Reducing Suicidal Ideation Through Insomnia Treatment

Primary Aim: We will assess the effect of treating insomnia with hypnotic medication on the
intensity of suicidal ideation in depressed outpatients with insomnia and suicidal ideation.

-Hypothesis 1. Treatment of depressed, insomniac and suicidal outpatients with open-label
fluoxetine (FLX) and blinded zolpidem controlled release (ZOL) will reduce suicidal ideation
more than treatment with FLX and blinded placebo.

Secondary Aim: We will examine whether reduced suicidal ideation in depressed insomniacs is
mediated through reduced dysfunctional beliefs about sleep, reduced hopelessness, or fewer
nightmares.

- Hypothesis 2a. Reduction in suicidal ideation will be mediated through reductions in
dysfunctional beliefs about sleep.

- Hypothesis 2b. Reduction in suicidal ideation will be mediated through reductions in
hopelessness.

- Hypothesis 2c. Reduction in suicidal ideation is mediated through fewer nightmares.

Tertiary Aim: We will confirm findings from our prior pilot studies that treatment of
insomnia in depressed insomniacs leads to improvements in health-related quality of life,
especially in women.

Exploratory Aim: We will archive actigraphy data to permit future examination to confirm our
preliminary data that actigraphic activity decreases as suicidal ideation resolves.

Overview of the Need for and Management of a Collaborative Application: The sample sizes
required to satisfy the Aims are relatively large, necessitating the pooled recruiting
resources of 3 sites. Georgia Regents University (GRU) will serve both as the
coordinating/data management site, as well as a recruiting site, with Duke and Wisconsin as
recruiting sites. Project management will be coordinated through an Executive Committee of
site principal investigators, under the supervision of a Data and Safety Monitoring Board.

Impact on the Field: This application has the potential to change providers' practice in the
approach to treating insomnia in depressed patients with mild-moderate suicidal ideation. It
may also reveal the mechanisms whereby insomnia increases the risk for suicidal ideation and
behavior, and begin to examine whether there is an actigraphic "signature" for reductions in
suicidal ideation. When these lessons are applied to the clinical world, they can be applied
with low cost.

Inclusion Criteria:

- Persons 18-65 years of age

- Persons with confirmed DSM-IV diagnosis of MDE by SCID

- Persons with Research Diagnostic Criteria diagnosis of insomnia

- Persons free of all psychotropic medications for one week before baseline assessment,
except that prior FLX treatment will require 4 weeks of abstinence, and MAOIs will
require 2 weeks of abstinence.

- Persons with Scale for Suicide Ideation (SSI) scores >2

- Persons with Hamilton Rating Scale for Depression (HRSD24) score >20

- Persons with Mini Mental State Exam (MMSE) score >24

- Persons with Insomnia Severity Index (ISI) score > 7

- Persons with habitual sleep latency > or = 30 minutes or wake time in the middle of
the night of > or = 30 minutes, and sleep efficiency < 85%

Exclusion Criteria:

- Non-English speaking, reading, writing persons

- Persons who pose imminent danger to self or others

- Persons with severe suicidal ideation (C-SSRS Suicidal Ideation Score >3)

- Persons with clinical diagnosis of dementia

- Persons with active or past diagnosis of alcohol or substance abuse, bipolar disorder,
schizophrenia per the SCID

- Persons who screen positive for moderate-severe sleep apnea (AHI >10) or a prior sleep
laboratory-confirmed diagnosis of a primary sleep disorder, such as sleep apnea or
periodic limb movement disorder.

- Persons with BMI > 50

- Persons with a self-reported history of napping > 2 times per week (as these are
associated with sleep apnea and periodic limb movement disorder in depressed
insomniacs)

- Persons who might be harmed by exposure to hypnotics, including pregnant women and
patients with respiratory conditions