Is the polarity of relapse/recurrence in bipolar-I disorder patients related to serum lithium levels? Results from an empirical study

Background: Preliminary evidence suggests that the polarity of relapse/recurrence (depressive vs. hypomanic/manic/mixed) in bipolar patients on lithium might be related to serum lithium levels. Methods: Polarity of episodes in 64 bipolar-I patients on lithium monotherapy during a prospective 18-month maintenance trial was predicted from (a) intra-individual oscillations of lithium levels overtime and from (b) absolute lithium levels preceding relapse/recurrence. Results: On an individual basis, depressive (vs. hypomanic/manic/mixed) episodes were mostly preceded by lithium levels above the individual means (p<0.001). Relapse/recurrence occurring at lithium levels above the overall mean serum level of 0.66 mmol/l was depressive (not hypomanic/manic/mixed) in most cases (odds-ratio 3.86, p=0.032). Lithium levels before depressive episodes were numerically higher than before hypomanic/manic/mixed episodes (0.769 +/- 0.242 vs. 0.675 +/- 0.262 mmol/l,p=0.13). Cox-regression including current lithium levels as time-dependent predictor essentially confirmed these results. Limitations: As patients were not randomized to specific lithium levels, potential confounders could not be completely ruled out. Furthermore, a closer than monthly assessment of both lithium levels and psychopathology would have been desirable to better understand the interplay between lithium levels and new mood episodes. Conclusions: The results indicate that within the currently accepted therapeutic range (0.4-1.1 mmol/l), the relative risk for depressive vs. hypomanic/manic/mixed relapses/recurrences may be associated with higher lithium levels. Therefore, lithium levels at the lower range of the therapeutic spectrum may be sufficient for the optimal prevention of depressive episodes whereas higher lithium levels within this range may be needed for optimal protection against manic/mixed episodes. (C) 2008 Published by Elsevier B.V.