3.2 The Mood Disorder Questionnaire (MDQ)

The MDQ, developed by Hirschfeld, Williams, Spitzer, and colleagues and published in 2000 in the American Journal of Psychiatry, remains the most widely recognized bipolar screening instrument in adult populations (Hirschfeld et al., 2000, American Journal of Psychiatry).

**Structure. **The MDQ is a brief self-report instrument with three sections: (1) 13 yes/no items assessing lifetime manic/hypomanic symptoms, (2) a single question about whether endorsed symptoms occurred during the same time period, and (3) a question about the degree of functional impairment caused by symptoms. A positive screen requires endorsing 7 or more of the 13 symptoms, confirming co-occurrence, and reporting at least moderate functional impairment (Hirschfeld et al., 2000, American Journal of Psychiatry).

**Psychometric Properties. **The MDQ's diagnostic performance varies substantially depending on the population in which it is used, which is itself an important clinical lesson about screening tool interpretation. The original 2000 validation in psychiatric outpatients showed sensitivity of 73% and specificity of 90%. However, subsequent meta-analytic data reveals significant heterogeneity across settings. A comprehensive meta-analysis examining 21 independent studies found that at the standard cutoff of 7 or more symptoms, the MDQ demonstrates a summary sensitivity of 0.62 and a summary specificity of 0.85 (Wang et al., 2015, Depression and Anxiety).

The MDQ's primary strength is its high specificity: it is effective at identifying classical, severe Bipolar I disorder and ruling out true negatives. Its critical limitation is poor sensitivity for Bipolar II disorder. Because the MDQ requires endorsement of "moderate or severe" functional impairment, and hypomanic episodes by definition do not cause severe functional impairment, Bipolar II patients routinely screen negative (Wang et al., 2015, Depression and Anxiety).

A particularly important finding from the Wang meta-analysis: MDQ sensitivity was 0.76 when comparing known bipolar patients against known unipolar depression patients, but dropped to 0.37 in studies excluding patients with established bipolar diagnoses (Wang et al., 2015). This indicates that the MDQ functions best as a case-finding instrument among established mood disorder patients, not as a general population screener.

**The False-Positive Problem. **The MDQ generates clinically significant false positives in patients with borderline personality disorder, PTSD, and ADHD. Research by Zimmerman and colleagues demonstrated that in outpatient psychiatric settings, MDQ-positive patients are equally likely to have BPD as they are to have true bipolar disorder (Zimmerman et al., 2010/2019, Comprehensive Psychiatry). A 2019 study by Balling, Chelminski, Dalrymple, and Zimmerman (N=476 depressed patients assessed via the SCID) found that a specific triad of MDQ items, consisting of elevated mood, increased goal-directed activity, and episodic co-occurrence, better differentiates bipolar disorder from BPD than the full 13-item scale (Balling et al., 2019, Comprehensive Psychiatry).

A 2024 genome-wide association study (N = 11,568-19,859) examined the genetic underpinnings of MDQ responses. This type of study scans the entire human genome across thousands of participants to identify which genetic variants are associated with a trait. The results were striking: the MDQ had a positive predictive value of only 0.29 for self-reported bipolar disorder. Lifetime manic symptoms endorsed on the MDQ showed the highest genetic correlation (shared genetic architecture) with PTSD, rather than bipolar disorder, followed by significant correlations with ADHD and MDD (Mundy et al., 2023, American Journal of Medical Genetics Part B). In other words, the MDQ may be capturing general emotional dysregulation and distress rather than specifically identifying bipolar neurobiology.

**Clinical Integration. **The MDQ is copyrighted (Hirschfeld, 2000) but widely available for clinical use without per-use licensing fees. The tool is best deployed in psychiatric outpatient settings among patients with established mood disorders, where its high specificity is most useful. It should not be used as the sole screening instrument when Bipolar II is a primary concern.

3.3 The Hypomania Checklist-32 (HCL-32)

Developed by Jules Angst and colleagues to address the MDQ's insensitivity to Bipolar II, the HCL-32 is a 32-item self-report questionnaire that assesses a broader array of hypomanic symptoms. Statistical analysis of the questionnaire's structure (principal component analysis) consistently reveals two underlying dimensions: an "active/elevated mood" dimension and a "risk-taking/irritable behavior" dimension (Angst et al., 2005, Journal of Affective Disorders).

**Psychometric Properties. **Meta-analytic data from Youngstrom and colleagues (2019, Psychiatry Research), analyzing 9 studies, found that the HCL-32 has a summary sensitivity of 82% (95% CI: 72%-89%) but a substantially lower specificity of 57% (95% CI: 48%-66%) (Youngstrom et al., 2019, Psychiatry Research). A comprehensive network meta-analysis across 93 studies (n=62,291) confirmed pooled sensitivity of 0.75 with specificities of 0.63 to 0.64 (Sayyah, Delirrooyfard, & Rahim, 2022, Brazilian Journal of Psychiatry).

The HCL-32 represents a deliberate psychometric trade-off: significantly higher sensitivity for detecting hypomania, at the cost of substantially more false positives. In one study of antidepressant-resistant depressed outpatients, the HCL-32 identified subthreshold bipolarity in 55% of patients versus only 18% identified by the MDQ.

