Older Brains Still Need Their Sleep. They Have Lost the Machinery to Make It

[RapAdmin]

This 2017 review from UC Berkeley argues that the lighter, shorter, more broken sleep of later life is not a sign that older people need less sleep. Deep-sleep brain waves weaken by 75 to 80 percent over the frontal lobes, the decline starts in the 30s, and it is far steeper in men. The authors link these losses to shrinkage in specific brain regions and to a dulled response to the brain’s own sleep-pressure signal, and they tie the weakened sleep to poorer memory. The evidence is mostly correlational, and no reliable way to restore deep sleep has been shown.

It is one of the most familiar assumptions about ageing: grandparents rise at dawn and doze after lunch because older people need less sleep. A review from the University of California, Berkeley, argues that this has the story backwards.

Bryce Mander, Joseph Winer and Matthew Walker draw together roughly 160 studies. The decline they describe begins earlier than most people expect. Deep sleep starts to erode in the 30s. By the 70s, the slow electrical waves that define deep sleep are 75 to 80 percent weaker over the front of the brain than in young adults. Sleep spindles, the brief bursts of activity linked to learning, fall by up to half late in the night. Sleep becomes shorter, lighter and more fragmented.

The sex difference is large. In a cohort of more than 2,500 adults, men over 70 had half the deep sleep of men under 55, while women showed no significant decline. Yet women report more sleep complaints, a mismatch nobody has explained.

The review traces these changes to wear in particular brain structures. Sleep-promoting neurons in the hypothalamus dwindle. The orexin cells that keep sleep and wake states stable thin out. The body clock’s signal flattens. And the medial prefrontal cortex, where slow waves are generated, loses grey matter. The more that region has shrunk, the weaker a person’s slow waves.

The most interesting argument concerns adenosine, the chemical that builds up during waking hours and creates the pressure to sleep. If older people needed less sleep, they should accumulate less of it. Rodent studies show the opposite: older brains carry more. What declines is the number of receptors that detect it. On this reading, the signal is still being sent but is poorly received.

The consequences show up in memory. Older adults with the weakest slow waves forget more overnight, and those with the fewest spindles learn less the next day. Statistically, the sleep deficit accounts for these memory gaps better than age does. In one experiment, disturbing deep sleep with sounds impaired older adults’ learning the next day even though total sleep time was unchanged.

Fixing the problem is harder. Conventional sleeping pills can increase time in deep sleep, but they generally fail to deliver the memory benefit and sometimes impair recall. Gentle electrical or sound stimulation timed to slow waves has improved memory in some small studies and failed in others.

The case has real weaknesses. This is a narrative review, not a systematic one, and several of its central claims rest on small studies from the authors’ own laboratory. Most of the evidence compares young and old people at a single point in time, which cannot show that poor sleep causes memory decline. The authors themselves say there is no consensus. The paper is also nine years old, and the science of sleep drugs and dementia has moved since.

Still, the central message is that feeling less sleepy with age is not good evidence of needing less sleep.

Insights

What this review offers is a better way to think about sleep in midlife and beyond.

1. Do not treat low sleepiness as proof of enough sleep. Older adults feel less sleepy even when memory measures suggest a shortfall.
2. Hours slept and tracker stage totals miss the main change. Deep-sleep wave strength falls 75 to 80 percent over the frontal lobes even when stage durations look similar.
3. Men should start early. Slow wave activity is already about 50 percent lower in men in their 30s than in their 20s (25 percent in women). By the 70s, men’s deep sleep falls from roughly 13 percent of the night to about 5 percent, or approximately 65 minutes down to 25 in an eight-hour night.
4. Address fixable disruptors: sleep apnea, pain, nighttime urination, alcohol, and medications such as some blood pressure drugs that disturb deep sleep. Ask a clinician before changing any prescription.
5. Be wary of sedative sleeping pills for cognitive purposes. They add deep-sleep time without the memory benefit and can impair recall.
6. Treat stimulation devices as experimental. One young-adult study nearly doubled overnight retention, but three trials in the review found nothing.

Context and Source

• Open Access Paper: Sleep and Human Aging
• Institution: Sleep and Neuroimaging Laboratory, Department of Psychology, and Helen Wills Neuroscience Institute, University of California, Berkeley
• Country: USA
• Journal: Neuron, 2017
  Impact evaluation: The impact score of this journal is 16.9, evaluated against a typical high-end range of 0 to 30 for specialist neuroscience journals (0 to 60+ for top general science), therefore this is an Elite impact journal within its field.

