N=1 After 40

Successful experiment · August 2026

Seltorexant / JNJ-42847922

A short sleep experiment produced fewer awakenings and more consolidated sleep. It did not meaningfully extend total sleep, and increasing the amount above 5 mg added no observable value. Personal experiment record, not medical advice, prescribing, sourcing guidance, or an instruction to copy.

Outcome: Helpful for sleep continuity, not total sleep duration. Sleep felt less fragmented and I noticed fewer awakenings, while higher amounts did not improve the result beyond what I observed at 5 mg.

Research Summary

What it is

Seltorexant, also known as JNJ-42847922 or MIN-202, is an investigational selective orexin-2 receptor antagonist. Orexin is part of the brain's wake-maintenance system. The basic idea is to reduce an arousal signal that keeps wakefulness active rather than create sleep through a broad sedative mechanism.

Why the selective mechanism caught my attention

The practical target was not simply feeling knocked out. It was quieter sleep: falling asleep cleanly, waking less during the night, and returning to sleep without carrying an obvious next-morning penalty. Seltorexant was designed to favor the orexin-2 receptor and has a relatively short human half-life of roughly two to three hours, making sleep continuity and next-morning alertness the useful real-world questions.

The strongest insomnia trial

A randomized, double-blind trial enrolled 364 adults with insomnia and compared 5 mg, 10 mg, and 20 mg of Seltorexant with placebo and zolpidem. The clearest effects appeared at 10 mg and 20 mg: participants fell asleep faster and spent less time awake after sleep onset, and those effects remained present with repeated treatment. The study's conclusion did not establish 5 mg as an efficacious insomnia dose.

The earlier sleep-efficiency signal

An earlier crossover study used 40 mg in 27 people with insomnia. Compared with placebo, Seltorexant increased total sleep by about 28 minutes after the first dose and 38 minutes after repeated dosing. Sleep efficiency improved by 5.8 and 7.9 percentage points, respectively. Sleep onset also became faster, while wake time after sleep onset fell by roughly 11 minutes.

Evidence in persistent insomnia with depression

A separate four-way crossover study in 20 antidepressant-treated people with persistent insomnia tested single doses of 10 mg, 20 mg, and 40 mg. All three amounts shortened sleep latency, increased total sleep, and improved sleep efficiency compared with placebo. That population is different from mine, but it reinforces the same expected signature: faster onset and better sleep maintenance.

Timing and residual-effect context

In a healthy-volunteer study spanning 5–60 mg, Seltorexant reached peak concentration in approximately 0.5–1.5 hours and had a mean half-life of two to three hours. Amounts above 5 mg registered as sedating four hours after daytime administration, while 20 mg and above consistently produced somnolence. The short profile is relevant to next-morning function, but it does not guarantee an absence of residual effects.

Safety and development status

Seltorexant remains investigational. Johnson & Johnson's current pipeline lists it in phase 3 for adjunctive treatment of major depressive disorder with insomnia symptoms. In the 364-person insomnia trial, most treatment-emergent adverse events were mild or moderate; headache was the most common, and somnolence was more frequent at 20 mg than at 5 or 10 mg. Uncommon rhythm or conduction findings also led several participants to discontinue treatment.

Sources

The Experiment

Protocol

I tested Seltorexant in August 2026. I began at 5 mg and also tested 10 mg and 15 mg to see whether increasing the amount strengthened the result. Sleep was evaluated through wearable-reported duration, efficiency, and performance alongside my perceived awakenings and next-morning experience.

Question

Would Seltorexant make sleep less fragmented, improve wearable-reported efficiency, or extend total sleep? The dose changes also tested whether moving above 5 mg produced a stronger practical result without an obvious next-morning tradeoff.

What changed

The positive result was sleep continuity. During the initial treatment period, sleep efficiency averaged 94.6%, with approximately 17 minutes awake while in bed. During the subsequent no-dose comparison, efficiency averaged 88.5%, with approximately 35 minutes awake while in bed. I also noticed fewer awakenings and a more consolidated night, which is the part of the result that mattered in practice.

What did not change

Seltorexant did not meaningfully extend total sleep. The initial treatment period and the subsequent no-dose comparison averaged 4:34 and 4:33 asleep, respectively. Across the full comparison, treatment sleep contained about 20 additional minutes, but time in bed also increased by about 22 minutes. The apparent duration gain therefore tracked the longer sleep opportunity rather than a clear extension of sleep itself.

Dose read

Increasing the amount did not improve the result. The continuity benefit was already present at 5 mg. Moving to 10 mg did not establish an incremental effect, and the 15 mg result was among the poorer sleep outcomes. For this experiment, 5 mg was sufficient to capture the observed benefit; that is a personal result, not a claim that 5 mg is a generally effective or minimum clinical dose.

Decision

I am closing the experiment as helpful. Seltorexant earned a positive result for fewer awakenings and more consolidated sleep, not for extending total sleep duration. Increasing the amount above 5 mg added no observable value, so the useful finding is narrow and clear: better continuity without a longer night.