Stress
Chronic stress is linked with faster biological aging, and the interventions that counter it are unglamorous: slow breathing, mindfulness, exercise, sleep, and other people. The "stress hacks" sold on the internet mostly don't work.
The aging body has two ways of running: parasympathetic-dominant (rest, repair, immune surveillance, autophagy) and sympathetic-dominant (fight-or-flight, glucose mobilization, inflammation primed). Acute stress is fine — the system was designed for it. The modern problem is chronic sympathetic activation: a low-grade, never-resolved stress response that locks the body out of repair mode for years at a time. The accumulated cellular cost of that lock-out is real but only partly measured.
What chronic stress actually does to the body
Acute stress is a survival circuit: the HPA (hypothalamic–pituitary–adrenal) axis fires, cortisol and catecholamines mobilize energy, attention narrows, immune resources reposition. When the threat ends, the parasympathetic side takes over and the system reconsolidates. Healthy stress physiology is fast on, fast off.
Chronic stress is the failure mode where the "off" never arrives. Systemic consequences pile up across organ systems:
- Oxidative stress. Stress hormones can switch off Nrf2, the cell's main antioxidant control switch: this has been shown in the livers of restrained rodents and in liver cells in the lab, and brain tissue from people who had depression shows lower Nrf2 levels [1].
- Mitochondrial. Energy demand rises, oxidative phosphorylation efficiency drops, reactive oxygen species (ROS) accumulate. Neurons and cardiomyocytes — high-energy cells with limited regeneration — pay the steepest cost.
- Immune. Initially activated, then exhausted. Chronic glucocorticoid exposure shifts immunity toward a pro-inflammatory, less-effective profile — the same pattern seen in normal aging ("inflammaging").
- Brain. Months of very high cortisol, as in Cushing's syndrome, go with a smaller hippocampus and poorer memory [2], as does long-term steroid medicine [3]; whether ordinary chronic stress does the same is not established. Prefrontal-amygdala connectivity weakens; rumination loops entrench.
- Behavioral mediation. Stress doesn't only act biologically — it pushes people toward worse sleep, worse diet, more alcohol and tobacco, less exercise. In a 2026 study of about 1,300 US adults, much of the link between stress and faster epigenetic aging ran through these behaviors: for day-to-day perceived stress they explained all of it, while major life events kept a direct link to one clock (GrimAge). The data were a single snapshot, and stress hormones were not measured, so a direct hormonal route remains possible [4]. The implication is practical: managing stress without also fixing the secondary behaviors likely leaves much of the damage on the table.
This is the framework geroscience now uses: chronic stress is upstream of several hallmarks of aging — altered intercellular communication (inflammaging), mitochondrial dysfunction, and epigenetic alterations — rather than being a separate "psychological" issue.
The epigenetic clock signal
The clearest signal linking chronic stress to faster biological aging comes from DNA methylation clocks — algorithmic estimates of biological age built from CpG-site methylation patterns. The second-generation clocks PhenoAge and GrimAge and the third-generation pace-of-aging measure DunedinPACE all predict death, cognitive loss, weaker strength and poorer mobility in about 6,300 older adults in the US, Ireland and Northern Ireland [5]. In US data, though, everyday facts such as income, education, mental health and health habits predicted later health as well as the clocks or better [6].
In the 2026 study described above (1,308 middle-aged US adults, average age 54), people who reported more stressful life events had faster-running epigenetic clocks on all three measures tested (GrimAge, PhenoAge and DunedinPACE), while higher day-to-day perceived stress was clearly linked only to GrimAge. The perceived-stress link ran entirely through worse sleep, poorer diet and less exercise; part of the life-events link ran through smoking, drinking and poor sleep. The data were a single snapshot of self-reports, so they cannot show which came first [7]. A companion analysis of 531 adults from the same US cohort measured gene-activity markers rather than clocks: more childhood adversity went with higher levels of the stress hormone noradrenaline (norepinephrine), which tracked markers of DNA damage and age-related inflammation, but stress in adulthood showed no such link [8]. Following the same people over time points the same way, with a caveat: in seven military and civilian cohorts (1,367 people), those who newly developed post-traumatic stress disorder (PTSD) aged faster on an older clock (Horvath) by their next blood test, but not on GrimAge [9].
