Implementation of Mindfulness-Based Emotion Regulation Strategies: A Systematic Review and Meta-analysis

Abstract

Among the strategies people can use to regulate their emotions, there is increasing interest in mindfulness. Although prior reviews support that mindful acceptance (equanimity) is an effective strategy, other components of mindfulness (monitoring/ mindful awareness) have received less attention. Further, a recent theoretical framework—the Dual-mode Model of Mindful Emotion Regulation—proposed that mindfulness has two “modes” of action in emotion regulation: (1) implementation to regulate emotions as time-limited strategies and (2) acting as a moderator that facilitates effective emotion regulation. The present systematic review and meta-analysis aimed to evaluate the implementation of mindfulness-based strategies and facilitation effects of mindfulness on emotion reactivity and strategy implementation. A total of 2037 records were screened, from which 110 studies with 767 effects (N = 8,105) were analyzed. There was a significant effect of mindfulness-based strategies (g = .28, 95% CI [.18, .38], I² = 83.3%). Components of monitoring (g = .17, 95% CI [.02, .32]) and equanimity (g = .3, 95% CI [.19, .41]) were also effective, although monitoring alone was significantly less effective than other strategies. The effect size was greater for within-subjects designs, text-based emotional stimuli, personally relevant stimuli, and behavioral outcomes (e.g., pain tolerance). Despite small-study effects and publication bias, the estimate was robust to sensitivity analyses (between .15 and .3). Meta-regression of facilitation supported that greater study-level equanimity was associated with reduced reactivity. As hypothesized, mindfulness-based strategies can effectively regulate emotions. Methodological considerations and directions for future study are discussed.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42761-024-00281-x.

Emotion regulation is the process by which someone changes the frequency, intensity, duration, or type of experienced emotion (Fernandez et al., 2016). While several models exist to organize emotion regulation into discrete strategies, one of the most popular is the Extended Process Model (Gross, 1998, 2015; McRae & Gross, 2020; Webb et al., 2012). This model organizes emotion regulation into five families based on how they affect emotional experience (situation selection, attentional deployment, cognitive change, and response modulation), each of which is composed of strategies that share cognitive processes and tactics—the specific ways emotions are regulated (Table 1). For example, the attentional deployment family is characterized by shifting the attentional focus to regulate emotions, such as with distraction. Distraction involves directing attention away from a stimulus in order to regulate emotions with tactics such as focusing on something enjoyable or doing a neutral cognitive task. While there are many ways one might distract themselves (tactics), they share the common process of distraction.

Table 1.

