ ,
Faedda GL ,
Offidani E ,
Viktorin A ,
Launders N ,
Osborn DPJ ,
Roshanaei-Moghaddam B ,
De Hert M ,
Detraux J ,
van Winkel R ,
Lomholt LH ,
Andersen DV ,
Sejrsgaard-Jacobsen C ,
Skjelstad DV ,
Gregersen M ,
Søndergaard A ,
Brandt JM ,
Van Meter AR ,
Youngstrom EA ,
Taylor RH ,
Ulrichsen A ,
Strawbridge R
Hafeman DM ,
Merranko J ,
Hirschfeld RM ,
Williams JB ,
Spitzer RL ,
Adolfsson R ,
Benazzi F ,
Regier DA ,
Johnson KR ,
Akinhanmi MO ,
Biernacka JM ,
Strakowski SM ,
Goldberg JF ,
Schettler PJ ,
Coryell W ,
Scheftner W ,
Endicott J ,
Zarate CA Jr . ,
Matsuda Y ,
Fountoulakis KN ,
Zarate CA Jr .
Bowden CL ,
Brugger AM ,
The Depakote Mania Study Group
Calabrese JR ,
Depakote ER Mania Study Group
Weisler RH ,
Kalali AH ,
Ketter TA ,
SPD417 Study Group
Keck PE Jr . ,
Cutler AJ ,
Caffey EM Jr . ,
Grossman F ,
Eerdekens M ,
Jacobs TG ,
Grundy SL ,
The Olanzipine HGGW Study Group
Versiani M ,
Ziprasidone in Mania Study Group
Sanchez R ,
Aripiprazole Study Group
McIntyre RS ,
Panagides J
Calabrese J ,
McElroy SL ,
EMBOLDEN I (Trial 001) Investigators
EMBOLDEN II (Trial D1447C00134) Investigators
Lamictal 606 Study Group
Lamictal 605 Study Group
Keramatian K ,
Chakrabarty T ,
Nestsiarovich A ,
Gaudiot CES ,
Neijber A ,
Hellqvist A ,
Paulsson B ,
Trial 144 Study Investigators
Schwartz JH ,
Szegedi A ,
Cipriani A ,
Salanti G ,
Dorsey ER ,
Rabbani A ,
Gallagher SA ,
Alexander GC
Cerqueira RO ,
Yatham LN ,
Kennedy SH ,
Parikh SV ,
Højlund M ,
Andersen K ,
Correll CU ,
Ostacher M ,
Schlueter M ,
Geddes JR ,
Mojtabai R ,
Nierenberg AA ,
Goodwin GM ,
Agomelatine Study Group
Vázquez G ,
Baldessarini RJ
Altshuler LL ,
Cuijpers P ,
Miklowitz DJ ,
Efthimiou O ,
Furukawa TA ,
Strawbridge R ,
Tsapekos D ,
Hodsoll J ,
Vinberg M ,
Kessing LV ,
Forman JL ,
Miskowiak KW
Lewandowski KE ,
Sperry SH ,
Torrent C ,
Bonnin C del M ,
Martínez-Arán A ,
Bonnín CM ,
Tamura JK ,
Carvalho IP ,
Leanna LMW ,
Karyotaki E ,
Individual Patient Data Meta-Analyses for Depression (IPDMA-DE) Collaboration
Vipulananthan V ,
Hurlemann R ,
UK ECT Review Group
Haskett RF ,
Mulsant B ,
Trivedi MH ,
Wisniewski SR ,
Espinoza RT ,
Vazquez GH ,
McClintock SM ,
Carpenter LL ,
National Network of Depression Centers rTMS Task Group ,
American Psychiatric Association Council on Research Task Force on Novel Biomarkers and Treatments
Levkovitz Y ,
Isserles M ,
Padberg F ,
Blumberger DM ,
Vila-Rodriguez F ,
Thorpe KE ,
Williams NR ,
Sudheimer KD ,
Bentzley BS ,
Konstantinou G ,
Toscano E ,
Husain MM ,
McDonald WM ,
International Consortium of Research in tDCS (ICRT)
Sampaio-Junior B ,
Tortella G ,
Borrione L ,
McAllister-Williams RH ,
Gippert SM ,
Switala C ,
Bewernick BH ,
Hidalgo-Mazzei D ,
Mariani MG ,
Fagiolini A ,
Swartz HA ,
Benedetti F ,
Barbini B ,
Fulgosi MC ,
Burdick KE ,
Diazgranados N ,
Ibrahim L ,
Brutsche NE ,
Sinclair J ,
Gerber-Werder R ,
Miller JN ,
Vázquez GH ,
Franklin JC ,
Ribeiro JD ,
Turecki G ,
Gunnell D ,
Hansson C ,
Pålsson E ,
Runeson B ,
Pallaskorpi S ,
Suominen K ,
Ketokivi M ,
Hadzi-Pavlovic D ,
Stanton C ,
Lewitzka U ,
Severus E ,
Müller-Oerlinghausen B ,
Rogers MP ,
Li+ plus Investigators
Manchia M ,
Michel CA ,
Auerbach RP
Altavini CS ,
Asciutti APR ,
Solis ACO ,
Casañas I Comabella C ,
Riblet NBV ,
Young-Xu Y ,
Shortreed SM ,
Rossom RC ,
Amsterdam JD ,
Brunswick DJ
Gustafsson U ,
Marangell LB ,
Bernstein IH ,
Karanti A ,
Kardell M ,
Collins FS ,
Armstrong K ,
Concato J ,
Singer BH ,
Ziegelstein RC
↵ Holmes JH, Beinlich J, Boland MR, et al. Why Is the Electronic Health Record So Challenging for Research and Clinical Care? Methods Inf Med 2021;60(1-02):32-48. doi: 10.1055/s-0041-1731784
García Rodríguez LA ,
Cantero OF ,
Martinotti G ,
Dell’Osso B ,
Di Lorenzo G ,
REAL-ESK Study Group
Palacios J ,
DelBello MP ,
Husain MI ,
Chaudhry IB ,
Subramaniapillai M ,
Jones BDM ,
Daskalakis ZJ ,
Carvalho AF ,
↵ Goodwin GM, Haddad PM, Ferrier IN, et al. Evidence-based guidelines for treating bipolar disorder: revised third edition Recommendations from the British Association for Psychopharmacology. J Psychopharmacol 2016;30:495-553. doi: 10.1177/0269881116636545 OpenUrl CrossRef PubMed
Verdolini N ,
Del Matto L ,
Regeer EJ ,
Hoogendoorn AW ,
Harris MG ,
WHO World Mental Health Survey Collaborators

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

View all journals
Explore content
About the journal
Publish with us
Sign up for alerts
Open access
Published: 25 March 2022

Characterizing mood disorders in the AFFECT study: a large, longitudinal, and phenotypically rich genetic cohort in the US

Maria Dalby ORCID: orcid.org/0000-0001-6018-3468 1 , 2 ,
Morana Vitezic 1 ,
Niels Plath 1 ,
Lene Hammer-Helmich 1 ,
Yunxuan Jiang 3 ,
Chao Tian 3 ,
Devika Dhamija 3 ,
Catherine H. Wilson 3 ,
David Hinds ORCID: orcid.org/0000-0002-4911-803X 3 ,
23andMe Research Team ,
Patrick F. Sullivan 2 , 4 ,
Joshua W. Buckholtz 5 , 6 na1 &
Jordan W. Smoller 7 , 8 na1

Translational Psychiatry volume 12 , Article number: 121 ( 2022 ) Cite this article

4594 Accesses

5 Citations

10 Altmetric

Metrics details

Bipolar disorder
Clinical genetics

There has recently been marked progress in identifying genetic risk factors for major depression (MD) and bipolar disorder (BD); however, few systematic efforts have been made to elucidate heterogeneity that exists within and across these diagnostic taxa. The Affective disorders, Environment, and Cognitive Trait (AFFECT) study presents an opportunity to identify and associate the structure of cognition and symptom-level domains across the mood disorder spectrum in a prospective study from a diverse US population.

Participants were recruited from the 23andMe, Inc research participant database and through social media; self-reported diagnosis of MD or BD by a medical professional and medication status data were used to enrich for mood-disorder cases. Remote assessments were used to acquire an extensive range of phenotypes, including mood state, transdiagnostic symptom severity, task-based measures of cognition, environmental exposures, personality traits. In this paper we describe the study design, and the demographic and clinical characteristics of the cohort. In addition we report genetic ancestry, SNP heritability, and genetic correlations with other large cohorts of mood disorders.

A total of 48,467 participants were enrolled: 14,768 with MD, 9864 with BD, and 23,835 controls. Upon enrollment, 47% of participants with MD and 27% with BD indicated being in an active mood episode. Cases reported early ages of onset (mean = 13.2 and 14.3 years for MD and BD, respectively), and high levels of recurrence (78.6% and 84.9% with >5 episodes), psychotherapy, and psychotropic medication use. SNP heritability on the liability scale for the ascertained MD participants (0.19–0.21) was consistent with the high level of disease severity in this cohort, while BD heritability estimates (0.16–0.22) were comparable to reports in other large scale genomic studies of mood disorders. Genetic correlations between the AFFECT cohort and other large-scale cohorts were high for MD but not for BD. By incorporating transdiagnostic symptom assessments, repeated measures, and genomic data, the AFFECT study represents a unique resource for dissecting the structure of mood disorders across multiple levels of analysis. In addition, the fully remote nature of the study provides valuable insights for future virtual and decentralized clinical trials within mood disorders.

The genetic basis of major depressive disorder

Genome-wide significant risk loci for mood disorders in the Old Order Amish founder population

Dimensional and transdiagnostic phenotypes in psychiatric genome-wide association studies

Introduction.

Mood disorders have a high lifetime prevalence in the general population and represent the leading cause of disability worldwide [ 1 , 2 ]. Moreover, mood disorders cause marked impairment in social and occupational functioning, resulting in a high burden for the individual and to society [ 3 , 4 ]. Twin and family studies show moderate-high heritability for these syndromes, indicating a prominent role for genetic variation in conferring susceptibility [ 5 , 6 , 7 , 8 ]. MD has a lifetime prevalence of 15% [ 9 ] and twin-heritability of 30–40% [ 5 , 10 ]. In contrast, BD has a lifetime prevalence of 2.4% [ 11 ] and twin-heritability ~70% [ 6 , 12 ]. Genomic analyses have shown that mood disorders are highly polygenic with likely thousands of small-effect loci contributing to susceptibility [ 13 , 14 ]. Significant progress has been made in identifying common genetic risk variants associated with MD and BD, most recently from the Psychiatric Genomics Consortium (PGC). The PGC Bipolar working group identified 40 independent BD loci in a sample of 40,000 BD cases [ 15 ], and the PGC MD working group identified 102 independent loci associated with MD from more than 246,000 cases [ 16 ]. Despite these successes, a major obstacle in psychiatric genetics is our inability to map these signals to the symptom patterns, cognitive deficits and maladaptive decision-making that characterize mood disorders.

One critical open question is how genetic risk affects human cognition to predispose the development of mood disorder symptoms and related behaviors. With up to 90% of patients with major depression (MD) or bipolar disorder (BD) exhibiting impairment in multiple domains of cognition, this represents an important diagnostic and symptomatic feature in mood disorders and a key determinant of functional recovery [ 17 , 18 ]. Much of the morbidity and mortality in mood disorders is due to behavioral factors, such as substance abuse, aggression, self-harm, and risky sexual behavior [ 19 , 20 , 21 ]. These behaviors, in turn, are thought to result from deficits in cognitive processes related to cost-benefit decision-making, reinforcement learning, social cognition, and executive function [ 22 ]. Many groups have reported phenotypic associations between mood disorders and some of these cognitive processes [ 23 , 24 ]. However, such studies are typically small in size, limited in scope, and genetically uninformative, limiting insight into the underlying causes of cognitive dysfunction and maladaptive behavior in mood disorders.

It is widely recognized that the DSM-based nosology of psychiatric illness poorly captures two important features of mental disorders: the high degree of comorbidity between diagnostic taxa, and the profound symptom-level heterogeneity that exists within a given diagnostic taxon [ 22 , 25 , 26 , 27 ]. These features suggest the existence of latent transdiagnostic symptom clusters in mood disorders and are consistent with evidence for shared genetic liability between otherwise categorically distinct psychiatric disorders [ 28 , 29 , 30 , 31 , 32 , 33 ]. To date, we know little about how much of the shared variance among mood disorder symptoms, cognitive function and maladaptive behavior is due to genetic factors. Likewise, GWAS estimate the proportion of variance in liability attributable to common variants genome-wide (SNP-heritability) to be ~9% for MD and 18% for BD [ 15 ], which are fractions of the pedigree-based estimated heritability. This accords with the significant role of non-genetic factors in mood disorder risk. In particular, a number of environmental risk factors have been identified for mood disorders, including poverty and traumatic life events, particularly in early life. Understanding the mechanisms through which such environmental influences interact with genetic susceptibility is key to elucidating the risk architecture of mood disorders. However, existent GWAS data sets are unable to answer these and other important open questions because of practical constraints that preclude the collection of an appropriately rich set of phenotypic data at scale.

To bridge these gaps, we leveraged technological advances in web-based participant recruitment, diagnostic assessment and cognitive testing to create the AFFECT study. The AFFECT study employed a longitudinal case-control design in nearly 50,000 US-based participants with BD, MD, and controls. Study participants were recruited from the 23andMe, Inc research participant database and through social media, representing a diverse sample that includes patients who may be underrepresented in clinical samples. A key innovation of this study is the depth of phenotypic data acquired, made practical through the use of online data collection. The study collected 9 months of remote phenotypic assessments, including recent and lifetime diagnostic evaluations, transdiagnostic symptom assessments, longitudinal measures of symptom state severity, and detailed medication profiling. Further, we obtained detailed information about environmental risk and protective factors, personality traits, and real-world maladaptive behaviors related to mood disorder morbidity and mortality. Finally, we measured task-based cognitive performance using an online testing battery. In this paper, we present the AFFECT study design, enrollment process, data collection, and characterize the MD, BD, and control groups based on baseline descriptive characteristics and genetic analysis. Lastly, we assess cohort representativeness and disorder severity and demonstrate the similarity of the case groups to those from prior large-scale genomic studies.

Cohort design

This genetic, case-control study was designed to enroll three cohorts: 15,000 participants with MD, 10,000 participants with BD, and 25,000 controls with no lifetime MD or BD. Of these, 1533 participants (3.06%) withdrew consent or failed to return the spit kit or intake survey before the study termination date and were excluded.

Participant eligibility criteria were: age between 18 and 50 years upon enrolment; residence in the United States; access to a desktop or laptop computer; and no reported diagnosis of Parkinson’s disease, essential tremor, schizophrenia, or Alzheimer’s disease. Enrollment required that the participants self-reported having been diagnosed with MD or BD by a medical professional and prescribed medication to treat such a disorder. Enrollment into the control cohort required that participants reported no lifetime diagnosis of BD, MD, generalized anxiety disorder, or post-traumatic stress disorder (PTSD) as well as never having been prescribed an antidepressant, mood stabilizer, or antipsychotic medication. All study participants had to provide informed consent and a saliva sample for SNP array genotyping, and be willing to complete the online study sessions over the course of 9 months.

The study was conducted between August 2017 and September 2019 and online recruitment of participants, genotyping, and survey data collection were performed by 23andMe. Figure 1 illustrates the enrollment flow and study setup . Participants were recruited through two channels: all controls and approximately one-fifth ( n = 4997) of all case participants were recruited from 23andMe’s existing customer database through email or logged-in website invitation. All other case participants ( n = 9635) were recruited through social media such as Facebook and enrolled as new 23andMe customers. Study participants who met the eligibility criteria received compensation depending on if they were existing or new 23andMe customers. Existing customers, who had purchased a 23andMe kit prior to joining the study, received a $20 Amazon gift card. New customers received the 23andMe ® Health + Ancestry Service, including a DNA test kit, at no cost.

The procedural steps were: Informed consent, apply for enrollment and meet study inclusion and no exclusion criteria, return a saliva kit for genotyping (except for excisting costumers who purchased and returned a 23andMe kit prior to joining the study), and answer the baseline questionnaire. In the 9 months after enrolment, participants were asked to complete monthly surveys and cognitive tests.

Study assessments

The study content was designed by the AFFECT investigators and administered by 23andMe. The self-reported survey and test battery (Table 1 ) was initiated at session 1 with an extensive background survey covering: demographics (i.e., age, gender, race, ethnicity), socioeconomic information (i.e., marital status, current employment, education, parental education, income), clinical details about the given disorder (cases only; e.g., age of onset, current and past episode characterization), family psychiatric history, the Self-rated Diagnostic and Statistical Manual of Mental Disorders (DSM-5) Level 1 Cross-Cutting Symptom Measure, and adverse childhood experiences (scale and scoring details in Supplementary Materials).

The mood and medication survey was also given at session 1 and repeated in sessions 2–5, 7, and 9. This survey included: medication history (session 1), changes in medications (all follow-up surveys), life events/life style (e.g., alcohol use and sleep patterns), Altman Self-rating of Mania (ASRM) scale, and Patient-Reported Outcomes Measurement Information System (PROMIS)-Depression scale (scale and scoring details in supplementary materials). The study battery further included standardized behavioral tasks assessing risk, impulsivity and psychopathic traits and five cognitive tools designed to assess different domains of functioning. The cognitive tests were either given at one or two time points as noted in Table 1 .

SNP genotyping

We evaluated common variant genetic contributions to risk for MD and BD using SNP array data. DNA extraction and genotyping were performed on saliva samples by the National Genetics Institute, a CLIA-licensed clinical laboratory and a subsidiary of the Laboratory Corporation of America. Samples were genotyped, phased and imputed by 23andMe standardized pipeline, as described in detail in Supplementary Methods. Roughly 9.22 million high-quality genotyped and imputed SNPs on autosomal and X chromosome were tested.

For each GWAS, we restrict participants to a set of individuals who had a specified ancestry determined through an analysis of local ancestry estimation [ 34 ] and a maximal set of unrelated individuals was chosen for each GWAS analysis using a segmental identity-by-descent (IBD) estimation algorithm [ 35 ].

Genome-wide associations

GWAS was performed on MD versus controls, BD versus controls, mood disorder (MD, BD) versus controls and MD versus BD using a logistic regression model: case/control ~ age + sex + top 5 Principal Components (PCs) + genotyping platforms + genotype . GWAS was first performed separately on individuals of European, African American, East Asian, Latino ancestry, and combined by fixed-effect meta-analysis using METAL [ 36 ]. GWAS results were adjusted for the genomic control inflation factors, which can be found under each Manhattan plot in Supplementary Figures . Note that the study enrollment channel (existing/enrolled customers) was embedded in the genotype platform term, where around 80% of existing customers were genotyped on 23andMe’s genotype platform v4, while all newly enrolled participants were genotyped on platform v5 (Supplementary Table 1 ). Across all results, we removed SNPs that had an available sample size of less than 20% of the total GWAS sample size; where logistic regression results that did not converge due to complete separation, identified by absolute value of effect size or standard error greater than 10 on the log-odds scale; or that had MAF < 0.1%.

SNP-heritability and genetic correlations

We used LD score regression (LDSC) [ 37 ] v1.0.1 to estimate SNP-heritability ( $h_{SNP}^2$ ) from GWAS summary statistics for European ancestry MD and BD including variants with r 2 > 0.8 and minor allele frequency ≥0.01. Estimates of $h_{SNP}^2$ on the liability scale depend on the assumed lifetime prevalence of each disorder in the population (K). We report $h_{SNP}^2$ with K = [0.001–0.3] for MD and K = [0.001–0.03] for BD.

Genetic correlations (r g ) to external summary statistics were also performed using LDSC [ 37 ]. External data included; the PGC MDD meta-analysis samples PGC-MDD1 (2013) [ 10 ], PGC-MDD2 excluding the 23andMe sample (2018) [ 38 ], and PGC-MDD3 excluding the 23andMe sample (2019) [ 16 ]; the 23andMe discovery sample of MDD (herein Hyde et. al, 2016; where 5.0% of MD cases and 4.3% of controls from the AFFECT study were also included in Hyde et al.) [ 39 ]; the two most recent PGC BD meta-analysis samples PGC-BD2 (2019) [ 40 ] and PGC-BD3 (including the PGC-BD3 type I and type II sub-cohorts (2020) [ 15 ]); the most recent PGC SCZ meta-analysis samples PGC-SCZ2 (2014) [ 41 ] and PGC-SCZ3 (2020) [ 41 , 42 ]. Data was obtained from https://www.med.unc.edu/pgc/download-results/ and through the 23andMe data-access portal.

Statistical analyses

Sample comparisons were conducted using R (v3.5.2). Descriptive statistics were performed on the total participation pool and on subgroups: the three cohorts of MD, BD, and controls; within subtype of BD diagnosis (BD1 vs. BD2) and, within each cohort subgroups based on enrollment strategy (i.e., participants drawn from the 23andMe database and participants enrolled through social media for this study). For categorical variables, the number and percentage were reported for each value. For quantitative variables, the mean, median, standard deviation, and ranges were reported. Differences in demographic and clinical covariates were compared using regression models (continuous or categorical variables) and Fisher’s exact tests for categorical variables.

Cohort characteristics

A total of 48,467 participants were included in these analyses: 14,768 reported that they had been diagnosed and treated for MD, 9864 had been diagnosed and treated for BD, and 23,835 were controls with no lifetime history of MD or BD (Fig. 1 ). The BD cohort contained 3070 (31.2%) BD subtype I (BD1), 5053 (51.3%) BD subtype II (BD2), and 1718 (17.5%) did not specify the latest type of BD diagnosis received (BD unspecified-type). Among all participants, 72% were female and the mean age was 32.3 years (range 18–52 years). Most participants were of European ancestry (71.9%) followed by Latino (14.2%), African American (3.8%), and East Asian (3.6%) ancestry (Table 2 ).

Participant completion rates ranged from 28 to 100% (mean 42.6%) per session and were lower for cognitive assessments than for surveys (Supplementary Table 2 ). Study retention (i.e., number of assessments completed) was highest for MD cases (mean 50.2%, SD 30.6) followed by BD cases (mean 45.2%, SD 30.6), lowest for controls (mean 38.2%, SD 28.8), and higher in females (mean 45.0%, SD 30.1) than males (mean 38.7%, SD 29.9) (Supplementary Fig. 1 ). Study retention was positively correlated with educational level and age, and negatively correlated with reported adverse childhood experience score, BMI, ASRM score, and the DSM-5 cross-cutting domains of substance use, anxiety, depression, anger, suicidal ideation, and sleep problems (Supplementary Fig. 2 ).

Marital status, highest education achieved, and current socioeconomic status were reported at baseline and followed-up by a brief status assessment during each longitudinal assessment. Overall, socioeconomic status was significantly lower for cases, especially BD participants (Supplementary Table 3 ). In particular, we found that 19.5% and 26.9% of MD and BD participants, respectively, were currently not in paid employment as compared to only 7.3% of the control cohort. We observed an ascertainment effect in which case participants drawn from the 23andMe database (existing consumers) showed higher yearly salary and educational level than those enrolled through social media (multivariate analysis, P < 1.0 × 10 −16 ). After adjusting for enrollment method, however, significant socioeconomic differences remained between cases and controls (multivariate analysis, P < 1.0 × 10 −16 , Supplementary Table 3 ).

Disease history

Figure 2A and Supplementary Table 4 summarizes the clinical features of MD and BD cases and highlights that both MD and BD presented with high disease severity. Most participants reported symptom onset in adolescence (MD; mean 13.2 (SD 5.1), BD; mean 14.3 (SD 5.2)) while formal psychiatric diagnosis was not typically received until early adulthood (MD; mean 19.5 (SD 6.6), BD; mean 23.2 (SD 7.6)), consistent with prior studies [ 43 , 44 , 45 ]. The course of illness differed between the disorders; BD cases tended to report short but recurrent episodes: 52.2% of the participants had experienced >10 episodes and 80.0% reported a typical episode duration of <3 months. In contrast, MD cases had fewer episodes of longer duration: 59.7% had experienced ≤10 episodes, 47.7% reported a typical episode duration of 3–6 months or longer, and 10.0% reported episode duration ≥1 year (Fig. 2A , Supplementary Fig. 3A, B ).

A Summary of key clinical features in cases reporting a diagnosis of MDD, BD subtype 1 (BD1) and BD subtype 2 (BD2), as per latest diagnosis recieved. Mood disorders cases in this 23andMe sub-cohort show high burdens of illness. Any medication class refers to medication received over the last 5 years and during the study. Percentage of those who answered one or several treatment questions in the medication survey. B Transdiagnostic symptoms. Radar plot of median score pr. symptom domain within controls (purple), MD (blue), and BD (orange) participants. Scores are based on the DSM-5 cross-cutting symptom measures, where max item score (ranging from 0 to 4) within each domain is reported and summarized.