Whether low heart-rate variability (HRV) predicts faster epigenetic aging has barely been studied: the only small study found (86 adolescents) linked faster GrimAge to higher, not lower, heart-rate variability during stress tasks [10].
Stress and hard health endpoints
Beyond biomarkers, psychosocial stress tracks with cardiovascular disease and mortality — moderately, and with the usual confounding caveats (reverse causation, residual confounding by health behaviors, documented publication bias). Evidence rating: Moderate. The effects below are given as hazard, risk, or odds ratios, where a value above 1.0 means higher risk; the bracketed 95% confidence interval (CI) is the range where the true value most plausibly lies, and when it crosses 1.0 the result is compatible with no effect.
- Heart attack. In a 52-country case-control study of 11,119 patients with a first myocardial infarction (MI, heart attack) and 13,648 controls, chronic stress roughly doubled the odds of a first heart attack — permanent work stress carried an odds ratio (OR) of 2.14 and permanent home stress 2.12, consistent across regions and sexes — though the case-control design is vulnerable to recall bias.[11]
- Job strain. High-demand, low-control jobs are linked to somewhat more heart disease, but the effect is small and may not be causal. The 2012 pooled analysis of 197,473 workers found about 23% more coronary heart disease (HR 1.23, 95% CI 1.10–1.37); the effect was larger in published than unpublished cohorts (HR 1.43 vs 1.16), and even if fully causal, job strain would explain only about 3.4% of coronary disease [12]. A 2026 review pooled 21 studies at about 14% more ischemic heart disease (relative risk 1.14, 95% CI 1.06–1.23) and about 8% more heart attacks (1.08, 1.00–1.15); studies that rated jobs by job title rather than self-report found no clear link, and the authors judged the evidence for cause limited, a conclusion other researchers have disputed [13]. In four cohorts of employed adults aged 30–60 followed for more than 23 years, high job strain went with about 14% more cardiovascular deaths (HR 1.14, 95% CI 1.01–1.27) [14].
- Perceived stress. Across 6 prospective studies (n≈119,696), high vs low perceived stress predicted incident CHD about 27% more often (RR 1.27, 1.12–1.45) — roughly the increment of smoking a few more cigarettes a day.[15]
- Allostatic load — the cumulative multi-system "wear and tear" of chronic stress (a composite of neuroendocrine, cardiovascular, metabolic, and immune markers) — predicts about 22% higher all-cause mortality (HR 1.22, 95% CI 1.14–1.30) and 31% higher CVD mortality (HR 1.31).[16] Heterogeneity is high and there is no consensus way to compute the index; read it as directional, not a clinical decision tool.
- Diagnosed stress disorders. A 2026 review of 165 studies found that people with PTSD and other stress-related disorders died about 39% more often than the general population (RR 1.39, 95% CI 1.15–1.67, from 21 studies), including more deaths from natural causes, not only suicide. Compared with people who had the same other physical or mental illnesses, the gap was no longer clear, so much of the extra risk travels with those illnesses [17].
The same behavioral-mediation lens applies here as to the epigenetic-clock data: much of the stress→disease link runs through sleep, diet, alcohol, tobacco, and exercise, not stress hormones alone.
How stress is actually measured
There is no single number for stress. The practical toolkit:
- Perceived Stress Scale (PSS-10/PSS-14) — the most widely used validated self-report of global perceived stress; public domain.[18]
- Cortisol dynamics. The better-validated marker is not high total cortisol but a flattened diurnal slope — across 80 studies, flatter slopes tracked with worse outcomes in 10 of 12 health domains (strongest for immune/inflammatory) and higher mortality.[19] The cortisol awakening response and hair cortisol (chronic exposure) are complementary. Chronic stress dysregulates cortisol rhythm rather than simply raising it.
- HRV — a direct autonomic read; see HRV.
- Allostatic load index — the multi-system composite above.
Appraisal and mindset — weak / preliminary
How you interpret stress matters a little. A "stress-is-enhancing" mindset, manipulable with brief videos, shifts cortisol reactivity toward a more moderate range under acute stress, though effects are small and lab-bound.[20] In a US national sample, people reporting a lot of stress and that stress harmed their health had about 43% higher risk of premature death (HR 1.43) — but adjusting for self-rated health cut that to a borderline 18% (HR 1.18), suggesting the perception partly reflects already-poor health.[21] The appraisal literature is real but modest, and is no substitute for changing the objective stressor.
