Back to Stress

The science · evidence review

The science of stress

How the stress response works in the body, why recovery is the real lever, and what trained resilience actually looks like at a physiological level.

9 min read2 sourced studiesBy Healicus editorial

If you ask a careful clinician what variable they wish their patients took more seriously, "chronic stress" is on most short lists. The reason isn't soft. The mechanisms by which chronic stress accelerates ageing (sustained sympathetic activation, cortisol dysregulation, low-grade inflammation, sleep disruption, blood pressure elevation) show up in cohort data as meaningful predictors of mortality risk. The framework of allostatic load, the cumulative cost to the body of repeatedly mobilising stress responses, has been studied across decades and consistently associates with worse cardiovascular, metabolic, and cognitive outcomes.

The intervention is unromantic. It doesn't require a supplement, a wearable, or a subscription. It does require something that's often harder than buying a thing: a daily practice, sustained over years, of doing less and breathing slower. The honest summary of the literature is that this works, and that almost everyone underweights it.

This guide covers what the research actually says about stress and ageing, the practices that have accumulated evidence, a short note on how older traditions framed the territory, and where the contemporary discourse, particularly around heart rate variability and "vagal tone," has run ahead of what the evidence cleanly supports.

The acute stress response is, in evolutionary terms, well-engineered. Faced with an immediate threat, the sympathetic nervous system mobilises rapid resources: heart rate up, blood pressure up, attention narrowed, glucose mobilised, digestion paused, immune function temporarily suppressed. The system is designed to switch off when the threat resolves and to hand control back to the parasympathetic ("rest and digest") branch.

Chronic stress is what happens when the system fails to switch off. The threat may be subtle (financial worry, social isolation, sleep deprivation, work overload) but the physiological response runs anyway, hour after hour, day after day. The accumulated cost is what the allostatic load framework names: glycaemic dysregulation, hypertension, low-grade inflammation, immune dysfunction, sleep architecture damage, accelerated telomere shortening, hippocampal volume loss.

The cohort data is consistent. Higher allostatic load, assembled from multiple biomarkers including cortisol, inflammatory markers, blood pressure, and metabolic indicators, predicts cardiovascular disease, cognitive decline, and all-cause mortality across populations. The effect sizes are not heroic, but they are real and they are independent of other risk factors.

The longevity-relevant adaptations therefore live not in eliminating stress (impossible) but in restoring the parasympathetic balance more reliably. The research-supported practices are several, and they're surprisingly accessible.

If you read longevity content from the past decade, you've encountered heart rate variability (HRV) framed as a key biomarker. The basic premise is real: HRV, specifically the higher-frequency component reflecting parasympathetic tone, has been studied across cohorts and lower HRV consistently associates with cardiovascular mortality and all-cause mortality, particularly in older adults. The 2018 cohort literature on HRV-mortality associations is strong enough that HRV has clinical application in cardiac risk stratification.

The question is what to do with this as an individual outside of clinical care. Here the discourse has run ahead of the evidence in important ways.

What's well-supported:

  • HRV is meaningfully trainable through breathwork and consistent autonomic nervous system practices.
  • Higher resting HRV (over months, not days) is a reasonable proxy for parasympathetic recovery capacity.
  • Acute drops in HRV reasonably indicate acute physiological stress (illness, sleep deprivation, training load).

What's less supported:

  • That a daily HRV score from a wearable provides actionable individual guidance better than self-report (sleep, mood, training-as-felt).
  • That an "HRV optimisation" protocol produces longevity benefits beyond what the underlying practices (sleep, breath, exercise) produce on their own.
  • That HRV-derived "biological age" or "recovery score" metrics from consumer wearables are accurate enough to drive day-to-day decisions.

The honest take: HRV is a real signal, useful at population level and in clinical research. The consumer-wearable HRV ecosystem has run ahead of what the underlying signal can support at the individual day-to-day level. Trend over months: meaningful. Daily fluctuation: noisy. Don't optimise against the noise.

Several stress-regulation practices have produced consistent evidence across clinical trials. None requires equipment. All are free. The differences between them are smaller than their proponents suggest.

Slow paced breathing

Of all the breath practices, the strongest research signal is for slow paced breathing, typically 5 to 6 breaths per minute, with extended exhalation, performed for 5 to 20 minutes at a time. Mechanistically, slow breathing engages baroreceptor reflexes and increases vagal tone reasonably acutely (Russo et al., Breathe 2017). Clinically, small and mostly short-term trials have reported effects on resting blood pressure and anxiety; the evidence is mixed.

The specific protocol matters less than the consistency. 4-7-8 breathing (inhale 4, hold 7, exhale 8), box breathing (4-4-4-4), and coherent breathing (5.5 breaths per minute, equal inhale/exhale) all engage the same physiological systems. Pick one. Do it daily.

Meditation, particularly mindfulness

Mindfulness-based stress reduction (MBSR) is now in widespread clinical use. A meta-analysis of 47 trials of meditation programmes (n=3,515) found moderate-strength evidence for small improvements in anxiety, depression and pain, low-strength evidence for stress and distress, and insufficient evidence for sleep (Goyal et al., JAMA Intern Med 2014). The standard MBSR protocol is 8 weeks of weekly sessions plus daily practice; effect sizes are modest.

The longevity-relevant point is not that meditation is uniquely beneficial (it isn't) but that it provides a structured way to reliably engage parasympathetic recovery, daily, in a form that is sustainable across decades. Other practices that engage the same physiology (long walks, contemplative reading, time in nature) work via similar mechanisms, and the research signal for them is comparable when controlled for time spent.

