Emotional Resilience

Last reviewed 5 October 2026

What it is

Emotional resilience is the capacity to adapt to stress, adversity, or setback and return to stable psychological functioning. Researchers usually break it into related parts: cognitive reappraisal, which is the ability to reinterpret a stressful situation, distress tolerance, which is staying functional while upset, and recovery time, which is how quickly mood and functioning return to baseline.

Even the basic definition is contested. Bonanno argued that resilience is a distinct trajectory from recovery (minor, transient disruption rather than a dip into serious symptoms and recovery) and, after loss or potential trauma, is more common than often believed, with several pathways to it [1]. Kalisch and colleagues treat it as a process of adapting to stressors, not a trait, which cannot be captured by any single cross-sectional measure or biomarker [2]. Panels in 2014 and 2020 agree there is no single definition, and the 2020 panel questions the cross-cultural validity of the Connor-Davidson scale and notes that self-reported and functional resilience do not line up [3] [57].

Most studies below measure a feeling, a score, or a brain response at one moment, and very few watch anyone meet an actual setback. That gap is the central limit of the field.

How it is measured

Kalisch and colleagues warn that no resilience questionnaire has been validated as a predictor of outcome after hardship in a prospective study. Their own review of 13 prospective studies, which was not systematic and did not appraise study quality, found weak predictors, four studies with null results, and 5 to 13 percent of the variance explained in samples of up to 94 people [2]. A single questionnaire score should not be read as "resilience."

The Connor-Davidson Resilience Scale showed sensitivity to treatment effects only in two PTSD trials. We read its original paper's abstract only, so we do not report its sample means or reliability figures [4]. A methodological review scored 15 scales out of 18 (its search ended in September 2009, and its first author developed one of the included scales), with the Connor-Davidson and Brief Resilience scales at only a moderate 7. No scale reported floor or ceiling effects (scores bunching at the bottom or top of the range, which hides change), and only the Connor-Davidson scale had been tested for sensitivity to change, in PTSD trials. Its retest reliability of 0.87 came from only 24 people [5]. The Brief Resilience Scale was built to capture bouncing back [6], but its German validation was cross-sectional and did not test sensitivity to change [7].

The Emotion Regulation Questionnaire records habits, not behavior when something goes wrong [8], and its validation used an online panel, a single self-report criterion, and an author who developed it [9]. The Difficulties in Emotion Regulation Scale performed poorly on its awareness subscale in treatment-seeking adults, and its validity against other regulation measures was not tested [10]. The Challenging Experience Questionnaire captures difficult experiences, not resilience, and its authors note recruitment biased toward favorable views of psychedelics [62].

A review of 43 randomized trials of resilience programs from 1979 to 2014 found the field's concepts, methods, and designs of limited use for judging whether such programs work, and proposed assessing resilience as change in mental health in relation to the stress people actually face. We read its abstract only [11]. A Cochrane review found that only five of 44 trials had measured exposure to stressors [44].

What is known

In a crossover trial, every volunteer receives each condition on separate days.

In the hours after ingestion: how the brain handles emotional pictures

The most consistent controlled finding in healthy adults is a lower response to negative images in the acute state in the amygdala, a deep brain structure involved in detecting threat. In a double-blind crossover of 25 healthy volunteers, psilocybin lowered amygdala response to negative and neutral images compared with placebo. We read its abstract only [12]. A preregistered LSD crossover (20 analyzed of 24 recruited) reduced responses to fearful faces in the left amygdala and right medial prefrontal cortex, but only in tiny clusters under a small-volume correction, with the correlation computed inside the selected cluster [13]. A study of 20 volunteers with psilocybin-session data had no placebo condition, and scans were taken about 169 minutes into the session. Amygdala response fell for angry faces but not for fearful or neutral ones [14].

The mechanism looks different from what a skilled reappraiser does. In a meta-analysis of 48 neuroimaging studies of healthy people, reappraisal engaged prefrontal and parietal regions and lowered amygdala activity, but only one of 838 coordinates fell in the ventromedial prefrontal cortex, and every study used picture paradigms [18]. In the psilocybin reanalysis of about 23 volunteers, the drug reduced threat-related feedback from the amygdala to the visual cortex, but the expected strengthening of prefrontal control did not appear, and the connection changes did not correlate with mood or behavior [15]. So the drug-state amygdala drop is not the brain regulating emotion more effectively: participants viewed images passively. In two pooled LSD crossovers (38 analyzed), recognition of fearful faces dropped, and recognition of sad faces dropped overall, but follow-up tests within the dose groups were not significant. Cognitive empathy dropped as well [16].

