AXIOM SELENE

Evidence Library · Mindfulness & Stress Science

Mindfulness & Stress Science · AXIOM SELENE

Meditation & Stress: What the Evidence Actually Shows

Mindfulness meditation has moderate evidence for reducing anxiety and depression. The cortisol picture is more complicated: 71% of studies find reductions, but effect sizes are likely small, study quality is inconsistent, and many wellness claims go further than the data warrant. Here is an honest read of where the science stands.

Evidence grade

Moderate evidence — Some RCTs or robust observational evidence

6 sources7 documented gapsLast verified: 2026-06-28

What We Mean by Meditation Here

The research base is not about 'meditation' as a single thing. The most studied form is Mindfulness-Based Stress Reduction (MBSR) — an 8-week structured programme of 2–2.5-hour weekly group sessions, a full-day retreat, and daily home practice. Mindfulness-Based Cognitive Therapy (MBCT) is a close variant. Most of what the science actually shows comes from MBSR and MBCT. App-based meditation, brief breathing exercises, and other informal practices have far less rigorous evidence and are not the same intervention.

The Clearest Evidence: Psychological Wellbeing

The most robust evidence for meditation is on anxiety and depression — not cortisol. A landmark 2014 JAMA Internal Medicine meta-analysis pooled 47 RCTs and 3,320 participants to compare mindfulness meditation against active controls (education, relaxation, supportive therapy). The findings hold up in the way meta-analyses with large sample pools tend to: the effects are real but modest.

~ Probable (incomplete evidence)
Mindfulness meditation programmes reduce anxiety symptoms compared with active controls — effect size 0.38 at 8 weeks, declining to 0.22 at 3–6 months.

🅰 Goyal et al. 2014 — JAMA Internal Medicine meta-analysis: 47 RCTs, 3,320 participants (PMC4142584)— Effect sizes are statistically significant but 'small to moderate' by conventional standards. The review describes this as 'moderate evidence'. Effects vs inactive controls would look larger; the active-control comparison is the more conservative and informative benchmark.

~ Probable (incomplete evidence)
Mindfulness meditation programmes reduce depression symptoms compared with active controls — effect size 0.30 at 8 weeks, declining to 0.23 at 3–6 months.

🅰 Goyal et al. 2014 — JAMA Internal Medicine meta-analysis: 47 RCTs, 3,320 participants (PMC4142584)— Authors note the effect is 'comparable with what would be expected from an antidepressant in primary care.' That is a helpful anchor: real, clinically meaningful for some people, but not a dramatic transformation for all.

✗ No evidence found
Mindfulness meditation meaningfully improves positive mood, attention, sleep quality, or substance use — common wellness claims.

🅰 Goyal et al. 2014 — JAMA Internal Medicine meta-analysis: 47 RCTs, 3,320 participants (PMC4142584)— The same 2014 meta-analysis found 'low evidence of no effect or insufficient evidence of any effect' for these outcomes. The absence of evidence is not evidence of absence — but wellness marketing frequently asserts these benefits without adequate support.

Cortisol — What the Studies Actually Find

Cortisol is produced by the adrenal glands as the body's primary stress hormone. Chronically elevated cortisol is associated with cardiovascular risk, immune suppression, and poor sleep. Does meditation reduce it? A 2024 systematic review of 35 studies spanning 1990–2024 is the most comprehensive synthesis available. Its finding is nuanced: a majority of studies show reductions, but the quality of evidence is not strong enough to say why or for whom.

~ Probable (incomplete evidence)
71% of studies (25 of 35) found significant cortisol reductions following mindfulness-based interventions; 29% found no significant change.

🅰 2024 Systematic Review — MBIs and the HPA Axis, 35 studies 1990–2024 (PMC11587421)— Reviewers caution that 'the magnitude of these effects is likely to be small' and that 'the weakness of the studies and the absence of robust designs makes it difficult to establish a causal association.' Some studies also observed cortisol increases in individuals with initially low baseline levels — suggesting a normalisation effect rather than simple suppression.

~ Probable (incomplete evidence)
An earlier 2016 meta-analysis of 5 RCTs (190 participants) found a 'moderately low' overall effect (g = 0.41) of mindfulness programmes on salivary cortisol in healthy adults.

🅱 Frontiers in Physiology 2016 meta-analysis — MBIs and salivary cortisol in healthy adults: 5 RCTs, 190 participants— The small evidence base (5 studies, 190 people total) and high variability in measurement methods limit confidence. Effect sizes ranged from g = 0.03 (raw cortisol data) to g = 0.81 (standardised protocols), showing how sensitive findings are to how you measure.