Strategy families, strategies, and tactics with meta-analysis codes

Strategy	Tactic	Definition
Attentional deployment—changes emotions through shifting attention relative to stimulus
Distraction	Active positive distraction (D1)	Instructed to think about something positive unrelated to the outcome or stimulus
	Passive positive distraction (D2)	Provided with positive materials unrelated to the outcome or stimulus
	Active neutral distraction (D3)	Instructed to think about something neutral unrelated to the outcome or stimulus
	Passive neutral distraction (D4)	Provided with neutral materials unrelated to the outcome or stimulus
	Distract—mixed (DM)	Instructions blend D1–4 or are not sufficiently specific to differentiate
Concentration	Concentrate on feelings (C1)	Instructed to attend to, focus on, or relive the outcome or stimulus
	Concentrate on meaning (C2)	Instructions to focus on causes, meanings, or consequences related to outcome or stimulus. Includes rumination
	Concentrate—mixed (CM)	Instructions blend C1 and C2 or are not sufficiently specific to differentiate
Monitoring	Physical awareness (M1)	Instructed to focus and maintain awareness specifically on breath and physical sensations without inclusion of emotions; focus should be gently returned to current physical sensations if it wanders
	Emotional awareness (M2)	Instructions to be aware of reactions and emotions without “holding onto them” or with allowing them to pass naturally; focus should be gently returned to current reactions and emotions if it wanders
	Focused awareness (M3)	Instructions to maintain focus on awareness itself without a specific focus on the body, emotions, or thoughts. May include focusing on a specific mantra or topic
	Monitoring—mixed (MM)	Instructions blend M1-3 or are not sufficiently specific to differentiate. Includes general emphasis on awareness itself as the focus of attention rather than specific content of attention (which also differentiates monitoring from distraction or concentration—attention towards process rather than content)
Cognitive change—changes emotions through changing appraisals
Equanimity	Acceptance (E1)	Instructions to change how one thinks about their reactions to a stimulus, including adopting an accepting or non-judgemental stance
	De-fusion (E2)	Instructions to attend to reactions and emotions as separate from oneself. Differentiated from perspective taking through an emphasis on de-identification with the reaction rather than distancing from the stimulus
	Self-compassion (E3)	Instructions to adopt an empathetic or compassionate focus towards oneself without instructions to reappraise the stimulus
	Equanimity—mixed (EM)	Instructions blend A1–3 or are not sufficiently specific to differentiate
Reappraisal	Reappraise stimulus meaning (R1)	Instructions to reinterpret the stimulus itself, such as through reinterpretation of the context, cause, or outcome of a situation
	Reappraise via perspective taking (R2)	Instructions to reinterpret the stimulus itself through changing individual perspective, such as through imagining themselves in the situation or to be “objective.”
	Reappraisal—mixed (RM)	Instructions blend R1–3 or are not sufficiently specific to differentiate
Response modulation—changes emotions through altering behavioral components of emotional response
Suppression	Expressive suppression (S1)	Instructions to hide their emotions such that an outside observer could not know what they were feeling
	Experiential suppression (S2)	Instructions to prevent or inhibit internal affective reactions to stimulus
	Thought suppression (S3)	Instructions to prevent or inhibit thoughts, urges, cravings, or other cognitions related to stimulus
	Suppression—mixed (SM)	Instructions blend S1–3 or are not sufficiently specific to differentiate
Hybrid—strategies that blend aspects from more than one family
Mindfulness (H1)		Combination of any among mindfulness-based strategies (in italics). Includes attending to experience (monitoring) using an open, accepting, or non-judgemental stance (equanimity). Also used when details are insufficient to distinguish between specific strategies
Other		Blend of two or more tactics from distinct strategies (e.g., S3, D4)
Reactivity
	None (NR1)	No instructions provided during baseline consisting of valenced stimuli (e.g., positive or negative images)
	Non-regulation (NR2)	Instructions specify not to regulate in response to stimulus but otherwise are not prescriptive
	Maintain (NR3)	Instructions to maintain reaction to stimulus
	React naturally (NR4)	Instructions to respond naturally to stimulus without efforts to control or apply a mindfulness-based strategy. Includes mind-wandering or unfocused instructions
	Control—mixed (NRM)	Instructions blend NR1–4 or are not sufficiently specific to differentiate

Subheadings of attentional deployment, cognitive change, and response modulation are families of emotion regulation strategies from the Extended Process Model. Table modified from Webb et al. (2012) with additions and revisions made regarding mindfulness-based strategies based on Raugh and Strauss (2023)

Although no family of strategies is considered “better” than any other, some strategies are considered more effective (i.e., able to influence emotions to meet hedonic goals) or adaptive (i.e., greater use associated with greater long-term well-being or reduced psychopathology) (Aldao et al., 2010; Naragon-Gainey et al., 2017; Webb et al., 2012). Among various strategies, theoretical (Chambers et al., 2009; Hölzel et al., 2011; Lindsay & Creswell, 2017; Raugh & Strauss, 2023), experimental (Chiesa et al., 2011; Kohl et al., 2012; Naragon-Gainey et al., 2017; Van Dam et al., 2018; Weder, 2022), and clinical research (Goldberg et al., 2018; Gu et al., 2015; Hayes et al., 2011; Renna et al., 2017) suggests that strategies grounded in mindfulness practice improve symptoms and well-being; however, mindfulness also presents challenges to study.

Mindfulness is the process of attending to the present moment with an orientation of openness, curiosity, compassion, acceptance/non-judgment, non-evaluation, and non-identification (summarized as awareness with acceptance; Bishop et al., 2006; Chambers et al., 2009; Desbordes et al., 2014; Grossman & Van Dam, 2011; Lindsay & Creswell, 2017). While some studies consider mindfulness a specific strategy that can be applied (Chiesa et al., 2013; Keng et al., 2013; J. Lutz et al., 2014), others consider it a trait associated with reduced emotional reactivity or more effective regulation (Egan et al., 2018; Quaglia et al., 2016; Rowland et al., 2020).

