Sleep stages after 40 can start to feel like a nightly report card. You wake up, check your watch and see a neat stack of light, deep and REM sleep. One number looks “low,” and suddenly a normal morning becomes a health worry.
A large new study gives those graphs some scientific context—but not the meaning a wearable dashboard may seem to promise. This was an observational study, so it cannot establish cause and effect. It found associations between wrist-estimated sleep patterns and later health diagnoses. It did not show that a consumer tracker can predict disease, diagnose a sleep disorder or tell you how to engineer a healthier future by chasing a particular stage percentage.
The quick answer: Sleep trackers may help you notice broad patterns, such as sleep timing or recurring nights that feel disrupted. Their REM and deep-sleep numbers are estimates, not direct measurements of brain activity. The new research is observational, so it cannot establish cause and effect. A consumer sleep tracker cannot diagnose a sleep disorder or predict your future health from a stage score.
Sleep stages after 40, explained without the dashboard jargon
Sleep is not one unchanging state. You move through non-REM and REM sleep several times during a typical night.
The National Institute of Neurological Disorders and Stroke describes four stages:
| Stage | Plain-English description |
|---|---|
| Non-REM stage 1 | The transition from wakefulness into sleep. |
| Non-REM stage 2 | Light sleep. |
| Non-REM stage 3 | Deep sleep, which tends to occur more in the first half of the night. |
| REM sleep | The first period typically begins about 90 minutes after sleep starts, with longer REM periods later in the sleep session. |
REM and non-REM stages repeat in cycles through the night.
That timing matters. A short or frequently interrupted night can change the pattern your device displays. Age matters too: NINDS notes that people generally spend less time in REM sleep as they get older. But sleep also varies substantially among people of the same age, which is one reason a single “ideal” stage percentage is not a useful target for everyone over 40.
What the 2026 sleep study actually measured
The new PLOS Medicine study analyzed data from 95,559 UK Biobank participants. Their average age at baseline was 56.2 years.
Participants wore a research wrist accelerometer for seven consecutive days during 2013–2015. A deep-learning system called SleepNet used the movement data to infer REM and the three non-REM stages: N1, N2 and N3.
Researchers then examined links between those estimated sleep patterns and new inpatient diagnoses across 1,049 disease categories. The median follow-up was 8.9 years.
That is an impressive amount of health-record data. It is also important to be precise about what it represents. The researchers did not bring nearly 96,000 people into sleep laboratories. They did not directly record everyone’s brain waves or eye movements overnight. They used a research algorithm to infer sleep stages from wrist movement.
The study also did not evaluate Apple Watch, Fitbit, Oura or any other named consumer product. Its findings cannot be used as an accuracy certificate for the device on your wrist.
What the study found—and why the headline numbers need context
The analysis identified 156 statistically significant associations between estimated sleep patterns and later disease categories after a strict correction for testing many outcomes.
Greater estimated REM duration was associated with lower risk across 83 disease categories. Greater estimated deep sleep was associated with lower risk across seven.
Those numbers sound dramatic, but they are not 83 predictions that your watch can make about you.
This was a broad phenome-wide association study. The researchers looked across more than a thousand outcome categories to identify patterns worth investigating. An association can be scientifically useful without proving that one factor produced the other.
When the team added a two-year washout period to reduce the possibility that early, undiagnosed disease had already affected sleep, 95 of the original 156 associations remained statistically significant. That does not erase the findings. It does show why reverse causation is more than a technical footnote.
Why your sleep graph is not a personal health forecast
Three limitations should stay beside every discussion of this study.
Association is not cause and effect
The study followed people; it did not assign them to get more REM or deep sleep. Poor sleep could contribute to health problems, health problems could disturb sleep before diagnosis, or other factors could influence both. This design cannot tell us that increasing one sleep stage would prevent disease.
Wrist movement is not a laboratory sleep study
A clinical polysomnogram records brain waves, eye and limb movements, breathing, oxygen levels and heart rate. Wrist accelerometry does not collect that full set of signals.
Independent experts commenting through the Science Media Centre emphasized that the study’s stages were movement-based estimates rather than exact physiological measurements. They also described the model’s performance as moderate.
Seven days cannot represent every season of life
Seven days of wrist monitoring may not represent long-term habitual sleep. The UK Biobank sample also limits how broadly the results can be applied: 96.9% of participants identified as White, and the cohort is not a representative sample of adults in the United States or Canada.
What a sleep tracker can reasonably tell you
Consumer devices may collect sounds, movement, hours slept, heartbeat and breathing. That can make sleep less invisible.
A tracker may help you notice questions worth exploring:
- Does your sleep opportunity shrink on work nights?
- Are your bed and wake times changing substantially?
- Do nights that look disrupted match how rested you feel?
- Is a recurring pattern appearing alongside an ongoing sleep complaint?
The useful word is pattern. An isolated score can change for many reasons and is not a diagnosis. A recurring observation can be a starting point for a better conversation about your routine and symptoms.
The American Academy of Sleep Medicine says consumer sleep data should be considered as part of a comprehensive sleep evaluation and should not replace validated diagnostic testing. The organization also notes that wearable data can raise awareness and support conversations between patients and clinicians.
Think of your device as a notebook that makes estimates—not a miniature sleep laboratory.
