In This Article
The short answer: Slow-wave sleep and the growth hormone pulse tied to it begin declining well before old age, with the steepest drop happening between your mid-20s and mid-40s (Van Cauter et al., 2000, JAMA). REM sleep declines too, but more gradually, continuing into old age. At the same time, your circadian clock shifts earlier, which is why many older adults fall asleep and wake up earlier without meaning to. Some of what gets blamed on unavoidable aging is really driven by comorbid conditions layered on top of it, which means more of this is modifiable than most people assume.
- What Changes
- Growth Hormone
- Circadian Shift
- The Misconception
- Wearable Data
- What to Do
- FAQ
- Key Takeaways
- References
Read key takeaways →
What Actually Changes in Your Sleep As You Age
Sleep does not decline as one smooth line from age 20 to 80. It is several overlapping processes that change on different schedules. The clearest evidence comes from Eve Van Cauter's team at the University of Chicago, who tracked overnight sleep studies in 149 healthy men aged 16 to 83 (Van Cauter, Leproult and Plat, 2000, JAMA). They found that age-related sleep deterioration happens in at least two stages: slow-wave sleep (SWS) falls off sharply in the decades before age 45, while REM sleep declines more slowly and keeps declining well into old age.
A separate meta-analysis pooling 65 studies and 3,577 subjects aged 5 to 102 (Ohayon, Carskadon, Guilleminault and Vitiello, 2004, Sleep) confirmed the broader pattern: sleep latency and the percentage of lighter stage 1 and stage 2 sleep both increase with age, while REM percentage decreases. For a deeper walkthrough of what these stages actually are, see Sleep Stages Explained.
How Sleep Architecture Shifts With Age
20s to early 30s
Deep sleep near its lifetime peak
Slow-wave sleep and the growth hormone pulses tied to it are close to their highest lifetime levels. Sleep is typically consolidated with few awakenings.
Mid-30s to mid-40s
The steepest slow-wave decline
Van Cauter's data identified this window as the first and sharpest stage of sleep deterioration, with SWS falling faster here than at any later stage of life.
50s to 60s
Fragmentation becomes the dominant issue
SWS decline slows compared to the prior stage, but sleep latency lengthens and nighttime awakenings become more frequent, per the Ohayon meta-analysis.
70s and beyond
REM continues its slow decline
Unlike SWS, REM sleep percentage keeps declining gradually into this decade rather than plateauing, and light stage 1 and 2 sleep make up a larger share of the night.
The Growth Hormone Connection
The reason slow-wave sleep matters so much for aging is what happens during it. The largest daily pulse of growth hormone (GH) is released during the first few hours of SWS, not spread evenly across the day. Van Cauter's team measured this directly and found that the age-related decline in SWS and the age-related decline in GH secretion track each other closely across the lifespan, alongside a rise in evening cortisol.
This is one reason "deep sleep" gets so much attention in longevity circles. It is not just subjective rest. It is the window when a major anabolic and repair-signaling hormone does most of its daily work.
What Drives the SWS to GH Relationship
- →Timing: The largest GH pulse of the day is concentrated in the first SWS-heavy sleep cycles, which is also when SWS is most front-loaded.
- →Parallel decline: Van Cauter et al. (2000) found SWS and 24-hour GH secretion decline together across age groups, not on separate timelines.
- →Cortisol counterpart: The same study found evening cortisol levels rise as SWS falls, shifting the hormonal balance of the night toward stress signaling and away from repair.
- →Not just a number: A wearable cannot measure GH directly, but a consistently low deep sleep percentage is a reasonable proxy for a blunted overnight GH pulse.
For more on how the brain uses deep sleep beyond hormone release, including waste clearance, see What the Glymphatic System Is and Why Sleep Is Your Brain's Cleaning Cycle.
Your Circadian Clock Shifts Earlier
Aging does not just change how much deep sleep you get. It changes when your body wants to sleep in the first place. Reviews of the aging circadian system (Duffy, Zitting and Chinoy, 2015, Sleep Medicine Clinics) describe a consistent phase advance: older adults tend to feel sleepy earlier in the evening and wake up earlier in the morning, even when total sleep opportunity is unchanged.