In a recent review, we proposed the Dual-mode Model of Mindful Emotion Regulation (D-MER) to address some of these challenges (Raugh & Strauss, 2023). This model—which labels the two components of mindfulness as monitoring (for awareness) and equanimity (for acceptance)—posits that mindfulness acts through two modes: implementation as emotion regulation strategies (similar to reappraisal or distraction) and facilitation of effective emotion regulation (i.e., as a moderator that increases the effectiveness of emotion regulation).

Dual-mode Model of Mindful Emotion Regulation

Implementation

The implementation mode proposes that mindfulness components of monitoring (awareness) and equanimity (acceptance) can be applied to change emotional experience as strategies within the Extended Process Model (Table 1). Within this mode, mindfulness-based techniques function to regulate emotions similarly to more traditionally studied strategies (e.g., cognitive reappraisal, distraction, expressive suppression). This model categorizes monitoring as a form of attentional deployment which shifts attention from the content of experience to the process of experience (i.e., how one is attending rather than what they are attending to) to influence perception of affective stimuli. Where distraction would shift attention away from an experience entirely, monitoring maintains that attention but adopts a broader attentional lens. Equanimity is categorized as a form of cognitive change which targets how reactions to a stimulus are appraised rather than the meaning or consequences of the stimulus itself. Cognitive reappraisal of an anxiety-inducing stimulus may challenge the accuracy of the appraisal (“I’m not in danger”), whereas equanimity-based cognitive change would prompt self-compassion or emotional acceptance (“it’s okay to feel anxious”). In this model, monitoring is considered a form of attentional deployment and equanimity a form of cognitive change, rather than proposing that mindfulness-based strategies are an independent family.

The D-MER further posits that mindfulness-based strategies may serve the function of regulating emotions (i.e., decreasing negative affect or promoting adaptive behavior) even in the absence of a specific regulatory goal (for detailed discussion, see Raugh & Strauss, 2023). Further, empirical evidence suggests that mindfulness components effectively regulate emotions.

Prior reviews have evaluated the effective implementation of emotion regulation strategies broadly (Webb et al., 2012) or equanimity/acceptance in particular (Hernández-Gómez & Hervas, 2022; Kohl et al., 2012) and found them to be effective (g ≈ .3). Equanimity-based strategies (specifically acceptance) are similar in effectiveness to other strategies, including cognitive reappraisal (Kohl et al., 2012; Webb et al., 2012). However, this effect was larger for pain tolerance than negative emotion outcomes (Kohl et al., 2012). These reviews also indicated that effectiveness may vary based on a range of factors including the emotion(s) induced, the outcome being regulated (pain, negative emotion, positive emotion), control comparison used, and number of regulation attempts (Webb et al., 2012). Further, heterogeneity in instructional quality or detail may influence how effectively participants can apply mindfulness-based strategies (Hernández-Gómez & Hervas, 2022).

These reviews have provided valuable insights into equanimity as an effective emotion regulation strategy; however, several directions remain unexplored. How effective is equanimity as a strategy in clinical populations for whom it may be most relevant, but also more difficult to use? Prior literature is also limited by heterogeneity in the use of “acceptance” to describe a wide range of strategies (Hernández-Gómez & Hervas, 2022; Wojnarowska et al., 2020). Given this heterogeneity, it is unclear how precisely earlier studies evaluated equanimity. Finally, although no meta-analyses to date have comprehensively evaluated the effects of monitoring or mindfulness as emotion regulation strategies, substantial research on both exists which support their effectiveness for changing emotional experience (Beblo et al., 2018; Brown et al., 2022; Fitzpatrick et al., 2019; Forsyth & Hayes, 2014; Keng et al., 2013; A. Lutz et al., 2009; Murakami et al., 2015; Rosenbaum et al., 2020; Uusberg et al., 2016; Zhang et al., 2019). However, some research suggests that equanimity is the component of mindfulness that downregulates negative emotion and that using monitoring alone may increase negative emotion (Lindsay & Creswell, 2017; Ortner, 2015; Thompson et al., 2009). Although existing evidence supports the effectiveness of mindfulness-based strategies, uncertainty remains regarding the effects of specific strategies among different populations or for different purposes and the ways in which mindfulness skills may facilitate effective regulation.

