Concurrent Changes in Self-Reported Sleep Disturbance During At-Home Ketamine-Assisted Therapy for Depression: A Retrospective Analysis of 13,963 Adults

Abstract Background Sleep disturbances are a significant public health burden affecting 50 to 70 million adults in the United States and contributing to impaired cognition, accelerated disease progression, and increased mortality risk. 1,2 Available pharmacological treatments carry notable safety and tolerability risks, while access to the leading evidence-based behavioral intervention, cognitive behavioral therapy for insomnia, remains limited by provider shortages and geographic barriers. 3 Although preliminary evidence suggests ketamine may influence sleep through neurobiological mechanisms such as BDNF-mediated synaptic plasticity, 4 and although sleep and depression are bidirectionally linked through both neurobiological and cognitive pathways, 5,6 large-scale real-world data on sleep outcomes remain limited, particularly for at-home, remotely supervised delivery models. Sleep disturbances also frequently co-occur with depression, 5 making ketamine-treated populations clinically relevant for sleep outcome investigation.

Objective: This study describes sleep disturbance outcomes among adults with comorbid sleep disturbances who received at-home ketamine-assisted therapy for depression through Mindbloom, a telehealth ketamine therapy platform.

Methods: We conducted a retrospective cohort analysis of 13,963 adults with moderate-to-severe baseline sleep disturbance (PHQ-9 item 3 score ≥ 2; mean baseline = 2.55) who received guided at-home sublingual or subcutaneous ketamine therapy for depression. Sleep disturbance was assessed using PHQ-9 item 3 (a sub-item of a depression screening instrument, not a validated standalone sleep measure) at post-session 2 (n = 13,963), post-session 4 (n = 10,556), and post-session 6 (n = 1,340). Response was defined as ≥ 1-point improvement from baseline.

Results: Study-defined response (≥ 1-point improvement in PHQ-9 item 3) was achieved by 67.4% of participants at post-session 2 (n = 13,963), 75.1% at post-session 4 (n = 10,556), and 76.8% at post-session 6 (n = 1,340). Among six-session completers (mean baseline = 2.56), mean PHQ-9 item 3 score declined to 1.31 (mean change = 1.25; Cohen’s d = 1.32 [95% CI: 1.25, 1.39]), representing a 48.8% mean improvement; in the full cohort (n = 13,963), mean improvement from baseline was 38.8% at post-session 2. 4.3% of participants reported worsened sleep at post-session 2. Side effects were reported by 4.4% at post-session 2, 5.0% at post-session 4, and 3.7% at post-session 6.

Conclusions: In this large retrospective cohort, 67.4% of adults with moderate-to-severe comorbid sleep disturbances showed study-defined response (≥ 1-point improvement in PHQ-9 item 3) by post-session 2 of at-home ketamine-assisted therapy, increasing to 76.8% among the 1,340 (9.6%) six-session completers. Because sleep and depression are bidirectionally linked and this cohort was treated for depression, observed sleep changes cannot be disentangled from concurrent mood improvement in this uncontrolled design. Randomized controlled trials with validated sleep instruments and depression-independent sleep endpoints are needed to confirm these findings and to characterize the mechanisms and durability of any observed improvements.