As expected, psychotropic medication use was common, since this was an inclusion criterion: 23,202 (96.4%) of cases reported having taken medication for a mood disorder in the prior 5 years, 17,292 (70.2 %) were taking medication at baseline, and 7726 (31.4%) began or restarted a medication during the study. MD and BD participants (respectively) reported use of the following treatments in the prior 5 years and/or at present: antidepressants 13,803 (95.4%) and 8508 (88.1%); mood stabilizers 4875 (33,8%) and 8394 (86.9%); antipsychotics 6133 (42.4%) and 7972 (82.5%); and electroconvulsive therapy 107 (0.9%) and 144 (2.0%). Most cases had received cognitive or behavioral psychotherapy in the past 5 years (MD 9770, 67.7%, BD 7235, 79.3%; Supplementary Fig. 3C, D ), most commonly 1–2 times a week. BD1 cases had the highest rates of symptom-related hospitalization (63.6%), although the rates were also high for the other mood disorder diagnoses (BD2, 46.1%; MD, 29.0%).

Symptom state

Nearly half of the MD cases ( N = 6971, 47.5%) and about a quarter of the BD cases (2729, 27.8%) reported that they were experiencing an episode at baseline (Table 3 , Supplementary Fig. 4 ). Most BD participants reported their current episode as depressive (1694, 62.7%). A current manic episode was reported in 219 (7.13%) BD1 participants, 106 (6.17%) unspecified-type BD participants, and a current hypomanic episode was reported across BD type: BD1 140 (16.0%), BD2 407 (29.4%), and 55 (11.9%) unspecified-type. We further observed that participants enrolled through social media exhibite greater disease burden (Supplementary Fig. 5 , Supplementary Table 4 ) and were more likely to be in active mood episode compared to participants drawn from the 23andMe research participant database (41.0% versus 34.0%).

We defined probable depressive episodes using the Level 1 DSM-5 cross-cutting measure—depressive domain (score ≥ 2) and the PROMIS-depression scale ( T -score ≥60), which identified 71.7% of all cases being in a depressive episode at baseline. Additionally, we defined a probable manic/hypomanic episode from the Level 1 DSM-5 cross-cutting measure—manic domain (score ≥ 2) and the ASRM scale (score > 5), identifying 28.5% of BD participants being in an episode at baseline (Table 3 ). When comparing the symptom scale-based episodes with the self-identified episodes at baseline, we found reasonable correspondence for depressive episodes (κ = 0.43 and κ = 0.36 respectively for MD and BD) and a more modest correspondence for manic or hypomanic episodes (κ = 0.22).

Symptom-level comorbidities

The DSM-5 self-rated cross-cutting symptom measure assesses 13 transdiagnostic symptom domains of relevance across psychiatric diagnosis [ 34 ] (scoring details given in Supplementary Materials). We found that both MD and BD participants exhibited a wide-range of transdiagnostic symptoms (median number of positively screened symptom domains = 9), a clear distinction to the control cohort (median number of positively screened symptom domains = 2) (Fig. 2B , Supplementary Table 5 ). The most common symptom domains in cases were depression, mania, somatic symptoms (i.e. aches and pains), and anxiety. Furthermore, sleep problems and substance use symptoms provided the strongest differentiation of BD from MD (multivariable analysis, coefficient 0.43 (95% CI ±0.03) P < 2.2e −16 , coefficient 0.41 (95% CI ±0.05) P < 2.2e −16 , respectively).

Regarding non-psychiatric conditions, MD and BD participants reported higher rates of comorbidities compared to controls. This was particularly evident for inflammatory and neurological disorders (multivariable analysis, OR ≥ 3.03 P < 0.001, Supplementary Table 3 ).

Family psychiatric history

Family history prevalence of anxiety disorder, MD, BD, or PTSD in first-degree relatives is shown in Supplementary Table 6 . Rates were significantly higher for all disorders among cases (78.4 %) compared to controls (Fisher’s exact OR = 4.2 (95% CI ±0.1), P < 2.2e −16 ), particularly for the same disorder and within BD subtypes (Fisher’s exact OR (95% CI) MD = 6.6 (0.6), BD1 = 3.1 (±0.4), OR = 5.0 (±1.0), P < 2.2e −16 , Supplementary Fig. 6 ). The prevalence of mental disorders in first-degree relatives of controls (33.0%) was comparable to rates reported in population-based samples [ 46 ].

Environmental influences

Reported adverse childhood experiences (ACE) were assessed across multiple domains (i.e., psychological and sexual abuse, neglect, and household dysfunction) [ 47 , 48 ]. Childhood adversity was common, with 63.9% of participants reporting at least one ACE. The total ACE score was significantly higher in cases than controls, with almost twice as many ACEs reported (case mean = 3.96, control mean = 2.00, P < 1.0 × 10 −16 ). Moreover, BD cases reported more ACEs than MD cases (Supplementary Table 7 ). Within ACE domains, physical and emotional neglect showed the largest association with mood disorders (OR = 5.6, 95% CI ±0.4); again, these associations were considerably stronger in BD cases (OR = 6.54, 95% CI ±0.34).

SNP-heritability and genetic comparability

GWAS was conducted in European ancestry participants for mood disorder (MD + BD), each disorder separately, BD subtypes, and comparing MD versus BD. Furthermore, a trans-ethnic meta-analysis of European, Latino, African American and East Asian GWAS was conducted for MD and for BD. Variant-level analysis, which was not the focus of this paper, is provided in Supplementary Figs. 7 – 22 and sample sizes for each GWAS can be found in Supplementary Table 8 .

The SNP-heritability ( $h_{SNP}^2$ ) on the liability scale for European ancestry MD was 0.19 (SE 0.02) and 0.21 (SE 0.03) for a population prevalence of 0.10 and 0.15, respectively. These estimates are higher than those reported in previous self-reported or broadly ascertained MD cohorts [ 39 , 49 ]. The SNP-heritability for European ancestry BD was comparable to previous large cohorts [ 1 , 15 , 40 ] with $h_{SNP}^2$ estimates of 0.16 (SE 0.02) and 0.22 (SE 0.02) on the liability scale assuming population prevalence of 0.005 and 0.02, respectively (Fig. 3A ).

A Liability-scale SNP-heritability of AFFECT BD and MD as a function of population prevalence, ranging from 0.001 to 0.03 for BD and 0.001–0.3 for MD with r g estimates at every 0.001 step-wise increase. Dotted line represents s.e. B Estimated genetic correlations of European ancestry AFFECT BD and MD with PGC GWAS of MDD3 (excluding the 23andMe cohort), the 23andMe MD discovery cohort (Hyde et al, 2016), PGC-BD2, and of PGC-BD3, which is further divided into BD3 type I and type II. Correlations in AFFECT BD were performed with the full cohort (BD) and within BD type (BD1, BD2). All correlations were significant, circle size and values indicate r g . P -values, Z -scores and s.e are reported in Supplementary Table 9 .

To further compare the MD and BD cohorts to other mood disorder studies, we estimated genetic correlations (r g ) to the most recent and largest meta-analysis samples (Fig. 3B , Supplementary Table 9 ). We found that r g for AFFECT-MD was highest with PGC MD2 (0.85 (SE 0,06), P = 2.1 × 10 −40 ), followed by significant correlations to the other MD cohorts, then PGC BD2 type II. We found significant, but moderate, genetic correlation between and the PGC3 BP cohort (0.43 (SE 0.04), P = 5.3 × 10 −22 ). Of note, stronger genetic correlations were observed between the AFFECT-BD cohort and prior MD samples (0.61 (SE 0.1) – 0.78 (SE 0.08)), suggesting that the current BD cohort is genetically different than previously published BD cohorts that used more traditional clinical ascertainment (see Discussion). Genetic correlations of AFFECT-BD1 and BD2 cases to external data showed an increased positive correlation between BD1 and external BD cohorts (0.42 (SE 0.07) – 0.59 (SE 0.12)) and SCZ cohorts (0.30 (SE 0.07) – 0.33 (SE 0.06)), while the genetic correlations of BD2 was greater for external MD cohorts (0.46 (SE 0.1) – 0.71 (SE 0.06) and the PGC3 BP type II cohort (0.56 (SE 0.08), P = 1.2e−10).

The AFFECT study was initiated to advance our understanding of phenotypic and genetic heterogeneity in MD and BD and to clarify the role of shared genomic and environmental risk factors that may transcend their diagnostic boundaries. Several aspects of AFFECT are notable including the administration of task-based measures indexing multiple domains of cognition (e.g. executive, motivational, and social) that capture key facets of the Research Domain Criteria (RDoC) [ 50 ] framework; transdiagnostic symptom assays; the assessment of trait and environmental risk and resilience factors; and the repeated measures design enabling analysis of change in symptoms and multi-domain cognitive task performance. Here, we have presented baseline characterization of the cohort and summarized the clinical features of MD and BD cases.

The US-based study participants were ascertained from the general 23andMe participant database and from social media. Control participants did not self-report diagnosis of or treatment for mood disorders. Case participants self-reported a clinican-ascertained diagnosis of MDD or BD (I or II) and were currently using one or more prescribed medications to manage their symptoms. Additional study ascertainment criteria pertained to age (18–50 years old) and the absence of of Parkinsons disease, Alzheimers disease, essential tremor, or schizophrenia diagnosis. Demographic and socio-economic features of BD and MD cases in the AFFECT study were largely comparable to those reported in epidemiologic and clinical samples [ 51 , 52 , 53 ] with a substantial female predominance among cases. Consistent with prior research [ 54 , 55 ], reported adverse childhood experiences were relatively common and associated with significantly increased risk of mood disorder.

Prior studies have shown that selective participation represents a potential source of bias in both epidemiological and genetic association studies [ 56 , 57 ]. Consistent with this, several features of the cohort differ from those seen in many clinically ascertained mood disorder cohorts. For example, educational attainment and income levels among MD cases were somewhat higher than reported in population-based samples [ 52 ] as might be expected given the ascertainment through a direct-to-consumer genomics company. Interestingly, we observed some differences within the sample: lower socioeconomic status and greater illness severity were observed among those recruited through social media compared to participants drawn from the existing 23andMe consumer database. Although it might be expected that cases recruited through direct-to-consumer genomics and social media platforms would have less burden of illness compared with those ascertained clinically, this was not the case. In fact, most mood disorder cases in this study reported early-onset illness, recurrent episodes, positive family history, and treatment with medication and psychotherapy. Indeed, a history of psychiatric hospitalization among MD cases was higher (29%) than that reported in a representative sample of US adults (12%) [ 52 ]. Together, these suggest a high disease burden (significant impairment and dysfunction) in our cohort.

Overall, 71.7% of AFFECT participants reported symptoms of a current depressive episode at baseline, and 28.1% of BD cases reported current manic or hypomanic symptoms. This likely reflects the fact that BD2 was overrepresented in our BD cohort (51.3%) relative to population-based samples [ 11 , 53 ], but may also suggest that remote study participation is more likely for euthymic and depressive BD patients. We found that the agreement between self-reported and mood scale ratings for mania was limited. This underlines the limitations of self-reported assessments and symptom-based outcomes as discussed elsewhere [ 58 ].

Despite these considerations, we expect the AFFECT study to contribute importantly to understanding the genetic basis of mood disorders. The incorporation of transdiagnostic symptom and behavior measures, longitudinal symptom assessments, and task-based measures of neuro- and social cognition, make this a unique resource for genomic studies. In the initial GWAS of the AFFECT mood disorders, we identified several genome-wide significant loci; the strongest association was between MD and SNPs within NEGR1 , a gene encoding a synaptic adhesion protein that has been robustly associated with depression in prior studies [ 16 , 59 ]. Recent analyses have found that GWAS of MD samples characterized by “minimal phenotyping” (e.g. based on self-report of prior diagnosis and/or treatment for depression) show lower heritability and are enriched for less specific genetic effects on MD compared with samples diagnosed using strict syndromal criteria [ 60 ]. In this context, it is notable that the estimated liability scale h 2 SNP for AFFECT MD (0.19–0.21) is in the same range as “strictly-defined lifetime MDD” in that analysis and higher than what is seen in broadly-defined MD cohorts, including the previous 23andMe self-reported depression cohort [ 16 , 38 , 39 ]. As demonstrated in previous work [ 61 , 62 ], SNP heritability is a consequence of several known and unknown effects, including the exclusion of specific comorbidities, disease severity, and the use of controls from which other psychiatric disorders have been excluded [ 63 ].

Genetic correlation analyses indicate that AFFECT MD is highly correlated (r g = 0.71–0.85) with MD ascertained in studies included in the PGC. Unexpectedly, however, genetic correlations between AFFECT-BD and published PGC GWAS of BD were relatively modest (r g s = 0.38–0.43) while the genetic correlation between the AFFECT MD and AFFECT BD was approximating 1. Indeed, the pattern of genetic correlations seen with AFFECT-BD closely resembled those of AFFECT-MD and did not vary substantially by AFFECT-BD subtype 1 or 2. While recent genetic studies have shown that depression and bipolar depression have a large genetic overlap and many symptoms co-occur [ 17 ], we speculate that study exclusion of comorbid SCZ diagnosis and the fully remote ascertainment and follow-up strategy might have affected study participation, e.g deselected BD cases with psychotic features. Furthermore, the high genetic correlation within the AFFECT study sub-cohorts may have been affected by the use of fully shared controls that were screened for both MD and BD (i.e. “extreme” controls). Together, these results suggest a large genetic overlap with depression and high variability between different BD samples, further underlining the importance of understanding heterogeneity within and across diagnostic taxa.

The AFFECT study represents a unique cohort of remotely recruited individuals with MD and BD and controls. The availability of repeated measures over time as well as task-based cognitive domains will provide an important opportunity to examine the genomic basis of mood disorders and underlying traits. More in-depth analyses of these phenotypes and shared or unique contributions to BD and MD are forthcoming.

Data availability

The top 10,000 SNPs for each GWAS are provided in Supplementary Tables 10 – 15 . Participants provided informed consent and participated in the research online, under a protocol approved by the external AAHRPP-accredited IRB, Ethical & Independent Review Services (E&I Review). Participants were included in the analysis on the basis of consent status as checked at the time data analyses were initiated.

The full GWAS summary statistics for the 23andMe discovery data set will be made available through 23andMe to qualified researchers under an agreement with 23andMe that protects the privacy of the 23andMe participants. Please visit https://research.23andme.com/collaborate/#dataset-access/ for more information and to apply for access. Individual-level data are not publicly available due to participant confidentiality, and in accordance with the IRB-approved protocol under which the study was conducted. Researchers interested in the study’s individual-level data may apply to the 23andMe Research Innovation Collaborations program.

Sklar P, Ripke S, Scott LJ, Andreassen OA, Cichon S, Craddock N, et al. Large-scale genome-wide association analysis of bipolar disorder identifies a new susceptibility locus near ODZ4. Nat Genet. 2011;43:977–85. https://doi.org/10.1038/ng.943

Article CAS PubMed Central Google Scholar

World Health Organization. Depression and Other Common Mental Disorders—Global Health Estimates. World Health Organization; 2017. p24.

Romera I, Perez V, Menchón JM, Delgado-Cohen H, Polavieja P, Gilaberte I. Social and occupational functioning impairment in patients in partial versus complete remission of a major depressive disorder episode. A six-month prospective epidemiological study. Eur Psychiatry. 2010;25:58–65. https://doi.org/10.1016/j.eurpsy.2009.02.007 .

Article CAS PubMed Google Scholar

Rosa AR, Reinares M, Michalak EE, Bonnin CM, Sole B, Franco C. et al. Functional Impairment and Disability across Mood States in Bipolar Disorder. Value Heal.2010;13:984–8. https://doi.org/10.1111/j.1524-4733.2010.00768.x .

Article Google Scholar

Sullivan PF, Neale MC, Kendler KS. Genetic epidemiology of major depression: Review and meta-analysis. Am J Psychiatry. 2000;157:1552–62. https://doi.org/10.1176/appi.ajp.157.10.1552 .

McGuffin P, Rijsdijk F, Andrew M, Sham P, Katz R, Cardno A The heritability of bipolar affective disorder and the genetic relationship to unipolar depression. Arch Gen Psychiatry. 2003. https://doi.org/10.1001/archpsyc.60.5.497 .

Kendler KS, Gatz M, Gardner CO, Pedersen NL. A Swedish national twin study of lifetime major depression. Am J Psychiatry. 2006;163:109–14. https://doi.org/10.1176/appi.ajp.163.1.109 .

Article PubMed Google Scholar

Smoller JW, Finn CT. Family, Twin, and Adoption Studies of Bipolar Disorder. Am J Med Genet Semin Med Genet. 2003;123 C:48–58. https://doi.org/10.1002/ajmg.c.20013 .

Bromet E, Andrade LH, Hwang I, Sampson NA, Alonso J, de Girolamo G, et al. Cross-national epidemiology of DSM-IV major depressive episode. BMC Med. 2011. https://doi.org/10.1186/1741-7015-9-90 .

Sullivan PF, Daly M, Ripke S, Lewis CM, Wray NR, Hamilton SP. et al. A mega-analysis of genome-wide association studies for major depressive disorder. Mol Psychiatry. 2013;18:497–511. https://doi.org/10.1038/mp.2012.21 .

Merikangas KR, Jin R, He JP, Kessler RC, Lee S, Sampson NA, et al. Prevalence and correlates of bipolar spectrum disorder in the World Mental Health Survey Initiative. Arch Gen Psychiatry. 2011. https://doi.org/10.1001/archgenpsychiatry.2011.12 .

Edvardsen J, Torgersen S, Røysamb E, Lygren S, Skre I, Onstad S, et al. Heritability of bipolar spectrum disorders. Unity or heterogeneity? J Affect Disord. 2008. https://doi.org/10.1016/j.jad.2007.07.001 .

Sullivan PF, Daly MJ, Ripke S, Lewis CM, Lin DY, Wray NR. et al. A mega-Analysis of genome-wide association studies for major depressive disorder. Mol Psychiatry. 2013;18:497–511. https://doi.org/10.1038/mp.2012.21 .

Neale BM, Sklar P. Genetic analysis of schizophrenia and bipolar disorder reveals polygenicity but also suggests new directions for molecular interrogation. Curr Opin Neurobiol. 2015;30:131–8. https://doi.org/10.1016/j.conb.2014.12.001 .

Mullins N, Forstner AJ, O KS, Sloofman LG, Steinberg S, Trubetskoy V. et al. Genome-wide association study of over 40,000 bipolar disorder cases provides novel biological insights. medRxiv. 2020;17:202. https://doi.org/10.1101/2020.09.17.20187054 .

Howard DM, Adams MJ, Clarke TK, Hafferty JD, Gibson J, Shirali M. et al. Genome-wide meta-analysis of depression identifies 102 independent variants and highlights the importance of the prefrontal brain regions. Nat Neurosci. 2019;22:343–52. https://doi.org/10.1038/s41593-018-0326-7 .

Article CAS PubMed PubMed Central Google Scholar

Zuckerman H, Pan Z, Park C, Brietzke E, Musial N, Shariq AS, et al. Recognition and Treatment of Cognitive Dysfunction in Major Depressive Disorder. Front Psychiatry. 2018;9:1–11. https://doi.org/10.3389/fpsyt.2018.00655 .

Zubieta JK, Huguelet P, O’Neil RL, Giordani BJ. Cognitive function in euthymic Bipolar I Disorder. Psychiatry Res. 2001;102:9–20. https://doi.org/10.1016/S0165-1781(01)00242-6 .

Fagiolini A, Forgione R, Maccari M, Cuomo A, Morana B.Dell’Osso B, et al. Prevalence, chronicity, burden and borders of bipolar disorder. J Affect Disord. 2013;148:161–9. https://doi.org/10.1016/j.jad.2013.02.001 .

Hirschfeld RMA, Cass AR, Holt DCL, Carlson CA. Screening for bipolar disorder in patients treated for depression in a family medicine clinic. J Am Board Fam Pract. 2005;18:233–9. https://doi.org/10.3122/jabfm.18.4.233 .

Kleinman LS, Lowin A, Flood E, Gandhi G, Edgell E, Revicki DA. Costs of bipolar disorder. Pharmacoeconomics. 2003;21:601–22. https://doi.org/10.2165/00019053-200321090-00001 .

Buckholtz JW, Meyer-Lindenberg A. Psychopathology and the Human Connectome: Toward a Transdiagnostic Model of Risk For Mental Illness. Neuron. 2012;74:990–1004. https://doi.org/10.1016/j.neuron.2012.06.002 .

Goldberg JF, Chengappa KNR. Identifying and treating cognitive impairment in bipolar disorder. Bipolar Disord. 2009;11:123–37. https://doi.org/10.1111/j.1399-5618.2009.00716.x . SUPPL. 2.

Quraishi S, Frangou S. Neuropsychology of bipolar disorder: A review. J Affect Disord. 2002;72:209–26. https://doi.org/10.1016/S0165-0327(02)00091-5 .

Hyman SE. The Diagnosis of Mental Disorders: The Problem of Reification. Annu Rev Clin Psychol. 2010;6:155–79. https://doi.org/10.1146/annurev.clinpsy.3.022806.091532 .

Lichtenstein P, Yip BH, Björk C, Pawitan Y, Cannon TD, Sullivan PF, et al. Common genetic determinants of schizophrenia and bipolar disorder in Swedish families: a population-based study. Lancet. 2009. https://doi.org/10.1016/S0140-6736(09)60072-6 .

Hakulinen C, Musliner KL, Agerbo E. Bipolar disorder and depression in early adulthood and long‐term employment, income, and educational attainment: A nationwide cohort study of 2,390,127 individuals. Depress Anxiety. 2019:da.22956. https://doi.org/10.1002/da.22956 .

Anttila V, Bulik-Sullivan B, Finucane HK, Walters RK, Bras J, Duncan L, et al. Analysis of shared heritability in common disorders of the brain. Science. 2018;360. https://doi.org/10.1126/science.aap8757 .

Lee PH, Anttila V, Won H, Feng YCA, Rosenthal J, Zhu Z. et al. Genomic Relationships, Novel Loci, and Pleiotropic Mechanisms across Eight Psychiatric Disorders. Cell. 2019;179:1469–1482.e11. https://doi.org/10.1016/j.cell.2019.11.020 .

Article CAS Google Scholar

Coleman JRI, Gaspar HA, Bryois J, Breen G, Byrne EM, Forstner AJ, et al. The Genetics of the Mood Disorder Spectrum: Genome-wide Association Analyses of More Than 185,000 Cases and 439,000 Controls. Biol Psychiatry. 2019. https://doi.org/10.1016/j.biopsych.2019.10.015 .

Lee SH, Ripke S, Neale BM, Faraone SV, Purcell SM, Perlis RH. et al. Genetic relationship between five psychiatric disorders estimated from genome-wide SNPs. Nat Genet. 2013;45:984–94. https://doi.org/10.1038/ng.2711 .

Smoller JW, Andreassen OA, Edenberg HJ, Faraone SV, Glatt SJ, Kendler KS. Psychiatric Genetics and the Structure of Psychopathology. Mol Psychiatry. 2018:617–43. https://doi.org/10.1038/s41380-017-0010-4 .

Ruderfer DM, Fanous AH, Ripke S, McQuillin A, Amdur RL, Gejman PV. et al. Polygenic dissection of diagnosis and clinical dimensions of bipolar disorder and schizophrenia Cross-Disorder Working Group of the Psychiatric Genomics Consortium. Mol Psychiatry. 2014;19:1017–24. https://doi.org/10.1038/mp.2013.138 .

Durand EY, Do CB, Mountain JL, Macpherson JM. Ancestry Composition: A Novel, Efficient Pipeline for Ancestry Deconvolution. bioRxiv. 2014;010512. https://doi.org/10.1101/010512 .

Henn BM, Hon L, Macpherson JM, Eriksson N, Saxonov S, Pe’er I, et al. Cryptic distant relatives are common in both isolated and cosmopolitan genetic samples. PLoS One. 2012;7. https://doi.org/10.1371/journal.pone.0034267 .

Willer CJ, Li Y, Abecasis GR. METAL: Fast and efficient meta-analysis of genomewide association scans. Bioinformatics. 2010;26:2190–1. https://doi.org/10.1093/bioinformatics/btq340 .

Bulik-Sullivan B, Loh PR, Finucane HK, Ripke S, Yang J, Patterson N. et al. LD score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. 2015;47:291–5. https://doi.org/10.1038/ng.3211 .

Wray NR, Ripke S, Mattheisen M, Trzaskowski M, Byrne EM, Abdellaoui A. et al. Genome-wide association analyses identify 44 risk variants and refine the genetic architecture of major depression. Nat Genet. 2018;50:668–81. https://doi.org/10.1038/s41588-018-0090-3 .

Hyde CL, Nagle MW, Tian C, Chen X, Paciga SA, Wendland JR. et al. Identification of 15 genetic loci associated with risk of major depression in individuals of European descent. Nat Genet. 2016;48:1031–6. https://doi.org/10.1038/ng.3623 .