**Clinical Application. **The HCL-32 is best deployed in populations where the prior probability of bipolarity is already elevated, such as treatment-resistant depression cohorts in specialty clinics. Its low specificity (often around 50%-60%) makes it inappropriate as a mass screener in general primary care, where it would generate unacceptably high false-positive rates among individuals with ADHD, borderline traits, or agitated depression (Youngstrom et al., 2019, Psychiatry Research).

3.4 The Bipolar Spectrum Diagnostic Scale (BSDS)

The BSDS, developed by Ronald Pies and validated by Nassir Ghaemi, takes a different approach from symptom checklists. Instead of presenting isolated symptoms, it uses a narrative format: a descriptive paragraph outlining the typical experience of cyclothymic mood shifts. Patients rate how well the narrative describes their own life experience on a sentence-by-sentence basis (Ghaemi et al., 2005, Journal of Affective Disorders).

**Psychometric Properties. **In the original validation study (68 bipolar patients and 27 unipolar depressed patients), the BSDS demonstrated sensitivity of 0.76, with approximately equal sensitivity for Bipolar I (0.75) and Bipolar II/Bipolar NOS (0.79). Specificity was 0.85 (Ghaemi et al., 2005, Journal of Affective Disorders). A 2022 meta-analysis of 93 studies reported pooled sensitivity of 0.71 and pooled specificity of 0.73.

The BSDS's clinical utility lies in its ability to bypass a patient's lack of psychiatric vocabulary. Because the narrative format normalizes the experience of mood cycling, it captures patients who might not recognize or endorse clinical terms like "racing thoughts" or "elevated mood" when presented in isolation on a checklist. The BSDS is an especially valuable complementary tool when evaluating patients with suspected treatment-resistant unipolar depression who may harbor a subtle bipolar spectrum condition. The developers permit free clinical use (Pies).

3.5 The Rapid Mood Screener (RMS)

The RMS, developed by Roger McIntyre and colleagues in 2021, represents the most pragmatically designed recent addition to bipolar screening. It consists of 6 items requiring simple yes/no responses, deliberately incorporating longitudinal risk factors alongside manic symptoms (McIntyre et al., 2021, Current Medical Research and Opinion).

The six items ask patients to endorse: (1) at least 6 different periods of deep depression, (2) problems with depression before age 18, (3) having to stop or change an antidepressant because it caused irritability or hyperactivity, (4) periods of talkativeness with racing thoughts, (5) periods of feeling unusually happy, outgoing, or energetic, and (6) periods of needing much less sleep than usual (McIntyre et al., 2021).

**Psychometric Properties. **At a cutoff of 4 or more endorsed items, the RMS demonstrates sensitivity of 0.88 and specificity of 0.80, with a positive predictive value of 0.80 and negative predictive value of 0.88 in validation populations (McIntyre et al., 2021). Note that these PPV figures were derived from controlled validation cohorts; in general primary care populations where the base rate of bipolar disorder is lower, the PPV would be correspondingly reduced, consistent with the Bayesian principles discussed in Section 3.1. In direct meta-analytic comparison, the RMS demonstrated statistically significant superiority in diagnostic accuracy over the MDQ for detecting Bipolar I (Sayyah, Delirrooyfard, & Rahim, 2022, Brazilian Journal of Psychiatry).

**Clinical Adoption. **A 2023 nationwide survey of 200 healthcare providers, published by Thase and colleagues in The Primary Care Companion for CNS Disorders, found that 81% of clinicians preferred the RMS over the MDQ due to brevity, practicality, and ease of scoring (Thase et al., 2023, Primary Care Companion for CNS Disorders). The RMS uses 60% fewer items than the MDQ and takes less than two minutes to complete, making it particularly suited to time-constrained primary care encounters.

**Important limitation: **The RMS is a newer tool with a smaller longitudinal evidence base than the MDQ. It was primarily validated for discriminating Bipolar I from MDD and does not provide a comprehensive symptom map. The RMS development and initial validation study were funded by Allergan (now AbbVie), with AbbVie employees among the co-authors; the clinician preference survey by Thase et al. (2023) was also industry-supported. These industry relationships do not invalidate the psychometric data but should be considered when evaluating the promotional enthusiasm around this tool relative to established alternatives.

Clinical Pearl: Matching the Screener to the Clinical Question

When the primary question is "Does this depressed patient have unrecognized Bipolar II?", the RMS may be insufficient on its own because it was primarily validated for Bipolar I detection. For this specific clinical scenario, which is the most common and consequential diagnostic challenge described in this module, the BSDS may be a more appropriate choice given its near-equal sensitivity for Bipolar I (0.75) and Bipolar II/BD-NOS (0.79). A practical two-step workflow for primary care: use the RMS as a rapid initial screen for general bipolar risk, and add the BSDS when Bipolar II is specifically suspected (e.g., treatment-resistant depression, atypical features, family history).

PsychHQ

Screening Instruments: MDQ, HCL-32, BSDS, RMS