Staging bipolar disorder: what data and what models are needed?

Although bipolar disorder is increasingly recognised as a spectrum of multisystem disorders (ie, bipolar disorders), proposed staging models and theories of bipolar disease progression often fail to incorporate longitudinal data or data from multiple domains of dysfunction. We propose that bipolar disorders are best thought of as syndromes, with different trajectories of development and progression for various symptoms and demographic groups. This inherent complexity might be better suited to non-traditional modelling techniques, potentially derived from chaos theory. In this Personal View, we propose an allostatic load framework to account for biomarkers of physiological symptom progression. We then suggest integration of two potential domains of biobehavioural markers: sleep and wake and circadian rhythm regulation and the behavioural activation system. A satisfactory model should account for the effects of developmental stage as well as demographic characteristics, including but not limited to sex, culture, ethnicity, and socioeconomic status. The ultimate goal of a staging model has to be to inform the development of targeted, stage-appropriate interventions to reduce the substantial burden of bipolar disorders on individuals and societies.