Melatonin, the hormone that signals the timing of sleep, changes on a similar path. Skene and Swaab (2003, Experimental Gerontology) reviewed evidence that both the amplitude and the timing of the melatonin rhythm shift with age, generally toward a smaller peak and an earlier onset.
Younger vs. Older Circadian Pattern
Melatonin onset
Younger adult: later evening onset, larger peak amplitude.
Older adult: earlier onset, blunted peak amplitude.
Preferred bedtime
Younger adult: circadian drive to sleep arrives later at night.
Older adult: circadian drive to sleep arrives earlier, phase-advanced.
Early waking
Younger adult: less likely without an external trigger.
Older adult: common on a phase-advanced clock, independent of sleep quality.
This distinction matters practically. Waking at 5am is not automatically a sign of poor sleep or a health problem. On a phase-advanced circadian clock, it can simply be what a full night looks like when it starts earlier too.
A Common Misconception: Aging Alone Wrecks Your Sleep
Common Misconception
A lot of what gets attributed to aging alone is really the accumulation of conditions that happen to become more common with age. In the Ohayon et al. (2004) meta-analysis, the size of the age-sleep relationship changed substantially depending on how carefully studies screened out participants with sleep disorders, pain, medication use, and other confounders. Better-screened studies showed smaller age effects. That does not mean sleep architecture is unchanged by age; the SWS and REM changes above are real. It means the popular idea that bad sleep is simply an inevitable, untouchable cost of getting older overstates the case. Untreated sleep apnea, chronic pain, nocturia, and certain medications explain a meaningful share of what looks like "normal aging."
This is a heuristic, not a guarantee: some sleep decline is structural and will happen regardless of how well you manage everything else. But it is worth ruling out the modifiable pieces before accepting a bad night as unavoidable.
What Your Wearable Data Shows
Your device cannot measure growth hormone or melatonin directly, but the sleep stage breakdown it reports is a reasonable window into the same underlying architecture the research above describes. Deep sleep percentage and sleep timing consistency are the two signals worth tracking over months, not nights, since age-related change is slow and a single bad night rarely means anything on its own.
Deep sleep tracking near your personal historical baseline, wake time stable within an hour
Sleep architecture and circadian timing look consistent with a normal pattern for your age. Keep the routine that is working.
Deep sleep trending down over several months, or wake time drifting earlier and earlier
Worth a closer look. Rule out alcohol, evening light exposure, later caffeine intake, and inconsistent bed and wake times before assuming it is simply age.
Frequent nighttime awakenings, snoring flagged by your device, or persistent daytime fatigue despite adequate time in bed
These are signs worth discussing with a clinician, not just optimizing on your own. Undiagnosed sleep apnea becomes more common with age and is treatable.
If your wearable is flagging elevated overnight heart rate or irregular breathing alongside declining deep sleep, read Snoring, Airway, and Undiagnosed Sleep Apnea for what those signals can and cannot tell you.
What to Do About It
None of the interventions below reverse the underlying biology. What they do is protect the sleep architecture you still have, and rule out the modifiable problems that get lumped in with normal aging.
Anchor a consistent wake time, especially as your clock advances
A phase-advanced circadian rhythm works with you if you stop fighting it. Fixing your wake time (rather than forcing a later bedtime) tends to be easier and more sustainable than resisting an earlier chronotype.
Get bright light exposure earlier in the day
Morning light exposure helps anchor a circadian rhythm that naturally drifts earlier with age, and reviewed evidence links stronger light-dark contrast to more stable sleep timing in older adults.
Protect the first few hours of the night
Since SWS and its GH pulse concentrate early in the sleep period, avoid alcohol and late heavy meals in the hours before bed. Both are known to blunt slow-wave sleep specifically.
