PROJECT SUMMARY
Insomnia is a prevalent health problem associated with adverse cardiovascular, metabolic, and mental health
outcomes. Previously proposed subtypes, based on traditional clinical measures, have poor reliability and validity
and have not proven useful for guiding insomnia treatment decisions. Based on a large base of preliminary data
from various domains and several investigative groups, we have identified a particular phenotype, insomnia with
short sleep duration (ISS), that is associated with increased risk for adverse health outcomes, greater
physiological hyperarousal as indicated by hypothalamic-pituitary-adrenal (HPA) axis activation, and worse
response to Cognitive-Behavioral Treatment for Insomnia (CBT-I). The proposed study represents the next
logical extension of our previous observations: To determine the efficacy of CBT-I in individuals with
ISS vs. Insomnia with normal sleep duration (INS) among adults with elevated blood pressure (BP), and
to examine the efficacy of trazodone among non-remitters to CBT-I. CBT-I is recommended as first-line
treatment for insomnia, and trazodone is a widely-prescribed but grossly understudied medication for insomnia.
In addition, our pilot data demonstrate differential efficacy of CBT-I and trazodone in ISS and INS: trazodone,
but not CBT-I, increases objective total sleep time (TST), and lowers BP and evening cortisol in ISS. We will
conduct a 4-site cohort study followed by a placebo-controlled RCT in 600 adults (≥18y) with insomnia. The
cohort study will examine the efficacy of CBT-l among individuals with ISS vs. INS phenotypes (n=300 each),
defined by polysomnographic (PSG) TST. The subsequent RCT will compare the efficacy of trazodone vs.
placebo among CBT-I non-remitters. Investigators at the 4 study sites (Hershey, Denver, Pittsburg, and Quebec)
have a long history of collaboration and successful completion of NIH-funded mechanistic and clinical trial
studies. Our primary outcome will be the insomnia remission at 8 weeks, defined as Insomnia Severity Index
(ISI) <8; ISI is the gold-standard self-report measure of insomnia symptoms. Secondary outcomes will include
ISI (continuous), objective (i.e., PSG and actigraphy) measures of sleep efficiency (in the CBT-I cohort study)
and TST, HBP, and evening cortisol (in the trazodone-placebo RCT). In exploratory analyses, we will test
whether changes in evening cortisol mediate the effect of trazodone on objective TST and HBP. Outcomes will
be assessed at 8 weeks and 6 months following the end of treatment to evaluate the durability of treatment
effects. Demonstrating a differential efficacy of CBT-I as a function of insomnia phenotype would aid the goals
of precision medicine, which directs therapy on the basis of clinical phenotypes and physiology as well as
genetics. Although CBT-I is recommended as first-line treatment for all adults with insomnia, finding a worse
response in the ISS phenotype will lead to reconsidering this current guideline in lieu of matching patients’
phenotype to treatment. Demonstrating the efficacy of trazodone among CBT-I non-remitters will fill an obvious
and important knowledge gap in insomnia treatment l as it pertains to its current wide-spread off-label use.
