Results from the current study revealed that depression symptom expression in BP II was remarkably similar to that in BP I and unipolar depression. A visual inspection of the item response functions (see & ) indicated that, with the exception of suicidal ideation in BP I, symptom expression was virtually identical in the three conditions. Although additional research is necessary to confirm such a conclusion, it is notable that these data are in keeping with recent arguments (Cuellar et al., 2005
; Joffe et al., 1999
; Swann, 1994
) that depression may be understood as a clinical phenomenon that is relatively consistent across mood disorders. For example, using IRT methodology, our research group previously reported very few differences between BP I and unipolar depression, and the differences that did emerge fell into the small-to-medium effect size range (Weinstock et al., 2009
). Nevertheless, as suggested by Brugue et al. (2008)
, it is also important to note that there may be clinical features that differentiate BP II from BP I and unipolar depression that are otherwise not DSM-IV symptoms of depression (e.g., anxiety) or that were not assessed in the NESARC (e.g., mood reactivity).
Indeed, there is a growing body of literature focused on atypicality of depression in BP II (Akiskal & Benazzi, 2005
), and it is important to note that current study cannot be generalized to form conclusions about depressive subtypes. To the extent that we did have data on atypical symptoms (i.e., hypersomnia and hyperphagia), we could not evaluate these symptoms separately from their counterparts (i.e., insomnia and loss of appetite) due to the assumption of local independence described in the Methods. Of note, however, is that frequency of endorsement of the component parts of these items (see ) reveals highest rates of atypical symptom endorsement in BPI and not BPII. Although these endorsement rates are unadjusted, they nevertheless provide some preliminary descriptive data that can be used to inform future research on depressive features specifiers across the mood disorders. Such research will be necessary in order to fully characterize any differential phenomenology of depression in BP II.
As noted above, current study analysis revealed a significant difference between BP II and BP I in the endorsement of suicidal ideation/attempt. The direction of this effect suggested that individuals with bipolar I depression were more likely to experience suicidal ideation at lower levels of depression severity, and thus more frequently than those with bipolar II depression. This finding runs counter to recent arguments that suicidal ideation and behaviors occur more frequently in bipolar II depression (Rihmer & Pestality, 1999
). Also counter to the argument that suicide risk may be greatest in BP II, there were no significant differences in suicidal ideation/attempt between bipolar II and unipolar depression, and our prior IRT research evaluating bipolar I versus unipolar depression revealed greater likelihood of endorsement of suicidal ideation among those with BPI (Weinstock et al., 2009
). Taken together, this pattern of findings suggests that the likelihood of endorsing suicidal ideation/attempt may actually be highest in bipolar I relative to bipolar II and unipolar depression, and no different between bipolar II and unipolar depression.
One possible explanation for this discrepancy from the published literature is that IRT analysis adjusts for overall depression severity, whereas most prior research comparing mood disorders has not (cf., Rihmer & Pestality, 1999
). It is also important to note that mixed data concerning suicidality across the mood disorders may be related to how suicidal ideation and behaviors are measured in the extant literature (MacQueen & Young, 2001
; Valtonen et al., 2009
). Nevertheless, a visual inspection of the item response functions suggested that, at an average level of depression severity
(i.e., latent trait = 0), the probability of endorsing suicidal ideation was approximately 76% for individuals with bipolar I depression in comparison to approximately 70% for individuals with bipolar II depression (see ). Although we set a minimum effect size threshold for interpretation of clinical significance in the current study, it is notable that the effect size for the difference in suicidal ideation/attempt between BPII and BPI fell into the small effect size range. Such a subtle difference may not directly guide clinical decision making in applied settings (e.g., differential diagnosis), yet it is nevertheless important from a public health and empirical perspective. Perhaps most striking are the high rates of suicidal ideation endorsement in both
groups, which clearly warrant continued clinical and empirical attention.
When interpreting the findings above, it is important to acknowledge study limitations. First, in order to be included in the data analysis, individuals must have endorsed either depressed mood or anhedonia. In the NESARC, the remaining DSM-IV depression symptoms were not assessed if one or both of these symptoms had not been endorsed. Thus, given that all of the current study sample endorsed either depressed mood or anhedonia, and most endorsed both, it would not be terribly meaningful, from a statistical perspective, to conduct DIF analyses on these items. It is also important to acknowledge that current study analyses did not account for clinical course characteristics (e.g., length of illness, rates of depressive or (hypo)manic episode recurrence or hospitalization, or medication regimen) that might have potentially influenced symptom profiles. Indeed, the data used in this study were cross-sectional, allowing for the evaluation of endorsement patterns of lifetime depressive symptoms only. Future research will be necessary in order to evaluate differences in depression symptom expression across mood disorders over longitudinal course of illness. Finally, it is important to reiterate that current study analyses cannot be used to form conclusions regarding depressive subtypes, nor can these data be used to form conclusions regarding bipolar mixed states, when individuals present with concurrent depression and (hypo)mania.
In conclusion, by addressing several limitations of the existing research, and employing a methodology grounded in Item Response Theory, current study results add to a small, yet growing literature focused on the phenomenology of bipolar II depression. Consistent with some recent assertions (Cuellar et al., 2005
; Joffe et al., 1999
; Swann, 1994
), data suggested that DSM-IV depression symptom expression in BP II was very similar to that in BP I and unipolar depression. Although the frequency of suicidal ideation/attempt was higher in BP I relative to BP II, it should be noted that the effect size for this difference was small and that rates of endorsement in both groups were quite high and deserve continued clinical and empirical attention. Future research that continues to explore depression features specifiers in BP II remains an important area of inquiry; however, it is imperative that this research account for underlying depression severity when evaluating the distinct versus shared characteristics of BP II relative to its BP I and unipolar counterparts.