Orexin Receptor Antagonists as Modulators of Threat Sensitivity in Individuals With Alcohol Use Disorder

Percentage of heavy drinking days (PHDD) was calculated each week for all four weeks of the trial. A heavy drinking day was defined using NIH criteria: 5+ drinks for males and 4+ drinks for females in a single day. We conducted a multilevel mixed model with PHDD from week 1 to 4 as the within-subjects variable and treatment arm as the between subjects variable.

Time frame: Change from baseline to post-treatment, over the course of 4 weeks.

Population: All 64 participants randomized to treatment were analyzed.

Startle EMG responses were collected during the well-validated No-Predictable-Unpredictable threat paradigm. The task includes within-subjects conditions and raw EMG responses were averaged for each condition. To quantify the difference between threat and no-threat periods, we calculated a standardized residual score for unpredictable threat (U-threat) by saving the variance leftover (i.e., the amount of variability in a dependent variable \[DV\] that is not explained by an independent variable \[IV\]) in a simple linear regression, where the no-threat EMG startle average (IV) was entered to predict the U-threat EMG startle average (DV). The U-threat residual score was used as the primary variable. It has a mean of zero and higher scores reflect greater startle reactivity to U-threat. U-threat residual scores were calculated for the baseline session and the acute challenge.

Time frame: Change from baseline to 2 hours post-ingestion of an acute dose of suvorexant.

Population: A total of 57 participants were included. Missing data was due to attrition, poor quality startle eyeblink data, or technical errors during startle data collection.

Startle EMG responses were collected during the well-validated No-Predictable-Unpredictable threat paradigm. The task includes within-subjects conditions and raw EMG responses were averaged for each condition. To quantify the difference between threat and no-threat periods, we calculated a standardized residual score for unpredictable threat (U-threat) by saving the variance leftover (i.e., the amount of variability in a dependent variable \[DV\] that is not explained by an independent variable \[IV\]) in a simple linear regression, where the no-threat EMG startle average (IV) was entered to predict the U-threat EMG startle average (DV). The U-threat residual score was used as the primary variable. It has a mean of zero and higher scores reflect greater startle reactivity to U-threat.

Time frame: Change from baseline to post-treatment, up to 2 months.

Population: A total of 54 participants were included. Missing data was due to attrition, poor quality startle eyeblink data, or technical errors during startle data collection.

Activation parameter estimates (arbitrary units) were extracted from anatomical bilateral aINS masks for each individual, pre and post treatment. Specifically, activation parameter estimates were extracted from anticipation of unpredictable threat \> anticipation of no-threat individual contrast maps for each scanning session. Greater values reflect greater activation in the bilateral insula during the anticipation of unpredictable threat. A repeated measures analysis of variance was used to test the session (time 1 vs. time 2) by treatment arm interaction on bilateral aINS activation during unpredictable threat. Significance was set at p \<.05

Time frame: Change from baseline to post-treatment, up to 2 months.

Population: A subset of participants were randomized to complete functional magnetic resonance imaging (fMRI) pre- and post-treatment. Only individuals with two scanning sessions were analyzed.