A downscaled practical measure of mood lability as a screening tool for bipolar II

Abstract

Background: Current data indicate a strong association between Cyclothymic temperament (and its more ultradian counterpart of mood lability) and Bipolar II (BPII). Administration of elaborate measures of temperament are cumbersome in routine practice. Accordingly, the aim of the present analyses was to test if a practical measure of mood lability was unique to BPII, in comparison with major depressive disorder (MDD).

Methods: Using the Structured Clinical Interview for DSM-IV Axis I Disorders, Clinician Version as modified by us [J. Affect. Disord. 73 (2003) 33; Curr. Opin. Psychiatry 16 (2003) S71], we interviewed 62 consecutive BPII outpatients, as well as their 59 MDD counterparts during a major depressive episode (MDE). Hypomanic symptoms during MDE were systematically assessed: three or more such symptoms defined depressive mixed state (DMX3) on the basis of previous work by us [J. Affect. Disord. 73 (2003) 113]. A downscaled definition of trait mood lability was adapted from Akiskal et al. [Arch. Gen. Psychiatry 52 (1995) 114] and Angst et al. [J. Affect. Disord. 73 (2003) 133], requiring a positive response to one of two queries on whether one is a person with frequent "ups and downs" in mood, and whether such mood swings occur for no reason. The patients selected for inclusion had not received neuroleptics and antidepressants for at least 2 weeks prior to the index episode, they were free of substance and alcohol abuse, and did not meet the DSM-IV criteria for borderline personality disorder (BPD). Associations between mood swings and clinical variables were tested by logistic regression (STATA 7).

Results: Mood swings were endorsed by 50.4% of the entire sample: 62.9% of BPII and 37.2% of MDD (p = 0.0047). This practical measure of mood lability was significantly associated with BPII, lower age at onset, high depressive recurrences, atypical features, and DMX3. When controlled for number of major affective episodes, mood swings were still significantly associated with BP-II. Sensitivity and specificity of mood swings for predicting BPII were 62.9% and 62.7%, respectively.

Limitation: The low specificity of trait mood lability for BPII diagnosis is probably due to the fact that we used a downscaled simplified measure of this trait.

Conclusions: On the other hand, the relatively high sensitivity of our downscaled measure of mood lability for predicting BPII supports its usefulness as a screening tool for this diagnosis. The lack of association between self-reported mood lability and number of major mood episodes indicates that such lability does not reflect the perception of history of frequent episodes, and that it has some validity as a trait indicator. Given that our sample excluded patients meeting the DSM-IV criteria for BPD, contradicts the opinion of the latter manual that such mood lability represents its pathognomonic characteristic that distinguishes it from BPII. The bipolar nature of mood lability is further supported by significant associations with external validating criteria for bipolarity. Overall, these data indicate that in the differential diagnosis between MDD and BPII, trait mood lability favors the latter at a significant statistical level.