Background: Seasonality of mania has been previously reported in several world regions. A spring and/or summer peak has been the most frequent finding, correlating to climatic variables, especially luminosity. There are, however, no South American studies on this association.
Methods: The charts of 269 manic patients admitted from 1996 to 2000 in a psychiatric hospital at Belo Horizonte, Brazil, were reviewed. Seasonality was assessed with Cosinor Analysis. Correlations of the rate of admissions for mania to climatic variables were performed, including lagged and differenced data.
Results: A circannual pattern was evident, with a late winter-spring peak and a late summer-autumn trough. The rate of admissions for mania correlated positively to: (a) average index and previous months' hours of sunshine, and (b) differenced mean temperature; and negatively to: (a) index and previous months' rainfall, (b) index months' relative humidity, and (c) previous months' duration of days and mean temperature. Altogether, climatic variables explained 23.7% of the variance in the rate of admissions for mania.
Limitations: This was a retrospective study conducted in a single institution.
Conclusions: The fact that climatic variables are associated to the course of bipolar disorder even in subtropical regions indicate that this effect may be more subtle and extent than previously thought. Further exploration of the biological mechanisms of this association is necessary.