What your tracker cannot tell you
Even an attractive, detailed dashboard cannot establish:
- that you were definitely in a particular physiological stage at a particular minute;
- that a low deep-sleep or REM number caused today’s fatigue;
- that you have insomnia, sleep apnea or another sleep disorder;
- that your current stage pattern will lead to a future disease; or
- that changing one stage number will prevent a health condition.
A “good” score should not override persistent symptoms, and a “poor” score should not be treated as a diagnosis.
Do you need a perfect amount of REM or deep sleep after 40?
There is no fixed tracker-derived REM or deep-sleep percentage that every adult over 40 should target.
Sleep stages naturally shift across the night. They also change with age and vary between individuals. A target copied from an app, influencer or online chart may create precision that the estimate cannot justify.
A more useful question is whether you are allowing enough time for sleep and whether your sleep is generally regular, refreshing and compatible with daytime functioning.
The six-to-eight-hour finding does not mean six hours is enough
The study found nonlinear associations between total sleep duration and 86 disease categories. For 69 of them, the modeled lowest-risk point fell within a six-to-eight-hour window.
That result should not be turned into an instruction to shorten your night.
The Centers for Disease Control and Prevention recommends:
- at least seven hours for adults ages 18–60;
- seven to nine hours for adults ages 61–64; and
- seven to eight hours for adults 65 and older.
The National Heart, Lung, and Blood Institute says most adults need seven to nine hours.
Do not shorten your sleep to chase the study’s six-to-eight-hour window; official guidance still generally starts at seven hours for adults, with age-specific ranges. The research reported group-level associations, not a tested sleep prescription.
Better questions to ask than “Was my deep sleep score good?”
If the dashboard is becoming the first thing you judge every morning, step back and ask:
- Did I give myself enough time to sleep? A perfect-looking graph cannot replace an adequate sleep opportunity.
- Is my schedule reasonably consistent? Bed and wake times provide practical context that a stage score alone does not.
- Do I feel refreshed and able to function during the day? Your lived experience is relevant even when the app looks reassuring.
- Is the problem ongoing? Regular sleep problems, persistent dissatisfaction with sleep, unusual daytime tiredness, fatigue or excessive sleepiness deserve more attention than one odd night.
- Could my environment be part of the picture? For more on a related topic, read our article about sleeping with the lights on after 40.
For a related women’s health topic, read our guide to menopause night sweats and sleep after 40.
Practical habits that do not require a sleep score
CDC guidance emphasizes simple conditions that support sleep:
- keep a consistent bedtime and wake time;
- make the bedroom quiet, relaxing and comfortably cool;
- turn off electronic devices at least 30 minutes before bed;
- avoid large meals and alcohol before bedtime;
- avoid caffeine in the afternoon or evening; and
- exercise regularly and maintain a healthy eating pattern.
These are general habits, not treatments or guarantees. If you have ongoing sleep problems, remain dissatisfied with your sleep or experience unusual daytime tiredness, excessive sleepiness or fatigue, discuss it with a healthcare professional. A clinical evaluation may include a sleep study when appropriate.
The bottom line
The study found associations between algorithmically estimated sleep patterns and later diagnoses. It did not convert a wrist device into a disease detector.
Use a tracker to support curiosity, not certainty. Pay attention to adequate sleep opportunity, a workable routine, how you feel during the day and persistent problems that deserve care. Your long-term health is not contained in one morning’s REM bar.
For more evidence-led guidance without miracle claims, explore the Healthy Aging section.
Frequently asked questions
How much REM and deep sleep should adults over 40 get?
There is no fixed tracker-derived REM or deep-sleep percentage that every adult over 40 should target. Sleep stages vary across the night, with age and between individuals. Consumer devices also estimate stages rather than directly measuring all the signals used in a laboratory sleep study.
Can a smartwatch diagnose a sleep disorder?
No. The American Academy of Sleep Medicine says consumer sleep technology should not replace validated diagnostic testing. Ongoing sleep problems or excessive daytime sleepiness or fatigue should be discussed with a licensed healthcare professional regardless of what a device reports.
Should I aim for six hours because the new study found a six-to-eight-hour window?
No. That window came from observational models across groups and was not a tested sleep prescription. Official guidance still generally starts at seven hours for adults, with age-specific ranges. Do not shorten your sleep to chase the study result.
Why do sleep stages change as we age?
Sleep architecture changes across adulthood, and NINDS notes that people generally spend less time in REM sleep as they age. Individual variation remains substantial, so age does not provide a fixed personal target for a wearable score.
When should I discuss sleep with a clinician?
Consider seeking care when sleep problems are ongoing, you remain dissatisfied with your sleep, or you experience unusual daytime tiredness, excessive sleepiness or fatigue. A tracker can provide context, but symptoms and a clinical evaluation matter more than the app’s stage label.
Sources
- Luo et al., PLOS Medicine: Accelerometer-derived real-world sleep stages and risk of incident diseases
- Science Media Centre: Independent expert reaction to the study
- CDC: About Sleep
- NHLBI: How Much Sleep Is Enough?
- NINDS: Brain Basics—Understanding Sleep
- American Academy of Sleep Medicine: Consumer sleep technology position statement
This article is for general education and does not replace individualized medical advice or a clinical sleep evaluation.