Related Reading:

• Why Sleep Regularity Beats Sleep Duration as the Ultimate Predictor of Longevity
• Exercise Deployed as a "Shield" Against the Mortality Risks of Bad Sleep
• The Orexin Neuropeptide System; Why Your Wakefulness & Activity Neurons Fade With Age—and How to Turn Them Back On
• Breaking the Nap Habit: How Kicking the Afternoon Snooze Safeguards the Aging Brain
• Bursts of ‘pink noise’ during sleep seem to help clear waste from brain
• Sleep supplements: what is most effective, least habit forming, and safest?

[RapAdmin]

After 60, the Body Clock Loosens Its Hold on Sleep

This 2021 French review asks which sleep changes continue after age 60 and why. It finds that most deterioration in sleep architecture happens before 60. What keeps worsening afterwards is early waking, broken sleep, and the loss of sleep spindles and their timing against slow waves. The authors attribute this to a weaker circadian signal combined with lower sleep pressure, and they report that healthy older adults are no sleepier by day and tolerate sleep deprivation better than the young. They lean toward the view that sleep need falls with age, which is the opposite of the Mander and Walker position.

Most accounts of ageing and sleep describe a long slide that gets worse every decade. A review from sleep researchers in Bordeaux and Lyon suggests the timeline is different. Drawing on a meta-analysis of 65 studies, they note that the bulk of the change in sleep stages is already complete by 60. After that, only a few things continue to decline.

The first is timing. Older adults wake one to two hours earlier than younger adults. In a study that followed more than 6,000 people for up to 35 years, the midpoint of sleep crept earlier by about six minutes per decade.

The second is continuity. Time spent awake during the night grows by roughly ten minutes per decade from the age of 30, and older sleepers are far more likely to be roused from light and deep sleep.

The third is in the fine structure of the brain’s electrical activity. Sleep spindles become fewer, smaller and shorter, and they drift out of step with the slow waves they normally ride on. That coupling is thought to support memory, and the authors suggest it may be an early marker of neurodegeneration.

The review then turns to cause. Sleep is governed by two forces: pressure that builds the longer a person is awake, and a roughly 24-hour clock in the hypothalamus that promotes wakefulness by day and sleep at night. Contrary to a common belief, the clock does not speed up with age. Its period stays the same. What weakens is its amplitude. Older people can no longer hold on to sleep in the early morning once melatonin has peaked, and the evening signal that keeps younger adults alert is also blunted.

The ageing eye may contribute. The lens yellows and the pupil shrinks, so less blue light reaches the cells that set the clock. Older adults respond less to moderate indoor light, though the difference disappears under bright light.

The most surprising section concerns sleep loss. Kept awake all night, older adults show smaller drops in reaction time and attention than younger adults. Healthy older people are also no sleepier by day. The authors read this as evidence that sleep pressure, and therefore sleep need, declines with age.

That reading is disputed. The influential 2017 Berkeley review argued that the need persists and the ability to generate sleep fails. The French team acknowledge the disagreement and concede that the question is unresolved.

The review has clear weaknesses. It is a narrative survey with no systematic method, it appeared in a mid-tier journal after a 26-day review, and it contains several internal inconsistencies and at least two apparent citation errors. It proposes no tested treatment.

Its useful contribution is to move attention from the amount of deep sleep to the timing system that organises it.

Actionable Insights

This paper tests no intervention. The points below are reasonable inferences, not proven treatments.

1. Early waking after 60 is mostly a clock problem, not a sign of illness. The shift is real but modest: about six minutes per decade in sleep midpoint, and one to two hours versus young adults.
2. Daytime light probably matters more with age. Older adults respond less to indoor light of 100 to 1,000 lux but normally above 2,000 lux. Ordinary rooms sit at the low end of that range; outdoor daylight is far brighter.
3. Physical activity is described as a second time cue for the clock, though the review gives no effect size.
4. Persistent daytime sleepiness is not normal ageing. The review treats it as a signal of disease, medication effects or a sleep disorder that warrants evaluation.
5. A sleep complaint is not a disorder. About 50 percent of older adults report difficulty, 30 to 48 percent have insomnia symptoms, but only 10 to 20 percent meet criteria for insomnia disorder.
6. Target six to nine hours. That range is associated with better cognition and health in older adults.
7. Fragmentation is the measure to watch: roughly ten extra minutes awake per night for each decade after 30.