Stahl EA, Breen G, Forstner AJ, McQuillin A, Ripke S, Trubetskoy V. et al. Genome-wide association study identifies 30 loci associated with bipolar disorder. Nat Genet. 2019;51:793–803. https://doi.org/10.1038/s41588-019-0397-8 .

Ripke S, Neale BM, Corvin A, Walters JTR, Farh KH, Holmans PA. et al. Biological insights from 108 schizophrenia-associated genetic loci. Nature. 2014;511:421–7. https://doi.org/10.1038/nature13595 .

Schizophrenia Working Group of the Psychiatric Genomics Consortium., Ripke S, Walters JT, O’Donovan MC Mapping genomic loci prioritises genes and implicates synaptic biology in schizophrenia. medRxiv. 2020. https://doi.org/10.1101/2020.09.12.20192922 .

Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the national comorbidity survey replication. Arch Gen Psychiatry. 2005;62:593–602. https://doi.org/10.1001/archpsyc.62.6.593 .

Battle DE Diagnostic and Statistical Manual of Mental Disorders (DSM). CoDAS. https://doi.org/10.1007/978-3-642-28753-4_1094 .

Robert M A Hirschfeld, Lydia L, Lana A Vornik. Perceptions and Impact of Bipolar Disorder: How Far Have We Really Come? Results of the National Depressive and Manic-Depressive Association 2000 Survey of Individuals With Bipolar Disorder |J Clin Psychiatry. J Clin Psychiatry. Accessed 13 Oct 2020. https://www.psychiatrist.com/JCP/article/Pages/perceptions-impact-bipolar-disorder-far-really-results.aspx .

Steel Z, Marnane C, Iranpour C, Chey T, Jackson JW, Patel V. et al. The global prevalence of common mental disorders: A systematic review and meta-analysis 1980-2013. Int J Epidemiol. 2014;43:476–93. https://doi.org/10.1093/ije/dyu038 .

Article PubMed PubMed Central Google Scholar

Felitti VJ, Anda RF, Nordenberg D, Williamson DF, Spitz AM, Edwards V. et al. Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults: The adverse childhood experiences (ACE) study. Am J Prev Med. 1998;14:245–58. https://doi.org/10.1016/S0749-3797(98)00017-8 .

Dube SR, Felitti VJ, Dong M, Chapman DP, Giles WH, Anda RF. Childhood abuse, neglect, and household dysfunction and the risk of illicit drug use: The adverse childhood experiences study. Pediatrics. 2003;111:564–72. https://doi.org/10.1542/peds.111.3.564 .

Howard DM, Adams MJ, Shirali M, Clarke TK, Marioni RE, Davies G. et al. Genome-wide association study of depression phenotypes in UK Biobank identifies variants in excitatory synaptic pathways. Nat Commun. 2018;9:1–10. https://doi.org/10.1038/s41467-018-03819-3 .

Insel T, Cuthbert B, Garvey M, Heinssen R, Pine DS, Quinn K. et al. Research Domain Criteria (RDoC): Toward a new classification framework for research on mental disorders. Am J Psychiatry. 2010;167:748–51. https://doi.org/10.1176/appi.ajp.2010.09091379 .

Kessler RC, Bromet EJ. The Epidemiology of Depression Across Cultures. Annu Rev Public Health. 2013;34:119–38. https://doi.org/10.1146/annurev-publhealth-031912-114409 .

Hasin DS, Sarvet AL, Meyers JL, Saha TD, Ruan WJ, Stohl M. et al. Epidemiology of adult DSM-5 major depressive disorder and its specifiers in the United States. JAMA Psychiatry. 2018;75:336–46. https://doi.org/10.1001/jamapsychiatry.2017.4602 .

Rowland TA, Marwaha S. Epidemiology and risk factors for bipolar disorder. Ther Adv Psychopharmacol. 2018;8:251–69. https://doi.org/10.1177/2045125318769235 .

Nelson J, Klumparendt A, Doebler P, Ehring T. Childhood maltreatment and characteristics of adult depression: Meta-analysis. Br J Psychiatry. 2017;210:96–104. https://doi.org/10.1192/bjp.bp.115.180752 .

Gilman SE, Ni MY, Dunn EC, Breslau J, Mclaughlin KA, Smoller JW. et al. Contributions of the social environment to first-onset and recurrent mania. Mol Psychiatry. 2015;20:329–36. https://doi.org/10.1038/mp.2014.36 .

Batty GD, Gale CR, Kivimäki M, Deary IJ, Bell S. Comparison of risk factor associations in UK Biobank against representative, general population based studies with conventional response rates: prospective cohort study and individual participant meta-analysis. BMJ. 2020;368. https://doi.org/10.1136/bmj.m131 .

Taylor AE, Jones HJ, Sallis H, Euesden J, Stergiakouli E, Davies NM, et al. Exploring the association of genetic factors with participation in the Avon Longitudinal Study of Parents and Children. https://doi.org/10.1093/ije/dyy060 .

Davis KAS, Cullen B, Adams M, Brailean A, Breen G, Coleman JRI, et al. Indicators of mental disorders in UK Biobank—A comparison of approaches. Int J Methods Psychiatr Res. 2019;28. https://doi.org/10.1002/mpr.1796 .

Dall’Aglio L, Lewis CM, Pain O. Delineating the Genetic Component of Gene Expression in Major Depression. Biol Psychiatry. 2021;89:627–36. https://doi.org/10.1016/j.biopsych.2020.09.010 .

Cai N, Revez JA, Adams MJ, Andlauer TFM, Breen G, Byrne EM, et al. Minimal phenotyping yields GWAS hits of low specificity for major depression. bioRvix. 2018:1–34. https://doi.org/10.1101/440735 .

Schork A, Hougaard D, Nordentoft M, Mors O, Boerglum A, Mortensen PB, et al. Exploring contributors to variability in estimates of SNP-heritability and genetic correlations from the iPSYCH case-cohort and published meta-studies of major psychiatric disorders. bioRxiv. 2019:487116. https://doi.org/10.1101/487116 .

Wray NR, Lee SH, Kendler KS. Impact of diagnostic misclassification on estimation of genetic correlations using genome-wide genotypes. Eur J Hum Genet. 2012;20:668–74. https://doi.org/10.1038/ejhg.2011.257 .

Kendler KS, Chatzinakos C, Bacanu SA. The impact on estimations of genetic correlations by the use of super-normal, unscreened, and family-history screened controls in genome wide case–control studies. Genet Epidemiol. 2020;44:283–9. https://doi.org/10.1002/gepi.22281 .

Altman EG, Hedeker D, Peterson JL, Davis JM. The altman self-rating Mania scale. Biol Psychiatry. 1997;42:948–55. https://doi.org/10.1016/S0006-3223(96)00548-3 .

Pilkonis PA, Choi SW, Reise SP, Stover AM, Riley WT, Cella D. Item banks for measuring emotional distress from the patient-reported outcomes measurement information system (PROMIS®): Depression, anxiety, and anger. Assessment. 2011;18:263–83. https://doi.org/10.1177/1073191111411667 .

Schuster TL, Kessler RC, Aseltine RH. Supportive interactions, negative interactions, and depressed mood. Am J Community Psychol. 1990;18:423–38. https://doi.org/10.1007/BF00938116 .

Sadeh N, Baskin-Sommers A. Risky, Impulsive, and Self-Destructive Behavior Questionnaire (RISQ): A Validation Study. Assessment. 2017;24:1080–94. https://doi.org/10.1177/1073191116640356 .

Neumann CS, Pardini D. Factor structure and construct validity of the self-report psychopathy (SRP) scale and the youth psychopathic traits inventory (YPI) in young men. J Pers Disord. 2014;28:419–33. https://doi.org/10.1521/pedi_2012_26_063 .

Lezak MD. Neuropsychological Assessment. 3rd ed. Oxford University Press; 1995. p24.

Miedl SF, Peters J, Büchel C. Altered neural reward representations in pathological gamblers revealed by delay and probability discounting. Arch Gen Psychiatry. 2012;69:177–86. https://doi.org/10.1001/archgenpsychiatry.2011.1552 .

Peters J, Büchel C. The neural mechanisms of inter-temporal decision-making: Understanding variability. Trends Cogn Sci. 2011;15:227–39. https://doi.org/10.1016/j.tics.2011.03.002 .

McClure SM, Laibson DI, Loewenstein G, Cohen JD. Separate neural systems value immediate and delayed monetary rewards. Science. 2004;306:503–7. https://doi.org/10.1126/science.1100907 .

Kable JW, Glimcher PW. The neural correlates of subjective value during intertemporal choice. Nat Neurosci. 2007;10:1625–33. https://doi.org/10.1038/nn2007 .

Rosenberg M, Noonan S, DeGutis J, Esterman M. Sustaining visual attention in the face of distraction: A novel gradual-onset continuous performance task. Attention, Perception, Psychophys. 2013;75:426–39. https://doi.org/10.3758/s13414-012-0413-x .

McIntyre RS, Best MW, Bowie CR, Carmona NE, Cha DS, Lee Y, et al. The THINC-Integrated Tool (THINC-it) Screening Assessment for Cognitive Dysfunction: Validation in Patients With Major Depressive Disorder. J Clin Psychiatry. 2017;1–4. https://doi.org/10.4088/JCP.16m11329

Lam RW, Saragoussi D, Danchenko N, Rive B, Lamy FX, Brevig T. Psychometric Validation of Perceived Deficits Questionnaire – Depression (PDQ-D) in Patients with Major Depressive Disorder (MDD). Value Heal. 2013;16:A330. https://doi.org/10.1016/j.jval.2013.08.046 .

Lejuez CW, Richards JB, Read JP, Kahler CW, Ramsey SE, Stuart GL. et al. Evaluation of a behavioral measure of risk taking: The balloon analogue risk task (BART). J Exp Psychol Appl. 2002;8:75–84. https://doi.org/10.1037/1076-898X.8.2.75 .

Baron-Cohen S, Jolliffe T, Mortimore C, Robertson M. Another advanced test of theory of mind: Evidence from very high functioning adults with autism or Asperger syndrome. J Child Psychol Psychiatry Allied Discip. 1997;38:813–22. https://doi.org/10.1111/j.1469-7610.1997.tb01599.x .

Download references

Acknowledgements

We thank the project coordinator STF (23andMe Inc.), Lars-Peder Haahr (former emplyee at Lundebck A/S), all AFFECT-study scientists from Lundbeck A/S, Massachusetts General Hospital, and 23andMe Inc. for valuable discussion and input. This work was supported by a Post Doc grant (8054-00026B) from the Innovation Fund Denmark (MD). Finally, we thank all study participants, who made this work possible.

Author information

These authors contributed equally: Joshua W. Buckholtz, Jordan W. Smoller.

Authors and Affiliations

H. Lundbeck A/S, Valby, Denmark

Maria Dalby, Morana Vitezic, Niels Plath & Lene Hammer-Helmich

Department of Medical Epidemiology and Biostatistics, Karolinska Institutete, Stockholm, Sweden

Maria Dalby & Patrick F. Sullivan

23andMe Inc, Sunnyvale, CA, USA

Yunxuan Jiang, Chao Tian, Devika Dhamija, Catherine H. Wilson, David Hinds, Stella Aslibekyan, Adam Auton, Elizabeth Babalola, Robert K. Bell, Jessica Bielenberg, Katarzyna Bryc, Emily Bullis, Daniella Coker, Gabriel Cuellar Partida, Sayantan Das, Sarah L. Elson, Teresa Filshtein, Kipper Fletez-Brant, Pierre Fontanillas, Will Freyman, Anna Faaborg, Shirin T. Fuller, Pooja M. Gandhi, Julie M. Granka, Karl Heilbron, Alejandro Hernandez, Barry Hicks, Ethan M. Jewett, Katelyn Kukar, Keng-Han Lin, Maya Lowe, Jey C. McCreight, Matthew H. McIntyre, Steven J. Micheletti, Meghan E. Moreno, Joanna L. Mountain, Priyanka Nandakumar, Elizabeth S. Noblin, Jared O’Connell, Yunru Huang, Joanne S. Kim, Vanessa Lane, Aaron A. Petrakovitz, G. David Poznik, Morgan Schumacher, Anjali J. Shastri, Janie F. Shelton, Jingchunzi Shi, Suyash Shringarpure, Christophe Toukam Tchakouté, Vinh Tran, Joyce Y. Tung, Xin Wang, Wei Wang, Peter Wilton & Corinna Wong

Department of Genetics and Psychiatry, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

Patrick F. Sullivan

Department of Psychology, Harvard University, Cambridge, MA, USA

Joshua W. Buckholtz

Department of Psychiatry, Massachusetts General Hospital, Boston, MA, USA

Stanley Center for Psychiatric Research, Broad Institute of MIT and Harvard, Cambridge, MA, USA

Jordan W. Smoller

Psychiatric and Neurodevelopmental Genetics Unit, Massachusetts General Hospital, Boston, MA, USA

You can also search for this author in PubMed Google Scholar

23andMe Research Team

Stella Aslibekyan
, Adam Auton
, Elizabeth Babalola
, Robert K. Bell
, Jessica Bielenberg
, Katarzyna Bryc
, Emily Bullis
, Daniella Coker
, Gabriel Cuellar Partida
, Sayantan Das
, Sarah L. Elson
, Teresa Filshtein
, Kipper Fletez-Brant
, Pierre Fontanillas
, Will Freyman
, Anna Faaborg
, Shirin T. Fuller
, Pooja M. Gandhi
, Julie M. Granka
, Karl Heilbron
, Alejandro Hernandez
, Barry Hicks
, Ethan M. Jewett
, Katelyn Kukar
, Keng-Han Lin
, Maya Lowe
, Jey C. McCreight
, Matthew H. McIntyre
, Steven J. Micheletti
, Meghan E. Moreno
, Joanna L. Mountain
, Priyanka Nandakumar
, Elizabeth S. Noblin
, Jared O’Connell
, Yunru Huang
, Joanne S. Kim
, Vanessa Lane
, Aaron A. Petrakovitz
, G. David Poznik
, Morgan Schumacher
, Anjali J. Shastri
, Janie F. Shelton
, Jingchunzi Shi
, Suyash Shringarpure
, Christophe Toukam Tchakouté
, Vinh Tran
, Joyce Y. Tung
, Peter Wilton
& Corinna Wong

Contributions

Conceptualization: CHW, JWS, JWB. Data curation: MD, YJ, DD. Formal analysis: MD, YJ. Funding acquisition: 23andMe Research Team, NP. Project administration: 23andMe Research Team, NP, JWS, JWB. Supervision: PFS, LH-H, MV, DH, JWS, JWB. Writing the original draft: MD, PFS, JWS, JWB. Reviewing and editing: all authors. 23andMe Research Team contributed to this study: SA, AA, EBabalola, RKB, JB, KB, EBullis, DC, GCP, DD, SD, SLE, TF, KF-B, PF, WF, AF, STF, PMG, KH, BH, EMJ, KK, K-HL, ML, JCMcC, MHM, SJM, MEM, JLM, PN, ESN, JO’C, YH, AAP, VL, JSK, GDP, MS, AJS, JFS, JS, SS, VT, JYT, XW, WW, PW, AH, CWong, CTT.

Corresponding authors

Correspondence to Maria Dalby or David Hinds .

Ethics declarations

Competing interests.

The study was funded by H. Lundbeck A/S and the Milken Institute. MD, MV, NP, and LH-H are employees of H. Lundbeck A/S. DH, YJ, CTT, DD, CHW, and members of the 23andMe Research Team are employees of 23andMe, Inc. JWS is a member of the Leon Levy Foundation Neuroscience Advisory Board and received an honorarium for an internal seminar at Biogen, Inc.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary materials, supplementary tables 1-9, supplementary table 10, supplementary table 11, supplementary table 12, supplementary table 13, supplementary table 14, supplementary table 15, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Dalby, M., Vitezic, M., Plath, N. et al. Characterizing mood disorders in the AFFECT study: a large, longitudinal, and phenotypically rich genetic cohort in the US. Transl Psychiatry 12 , 121 (2022). https://doi.org/10.1038/s41398-022-01877-2

Download citation

Received : 06 September 2021

Revised : 23 February 2022

Accepted : 24 February 2022

Published : 25 March 2022

DOI : https://doi.org/10.1038/s41398-022-01877-2

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Olfactory genes affect major depression in highly educated, emotionally stable, lean women: a bridge between animal models and precision medicine.

Nora Eszlari
Gabor Hullam
Gabriella Juhasz

Translational Psychiatry (2024)

Polygenic heterogeneity in antidepressant treatment and placebo response

Anne Krogh Nøhr
Annika Forsingdal
Maria Dalby

Translational Psychiatry (2022)

Quick links

Explore articles by subject
Guide to authors
Editorial policies

Introduction
Conclusions
Article Information

A, There was no association between COVID-19 susceptibility and preexisting mood disorders (n = 65 514 469). B, The odds of COVID-19 hospitalization were significantly greater for individuals with preexisting mood disorders when compared with those without mood disorders (n = 26 554 397). C, There was no association between COVID-19 severe events or preexisting mood disorders (n = 83 240). D, The odds of COVID-19-related death were significantly greater for individuals with preexisting mood disorders when compared with those without mood disorders (n = 25 808 660). Squares represent effect sizes of individual studies, lines indicate the 95% CIs, and the diamond represents the summary effect size (ie, statistical combination of the effect sizes of component studies via the random-effects model).

eAppendix 1. Sample Search Strategies

eAppendix 2. Modified Newcastle-Ottawa Scales

eTable 1. Methodological Quality and Risk of Bias Assessment of Each Component Study Using the Modified Newcastle-Ottawa Quality Assessment Scales

eTable 2. Random Effects Meta-Regression Results

eFigure 1. Funnel Plot of Standard Error for All Component Studies Included in All Four Meta-analyses (Both Adjusted and Unadjusted Odds Ratios)

eFigure 2. Funnel Plot of Standard Error for Studies Included in the COVID-19 Hospitalization Meta-analysis (Both Adjusted and Unadjusted Odds Ratios)

eFigure 3. Funnel Plot of Standard Error for Studies Included in the COVID-19 Death Meta-analysis (Both Adjusted and Unadjusted Odds Ratios)

See More About

Select your interests.

Customize your JAMA Network experience by selecting one or more topics from the list below.

Academic Medicine
Acid Base, Electrolytes, Fluids
Allergy and Clinical Immunology
American Indian or Alaska Natives
Anesthesiology
Anticoagulation
Art and Images in Psychiatry
Artificial Intelligence
Assisted Reproduction
Bleeding and Transfusion
Caring for the Critically Ill Patient
Challenges in Clinical Electrocardiography
Climate and Health
Climate Change
Clinical Challenge
Clinical Decision Support
Clinical Implications of Basic Neuroscience
Clinical Pharmacy and Pharmacology
Complementary and Alternative Medicine
Consensus Statements
Coronavirus (COVID-19)
Critical Care Medicine
Cultural Competency
Dental Medicine
Dermatology
Diabetes and Endocrinology
Diagnostic Test Interpretation
Drug Development
Electronic Health Records
Emergency Medicine
End of Life, Hospice, Palliative Care
Environmental Health
Equity, Diversity, and Inclusion
Facial Plastic Surgery
Gastroenterology and Hepatology
Genetics and Genomics
Genomics and Precision Health
Global Health
Guide to Statistics and Methods
Hair Disorders
Health Care Delivery Models
Health Care Economics, Insurance, Payment
Health Care Quality
Health Care Reform
Health Care Safety
Health Care Workforce
Health Disparities
Health Inequities
Health Policy
Health Systems Science
History of Medicine
Hypertension
Images in Neurology
Implementation Science
Infectious Diseases
Innovations in Health Care Delivery
JAMA Infographic
Law and Medicine
Leading Change
Less is More
LGBTQIA Medicine
Lifestyle Behaviors
Medical Coding
Medical Devices and Equipment
Medical Education
Medical Education and Training
Medical Journals and Publishing
Mobile Health and Telemedicine
Narrative Medicine
Neuroscience and Psychiatry
Notable Notes
Nutrition, Obesity, Exercise
Obstetrics and Gynecology
Occupational Health
Ophthalmology
Orthopedics
Otolaryngology
Pain Medicine
Palliative Care
Pathology and Laboratory Medicine
Patient Care
Patient Information
Performance Improvement
Performance Measures
Perioperative Care and Consultation
Pharmacoeconomics
Pharmacoepidemiology
Pharmacogenetics
Pharmacy and Clinical Pharmacology
Physical Medicine and Rehabilitation
Physical Therapy
Physician Leadership
Population Health
Primary Care
Professional Well-being
Professionalism
Psychiatry and Behavioral Health
Public Health
Pulmonary Medicine
Regulatory Agencies
Reproductive Health
Research, Methods, Statistics
Resuscitation
Rheumatology
Risk Management
Scientific Discovery and the Future of Medicine
Shared Decision Making and Communication
Sleep Medicine
Sports Medicine
Stem Cell Transplantation
Substance Use and Addiction Medicine
Surgical Innovation
Surgical Pearls
Teachable Moment
Technology and Finance
The Art of JAMA
The Arts and Medicine
The Rational Clinical Examination
Tobacco and e-Cigarettes
Translational Medicine
Trauma and Injury
Treatment Adherence
Ultrasonography
Users' Guide to the Medical Literature
Vaccination
Venous Thromboembolism
Veterans Health
Women's Health
Workflow and Process
Wound Care, Infection, Healing

Others Also Liked

Download PDF
X Facebook More LinkedIn

Ceban F , Nogo D , Carvalho IP, et al. Association Between Mood Disorders and Risk of COVID-19 Infection, Hospitalization, and Death : A Systematic Review and Meta-analysis . JAMA Psychiatry. 2021;78(10):1079–1091. doi:10.1001/jamapsychiatry.2021.1818

Manage citations:

Permissions

Association Between Mood Disorders and Risk of COVID-19 Infection, Hospitalization, and Death : A Systematic Review and Meta-analysis

1 Department of Immunology, University of Toronto, Toronto, Ontario, Canada
2 Mood Disorders Psychopharmacology Unit, Poul Hansen Family Centre for Depression, University Health Network, Toronto, Ontario, Canada
3 Braxia Health, Mississauga, Ontario, Canada
4 Department of Psychiatry, University of Toronto, Toronto, Ontario, Canada
5 Department of Affective Disorders, the Affiliated Brain Hospital of Guangzhou Medical University (Guangzhou Huiai Hospital), Guangzhou Medical University, Guangzhou, China
6 Laboratory of Emotion and Cognition, the Affiliated Brain Hospital of Guangzhou Medical University (Guangzhou Huiai Hospital), Guangzhou Medical University, Guangzhou, China
7 Department of Psychological Medicine, Yong Loo Lin School of Medicine, National University of Singapore, Singapore
8 Institute for Health Innovation and Technology, National University of Singapore, Singapore

Question Are preexisting mood disorders associated with higher risk of COVID-19 infection, hospitalization, severe complications, and death?

Findings In this systematic review and meta-analysis of more than 91 million people, individuals with preexisting mood disorders, compared with those without mood disorders, had significantly higher pooled odds ratios for COVID-19 hospitalization and death. There were no associations between preexisting mood disorders and risk of COVID-19 infection or severe events.

Meaning These results suggest that individuals with mood disorders should be categorized as an at-risk group for COVID-19 hospitalization and death, providing basis for vaccine prioritization.

Importance Preexisting noncommunicable diseases (eg, diabetes) increase the risk of COVID-19 infection, hospitalization, and death. Mood disorders are associated with impaired immune function and social determinants that increase the risk of COVID-19. Determining whether preexisting mood disorders represent a risk of COVID-19 would inform public health priorities.

Objective To assess whether preexisting mood disorders are associated with a higher risk of COVID-19 susceptibility, hospitalization, severe complications, and death.

Data Sources Systematic searches were conducted for studies reporting data on COVID-19 outcomes in populations with and without mood disorders on PubMed/MEDLINE, The Cochrane Library, PsycInfo, Embase, Web of Science, Google/Google Scholar, LitCovid, and select reference lists. The search timeline was from database inception to February 1, 2021.

Study Selection Primary research articles that reported quantitative COVID-19 outcome data in persons with mood disorders vs persons without mood disorders of any age, sex, and nationality were selected. Of 1950 articles identified through this search strategy, 21 studies were included in the analysis.

Data Extraction and Synthesis The modified Newcastle-Ottawa Scale was used to assess methodological quality and risk of bias of component studies. Reported adjusted odds ratios (ORs) were pooled with unadjusted ORs calculated from summary data to generate 4 random-effects summary ORs, each corresponding to a primary outcome.

Main Outcomes and Measures The 4 a priori primary outcomes were COVID-19 susceptibility, COVID-19 hospitalization, COVID-19 severe events, and COVID-19 death. The hypothesis was formulated before study search. Outcome measures between individuals with and without mood disorders were compared.