What actually works to lower it
The interventions with the best evidence are pleasingly old-fashioned. There is no nootropic shortcut.
Mindfulness and meditation — moderate
Formal mindfulness programs (Mindfulness-Based Stress Reduction, MBSR, and Mindfulness-Based Cognitive Therapy, MBCT) have been studied extensively; MBSR, the older of the two, for over four decades. In adults without a mental-health diagnosis, teacher-led mindfulness courses reliably lower psychological distress compared with doing nothing, by a small-to-moderate amount (a standardized effect of 0.32 across 2,371 people in 13 trials) [22]. Across 136 trials, though, they did not clearly beat other purpose-built active programs, so mindfulness is one good option among several rather than a uniquely powerful one [23]. It is low-risk but not risk-free: pooling 83 studies, about 1 in 12 meditators reported an unwanted effect (about 1 in 27 in experimental studies), most often anxiety, low mood or unusual thinking, sometimes in people with no history of mental-health problems [24]. Against active controls (the gold standard), the effects are modest: a rigorous 2014 meta-analysis found a small-to-moderate benefit for anxiety (standardized mean difference, SMD, 0.38; 95% CI 0.12–0.64) at 8 weeks, fading to 0.22 at 3–6 months, plus depression 0.30 and pain 0.33, with only low-strength evidence of less stress and distress, low or insufficient evidence for mood, sleep, or weight — and no evidence that meditation outperformed any other active treatment.[25] It does hold its own against a standard drug, though: in 276 adults with a diagnosed anxiety disorder, an 8-week MBSR course reduced anxiety about as much as the antidepressant escitalopram, with far fewer side effects (15% vs 79% had at least one study-related side effect); patients knew which treatment they got and there was no placebo group [26]. Much larger figures sometimes cited (Hedges' g ≈ 0.97 anxiety, 0.95 mood) are uncontrolled, within-group pre-to-post estimates in patients with diagnosed disorders, not stress-reduction effects against a control.[27]
Mechanism, in short: deliberate present-moment attention with non-judgmental acceptance interrupts the prefrontal-amygdala-default-mode rumination loop, and may lower tonic cortisol output; whether it shifts resting autonomic balance is unclear (see HRV). Pooled trials show meditation lowering several physiological stress and inflammation markers compared with active control activities [28].
A 2025 systematic review of lifestyle interventions for healthy aging (Joshi et al.) found a moderate reduction in stress from mindfulness-based interventions (SMD −0.65) alongside a 25% reduction in dementia risk for sustained cognitive-stimulation programs.[29] The studied course runs 8 weeks; in these courses people typically manage about 30 minutes most days, about two-thirds of what is assigned [30].
Slow-paced ("resonance") breathing — moderate, fast-acting
Breathing at ~6 breaths per minute (about 5 seconds in, 5 seconds out) hits the resonance frequency of the human cardiovascular reflex loop, maximizes respiratory sinus arrhythmia (the natural rise and fall of heart rate with each breath), and acutely raises HRV. The acute physiology is well-demonstrated: in 20 people with high blood pressure, a single session of breathing at 6 breaths per minute raised baroreflex sensitivity (the reflex that buffers blood-pressure swings) and lowered blood pressure by about 9/5 mmHg during the session [31]. Across 15 randomized trials of various breathing exercises (slow, deep, yogic pranayama and alternate-nostril), regular practice lowered blood pressure by about 7/3 mmHg, though the trials carry some risk of bias [32]. Practiced over several weeks, slow breathing raises vagal heart-rate variability [33], and breathing exercises (mostly slow-paced) lower self-reported stress by a small-to-medium amount, about a third of a standard deviation (g −0.35), roughly the size of the mindfulness effect [34]. Whether resting HRV stays higher is less clear: in a 4-week slow-breathing biofeedback trial of 88 people, stress fell but resting HRV did not change [35]. Its real advantage is that it is cheap, safe and quick. Full mechanism in HRV.
Exercise — strong
Fitter, more active people react less to acute stress: their heart rate and systolic blood pressure rise less [36], and about half of studies also find a smaller cortisol surge, mostly in snapshot comparisons rather than trials [37]. Across about 1,000 trials in adults, being more active gave medium-sized improvements in depression, anxiety and psychological distress, and harder exercise helped more than simply adding minutes; in depression trials, programs of 150 minutes a week or less did at least as well as longer ones [38], [39]. Most of that evidence is low quality and measures mood and distress rather than stress itself. The 150 min/week target still matters for heart and metabolic health. Combining meditation with aerobic exercise is promising but rests on small pilots: in women with a history of sexual violence, meditation followed by aerobic exercise twice a week reduced rumination, while either alone did not [40]. See Zone 2, Resistance training, and VO₂ max.
