Bipolar disorder diagnosis


How is a diagnosis of bipolar disorder made?

Bipolar disorder is characterised by episodes of mania or less severe hypomania, and depression. Bipolar I disorder is determined by the existence of mania, which may include psychotic features, while bipolar II disorder is determined by less severe hypomania. Cyclothymic disorder is an overall milder form of bipolar disorder, however symptoms occur fairly often and constantly. 

A manic episode is at least one week of extremely high spirited or irritableness most of the time. There are changes in normal behaviour. These include exaggerated self-esteem/grandiosity, less sleep, talking more, talking loudly and quickly, being easily distracted, doing many activities at once, very risky behaviour, having uncontrollable and racing thoughts, and/or quickly changing ideas or topics. A depressive episode is a period of at least two weeks of intense sadness, despair, helplessness, hopelessness or worthlessness. There may be loss of interest in activities once enjoyed, feelings of guilt, restlessness or agitation, sleep problems, slowed speech or movements, changes in appetite, loss of energy, difficulty concentrating, remembering or making decisions, and/or thoughts of death or suicide.

What is the evidence regarding diagnosis of bipolar disorder?

For adults

Moderate to high quality evidence finds reasonable diagnostic stability of bipolar disorder over time. There was better inter-rater and test-retest reliability for diagnosing bipolar disorder using any method, than for diagnosing schizoaffective disorder, schizophrenia, or unipolar depression. Compared to people with schizoaffective disorder, people with bipolar disorder are usually older, married, have a later age of illness onset, a shorter duration of illness, have less psychotic and negative symptoms (e.g. social withdrawal), less depression, more years of education, and more likelihood of being Caucasian.

Moderate to high quality evidence finds the Hypomania Checklist, the Bipolar Spectrum Diagnostic Scale and the Mood Disorder Questionnaire have good accuracy for detecting bipolar disorder. The Hypomania Checklist was better at detecting bipolar disorder II than the Mood Disorder Questionnaire. Moderate quality evidence finds reasonable predictive value and moderate agreement for bipolar disorder diagnosis between administrative databases using the ICD-10, and clinical or research diagnoses. However, an estimated 17% of people in primary care settings that were diagnosed with depression had undiagnosed bipolar disorder. Results from structural and functional neuroimaging studies analysed using machine learning show similar, moderate levels of accuracy for determining bipolar disorder diagnosis from other psychiatric diagnoses or from healthy controls.

For children and adolescents

Moderate quality evidence finds the clinical features associated more often in children or youth with bipolar depression than in children or youth with unipolar depression include; more psychiatric comorbidities and behavioural problems (i.e. oppositional disorder, conduct disorder, anxiety disorders, irritability, suicidal/self-harm, social impairment, substance use); earlier onset of mood symptoms; more severe depression; and having a family history of any psychiatric illness.

There is good reliability of checklists for identifying bipolar disorder in children. Checklists are better at detecting bipolar disorder than at detecting schizophrenia or schizoaffective disorder, but not as good as detecting unipolar depression. Caregiver report was more accurate at detecting bipolar disorder than youth self-report or teacher report, and checklists that focus on manic symptoms were most accurate.

August 2021

Image: ©designer491 –

Last updated at: 2:31 pm, 15th February 2022
To view documentation related to this topic download the files below
Fact Sheet Technical Commentary

NeuRA Libraries

Title Colour Legend:
Green - Topic summary is available.
Orange - Topic summary is being compiled.
Red - Topic summary has no current systematic review available.