Facilitation

The facilitation mode proposes that mindfulness components of monitoring and equanimity improve emotion regulation via cognitive control (e.g., greater ability to remain focused on a regulatory goal or represent more regulatory options) and reduced reactivity to negative thoughts and stimuli, respectively. These effects are posited to differ for monitoring and equanimity. Monitoring, or awareness of the present moment, can aid emotion regulation through improved cognitive control and ability to orient to relevant stimuli; however, it can also increase emotional reactivity due to a focus on negative stimuli when not paired with equanimity (Fletcher et al., 2010; Hölzel et al., 2011; Inzlicht et al., 2014; Lindsay & Creswell, 2017, 2019; Raugh & Strauss, 2023; Teper et al., 2013). Equanimity is thought to reduce emotional reactivity through lowering neurophysiological responses to stimuli, reducing initial negative appraisals, and reducing emotional meta-cognitions (e.g., “I shouldn’t feel sad”) (Chambers et al., 2009; Delgado et al., 2010; Ford et al., 2018; Kiken & Shook, 2014).

However, the available research generally lacks the specificity to identify if effects are due to monitoring, equanimity, or the interaction of the two (c.f., Lindsay & Creswell, 2017) and which stages of the emotion regulation process are affected by mindfulness-based interventions. As such, these studies provide broad indication that facilitation is likely to occur but do not evaluate specific hypotheses of facilitation. If the specific hypotheses of facilitation hold, they provide clear guidance for treatments which use mindfulness skills as stand-alone treatment targets or to supplement emotion regulation skills. For example, support that equanimity reduces negative appraisals during emotion generation (as evidenced by reduced reactivity to negative stimuli) could be used to guide clients to not engage in overly costly, effortful, or harmful regulation strategies.

The Present Study

A meta-analysis was conducted to evaluate key hypotheses of the D-MER using laboratory studies which instruct the use of emotion regulation. To address gaps in prior meta-analyses, the present study used a priori strategy categorization (Table 1), included a wide range of mindfulness-based strategies, and included non-clinical and clinical samples. For strategy categorization, a version of the taxonomy from Webb and colleagues (2012) was modified to better represent mindfulness-based strategies based on Raugh and Strauss (2023). As evaluation of every facilitation hypothesis in a single study is not currently possible due to the scope of such a project and a dearth of available studies for identification and selection, this meta-analysis focused on emotional reactivity and effective implementation of instructed emotion regulation strategies. The following hypotheses were made:

1. Mindfulness-based strategies would demonstrate effectiveness for regulating emotions. Further, this effect would be greater among those provided with more extensive training or instructions (H1A) and for behavioral outcomes (e.g., pain tolerance) compared to other outcomes (e.g., positive or negative emotion, physiological responses) (H1B). It was hypothesized that monitoring and equanimity would show similar effects to other strategies within their same families (attentional deployment and cognitive change, respectively; H1C). A number of other methodological moderators were evaluated without directional hypotheses (Webb et al., 2012).

2. Greater self-reported mindfulness factors of monitoring and equanimity would be associated with more effective regulation (H2). Specifically, greater monitoring would be associated with increased emotional reactivity (H2A), greater acceptance would be associated with reduced reactivity (H2B), and both would be associated with greater effectiveness across strategies (H3C).

Method

Review protocol was preregistered with PROSPERO (CRD42022340089) and on OSF (osf.io/qv43d/); article screening, data extracted, codebooks, and analytic code are provided on OSF. See the Appendix for the PRISMA checklist (Page et al., 2021) and additional methodological details.

Search Strategy

In accordance with PRISMA guidelines (Page et al., 2021), a database search was conducted through July 2024 using PubMed, PsychInfo/PsychArticles, and Web of Science. Search terms were included for emotion regulation (“emotion”; “regulat*”); mindfulness (“mindful*”, “accept*”, “aware*”, “decenter”, or “defus*”); and experimental design (“task”, “condition”, “assigned”, or “instructed”). Additional records were included when identified in relevant reviews (Hernández-Gómez & Hervas, 2022; Kohl et al., 2012; Messina et al., 2021; Rompilla et al., 2022; Webb et al., 2012; Wojnarowska et al., 2020) or other sources.