Results This review included 21 studies that involved more than 91 million individuals. Significantly higher odds of COVID-19 hospitalization (OR, 1.31; 95% CI, 1.12-1.53; P = .001; n = 26 554 397) and death (OR, 1.51; 95% CI, 1.34-1.69; P < .001; n = 25 808 660) were found in persons with preexisting mood disorders compared with those without mood disorders. There was no association between mood disorders and COVID-19 susceptibility (OR, 1.27; 95% CI, 0.73-2.19; n = 65 514 469) or severe events (OR, 0.94; 95% CI, 0.87-1.03; n = 83 240). Visual inspection of the composite funnel plot for asymmetry indicated the presence of publication bias; however, the Egger regression intercept test result was not statistically significant.

Conclusions and Relevance The results of this systematic review and meta-analysis examining the association between preexisting mood disorders and COVID-19 outcomes suggest that individuals with preexisting mood disorders are at higher risk of COVID-19 hospitalization and death and should be categorized as an at-risk group on the basis of a preexisting condition.

Despite ongoing public health efforts, the devastating impact of COVID-19 continues to be observed worldwide. In October 2020, the World Health Organization estimated that approximately 10% of the global population had been infected with COVID-19, representing 20 times the number of recorded cases. 1 An early study 2 reported that approximately 14% of affected individuals experienced severe COVID-19–associated symptoms, whereas 5% presented as critically unwell and required intensive care. Established risk factors for severe COVID-19 include preexisting cardiovascular disease, obesity, diabetes, cancers, and respiratory disease. 3 , 4

Individuals with mood disorders may be at greater risk (and vice versa) for COVID-19 because of a confluence of factors known to increase the risk in the general population. 5 For example, the finding that a subpopulation of individuals with mood disorders exhibit evidence of dysregulated immune function has been replicated in several studies. 6 , 7 Moreover, persons with mood disorders are differentially affected by noncommunicable diseases (eg, obesity and cardiovascular disease) known to increase the risk of COVID-19. 8 , 9 In addition, social determinants of risk (eg, poverty and insufficient access to timely and preventive health care) are also more commonly observed in persons with mood disorders.

The fact that COVID-19 is associated with complex and, in some cases, enduring neuropsychiatric manifestations has been amply documented. 10 - 12 A related but separate question is whether individuals with mood disorders are at higher risk of contracting COVID-19 and/or experiencing complications and death from the disease. 13 - 15 A higher risk of infection and/or complications attributable to COVID-19 in individuals with psychiatric illnesses has been reported 16 ; however, to our knowledge, no meta-analysis has delimited its focus to persons with mood disorders. Empirical evidence addressing this question is crucial given the high global lifetime prevalence of mood disorders in the general population and the need to prioritize public health strategies to mitigate the risk of COVID-19 and associated complications. 17 In this study, we hypothesized that individuals with preexisting mood disorders are at higher risk of COVID-19 susceptibility, hospitalization, severe events, and death.

The protocol pertaining to this study was registered on PROSPERO ( CRD42020220502 ). A systematic search was conducted from database inception to February 1, 2021, for primary research articles. The following databases were searched systematically: PubMed/MEDLINE, the Cochrane Library, PsycInfo, Embase, and Web of Science. We manually searched the references of relevant articles as well as Google Scholar/Google and LitCovid for additional studies. No language restrictions were imposed. Sample search strategies are provided in eAppendix 1 in the Supplement . A total of 1950 articles were identified through this search strategy. This study followed the Meta-analysis of Observational Studies in Epidemiology ( MOOSE ) 18 and Preferred Reporting Items for Systematic Reviews and Meta-analyses ( PRISMA ) 19 reporting guidelines.

Titles and abstracts were independently screened by 2 reviewers (F.C. and D.N.), articles identified as potentially relevant by at least 1 reviewer were retrieved, and duplicates were removed. Full-text articles were independently screened by 2 reviewers (F.C. and D.N.), with discrepancies resolved through discussion.

We compared COVID-19 test positivity, hospitalizations, severe events (including intensive care unit admission, mechanical ventilatory support, oxygen therapy, extracorporeal membrane oxygenation, acute respiratory distress syndrome, and/or cardiopulmonary resuscitation), and/or death in populations with mood disorders vs without mood disorders of any age, sex, and nationality. Inclusion criteria were established before article review and were as follows: (1) a diagnosis of depression or bipolar disorder using standardized diagnostic criteria (eg, DSM-5 or International Statistical Classification of Diseases and Related Health Problems, Tenth Revision [ICD-10] ); (2) mood disorder diagnosis documented before COVID-19 event or present as of the index date (ie, comorbid); (3) COVID-19 ascertained according to laboratory testing, ICD-10 , electronic health record (EHR), and/or clinical judgment; (4) available complete quantitative data relevant to at least 1 primary outcome ( Table 1 ); (5) discrete outcome data for mood disorders; (6) primary research; and (7) presentation as a full-text article (which included preprints). Exclusion criteria were as follows: (1) mood disorders that were self-reported or otherwise did not follow standardized diagnostic criteria; (2) mood disorder outcomes grouped with those for other mental illnesses; (3) unpublished study or abstract; (4) and nonprimary research.

Methodological quality and risk of bias were assessed using the Newcastle-Ottawa Scale (NOS), 20 modified for applicable retrospective cohort and case-control studies using patient EHR data (eAppendix 2 in the Supplement ). Each included study was independently rated by 2 reviewers (F.C. and D.N.) and results were corroborated, with discrepancies resolved through discussion. Studies ranking 7 to 9 stars on the case-control NOS or 6 to 7 stars on the cohort NOS were deemed high quality, those ranking 5 to 6 (case-control) or 5 (cohort) stars were moderate, and 4 or fewer stars were low quality.

Data from included articles were independently extracted by 2 reviewers (F.C. and D.N.) using a pilot-tested data extraction form and then corroborated, with discrepancies resolved through discussion. Information to be extracted was established a priori and included study characteristics, participant characteristics and subgroups, sample source and collection period, modes of ascertainment, methods of data analysis, selection of cases and controls, and quantitative data pertaining to any primary outcomes along with adjusted factors.

Odds ratios (ORs) with 95% CIs and/or summary data for COVID-19 test positivity, hospitalizations, severe events, and/or death in corresponding study populations with and without preexisting mood disorders were extracted. Where available, adjusted ORs (aORs) were preferentially used to reduce confounding, as recommended by the Cochrane Handbook . 21 When relevant studies did not report an applicable OR, unadjusted ORs were calculated from summary data and combined with reported aORs. The meta-analyses included 21 studies; aORs were derived from 11 studies, 22 - 32 crude summary data from 6 studies, 33 - 38 and both measures from 4 studies. 39 - 42 Authors of potentially eligible studies were contacted and asked to provide clarification and/or supplementary data where necessary; 4 study authors provided eligible summary data 34 - 36 and 1 provided aORs. 42 Among multiple statistical models within 1 study, the aOR most similar to that adjusted for age, sex, and comorbidities was used because these were the most common corrections and represent established COVID-19 risk factors. When aORs were reported only for subgroups, composite values for the aggregate research populations were calculated. Recent diagnoses of mood disorders were preferred over lifetime diagnoses if both were reported.

Pooled ORs with 95% CIs and prediction intervals (PIs) representing the range of values expected to contain a future observation have been provided (except for analyses in which all studies shared a common effect size), and forest plots were generated for 4 prespecified COVID-19 events: (1) susceptibility, (2) hospitalization, (3) severe events, and (4) deaths ( Table 1 ). An α level of .05 was chosen to indicate statistical significance. The comparative groups were individuals with preexisting mood disorders (ie, exposure) vs those without mood disorders as of the index date. Study participants who experienced more than 1 COVID-19 outcome were not censored from precursory event counts. Summary data that encompassed varying intrasample follow-up times 31 , 36 (ie, where hazards ratios would be appropriate) were not used to calculate unadjusted ORs owing to high daily COVID-19 event rates. Measures that pertain to specific mood disorders (assessed independently) were combined with those for mood disorder subgroups and all mood disorders grouped. When measures for overlapping categories were available (eg, major depressive disorder and depression), the data that would encompass a greater range of exposures were used.

Observational studies are susceptible to interstudy differences in exposure and participants, justifying the use of the random-effects model of DerSimonian and Laird, which assumes varying effect sizes that arise from heterogeneity. 43 Furthermore, heterogeneity was evaluated using the I 2 statistic, where a 30% cutoff denotes moderate heterogeneity, 50% denotes substantial heterogeneity, and 75% denotes considerable heterogeneity, as recommended by GRADE (grading of recommendations, assessment, development and evaluation) criteria and the Cochrane Handbook ’s interpretation of heterogeneity scores. 44 , 45 Exploratory post hoc sensitivity analyses were undertaken for subgroups of studies to ascertain the impact of moderator variables, and a meta-regression was performed to investigate heterogeneity.

To consider the impact of possible data duplication caused by identical sample source, sensitivity analyses retaining the study with the largest sample size derived from each duplicate database were conducted. The Egger regression intercept test and visual inspection of funnel plots for each COVID-19 event, as well as for all component data (including some overlap if 1 study reported multiple types of COVID-19 events as defined in Table 1 ), was performed to assess publication bias. All statistical analyses were conducted using Comprehensive Meta-Analysis, version 3.0 (BioStat Inc). A narrative synthesis was included for relevant results excluded from the meta-analyses. 31 , 36

The literature search yielded 1950 articles of which 59 were eligible after screening titles and abstracts and removing duplicates. Of these eligible studies, 38 were further excluded after full-text screening. Details of study selection are provided in Figure 1 . Twenty-one studies were included in the review: 8 retrospective case-control studies 22 - 24 , 27 , 28 , 38 , 41 (including 1 phenome-wide association study 42 ), 12 retrospective cohort studies, 25 , 26 , 29 - 32 , 34 - 37 , 39 , 40 and 1 exposure-crossover study. 33 All studies contributed to at least 1 meta-analysis. Twelve studies analyzed data from the US, 22 - 25 , 31 , 33 , 34 , 38 - 42 2 from South Korea, 26 , 27 2 from Spain, 29 , 30 and 1 each from Italy, 28 Turkey, 37 the UK, 35 England, 32 and Israel. 36 Sample sizes ranged from 398 to 69.8 million, and data acquisition periods spanned from December 1, 2019, to September 30, 2020. Table 2 provides detailed characteristics and summaries of applicable findings for all component studies.

Methodological quality of the included studies was moderate to high, evidenced by a mean score of 6.7 of 9 for case-control studies and 5.7 of 7 for cohort studies. The NOS rankings within each category for individual studies organized by design are provided in eTable 1 in the Supplement .

Meta-analyses of the 4 primary outcomes revealed significantly higher odds of COVID-19 hospitalization and death among individuals with preexisting mood disorders compared with those without mood disorders. No significant associations between COVID-19 susceptibility or severe events were observed. Visual inspection of the composite funnel plot for asymmetry indicated the presence of publication bias. However, the Egger regression intercept test result was not statistically significant (eFigure 1 in the Supplement ).

There was no association between COVID-19 susceptibility and preexisting mood disorders (OR, 1.27; 95% CI, 0.73-2.19; P = .40; PI, 0.11-14.23; n = 65 514 469) ( Figure 2 A). Furthermore, no sensitivity analyses produced any statistically significant associations ( Table 3 ), with the exception of the analysis restricted to moderate- to low-quality studies (OR, 2.16; 95% CI, 1.28-3.66; P = .004) ( Table 3 ).

The odds of COVID-19 hospitalization were significantly greater for individuals with preexisting mood disorders when compared with those without mood disorders (OR, 1.31; 95% CI, 1.12-1.53; P = .001; PI, 0.78-2.20; n = 26 554 397) ( Figure 2 B). All sensitivity analyses produced statistically significant higher odds of COVID-19 hospitalization ( Table 3 ), with the exception of the analysis restricted to case-control studies (OR, 1.16; 95% CI, 0.95-1.43; P = .15) ( Table 3 ).

There was no association between COVID-19 severe events and preexisting mood disorders (OR, 0.94; 95% CI, 0.87-1.03; P = .19; n = 83 240) ( Figure 2 C). Furthermore, sensitivity analyses restricted to depression, comparing unadjusted vs aORs, based on study design or quality, and excluding possible data duplication did not find any associations ( Table 3 ).

The odds of COVID-19–related death were significantly greater for individuals with preexisting mood disorders when compared with those without mood disorders (OR, 1.51; 95% CI, 1.34-1.69; P < .001; PI, 1.09-2.07; n = 25 808 660) ( Figure 2 D). Compared with the foregoing OR, the summary effect size was greater when the meta-analysis was restricted to aORs (OR, 1.57; 95% CI, 1.26-1.95; P < .001) ( Table 3 ) and markedly so when restricting analysis to high-quality studies (OR, 1.80; 95% CI, 1.30-2.51; P < .001) ( Table 3 ). Furthermore, the effect size modestly increased when the exposure was restricted to depression (OR, 1.55; 95% CI, 1.34-1.79; P < .001) and did not markedly change depending on study design ( Table 3 ).

The meta-analysis for COVID-19 hospitalization exhibited considerable heterogeneity ( I 2 = 98.5%). The meta-analysis for COVID-19–related death exhibited substantial heterogeneity ( I 2 = 67.2%). Meta-regression results are provided in eTable 2 in the Supplement .

Visual inspection of funnel plot asymmetry for the COVID-19 hospitalization meta-analysis indicated the presence of publication bias, and the Egger regression intercept test was statistically significant (eFigure 2 in the Supplement ). Likewise, visual inspection of funnel plot asymmetry for the COVID-19–related death meta-analysis indicated the presence of publication bias, and the Egger regression intercept test result was statistically significant (eFigure 3 in the Supplement ). Although the Egger regression intercept test result was not statistically significant (eFigure 1 in the Supplement ), visual inspection of the composite funnel plot (ie, for all data sets included in the review, including both aORs and unadjusted ORs) indicated the presence of publication bias.

Relevant results from Tang et al 31 could not be incorporated into the COVID-19 hospitalization meta-analysis because of variant intrastudy follow-up times; however, this study similarly established a higher subhazard ratio of 1.33 (95% CI, 1.01-1.75; P < .05) of COVID-19 hospitalization for skilled nursing facility residents with preexisting mood disorders vs those without mood disorders. Relevant results from Yanover et al 36 could not be incorporated into the COVID-19 severe events meta-analysis because of variant intrastudy follow-up times; however, there was no association between preexisting depression and COVID-19 complications (including death) (aOR, 1.58; 95% CI, 0.70-3.70). Finally, relevant results from Tang et al 31 could not be included in the COVID-19–related death meta-analysis because of variant intrastudy follow-up times; however, the reported result was not statistically significant (hazard ratio, 1.08; 95% CI, 0.77-1.50) for COVID-19–related death.

This systematic review and meta-analysis identified higher odds of COVID-19 hospitalization and death in individuals with preexisting mood disorders compared with individuals without mood disorders. However, no significant associations between COVID-19 susceptibility and severe events were identified. The foregoing results are in accordance with a recent meta-analysis that examined all grouped psychiatric illnesses by Toubasi et al, 16 which reported a higher risk of COVID-19 death, although the current meta-analysis does not similarly report a higher risk of severe events.

There are multiple pathways by which persons with mood disorders may be at greater risk for COVID-19 hospitalization and death. Social determinants, including economic insecurity, insufficient access to primary preventive health care, and lower health literacy, may portend COVID-19 risk. 46 , 47 For example, many individuals with mood disorders reside in congregate facilities, such as psychiatric inpatient units, homeless shelters, community housing, and prisons, where risk of COVID-19 transmission is increased because of the inability to effectively socially distance and/or quarantine. 48 , 49 Moreover, symptoms of mood disorders, including disinhibition, apathy, avolition, and cognitive deficits, may presage nonconcordance with healthy behaviors and possibly public health directives. However, some of the possible mediators discussed may act as confounders; hence, causal inferences with respect to social determinants of health, mood disorders, and COVID-19 outcomes cannot be established.

Furthermore, cigarette smoking 50 , 51 and substance use disorders, 52 established risk factors for COVID-19 infection and complications, are significantly more prevalent among individuals with mood disorders. In addition, persons with mood disorders are differentially affected by noncommunicable diseases that are established risk factors for COVID-19 (eg, obesity and cardiovascular disease 8 , 9 ) as well as behaviors (eg, sleep dysregulation and habitual inactivity) that may presage COVID-19 risk. 53

Disturbance in immune regulation is also a well-documented abnormality in subpopulations of persons with mood disorders. 6 , 7 For example, it has been reported that subsets of populations with mood disorders exhibit increases in central and peripheral levels of acute-phase proteins, such as C-reactive peptide, and proinflammatory cytokines, such as interleukin 6 and tumor necrosis factor α. 6 , 54 - 57 The inflammatory signature of mood disorders overlaps with clinical reports 57 , 58 of increased cytokine activity in persons affected with severe COVID-19. It is hypothesized that the cytokine disturbance intrinsic to mood disorders may be exacerbated in situations of COVID-19 infection, increasing the risk of death and complications from COVID-19.

Pharmacotherapy prescribed to individuals with mood disorders exerts disparate effects on the immune inflammatory system. Benzodiazepines and select atypical antipsychotics have been associated with a higher risk of pneumonia and/or COVID-19. 59 - 62 Interpretation of this finding is, however, complicated by confounding by indication as well as by other lines of research reporting that some antipsychotics exert anti-inflammatory effects. 63 Moreover, lithium exerts substantial immune-modulating and anti-inflammatory effects, 64 and valproate has been associated with a lower risk of respiratory infections. 65 Preliminary results from controlled and observational studies 66 - 70 also suggest that conventional antidepressants provide protective effects against COVID-19 complications. The foregoing finding would be in accordance with preclinical work documenting anti-inflammatory and/or anti–SARS-CoV-2 replication effects of select selective serotonin reuptake inhibitors. 71 , 72

It warrants consideration that the risk of being infected and possibly experiencing complications from COVID-19 in adults with mood disorders may have overlapping but different determinants. For example, individuals with mood disorders report higher rates of social isolation, unemployment, and reduced interpersonal contact, which would be hypothesized to decrease their risk of COVID-19 exposure and/or complications. In addition, individuals with mood disorders are also more likely to have insufficient access to primary preventive health care, live in congregate settings, and engage in behaviors (eg, smoking cigarettes) that would place them at higher risk of contracting COVID-19 and experiencing complications, as previously discussed.

Moreover, it is counterintuitive that the OR for the intermediary of COVID-19 severe events was not statistically significant, whereas the ORs for hospitalization and death were statistically significant. Possible explanations for this finding include interstudy variation in how severe events were defined ( Table 1 ), differences across studies in event reporting and coding, and heterogeneity in statistical approaches. Furthermore, the COVID-19 severe events analysis included a relatively small sample size.

The calculated OR for COVID-19–related death is comparable to reported unadjusted ORs and aORs for other preexisting conditions that are risk factors for COVID-19 (eg, diabetes, 73 , 74 liver disease 74 cancer, 75 and obesity 4 , 76 ). Notably, the higher COVID-19 risk in individuals with mood disorders cannot be fully accounted for by medical comorbidity.

This study has limitations. There are several methodological aspects that may affect inferences and interpretations of the results. The pathways highlighted are not definitive and thus may not adequately address associations among the primary outcomes. First, component studies were observational, and therefore causal relationships cannot be inferred. Second, data missing because of factors such as publication bias and the possibility of grouping of mood disorders with other mental disorders could confound interpretation. Third, the lack of distinction in multiple studies as to whether an individual with a mood disorder had major depressive disorder or bipolar disorder is also a substantial limitation insofar as there are differences between the 2 conditions in their overall risk of medical comorbidity. 77 Fourth, it was not possible to quantify the extent of unexplained heterogeneity or assess the causes of the substantial to considerable heterogeneity among studies included in our random-effects meta-analytic model. We recognize that high heterogeneity in meta-analytic studies may belie study findings. Fifth, it is not certain whether, in some cases, the dependent measures (eg, hospitalization) were confounded by other factors (eg, health care systems, staffing, local standard operating procedures, policies regarding hospitalization and testing, and attribution of cause of death), 78 and a quantitative characterization of the influence of mediators (eg, comorbidities) was not possible because sufficient information was not provided by the included studies. Sixth, many component studies had insufficient characterization of patient sociodemographic characteristics, as well as clinical, psychiatric, comorbidity, and smoking histories, which may have confounded or limited results. Seventh, there is a possibility that the component studies may have misdiagnosed patients or misentered patient information into administrative data records, which is an inherent limitation of observational data and administrative data sets. In addition, there was variability in how controls were defined ( Table 1 ); notably, this study included individuals ascertained as not having COVID-19 who may not have had a laboratory confirmation. Furthermore, the Egger regression test is suboptimal toward the characterization of publication bias in analyses that involve ORs. 79 An additional methodological aspect that may affect interpretation of our findings relates to our inability to fully characterize the temporality of events. For example, it is possible that, in some cases, COVID-19 was not a subsequent event in an individual with a previously declared mood disorder. Instead, in some cases COVID-19 may have been an antecedent to the onset of mood disorders because it is well established that COVID-19 infection results in a significant increase in a variety of neuropsychiatric disorders. 12 Moreover, some of the included studies were preprints ( Table 2 ) and thus not yet peer reviewed. In addition, the possibility that some studies included overlapping samples cannot be excluded. Notwithstanding, results from the sensitivity analyses, wherein studies that may have had overlapping samples were excluded, were consistent with the overall findings ( Table 3 ). Most of the included studies were conducted within the first 6 to 9 months of the pandemic, which possibly oversampled persons with symptomatic COVID-19 because of prioritized testing and thus may underestimate the associations between mood disorders and COVID-19 complications. Selective COVID-19 testing and a lack of adjudication in databases or accuracy of the original clinical diagnoses were not accounted for by either of the modified NOSs.

In this systematic review and meta-analysis examining the association between preexisting mood disorders and COVID-19 outcomes, results of analyses of more than 91 million people indicated that individuals with preexisting mood disorders are at a higher risk of COVID-19 hospitalization and death. These results suggest that individuals with mood disorders, like persons with other preexisting conditions (eg, obesity), should be categorized as an at-risk group on the basis of a preexisting condition. Future research should address whether COVID-19 vaccinations exhibit differential efficacy in persons with mood disorders and whether COVID-19 infection affects the longitudinal trajectory of the underlying mental disorder. 80 , 81

Accepted for Publication: May 25, 2021.

Published Online: July 28, 2021. doi:10.1001/jamapsychiatry.2021.1818

Corresponding Author: Roger S. McIntyre, MD, University Health Network, 399 Bathurst St, MP 9-325, Toronto, ON M5T 2S8, Canada ( [email protected] ).

Author Contributions: Dr McIntyre had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Ceban, Lee, Xiong, Gill, Mansur, Rosenblat, Ho, McIntyre.

Acquisition, analysis, or interpretation of data: Ceban, Nogo, Carvalho, Lee, Nasri, Lui, Subramaniapillai, Liu, Joseph, Teopiz, Cao, Lin, Rosenblat, McIntyre.

Drafting of the manuscript: Ceban, Nogo, Carvalho, Lui, Liu, Teopiz, McIntyre.

Critical revision of the manuscript for important intellectual content: Ceban, Nogo, Carvalho, Lee, Nasri, Xiong, Subramaniapillai, Gill, Joseph, Cao, Mansur, Lin, Rosenblat, Ho, McIntyre.

Statistical analysis: Ceban, Carvalho.

Administrative, technical, or material support: Nogo, Carvalho, Nasri, Xiong, Lui, Subramaniapillai, Gill, Liu, Joseph, Rosenblat.

Supervision: Nogo, Lee, Mansur, Lin, Rosenblat, Ho, McIntyre.

Conflict of Interest Disclosures: Ms Lee reported receiving personal fees from Braxia Scientific Corp outside the submitted work. Ms Lui is a contractor to Braxia Scientific Corp. Ms Teopiz reported receiving personal fees from Braxia Scientific Corp outside the submitted work. Dr Rosenblat is the medical director of Braxia Health (formally known as the Canadian Rapid Treatment Center of Excellence and is a fully owned subsidiary of Braxia Scientific Corp), which provides ketamine and esketamine treatment for depression, and has received research grant support from the American Psychiatric Association, American Society of Psychopharmacology, Canadian Cancer Society, Canadian Psychiatric Association, Joseph M. West Family Memorial Fund, Timeposters Fellowship, University Health Network Centre for Mental Health, and University of Toronto and speaking, consultation, or research fees from Allergan, COMPASS, Janssen, Lundbeck, and Sunovion. Dr McIntyre reported receiving grant support from Canadian Institutes of Health Research, Global Alliance for Chronic Diseases, and Chinese National Natural Research Foundation and speaker/consultation fees from Lundbeck, Janssen, Purdue, Pfizer, Otsuka, Takeda, Neurocrine, Sunovion, Bausch Health, Novo Nordisk, Kris, Sanofi, Eisai, Intra-Cellular, NewBridge Pharmaceuticals, and AbbVie. Dr McIntyre is a chief executive officer of Braxia Scientific Corp.

