Observational evidence and findings from clinical trials conducted for other reasons suggest
that lithium, a drug used for the treatment of bipolar disorder, and, to a lesser extent,
depression, may reduce rates of suicides and suicide attempts. However, this hypothesis has
not yet been adequately examined in a randomized clinical trial conducted specifically to
test lithium's efficacy in preventing suicides. This clinical trial fills this gap.
This study is feasible within the Department of Veterans Affairs (VA) because it is a large,
integrated health system with existing programs for identifying patients at risk for suicide
and delivering enhanced services. In VA, approximately 12,000 patients with depression or
bipolar disorder survive a suicide attempt or related behavior each year, and 15% of them
repeat within one year. Experimental treatment in this study will supplement usual care for
major depression or bipolar disorder, as well as VA's standard, enhanced management for
patients at high risk.
The investigators will recruit 1862 study participants, from approximately 30 VA Hospitals.
Participants will be patients with bipolar disorder or depression who have survived a recent
episode of suicidal self-directed violence or were hospitalized specifically to prevent
suicide. Randomly, half will receive lithium, and half will receive placebo. Neither the
patients nor their doctors will know whether a particular person has received lithium or
placebo. The treatment will be administered and the patients will be followed for one year,
after which patients will go back to usual care. Recruitment will occur over 3 years.
The investigators are primarily interested in whether lithium leads to increases in the time
to the first repeated episode of suicidal behavior, including suicide attempts, interrupted
attempts, hospitalizations specifically to prevent suicide, and deaths from suicide. In
addition, this study will allow us to explore whether lithium decreases the total number of
suicidal behaviors, and whether it has comparable effects on impulsive and non-impulsive
behaviors. If there is an effect of lithium, the investigators will be interested in whether
or not it could be attributed to improved control of the underlying mental health condition,
or, alternatively, whether it represents a direct effect of suicide-related behavior.