Two insomnia drugs alter the sleeping brain differently. Only one tracked with relief.

Lemborexant and zolpidem treat insomnia through different mechanisms — lemborexant blocks orexin receptors, which promote wakefulness, while zolpidem enhances GABA-A receptor activity, the brain's primary inhibitory signaling system. A new analysis published in SLEEP Advances used detailed EEG recordings from an existing phase 3 trial to compare how the two drugs actually change brain activity during sleep in older adults, and whether those changes relate to symptom improvement [s1].

The design

The analysis draws on data from a randomized, double-blind, placebo-controlled, active-comparator phase 3 trial conducted between May 2016 and January 2018 [s1]. A subset of 249 participants, all meeting DSM-5 criteria for insomnia, had detailed EEG outcomes analyzed as part of this new paper [s1]. Participants received 5 mg or 10 mg lemborexant, 6.25 mg extended-release zolpidem, or placebo nightly for 30 days, with polysomnography — full EEG-based sleep monitoring — collected at baseline, after acute exposure (days 1–2), and after chronic exposure (days 29–30) [s1]. The analysis tracked how EEG measures changed from acute to chronic use, and how those within-person changes related to scores on the Insomnia Severity Index (ISI), a standard symptom questionnaire [s1].

What it found

Zolpidem produced a distinct spectral signature from the start: significant increases in slow-oscillation, sigma, beta, and gamma brain-wave power, alongside decreases in delta and theta power, compared with all other groups — an effect present from the first dose and persisting with continued use, with an added increase in alpha power appearing between acute and chronic exposure [s1].

Lemborexant at 10 mg showed a different trajectory: minimal changes after the first doses, but with continued use, an increase in slow-oscillation power and a decrease in alpha power emerged [s1]. Notably, the size of that slow-oscillation increase from acute to chronic exposure predicted improvement in Insomnia Severity Index scores — and this relationship was driven specifically by the 10 mg lemborexant group [s1]. Zolpidem produced its own distinct effects on REM sleep, and 5 mg (but not 10 mg) lemborexant reduced REM sigma and beta power [s1].

Why the symptom link matters more than the brain-wave changes themselves

The most substantive finding here isn't that the two drugs affect the sleeping brain differently — that was already established by prior pharmacodynamic research in younger populations, which the paper's authors cite as the basis for this study [s1]. It's that only one of the observed EEG changes — the emergence of increased slow-oscillation power with continued lemborexant 10 mg use — actually tracked with how much better people said they were sleeping, based on the Insomnia Severity Index [s1]. That's a more specific and clinically relevant finding than a spectral difference alone, because it links a measurable brain change to the outcome patients and clinicians actually care about, at least in this dataset and for this dose.

What this doesn't establish

This is a secondary analysis of an existing trial, using data collected between 2016 and 2018 and analyzed now — the underlying trial was designed around clinical insomnia outcomes, not primarily around this EEG relationship, and the correlation between slow-oscillation change and symptom improvement, while statistically identified, is not evidence that the brain-wave change causes the symptom relief rather than both reflecting some other underlying process. The study population was specifically older adults with insomnia; whether the same EEG-symptom relationship holds in younger adults isn't addressed. And this analysis doesn't compare the two drugs' relative clinical effectiveness — the source trial did that separately — only how their physiological signatures diverge.

What to watch

Whether the slow-oscillation/symptom-improvement relationship replicates in an independent cohort, and whether it could eventually help predict, earlier in treatment, which patients are responding to lemborexant specifically. This article describes results from approved insomnia medications' comparative pharmacology; it is not medical advice.

Sources

1. Effects of lemborexant and zolpidem on sleep electroencephalography in older adults with insomnia: a randomized trial — SLEEP Advances, 12 March 2026