Register for email alerts with links to free full-text articles
Access PDFs of free articles
Manage your interests
Save searches and receive search alerts

Masks Strongly Recommended but Not Required in Maryland

Respiratory viruses continue to circulate in Maryland, so masking remains strongly recommended when you visit Johns Hopkins Medicine clinical locations in Maryland. To protect your loved one, please do not visit if you are sick or have a COVID-19 positive test result. Get more resources on masking and COVID-19 precautions .

Vaccines
Masking Guidelines
Visitor Guidelines
Mood Disorders

A mood disorder is a type of mental health condition where there is a disconnect between actual life circumstances and the person's state of mind or feeling. A mood disorder can negatively affect your ability to function normally. It can have serious consequences in all aspects of life, from personal to professional.

Children, teens, and adults can all have mood disorders. But children and teens don’t always have the same symptoms as adults. It’s harder to diagnose mood disorders in children. That's because they can't always express how they feel, and symptoms may look different in children from how they look in adults.

Therapy, medicines, and support and self-care can help treat mood disorders.

What are the different types of mood disorders?

These are the most common types of mood disorders:

Major depression. Having less interest in normal activities, feeling sad or hopeless, and other symptoms for at least 2 weeks may mean depression.

Dysthymia. This is an ongoing (chronic), low-grade, depressed, or irritable mood that lasts for at least 2 years.

Bipolar disorder. With this condition, a person has times of depression alternating with times of mania or a higher mood.

Mood disorder linked to another health condition. Many health conditions (including cancer, injuries, infections, and chronic illnesses) can trigger symptoms of depression.

Substance-induced mood disorder. Symptoms of depression may be caused by drug abuse, alcohol use disorder, exposure to toxins, or side effects of medicines.

What causes mood disorders?

Many factors help lead to mood disorders. They are likely caused by an imbalance of brain chemicals. Life events (such as stressful life changes) may also help lead to a depressed mood. Mood disorders also tend to run in families.

Who is at risk for mood disorders?

Anyone can feel sad or depressed at times. But mood disorders are more intense and last longer. They are also harder to manage than normal feelings of sadness. Children, teens, or adults who have a parent with a mood disorder have a greater chance of also having a mood disorder. But life events and stress can expose or worsen feelings of sadness or depression. This makes the feelings harder to manage.

Sometimes life's problems can trigger depression. Things such as being fired from a job, getting divorced, losing a loved one, having a death in the family, and financial trouble can be difficult. Coping with the pressure may be troublesome. These life events and stress can bring on feelings of sadness or depression. Or they can make a mood disorder harder to manage.

The risk for depression in women is nearly twice as high as it is for men. Once a person in the family has this diagnosis, their siblings and their children have a higher chance of the same diagnosis.

What are the symptoms of mood disorders?

Depending on age and the type of mood disorder, a person may have different symptoms when they become depressed. The following are the most common symptoms of a mood disorder:

Ongoing sad, anxious, or “empty” mood

Feeling hopeless or helpless

Having low self-esteem

Feeling inadequate or worthless

Excessive guilt

Not interested in normal activities or activities that were once enjoyed, including sex

Relationship problems

Trouble sleeping or sleeping too much

Changes in appetite or weight

Decreased energy

Trouble focusing

Less able to make decisions

Frequent physical complaints (for example, headache, stomachache, or tiredness) that don’t get better with treatment

Running away or threats of running away from home

Very sensitive to failure or rejection

Irritability, hostility, or aggression

Repeated thoughts of death or suicide, planning for death, or wishing to die ( Note: People with this symptom should get treatment right away!)

In mood disorders, these feelings are more intense than what a person may normally feel from time to time. It’s also of concern if these feelings continue over time. Or if they interfere with someone's interest in family, friends, community, or work.

Any person who has thoughts of suicide should get medical help right away. If you can't get in immediately to your primary care provider, go to a reputable mental health facility in your community. Don't put it off.

The symptoms of mood disorders may seem like other conditions or mental health problems. Always talk with a healthcare provider for a diagnosis.

How are mood disorders diagnosed?

Mood disorders are serious illnesses. A psychiatrist, clinical psychologist, advanced practice registered nurse, or licensed clinical social worker can diagnose mood disorders after completing a complete health history and psychiatric evaluation.

How are mood disorders treated?

Mood disorders can often be treated with success. Treatment may include:

Antidepressant and mood-stabilizing medicines. These medicines work very well in treating mood disorders, especially when combined with psychotherapy.

Psychotherapy ( most often cognitive-behavioral or interpersonal therapy). This kind of therapy is focused on changing the person’s distorted view of themselves and their environment. It also helps to improve relationship skills. And it can help the person identify stressors in the environment and learn how to avoid or manage them.

Family therapy. A mood disorder can affect all aspects of a family (emotional, physical, occupational, and financial). Professional support can help both the person with the diagnosis and family members.

Other therapies. These may include transcranial stimulation and electroconvulsive therapy for refractory depression (treatment-resistant depression).

Families play a vital supportive role in any treatment process.

Someone with a mood disorder may have times of stability and times when symptoms return. Long-term, continuous treatment can help the person stay healthy and control symptoms.

When correctly diagnosed and treated, people with mood disorders can live stable, productive, healthy lives.

Can mood disorders be prevented?

At this time, there are no ways to prevent or reduce mood disorders. But early diagnosis and treatment can reduce the severity of symptoms. It can also enhance the person’s normal growth and development, and improve their quality of life. If you or someone you care about has symptoms of a mood disorder, talk to your healthcare provider.

Key points about mood disorders

A mood disorder is a class of serious mental illnesses. The term broadly describes all types of depression and bipolar disorders.

Children, teens, and adults can all have mood disorders.

Many factors help lead to mood disorders. They are likely caused by an imbalance of brain chemicals.

Most people with a mood disorder have ongoing feelings of sadness. They may feel helpless and hopeless.

Without treatment, symptoms can last for weeks, months, or years. They can affect quality of life.

Mood disorders are most often treated with medicine, psychotherapy or cognitive behavioral therapy, family therapy, or a combination of medicine and therapy.

Long-term, comprehensive follow-up care will help ensure the support needed for a full, productive life.

Research Shows Identifying More Personalized and Successful Treatments for Mood Disorders

Anxiety Disorders
Major Depression
Bipolar Disorder

Treatments, Tests and Therapies

How to Get a Toddler to Listen
Esketamine for Treatment-Resistant Depression
TomorrowsDiscoveries A Different Approach to Treating Psychiatric Disorders Atsushi Kamiya MD

Children's Health

Mental Health Disorders in Children and Teens
Childhood Depression
How to Get Your Kids to Do Chores
How to Foster Cultural Awareness at Home
7 Tips for Helping Your Child Deal with Bullying
Fentanyl Overdoses: Is My Child at Risk?
Anorexia Nervosa in Children

Find a Doctor

Specializing In:

Behavioral Disorder
Mood Disorders in Children
Multiple Endocrine Neoplasia (MEN)
Affective Disorders

Request an Appointment

Specializing In

At Another Johns Hopkins Member Hospital:

Howard County Medical Center
Sibley Memorial Hospital
Suburban Hospital

Louisville.edu
Health Sciences Center
PeopleSoft HR
PeopleSoft Campus Solutions
PeopleSoft Financials
Business Ops
Cardinal Careers

Department of Psychiatry and Behavioral Sciences

Departments /
Department of Psychiatry and Behavioral Sciences /

Mood Disorders Research Program

The Mood Disorders Research Program, under the direction of Dr. Rif El-Mallakh, pursues basic and clinical research in the major mood disorders, focusing on bipolar illness. Most basic research projects focus on mood-state-related abnormalities in nerve or membrane function. Studies include: transmembrane potential changes in Lymphocytes and Lymphoblasts, nerve conduction velocity determinations, measurement of endogenous digoxin-like compounds, determination of sodium pump isoform expression in brains of bipolar patients, and the development of an animal model for bipolar illness. Several clinical studies range from pharmacologic studies in acute mania, bipolar depression, and prophylaxis to genetic studies.

In addition to pharmaceutical company funding (from Eli Lilly, Pfizer, Glaxo-Wellcome) the program has NIMH funding for a genetics study. Our laboratory has been funded by NARSAD, Alliant, and The Gheens Foundation.

Education is a major focus of the program and residents and medical students perform elective research projects. Graduate students in basic science departments have performed their research work in our laboratory. The Mood Disorders Research Program sponsors an annual September bipolar symposium for physician continuing medical education.

The Bipolar Clinic at the UofL HealthCare Outpatient Center provides evaluation and a wide array of psychiatric services to bipolar patients and their families.

University of Louisville Physicians Outpatient Center

401 East Chestnut Street, Suite 600 Louisville, KY 40202

Office Hours

M-F 8:00 am to 4:00 pm

No Holiday Hours

View contact information for the Department of Psychiatry and Behavioral Sciences .

Social Media

Patient Care & Health Information
Diseases & Conditions
Mood disorders

Mood disorders, such as depression and bipolar disorder, affect people emotionally. If you have depression, you may constantly feel sad. You also may be anxious. If you have bipolar disorder, you'll likely have extreme mood swings. Your feelings may range from being very sad, empty or cranky to being very happy — going back and forth between each mood. Mood disorders are more common in women.

Having a mood disorder may raise your risk of suicide. This risk is higher if the mood disorder is serious and you also have problems with alcohol or drugs.

If you're thinking about suicide, contact a hotline for help. In the U.S., call or text 988 to reach the 988 Suicide & Crisis Lifeline . It's available 24 hours a day, every day. Or use the Lifeline Chat . Services are free and private. The Suicide & Crisis Lifeline in the U.S. has a Spanish-language phone line at 1-888-628-9454 (toll-free).

Types of mood disorders

Mood disorders are divided into two major groups: depressive disorders and bipolar disorders. Each group includes several different types.

Depressive disorders

Depressive disorders cause loss of pleasure in most or all activities and ultimately affect your quality of life. You could have less energy, trouble sleeping, trouble concentrating, changes in appetite and lack interest. You also could have feelings of worthlessness or guilt and be in pain and tired.

Types of depressive disorders include:

Major depression — typically lasts for at least two weeks and often longer than four weeks.
Seasonal affective disorder — occurs at certain times of the year, typically with a change of season.
Persistent depressive disorder — a long-term form of depression that causes feelings of sadness, emptiness and often hopelessness.
Disruptive mood dysregulation disorder — a diagnosis used for children and teenagers. It features constant, serious and lasting testiness with frequent temper outbursts that are not consistent with the age of the child.
Premenstrual dysphoric disorder — features mood changes, hopelessness and feelings of being overwhelmed or out of control. These symptoms occur in the 10 days before a menstrual period and go away within a few days after a period begins.
Depression related to a medical condition — features a great loss of pleasure in most or all activities due to the physical effects of another medical health problem.
Depression related to substance or medicine use — features depression symptoms that start during or soon after using a street drug or medicine, or after withdrawal from these substances.

Bipolar disorders

Bipolar disorders feature mood swings that include emotional highs called manic or hypomanic episodes, and lows, called depressive episodes. These highs and lows are usually continuous. But they also can change from high to low or low to high — or shift into a normal mood. Sometimes both the highs and lows might occur together. This is called a mixed episode. You could be easily distracted and have racing thoughts. Your sleep also could be affected.

Types include:

Bipolar I disorder — features a constantly elevated mood that lasts for at least one week. This is called a manic episode. It affects your overall ability to function and makes it more likely that you'll take part in risky behavior.
Bipolar II disorder — features constantly elevated moods — called hypomanias — that last at least four days and less than one week. There may be risky behaviors, but usually hypomania does not greatly affect your ability to function. But other people should be able to notice that something is different about you.
Cyclothymia — features shifts from emotional highs to emotional lows that can affect your ability to function. The emotional ups and downs are not as extreme as those in bipolar I or II disorder.
Bipolar related to a medical condition — features symptoms that are the same as bipolar disorder, but they can be due to a medical condition. For example, Cushing's disease, multiple sclerosis, stroke and traumatic brain injury can cause bipolar mania or hypomania.
Bipolar related to the use of certain substances — features symptoms that are the same as bipolar disorder, but they can be due to alcohol, street drugs or medicine.

Products & Services

A Book: Mayo Clinic Family Health Book
Newsletter: Mayo Clinic Health Letter — Digital Edition

Symptoms depend on the type of mood disorder.

Depressive disorders are common and often long-lasting. They can:

Cause you to feel sad, empty, anxious and cranky.
Affect your ability to focus and function.
Cause loss of pleasure in most or all activities.
Affect your energy level and quality of life.
Make you feel worthless or guilty.
Affect how much you eat and sleep.
Raise thoughts about suicide.

Bipolar disorders may feature:

Moods that go back and forth between emotional highs, called mania or hypomania, and lows, called depression.
Feeling on top of the world, superior to others, or that you're so strong that nothing can harm or change you.
Racing thoughts.
Increased energy.
Disrupted sleep, usually a decreased need for sleep, but a continued high energy level.
Impulsive behaviors.

You also could be easily distracted and more likely to think about suicide or plan for suicide, depending on the seriousness of symptoms.

Other types of mood disorders may include other symptoms.

When to see a doctor

If you're concerned that you may have a mood disorder, see your doctor or a mental health professional as soon as you can. If you're not sure you want to seek treatment, talk to a friend or loved one, a faith leader, or someone else you trust.

Talk to a health care professional if you:

Feel like your emotions are getting in the way of work, how you get along with others or other areas of your life, or you're not taking part in social activities.
Have trouble with alcohol or drugs.
Are thinking about taking your own life. If this is the case, seek emergency treatment at once.

Your mood disorder is not likely to go away on its own. And it may get worse over time. Get professional help before your mood disorder becomes serious. It may be easier to treat early on.

There is a problem with information submitted for this request. Review/update the information highlighted below and resubmit the form.

From Mayo Clinic to your inbox

Sign up for free and stay up to date on research advancements, health tips, current health topics, and expertise on managing health. Click here for an email preview.

Error Email field is required

Error Include a valid email address

To provide you with the most relevant and helpful information, and understand which information is beneficial, we may combine your email and website usage information with other information we have about you. If you are a Mayo Clinic patient, this could include protected health information. If we combine this information with your protected health information, we will treat all of that information as protected health information and will only use or disclose that information as set forth in our notice of privacy practices. You may opt-out of email communications at any time by clicking on the unsubscribe link in the e-mail.

Thank you for subscribing!

You'll soon start receiving the latest Mayo Clinic health information you requested in your inbox.

Sorry something went wrong with your subscription

Please, try again in a couple of minutes

Mood disorders are caused by traits passed down to you, as well as environmental factors and life events. Environmental factors can include, for example, childhood experiences and stressful life events. Some prescription drugs, such as corticosteroids and medicines for Parkinson's disease, and street drugs also can cause mood disorders.

Risk factors

Risk factors include life experiences and stressful life events that increase the risk of certain types of mood disorders.

Mood disorders may occur along with neurological disorders. These are conditions that affect the brain and the nervous system. For example, depression is common among people with multiple sclerosis, dementia, traumatic brain injury, stroke and epilepsy. Depression also often occurs in people who have movements disorders like Parkinson's disease and those who have other long-term health conditions.

Mood disorders care at Mayo Clinic

Mood disorders fact sheet. National Institutes of Health. https://www.nimh.nih.gov/health/statistics/any-mood-disorder. Accessed March 15, 2023.
Allscripts EPSi. Mayo Clinic, Rochester, Minn.
Overview of mood disorders. Merck Manual Professional Version. https://www.merckmanuals.com/professional/psychiatric-disorders/mood-disorders/overview-of-mood-disorders. Accessed March 15, 2023.
Mood disorders. MentalHealth.gov. https://www.mentalhealth.gov/what-to-look-for/mood-disorders. Accessed March 15, 2023.
Personality disorders. In: Diagnostic and Statistical Manual of Mental Disorders DSM-5-TR. 5th ed. American Psychiatric Association; 2022; 10.1176/appi.books.9780890425787.x04_Depressive_Disorders.
Personality disorders. In: Diagnostic and Statistical Manual of Mental Disorders DSM-5-TR. 5th ed. American Psychiatric Association; 2022; https://dsm.psychiatryonline.org. Accessed March 15, 2023.
Kung S (expert opinion). Mayo Clinic. April 16, 2023.
Rakofsky J, et al. Mood disorders. Continuum: Lifelong Learning in Neurology. 2018; doi:10.1212/CON.0000000000000604.
Datta S, et al. Mood disorders. Continuum: Lifelong Learning in Neurology. 2021; doi:10.1212/CON.0000000000001051.
The lifeline and 988. 988 Suicide & Crisis Lifeline. https://988lifeline.org/current-events/the-lifeline-and-988/. Accessed March 16, 2023.
Galima SV, et al. Seasonal affective disorder: Common questions and answers. American Family Physician. https://www.aafp.org/pubs/afp/issues/2020/1201/p668.html. Accessed April 17, 2023.
Revadigar N, et al. Substance induced mood disorders. Stat Pearls. https://www.ncbi.nlm.nih.gov/books/NBK555887/. Accessed April 17, 2023.
Bipolar and related disorder due to another medical condition. PsychDB. https://www.psychdb.com/bipolar/z-bipolar-medical. Accessed April 18, 2023.
Substance/medication-induced bipolar and related disorder. PsychDB. https://www.psychdb.com/bipolar/substance-medication. Accessed April 18, 2023.
Symptoms & causes
Diagnosis & treatment
Doctors & departments
Care at Mayo Clinic

Mayo Clinic does not endorse companies or products. Advertising revenue supports our not-for-profit mission.

Opportunities

Mayo Clinic Press

Check out these best-sellers and special offers on books and newsletters from Mayo Clinic Press .

Mayo Clinic on Incontinence - Mayo Clinic Press Mayo Clinic on Incontinence
The Essential Diabetes Book - Mayo Clinic Press The Essential Diabetes Book
Mayo Clinic on Hearing and Balance - Mayo Clinic Press Mayo Clinic on Hearing and Balance
FREE Mayo Clinic Diet Assessment - Mayo Clinic Press FREE Mayo Clinic Diet Assessment
Mayo Clinic Health Letter - FREE book - Mayo Clinic Press Mayo Clinic Health Letter - FREE book

5X Challenge

Thanks to generous benefactors, your gift today can have 5X the impact to advance AI innovation at Mayo Clinic.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Publications
Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

Advanced Search
Journal List
HHS Author Manuscripts

The genetics of the mood disorder spectrum: genome-wide association analyses of over 185,000 cases and 439,000 controls

Jonathan r. i. coleman.

1. Social, Genetic and Developmental Psychiatry Centre, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, United Kingdom

2. NIHR Maudsley Biomedical Research Centre, King’s College London, London, United Kingdom

Héléna A. Gaspar

Julien bryois.

3. Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden

Bipolar Disorder Working Group of the Psychiatric Genomics Consortium

4. Full consortium authorship listed in Article Information

Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium

Gerome breen, associated data.

Mood disorders (including major depressive disorder and bipolar disorder) affect 10–20% of the population. They range from brief, mild episodes to severe, incapacitating conditions that markedly impact lives. Despite their diagnostic distinction, multiple approaches have shown considerable sharing of risk factors across the mood disorders.

To clarify their shared molecular genetic basis, and to highlight disorder-specific associations, we meta-analysed data from the latest Psychiatric Genomics Consortium (PGC) genome-wide association studies of major depression (including data from 23andMe) and bipolar disorder, and an additional major depressive disorder cohort from UK Biobank (total: 185,285 cases, 439,741 controls; non-overlapping N = 609,424).

Seventy-three loci reached genome-wide significance in the meta-analysis, including 15 that are novel for mood disorders. More genome-wide significant loci from the PGC analysis of major depression than bipolar disorder reached genome-wide significance. Genetic correlations revealed that type 2 bipolar disorder correlates strongly with recurrent and single episode major depressive disorder. Systems biology analyses highlight both similarities and differences between the mood disorders, particularly in the mouse brain cell types implicated by the expression patterns of associated genes. The mood disorders also differ in their genetic correlation with educational attainment – positive in bipolar disorder but negative in major depressive disorder.

Conclusions

The mood disorders share several genetic associations, and can be combined effectively to increase variant discovery. However, we demonstrate several differences between these disorders. Analysing subtypes of major depressive disorder and bipolar disorder provides evidence for a genetic mood disorders spectrum.

Introduction

Mood disorders affect 10–20% of the global population across their lifetime, ranging from brief, mild episodes to severe, incapacitating conditions that markedly impact lives ( 1 – 4 ). Major depressive disorder and bipolar disorder are the most common forms and have been grouped together since the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III) ( 5 ). Although given dedicated chapters in DSM5, they remain grouped as mood disorders in the International Classification of Disorders (ICD11) ( 6 , 7 ).

Depressive episodes are common to major depressive disorder and type 2 bipolar disorder, and are usually present in type 1 bipolar disorder ( 7 ). The bipolar disorders are distinguished from major depressive disorder by the presence of mania in type 1 and hypomania in type 2 ( 7 ). However, these distinctions are not absolute – some individuals with major depressive disorder may later develop bipolar disorder, and some endorse (hypo)manic symptoms ( 8 – 10 ). Following their first depressive episode, a non-remitting individual might develop recurrent major depressive disorder or bipolar disorder. Treatment guidelines for these disorders differ ( 11 , 12 ). Identifying shared and distinct genetic associations for major depressive disorder and bipolar disorder could aid our understanding of these diagnostic trajectories.

Twin studies suggest that 35–45% of variance in risk for major depressive disorder and 65–70% of the variance in bipolar disorder risk is accounted for by additive genetic factors ( 13 ). These genetic components are partially shared, with a twin genetic correlation (r g ) of ~65%, and common variant based r g derived from the results of genome-wide association studies (GWAS) of 30–35% ( 14 – 17 ). Considerable progress has been made in identifying specific genetic variants that underlie genetic risk. Recently, the Psychiatric Genomics Consortium (PGC) published a GWAS of bipolar disorder, including over 20,000 cases, with 30 genomic loci reaching genome-wide significance ( 16 ). They also performed a GWAS of major depression, including over 135,000 individuals with major depressive disorder and other definitions of depression, with 44 loci reaching genome-wide significance ( 15 ). The PGC GWAS of major depression has since been combined with a broad depression GWAS ( Supplementary Note ).

GWAS have identified statistical associations with major depressive disorder and with bipolar disorder individually, but have not explored the genetic aspects of the relationship between these disorders. In addition, both major depressive disorder and bipolar disorder exhibit considerable clinical heterogeneity and can be separated into subtypes. For example, the DSM5 includes categories for bipolar disorder type 1 and type 2, and for single episode and recurrent major depressive disorder ( 7 ). We use the PGC analyses of major depression and bipolar disorder, along with analyses of formally-defined major depressive disorder from UK Biobank, to explore two aims ( 18 , 19 ). Firstly, we seek to identify shared and distinct mood disorder genetics by combining studies of major depressive disorder and bipolar disorder. We then explore the genetic relationship of mood disorders to traits from the wider GWAS literature. Secondly, we assess the overall genetic similarities and differences of bipolar disorder subtypes (from the PGC) and major depressive disorder subtypes (from UK Biobank), through comparing their genetic correlations and polygenic risk scores from GWAS.

Materials and Methods

Participants.

Our primary aim was to combine analyses of bipolar disorder and major depression to examine the shared and distinct genetics of these disorders. Summary statistics were derived from participants of Western European ancestries. Full descriptions of each study and their composite cohorts are provided in each paper ( 15 , 16 , 19 ). Brief descriptions are provided in the Supplementary Methods . Except where otherwise specified, summary statistics are available (or will be made available) at https://www.med.unc.edu/pgc/results-and-downloads .

Major depression data were drawn from the full cohort (PGC MDD: 135,458 cases, 344,901 controls) from ( 15 ). This included data from 23andMe ( 20 ), access to which requires a Data Transfer Agreement; consequently, the data analysed here differ from the summary statistics available at the link above. Data for bipolar disorder were drawn from the discovery analysis previously reported (PGC BD: 20,352 cases, 31,358 controls), not including replication results ( 16 ).

Secondly, we wished to examine genetic correlations between mood disorder subtypes. Summary statistics were available for the primary bipolar disorder subtypes, type 1 bipolar disorder (BD1: 14,879 cases, 30,992 controls) and type 2 bipolar disorder (BD2: 3,421 cases, 22,155 controls), and for schizoaffective bipolar disorder (SAB: 977 cases, 8,690 controls), a mood disorder including psychotic symptoms. Controls are shared across these subtype analyses.