Context and Source

• Open Access Paper: Sleep in Normal Aging, Homeostatic and Circadian Regulation and Vulnerability to Sleep Deprivation
• Institutions: Université de Bordeaux and CNRS (SANPSY, USR 3413); CHU de Bordeaux; Lyon Neuroscience Research Center (Inserm, CNRS, Université Claude Bernard Lyon 1)
• Country: France
• Journal: Brain Sciences (MDPI), 2021
  Impact evaluation: The impact score of this journal is 3.4, evaluated against a typical high-end range of 0 to 30 for specialist neuroscience journals (0 to 60+ for top general science), therefore this is a Medium impact journal.

[RapAdmin]

Healthy Older Sleepers: A 20-Study Map With Headline Numbers That Rest on a Few Small Studies

This 2025 scoping review from a largely Italian team searched four databases for observational studies of sleep in disease-free adults aged 60 and over, and included 20. It reports earlier sleep timing, shorter and more broken sleep, less deep and REM sleep, and a U-shaped link between sleep duration and healthy aging. It also finds that many older adults with objectively poor sleep do not complain about it. The method is more transparent than in the two narrative reviews already covered, but the precise figures in its abstract each come from one small study.

Most research on sleep and ageing studies people who are already unwell. A team led from the University of Parma tried to isolate something cleaner: what sleep looks like in older adults who are free of physical and mental disease, and how it relates to ageing well.

They screened 3,832 records and kept 20 studies, published between 1988 and 2023. Together these cover more than 20,000 people, though almost all of that number comes from four large questionnaire surveys. The studies that measured sleep in a laboratory involve only a few hundred participants in total.

The broad pattern is familiar. Healthy older adults go to bed and wake earlier. They spend more of the night awake. Deep sleep and REM sleep shrink. Night-time cortisol runs higher.

The more useful findings are the less expected ones. First, sleep duration and healthy ageing follow a U-shaped curve in three large cohorts. People who sleep much less or much more than average are less likely to be ageing well. Long sleep may be a marker of poor health and not a cause of it, and the review cannot separate the two.

Second, perception and measurement diverge. In one study, 26 percent of older men and 34.5 percent of older women reported no sleep problems despite scoring in the abnormal range on a standard sleep questionnaire. The authors suggest older people adapt to disturbed sleep and stop noticing it. That is consistent with the argument, made by the Berkeley group, that feeling fine is weak evidence of sleeping well.

Third, napping cuts both ways. In a Belgian study, habitual nappers had a lower melatonin peak, hinting at a weaker body clock. In a Chinese cohort of 7,469 people, long naps were tied to lower odds of successful ageing.

Fourth, one 14-year study found that people who aged well cognitively had denser eye movements during REM sleep but spent less total time in REM. That cuts against the simple idea that more REM is better.

The review needs to be read with care. Its abstract states that older adults go to bed 39 minutes earlier, wake 76 minutes earlier, sleep 2.4 hours less, lose 13 percent of sleep efficiency and spend four times as long awake at night. Each of those figures comes from a single study, and one of them compared 150 older adults with just nine young men. The paper offers no pooled estimates and no assessment of study quality.

Its practical advice, covering exercise, relaxation, music therapy and acupuncture, comes from a separate review of trials. The authors excluded all intervention studies from their own analysis.

What the review does well is show how thin the evidence base on healthy aging and sleep is.

Actionable Insights

1. Do not rely on self-rating. Between a quarter and a third of healthy older adults reported no complaints while scoring as poor sleepers. If cognition or mood is slipping, objective assessment is worth considering even without a complaint.
2. Treat very short and very long sleep as signals. Both ends were linked to lower odds of healthy ageing in cohorts of 12,304, 7,469 and 5,616 people. The review gives no odds ratios, so the size of this effect is not available from the paper.
3. Keep naps short. Naps over 60 minutes a day were associated with lower odds of successful ageing in one cross-sectional study. Long naps may reflect existing ill health.
4. Expect less continuous sleep. In one laboratory study, healthy adults in their late 60s were asleep for 82 percent of their time in bed versus 95 percent in young men. Over eight hours in bed, that is about one hour less sleep.
5. Night-time waking was about 64 minutes versus 15, roughly four times longer.
6. Sex matters, but the direction is unclear. Studies disagree on whether older women sleep more or less than men.