Selection of Studies

The first author (IMR) reviewed all articles for inclusion through the title, abstract, then full manuscript. Inclusion criteria for studies were (1) available in English; (2) evaluated the effectiveness of at least one mindfulness-based emotion regulation strategy; (3) compared mindfulness-based emotion regulation to a reactivity control condition (e.g., passive viewing, instructions to not regulate); (4) indexed effectiveness via subjective experience (i.e., self-report ratings), physiological response (e.g., heart rate variability, electrodermal activity), or goal-directed behavior (e.g., pain tolerance, task persistence). Specific examples of study designs which did not meet inclusion criteria are included in the Supplement.

Coding of Studies

IMR coded all articles for sample characteristics, methodology, mindfulness, emotion regulation, and quality. Secondary coding was conducted by AMB with perfect consistency for numeric ratings and acceptable consistency for categorical ratings (ICC = 1; kappa = .82). See Supplement for additional coding details and OSF materials for full codebook.

Sample characteristics included age, gender, clinical status, country of participation, and race. Demographic characteristics were only coded for within-subjects designs as between-subjects designs were initially included for exploratory purposes. Clinical status was considered in four categories: (1) healthy (no noted medical or psychological illness), (2) at-risk (elevated risk for a clinical condition, e.g., family history of clinical disorder or trauma exposure), (3) clinical (meeting criteria for physical or mental health conditions), (4) general (no clear indication of clinical status). Clinical presentation was coded at four levels of specificity based on the HiTOP model (Kotov et al., 2017, 2021; Ringwald et al., 2021). Coding of clinical presentations at multiple levels and within a dimensional model allows for evaluation of transdiagnostic patterns and can improve power for moderation among spectra or subfactors which would not be available by evaluating individual syndromes.

Methodological characteristics coded included study design, order of reactivity and regulation conditions, stimulus type (image, audio, video, text, experiential), duration of emotion regulation training, detail and clarity of emotion regulation instructions, and number of trials where regulation was assessed. Mindfulness characteristics coded included the mean and standard deviation of mindfulness or mindfulness factors (awareness or acceptance), the measure used, and mindfulness experience. Mindfulness experience was categorized as consistent with the general population (i.e., exposure to concepts and likely past, infrequent practice), those trained in mindfulness (such as through a mindfulness-based intervention) but without long-standing practice, experts identified by prolonged practice (greater than 1 year or with frequency more than once per week) or by study authors, or mixed.

Emotion regulation effectiveness was considered across a range of outcomes including negative or positive emotions, syndromal symptoms (e.g., pain, social exclusion, cravings), psychophysiology, and goal-directed behaviors (e.g., pain tolerance, task persistence). Each outcome reported in a given study was coded separately. Further, coding included emotion regulation goals (i.e., upregulation or downregulation) and hedonic direction (contra-hedonic or pro-hedonic). Each emotion regulation strategy and the reactivity condition were coded according to Table 1. Quality was assessed using the Standard Quality Assessment Criteria For Evaluating Primary Research Paper (Kmet et al., 2004). No studies were excluded due to quality.

Effect Size Calculation

Within-group effectiveness was calculated using Cohen’s d_av (Cumming, 2011; Lakens, 2013). Where sd_av could not be calculated, sd_av was estimated from inferential statistics (i.e., t test p value) or the standard deviation of the difference score (sd_diff), if no other options were available. Between-group effectiveness was calculated using Cohen’s d and the standard formula for d_SE. For both effects, d_av and d were converted to Hedges’ g; g was used for calculation of standard errors. Effectiveness was calculated such that negative values reflect that successful regulation decreased outcome relative to reactivity condition (m_effect = m_regulation − m_reactivity). To make effects comparable across regulatory goals, all effects used a directional correction (downregulation =  − 1, upregulation = 1). Thus, effectiveness was calculated such that greater values indicate more successful regulation.

Statistical Analysis

Analyses were conducted using meta (Schwarzer et al., 2015) and metafor (Viechtbauer, 2010) packages in R (R Core Team, 2023). Effect sizes were pooled using random effects models. To account for studies reporting multiple samples and effects, multilevel meta-analysis was conducted to determine the best-fitting model. See Supplement for power analyses.

Moderation analyses for hypothesis one evaluated the effect of moderators on the effect of mindfulness-based strategies using the best-fitting model for the nested data. All moderators evaluated are listed in Table 2 with the results; exploratory moderators are indicated in the table.