The direction depends on who is studied. In an open-label study without placebo, 19 scanned people with treatment-resistant depression showed a rise in right amygdala response to fearful faces the day after the second psilocybin session, the opposite of the healthy acute data. This was a clinical sample, so the healthy-volunteer drop cannot be assumed to carry over [17].

Days to months: controlled and open-label follow-up

In one randomized, double-blind crossover (43 analyzed healthy adults), perceived stress one week after a single full LSD session was lower than after placebo, and the Alternatives subscale of a cognitive flexibility questionnaire was higher, while the Control subscale was not (P = 0.42). These are differences from placebo, not evidence of functioning above the normal range, and baseline stress was already near the minimum and flexibility near the maximum. Every participant correctly identified the LSD session, and 79 percent of those given a placebo presented as a low dose guessed that was what they had received. The authors say some self-report changes may be expectation effects, and the questionnaire results were not corrected for multiple comparisons. The day after LSD, participants also reported more fatigue and low mood, and one possibly treatment-related serious adverse event occurred and resolved [19]. It is the strongest controlled post-acute stress result we found in healthy adults, yet a single secondary outcome in a trial designed for other questions.

In an open-label study of 12 healthy volunteers, negative affect and amygdala response in the emotion identification task (not the discrimination task) were lower at one week, with an effect size around 0.8, and both had returned to baseline by one month. Trait anxiety stayed lower, and neuroticism did not change. A prefrontal increase was reported only at an uncorrected threshold, the personality tests were not corrected for multiple comparisons, there was no control group, and the authors name expectancy and habituation as possibilities [20].

Controlled studies with longer follow-up are less encouraging on standardized measures. In a placebo-controlled trial of 89 healthy adults, neuroticism and symptom scales did not differ from placebo up to day 85, nor did a cognitive flexibility task at day 8, while a retrospective, unvalidated measure of life changes did (the authors call the analysis exploratory and tentative). The trial was run by a company developing psilocybin treatments, and the authors declare industry ties [21]. In a randomized crossover of a purified DMT and harmine formulation in 31 healthy men, acceptance, well-being, and neuroticism did not change at one and four months (22 completed the trait follow-up, there was no DMT-alone arm, and some authors are company founders) [22]. Sixteen healthy volunteers rated lasting positive changes at 12 months on a questionnaire built to attribute change to the session, while a trait anxiety scale did not move (14 analyzed, no questionnaire baseline, conscientiousness rose uncorrected, expectancy not excluded) [23].

The same split shows up in the classic 14-month psilocybin follow-up of 36 healthy, spiritually active adults. At 14 months, 58 percent rated the experience among the five most meaningful of their lives, and most said it increased their well-being. Yet of eight standardized questionnaires, only the one measuring mystical experience changed, and personality scores were already near the ends of their ranges at baseline. Eleven of the 36 had significant fear during a session, and none had it persist [24].

Uncontrolled follow-up of people who chose to use

A retreat cohort of 52 people showed lower state and trait anxiety the morning after the ceremony (effect sizes 0.77 and 0.52) and larger drops at one week (0.87 and 0.77). More than half had dropped out by one week, and dropouts had started with higher state anxiety (37.6 against 30.7). The trait anxiety difference was not significant, and the analysis treated sessions as independent and did not correct for multiple tests [25]. A cohort of 261 ayahuasca retreat attendees, not all healthy (13 percent reported depression), showed rising psychological flexibility (0.60) and falling negative affect (0.46) at three months, with larger gains in people reporting more reappraisal-like experiences and higher expectancy. There was no control group, 18 percent were lost to follow-up, and the reappraisal scale was new, unvalidated, and measured only at follow-up [26].

The controlled counterpoint is a placebo-controlled study of 30 experienced participants at ayahuasca retreats, students of an ayahuasca school, in which stress, depression, and anxiety scores fell the morning after a ceremony whether people had received ayahuasca or placebo, measured only once. The ayahuasca group did show greater emotional empathy toward negative stimuli on an implicit measure [56]. The setting may carry part of what uncontrolled cohorts attribute to the substance.