Subtype GWAS are not yet available from PGC MDD. As such, a major depressive disorder cohort was derived from the online mental health questionnaire in the UK Biobank (UKB MDD: 29,475 cases, 63,482 controls; Resource 22 on http://biobank.ctsu.ox.ac.uk ) ( 18 ). The definition of major depressive disorder in this cohort is based on DSM-5, as described in full elsewhere ( 18 ), and in Supplementary Table 1 ( 7 ). We defined three major depressive disorder subtypes for analysis. Individuals meeting criteria for major depressive disorder were classed as recurrent cases if they reported multiple depressed periods across their lifetime (rMDD, N = 17,451), and single-episode cases otherwise (sMDD, N = 12,024, Supplementary Table 1 ). Individuals reporting depressive symptoms, but not meeting case criteria, were excluded from the main analysis but used as a “sub-threshold depression” subtype to examine the continuity of genetic associations with major depressive disorder below clinical thresholds (subMDD, N = 21,596). All subtypes were analysed with the full set of controls. Details on the quality control and analysis of the UK Biobank phenotypes is provided in the Supplementary Methods .

Meta-analysis of GWAS data

We meta-analysed PGC MDD and UKB MDD to obtain a single major depressive disorder GWAS (combined MDD). We meta-analysed combined MDD with PGC BD, comparing mood disorder cases to controls (MOOD). Further meta-analyses were performed between PGC MDD and each bipolar disorder subtype and major depressive disorder subtype to assess the relative increase in variant discovery when adding different mood disorder definitions to PGC MDD ( Supplementary Results ).

Summary statistics were limited to common variants (MAF > 0.05; Supplementary Methods ) either genotyped or imputed with high confidence (INFO score > 0.6) in all studies. Controls were shared between PGC MDD and PGC BD, and (due to the inclusion of summary data in PGC MDD) the extent of this overlap was unknown. Meta-analyses were therefore performed in METACARPA, which controls for sample overlap of unknown extent between studies using the variance-covariance matrix of the observed effect sizes at each variant, weighted by the sample sizes ( 21 , 22 ). METACARPA adjusted adequately for known overlap between cohorts ( Supplementary Methods ). For later analyses (particularly linkage disequilibrium score regression) we used as the sample size a “non-overlapping N” estimated for each meta-analysis ( Supplementary Methods ). The definition, annotation and visualisation of each meta-analysis is described in the Supplementary Materials .

Sensitivity analysis using down-sampled PGC MDD

Results from MOOD showed greater similarity to PGC MDD than to PGC BD. Cross-trait meta-analyses may be biased if the power of the composite analyses differs substantially ( 23 , 24 ). The mean chi-square of combined MDD [1.7] exceeded that of PGC BD [1.39], suggesting this bias may affect our results ( Supplementary Table 2 ). We therefore repeated our analyses, meta-analysing UKB MDD with summary statistics for PGC MDD that did not include participants from 23andMe nor the UK Biobank (mean chi-square = 1.35). All analyses were performed on the full and the down-sampled analyses, with the exception of GSMR analyses. Full results of the down-sampled analyses are described in the Supplementary Materials .

Estimation of SNP-based heritability captured by common variants and genetic correlations with published GWAS

The SNP-based heritability captured by common variants was assessed using linkage disequilibrium score regression (LDSC) for each meta-analysed set of data ( 25 ). SNP-based heritability estimates were transformed to the liability scale, assuming population prevalences of 15% for combined MDD, 1% for PGC BD, and 16% for MOOD, and lower and upper bounds of these prevalences for comparison ( Supplementary Methods ). LDSC separates genome-wide inflation into components due to polygenicity and confounding ( 25 ). Inflation not due to polygenicity was quantified as (intercept-1)/(mean observed chi-square-1) ( 26 ). Genetic correlations were calculated in LDSC between each analysis and 414 traits curated from published GWAS. Local estimates of SNP-based heritability and genetic covariance were obtained using HESS v0.5.3b ( Supplementary Methods and Results ) ( 27 , 28 ).

Genetic correlations between subtype analyses

To assess the structure of genetic correlations within the mood disorders, SNP-based heritabilities and genetic correlations were calculated in LDSC between bipolar disorder subtypes (BD1, BD2, SAB), and major depressive disorder subtypes (rMDD, sMDD, subMDD). Putative differences between genetic correlations were identified using a z-test (p < 0.05), and formally tested by applying a block-jackknife, with Bonferroni correction for significance (p < 8.3×10 −4 ; Supplementary Methods ). Differences between the genetic correlations of PGC MDD and each bipolar disorder subtype, and between PGC BD and each major depressive disorder subtype were also tested (Bonferroni correction for significance, p < 0.0083). Genetic correlations were hierarchically clustered using the gplots package in R v1.4.1 ( 29 , 30 ). Hierarchical clustering was performed using just the subtypes, and including results from six external GWAS relevant to mood disorders ( Supplementary Methods ). To validate our conclusion of a genetic mood disorder spectrum, we performed principal component analysis of the genetic correlation matrix including the six external GWAS ( Supplementary Methods and Results ).

Association of PGC BD polygenic risk scores with major depressive disorder subtypes

Polygenic risk score analyses were performed using PRSice2 to assess whether rMDD was genetically more similar to PGC BD than were sMDD or subMDD ( Supplementary Methods ) ( 36 ).

Gene-wise, gene-set, and tissue and single-cell enrichment analyses

For all analyses, gene-wise p-values were calculated as the aggregate of the mean and smallest p-value of SNPs annotated to Ensembl gene locations using MAGMA v1.06 ( Supplementary Methods and Results ) ( 37 ). Gene set analysis was performed in MAGMA ( Supplementary Methods and Results ). Further analyses were performed to assess the enrichment of associated genes with expression-specificity profiles from tissues (Genotype-Tissue Expression project, version 7) and broadly-defined (“level 1”) and narrowly-defined (“level 2”) mouse brain cell-types ( 38 , 39 ). Analyses were performed in MAGMA following previously described methods with minor modifications, with Bonferroni-correction for significance ( Supplementary Methods ) ( 38 ). Similar analyses can be performed in LDSC-SEG – we report MAGMA results, which reflect specificity of expression across the range, whereas LDSC-SEG compares the top 10% of the range with the remainder ( 40 ). Results using LDSC are included in the Supplementary Tables .

Mendelian randomisation (GSMR)

Bidirectional Mendelian randomisation analyses were performed using the GSMR option in GCTA to allow exploratory inference of the causal direction of known relationships between mood disorder traits and other traits ( 41 , 42 ). Specifically, the relationship between the mood disorder analyses (MOOD, combined MDD, PGC BD) and schizophrenia, intelligence, educational attainment, body mass index, and coronary artery disease were explored ( Supplementary Methods ) ( 32 , 43 – 46 ). These traits were previously examined in the PGC major depression GWAS – we additionally tested intelligence following the results of our genetic correlation analyses ( 15 ).

Conditional and reversed-effect analyses

Additional analyses were performed to identify shared and distinct mood disorder loci, using mtCOJO, an extension of GSMR ( Supplementary Methods ) ( 41 , 42 ). Analyses were performed on combined MDD conditional on PGC BD, and on PGC BD conditional on combined MDD ( Supplementary Results ). To identify loci with opposite directions of effect between combined MDD and PGC BD, the MOOD meta-analysis was repeated with reversed direction of effects for PGC BD ( Supplementary Methods and Results ).

Evidence for confounding in meta-analyses

Meta-analysis results were assessed for genome-wide inflation of test statistics using LDSC ( 25 ). The LDSC intercept was significantly >1 in most cases (1.00–1.06), which has previously been interpreted as confounding ( Supplementary Table 2 ). However, such inflation can occur in large cohorts without confounding ( 47 ). Estimates of inflation not due to polygenicity were small in all meta-analyses (4–7%, Supplementary Table 2 ).

Combined MOOD meta-analysis

We meta-analysed the PGC MDD, PGC BD and the UKB MDD cohorts (MOOD, cases = 185,285, controls = 439,741, non-overlapping N = 609,424). 73 loci reached genome-wide significance, of which 55 were also seen in the meta-analysis of PGC MDD and UKB MDD (combined MDD, Supplementary Table 3 , Supplementary Figures 1 and 2 ). Results are summarised in Table 1 : 39 of the 44 PGC MDD loci reached genome-wide significance in MOOD ( Supplementary Table 3 , Supplementary Figures 1 – 8 ). In comparison, only four of the 19 PGC BD loci reached genome-wide significance in MOOD. MOOD loci overlapped considerably with previous studies of depression and depressive symptoms (51/73) ( 20 , 23 , 48 – 52 ), bipolar disorder (3/73) ( 53 – 56 ), neuroticism (32/73) ( 23 , 57 – 59 ), and schizophrenia (15/73) ( 32 , 60 ), although participants overlap between MOOD and many of these studies. Locus 52 (chromosome 12) passed genome-wide significance in a previous meta-analysis of broad depression and bipolar disorder, although the two other loci from this study did not replicate ( 51 ). Six of the 73 associations are entirely novel (p > 5×10 −8 in previous studies of all phenotypes; Table 1 , Supplementary Table 4 ).

Loci genome-wide significant (p < 5×10 −8 ) in the MOOD meta-analysis.

Locus	Chr	BP	Index SNP	A1	A2	OR	SE	p	Previous report
1	1	37192741	rs1002656	T	C	0.97	0.005	2.71×10	DO, N
2	1	72837239	rs7531118	T	C	0.96	0.004	1.05×10	D, DO, S, O
4	1	80795989	rs6667297	A	G	0.97	0.005	5.86×10	D, DO
5	1	90796053	rs4261101	A	G	0.97	0.005	1.78×10	D
6	1	175913828	rs10913112	T	C	0.97	0.005	1.46×10	DO, O
7	1	177370033	rs16851203	T	C	0.96	0.007	2.38×10	DO, S, O
9	2	22582968	rs61533748	T	C	0.97	0.004	3.84×10	DO, N
10	2	57987593	rs11682175	T	C	0.97	0.004	2.18×10	D, DO, BS, N, S, O
11	2	157111313	rs1226412	T	C	1.03	0.005	1.27×10	D, DO, N, O
12	2	198807015	rs1518367	A	T	0.97	0.005	1.18×10	BS, S, O
13	3	108148557	rs1531188	T	C	0.96	0.006	1.61×10	O
14	3	158107180	rs7430565	A	G	0.97	0.004	2.30×10	D, DO, N, O
16	4	42047778	rs34215985	C	G	0.97	0.006	1.72×10	D, DO, N
17	5	77709430	rs4529173	T	C	0.97	0.005	4.29×10	O
18	5	88002653	rs447801	T	C	1.03	0.004	2.29×10	D, DO, N, O
19	5	92995013	rs71639293	A	G	1.03	0.005	5.85×10	DO, N
20	5	103904226	rs12658032	A	G	1.04	0.005	2.19×10	D, DO, N, O
21	5	106603482	rs55993664	A	C	0.97	0.006	1.87×10	NOVEL LOCUS
22	5	124251883	rs116755193	T	C	0.97	0.005	1.47×10	D, O
23	5	164523472	rs11135349	A	C	0.97	0.004	2.96×10	D, DO, N
24	5	166992078	rs4869056	A	G	0.97	0.005	5.21×10	D
25	6	28673998	rs145410455	A	G	0.94	0.007	7.17×10	D, DO, BO, BS, DS, N, S, O
26	6	101339400	rs7771570	T	C	0.97	0.004	9.68×10	DO, N, O
27	6	105365891	rs1933802	C	G	0.98	0.004	1.05×10	DO, S, O
28	7	12267221	rs4721057	A	G	0.97	0.004	7.31×10	D, DO, N, O
29	7	24826589	rs79879286	C	G	1.04	0.006	1.97×10	B, BS, DO, S
30	7	82514089	rs34866621	T	C	1.03	0.005	2.21×10	DO, O
31	7	109099919	rs58104186	A	G	1.03	0.004	7.12×10	D, DO
34	9	11379630	rs10959753	T	C	0.96	0.005	1.45×10	D, DO, N, O
35	9	37207269	rs4526442	T	C	0.96	0.006	7.97×10	DO, O
36	9	81413414	rs11137850	A	G	1.03	0.005	1.25×10	NOVEL LOCUS
38	9	119733380	rs10759881	A	C	1.03	0.005	8.56×10	D, DO
40	9	122664468	rs10818400	T	G	0.98	0.004	1.29×10	N
41	9	126682068	rs7029033	T	C	1.04	0.008	2.61×10	D, DO, O
42	10	104684544	rs78821730	A	G	0.96	0.007	2.95×10	N, BS, S, O
43	10	106563924	rs61867293	T	C	0.96	0.005	5.64×10	D, DO, N, O
44	11	16293680	rs977509	T	C	0.97	0.005	1.19×10	DO, N, O
45	11	31850105	rs1806153	T	G	1.03	0.005	2.81×10	D, DO, N, O
46	11	32765866	rs143864773	T	C	1.04	0.008	1.70×10	NOVEL LOCUS
47	11	61557803	rs102275	T	C	0.97	0.005	5.04×10	B, DO, BO, O
48	11	63632673	rs10792422	T	G	0.98	0.004	2.18×10	O
49	11	88743208	rs4753209	A	T	0.97	0.004	4.15×10	DO, N, O
50	11	99268617	rs1504721	A	C	0.98	0.004	2.24×10	O
51	11	113392994	rs2514218	T	C	0.97	0.005	3.22×10	DO, BS, N, S, O
52	12	2344644	rs769087	A	G	1.03	0.005	3.27×10	B, BD, BO, DS, BS, S, O
53	12	23947737	rs4074723	A	C	0.97	0.004	3.18×10	D, DO, N, O
54	12	121186246	rs58235352	A	G	0.95	0.009	1.64×10	DO, O
55	12	121907336	rs7962128	A	G	1.02	0.004	3.63×10	NOVEL LOCUS
56	13	44327799	rs4143229	A	C	0.95	0.008	2.73×10	D
57	13	53625781	rs12552	A	G	1.04	0.004	1.25×10	D, DO, O
58	14	42074726	rs61990288	A	G	0.97	0.004	2.29×10	D, DO, O
60	14	64686207	rs915057	A	G	0.98	0.004	1.92×10	D, DO, O
61	14	75130235	rs1045430	T	G	0.97	0.004	9.83×10	D, DO, N, O
62	14	104017953	rs10149470	A	G	0.97	0.004	1.15×10	D, DS, DO, BS, S, O
63	15	36355868	rs1828385	A	C	0.97	0.004	1.15×10	NOVEL LOCUS
64	15	37643831	rs8037355	T	C	0.97	0.004	4.09×10	D, DO, O
65	16	6310645	rs8063603	A	G	0.97	0.005	5.36×10	D, DO
66	16	7667332	rs11077206	C	G	1.03	0.004	5.49×10	D, DO, N, O
67	16	13038723	rs12935276	T	G	0.97	0.005	4.75×10	D, DO, N, O
68	16	13750257	rs7403810	T	G	1.03	0.005	7.52×10	DO, BS, S, O
69	16	72214276	rs11643192	A	C	1.03	0.004	1.46×10	D, O
70	17	27363750	rs75581564	A	G	1.04	0.006	2.47×10	D, DO, O
71	18	31349072	rs4534926	C	G	1.03	0.004	9.14×10	DO, N
72	18	36883737	rs62099069	A	T	0.97	0.004	9.52×10	D, O
73	18	42260348	rs117763335	T	C	0.97	0.005	1.33×10	O
74	18	50614732	rs11663393	A	G	1.03	0.004	1.56×10	D, DO, N, O
75	18	52517906	rs1833288	A	G	1.03	0.005	4.54×10	D, DS, DO, N, S, O
76	18	53101598	rs12958048	A	G	1.04	0.005	4.86×10	D, DO, BS, N, S, O
77	19	30939989	rs33431	T	C	1.02	0.004	4.04×10	DO, O
78	20	45841052	rs910187	A	G	0.97	0.005	3.09×10	DO, O
79	22	41621714	rs2179744	A	G	1.03	0.005	3.83×10	D, B, DO, BS, N, S, O
80	22	42815358	rs7288411	A	G	1.03	0.005	3.86×10	NOVEL LOCUS
81	22	50679436	rs113872034	A	G	0.96	0.006	1.10×10	O

Locus – shared locus number for annotation ( Supplementary Table 3 ), Chr – chromosome, BP – base position, A1 – effect allele, A2 – non-effect allele, Previous report – locus previously implicated in PGC MDD (D), PGC BD (B), previous combined studies of bipolar disorder and major depressive disorder (BD), other studies of major depressive disorder or depressive symptoms (DO), other studies of bipolar disorder (BO), previous combined studies of bipolar disorder and schizophrenia (BS), previous combined studies of major depressive disorder and schizophrenia (DS), neuroticism (N), schizophrenia (S), or other studies (O – see Supplementary Table 4 ).

The down-sampled MOOD (cases = 95,481, controls = 287,932, non-overlapping N = 280,214) showed increased similarity to PGC BD compared to MOOD, but remained more similar to PGC MDD. Nineteen loci reached genome-wide significance in down-sampled MOOD, including nine (20%) from PGC MDD, compared with two (11%) reported in PGC BD ( Supplementary Table 3 ). 17/19 loci were also observed in MOOD. Of the two loci not observed in MOOD, one passed genome-wide significance in PGC BD.

SNP-based heritability and genetic correlations

The estimate of SNP-based heritability for MOOD (8.8%) was closer to PGC MDD (9%) than to PGC BD (17–23%) ( 15 , 16 ). Significant genetic correlations between MOOD and other traits included psychiatric and behavioural, reproductive, cardiometabolic, and sociodemographic traits ( Figure 1 , Supplementary Table 5 ). Genetic correlations with psychiatric and behavioural traits are consistently observed across psychiatric traits ( 17 , 61 ). The genetic correlation with educational attainment differs, being negative in combined MDD, but positive in PGC BD ( Supplementary Table 6 ). The genetic correlation (r g ) between MOOD and educational attainment was −0.058 (p=0.004), intermediate between the results of combined MDD and of PGC BD. Notably, the genetic correlation with intelligence (IQ) was not significant in combined MDD, PGC BD, nor MOOD (p>1.27×10 −4 ). However, sensitivity analyses (see below), indicated that including 23andMe in the PGC MDD sample obscured a negative genetic correlation of MDD with IQ.

An external file that holds a picture, illustration, etc.
Object name is nihms-1669830-f0001.jpg

Selected genetic correlations of psychiatric traits with the main meta-analysis (MOOD), the separate mood disorder analyses (combined MDD and PGC BD), and the down-sampled analyses (down-sampled MOOD, down-sampled MDD). Full genetic correlation results are provided in Supplementary Table 5 .

The SNP-based heritability of down-sampled MOOD from LDSC was 11%, closer to PGC MDD than to PGC BD ( Supplementary Table 2 ). Genetic correlations varied ( Supplementary Tables 5 and 7 ) with some more similar to PGC BD (schizophrenia: down-sampled rg = 0.61, combined MDD rg = 0.35, PGC BD rg = 0.7), and others more similar to combined MDD (ADHD: down-sampled rg = 0.48, combined MDD rg = 0.45, PGC BD rg = 0.14). The genetic correlation with IQ was significant (rg = −0.13, p = 5×10 −7 ), because the excluded 23andMe depression cohort has a positive genetic correlation with IQ (rg = 0.06, p = 0.01). The greater genetic correlation of MOOD with combined MDD (0.98) compared to PGC BD (0.55) persisted when comparing down-sampled MOOD to combined MDD (0.85) and PGC BD (0.75; Supplementary Table 6 ).

Relationship between mood disorder subtypes

Analyses were performed using GWAS data from subtypes of bipolar disorder (BD1, BD2, SAB) and major depressive disorder (rMDD, sMDD, subMDD). SNP-based heritability for the subtypes ranged from subMDD and sMDD (8%), through BD2 and rMDD (10% and 12%, respectively) to BD1 and SAB (22% and 29% respectively, Figure 2 , Supplementary Table 2 ).

An external file that holds a picture, illustration, etc.
Object name is nihms-1669830-f0003.jpg

SNP-based heritability estimates for the subtypes of bipolar disorder and subtypes of major depressive disorder. Points = SNP-based heritability estimates. Lines = 95% confidence intervals. Full SNP-based heritability results are provided in Supplementary Table 2 .

The major depressive disorder subtypes were strongly and significantly genetically correlated (r g = 0.9–0.94, p rg = 0 < 8.3×10 −4 ). These correlations did not differ significantly from 1 (all p rg = 1 > 0.3), nor from each other (all p Δrg = 0 > 0.5, Figure 2 , Supplementary Table 8 ). BD1 and SAB were strongly correlated (r g = 0.77, p rg = 0 = 6×10 −13 , p rg = 1 = 0.03), as were BD1 and BD2 (r g = 0.86, p rg = 0 = 3×10 −16 , p rg = 1 = 0.2). However, BD2 was not significantly correlated with SAB (r g = 0.22, p rg = 0 = 0.02).

In hierarchical clustering, BD2 clustered with the major depressive disorder subtypes rather than the bipolar disorder subtypes. The strength of correlation between BD2 and BD1 did not differ from that between BD2 and rMDD (r g = 0.68, p rg = 0 = 3×10 −8 , p rg = 1 = 0.01), following multiple testing correction (Δr g = 0.18, p = 0.02). Overall, these results suggest a spectrum of genetic relationships between major depressive disorder and bipolar disorder, with BD2 bridging the two disorders ( Figure 3 ; Supplementary Figure 9 ). This spectrum remained when six external phenotypes were added, and was supported by results from principal component analysis ( Supplementary Results , Supplementary Figure 10 ).

An external file that holds a picture, illustration, etc.
Object name is nihms-1669830-f0004.jpg

Genetic correlations across the mood disorder spectrum. Labelled arrows show genetic correlations significantly different from 0. Solid arrows represent genetic correlations not significantly different from 1 (p < 0.00333, Bonferroni correction for 15 tests). Full results are provided in Supplementary Table 8 .

Polygenic risk score analyses showed that individuals with high polygenic risk scores for PGC BD were more likely to report rMDD than sMDD, and more likely to report sMDD than subMDD ( Supplementary Results ).

Tissue and cell-type specificity analyses

The results of gene-wise and gene set analyses are described in the Supplementary Results . The tissue-specificity of associated genes differed minimally between the analyses ( Supplementary Table 9 ). All brain regions were significantly enriched in all analyses, and the pituitary was also enriched in combined MDD and PGC BD (p < 9.43×10 −4 , Bonferroni correction for 53 regions, Supplementary Table 9 ). Results from down-sampled MOOD and down-sampled MDD were generally consistent with the main analyses, except spinal cord was not enriched in either, nor was the cordate in the down-sampled MDD analysis.

In contrast, cell-type enrichments differed between combined MDD and PGC BD ( Figure 4 , Supplementary Tables 10 and 11 ). Genes associated with PGC BD were enriched for expression in pyramidal cells from the CA1 region of the hippocampus and the somatosensory cortex, and in striatal interneurons. None of these enrichments were significant in combined MDD. Genes only associated with combined MDD were significantly enriched for expression in neuroblasts and dopaminergic neurons from adult mice. Further cell-types (dopaminergic neuroblasts; dopaminergic, GABAergic and midbrain nucleus neurons from embryonic mice; interneurons; and medium spiny neurons) were enriched for both combined MDD and PGC BD, but the rank and strength of enrichment differed, most notably for medium spiny neurons. The general pattern of differences persisted when comparing PGC BD with down-sampled MDD, although genes associated with down-sampled MDD were not enriched for expression in adult dopaminergic neurons, embryonic midbrain nucleus neurons, interneurons, nor medium spiny neurons ( Supplementary Figure 11 ).

An external file that holds a picture, illustration, etc.
Object name is nihms-1669830-f0005.jpg

Cell-type expression specificity of genes associated with bipolar disorder (PGC BIP, left) and major depressive disorder (combined MDD, right). Black vertical lines = significant enrichment (p < 2×10 −3 , Bonferroni correction for 24 cell types). See Supplementary Table 10 for full results.

Shared and distinct relationships with mood disorders and inferred causality

Bidirectional Mendelian randomisation was used to investigate previously-described relationships between mood disorder phenotypes (combined MDD, PGC BD) and external traits: schizophrenia, educational attainment, IQ, body mass index (BMI) and coronary artery disease (CAD; Figure 5 , Supplementary Table 12 ). Associations with PGC BD should be interpreted cautiously, as only 19 loci reached genome-wide significance, several of which were removed as potentially pleiotropic in the analyses below.

An external file that holds a picture, illustration, etc.
Object name is nihms-1669830-f0006.jpg

GSMR results from analyses with the main meta-analysis (MOOD), and the major depression and bipolar disorder analyses (combined MDD, PGC BD). External traits are coronary artery disease (CAD), educational attainment (EDU), body mass index (BMI), and schizophrenia (SCZ). Betas are on the scale of the outcome GWAS (logit for binary traits, phenotype scale for continuous). * p < 0.004 (Bonferroni correction for two-way comparisons with six external traits). For figure data, including the number of non-pleiotropic SNPs included in each instrument, see Supplementary Table 12 .

