A large-scale analysis published in PLOS Medicine recently put a number on something sleep researchers have suspected for years: not all sleep stages protect your health equally. The study, drawing on data from nearly 95,000 participants in the UK Biobank, found that REM sleep was associated with a lower risk of 83 distinct health conditions, while deep sleep was linked to protection against 7. That's a striking gap. But before you start panicking about your slow-wave percentages or chasing more dream-filled nights, here's what those numbers actually mean.
What the Research Actually Found
The PLOS Medicine study is one of the largest sleep-health analyses ever conducted. Researchers used accelerometer data and self-reported sleep information from UK Biobank participants, then cross-referenced sleep stage patterns against medical records spanning dozens of disease categories. You can read a full breakdown of the findings in this report on REM sleep's link to 83 fewer diseases.
The 83 conditions associated with REM sleep span a wide range, including mental health disorders, neurological conditions, and metabolic diseases. The 7 conditions linked to deep sleep include cardiovascular and metabolic conditions. Notably, the two lists don't overlap neatly. That's actually the point: these two sleep stages are doing very different work inside your body, and the study design captured that distinction at scale.
One important caveat: this is an observational study. It identifies associations, not causes. People who get more REM sleep may have other lifestyle habits, lower baseline stress levels, or genetic factors that also protect against disease. The data is compelling, but it doesn't mean engineering more REM will automatically reduce your disease risk by the same margin.
What REM Sleep Is Actually Doing
REM (rapid eye movement) sleep is the stage most people associate with vivid dreaming. But the biological work happening during REM extends well beyond processing last night's dinner conversation into surreal imagery.
During REM, your brain is highly active. Neural circuits involved in emotional processing are firing, your hippocampus is consolidating memories, and your prefrontal cortex is essentially rehearsing emotional responses to past and future experiences. Research consistently shows that REM deprivation amplifies emotional reactivity, reduces empathy, and impairs decision-making in ways that persist for days.
The connection to 83 disease conditions makes more sense when you understand that chronic emotional dysregulation, poor memory consolidation, and impaired prefrontal function are upstream drivers of a wide range of mental and physical conditions. Poor emotional regulation alone is implicated in anxiety disorders, depression, cardiovascular disease, and immune dysfunction. REM isn't just protecting your mood. It's protecting your physiology.
REM sleep also tends to concentrate in the second half of the night. That's why cutting sleep short by even 60 to 90 minutes, even if total sleep feels adequate, disproportionately slashes your REM exposure. This is one reason sleep consistency matters more than most people expect.
What Deep Sleep Is Actually Doing
Deep sleep, or slow-wave sleep, is when your body does its most intensive physical repair. Growth hormone secretion peaks during deep sleep. Tissue is rebuilt. The immune system ramps up production of cytokines, which are proteins that help fight infection and inflammation. Your cerebrospinal fluid flushes metabolic waste products from the brain through the glymphatic system, including compounds associated with neurodegenerative disease.
The fact that deep sleep was linked to only 7 conditions in the study doesn't mean it's less critical. It may simply mean that its effects are more concentrated on specific physiological systems rather than broadly distributed across disease categories. The cardiovascular and metabolic conditions on that list are among the leading causes of death globally. Protecting against them is no minor achievement.
Deep sleep declines sharply with age. A healthy 20-year-old may spend 20 to 25 percent of their night in slow-wave sleep. By their 60s, that figure can drop to less than 5 percent. This matters for coaches working with older clients, where physical recovery capacity is already under pressure from sarcopenia and reduced hormonal output. Understanding what coaches should tell aging clients about one-session-a-week training becomes inseparable from understanding their sleep physiology.
Why Your Tracker Is a Signal, Not a Sentence
Consumer sleep trackers have gotten better. Devices from Oura, Garmin, Apple, and others now use a combination of heart rate variability, movement, skin temperature, and blood oxygen data to estimate sleep stages. The estimates are reasonably useful at the population level. At the individual level, they're less precise.