Sleep — strong, bidirectional
Stress fragments sleep; fragmented sleep amplifies stress reactivity the next day; the loop is what entrenches the chronic state. For someone in chronic stress, fixing sleep timing and consistency is a sensible first step before adding a meditation practice on top of broken sleep. See Sleep, and specifically Insomnia treatment: cognitive behavioral therapy for insomnia (CBT-I) improves sleep quality substantially (SMD 0.74) [41], and better sleep in turn lowers stress: across 65 trials (8,608 people), interventions that improved sleep reduced stress by a small-to-medium amount (g −0.42) and rumination slightly more (g −0.49), with larger sleep gains bringing larger benefits [42].
Social connection — strong (and underrated)
Holt-Lunstad's meta-analysis of 148 cohorts established that strong social relationships are associated with a 50% greater likelihood of survival vs. weaker ones — an effect its authors judged comparable with quitting smoking and larger than well-known risk factors such as obesity and physical inactivity.[43] Mechanistically, supportive social interaction directly downregulates the HPA axis through the same neural pathways that the relaxation response engages — the brain's reward-and-safety circuitry (ventral striatum, ventral tegmental area, and medial prefrontal cortex). A later meta-analysis quantified the downside: social isolation, loneliness, and living alone each raised mortality risk by roughly a quarter to a third (OR 1.29, 1.26, and 1.32 respectively) — comparable to established risk factors.[44] Depth matters more than headcount: in the 2010 pooling, measures of how deeply someone is woven into a network predicted survival best (OR 1.91), while simply living with others was a weak, uncertain predictor (OR 1.19, 95% CI 0.99–1.44) [45]. Living alone is not harmless, though: its 2015 figure (OR 1.32) is in line with loneliness and isolation. Fuller treatment in Purpose.
Purpose in life — moderate (association only)
People with a stronger sense of purpose are less likely to develop dementia: pooled across six samples (about 53,000 people), higher purpose went with about 23% lower dementia risk (HR 0.77) [46]. In one 28-year Wisconsin cohort, purpose measured at 63 and at 70 predicted a dementia-range score on a short telephone test of memory and thinking at about 80 (about 15% lower odds per standard deviation of purpose, OR 0.85), while purpose at 52 did not clearly do so. Those estimates overlap and were never formally compared, so this is not proof of a special late-midlife window, and part of the link may run backwards, with early, still-silent decline lowering purpose [47]. No trial has yet shown that raising purpose lowers dementia risk. Full treatment in Purpose.
Time in nature, journaling, gratitude — modest but real
Short, replicable: 20-minute "nature dose" lowers salivary cortisol; expressive writing 15 min/day for several days has modest effects on rumination; gratitude practices act as emotional amplifiers when added to existing mindfulness work. Useful adjuncts; not load-bearing.
Emerging frontier: dyadic and "social" meditation
Solitary practice fails for many people simply because adherence collapses. Two recent directions are worth knowing about:
- Partner-based "contemplative dyads" — two people take turns speaking and listening mindfully, about 12 minutes a day through an app, with weekly online coaching. In a 10-week randomized trial in 253 healthy adults, the partner practice reduced loneliness more than solo mindfulness training did [48]. People kept up the partner practice at least as well as solo meditation [49]. In 140 participants from the same project, tested after training, the dyad group (but not the mindfulness group) had lower levels of the stress hormone cortisol than an untrained control group during a lab public-speaking test, though they did not feel less stressed [50].
- Shared-biofeedback virtual-reality (VR) meditation — pairs see each other's breathing in VR. Studies so far measured short-term effects, such as feeling more "present" with a partner right after a session; the review cites no data on adherence or stress hormones [51].
These are early-evidence modalities, not yet ready for "do this" prescriptive advice. But they answer a real problem: in real-world usage data on 93 popular mental-health apps, a median of only about 3% of users were still using an app 30 days after installing it (about 5% for meditation apps) [52], and a practice that pairs mindfulness with social connection adds one of the strongest non-drug mortality levers, close relationships, to a useful stress-reduction tool.
