Sleep is the underestimated longevity tool

Some levers of life expectancy appear in every textbook. Not smoking, staying active, maintaining social connections. Sleep usually makes this list, but rarely at the top. A 2025 analysis published in Nature Medicine suggests that is a mistake.

Argentieri et al. from the University of Oxford analysed data from 492,567 UK Biobank participants over more than twelve years. The finding: sleep belongs in the top tier of longevity factors, alongside smoking status, physical activity, and social connectedness.

What makes this study different

Three points set it apart from the usual body of sleep research.

First, scale and statistics. Nearly half a million individuals, with genetic background and plasma proteomics. That makes the effect estimates robust.

Second, the comparative perspective. The 25 identified lifestyle and environmental factors explained 17 percentage points more of the variation in mortality than polygenic risk scores for 22 major diseases. Lifestyle clearly outperforms genetics.

Third, the multiplier logic. Each of the 25 top exposures was associated on average with 22 ageing biomarkers and 15 ageing-related diseases. Sleep was among those with a particularly broad reach. Poor sleep quality does not act on one disease. It acts on many simultaneously.

What happens during sleep

Sleep is the phase in which the body performs repair work that would not be possible while awake:

  • Clearance of metabolic waste from the brain via the glymphatic system
  • Hormonal regeneration (growth hormone, testosterone, leptin, ghrelin)
  • Modulation of the inflammatory response
  • Consolidation of memory content
  • Rebuilding of immunological reserves

Those who sleep too little or too restlessly pay the price not in one system, but in several at once.

The range that matters

The study and the broader epidemiological evidence converge on a window of seven to nine hours per night. Fewer than six hours consistently shows elevated risks across multiple endpoints. More than nine hours is not trivially beneficial. It is frequently a marker for undetected illness.

More important than the raw hour count is quality. Sleep architecture, deep sleep proportion, REM phases, wake frequency, and HRV trajectory across the night say more than the time between lights out and the alarm.

Measuring precisely rather than guessing

This is exactly where the SLOW approach becomes visible. Holistic and precise.

Precise means that sleep does not work as self-reported data. It works as a measurable system. Wearable sensors (polysomnography-validated wearables), HRV trajectories, cortisol diurnal profile, melatonin and iron status, thyroid values. Each client receives not the average tip, but the data picture of their own night.

Holistic means that sleep is not treated in isolation. It is interlocked with stress load, training intensity, light exposure, caffeine and alcohol patterns, hormonal cycles, and nutrition. Pulling one sleep lever changes several systems at once. And the reverse is equally true.

Relevant biomarkers for practice

When the hour count is right but quality is not, we draw on these markers:

Hormonal markers:

  • Cortisol diurnal profile (4-point saliva test)
  • Melatonin (saliva, evening)
  • Testosterone/oestradiol (sex-dependent)
  • Thyroid values (TSH, fT3, fT4)

Nutrient status:

  • Magnesium (intraerythrocytic)
  • Vitamin D3
  • B vitamins (particularly B6, B12, folate)
  • Iron status (ferritin, transferrin saturation)

Inflammatory markers:

  • hsCRP
  • IL-6
  • TNF-alpha

Combined with continuous HRV monitoring and validated wearable technology, a precise picture of sleep quality emerges. Along with the levers where meaningful change is possible.

From data to results

The Oxford study confirms what we see in practice: sleep is not a nice-to-have. It is a central longevity lever. But only when we measure it precisely and optimise it systematically.

How do you measure sleep in practice? Which markers do you draw on when the hour count is right but quality is not?