The Mood Disorder Questionnaire (MDQ) is a lifetime screen for bipolar spectrum disorder. Its 13 yes/no items ask whether there has ever been a period, unlike the person's usual self, marked by the DSM-IV features of mania or hypomania: elevated or irritable mood, inflated self-confidence, reduced need for sleep, pressured speech, racing thoughts, distractibility, increased energy and activity, heightened sociability and libido, spending sprees and risk-taking. Two further questions ask whether several of these experiences clustered in the same period and how much of a problem they caused.
The MDQ measures history, not present state. It does not grade the severity of a current episode, and it was not designed to track change over treatment - a person who has been well for years will still screen positive if they once met the criteria. A positive screen indicates that a structured diagnostic interview for bipolar disorder is warranted; it is not itself a diagnosis.
Where it comes from.
The MDQ was developed by Robert Hirschfeld and colleagues and published in the American Journal of Psychiatry in 2000. The items were drafted by a committee of psychiatrists, researchers and consumer advocates, including representatives of NAMI and the National Depressive and Manic-Depressive Association, and the instrument was validated in 198 outpatients attending five mood-disorder clinics, using a telephone SCID bipolar module as the reference standard.
Its purpose was practical: bipolar disorder is frequently misdiagnosed as unipolar depression, often for years, because patients present when depressed and do not volunteer past hypomania. The MDQ gave clinicians a two-minute self-report that could be handed to any patient presenting with depression. It was subsequently tested in the US general population (2003), in primary-care patients being treated for depression (2005) and, as the MDQ-A, in adolescents and their parents (2006).
How scoring works.
The MDQ has three parts. Part one lists 13 lifetime manic or hypomanic experiences answered yes or no; the number of "yes" answers gives a 0-13 count. Part two asks whether several of the endorsed experiences happened during the same period of time. Part three asks how much of a problem these experiences caused (no problem, minor, moderate or serious).
A positive screen requires all three criteria at once: seven or more "yes" answers in part one, "yes" to the co-occurrence question, and "moderate" or "serious" problems in part three. The symptom count alone is not a severity score, and a positive screen is not a diagnosis - it should be followed by a clinical interview for bipolar disorder. Some distributed versions add unscored questions on family history and previous diagnosis.
Score
Severity
Interpretation
0–6
Below symptom threshold
Fewer than seven lifetime symptoms endorsed. Screen negative regardless of the other two questions.
7–13
Symptom threshold met
Seven or more symptoms endorsed. The screen is positive only if the symptoms clustered in the same period AND caused moderate or serious problems; otherwise it is negative. A positive screen warrants a clinical interview for bipolar disorder.
How well it performs.
In the original outpatient validation (Hirschfeld et al., 2000), the standard positive-screen rule gave a sensitivity of 0.73 and a specificity of 0.90 against a SCID diagnosis of bipolar disorder, with a Cronbach's alpha of 0.90. Accuracy falls sharply outside specialist settings: in a US general-population sample the sensitivity was 0.28 with a specificity of 0.97 (Hirschfeld et al., 2003), and in primary-care patients treated for depression it was 0.58 and 0.93 respectively (Hirschfeld et al., 2005).
A meta-analysis of 21 studies (Wang et al., 2015) found a pooled sensitivity of 0.62 and specificity of 0.85 at the standard cutoff of seven items; performance was best when distinguishing bipolar from unipolar depression (0.76 and 0.81) and weakest in samples that excluded patients already known to have bipolar disorder (sensitivity 0.37). The MDQ is markedly less sensitive to bipolar II disorder: in a psychiatric outpatient sample sensitivity was 0.69 for bipolar I but 0.30 for bipolar II or NOS (Miller et al., 2004).
0.90
CRONBACH'S α
73%
SENSITIVITY (≥7, OUTPATIENTS)
90%
SPECIFICITY (≥7, OUTPATIENTS)
62%
POOLED SENSITIVITY (21 STUDIES)
How it compares to the alternatives.
Instrument
Items
Time
When to reach for it
MDQ
13 + 2
~5 min
Lifetime screen for a bipolar history in patients presenting with depression. Specific but misses many bipolar II presentations. Licensed.
Depression severity. Pair with the MDQ when a depressed patient has not been asked about past hypomania.
Right tool, wrong tool.
The MDQ is intended for adults presenting with depressive symptoms in psychiatric, primary-care and general-medical settings, where the question is whether a bipolar history has been missed. It has been validated in adult psychiatric outpatients, in the US general population and in primary care, and has been translated into many languages. The adolescent version (MDQ-A) performs best as a parent-report instrument; the self-report adult form is not validated in children.
Reach for it when
-A patient presenting with depression who has never been asked about past manic or hypomanic periods.
-Before starting or escalating an antidepressant, where a missed bipolar history would change the plan.
-Depression that has not responded to two or more antidepressants, or a family history of bipolar disorder.
-Structuring the history-taking around lifetime hypomania before a diagnostic interview.
Reach for something else when
-Grading the severity of a current manic episode - use the ASRM or the clinician-rated YMRS.
-Tracking response to treatment; the MDQ measures lifetime history and does not change.
-Ruling out bipolar II disorder - a negative MDQ is weak evidence against it.
-Confirming a diagnosis; the MDQ is a screen and cannot replace a structured interview.
-Self-report in children or younger adolescents - the parent-report MDQ-A performs better.
Reading the score with care.
The MDQ is a categorical lifetime screen rather than a severity scale, and no standard error of measurement or minimal clinically important difference has been published for it; a change in the symptom count between administrations has no established meaning. What matters clinically is the positive predictive value, which depends heavily on how common bipolar disorder is in the setting. In a psychiatric outpatient sample, only 23.5% of MDQ-positive patients had bipolar disorder while 27.6% had borderline personality disorder, and false positives outnumbered false negatives by more than three to one (Zimmerman et al., 2010). In low-prevalence settings such as general practice, most positive screens will not be bipolar disorder. Treat a positive result as a prompt for a careful diagnostic interview, and a negative result as reassuring for bipolar I but not for bipolar II.
What it cannot tell you.
The MDQ's central weakness is bipolar II disorder: with the recommended three-part rule it misses a majority of bipolar II cases, and in community samples it has detected almost none. Dropping the co-occurrence and impairment criteria raises sensitivity substantially but at the cost of specificity, so different criterion variants in the literature are not directly comparable.
False positives are common in borderline personality disorder, PTSD, substance-use disorders, ADHD and eating disorders, all of which share irritability, impulsivity and mood instability with hypomania. The yes/no lifetime format relies on recall and insight, is open to over- and under-reporting, and cannot distinguish a single medication- or substance-induced episode from a recurrent illness. The instrument is copyrighted, so it cannot be embedded in electronic systems without permission.
Get our case study pack by email, with practical examples from clinical teams.
We will only use your details to send relevant Aisel resources. Unsubscribe at any time.
We value your privacy
We use cookies to analyse site usage and improve your experience. Analytics and embedded media (e.g. YouTube) only load if you accept. Read our cookie policy.