Rule out sleep apnea if snoring or fatigue is present
Untreated sleep apnea is one of the most common confounders in aging-and-sleep research and is directly treatable, unlike the underlying aging process itself.
Review medications and evening habits with a clinician
Some common medications for blood pressure, allergies, and pain fragment sleep or suppress SWS. A yearly review of what you are taking and when can meaningfully change your sleep data.
Track trends over months, not nights
Because age-related change is gradual, month-over-month deep sleep and consistency trends are far more informative than any single night on your wearable.
For the full framework on building a wind-down routine that protects early-night deep sleep, see the Sleep Protocol.
Frequently Asked Questions
At what age does slow-wave sleep start declining?
Is it normal to wake up earlier as you get older?
Can you increase growth hormone by improving deep sleep?
Should I be worried if my wearable shows very little deep sleep?
Does everyone's sleep decline the same way with age?
Is REM sleep affected the same way as deep sleep with age?
What to Remember
- →Slow-wave sleep declines sharply between roughly age 25 and 45, then more gradually afterward, while REM sleep declines more slowly but keeps falling into old age (Van Cauter et al., 2000, JAMA).
- →The largest daily growth hormone pulse is tied to early-night slow-wave sleep, and the two decline together across the lifespan alongside a rise in evening cortisol.
- →Aging shifts your circadian clock earlier: melatonin onset and the drive to sleep both advance, which explains earlier bedtimes and earlier waking independent of sleep quality (Duffy, Zitting and Chinoy, 2015; Skene and Swaab, 2003).
- →A meaningful share of what looks like inevitable age-related sleep decline is driven by comorbid, treatable conditions rather than age itself (Ohayon et al., 2004).
- →Track deep sleep and wake-time consistency as month-over-month trends, not single nights, since the underlying biological change is gradual.
- →Protecting the first few hours of sleep (avoiding alcohol and late heavy meals, keeping a consistent wake time, getting morning light) targets the exact window where SWS and its GH pulse are concentrated.
Related on Protocol
Sleep Stages Explained: SWS, REM, and Light Sleep
What each stage does and how your wearable measures it
What the Glymphatic System Is and Why Sleep Is Your Brain's Cleaning Cycle
How slow-wave sleep clears metabolic waste from the brain
What Epigenetic Age Means and Whether You Can Actually Reverse It
How biological aging is measured and what slows it
See how your deep sleep and sleep timing are trending as you age
Protocol tracks your sleep stages and bedtime consistency over months, not nights, so you can tell a real shift in your sleep architecture from ordinary night-to-night noise.
Get started freeReferences
Key Researchers
- Eve Van Cauter (University of Chicago) Sleep endocrinologist whose research established the parallel age-related decline of slow-wave sleep and growth hormone secretion, and the rise in evening cortisol.
- Matthew P. Walker (UC Berkeley) Co-author of the 2017 Neuron review on sleep and human aging, synthesizing decades of research on how sleep architecture and brain function change with age.
- Maurice Ohayon (Stanford Sleep Epidemiology Research Center) Led the largest meta-analysis of normative sleep parameters across the human lifespan, spanning 65 studies and 3,577 subjects.
Key Studies
- Van Cauter, Leproult and Plat (2000) JAMA. Tracked sleep and hormone data in 149 healthy men aged 16 to 83, establishing the two-stage pattern of SWS and REM decline and its link to growth hormone and cortisol.
- Ohayon, Carskadon, Guilleminault and Vitiello (2004) Sleep. Meta-analysis of 65 studies developing normative sleep values across the lifespan, and showing that study screening quality changes the measured size of age-related sleep effects.
- Mander, Winer and Walker (2017) Neuron. Review connecting age-related decline in slow-wave activity to changes in memory consolidation and underlying neural mechanisms.
- Duffy, Zitting and Chinoy (2015) Sleep Medicine Clinics. Review of age-related changes in the human circadian timing system, including the phase advance in sleep timing.
- Skene and Swaab (2003) Experimental Gerontology. Reviewed age-related changes in melatonin rhythm amplitude and timing.