Diagnostic guidelines for bipolar depression: a probabilistic approach

Objectives: There are currently no accepted diagnostic criteria for bipolar depression for either research or clinical purposes. This paper aimed to develop recommendations for diagnostic criteria for bipolar I depression. Methods: Studies on the clinical characteristics of bipolar and unipolar depression were reviewed. To identify relevant papers, literature searches using PubMed and Medline were undertaken. Results: There are no pathognomonic characteristics of bipolar I depression compared to unipolar depressive disorder. There are, however, replicated findings of clinical characteristics that are more common in both bipolar I depression and unipolar depressive disorder, respectively, or which are observed in unipolar-depressed patients who 'convert' (i.e., who later develop hypo/manic symptoms) to bipolar disorder over time. The following features are more common in bipolar I depression (or in unipolar 'converters' to bipolar disorder): 'atypical' depressive features such as hypersomnia, hyperphagia, and leaden paralysis; psychomotor retardation; psychotic features, and/or pathological guilt; and lability of mood. Furthermore, bipolar-depressed patients are more likely to have an earlier age of onset of their first depressive episode, to have more prior episodes of depression, to have shorter depressive episodes, and to have a family history of bipolar disorder. The following features are more common in unipolar depressive disorder: initial insomnia/reduced sleep; appetite, and/or weight loss; normal or increased activity levels; somatic complaints; later age of onset of first depressive episode; prolonged episodes; and no family history of bipolar disorder. Conclusions: Rather than proposing a categorical diagnostic distinction between bipolar depression and major depressive disorder, we would recommend a 'probabilistic' (or likelihood) approach. While there is no 'point of rarity' between the two presentations, there is, rather, a differential likelihood of experiencing the above symptoms and signs of depression. A table outlining draft proposed operationalized criteria for such an approach is provided. The specific details of such a probabilistic approach need to be further explored. For example, to be useful, any diagnostic innovation should inform treatment choices.