Context and Source

• Open Access Paper: The Complex Interplay Between Sleep and Healthy Aging: A Scoping Review (2025)
• Institutions: University of Parma and Parma University Hospital (lead); University of Bologna and IRCCS Istituto delle Scienze Neurologiche di Bologna; University of Catania; University of Modena and Reggio Emilia; Neurocenter of Southern Switzerland; University of Oxford
• Country: Italy (with Swiss and UK affiliations)
• Journal: Nature and Science of Sleep (Dove Medical Press, Taylor and Francis), 2025
• Impact evaluation: The impact score of this journal is approximately 3.6, evaluated against a typical high-end range of 0 to 30 for specialist neuroscience journals (0 to 60+ for top general science), therefore this is a Medium impact journal.

[RapAdmin]

Is it Still True? That no reliable way to restore deep sleep has been shown?

Claim under test, from the Mander, Winer and Walker review: brain stimulation methods “require further refinement and demonstration of efficacy and reproducibility”, and sleep drugs “often fail to trigger any corresponding sleep-dependent memory benefit in the elderly”. My one-line summary of this was “no reliable way to restore deep sleep has been shown”. Source: Sleep and Human Aging (2017).

Verdict: the claim holds in 2026, with one refinement. Several methods raise slow wave measures for a night. None has Level A support, or replicated Level B support, for a durable functional benefit in older adults. Two of the paper’s supporting sub-claims are weaker than presented.

State of the evidence on restoring deep sleep in older adults (as of October 2026)

The 2017 statement still holds. Several interventions raise slow wave activity on an EEG, but none has shown a durable, replicated functional benefit in older adults. The field has also learned that “more slow waves” and “better outcomes” are separable, which weakens the original premise.

Grades follow your A to E scale; I have assumed A = meta-analyses of human trials, B = individual human RCTs, C = human observational, D = preclinical, E = expert opinion.

Why the question matters more now

In 346 Framingham participants aged 60 and over, 52 developed dementia across 17 years, and each percentage-point annual loss of deep sleep was tied to a 27 percent higher risk (hazard ratio 1.27, 95% CI 1.06 to 1.54). The lower bound sits close to 1, the event count is small, and the authors acknowledge that dementia-related brain changes may cause the sleep loss instead. [Grade C] [Confidence: Medium for association, Low for causation] improving deep sleep older adults may lower dementia risk +2

Evidence by approach

Approach	Raises slow waves in older adults?	Functional benefit in older adults?	Grade
Closed-loop acoustic stimulation	Yes, oscillation power; not stage minutes	Not established; effect has shrunk with replication	A (negative-leaning)
Slow-oscillatory electrical stimulation	Yes, acutely	Mixed, small trials	B
Orexin antagonists	No	Biomarker shifts only, short term	B
GABAergic enhancers (tiagabine, gaboxadol, sodium oxybate)	Yes in younger adults	Untested or negative in older adults	B, with translational gap
Trazodone	Suggested	Trial running	B pending
Exercise	Small	Mainly sleep continuity	A for sleep, weak for deep sleep
Warm bath before bed	Reported	Not tested	A, low quality

Acoustic stimulation

Claims on electrical stimulation

• Claim 1: slow-oscillation electrical stimulation in young adults “almost doubles” overnight memory retention.
  • Evidence level: B, single small trial, not reliably replicated.
  • The original study reported enhanced declarative retention with stimulation at 0.75 Hz: Boosting slow oscillations during sleep potentiates memory (2006). A contemporaneous summary gives the sample as 13 participants.
  • A sham-controlled replication in 12 students found no benefit for declarative or procedural memory: Oscillating square wave tDCS during slow wave sleep does not improve declarative memory (2015).
  • A further re-test in 23 young adults is reported as null for memory and for sleep structure. Primary link: source unverified in live search.
  • A 2016 meta-analysis (Barham et al.) is described by a later paper as finding a benefit for declarative memory: Short Duration Repetitive Transcranial Electrical Stimulation During Sleep (2019). Direct link to the meta-analysis: source unverified in live search.
  • Assessment: “almost doubles” is an early, small-sample estimate and should not be quoted as an expected effect.
• Claim 2: the same stimulation during a nap improves memory in older adults.
  • Evidence level: B, small and mixed.
  • Positive nap study: Memory improvement via slow-oscillatory stimulation during sleep in older adults (2015).
  • Null in healthy elderly: No effects of slow oscillatory tDCS on sleep-dependent memory consolidation in healthy elderly subjects (2013).
  • The second positive study (Ladenbauer 2016) and the second null (Passmann 2016): source unverified in live search.
  • Assessment: one-session nap effects, never shown to persist.

