Lemborexant Versus Zolpidem: Why Sleep Recordings and Diaries Gave Different Answers

Before comparing two insomnia pills, decide what improvement would matter to you: less time awake, fewer difficult mornings, or both. Sleep recordings and a person’s account of the night can tell different stories.

The trial mostly enrolled women with trouble staying asleep

SUNRISE 1 randomized 1,006 people; 86.4% were women. Women qualified from age 55 and men from age 65. Treatment lasted 30 nights, comparing lemborexant 5 mg, lemborexant 10 mg, zolpidem extended release 6.25 mg and placebo. Eisai sponsored it.¹

At month’s end, lemborexant reduced recorded time awake after initially falling asleep by an adjusted 7.7 minutes at 5 mg and 9.1 minutes at 10 mg more than zolpidem. These are between-treatment differences, not total improvements from baseline.¹

In diaries, zolpidem reduced reported wake time 14.5 minutes more than lemborexant 5 mg. Differences in insomnia-severity and daytime-function scores were not significant. These results concern the tested doses and formulations.¹

That discrepancy is a reason to bring both your night-time observations and your morning symptoms to follow-up. A favorable recording cannot tell the prescriber whether you feel better enough to keep taking a medicine.

The 2026 brain-wave paper used the same trial

In 2026, Sattari² and colleagues published a secondary analysis of SUNRISE 1 recordings. It did not recruit another independent group. Investigators selected 249 participants aged 65 or older, using criteria that included treatment adherence and baseline memory scores. After recording-quality checks, 158 contributed usable non-REM data and 128 usable REM data. Those samples partly overlapped.²

The drugs produced different patterns of electrical activity during sleep. With continued 10 mg lemborexant treatment, increases in very slow brain-wave activity, called slow oscillations, were associated with improved insomnia scores. That association does not establish that the wave change caused the symptom improvement.²

The analysis covered the same one-month exposure. It did not establish protection against dementia, better long-term memory or a superior overall balance of benefit and harm. Eisai funded the work, with additional US National Institute on Aging support.²

Ask what a claim about “more natural sleep” means in measurable terms. A brain-wave pattern, an insomnia score and the ability to drive safely are separate outcomes.

The fall-risk comparison also needs qualification

Canada’s Drug Agency³ reviewed additional evidence from two retrospective cohort studies in a resubmission assessment. One suggested fewer falls with lemborexant than with benzodiazepines as a group. Its comparisons with zolpidem and trazodone remained inconclusive, with confidence intervals crossing the no-difference value.³

The agency identified concerns including selection bias, possible errors in medication or outcome classification, and residual confounding. People prescribed different medicines may differ in ways the analysis cannot fully account for. It concluded that uncertainty remained about whether lemborexant offered a preferable balance of benefits and harms over other insomnia treatments.³

This does not mean a person must fall before treatment can be judged unsuitable. Tell the prescriber about unsteadiness, nighttime bathroom trips and previous falls when setting the plan.

Check the current US label before a switch

The DAYVIGO prescribing information recommends 5 mg no more than once nightly, immediately before bed, with at least 7 hours before planned waking. An increase to 10 mg depends on response and tolerability. This is not a rescue dose to add during a short remaining night.⁴

The label warns about next-day impairment and increased fall risk, especially in older people. Some participants taking 10 mg had impaired driving ability, despite the driving study’s average result not showing significant impairment. Avoid alcohol; other sedating medicines can add to the effects.⁴

It is contraindicated in narcolepsy, and medicines affecting CYP3A can require avoidance or a dose limit. If sleep-walking, sleep-driving or another complex sleep behavior occurs, the label says to stop DAYVIGO immediately. Contact the prescriber promptly.⁴

For the broader label comparison, see orexin-blocking sleep pills.

Ask how CBT-I fits into the plan

The 2026 American Academy of Sleep Medicine⁵ guideline conditionally favors starting cognitive behavioral therapy for insomnia, CBT-I, alongside medication over starting medication alone. It conditionally advises against routinely preferring that combination over CBT-I alone. Both recommendations rely on low-certainty evidence and require attention to individual preferences. They concern treatments started concurrently, not an instruction to abruptly stop an existing prescription.⁵

The “Medicines Belong in Shared Decisions” section of Did I Do That in My Sleep? sets out the review conversation: expected benefit, adverse effects, interactions, next-day impairment, cost and follow-up.

For an appointment about lemborexant, bring the current pill’s name, dose and timing, approximate time awake during the night, and any morning grogginess. Ask which outcome you will track, when to review it, and how a supervised switch would work if you decide together to make one.