Cellular signal: telomerase and inflammaging
The cellular story splits sharply by evidence quality. The observational link is real and foundational: in premenopausal women, higher perceived stress and longer caregiving chronicity tracked with shorter telomeres, lower telomerase, and more oxidative stress — those with the highest stress had telomeres shorter by roughly a decade of additional aging.[53] That is cross-sectional and confounded, and cannot show stress caused the shortening.
The intervention evidence is weak. Telomerase / telomere lengthening — weak / preliminary. The most rigorous, well-powered, long trial — 137 older adults randomised to 18 months of meditation training, non-native language training or a passive control — was null for telomere length.[54] A 2023 pooling of 25 studies (2,099 people) found a small benefit for telomere length, but it came mainly from studies comparing experienced meditators with non-meditators, not from trials of meditation programs; most studies were small, the authors saw signs of publication bias and concluded that telomere length "may not be a useful outcome" for these interventions [55]. Earlier positive pilots are small and often non-blinded. Treat any telomerase signal as preliminary and largely confounded.
- Inflammatory markers. Across 45 randomized trials that compared meditation with another active activity, meditation lowered C-reactive protein (CRP), tumor necrosis factor-α (TNF-α), cortisol, blood pressure, heart rate and triglycerides [56]. A pooling of 48 randomized trials (4,683 people) found small drops in CRP and interleukin-6 (IL-6) at the end of mindfulness programs, with wide uncertainty for IL-6 [57]. Whether trained people mount a smaller inflammatory response to an acute stressor has not been shown in pooled trials. This inflammatory signal is better supported than the telomere one.
On epigenetic clocks specifically: no completed RCT shows a stress-reduction intervention slows a clock. The strongest causal evidence in this whole domain is for caloric restriction, which slowed only DunedinPACE (~2–3%) over two years with null effects on GrimAge and PhenoAge.[58] These are mechanistic surrogates, not longevity outcomes.
What's overhyped or doesn't work
- Stress-hack supplements without behavioral change — adaptogens (ashwagandha, rhodiola), various nootropics. Ashwagandha is one of the more-studied options: in 14 small trials in people with anxiety and other mental disorders (about 700 people, median 600 mg/day for 8 weeks) it eased anxiety [59], and a separate pooling of 488 people found it lowered cortisol by about 32 nmol/L, though its effect on how stressed people felt was unclear (the larger review above did find less stress in people with mental disorders) [60]. The trials are small and short, and a meta-analysis of 12 of them (about 1,000 people) rated the certainty of the evidence low for both stress and anxiety [61], so treat this as preliminary. It also carries a rare but real risk of liver injury: among 25 published cases most recovered, but one person needed a liver transplant and three people with existing cirrhosis died [62]. Denmark banned it in 2023 [63], and EU regulators have recommended a procedure that could lead to an EU-wide ban [64]. Skip it if you are pregnant or breastfeeding or have liver or thyroid disease. None of these replace the autonomic retraining that practice and sleep deliver. See Sleep & anxiety supplements for the full evidence on ashwagandha.
- One-off retreats as the whole plan. A retreat is not wasted: pooled studies of meditation retreats in the general population found moderate benefits overall, with larger effects on anxiety, depression and stress, still present at follow-up, though the studies varied a lot [65]. Keeping up some daily practice afterwards is the sensible way to hold on to them.
- Short, app-only "1-minute breathing" sessions as the only intervention. Better than nothing, but well below the dose of the full MBSR-style protocols.
- Cold plunges as a "stress reset." Cold immersion produces an acute sympathetic surge with a parasympathetic rebound, but whether it improves mood or stress is unproven: a 2026 review of 13 studies in 4,105 adults found results leaning favorable yet concluded the evidence does not establish an improvement in mood or perceived stress, with most studies at high risk of bias [66]. It is not a substitute for the slower, sustained autonomic retraining of breath and meditation. See Cold exposure.