A cross-sectional comparison of 19 experienced ayahuasca users and 19 non-users found higher ego-resiliency scores and no difference on mood or anxiety scales, but ego-resiliency is a personality trait, distinct from resilience as a process that requires exposure to threat [54]. Only this one of seven psychometric comparisons differed, and in our reading it would not survive correction. A brain-scan classifier separated users from controls with 65 percent accuracy, not significant (P = 0.10), and the design cannot say whether the practice produced the difference or whether more resilient people join and stay [27]. A non-randomized comparison of 10 people after four ayahuasca sessions with 10 who completed an eight-week mindfulness course found improvement on every questionnaire facet in the mindfulness group and on one in the ayahuasca group (Non-Judging group-by-time interaction P = 0.08) [28]. A 45-person study with no control group found lower emotion dysregulation within 24 hours of a session (abstract only), but 12 of the 45 had borderline-like traits, so this is not a healthy-only sample [29].

Repeated very small amounts

A preregistered, double-blind crossover (analyzed samples of 40 to 52) found no effect on anxiety, depression, stress, or an emotional processing task compared with placebo. The go/no-go task was underpowered (power about 0.12), and participants were workshop attendees with prior experience [30]. In a placebo-controlled study of 56 healthy adults, scores fell in every group whatever the drug and nothing persisted at follow-up. All had prior experience, the analyses were not corrected for multiple tests, and fewer than half identified the higher active condition correctly [31]. In a self-blinding citizen-science trial with 191 completers, the placebo group improved on mindfulness, paranoia, and neuroticism just as the active group did, and well-being and life satisfaction rose significantly only within the active group. The active group did not beat placebo, and the differences vanished after accounting for what people guessed [32]. A six-week trial in 80 healthy men found same-day improvements that held up after adjusting for expectation, but no credible change between baseline and week six. The same-day findings came from daily questionnaires analyzed without a prior hypothesis, and blinding was only partial. Four people left the active arm because of treatment-related anxiety, which followed a feeling of overarousal amid stressful life events and resolved within two weeks with light support [33].

A 2026 systematic review (24 studies, 3,681 participants) pooled only six, and its randomized pool held just two parallel trials (three comparisons, 117 participants). The pooled effect was 0.02 for stress and -0.20 for anxiety, with a very wide interval and high heterogeneity (I² 82.3 percent), and the authors call the randomized evidence inconclusive. We read the abstract only [34]. In one online cohort, the Brief Resilience Scale rose (d = 0.47) and people with higher expectations improved more, but the analysis was uncontrolled and exploratory, with one-sided expectation tests, and the cohort lost 68 percent of its sample by week four and had 46 percent with a psychiatric diagnosis, so it is not a healthy-adult sample [35].

Blinding, expectation, and what they do to every number above

In large studies of full psychedelic experiences, participants guess correctly that they received the drug at rates approaching 95 percent, according to a 2026 perspective on well-being endpoints [36]. The expectancy analysis it points to comes from one depression trial of psilocybin against escitalopram, and we read the primary report [59]. In that analysis (55 participants), expectation interacted with time on several depression scales for escitalopram and on none for psilocybin. The authors' equivalence tests were not significant, and they state that they could neither rule out nor confirm an expectancy effect in the psilocybin arm. The analysis was exploratory and not preregistered, and expectation was measured with two unvalidated items. According to the authors, the narrowing of the gap between arms after adjusting for expectation came from a nocebo-like effect among escitalopram participants with low expectations, not from inflation in the psilocybin arm. This is one trial in a clinical sample, and it does not show that expectation plays no part. A review of psychedelic trials found that most do not report expectancy or blinding success, and that effect sizes are probably overestimated (abstract only) [37]. No one knows how much of any self-reported change in stress or emotion handling comes from the substance.

Harms that bear on emotional stability

A pooled analysis of eight double-blind studies in 110 healthy volunteers found acute anxiety and dysphoria in a relatively small proportion of participants, managed with support, with no lasting impairment at follow-up. We read the abstract only [38].

A self-selected online survey recruited 1,993 people because of a difficult experience with psilocybin. Of the 1,339 whose session was at least a year earlier, 24 percent reported symptoms lasting a week or more that they attributed to it, 10 percent said symptoms lasted a year or longer, and 7.6 percent sought professional treatment. During the sessions, 10.7 percent of all respondents put themselves or others at risk of physical harm, and 2.7 percent received medical help at a hospital or emergency department. Three respondents described the onset of enduring, impairing psychotic symptoms. Five described increased suicidal thoughts or behavior, three of them involving attempts, while six said that existing suicidal thoughts had fully remitted. The authors stress that this is not a prevalence estimate [39].

