Sleep architecture and HRV: a mirror of the autonomic nervous system

Sleep is not homogeneous — it is a choreography of distinct phases, each with its own metabolic role. Morning heart rate variability (HRV) tells you precisely how well that choreography went and how resilient your autonomic nervous system is.

The four sleep phases and their metabolic role

• N1 — transition (< 5% of the night): the brain switches off
• N2 — light sleep (~50%): pulse slows, body temperature drops, memory consolidation begins
• N3 — deep slow-wave sleep (~20%): growth hormone, glymphatic clearance, declarative memory consolidation
• REM — paradoxical sleep (~25%): emotional processing, procedural memory integration, synaptic pruning

The length of N3 determines how well your brain clears beta-amyloid and how efficiently mitochondrial ATP production is restored. The breakthrough work of Xie and colleagues (Science, 2013) showed that the interstitial space of the brain expands by about 60% during sleep, dramatically accelerating glymphatic washout of waste products. Chronic N3 deficit is a strong long-term risk factor for metabolic disease and Alzheimer's (Mander et al., Neuron 2017).

Sleep duration and cardiometabolic risk

The Cappuccio et al. meta-analysis (Sleep, 2010) included over 470,000 participants: short sleep (< 6 h) raised the risk of death by 12% and of cardiovascular disease by 48%. Excessively long sleep (> 9 h) is also associated with elevated risk, often reflecting poor sleep quality or comorbidity.

HRV — a yardstick of vagal tone

HRV is the variation in time between individual heartbeats (RR intervals). The higher the HRV (e.g. RMSSD or SDNN), the stronger the parasympathetic (vagal) tone and stress resilience. Shaffer and Ginsberg (Front Public Health, 2017) provide a thorough overview of HRV as a marker of autonomic flexibility.

Low morning HRV predicts higher fasting insulin, worse glucose tolerance and elevated baseline cortisol. Thayer's meta-analysis (J Affect Disord, 2012) links low HRV with systemic inflammation and depression — all mechanistically connected through the cholinergic anti-inflammatory loop.

"If morning HRV drops for several consecutive days, it is not a bad day — it is a signal of systemic overload."

Practical protocol to improve sleep architecture

• Stable sleep window ± 30 min, weekends included — strengthens the circadian rhythm
• Morning daylight 10–20 min in the first hour (≥ 1000 lux)
• Last large meal at least 3 h before sleep (protects deep sleep)
• Bedroom temperature 17–19 °C and complete darkness
• Orange glasses or a software blue-light filter 2 h before sleep
• Caffeine before 14:00 (half-life 5–7 h)
• Avoid alcohol within 3 h of sleep — it suppresses REM and breaks the second half of the night

Breathing and HRV — a fast intervention

Slow resonant breathing (~6 breaths per minute, 5 seconds in, 5 out) raises HRV during the session itself, and with regular practice also raises baseline. Lehrer and Gevirtz (Front Psychol, 2014) have clinically validated the method for anxiety, hypertension and insomnia.

What to track next

I recommend tracking morning HRV and percentage of deep sleep (e.g. Whoop, Withings, Oura, Garmin) for 4 weeks before and after each intervention. That gives objective feedback on whether magnesium, breathwork or Pulsetto really affect your body. Watch the trend, not the single day — HRV is a noisy daily signal.

Red flags that require medical assessment

• Snoring + morning fatigue + daytime sleepiness → rule out sleep apnoea (polygraphy)
• HRV persistently < 20 ms RMSSD in a young adult — assess autonomic dysfunction
• Deep sleep chronically < 10% of total sleep time — a sleep study is justified

How caffeine and alcohol affect sleep architecture

Caffeine's half-life is 5–7 h, but in slow CYP1A2 metabolisers (~45% of Europeans) up to 10 h. That means a 14:00 coffee still contains ~25% of its active dose at 22:00 and cuts deep sleep by ~20% (Drake et al., J Clin Sleep Med 2013). Rule: caffeine stop 8–10 h before bed.

Alcohol shortens sleep onset but suppresses REM in the first half of the night and causes wake-ups after 3–4 h (Ebrahim et al., Alcohol Clin Exp Res 2013). Even 1–2 drinks at dinner lower next-night HRV by 15–25%.

Blue light and melatonin — what is proven, what is myth

Screen blue light 2 h before bed suppresses melatonin secretion by ~55% (Chang et al., PNAS 2015). Blue-blocker glasses reduce this by ~15–20% but do not replace cutting screen time. What actually works: dim room lighting 2 h before bed + screen blue-light filter + phone night mode.

• Melatonin 0.3–0.5 mg 4–5 h before bed helps jetlag and delayed sleep-phase disorder
• High doses (3–10 mg) are not more effective and often cause morning grogginess
• It is not a sleeping pill for a healthy person who does not want to sleep — it is a circadian signal
• Pregnancy and children — physician supervision only

Sleep apnoea — the most underdiagnosed cause of chronically low HRV

If you snore + wake tired + weight is climbing + HRV stays low despite doing everything 'right' — suspect obstructive sleep apnoea. Home polygraphy gives AHI (apnea-hypopnea index) and oxygen desaturation data. Untreated apnoea roughly doubles cardiovascular and diabetes risk (Marin et al., Lancet 2005).