- Heart-rate-variability biofeedback gadgets (HeartMath, etc.) — the underlying physiology (paced breathing) works; the proprietary device adds little over a metronome and a free 6-breath-per-minute audio track. A meta-analysis of 58 trials found HRV-biofeedback effects that "do not differ from that of other effective treatments," with larger effects vs inactive than active controls — consistent with the breathing, not the device, doing the work.[67]
- "Adrenal fatigue." A popular driver of cortisol-testing and supplement purchases — but a systematic review concluded there is "no substantiation that 'adrenal fatigue' is an actual medical condition." It is recognized by no endocrinology society, and cortisol-profile testing has no validated utility for fatigue.[68]
- Reframing chronic life-situation stress as "mindset" — when the stressor is a punishing job, a hostile relationship, or financial precarity, breathing exercises are a partial buffer at best. Structural change (job, relationship, finances) often dwarfs any contemplative practice in effect size.
A practical stress protocol
For a healthy midlife adult who wants to push back against chronic stress:
- About 20 min/day of slow breathing (about 6 breaths/min), the home dose in a 4-week trial in which stress fell [69], or 5 minutes of "cyclic sighing" (breathe in, add a second short breath in to fill the lungs, then breathe out slowly and fully). In one small, exploratory one-month trial, 5 minutes a day of cyclic sighing raised positive mood more than the same time spent on mindfulness meditation, though both reduced anxiety similarly [70]. Cheap, fast-acting, no equipment beyond a timer.
- About 30 minutes of formal mindfulness practice most days, for at least 8 weeks before judging the effect. In MBSR and MBCT courses people typically manage about 30 minutes a day, six days a week (roughly two-thirds of what is assigned), and how much people say they practice is only modestly linked to how much they benefit [71].
- Fix sleep before adding more practice. Regular schedule, screen for sleep apnea, CBT-I if insomnia is the bottleneck.
- Aerobic + resistance exercise — even modest weekly amounts help mood and distress, and harder sessions help more than extra minutes [72]. Aim for about 150 min/week for heart and metabolic health, but don't wait until you can reach it.
- Defend two or three deep relationships actively (weekly contact, not annual). Social connection is the most underrated lever.
- Audit secondary behaviors. If stress is driving alcohol, late eating, screen time, or skipped training, the behavior is doing much of the cellular damage; addressing it is part of stress management, not separate from it.
- Reserve "biohacks" (cold plunges, breathwork variants, adaptogens, HRV gadgets) for after the basics are in place. Treat them as small, mostly unproven extras on top of a protocol that already includes the items above, not as substitutes for any of them.
Further reading
- Goering M et al. The role of health behaviors in links between stress and epigenetic aging. Brain Behav Immun Health 2026.[73]
- Hansen JL et al. Lifetime chronic stress exposures, stress hormones, and biological aging: Results from the Midlife in the United States (MIDUS) study. Brain Behav Immun 2025.[74]
- Molecular pathways linking chronic psychological stress to accelerated aging — review.[75]
- Holt-Lunstad J et al. Social relationships and mortality risk: a meta-analytic review. PLOS Med 2010.[76]
- Crego A et al. Relationships between Mindfulness, Purpose in Life, Happiness, Anxiety, and Depression: Testing a Mediation Model in a Sample of Women. Int J Environ Res Public Health 2021.[77]
- Goyal M et al. Meditation programs for psychological stress and well-being: a systematic review and meta-analysis. JAMA Intern Med 2014.[78]
- Joshi et al. Evidence-based pathways to healthy aging: a systematic review and meta-analysis of lifestyle interventions. Invest Educ Enferm 2025.[79]
- Epel ES et al. Accelerated telomere shortening in response to life stress. PNAS 2004.[80]
- Kaliman P et al. Effect of an 18-Month Meditation Training on Telomeres in Older Adults: A Secondary Analysis of the Age-Well Randomized Controlled Trial. Biol Psychiatry Glob Open Sci 2025.[81]
- Parker HW et al. Allostatic load and mortality: a systematic review and meta-analysis. Am J Prev Med 2022.[82]
- Rosengren A et al. Association of psychosocial risk factors with risk of acute myocardial infarction (INTERHEART). Lancet 2004.[83]
- Kivimäki M et al. Job strain as a risk factor for coronary heart disease (IPD-Work). Lancet 2012.[84]
- Adam EK et al. Diurnal cortisol slopes and mental and physical health outcomes. Psychoneuroendocrinology 2017.[85]
- Holt-Lunstad J et al. Loneliness and social isolation as risk factors for mortality. Perspect Psychol Sci 2015.[86]
- Yoga and telomeres — narrative review.[87]
- Singer T. A neuroscience perspective on the plasticity of the social and relational brain. Ann N Y Acad Sci 2025.[88]