A positive bidirectional relationship was observed between combined MDD and PGC BD, and between schizophrenia and both combined MDD and PGC BD. This is consistent with psychiatric disorders acting as causal risk factors for the development of further psychiatric disorders, or being correlated with other causal risk factors, including (but not limited to) the observed shared genetic basis.

The relationship with educational years differed between the mood disorders – there was a negative bidirectional relationship between educational years and combined MDD, but a positive bidirectional relationship with PGC BD (albeit with only nominal significance from PGC BD to educational years). In contrast, no significant relationship was observed between mood phenotypes and IQ. This is consistent with differing causal roles of education (or correlates of education) on the mood disorders, with a weaker reciprocal effect of the mood disorders altering the length of education.

A positive association was seen from BMI to combined MDD, but not from combined MDD to BMI. In contrast, only a nominally significant negative relationship was seen from PGC BD to BMI. A positive association was observed from combined MDD to CAD; no relationship was observed between CAD and PGC BD.

We identified 73 genetic loci by meta-analysing cohorts of major depression and bipolar disorder, including 15 loci novel to mood disorders. Our overall mood disorders meta-analysis results (MOOD) have more in common with our major depressive disorder analysis (combined MDD) than our bipolar disorder analysis (PGC BD). Partly, this results from the greater power of the major depressive disorder analysis compared to the bipolar disorder analysis. Nevertheless, genetic associations from our sensitivity analysis with equivalently powered cohorts (using down-sampled MDD in place of combined MDD) still showed a greater overall similarity to those from major depressive disorder rather than bipolar disorder.

This may reflect a complex genetic architecture in bipolar disorder, wherein one set of variants may be associated more with manic symptoms and another set with depressive symptoms. Variants associated more with mania (or psychosis) may have higher effect sizes, detectable at current bipolar disorder GWAS sample sizes, and may not be strongly associated with major depressive disorder. This could contribute to the observed higher heritability of bipolar disorder compared to major depressive disorder, and agrees with reports that most of the genetic variance for mania is not shared with depression ( 13 , 14 ). In this case, meta-analysis of bipolar disorder and major depressive disorder cohorts would support variants associated more with depression, but not those associated more with mania. This is consistent with our findings, and with depressive symptoms being both the unifying feature of the mood disorders and the core feature of major depressive disorder.

We assessed genetic correlations between mood disorder subtypes. We observed high, consistent correlations between major depressive disorder subtypes, including sub-threshold depression. Bipolar disorder type 2 showed greater genetic similarity to major depressive disorder compared to type 1. In this, we build on similar findings from polygenic risk scores analyses ( 16 , 56 ). Individuals with high polygenic risk scores for PGC BD were more likely to report recurrent than single-episode major depressive disorder. However, the genetic correlation of PGC BD with recurrent major depressive disorder was not significantly greater than that with single-episode major depressive disorder. This might reflect the difference in power between these methods. We also examined the genetic correlations between mood disorder subtypes in the context of relevant external traits ( Supplementary Results ). Our subtype analyses support a genetic mood spectrum consisting of the schizophrenia-like bipolar disorder type 1 and schizoaffective disorder at one pole, and the depressive disorders at the other, with bipolar disorder type 2 occupying an intermediate position.

Conditional and reversed-effect analyses ( Supplementary Results ) suggest that few of the loci we identified are disorder-specific. However, our results highlight some differences between the genetics of the mood disorders. The expression specificity of associated genes in mouse brain cell types differed between bipolar disorder and major depressive disorder analyses. Cell-types more associated with bipolar disorder (pyramidal neurons and striatal interneurons) were also enriched in analyses of schizophrenia ( 38 ). Cell-types more associated in major depressive disorder (neuroblasts, adult dopaminergic neurons, embryonic GABAergic neurons) had weaker enrichments in schizophrenia, but were enriched in analyses of neuroticism ( 57 ). The higher rank of the enrichment of serotonergic neurons with major depressive disorder compared to bipolar disorder is striking given the use of drugs targeting the serotonergic system in the treatment of depression ( 63 ). Nevertheless, cell-type enrichment analyses are still novel, and require cautious interpretation, especially given the use of non-human reference data ( 38 , 64 ).

We explored potential causal relationships between the mood disorders and other traits using Mendelian randomisation. The interpretation of these analyses is challenging, especially for complex traits, when the ascertainment of cases varies, and when there are relatively few (< 20) variants used as instruments (for example, in the PGC BD and down-sampled analyses presented) ( 41 , 67 , 68 ). Major depressive disorder and bipolar disorder demonstrate considerable heterogeneity (as our subtype analyses show for bipolar disorder types 1 and 2), potentially confounding the results of Mendelian randomisation. That said, our analyses are consistent with a bidirectional influence of educational attainment on risk for mood disorders (and vice versa), with different directions of effect in the two mood disorders. We found no significant relationship between IQ and either mood disorder. We also find results consistent with major depressive disorder increasing the risk for coronary artery disease in a relatively well powered analysis. This mirrors epidemiological findings, although the mechanism remains unclear ( 69 ).

Despite the presence of depressive episodes, the mood disorders are diagnostically distinct. This is reflected in their differing epidemiology – for example, more women than men suffer from major depressive disorder, whereas diagnoses of bipolar disorder are roughly equal between the sexes ( 3 ). Differences in our genetic results between major depressive disorder and bipolar disorder may result from epidemiological heterogeneity, rather than distinct biological mechanisms ( 70 ). Deeper phenotyping of GWAS datasets is ongoing, and will enable the effect of confounding factors such as sex to be estimated in future studies ( 71 ).

We extend previous findings showing genetic continuity across the mood disorders ( 15 – 17 , 56 ). Combined analyses of major depressive disorder and bipolar disorder may increase variant discovery, as well as the discovery of shared and distinct neurobiological gene sets and cell types. Our results also indicate some genetic differences between major depressive disorder and bipolar disorder, including opposite bidirectional relationships of each mood disorder with educational attainment, a possible influence of major depressive disorder on coronary artery disease risk and differing mouse brain cell types implicated by the enrichment patterns of associated genes in each disorder. Finally, our data are consistent with the existence of a genetic mood disorder spectrum with separate clusters for bipolar disorder type 1 and depressive disorders, linked by bipolar disorder type 2, and with depression as the common symptom. The mood disorders have a partially genetic aetiology that is partly shared. The identification of specific sets of genetic variants differentially associated with depression and with mania remains an aim for future research.

An external file that holds a picture, illustration, etc.
Object name is nihms-1669830-f0002.jpg

Selected genetic correlations of other traits with the main meta-analysis (MOOD), the separate mood disorder analyses (combined MDD and PGC BD), and the down-sampled analyses (down-sampled MOOD, down-sampled MDD). Full genetic correlation results are provided in Supplementary Table 5 .

Supplementary Material

Supplementary text, supplementary tables, acknowledgements.

This paper has previously been made available as a preprint on bioRxiv at https://www.biorxiv.org/content/10.1101/383331v1 .

We are deeply indebted to the investigators who comprise the PGC, and to the hundreds of thousands of subjects who have shared their life experiences with PGC investigators. This study represents independent research partly funded by the National Institute for Health Research (NIHR) Biomedical Research Centre (BRC) at South London and Maudsley NHS Foundation Trust and King’s College London. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care. High performance computing facilities were funded with capital equipment grants from the GSTT Charity (TR130505) and Maudsley Charity (980). The PGC has received major funding from the US National Institute of Mental Health (NIMH) and the US National Institute of Drug Abuse (NIDA) of the US National Institutes of Health (NIH; U01 MH109528 to PFS, U01MH109514 to MCO, and U01 MH1095320 to A Agrawal). We acknowledge the continued support of the NL Genetic Cluster Computer ( http://www.geneticcluster.org/ ) hosted by SURFsara in the management and curation of PGC data, with funding from Scientific Organization Netherlands (480-05-003 to DP). Central analysis of PGC data was funded by UK Medical Research Council (MRC) Centre and Program Grants (G0801418, G0800509 to PAH, MCO, MJO) and grants from the Australian National Health and Medical Research Council (NHMRC; 1078901,108788 to NRW). GB, JRIC, HG, CL were supported in part by the NIHR as part of the Maudsley BRC. DP is funded by the Dutch Brain Foundation and the VU University Amsterdam Netherlands. PFS gratefully acknowledges support from the Swedish Research Council (Vetenskapsrådet, award D0886501).

Acknowledgements and funding for individual cohorts follows. BD_TRS: This work was funded by the Deutsche Forschungsgemeinschaft (DFG, grants FOR2107 DA1151/5-1, SFB-TRR58, and Project C09 to UD) and the Interdisciplinary Center for Clinical Research (IZKF) of the medical faculty of Münster (grant Dan3/012/17 to UD). BiGS: Research was funded by the NIMH (Chicago: R01 MH103368 to ESG, NIMH: R01 MH061613 and ZIA MH002843 to FJM, Pittsburgh: MH63480 to VN, UCSD: MH078151, MH081804, MH59567 to JK). FJM was supported by the NIMH Intramural Research Program, NIH, DHHS. BOMA-Australia: Funding was supplied by the Australian NHMRC (1037196, 1066177, and 1063960 to JMF, 1103623 to SEM, 1037196 to PBM, 1078399 to GWM, 1037196 to PRS). JMF would like to thank Janette M O’Neil and Betty C Lynch for their support. BOMA-Germany I, BOMA-Germany II, BOMA-Germany III, PsyCourse, and Münster MDD Cohort: This work was supported by the German Ministry for Education and Research (BMBF) through the Integrated Network IntegraMent (Integrated Understanding of Causes and Mechanisms in Mental Disorders), under the auspices of the e:Med program (01ZX1314A/01ZX1614A to MMN and SC, 01ZX1314G/01ZX1614G to MR, 01ZX1314K to TGS) and through grants NGFNplus MooDS (Systematic Investigation of the Molecular Causes of Major Mood Disorders and Schizophrenia; grant 01GS08144, 01GS08147 to MMN, MR and SC). This work was also supported by the DFG (NO246/10-1 to MMN [FOR 2107], RI 908/11-1 to MR [FOR 2107], WI 3429/3-1 to SHW, SCHU 1603/4-1, SCHU 1603/5-1 [KFO 241] and SCHU 1603/7-1 [PsyCourse] to TGS), the Swiss National Science Foundation (SNSF, grant 156791 to SC) and the European Union (N Health-F2-2008-222963 to BTB and VA). MMN is supported through the Excellence Cluster ImmunoSensation. TGS is supported by an unrestricted grant from the Dr. Lisa-Oehler Foundation. AJF received support from the BONFOR Programme of the University of Bonn, Germany. MH was supported by the DFG. BOMA-Romania: The work was supported by Unitatea Executiva Pentru Finantarea Invatamantului Superior a Cercetarii (89/2012 to MG-S). Bulgarian Trios: Recruitment was funded by the Janssen Research Foundation, and genotyping was funded by multiple grants to the Stanley Center for Psychiatric Research at the Broad Institute from the Stanley Medical Research Institute, The Merck Genome Research Foundation, and the Herman Foundation to GK. CoFaMS – Adelaide: Research was funded by the Australian NHMRC (APP1060524 to BTB). CONVERGE: Research was funded by the Wellcome Trust (WT090532/Z/09/Z, WT083573/Z/07/Z and WT089269/Z/09/Z to J Flint) and the NIMH (MH100549 to KSK). Danish RADIANT: Research was funded by Højteknologifonden (0001-2009-2 to TW) and the Lundbeck Foundation, (R24-A3242 to TW). deCODE: Research was funded by FP7-People-2011-IAPP grant agreement PsychDPC, (286213 to KS), and NIDA (R01 DA017932 to KS, R01 DA034076 to TT). The authors are thankful to the participants and staff at the Patient Recruitment Center. Edinburgh: Genotyping was conducted at the Genetics Core Laboratory at the Clinical Research Facility (University of Edinburgh). Research was funded by the Wellcome Trust (104036/Z/14/Z to AMM, T-KC, and DJP). DJM is supported by an NRS Clinical Fellowship funded by the CSO. EGCUT: Research was funded by European Union Project, (EstRC-IUT20-60, No. 2014-2020.4.01.15- 0012, 692145 to AM). Fran: This research was supported by Foundation FondaMental, Créteil, France and by the Investissements d’Avenir Programs managed by the ANR (ANR-11-IDEX-0004-02 and ANR-10-COHO-10-01 to ML). GenPOD/Newmeds: Research was funded by MRC (G0200243 to GL and MCO), EU 6th Framework, (LSHB-CT-2003-503428 to RH), IMI-JU, (15008 to GL). GenScot: Research was funded by the UK Chief Scientist Office (CZD/16/6 to DJP) and the Scottish Funding Council (HR03006 to DJP). We are grateful to all the families who took part, the general practitioners and the Scottish School of Primary Care for their help in recruiting them, and the whole Generation Scotland team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists, healthcare assistants and nurses. Genotyping was conducted at the Genetics Core Laboratory at the Clinical Research Facility (University of Edinburgh). GERA: Participants in the Genetic Epidemiology Research on Adult Health and Aging Study are part of the Kaiser Permanente Research Program on Genes, Environment, and Health, supported by the NIA, NIMH, OD, (RC2 AG036607 to CS, NRisch) and the Wayne and Gladys Valley Foundation, The Ellison Medical Foundation, the Robert Wood Johnson Foundation, and the Kaiser Permanente Regional and National Community Benefit Programs. GSK_Munich: We thank all participants in the GSK-Munich study. We thank numerous people at GSK and Max-Planck Institute, BKH Augsburg and Klinikum Ingolstadt in Germany who contributed to this project. Halifax: Halifax data were obtained with support from the Canadian Institutes of Health Research to MA. Harvard i2b2: Research funded by NIMH (R01 MH085542 to JWS, R01 MH086026 to RHP). iPSYCH: The iPSYCH (The Lundbeck Foundation Initiative for Integrative Psychiatric Research) team acknowledges funding from The Lundbeck Foundation (grant no R102-A9118 and R155-2014-1724, R129-A3973 and R24- A3243), the Stanley Medical Research Institute, the European Research Council (294838), the Novo Nordisk Foundation for supporting the Danish National Biobank resource, the Capital Region of Denmark, (R144-A5327), and grants from Aarhus and Copenhagen Universities and University Hospitals, including support to the iSEQ Center, the GenomeDK HPC facility, and the CIRRAU Center. All funding was to the iPSYCH PIs: TW, ADB, OM, MN, DH, and PBM. Janssen: Funded by Janssen Research & Development, LLC. We are grateful to the study volunteers for participating in the research studies and to the clinicians and support staff for enabling patient recruitment and blood sample collection. We thank the staff in the former Neuroscience Biomarkers of Janssen Research & Development for laboratory and operational support (e.g., biobanking, processing, plating, and sample de-identification), and to the staff at Illumina for genotyping Janssen DNA samples. MARS/BiDirect: This work was funded by the Max Planck Society, by the Max Planck Excellence Foundation, and by a grant from the German Federal Ministry for Education and Research (BMBF) in the National Genome Research Network framework (NGFN2 and NGFN-Plus, FKZ 01GS0481), and by the BMBF Program FKZ01ES0811. We acknowledge all study participants. We thank numerous people at Max-Planck Institute, and all study sites in Germany and Switzerland who contributed to this project. Controls were from the Dortmund Health Study which was supported by the German Migraine & Headache Society, and by unrestricted grants to the University of Münster from Almirall, Astra Zeneca, Berlin Chemie, Boehringer, Boots Health Care, Glaxo-Smith-Kline, Janssen Cilag, McNeil Pharma, MSD Sharp & Dohme, and Pfizer. Blood collection was funded by the Institute of Epidemiology and Social Medicine, University of Münster. Genotyping was supported by the German Ministry of Research and Education (BMBF grant 01ER0816, 01ER1506 to KB). Mayo Bipolar Disorder Biobank: Research was funded by grants from the Marriot Foundation and the Mayo Clinic Center for Individualized Medicine to JMB and MF. Michigan (NIMH/Pritzker Neuropsychiatric Disorders Research Consortium): Research was funded by NIMH (R01 MH09414501A1, MH105653 to MB). We thank the participants who donated their time and DNA to make this study possible. We thank members of the NIMH Human Genetics Initiative and the University of Michigan Prechter Bipolar DNA Repository for generously providing phenotype data and DNA samples. Many of the authors are members of the Pritzker Neuropsychiatric Disorders Research Consortium which is supported by the Pritzker Neuropsychiatric Disorders Research Fund L.L.C. A shared intellectual property agreement exists between this philanthropic fund and the University of Michigan, Stanford University, the Weill Medical College of Cornell University, HudsonAlpha Institute of Biotechnology, the Universities of California at Davis, and at Irvine, to encourage the development of appropriate findings for research and clinical applications. Mount Sinai: This work was funded in part by a NARSAD Young Investigator award to EAS, and by NIH (R01MH106531, R01MH109536 to PS and EAS). NeuRA-CASSI-Australia: This work was funded by the NSW Ministry of Health, Office of Health and Medical Research, and by the NHRMC (568807 to CSW and TWW). CSW was a recipient of NHMRC Fellowships (#1117079, #1021970). NeuRA-IGP-Australia: Research was funded by the NHMRC (630471, 1061875, 1081603 to MJG. NESDA: Research was funded by Nederlandse Organisatie voor Wetenschappelijk (NOW; ZonMW Geestkracht grant to PWJHP). Norway: Research was funded by the Vetenskapsrådet to IA, the Western Norway Regional Health Authority to KJO, the Research Council of Norway (#421716 to IM, #249711, #248778, #223273, and #217776 to OAA), the South-East Norway Regional Health Authority (#2012-132 and #2012-131 to OAA, #2016-064 to OBS, #2017-004 to OAA and OBS, #2013-088, #2014-102, and #2011-085 to IM), and the KG Jebsen Stiftelsen to OAA. TE was funded by The South-East Norway Regional Health Authority (#2015-078) and a research grant from Mrs. Throne-Holst. NTR: Research was funded by NWO (480-15-001/674 to DIB). Pfizer: Research was funded by the EU Innovative Medicine Initiative Joint Undertaking (115008.5). PsyColaus: PsyCoLaus/CoLaus received additional support from research grants from GlaxoSmithKline and the Faculty of Biology and Medicine of Lausanne, and the SNSF (3200B0-105993, 3200B0- 118308, 33CSCO-122661, 33CS30-139468, 33CS30- 148401 to MP). QIMR: We thank the twins and their families for their willing participation in our studies. Research was funded by NHMRC (941177, 971232, 3399450 and 443011 to NGM) and NIAAA (AA07535, AA07728, and AA10249 to ACH). RADIANT: Research was funded by MRC (G0701420 to GB and CML, G0901245 to GB) and NIMH (U01 MH109528 to GB). Rotterdam Study: The Rotterdam Study is funded by Erasmus Medical Center and Erasmus University, and NWO (175.010.2005.011, 911-03- 012 to AGU). SHIP-LEGEND/TREND: SHIP is part of the Community Medicine Research net of the University of Greifswald which is funded by the DFG (GR 1912/5-1 to HJG), Federal Ministry of Education and Research (grants 01ZZ9603, 01ZZ0103, and 01ZZ0403), the Ministry of Cultural Affairs, and the Social Ministry of the Federal State of Mecklenburg-West Pomerania. Genotyping in SHIP was funded by Siemens Healthineers and the Federal State of Mecklenburg-West Pomerania. Genotyping in SHIP-TREND-0 was supported by the Federal Ministry of Education and Research (grant 03ZIK012). Span2: Research was funded by Instituto de Salud Carlos III (PI12/01139, PI14/01700, PI15/01789, PI16/01505), and cofinanced by the European Regional Development Fund (ERDF), Agència de Gestió d’Ajuts Universitaris i de Recerca-AGAUR, Generalitat de Catalunya (2014SGR1357), Departament de Salut, Generalitat de Catalunya, Spain, and a NARSAD Young Investigator Grant from the Brain & Behavior Research Foundation. This project has also received funding from the European Union’s Horizon 2020 Research and Innovation Programme under the grant agreements No 667302 and 643051. CSM is a recipient of a Sara Borrell contract (CD15/00199) and a mobility grant (MV16/00039) from the Instituto de Salud Carlos III, Ministerio de Economía, Industria y Competitividad, Spain. MR is a recipient of a Miguel de Servet contract (CP09/00119 and CPII15/00023) from the Instituto de Salud Carlos III, Ministerio de Economía, Industria y Competitividad, Spain. STAR*D: Research was funded by NIMH (R01 MH-072802 to SPH). The authors appreciate the efforts of the STAR*D investigator team for acquiring, compiling, and sharing the STAR*D clinical data set. SUNY DMC: Research was funded by NIMH (R01MH085542 to CP, MTP, JAK, and HM). SWEBIC: Research was funded by NIMH (MH077139), the Vetenskapsrådet (K2014-62X-14647-12-51 and K2010-61P-21568-01-4), the Swedish foundation for Strategic Research (KF10-0039) and the Stanley Center for Psychiatric Research, Broad Institute from a grant from Stanley Medical Research Institute, all to ML. We are deeply grateful for the participation of all subjects contributing to this research, and to the collection team that worked to recruit them. We also wish to thank the Swedish National Quality Register for Bipolar Disorders: BipoläR. Sweden: This work was funded by the Vetenskapsrådet (to MS and CL), the Stockholm County Council (to MS, CL, LB, LF, and UÖ) and the Söderström Foundation (to LB). TwinGene: Research was funded by GenomeEUtwin, (EU/QLRT-2001-01254; QLG2-CT-2002-01254 to NLP), Heart and Lung Foundation (20070481 to PKM), SFF and Vetenskapsrådet, (M-2005-1112 to U de Faire).We thank the Karolinska Institutet for infrastructural support of the Swedish Twin Registry. UCL: Research was funded by the MRC (G1000708 to AM). UCLA-Utrecht (Los Angeles): Research was funded by NIMH (R01MH090553, U01MH105578 to NBF, RAO, LMOL, and APSO). UK - BDRN: Research was funded by MRC Centre and Program Grants (G0801418, G0800509 to MCO and MJO), the Wellcome Trust (078901 to NC, IJ, LAJ), the Stanley Medical Research Institute (5710002223-01 to NC, IJ, LAJ), and a European Commission Marie Curie Fellowship (623932 to ADF). BDRN would like to acknowledge the research participants who continue to give their time to participate in our research. UK Biobank: This research has been carried out under application numbers 4844, 6818, and 16577, funded by the National Institute for Health Research under its Biomedical Research Centres funding initiative (to GB) and the Wellcome Trust (04036/Z/14/Z to AMM). UNIBO / University of Barcelona, Hospital Clinic, IDIBAPS, CIBERSAM: EV thanks the support of the Spanish Ministry of Economy and Competitiveness (PI15/00283 to EV) integrated into the Plan Nacional de I+D+I y cofinanciado por el ISCIII-Subdirección General de Evaluación y el Fondo Europeo de Desarrollo Regional (FEDER); CIBERSAM; and the Comissionat per a Universitats i Recerca del DIUE de la Generalitat de Catalunya to the Bipolar Disorders Group (2014 SGR 398). USC: Research funded by NIH (R01MH085542 to JLS). WTCCC: The principal funder of this project was the Wellcome Trust to NC and AHY. For the 1958 Birth Cohort, venous blood collection was funded by the UK MRC. AHY is funded by the National Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London. 23andMe: We thank the 23andMe research participants included in the analysis, all of whom provided informed consent and participated in the research online according to a human subjects protocol approved by an external AAHRPP-accredited institutional review board (Ethical & Independent Review Services), and the employees of 23andMe for making this work possible. 23andMe acknowledges the-invaluable contributions of Michelle Agee, Babak Alipanahi, Adam Auton, Robert K. Bell, Katarzyna Bryc, Sarah L. Elson, Pierre Fontanillas, Nicholas A. Furlotte, David A. Hinds, Bethann S. Hromatka, Karen E. Huber, Aaron Kleinman, Nadia K. Litterman, Matthew H. McIntyre, Joanna L. Mountain, Carrie A.M. Northover, Steven J. Pitts, J. Fah Sathirapongsasuti, Olga V. Sazonova, Janie F. Shelton, Suyash Shringarpure, Chao Tian, Joyce Y. Tung, Vladimir Vacic, and Catherine H. Wilson.