Table 2.

Moderation results of mindfulness-based strategy effectiveness

Moderator	Studies (K)	Effects (k)	Hedges’ g [95% CI]	Test statistic, p	b (SE)
Overall effect size and sensitivity analyses
Overall	110	417	.28 [.18, .38]	t = 5.51, p < .001
Overall (95% outliers removed)	102	346	.16 [.13, .2]	t = 8.63, p < .001
Overall (influential removed)	103	390	.17 [.11, .24]	t = 5.23, p < .001
Overall (only fully powered)E	31	126	.31 [.07, .54]	t = 2.55, p = .011
Overall (n ≥ 70, 10% largest)E	17	53	.15 [0, .31]	t = 2.03, p = .047
Overall ( ≥ 55, 20% largest)E	28	97	.2 [.06, .34]	t = 2.84, p = .006
PowerE	110	417		t = 0.49, p = .623	.03 (.07)
Publication status				F = 3.08, p = .08
Peer-reviewedI	97	373	.25 [.14, .35]	-	-
Dissertation	13	44	.52 [.23, .81]	t = 1.75, p = .08	.28 (.16)
Quality	110	417		t =  − 0.36, p = .715	− .13 (.374)
Methodological characteristics
Type				F = 4.85, p = .028
Between-subjectsI	39	140	.13 [− .03, .3]	−	−
Within-subjects	71	277	.36 [.24, .48]	t = 2.2, p = .028	.23 (.1)
Design				F = 0.65, p = .522
ObservationalI	90	346	.27 [.16, .38]	-	-
Brief intervention	13	43	.17 [− .13, .47]	t =  − 0.66, p = .508	− .10 (.16)
Post-interventionE	7	12	.48 [.02, .93]	t = 0.87, p = .386	.2 (.23)
InterventionE				F = 1.34, p = .264
Intervention BaselineI	5	6	.71 [.02, 1.4]	-	-
Post-intervention	7	12	.51 [− .15, 1.17]	t =  − 1.16, p = .264	− .2 (.18)
Training				F = 0.54, p = .584
BriefI	65	257	.28 [.15, .41]	-	-
Moderate	27	95	.25 [.04, .45]	t =  − 0.23, p = .815	− .03 (.12)
Extended	13	39	.43 [.13, .73]	t = .92, p = .356	.15 (.17)
Instruction detail				F = 0.74, p = .524
BriefI	23	94	.17 [− .05, .4]	-	-
Description	21	81	.31 [.07, .55]	t = 0.81, p = .418	.14 (.17)
Didactic	13	37	.46 [.15, .78]	t = 1.45, p = .147	.29 (.2)
Combination	27	114	.32 [.11, .53]	t = 0.92, p = .36	.14 (.15)
Instruction clarityE				F = 1.21, p = .307
UnclearI	12	52	.23 [− .05, .52]	-	-
Ambiguous	14	52	.19 [− .06, .44]	t =  − .24, p = .813	− .04 (.18)
Vague	32	102	.17 [− .01, .36]	t =  − 0.35, p = .728	− .06 (.17)
Clear	58	211	.37 [.23, .5]	t = 0.85, p = .396	.13 (.16)
Task characteristics
Task order				F = 0.71, p = .491
CounterbalancedI	20	94	.29 [.03, .55]	-	-
React first	31	107	.49 [.28, .71]	t = 1.19, p = .233	.20 (.17)
Random	17	63	.41 [.13, .69]	t = 0.62, p = .537	.12 (.19)
Duration of regulationE	51	207		t =  − 1.18, p = .238	− .0005 (.0004)
Induction stimuliE				F = 3.85, p = .004
ExperientialI	41	150	.34 [.2, .49]	-	-
Image	19	82	.13 [− .09, .36]	t = 1.54, p = .122	− .21 (.14)
Mixed	8	24	.33 [− .04, .69]	t = .09, p = .931	− .02 (.2)
Text	10	25	.77 [.44, 1.1]	t = 2.32, p = .021	.43 (.18)