Claims on acoustic stimulation

• Claim 3: closed-loop sound stimulation enhances slow oscillations and memory in young adults.
  • Evidence level: A, and the pooled result does not support a robust memory effect.
  • Original trial: Auditory closed-loop stimulation of the sleep slow oscillation enhances memory (2013).
  • Meta-analysis: pooled effect 0.25, p=0.07; 0.44 only in the young, phase-locked subgroup. Modulating overnight memory consolidation by acoustic stimulation during slow wave sleep (2021).
  • Re-analysis: predicted effect fell from 0.99 in 2013 to minus 0.39 in 2021, about a third of participants were probably unblinded, and the memory score had near-zero reliability. Memory retention following acoustic stimulation in slow-wave sleep (2023).
  • A third meta-analysis was retracted for substantial errors: Retraction notice (2022).
  • Assessment: the effect on sleep physiology is real. The effect on memory shows a classic decline pattern.
• Claim 4: sound stimulation increases slow oscillation power and next-day recall in older adults.
  • Evidence level: B, one night, 13 people. Not replicated.
  • Original: 13 participants aged 60 to 84, one stimulation night and one sham night. Acoustic Enhancement of Sleep Slow Oscillations and Concomitant Memory Improvement in Older Adults (2017).
  • Later trial: oscillations were induced, but declarative memory did not improve and retention was impaired. Susceptibility to auditory closed-loop stimulation of sleep slow oscillations changes with age (2020).
  • A 2026 preprint in 16 older adults again shows enhanced oscillations with unchanged sleep architecture: Closed-Loop Auditory Stimulation Reveals Differential Sleep Oscillatory Contributions to Memory in Healthy Older Adults (2026). Not peer reviewed.
  • Assessment: the paper labelled this “preliminary”, and that label was correct.
• Claim 5: some stimulation studies fail to replicate.
  • Evidence level: A. Fully supported by the sources under Claims 1 to 4.

Claims on drugs

• Claim 6: GABA-targeting hypnotics often fail to deliver a memory benefit in the elderly and can be amnestic.
  • Evidence level: B, but mostly in young adults. Extrapolation gap.
  • Supporting: tiagabine increased slow wave sleep without improving consolidation. Slow wave sleep induced by GABA agonist tiagabine fails to benefit memory consolidation (2013).
  • Contradicting: one of the four studies the review cites for this claim found the opposite. Zolpidem increased spindle density and improved verbal memory in 49 adults aged 18 to 39. The Critical Role of Sleep Spindles in Hippocampal-Dependent Memory: A Pharmacology Study (2013).
  • Cochrane found no randomised trials at all of benzodiazepines or related hypnotics for sleep in dementia: Pharmacotherapies for sleep disturbances in dementia (2020).
  • A systematic review notes that sodium oxybate improved some memory and cognitive measures: Can Slow-Wave Sleep Enhancement Improve Memory? (2019). The underlying trial enrolled adults aged 18 to 50: Enhancing Slow Wave Sleep with Sodium Oxybate (2010).
  • Assessment: the wording “in the elderly” is not backed by direct trials in older people. The honest statement is that this is largely untested in older adults.
• Claim 7: little is known about non-GABA sleep drugs in older people at risk of dementia.
  • Evidence level: A to B. Outdated since 2017, but the gap for deep sleep remains.
  • Cochrane: two orexin antagonist trials, 323 participants with Alzheimer’s disease, probable benefit on sleep. One trazodone trial of 30 people. Pharmacotherapies for sleep disturbances in dementia (2020). The outcomes were total sleep time and continuity, not deep sleep.
  • Biomarker trial: 38 adults aged 45 to 65; amyloid-beta fell 10 to 20 percent and one tau phosphorylation site fell 10 to 15 percent, with the tau effect gone by 24 hours. Insomnia drug may lower levels of Alzheimer’s proteins, NIH (2023).
  • Sleep architecture did not differ from placebo in that trial: Orexin Receptor Antagonists for the Prevention and Treatment of Alzheimer’s Disease (2024).
  • Trazodone for deep sleep and memory is still being tested: REST trial registration (ongoing).
  • Assessment: no cognitive endpoint data exist for any of these drugs.