Lab-grade polysomnography, the gold standard, uses electrodes directly measuring brain activity. Consumer wearables are inferring stage from proxy signals. Studies comparing wearable estimates to polysomnography show that trackers tend to overestimate light sleep and underestimate deep sleep, with accuracy varying significantly across individuals.
This creates a real problem. If your tracker tells you that you're getting 45 minutes of deep sleep on a night when you actually got 90 minutes, and you spend the following day anxious about your "deficit," you've introduced a stressor that will actively degrade your sleep the next night. Orthosomnia, the clinical term for tracker-induced sleep anxiety, is a documented phenomenon that sleep clinicians are encountering more frequently.
Use your tracker as a directional signal. Consistent patterns over weeks are meaningful. Single-night numbers are noise. If your deep sleep average has trended downward over three weeks alongside other symptoms like slower recovery or mood changes, that's worth investigating. If you had one bad night and your app gave you a low "sleep score," that's not a crisis.
How to Support Both Stages Without Obsessing
The good news is that most of the behaviors that support deep sleep also support REM. They're not in competition. You're not optimizing one at the expense of the other.
Here's what the evidence consistently supports:
- Consistent sleep and wake times. Your circadian rhythm regulates which sleep stages appear when. Irregular timing disrupts both REM and deep sleep architecture. Set a wake time and hold it, including on weekends, even if your bedtime varies slightly.
- Temperature management. Your core body temperature needs to drop to initiate and maintain deep sleep. A cool room, typically between 65 and 68°F (18 to 20°C), supports this. A warm shower 60 to 90 minutes before bed accelerates the core temperature drop through peripheral vasodilation.
- Limit alcohol. Alcohol suppresses REM sleep significantly. It may help you fall asleep, but it fragments the second half of the night when REM predominates. Even moderate consumption, two drinks in an evening, measurably reduces REM duration.
- Exercise timing and type. Regular physical activity increases slow-wave sleep. The timing matters somewhat. High-intensity exercise within 2 to 3 hours of bedtime can delay sleep onset, though this varies by individual. Morning or afternoon sessions are generally more sleep-neutral. For yoga practitioners specifically, research on morning versus evening yoga and sleep quality offers useful practical guidance on timing your practice.
- Manage stress before bed. REM sleep quality is directly influenced by your baseline stress and cortisol levels going into sleep. Techniques like coherent breathing can meaningfully shift your nervous system state. Evidence from a randomized controlled trial found that coherent breathing reduced stress by 31% in workplace settings, suggesting it's a practical tool, not just a wellness cliché.
- Protect your sleep window. This is the most underrated lever. REM concentrates late in the night. Deep sleep concentrates early. Cutting sleep from either end costs you something specific. A full 7 to 9 hours gives your body access to both.
The Bigger Picture: Sleep as Recovery Infrastructure
Sleep isn't a passive state. It's active biological infrastructure for everything your waking life demands of your body and mind. The PLOS Medicine data reinforces what clinical sleep medicine has argued for decades: no single stage is dispensable, and optimizing sleep isn't about gaming individual metrics.
If you're already training hard, the relationship between sleep and recovery is especially direct. What science says actually works for recovery in 2026 puts sleep at the top of the list, above supplements, cold therapy, and compression gear. That's not a soft finding. It's the consistent signal across decades of exercise science research.
And if you're approaching sleep optimization from a mental health angle, note that the same lifestyle factors that protect REM sleep quality, consistent schedules, stress management, limited alcohol, also independently reduce anxiety and depression risk. The mechanisms overlap. You're not managing separate systems. You're managing one integrated system that expresses itself in your sleep quality, your emotional resilience, and your long-term disease risk simultaneously.
The REM versus deep sleep debate is useful for understanding biology. It's less useful as a framework for daily decisions. What protects your health is consistently sleeping enough, sleeping regularly, and removing the obvious disruptors. Your tracker can help you notice patterns. It shouldn't be running your life at 6 a.m. on a Tuesday because you only hit 18 minutes of deep sleep.
Both stages matter. The question isn't which one to prioritize. The question is whether you're giving your body enough time and the right conditions for both to do their work.