A prospective online survey of people who planned to use psilocybin outside research had 2,833 at baseline, 1,182 at two to four weeks, and 657 at two to three months. About 11 percent and 7 percent of those who answered at the two later points reported at least one lasting negative effect, most often mood swings and depressive thoughts. It is not a healthy-adult sample, because 41.5 percent were above a depression cut-off at baseline, about 77 percent had dropped out by the end, and 26 and 50 percent had taken another dose by the two follow-ups [60]. In a retrospective survey of 2,510 people, 330 (13 percent) reported at least one negative outcome they attributed to psychedelics, most often cigarette smoking and problematic cannabis use. The eight-item instrument was created for that study, so the 13 percent is not a rate of emotional breakdown [61].

A small interview study of people with lasting negative responses found anxiety in 87 percent of completers and cannot say how common this is. Of 84 who agreed to take part, 32 completed (62 percent attrition), and recruitment was self-selected, including forums for hallucinogen persisting perception disorder [40]. In a general-population cohort of 12,345, the 505 who used a psychedelic started with higher depressive scores and had slightly higher depressive symptoms afterward than non-users (β = 0.12, after adjustment). Use in settings the authors classed as higher risk raised the effect, and use outside such settings showed no worsening (β = 0.03). The study was exploratory, with 44 percent attrition [41]. How harm is defined changes the rate a great deal, so these figures cannot be compared with each other.

What is not known

No controlled trial in healthy adults has made resilience or emotion regulation its main question. We found no registered healthy-adult randomized trial of a classic psychedelic with a primary endpoint on a resilience scale, reappraisal, or stress reactivity (three registry queries on 30 September 2026, run through a summarizing tool and not rechecked at the source). A PubMed search by entry date from 1 September to 4 October 2026 found no new human study in healthy adults with a resilience measure, and we found none that gave a classic psychedelic and then ran a standard laboratory stress test. None of these searches was exhaustive, and PsycINFO was not available to us.

Nobody has watched a real setback. The prospective design that Kalisch and colleagues call for [2] would follow healthy participants through real stressors. Their later proposal would monitor everyday stress exposure and mental health problems continuously and compute each person's reactivity relative to the sample, which works only if exposure and problems are adequately correlated and has not been tested with psychedelics [58]. We found no psychedelic study with either design.

Durability. The best amygdala data show an effect that fades by one month [20], and controlled follow-up to day 85 found no change on standard personality and symptom scales [21].

Direction across populations. The acute amygdala drop in healthy volunteers [12] [13] became a rise after treatment in depression [17], and we do not know why.

Expectation. With near-complete unblinding, we cannot separate the drug from the belief, and nobody has analyzed expectancy for resilience outcomes in a healthy-adult trial [59].

Floors and ceilings. Healthy adults start with low stress and high resilience scores [19], and no resilience scale in the methodological review reported ceiling data [5], so whether current scales can detect gains for people already doing well is unknown.

Which part is resilience. A drop in amygdala response, a lower perceived stress score, and a faster recovery after a real setback are different things. Evidence on the first is the best, on the second thin, and on the third nonexistent [2].

Harm rates. The harm data come from self-selected or harm-selected samples [39] [40] [60] [61], so the true rate of lasting distress in healthy adults is not known.

What Metanoia seeks to learn

Metanoia is a research and education charity, and it has no data of its own on this dimension yet. These are the questions it wants answered, through published research and through observational records kept under full legal authorization, without supplying any substance:

• Does a change on a validated resilience or recovery measure last at six and twelve months, and does someone who knows the participant confirm it?
• When participants later meet a real stressor, does recovery differ from comparable people without the experience? This needs a measurement before the stressor, a record of how much stress people faced, a measurement after, and a time window [2].
• Can self-report be paired with behavior, such as days on which plans were cancelled because of mood and time taken to return to routine after something upsetting?
• How do expectation, setting, and preparation, recorded under the international reporting guidelines for psychedelic settings, relate to outcomes in either direction?
• How often does emotional stability get worse, and under what circumstances? Harm needs a definition fixed in advance, because the definition changes the rate [60] [61].

The Brief Resilience Scale is the shortest of these scales and the one most directly aimed at recovery, which makes it a candidate anchor for studies of healthy adults. Whether it can detect gains in people who already score high is open, and by Kalisch's standard no single score should be presented as resilience. Null and negative results will be reported alongside positive ones.

How it compares with other paths

The comparison paths were mostly tested in different people from the psychedelic studies, with questionnaires that have their own limits [5] [11]. The numbers cannot be ranked head to head.