✗ No evidence found
Meditation reliably and permanently lowers baseline cortisol in stressed individuals — a claim common in wellness marketing.

🅰 2024 Systematic Review — MBIs and the HPA Axis, 35 studies 1990–2024 (PMC11587421)— No current evidence base establishes permanent or sustained cortisol lowering. The 2024 review could not pool studies into a meta-analysis due to high heterogeneity. Long-term follow-up data past 6 months is largely absent.

Sources in this section

Individual RCTs: What Happens in Controlled Settings

Two small but well-controlled trials give a concrete picture of what MBSR produces under real experimental conditions.

~ Probable (incomplete evidence)
An 8-week MBSR programme in Italian mental healthcare professionals produced large reductions in salivary cortisol (η²ₚ = 0.23) compared with a control group whose cortisol rose over the same period.

🅰 2025 RCT — MBSR in Italian mental healthcare professionals, n=37 (PMC12523626)— n=54 (37 completed; ~68% completion rate, 32% dropout). Very large effect size within this study. Significant caveats: single small trial, findings described by authors as 'preliminary evidence.' Results in mental healthcare workers under occupational stress may not generalise to healthy populations.

~ Probable (incomplete evidence)
An 8-week mindfulness programme significantly reduced hair cortisol (a measure of cortisol over weeks, not just at one moment) in stressed university workers, reducing the risk of worsening cortisol levels by 88.8% versus controls.

🅰 2023 RCT — 8-week mindfulness program and hair cortisol in stressed university workers, n=30 (PMC10648523)— n=30 (15 per arm). Hair cortisol is methodologically superior to salivary cortisol for measuring chronic stress load rather than acute spikes. However, the very small sample and 28.6% dropout limit how much weight to place on the effect size. Perceived stress and anxiety also fell significantly in the intervention group.

The Mechanism: Why Meditation Might Affect Cortisol

The biological rationale is coherent even if the evidence is inconsistent. Cortisol is released by the HPA (hypothalamic-pituitary-adrenal) axis in response to perceived threat. Mindfulness training — attention monitoring and experiential acceptance — appears to reduce reactivity in brain regions that trigger this cascade, particularly the amygdala. A 2025 review of RCTs on mindfulness mechanisms notes that reducing psychological appraisal of threat is a plausible path to lower cortisol, and that 'acute stress monitoring combined with acceptance produces the strongest physiological benefits.' The mechanism is biologically plausible. What is missing is a large, well-powered RCT that follows cortisol long enough to say the effect persists.

Sources in this section

What Wellness Marketing Gets Wrong

Several claims circulating in the wellness and meditation app industry go well beyond what the evidence supports. The 2014 JAMA meta-analysis found low evidence for positive mood, improved attention, and sleep benefits — yet these are frequently stated as established effects. The cortisol evidence base is particularly vulnerable to overstatement: 71% of studies finding 'significant reductions' sounds compelling, but when effect sizes are likely small, samples are tiny, and no meta-analysis is possible due to heterogeneity, the correct framing is 'cautiously promising — not proven.' Apps specifically have not been held to the same evidence standard as in-person MBSR programmes.

What we don't yet know

Honesty about gaps in the evidence is what distinguishes us from most wellness media.

  • Long-term cortisol effects beyond 3–6 months: the current evidence base lacks long-term follow-up. We do not know whether cortisol reductions from an 8-week programme persist at 12 or 24 months.
  • Whether app-based or self-guided meditation produces the same cortisol effects as structured in-person MBSR: most research uses in-person, instructor-led MBSR. App-based programmes have not been adequately compared.
  • Optimal dose: program duration and session count appear to influence outcomes, but no clear guidelines exist for the minimum or ideal 'dose' to produce measurable cortisol change.
  • Individual variation: some studies observed cortisol increases in individuals with initially low baseline levels. It is not yet known which individuals benefit most or who might see no effect.
  • Whether cortisol reductions translate to measurable long-term health outcomes (cardiovascular, immune, metabolic): the biological pathway is plausible but not established in human longitudinal data.
  • Active-control quality: many older studies compared meditation to waitlist or no-treatment controls, which inflates apparent effects. The true benefit vs an equally credible active comparator (e.g. structured relaxation, exercise) is less certain.
  • Effects in healthy, non-stressed populations: most studies with significant cortisol findings enrolled participants who were already stressed or at elevated cortisol levels. Findings may not apply to people with normal baseline stress.

All sources

This article summarises what published research has found about meditation and cortisol. It is educational information, not medical advice. It does not constitute a recommendation to start, stop, or change any health practice. If you have concerns about chronic stress or cortisol levels, speak with a healthcare professional.

Last verified: 2026-06-28 · ← Evidence Library