Discontinuing lithium maintenance treatment in bipolar disorders: risks and implications

Authors

  • Ross J Baldessarini,

    1. The International Consortium for Bipolar Disorders Research, Consolidated Department of Psychiatry and Neuroscience Program, Harvard Medical School, Boston, MA, Bipolar and Psychotic Disorders Program, McLean Division of Massachusetts General Hospital, Belmont, MA 02478, USA;
    Search for more papers by this author
  • Leonardo Tondo,

    1. The International Consortium for Bipolar Disorders Research, Consolidated Department of Psychiatry and Neuroscience Program, Harvard Medical School, Boston, MA, Bipolar and Psychotic Disorders Program, McLean Division of Massachusetts General Hospital, Belmont, MA 02478, USA;
    2. Departments of Psychiatry and Psychology, University of Cagliari, and the Lucio Bini–Stanley Foundation International Mood Disorders Research Center, Cagliari, Sardinia;
    Search for more papers by this author
  • Adele C Viguera

    1. The International Consortium for Bipolar Disorders Research, Consolidated Department of Psychiatry and Neuroscience Program, Harvard Medical School, Boston, MA, Bipolar and Psychotic Disorders Program, McLean Division of Massachusetts General Hospital, Belmont, MA 02478, USA;
    2. Perinatal and Reproductive Psychiatry Research Program, Massachusetts General Hospital, Boston, MA 02114, USA
    Search for more papers by this author

Dr. Ross Baldessarini, MRC-McLean Hospital, Belmont, MA 02478 USA. Fax +1 617 855 3479; e-mail: rjb@mclean.org

Abstract

Objective: To review research findings on clinical effects of discontinuing lithium maintenance treatment.

Methods: Data from studies reported since 1970 plus our recent findings were updated.

Results: Discontinuing lithium maintenance treatment led to marked increases of early affective morbidity and suicidal risk. Gradual discontinuation markedly reduced early recurrences of mania or depression, did so more in bipolar II than I disorder patients, and also tended to reduce suicidal risk. Similar effects were found in pregnant and nonpregnant women after lithium discontinuation. Long-term retreatment with lithium following discontinuation was only slightly less effective than in initial trials.

Conclusions: Recurrences increased sharply soon after discontinuing lithium, but were markedly limited and not merely delayed, by slow discontinuation. Similar reactions may follow discontinuation of other drugs, evidently as responses to long-term pharmacodynamic adaptations. Discontinuing treatment is not equivalent to not-treating. Post-discontinuation relapse risk has implications for the design, management, and interpretation of protocols involving discontinuation of long-term treatments that should be considered in both clinical management and research.

Ancillary