Rapid-cycling bipolar disorder. An overview of research and clinical experience

Although many studies of RCBD have been reported over the last 2 decades, knowledge remains limited. Higher incidence in women is the sole clearly replicated finding in most studies. This finding might be mediated by cyclothymia, a temperament that is of higher prevalence in women and that might be considered as a normal variant of RC. Many questions remain unanswered. Review of putative risk factors, such as hypothyroidism and treatment with antidepressants, provides no conclusive answers. There is clinical evidence to implicate both factors. In principle, the thyroid connection can be approached rationally, yet there seems to be no relationship between thyroid status and response to thyroid augmentation. For this reason and given the potential risks of long-term thyroid use, this strategy should not be the first one to be tried in RC. Cumulatively, naturalistic studies over the past 30 years have strongly implicated antidepressants in switching and cycle acceleration, yet the double-blind, controlled, prospective studies that are needed to provide definitive answers are unlikely to be conducted for ethical reasons discussed in this article. Bipolar family history of RC probands appears indistinguishable from non-RC probands, indicating that most likely RCBD does not breed true. Although RC seems to be more lithium resistant with less likelihood of being symptom-free after 2 to 5 years of follow-up, many of these patients nonetheless have resolution of the RC course. There is no marked difference in suicide rates. An association of RC with bipolar type II, D-M-I pattern and those who switch into mania or hypomania on antidepressants is a provocative possibility: Antidepressants might introduce RC by first inducing a switch during a depressive episode, creating a D-M-I pattern, a pattern that is poorly responsive to lithium, which eventually degenerates into RC. Again, this sequence might be mediated by the high prevalence of cyclothymia in bipolar II patients. Thus, data from phenomenology, family history, and long-term outcome do not support RC as a separate entity. RC appears to be a temporary complicated phase in the illness, not a stable feature. This was noted by Kraepelin: I think I am convinced that that kind of classification must of necessity wreck on the irregularity of the disease. The kind and duration of the attacks and the intervals by no means remain the same in the individual case but may frequently change, so that the case must be reckoned always to new forms. Data by Gottschalk et al testify to the chaotic mood swings of contemporary bipolar disorder. Moreover RC is seen in other medical diseases, such as epilepsy, in which patients have phases of increase in frequency of episodes (seizures) that become refractory to treatment. Further longitudinal prospective studies are required to understand the complexity of this intriguing phenomenon and to provide better treatments. Algorithms deriving from tertiary research or university-based clinical experience may not generalize to RC or otherwise treatment-resistant bipolar patients seen in more routine practice. Illness severity in RCBD generally precludes double-blind controlled investigations. Meanwhile, clinicians may rely on discontinuing antidepressants, maintaining patients on combined mood stabilizers--of which valproate is probably the most useful--and making judicious use of atypical neuroleptics. Benzodiazepines and alcohol (which produce withdrawal), caffeine, stimulants, exposure to bright light, and sleep deprivation during excited phases should be avoided. Thyroid and nimodipine augmentation can be considered in those with the most malignant course. These are patients who need the maximal support that their psychiatrist can provide them. Office visits must be arranged as the last appointment of the day.