Claims on mechanism and prevention

• Claim 8: NREM sleep drives clearance of amyloid-beta from the brain.
  • Evidence level: D. Translational Gap. Flag heavily.
  • The review rests this on two mouse studies (Kang 2009, Xie 2013). Primary links: source unverified in live search.
  • Direct contradiction in mice: clearance was markedly reduced, not increased, during sleep and anaesthesia. Brain clearance is reduced during sleep and anesthesia (2024). The same paper states the question is unresolved, with findings on both sides.
  • Assessment: the central rationale for restoring deep sleep to prevent dementia is preclinical and actively disputed.
• Claim 9: restoring sleep “may deliver preventative benefits” against cognitive decline and Alzheimer’s disease.
  • Evidence level: C for the association, E for the prevention claim.
  • Cohort: 346 participants, 52 dementia cases over 17 years, hazard ratio 1.27 (95% CI 1.06 to 1.54) per percentage-point annual loss of deep sleep. Improving Deep Sleep in Older Adults May Lower Dementia Risk, summary of Himali et al. (2023). Primary JAMA Neurology link: source unverified in live search.
  • No trial has tested whether increasing deep sleep changes dementia incidence.
  • The six-year sleep fragmentation cohort cited by the review (Lim 2013): source unverified in live search.
• Claim 10: experimentally disrupting deep sleep impairs next-day learning in older adults, which establishes causality.
  • Evidence level: B for disruption. This does not transfer to restoration.
  • Original study (Van Der Werf 2009): source unverified in live search.
  • Consistent recent result: randomly timed sounds reduced slow wave sleep and impaired declarative recall. Random auditory stimulation disturbs traveling slow waves and declarative memory (2025). Preprint.
  • Assessment: showing that damage hurts does not show that boosting helps.
• Claim 11: these methods have not been tested longitudinally.
  • Evidence level: A to B. Still true.
  • A two-week home acoustic study cautions its sample may have been too small to detect a memory effect: The Effect of Closed-Loop Auditory Stimulation on Memory Consolidation and Sleep Physiology in an Ecological Setting (2025).
  • The longest drug trials are still recruiting: suvorexant for 24 months and lemborexant for 6 months, per Orexin Receptor Antagonists review (2024).

Evidence the paper did not consider

• Exercise: Level A for sleep continuity in older adults, with benefits mainly in latency and efficiency. Physical exercise programmes to improve insomnia or poor sleep quality in non-hospitalised elderly people (2025). The effect on deep sleep specifically is small and not established in the healthy elderly.
• Warm bathing before bed: Level A, low quality. A meta-analysis of 17 studies lists slow wave sleep among the outcomes examined. Before-bedtime passive body heating by warm shower or bath to improve sleep (2019). I could not access the effect size or participant ages.

Summary of the hierarchy

• Level A support for restoring deep sleep with functional benefit in older adults: none.
• Level B: single-night or nap effects in samples of 12 to 30, with direct failures to replicate.
• Level C: deep sleep loss predicts dementia, with a confidence interval close to no effect.
• Level D: the clearance mechanism, now contradicted within mice.
• Largest Translational Gap: Claim 8, and through it Claim 9.

[RapAdmin]

GABAergic slow wave enhancers: what the trials show

All three drugs increase slow wave sleep in controlled trials, including in older adults. None has been shown to improve cognition or any health outcome in middle-aged or older people as a result. Tiagabine and gaboxadol research largely stopped around 2010. Sodium oxybate is the only one with trials in the last few years. No meta-analysis specific to these drugs and deep sleep surfaced in my searches, so the best evidence is Level B.

Tiagabine (GABA reuptake inhibitor)

• Elderly insomnia, Level B. A randomised, double-blind, placebo-controlled trial exists: Effect of tiagabine on sleep in elderly subjects with primary insomnia (2006). I could only reach the citation record. My recollection is that it showed a dose-related rise in slow wave sleep with next-day impairment at the highest dose; treat that as unverified.
• Adult insomnia, Level B. In 232 adults, the three highest doses raised slow wave sleep, with no difference from placebo in time awake, sleep latency or total sleep time. Tiagabine increases slow-wave sleep in a dose-dependent fashion without affecting traditional efficacy measures (2006). Conference coverage reported that patients did not rate their sleep as deeper.
• Sleep apnea, Level B. In 14 patients, 12 mg increased slow wave activity versus placebo: Effects of tiagabine on slow wave sleep and arousal threshold in patients with obstructive sleep apnea.
• Function, Level B, young adults only. In 38 healthy adults under sleep restriction, 8 mg increased slow wave sleep and ratings of restorative sleep: Enhancement of Slow Wave Sleep: Implications for Insomnia. In 14 young men, slow wave sleep rose but spindle activity fell and memory retention was not improved, and was significantly worse for a motor task: Slow wave sleep induced by GABA agonist tiagabine fails to benefit memory consolidation (2013).
• Assessment: more deep sleep on the recording, no benefit the sleeper can feel, and a possible cost to spindles.

