Nocturnal intravenous gamma-hydroxybutyrate and sleep architecture in non-sedated intensive care unit patients: a polysomnographic experience - Summary - MDSpire
To describe sleep architecture, perceived sleep quality, and tolerability after overnight intravenous gamma-hydroxybutyrate (GHB) in awake, non-sedated ICU patients.
Approach:
Design and setting: Small, uncontrolled observational experience in a general ICU; 13 patients had interpretable ambulatory polysomnography recordings.
Intervention and assessment: Eight patients received intravenous GHB from 22:00 to 06:00; five did not. Sleep was scored using American Academy of Sleep Medicine rules, with morning sleep-quality ratings, nurse assessments, and electrolyte measurements.
Key Findings:
Compared with patients not given GHB, recipients had higher median sleep efficiency (82% vs. 57%; difference 21 percentage points, 95% CI 7–35) and a greater proportion of sleep in N3 (58% vs. 8%; difference 48 percentage points, 95% CI 18–76).
REM sleep was less frequent with GHB (5% vs. 13%) and absent in two recipients.
All eight GHB recipients were judged by nurses to have slept, compared with two of five patients without GHB; median sleep-quality ratings were 6 vs. 4.
No bradycardia, hypotension, desaturation, vomiting, or seizures were observed. Morning sodium was higher after GHB (140 vs. 135 mmol/L); three patients had recurrent arousals from slow-wave sleep, and one became agitated late in the night.
Interpretation:
The authors report that GHB appeared feasible and was associated with longer, deeper, and better-perceived sleep, but note that polysomnography may not distinguish physiological N3 sleep from drug-induced delta activity. The clinical significance of the apparent N3 increase and the effect of reduced REM sleep remain uncertain.
Limitations:
The experience included only 13 patients and was uncontrolled.
GHB was preferentially given to patients who had slept poorly the preceding night, creating potential confounding by indication and regression to the mean.
No pretreatment polysomnographic recording was available.
A single night cannot establish whether REM sleep recovers or remains suppressed with repeated administration.
Conclusion:
The authors state that these findings require confirmation before routine use and report that the GAMMA-SLEEP randomized, placebo-controlled trial is scheduled to begin in October 2026.
