The investigators propose to explore the link between bipolar disorder, anxiety, and suicide by investigating intertemporal discounting in depressed, suicidal patients with bipolar I and II disorder who have various levels of anxiety. The investigators will determine the effect of anxiety on their intertemporal discounting (small rewards now compared to larger rewards later) in a decision-making paradigm and investigate the associated functional neuroanatomy using functional magnetic resonance imaging (fMRI).

Among the various psychological properties of the decision making process is temporal discounting, which is a decrease in the subjective value of a good as a function of the amount of and delay to reward. The ability to decide between immediate versus future rewards depends on self-control and consideration of the future. We can reconceptualize suicide as intertemporal discounting with an interaction between cognition and mood. Someone contemplating suicide weighs the time value of costs and benefits with shifting negative and positive valence systems. To the best of our knowledge, no one has assessed the interaction between mood symptoms, anxiety, and their impact on the temporal discounting paradigm in bipolar patients. If we better understood the difference between anxious, suicidal and non-anxious, suicidal bipolar patients, we could design more effective interventions to prevent this tragic outcome. We propose a novel paradigm to explore the link between bipolar disorder, anxiety, and suicide. If we conceptualize suicidal behavior and death by suicide as decisions, then it makes sense to examine key aspects of decision making in these patients. In particular, we can examine how mood, anxiety, and suicidal ideation and behaviors arise from patterns of decision making, along with neural correlates of decision-making, as assessed with fMRI.

Inclusion Criteria:

1. Meets DSM-IV criteria for BD I and II current depressive episode
2. Able to give written informed consent
3. Age > to 18 years and < 65 years
4. Currently suicidal as defined by a MADRS suicide item score of > 3 or previous history of serious suicidal ideation that required hospitalization.
5. All subjects need to have normal hearing and normal/corrected-to-normal vision.

Exclusion Criteria:

1. Medical illness or non-psychiatric medical treatment that would likely interfere with study participation
2. Neurologic disorder, previous ECT, or history of head trauma (i.e. known structural brain lesion potentially confounding MRI results)
3. Substance abuse within the past 3 months or current substance dependence (confirmed by MINI)
4. Left-handedness
5. Contraindications to MRI (metallic implants, claustrophobia, etc.)
6. Subjects who need urgent psychiatric care requiring hospitalization (evaluated by clinicians).

We will recruit 30 currently depressed patients (with DSM-IV bipolar I and II disorder who are currently or previously suicidal). Enrollment will be limited to adults aged 18-65.