Resilience training programs. Meta-analyses report small to moderate gains in resilience: 0.37 from 13 trials with 782 participants for stress-directed programs [42], 0.44 from 11 randomized trials, only 4 individually significant, by authors linked to a resilience-training company [43], and 0.45 in a Cochrane review of healthcare workers [44]. The trials were mostly small, at moderate to high risk of bias, and no longer than three months, with smaller effects against attention-control comparisons [42]. Confidence was low in the first review and very low in the Cochrane review, which found no difference after three to six months and no change in anxiety or well-being (we read its main report, not the appendices) [44]. No trial in the second review tested resilience after a real stressor, and it had no protocol in advance [43]. A meta-analysis of 101 digital programs in 20,010 non-clinical adults found resilience up by 0.22 and distress down by 0.24, but the distress effect was not significant against high-intensity active comparisons, certainty was very low, and the prediction interval for resilience included zero (-0.92 to 1.37) [45].

Cognitive and reappraisal-based methods. A meta-analysis of 306 comparisons found an effect of 0.36 for changing how one thinks about a situation (redirecting attention had none), and one of occupational stress-management programs found 0.53, largest for cognitive-behavioral ones. Both were abstracts only [46] [47].

Mindfulness, limited to emotion handling. A meta-analysis of 29 mindfulness-based stress reduction studies in healthy individuals found a between-group effect of 0.53 across 18 (abstract only) [48]. In 110 laboratory studies of mindfulness-based emotion regulation the effect was 0.28, and 0.22 in healthy participants, with high heterogeneity, signs of publication bias, no significant subgroup difference (P = 0.296), sensitivity analyses from 0.15 to 0.30, and the largest effect for pain tolerance, not emotion handling [49].

Exercise. A 2026 meta-analysis of 10 studies reported a moderate rise in resilience scores after exercise programs, around 0.5, but it contradicts itself: one scale's figure is 0.27 in one section and 0.34 in another, and a longer-versus-shorter comparison is significant after correction in the results but not in the discussion. It was registered on 25 January 2026, after its search of 7 January, and included single-group studies, combined treatments, mixed populations, and no follow-up, so we avoid leaning on an exact number [50]. In a laboratory stress test, 81 regular exercisers and 30 non-exercisers did not differ in heart rate, blood pressure, or cortisol reactivity, and differed only on one mood subscale, elation (composite P = 0.08), in groups unbalanced for sex and race [51]. The questionnaires and the physiology do not agree.

Sleep. A meta-analysis of 65 trials (53 percent in people with a coexisting condition, 43 percent not) found that better sleep lowered stress (-0.42, 95% CI -0.79 to -0.05, six comparisons, and -0.35 after adjusting for publication bias), with 45 percent of trials at high risk of bias [52]. A synthesis of 154 sleep-deprivation studies in 5,717 mostly young, healthy participants found that sleep restriction lowered positive emotion and raised anxiety by about 0.6 (arousal effects rested on two to four studies per analysis), but it excluded sleep extension and restorative sleep, so it cannot tell us whether improving sleep changes emotional reactivity [53].

Where psychedelics fit. The firmest finding is a laboratory one: a drop in amygdala response to negative pictures in the acute state, replicated in small samples but not for every image type and not seen in clinical samples. The one controlled post-acute stress result is small and cannot rule out expectation, and controlled studies at longer follow-up mostly found no change on standard trait scales. On present evidence, lasting improvement in emotional resilience in healthy adults is unproven for psychedelics, and the comparison paths are better documented, if often modest. Whether the psychedelic state adds something distinct is not settled.

History and tradition

Graded 2 and 4. The first paragraph describes how a scientific concept developed. The second describes a tradition. Neither is evidence of benefit or safety.

The word resilience entered developmental psychology in the 1970s through research on children growing up with a parent who had schizophrenia. A 2000 critical review describes the field's move from protective traits to processes, warns that early terms like "invulnerable" misled researchers, and separates resilience as a process, which requires exposure to significant threat, from ego-resiliency, a personality trait [54].

A scholarly review of archaeological, ethno-historical, and ethnographic evidence describes the use of psychedelic mushrooms in healing rituals and religious ceremonies in pre-Columbian Mesoamerica. It dates mushroom stones found in ritual contexts to about 3000 BC, cites 16th-century chroniclers, and reports that local healers still use them. We read its abstract only [55]. This page makes no health claim based on it, and practices differed by place and period.

The book on all twelve dimensions   Back to the twelve dimensions