A systematic review published in Sleep this May pooled the trials of dual orexin receptor antagonists in adults aged 65 and over, covering 2,235 treated patients against 1,409 controls. The results are worth quoting precisely, because precision is what this category usually lacks in coverage. At one month, time awake after falling asleep fell by about 13 minutes on the lower dose of daridorexant and 21 minutes on the higher, and by roughly 23 minutes on either dose of lemborexant. Total sleep time on lemborexant rose by 34 to 39 minutes. The effects held at three months.

Half an hour of sleep is a real gain for someone lying awake at four in the morning. It is also a long way from what the advertising implies, and the reason this class matters has more to do with what it replaces than with the size of the effect.

From the Lab to the Ledger

Older sleeping pills work by turning sedation up. Benzodiazepines and the so-called z-drugs act on the GABA system, the brain's main inhibitory channel, and push the whole nervous system toward sedation. That produces sleep, and it also produces the problems that made these drugs contentious in older adults: unsteadiness on the way to the bathroom, next-day grogginess, memory effects, tolerance, and a difficult withdrawal.

Orexin works the other way around. Orexin, also called hypocretin, is a neuropeptide that holds the brain in a waking state; people with narcolepsy lose the neurons that make it. Blocking its two receptors does not sedate the brain, it releases the brake holding wakefulness on. The theoretical advantage is that sleep architecture stays closer to normal and the drug is less likely to blunt everything else along with consciousness.

Whether that translates into fewer falls and less confusion in practice is the question, and it is the one the meta-analysis could not answer: it did not report on falls, cognition or next-day impairment. That is a substantial gap, since those outcomes are the entire reason for preferring a new class in this age group. What it did report is that daridorexant had a favourable safety profile, while lemborexant and suvorexant produced more treatment-emergent adverse events than placebo without more serious ones.

There is also a point about what insomnia in later life usually is. Sleep becomes lighter and more fragmented with age as normal physiology, and a substantial share of complaints involve something else: untreated sleep apnea, restless legs, nocturia, pain, an antidepressant taken at the wrong time of day, or alcohol used to get to sleep. A drug prescribed on top of an unrecognised cause treats the symptom and leaves the cause running.

Bio-Pipeline Ledger

Dual orexin receptor antagonists, meaning suvorexant, lemborexant and daridorexant: approved, effective at the modest magnitudes above, and the current preferred pharmacological option in older adults in most guidance.

Cognitive behavioural therapy for insomnia: the recommended first-line treatment in every major guideline, with effects that persist after treatment ends rather than requiring a nightly dose. Available in app and workbook form where a therapist is not.

Benzodiazepines and z-drugs in adults over 65: long listed as potentially inappropriate, with falls, fractures and cognitive effects as the reasons. Still widely prescribed, and not something to stop abruptly.

Melatonin: modest for shifting sleep timing, weak for staying asleep, unregulated in dose and content in the United States, with independent testing repeatedly finding large discrepancies between label and bottle.

Over-the-counter sleep aids containing diphenhydramine or doxylamine: strongly anticholinergic, on the same caution lists as the older prescription drugs, and the most commonly overlooked item on an older adult's medication list.

Treatable contributors including sleep apnea, restless legs, nocturia and pain: common, identifiable, and the reason a sleep complaint deserves a workup rather than a prescription.

The Clinical Reality Check

What is established is that this class improves sleep in older adults by measurable and unspectacular amounts, that the benefit holds at three months, and that its mechanism gives reason to expect fewer of the problems the older drugs cause.

What is not established is that expectation itself. The trials in this age group have not reported falls, next-day function or cognition, which is precisely where the case for switching rests. Anyone presenting these drugs as proven safe for older brains is ahead of the published evidence, and comparisons between the three of them come from network analyses rather than head-to-head trials.

The practical order is the one the guidelines already give and almost nobody follows. Cognitive behavioural therapy for insomnia first, because it works and it keeps working. A look for the treatable causes second, including the antihistamine in the drawer that most people do not count as a drug. Medication third, for a defined period, with a plan for reviewing it rather than an indefinite refill. If you are already on an older sleeping pill and it is working, that is a conversation about a gradual change rather than a reason for alarm.