Gaboxadol (extrasynaptic GABA-A agonist)

• Healthy elderly, Level B. A safety study states that 10 mg in the elderly reduced awakenings and time awake and increased sleep efficiency and slow wave sleep: Tolerability, pharmacokinetics and night-time effects of gaboxadol and zolpidem in elderly subjects. The original elderly studies (Lancel 2001, Mathias 2005): source unverified in live search.
• Mixed-age, Level B. In 101 healthy subjects aged 20 to 78 exposed to traffic noise for seven nights, slow wave sleep and slow wave activity stayed elevated throughout, with total sleep up 16 minutes and time awake down 11 minutes. Those 65 and over received 10 mg. Gaboxadol in a traffic noise model of transient insomnia.
• Insomnia, Level B. In 26 adults, 15 mg increased slow wave sleep by 21 percent over a placebo mean of 94 minutes, about 20 extra minutes: Novel Drug Shows Promise for Sleep Initiation, Maintenance. This is a conference report.
• Sleep restriction, Level B, younger adults. A trial of 15 mg found reduced daytime sleepiness, and notes that the drug consistently raised slow wave measures in adult and elderly subjects: Slow Wave Sleep Enhancement with Gaboxadol Reduces Daytime Sleepiness During Sleep Restriction.
• Development outcome. A six-week Phase 3 trial in elderly insomnia enrolled 465 and was terminated: Elderly Primary Insomnia Polysomnographic Study. The programme was dropped in 2007 because the overall profile did not support further development: Merck, Lundbeck End Development of Insomnia Drug (2007). Trade press cited hallucinations, disorientation and inconsistent efficacy.
• Assessment: the cleanest deep sleep signal in older adults of the three, but the drug is not available and failed at scale.

Sodium oxybate (GABA-B and GHB receptor agonist)

• Older patients, Level B. In 12 people with Parkinson’s disease, slow wave sleep rose by 72.7 minutes (95% CI 55.7 to 89.7), with new sleep apnea in two patients and parasomnia in one. Sodium oxybate for excessive daytime sleepiness and sleep disturbance in Parkinson disease (2018). This is the largest deep sleep increase reported for any intervention I found.
• Depression, Level B, recent. A crossover trial tested it specifically to promote slow wave sleep: Gamma-hydroxybutyrate to promote slow-wave sleep in major depressive disorder (2025). I could not extract the results or participant ages.
• Healthy adults, Level B, recent. One night increased NREM percentage and delta power and reduced REM, with sleep efficiency unchanged. The NREM gain showed no significant association with memory outcomes. Sodium Oxybate Alters Sleep Architecture and Reduces Emotional Selectivity in Memory-Related Responses (2026).
• REM sleep behaviour disorder, Level B. A 24-person trial is discussed in an editorial that also urges caution in sleep apnea and calls neuroprotection a hypothesis only: Sodium oxybate: a new horizon for symptomatic treatment of RBD? (2023).
• Sleep deprivation, Level B, younger adults. 58 adults aged 18 to 50: Enhancing Slow Wave Sleep with Sodium Oxybate Reduces the Behavioral and Physiological Impact of Sleep Loss.
• Alzheimer’s model, Level D. Translational Gap. Two weeks of treatment in mice at an early disease stage improved cognition and reduced amyloid pathology: Pharmacological enhancement of slow-wave activity at an early disease stage (2025). No human equivalent exists.
• Assessment: the strongest and most current evidence, but in small disease-specific samples, with respiratory risk that matters most in the older people who would be the target.

What is missing

• No trial of any of these drugs in healthy middle-aged or older adults with a cognitive or biomarker endpoint.
• No trial longer than a few weeks that tracks deep sleep.
• No evidence that drug-induced slow waves are functionally equivalent to natural ones. The tiagabine spindle finding suggests they may not be.