Video	24	109	.06 [− .14, .26]	t = 2.21, p = .028	− .28 (.13)
Induction familyE				F = 3.53, p = .03
Negative affectI	75	268	.2 [.08, .32]	-	-
Positive affect	8	22	.22 [− .01, .46]	t = 0.21, p = .836	.02 (.11)
Symptoms	31	106	.5 [.31, .68]	t = 2.66, p = .008	.3 (.11)
Outcome family				F = 5.59, p < .001
Negative affectI	96	329	.25 [.14, .36]	-	-
Positive affect	26	60	.26 [.09, .42]	t = .07, p = .941	.01 (.08)
Symptoms	36	107	.16 [0, .3]	t =  − 1.35, p = .177	− .09 (.07)
Psychophysiology	36	187	.33 [.19, .46]	t = 1.35, p = .177	.08 (.06)
Behavior	19	51	.69 [.48, .91]	t = 4.02, p < .001	.44 (.11)
ArousalE	8	33	.15 [− .08, .38]	t =  − 0.91, p = .368	− .1 (.11)
Specific outcome				F = 6.03, p < .001
Negative affectI	73	197	.24 [.13, .35]	-	-
Electrodermal activity	22	74	.32 [.16, .48]	t = 1.01, p = .316	.08 (.08)
Heart Rate	19	47	.18 [0, .36]	t =  − 0.65, p = .516	− .06 (.09)
Positive affect	20	47	.23 [.06, .4]	t =  − 0.13, p = .9	− .01 (.08)
Pain experience	17	41	.1 [− .13, .33]	t =  − 1.16, p = .248	− .14 (.12)
Pain tolerance	13	40	.93 [.67, 1.18]	t = 5.06, p < .001	.69 (.14)
Sadness	13	58	.22 [.03, .42]	t =  − 0.13, p = .894	− .01 (.1)
Anxiety	12	36	.13 [− .1, .35]	t =  − 0.98, p = .33	− .11 (.12)
Symptom intensity	10	25	.24 [0, .48]	t =  − .04, p = .97	0 (.12)
Personal relevanceE				F = 12.79, p < .001
ImpersonalI	83	327	.19 [.08, .29]	-	-
Personal	28	90	.56 [.38 .74]	t = 3.58, p < .001	.37 (.1)
Hedonic directionE				F = .01, p = .91
Contra-hedonicI	7	12	.22 [− .06, .5]	-	-
Pro-hedonic	104	375	.24 [.14, .33]	t = 0.11, p = .91	.02 (.14)
Reactivity condition				F = .02, p = .89
None (NR1)I	71	257	.3 [.17, .43]	-	-
React naturally (NR4)	31	122	.29 [.09, .48]	t =  − 0.14, p = .89	− .02 (.12)
Sample characteristics
Student sample				F = 0.38, p = .539
Non-studentI	62	250	.3 [.17, .43]	−	−
Students	51	167	.25 [.12, .38]	t =  − 0.62, p = .539	− .05 (.08)
Clinical status				F = 1.22, p = .296
HealthyI	36	130	.22 [.07, .37]	-	-
General	41	139	.4 [.23, .57]	t = 0.47, p = .64	.03 (.07)
Clinical	38	124	.25 [.1, .4]	t = 1.56, p = .119	.18 (.12)
Superspectra				F = 0.42, p = .516
ControlI	36	130	.21 [.09, .34]	-	-
Emotional dysfunction	29	99	.26 [.12, .39]	t = 0.65, p = .516	.04 (.06)
Subfactors				F = 2.78, p = .064
ControlI	36	130	.18 [.05, .31]	-	-
Distress	12	39	.15 [− .04, .34]	t =  − 0.32, p = .749	− .03 (.09)
Fear	13	37	.45 [.23, .67]	t = 2.28, p = .024	.27 (.12)
Proportion treated	16	46		t =  − 0.98, p = .333	− .22 (.23)
Country				F = 3.47, p = .032
EuropeanI	39	167	.13 [− .04, .3]	-	-
Americas	37	135	.43 [.27, .61]	t = 2.63, p = .009	.31 (.12)
Asia	12	32	.26 [− .06, .56]	t = 0.78, p = .438	.13 (.17)