Disclosures

OA Andreassen has received speaker fees from Lundbeck. ATF Beekman is on speaker’s bureaus for Lundbeck and GlaxoSmithKline. G Breen reports consultancy and speaker fees from Eli Lilly, Otsuka and Illumina and grant funding from Eli Lilly. G Crawford is a cofounder of Element Genomics. E Domenici was formerly an employee of Hoffmann–La Roche and a consultant to Roche and Pierre-Fabre. J Nurnberger is an investigator for Janssen and was an investigator for Assurex. SA Paciga is an employee of Pfizer. JA Quiroz was formerly an employee of Hoffmann–La Roche. S Steinberg, H Stefansson, K Stefansson and TE Thorgeirsson are employed by deCODE Genetics/Amgen. PF Sullivan reports the following potentially competing financial interests. Current: Lundbeck (advisory committee, grant recipient). Past three years: Pfizer (scientific advisory board), Element Genomics (consultation fee), and Roche (speaker reimbursement). AH Young has given paid lectures and is on advisory boards for the following companies with drugs used in affective and related disorders: Astrazenaca, Eli Lilly, Janssen, Lundbeck, Sunovion, Servier, Livanova. AH Young is Lead Investigator for Embolden Study (Astrazenaca), BCI Neuroplasticity study and Aripiprazole Mania Study, which are investigator-initiated studies from Astrazenaca, Eli Lilly, Lundbeck, and Wyeth. All other authors declare no financial interests or potential conflicts of interest.

The Bipolar Disorder and Major Depressive Disorder Working Groups of the Psychiatric Genomics Consortium are collaborative co-authors for this article. The individual authors are (numbers refer to affiliations listed in the Supplement ): Enda M. Byrne 4 , Andreas J. Forstner 5;6;7;8;9 , Peter A. Holmans 10 , Christiaan A. de Leeuw 11 , Manuel Mattheisen 12;13;14;15;16 , Andrew McQuillin 17 , Jennifer M. Whitehead Pavlides 18 , Tune H. Pers 19;20 , Stephan Ripke 21;22;23 , Eli A. Stahl 19;24;25 , Stacy Steinberg 26 , Vassily Trubetskoy 22 , Maciej Trzaskowski 4 , Yunpeng Wang 27;28 , Liam Abbott 21 , Abdel Abdellaoui 29 , Mark J. Adams 30 , Annelie Nordin Adolfsson 31 , Esben Agerbo 16;32;33 , Huda Akil 34 , Diego Albani 35 , Ney Alliey-Rodriguez 36 , Thomas D. Als 12;13;16 , Till F. M. Andlauer 37;38 , Adebayo Anjorin 39 , Verneri Antilla 23 , Sandra Van der Auwera 40 , Swapnil Awasthi 22 , Silviu-Alin Bacanu 41 , Judith A Badner 42 , Marie Bækvad-Hansen 16;43 , Jack D. Barchas 44 , Nicholas Bass 17 , Michael Bauer 45 , Aartjan T. F. Beekman 46 , Richard Belliveau 21 , Sarah E. Bergen 3 , Tim B. Bigdeli 41;47 , Elisabeth B. Binder 37;48 , Erlend Bøen 49 , Marco Boks 50 , James Boocock 51 , Monika Budde 52 , William Bunney 53 , Margit Burmeister 54 , Henriette N. Buttenschøn 3;12;55 , Jonas Bybjerg-Grauholm 16;43 , William Byerley 56 , Na Cai 57;58 , Miquel Casas 59;60;61;62 , Enrique Castelao 63 , Felecia Cerrato 21 , Pablo Cervantes 64 , Kimberly Chambert 21 , Alexander W. Charney 25 , Danfeng Chen 21 , Jane Hvarregaard Christensen 12;13;55 , Claire Churchhouse 21;23 , David St Clair 65 , Toni-Kim Clarke 30 , Lucía Colodro-Conde 66 , William Coryell 67 , Baptiste Couvy-Duchesne 18;68 , David W. Craig 69 , Gregory E. Crawford 70;71 , Cristiana Cruceanu 37;64 , Piotr M. Czerski 72 , Anders M. Dale 73;74;75;76 , Gail Davies 77 , Ian J. Deary 77 , Franziska Degenhardt 7;8 , Jurgen Del-Favero 78 , J Raymond DePaulo 79 , Eske M. Derks 66 , Nese Direk 80;81 , Srdjan Djurovic 82;83 , Amanda L. Dobbyn 24;25 , Conor V. Dolan 29 , Ashley Dumont 21 , Erin C. Dunn 21;84;85 , Thalia C. Eley 1 , Torbjørn Elvsåshagen 86;87 , Valentina Escott-Price 10 , Chun Chieh Fan 76 , Hilary K. Finucane 88;89 , Sascha B. Fischer 5;9 , Matthew Flickinger 90 , Jerome C. Foo 91 , Tatiana M. Foroud 92 , Liz Forty 10 , Josef Frank 91 , Christine Fraser 10 , Nelson B. Freimer 93 , Louise Frisén 94;95;96 , Katrin Gade 52;97 , Diane Gage 21 , Julie Garnham 98 , Claudia Giambartolomei 51 , Fernando S. Goes 99 , Jaqueline Goldstein 21 , Scott D. Gordon 66 , Katherine Gordon-Smith 100 , Elaine K. Green 101 , Melissa J. Green 102 , Tiffany A. Greenwood 75 , Jakob Grove 12;13;16;103 , Weihua Guan 104 , Lynsey S. Hall 30;105 , Marian L. Hamshere 10 , Christine Søholm Hansen 16;43 , Thomas F. Hansen 16;106;107 , Martin Hautzinger 108 , Urs Heilbronner 52 , Albert M. van Hemert 109 , Stefan Herms 5;7;8;9 , Ian B. Hickie 110 , Maria Hipolito 111 , Per Hoffmann 5;7;8;9 , Dominic Holland 73;112 , Georg Homuth 113 , Carsten Horn 114 , Jouke-Jan Hottenga 29 , Laura Huckins 24;25 , Marcus Ising 15 , Stéphane Jamain 116;117 , Rick Jansen 46 , Jessica S. Johnson 24;25 , Simone de Jong 1;2 , Eric Jorgenson 118 , Anders Juréus 3 , Radhika Kandaswamy 1 , Robert Karlsson 3 , James L. Kennedy 119;120;121;122 , Farnush Farhadi Hassan Kiadeh 123 , Sarah Kittel-Schneider 124 , James A. Knowles 125;126 , Manolis Kogevinas 127 , Isaac S. Kohane 128;129;130 , Anna C. Koller 7;8 , Julia Kraft 22 , Warren W. Kretzschmar 131 , Jesper Krogh 132 , Ralph Kupka 46;133 , Zoltán Kutalik 134;135 , Catharina Lavebratt 94 , Jacob Lawrence 136 , William B. Lawson 111 , Markus Leber 137 , Phil H. Lee 21;23;138 , Shawn E. Levy 139 , Jun Z. Li 140 , Yihan Li 131 , Penelope A. Lind 66 , Chunyu Liu 141 , Loes M. Olde Loohuis 93 , Anna Maaser 7;8 , Donald J. MacIntyre 142;143 , Dean F. MacKinnon 99 , Pamela B. Mahon 79;144 , Wolfgang Maier 145 , Robert M. Maier 18 , Jonathan Marchini 146 , Lina Martinsson 95 , Hamdi Mbarek 29 , Steve McCarroll 21;147 , Patrick McGrath 148 , Peter McGuffin 1 , Melvin G. McInnis 149 , James D. McKay 150 , Helena Medeiros 126 , Sarah E. Medland 66 , Divya Mehta 18;151 , Fan Meng 34;149 , Christel M. Middeldorp 29;152;153 , Evelin Mihailov 154 , Yuri Milaneschi 46 , Lili Milani 154 , Saira Saeed Mirza 80 , Francis M. Mondimore 99 , Grant W. Montgomery 4 , Derek W. Morris 155;156 , Sara Mostafavi 157;158 , Thomas W Mühleisen 5;159 , Niamh Mullins 1 , Matthias Nauck 160;161 , Bernard Ng 158 , Hoang Nguyen 24;25 , Caroline M. Nievergelt 75;162 , Michel G. Nivard 29 , Evaristus A. Nwulia 111 , Dale R. Nyholt 163 , Claire O’Donovan 98 , Paul F. O’Reilly 1 , Anil P. S. Ori 93 , Lilijana Oruc 164 , Urban Ösby 165 , Hogni Oskarsson 166 , Jodie N. Painter 66 , José Guzman Parra 167 , Carsten Bøcker Pedersen 16;32;33 , Marianne Giørtz Pedersen 16;32;33 , Amy Perry 100 , Roseann E. Peterson 41;168 , Erik Pettersson 3 , Wouter J. Peyrot 46 , Andrea Pfennig 45 , Giorgio Pistis 63 , Shaun M. Purcell 25;144 , Jorge A. Quiroz 169 , Per Qvist 12;13;55 , Eline J. Regeer 170 , Andreas Reif 124 , Céline S. Reinbold 5;9 , John P. Rice 171 , Brien P. Riley 41 , Fabio Rivas 167 , Margarita Rivera 1;172 , Panos Roussos 24;25;173 , Douglas M. Ruderfer 174 , Euijung Ryu 175 , Cristina Sánchez-Mora 59;60;62 , Alan F. Schatzberg 176 , William A. Scheftner 177 , Robert Schoevers 178 , Nicholas J. Schork 179 , Eva C. Schulte 52;180 , Tatyana Shehktman 75 , Ling Shen 118 , Jianxin Shi 181 , Paul D. Shilling 75 , Stanley I. Shyn 182 , Engilbert Sigurdsson 183 , Claire Slaney 98 , Olav B. Smeland 73;184;185 , Johannes H. Smit 46 , Daniel J. Smith 186 , Janet L. Sobell 187 , Anne T. Spijker 188 , Michael Steffens 189 , John S. Strauss 121;190 , Fabian Streit 91 , Jana Strohmaier 91 , Szabolcs Szelinger 191 , Katherine E. Tansey 192 , Henning Teismann 193 , Alexander Teumer 194 , Robert C Thompson 149 , Wesley Thompson 55;75;87;107 , Pippa A. Thomson 195 , Thorgeir E. Thorgeirsson 26 , Matthew Traylor 196 , Jens Treutlein 91 , André G. Uitterlinden 197 , Daniel Umbricht 198 , Helmut Vedder 199 , Alexander Viktorin 3 , Peter M. Visscher 4;18 , Weiqing Wang 24;25 , Stanley J. Watson 149 , Bradley T. Webb 168 , Cynthia Shannon Weickert 102;200 , Thomas W. Weickert 102;200 , Shantel Marie Weinsheimer 55;107 , Jürgen Wellmann 193 , Gonneke Willemsen 29 , Stephanie H. Witt 91 , Yang Wu 4 , Hualin S. Xi 201 , Wei Xu 202;203 , Jian Yang 4;18 , Allan H. Young 204 , Peter Zandi 205 , Peng Zhang 206 , Futao Zhang 4 , Sebastian Zollner 149 , Rolf Adolfsson 31 , Ingrid Agartz 14;49;207 , Martin Alda 98;208 , Volker Arolt 209 , Lena Backlund 95 , Bernhard T. Baune 210 , Frank Bellivier 211;212;213;214 , Klaus Berger 193 , Wade H. Berrettini 215 , Joanna M. Biernacka 175 , Douglas H. R. Blackwood 30 , Michael Boehnke 90 , Dorret I. Boomsma 29 , Aiden Corvin 156 , Nicholas Craddock 10 , Mark J. Daly 21;23 , Udo Dannlowski 209 , Enrico Domenici 216 , Katharina Domschke 217 , Tõnu Esko 19;147;154;218 , Bruno Etain 211;213;214;219 , Mark Frye 220 , Janice M. Fullerton 200;221 , Elliot S. Gershon 36;222 , EJC de Geus 29;223 , Michael Gill 156 , Fernando Goes 79 , Hans J. Grabe 40 , Maria Grigoroiu-Serbanescu 224 , Steven P. Hamilton 225 , Joanna Hauser 72 , Caroline Hayward 226 , Andrew C. Heath 171 , David M. Hougaard 16;43 , Christina M. Hultman 3 , Ian Jones 10 , Lisa A. Jones 100 , René S. Kahn 25;50 , Kenneth S. Kendler 41 , George Kirov 10 , Stefan Kloiber 115;121;190 , Mikael Landén 3;227 , Marion Leboyer 117;211;228 , Glyn Lewis 17 , Qingqin S. Li 229 , Jolanta Lissowska 230 , Susanne Lucae 115 , Pamela A. F. Madden 119 , Patrik K. Magnusson 3 , Nicholas G. Martin 66;231 , Fermin Mayoral 167 , Susan L. McElroy 232 , Andrew M. McIntosh 30;77 , Francis J. McMahon 233 , Ingrid Melle 234;235 , Andres Metspalu 154;236 , Philip B. Mitchell 102 , Gunnar Morken 237;238 , Ole Mors 16;239 , Preben Bo Mortensen 12;16;32;33 , Bertram Müller-Myhsok 37;240;241 , Richard M. Myers 139 , Benjamin M. Neale 19;21;23 , Vishwajit Nimgaonkar 242 , Merete Nordentoft 16;243 , Markus M. Nöthen 7;8 , Michael C. O’Donovan 10 , Ketil J. Oedegaard 244;245 , Michael J. Owen 10 , Sara A. Paciga 246 , Carlos Pato 126;247 , Michele T. Pato 126 , Nancy L. Pedersen 3 , Brenda W. J. H. Penninx 46 , Roy H. Perlis 248;249 , David J. Porteous 195 , Danielle Posthuma 11;250 , James B. Potash 79 , Martin Preisig 63 , Josep Antoni Ramos-Quiroga 59;60;61;62 , Marta Ribasés 59;60;62 , Marcella Rietschel 91 , Guy A. Rouleau 251;252 , Catherine Schaefer 118 , Martin Schalling 94 , Peter R. Schofield 200;221 , Thomas G. Schulze 52;79;91;97;233 , Alessandro Serretti 253 , Jordan W. Smoller 21;84;85 , Hreinn Stefansson 26 , Kari Stefansson 26;254 , Eystein Stordal 255;256 , Henning Tiemeier 80;257;258 , Gustavo Turecki 259 , Rudolf Uher 98 , Arne E. Vaaler 260 , Eduard Vieta 261 , John B. Vincent 190 , Henry Völzke 194 , Myrna M. Weissman 148;262 , Thomas Werge 16;107;263 , Ole A. Andreassen 184;185 , Anders D. Børglum 12;13;16 , Sven Cichon 5;7;9;159 , Howard J. Edenberg 264 , Arianna Di Florio 10;265 , John Kelsoe 75 , Douglas F. Levinson 176 , Cathryn M. Lewis 1;2;266 , John I. Nurnberger 92;267 , Roel A. Ophoff 50;51;93 , Laura J. Scott 90 , Pamela Sklar 24;25† , Patrick F. Sullivan 3;265;268 , Naomi R. Wray 4;18

Data availability

GWAS results from analyses including 23andMe are restricted by a data transfer agreement with 23andMe. For these analyses, LD-independent sets of 10,000 SNPs will be made available via the Psychiatric Genetics Consortium ( https://www.med.unc.edu/pgc/results-and-downloads ). Summary statistics not including 23andMe will be made available via the Psychiatric Genetics Consortium ( https://www.med.unc.edu/pgc/results-and-downloads ).

Quick Links
Make An Appointment
Our Services
Price Estimate
Price Transparency
Pay Your Bill
Patient Experience
Careers at UH

Schedule an appointment today

Mood Disorder Research

The Mood Disorders Program is committed to providing expert care to adult patients and their families while conducting research to advance scientific knowledge. We are dedicated to improving clinical outcomes of under-served patients with mood disorders, particularly those presenting with rapid cycling bipolar disorder, co-occurring anxiety and substance misuse/abuse, legal complications, and Ohio Guard and Reserve members. The research we do in the Mood Disorders Program focuses not only on symptom improvement but also on function and disability to ultimately improve patient quality of life.

Our mission is to identify the causes of bipolar disorder and other mood disorders, to find effective treatment, and to discover a cure. Our research is patient-centered. This means we have designed our research studies to improve the lives of people who choose to participate in them. Our research focuses on function and disability, not just symptom improvement.

Why We Do What We Do

Ohio army national guard ptsd study.

The Ohio Army National Guard Mental Health Initiative (OHARNG-MHI) began in 2008 to evaluate the relationships between resilience and risk factors of developing PTSD and other mental illness. The study, led by Dr. Joseph Calabrese, recently completed its ninth and final year of data collection.

Contact Us For More Information

Mood Disorders Program Interventions and Services Research University Hospitals Cleveland Medical Center Case Western Reserve University School of Medicine 10524 Euclid Ave., 12th Floor Cleveland, Ohio 44106-5923

Phone: 216-844-2865 Email: [email protected]

IMAGES

Mood Disorders
Yale Mood Disorders Research Study
Mood Disorders: Comprehensive Review
Yale Mood Disorders Research Study
Mood Disorders: Comprehensive Review
Mood Disorders: Clinical Management and Research Issues : Eric J.L

VIDEO

Supporting students with Anxiety & Mood Disorders 4-25-24
All mood disorders are curable by viewing my status. #viewmystatus#STAVA#amyloxy🌚💛
MOOD DISORDERS PART II DEPRESSION
Examining co-occurring OCD and Depression: Research and clinical strategies
"Depressive Disorder or depression by design?"
Understanding Mood Disorders: Depression

COMMENTS

Advances in Understanding and Treating Mood Disorders
This issue of the Journal also presents original research articles that address topics related to the treatment of bipolar disorder, the use of a new transcranial magnetic stimulation (TMS) strategy for treatment of refractory depression, and the effects of gender-affirming interventions on the treatment of mood and anxiety disorders in ...
Yale Mood Disorders Research Program (MDRP) < Psychiatry
The Yale Mood Disorders Research Program (MDRP) is dedicated to understanding the causes of mood and related disorders, and suicide risk, across the lifespan. The MDRP brings together a multi-disciplinary group of scientists from across the Yale campus in a highly collaborative research effort. We use a wide variety of scientific methods to ...
Mood Disorder
Mood Disorder - StatPearls
Advances in Understanding and Treating Mood Disorders
Environmental factors also play a prominent role in the ex-pression of mood disorders: advances are being made at the molecular level in understanding how environmental events are epigenetically programmed to result in altered gene ex-pression that is informative for understanding the gene-by-environment interactions relevant to mood disorders.
Mood Disorders Research Program
Program Chief. 6001 Executive Boulevard, Room 7131, MSC 9637. 301-827-7614, [email protected]. This program supports research on the etiology, core features, longitudinal course, assessment of, and interventions for mood and arousal/regulatory psychopathology as outlined in the RDoC initiative that emphasizes a mechanistic ...
Research
Research. Our vision for the Mood Disorders Center begins with identifying causes of illness through scientific research in the realms of genetics and epigenetics (the controls over the expression of genes), and environmental factors like stress. We are devoting extensive laboratory and computer-based analytic efforts towards this goal.
Mood Disorders
Through the research programs at the Stanford Mood Disorders Center, Stanford has led the quest for new knowledge and therapies for mood disorders. Today the center is expanding its reach and mobilizing Stanford's diverse expertise toward a powerful shared mission: to overcome mood disorders through innovation and compassion.
Precision medicine for mood disorders: objective assessment, risk
Precision medicine for mood disorders: objective ...
Mood Disorders & Cognition Research
Researchers in the Mood Disorders and Cognition (MDC) research program in NYU Langone's Department of Psychiatry investigate the neurobiology of mood and cognition disorders, including major depression, bipolar disorder, Alzheimer's disease, and novel treatments for these conditions, including pharmacological agents and brain stimulation strategies.
Mood Disorders
Purpose of review: This comprehensive review of mood disorders brings together the past and current literature on the diagnosis, evaluation, and treatment of the depressive and bipolar disorders. It highlights the primary mood disorders and secondary neurologic causes of mood disorders that are commonly encountered in a clinical setting.
Inflammation as a treatment target in mood disorders: review
Mood disorders, i.e. major depressive disorder (MDD) and bipolar disorders, are leading sources of disability worldwide. Currently available treatments do not yield remission in approximately a third of patients with a mood disorder. ... Using the Research Domain Criteria approach to examine specific symptom subsets (for example anhedonia ...
Diagnosis and management of bipolar disorders
Bipolar disorders (BDs) are recurrent and sometimes chronic disorders of mood that affect around 2% of the world's population and encompass a spectrum between severe elevated and excitable mood states (mania) to the dysphoria, low energy, and despondency of depressive episodes. The illness commonly starts in young adults and is a leading cause of disability and premature mortality.
Neurobiological mechanisms of mood disorders: Stress vulnerability and
Nevertheless, more research is needed to uncover the complex interaction between genetics and abnormal stressors that can lead to disease vulnerability, including psychiatric disorders. ... In other mood disorders, such as bipolar disorder, patients present with low levels of Treg cells along with immune and inflammatory imbalance (do Prado et ...
Characterizing mood disorders in the AFFECT study: a large
Characterizing mood disorders in the AFFECT study
Bipolar Breakthrough
Bipolar Breakthrough | Harvard Medical School
Association Between Mood Disorders and Risk of COVID-19 Infection
Study Selection Primary research articles that reported quantitative COVID-19 outcome data in persons with mood disorders vs persons without mood disorders of any age, sex, and nationality were selected. Of 1950 articles identified through this search strategy, 21 studies were included in the analysis.
Mood Disorders
Mood Disorders - Johns Hopkins Medicine ... Mood Disorders
Design and Methods of the Mood Disorder Cohort Research Consortium
The Mood Disorder Cohort Research Consortium (MDCRC) study is designed as a naturalistic observational prospective cohort study for early-onset mood disorders (major depressive disorders, bipolar disorders type 1 and 2) in South Korea. The study subjects consist of two populations: 1) patients with mood disorders under 25 years old and 2 ...
Mood Disorders Clinical Trials
However research has revealed that thoracic transplant caregivers have increased mood disorders, distress, and high caregiver burden, especially while caring for patients on the organ transplant waiting list. By monitoring caregiver traits and behaviors we will be better able to provide care and services for CG to improve both CG and patient ...
Mood Disorders Research Program
The Mood Disorders Research Program, under the direction of Dr. Rif El-Mallakh, pursues basic and clinical research in the major mood disorders, focusing on bipolar illness. Most basic research projects focus on mood-state-related abnormalities in nerve or membrane function. Studies include: transmembrane potential changes in Lymphocytes and ...
Mood disorders
Mood disorders - Symptoms and causes
The genetics of the mood disorder spectrum: genome-wide association
Introduction. Mood disorders affect 10-20% of the global population across their lifetime, ranging from brief, mild episodes to severe, incapacitating conditions that markedly impact lives (1-4).Major depressive disorder and bipolar disorder are the most common forms and have been grouped together since the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III
Mood Disorder Research
The Mood Disorders Program is committed to providing expert care to adult patients and their families while conducting research to advance scientific knowledge. We are dedicated to improving clinical outcomes of under-served patients with mood disorders, particularly those presenting with rapid cycling bipolar disorder, co-occurring anxiety and ...

In the News

Yale Mood Disorders Research Program (MDRP)

Examples of our Current Research Projects:

Participating in a Yale Mood Disorders Research Program Study

Celebrating 75 Years! Learn More >>

Mood Disorders Research Program

Areas of Emphasis

Pioneering Solutions for Depression and Bipolar Disorder

Stanford Mood Disorders Center at a Glance

Upcoming Event

In the News

Transformative Therapies & Technologies

Save citation to file

Add to My Bibliography

Mood Disorders

Similar articles

Publication types

Related information

Research Materials

Miscellaneous

Search form

Diagnosis and management of bipolar disorders

Introduction

Sources and selection criteria

Modern definitions of bipolar disorder

Epidemiology

Causes of bipolar disorder

Insights from genetics

Insights from neuroimaging

Environmental risk factors

Comorbidity

Precursors and prodromes: who develops bipolar disorder?

Screening for bipolar disorder

Racial/ethnic disparities

Acute treatment of mania

Pharmacological approaches to bipolar depression

Antidepressants

Pharmacological approaches to prevention of recurrent episodes

Psychotherapeutic approaches

Neurostimulation approaches

Treatment resistance in bipolar disorder

Treatment considerations to reduce suicide in bipolar disorder

Treatment consideration in BD-II and bipolar spectrum conditions

Precision medicine: can it be applied to improve the care of bipolar disorder?

Emerging treatments

Questions for future research

Characterizing mood disorders in the AFFECT study: a large, longitudinal, and phenotypically rich genetic cohort in the US

Similar content being viewed by others

The genetic basis of major depressive disorder

Genome-wide significant risk loci for mood disorders in the Old Order Amish founder population

Dimensional and transdiagnostic phenotypes in psychiatric genome-wide association studies

Cohort design

Study assessments

SNP genotyping

Genome-wide associations

SNP-heritability and genetic correlations

Statistical analyses

Cohort characteristics

Disease history

Symptom state

Symptom-level comorbidities

Family psychiatric history

Environmental influences

SNP-heritability and genetic comparability

Data availability

Acknowledgements

Author information

Authors and Affiliations

23andMe Research Team

Contributions

Corresponding authors

Ethics declarations

Additional information

Supplementary information

About this article

Share this article

This article is cited by

Polygenic heterogeneity in antidepressant treatment and placebo response

Quick links

See More About