What works in the trials is approximately:

  • Daily practice (5–20 minutes is sufficient for most of the documented benefit).
  • Sustained over months, not days.
  • Without competitive optimisation (the practice is undermined by treating it as a productivity activity).

Time in nature

The research base on nature exposure has grown in the past decade. Reported effects on cortisol, blood pressure, immune markers, and self-reported wellbeing are modest and come mostly from small or observational studies. The Japanese practice of shinrin-yoku ("forest bathing") has been studied, mostly in small trials.

Practical translation: regular outdoor time, particularly in green or blue spaces, is a low-friction practice that tends to come with other useful inputs (movement, sunlight exposure, social connection), even though the evidence for nature exposure on its own is modest.

Sauna and cold exposure

Both have research signals worth knowing about. In the Finnish sauna cohort (Laukkanen et al., JAMA Intern Med 2015) (n=2,315 men, about 20-year follow-up), regular sauna use (4+ times per week) was associated with lower cardiovascular mortality. This is an observational cohort; it does not show cause and effect. Proposed mechanisms include heart rate elevation, peripheral vasodilation and heat-shock protein responses.

Cold exposure has thinner cohort data but plausible mechanistic support: brief cold immersion produces a strong sympathetic stimulus followed by parasympathetic rebound, and there's evidence for mood and inflammatory marker effects in small trials. The "Wim Hof" protocols sit at the more aggressive end and have less rigorous evidence than their cultural prominence suggests.

The practical guidance is unspectacular. Sauna use is associated with lower cardiovascular mortality in a Finnish cohort; this does not show cause and effect. Avoid sauna with unstable heart disease, a recent heart attack, severe aortic stenosis, or after alcohol. Cold exposure is more variable, lower priority than the other practices, and should be used cautiously (cold-water immersion is not appropriate for everyone, particularly those with cardiovascular conditions).

Many cultures developed practices for what would now be called autonomic regulation: pranayama (breath work) and meditation in Ayurveda, qigong and tai chi in Chinese medicine, and structured attention training in contemplative traditions from vipassana to zen. Blue Zones populations have their own social rhythms, such as the Okinawan moai (small groups meeting weekly for support). This is cultural and historical context. It is not evidence that these practices work, and it is not a recommendation. Where individual practices such as tai chi or meditation have been tested in trials, the catalogue entries describe what those trials found and how certain the evidence is.

A few areas where the picture has changed meaningfully:

Vagal tone has become a popular but loosely-used concept. The underlying physiology is real; the consumer discourse has overgeneralised. Polyvagal theory, the framework popularised by Stephen Porges, has aspects that are well-supported and aspects that remain contested in the academic neuroscience literature. The practical practices that fall under its umbrella (breathing, social engagement, safe-environment exposure) have their own independent evidence and don't depend on polyvagal theory being wholly correct.

Wearable stress metrics are noisier than they appear. The HRV-based "stress scores" from consumer wearables have face validity but limited day-to-day individual guidance value. Use them for trends across months; don't use them to make individual daily decisions.

The gut-brain axis has matured as a research area. The bidirectional connection between gut microbiome and stress regulation is now substantial. The practical translation is consistent with everything else: support the microbiome via varied plant fibre and fermented foods, and the autonomic regulation downstream is incidentally supported.

Loneliness has been recognised as a chronic stressor. Cohort data link social isolation with higher mortality, which has moved social connection from a soft variable to a structural longevity factor in the research literature.

A handful of stress-adjacent claims worth flagging:

  • "Cortisol balancing" supplements marketed direct-to-consumer have weak evidence. The evidence for ashwagandha is of very low certainty, and liver injury has been reported with ashwagandha products.
  • Float tanks, sound therapy, biofeedback machines: face validity yes, longevity-altering evidence no. Pleasant practices that add modest value at best; not rate-limiting in someone who isn't doing the basics.
  • Aggressive HRV-coached training as a longevity intervention has not been demonstrated. The basics (sleep, breath, exercise, social connection) produce the HRV improvement that the advanced protocols later try to optimise from.

If you are dealing with severe chronic stress, anxiety, depression, or trauma, none of the above replaces clinical care. Stress-regulation practices are foundational, not corrective for psychiatric conditions, and they work best as part of a broader treatment plan when one is needed. If you're not sure whether what you're carrying is normal life stress or something requiring help, the answer is to ask your doctor. There is no virtue in handling things alone that don't need to be handled alone.

If you wanted to take the most-defensible stress-regulation practices and assemble a starting place, the shape is roughly:

  1. Two minutes of slow breathing daily. 4-7-8 or coherent breathing. The specific protocol matters less than the consistency.
  2. Ten minutes of meditation, daily, sustained for months. Pick a tradition or app that fits. Don't optimise against your own progress.
  3. Time outdoors regularly. Twenty minutes a day in green space if you can; weekly nature exposure if you can't.
  4. One stress-regulating practice a week that engages the body: yoga, tai chi, qigong, sauna, swimming, a long walk.
  5. Sufficient sleep, which is itself a stress-regulation intervention. See the sleep guide.
  6. Strong social connection, which the cohort data link with better health outcomes.

That's most of what the literature converges on. None of it is novel. The unromantic truth is that doing simple stress-regulation practices daily, for years, is the highest-leverage intervention available to most people.

The leverage is in the consistency.

Educational reference. Not medical advice. Not intended to diagnose, treat, cure, or prevent any disease. Speak with your physician before changing your diet, supplement, or exercise routine, especially if you have a medical condition or take prescription medication.