Abstract
OBJECTIVES: Alcohol consumption affects sleep both in healthy populations and in patients with alcohol use disorder (AUD). However, sleep has typically not been considered within AUD pharmacotherapy trials. We used data from a completed gabapentin clinical treatment trial to explore the medication's effect on patient-rated insomnia measured by a standard insomnia rating (Insomnia Severity Index [ISI]) and whether this influenced gabapentin's effects on alcohol consumption. METHODS: This study included 90 individuals with current Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition AUD criteria reporting current or past alcohol withdrawal. Participants were assigned to placebo or gabapentin (up to 1200 mg/day) for a 16-week randomized controlled trial with percent heavy drinking days (PHDD) and percent abstinent days (PDA) as outcomes. Utilizing mixed-effects models, this study assessed medication effects on ISI over the trial. We then examined the interaction of baseline ISI and medication on drinking. Finally, given our previous finding of alcohol withdrawal influencing gabapentin efficacy, we added change in ISI as a potential "moderator" of the interaction of medication effects and alcohol withdrawal on drinking. RESULTS: Sleep (ISI) improved more in those treated with gabapentin (60.6% reduction) compared with placebo (37.8% reduction; P = 0.013). Higher baseline ISI predicted drinking in gabapentin-treated individuals (lower PHDD [ P = 0.026] and higher (PDA [ P = 0.047]). ISI was an independent predictor of PHDD decrease and PDA increase ( P < 0.001; P = 0.002), but this did not significantly moderate gabapentin's effectiveness. CONCLUSIONS: Although gabapentin positively impacts both alcohol use and sleep, its effect on drinking is not fully dependent on sleep improvement, implying a direct biological mechanism on alcohol use.