^I = model intercept and reference category, ^E = exploratory analysis which was not preregistered or was included despite being underpowered. Bolded rows have an omnibus effect or specific contrast with p < .05

Hypothesis two (facilitation) was evaluated through meta-regression of the effect of study-level trait mindfulness variables (i.e., average mindfulness score for a given study, including mindfulness experience) on emotion regulation effectiveness across strategies. Meta-regression was considered less optimal to address this question than meta-analysis of Pearson’s correlations; however, no studies reported a correlation of trait mindfulness with emotion regulation effectiveness. In order to compare mindfulness scores on different measures, min–max normalization was implemented (normalized = (observed − minimum possible) / range) with a final scale of 0–10. Evaluation of relations between mindfulness and reactivity used a similar meta-regression approach and min–max normalization (outcomes without a specific upper or lower limit were excluded from this analysis).

Sensitivity analyses included assessment of outliers (defined using confidence intervals) and influential cases (determined by evaluation of study weights and leave-one-out analyses). Possible bias and small-study effects were evaluated through visual inspection of the funnel plot, Egger’s test of funnel plot asymmetry (Egger et al., 1997), rank correlation test (Begg & Mazumdar, 1994), trim-and-fill analysis (Duval & Tweedie, 2000), limit meta-analysis (Rücker, Carpenter et al., 2011; Rücker, Schwarzer et al., 2011), Rosenthal’s fail-safe N to reduce a significant effect to be nonsignificant, Orwin’s fail-safe N to reduce the observed effect size by 50% (Rosenberg, 2005), and p curve analysis (Simonsohn et al., 2014). See Supplement for additional details. Due to software limitations, some sensitivity and bias analyses required the use of the non-nested random effects model rather than multilevel meta-analysis; such instances are specified in the results section.

Results

Study Selection

A total of 2,037 records were screened for inclusion (see Fig. 1). Overall, 110 studies (K) were included consisting of 767 (k) effects. A total of 71 studies presented within-subjects effects (k = 535), while 39 studies presented between-subjects effects (k = 232). These effects reflect a total of 8105 individuals (within = 4,442; between, regulation = 2,433; between, control = 1,230). Of the total effects coded, 417 were coded as a mindfulness-based emotion regulation strategy (within = 277, between = 140). See the Appendix for a summary of included studies.

Fig. 1.

Study flowchart

The majority of coded samples (133 unique samples from within-subjects designs) reported age (k = 124) and gender (k = 107), but relatively few reported racial composition (k = 41). The mean age of participants was 29.5 (SD = 7.8) and the majority were women (2403 of 3711, 64.77%). Of those with reported race or ethnicity, the majority were White (793 or 1,586, 59.77%) or Asian (519 of 1586, 39.75%); all other categories were represented by fewer than 100 participants (see Supplement).

Effectiveness

Overall Mindfulness-Based Effect

A standard random effects model observed a significant effect of mindfulness-based strategies (g = .23, 95% CI [.17, .29], t[416] = 7.2, p < .001) with significant heterogeneity (Q[416] = 1,751.69, p < .001, I² = 76.3%). Multilevel meta-analysis with effects nested within study also observed a significant effect (Table 2 and Fig. 2); the total heterogeneity (I²_Overall = 83.3%) was primarily observed at the level of study (I²_Study = 62.13%, I²_Effect = 21.17%). The model of effects nested within studies showed the best fit among tested models (see Supplement) and thus was used as the base model for subsequent analyses.

Fig. 2.

Forest plot of within-subjects and between-subjects effects for mindfulness-based strategies. Note. Each point reflects the Hedges’ g with 95% confidence intervals from a given effect; the size of the center circle reflects the sample size used to determine the effect. Dashed line is at g = 0. Grey areas reflect pooled estimates of Hedges’ g with 95% confidence range; they do not include zero. See main text for Hedges’ g point estimates and confidence range. Overall effect plot includes all mindfulness strategies while sub-plots include only the strategies indicated

Moderators

A summary of moderators of the mindfulness-based effect is presented in Table 2.

Methodological Factors

Within-subjects designs demonstrated a greater effect size than between-subjects designs. Although significantly different, subsequent analyses pool within-subjects and between-subjects effects as applicable to maximize power and improve generalizability across designs. No other methodological characteristics showed significant moderation effects.

Task Characteristics

Effect size was moderated by the type of stimulus used, where text-based stimuli showed the greatest effect and video-based stimuli the smallest. Studies that induced symptom-specific experiences (e.g., social anxiety, chronic pain, compulsive urges) had a greater effect than those that induced negative affect. There were significant omnibus moderation effects of outcome, both overall families and specific outcomes. Compared to negative affect, mindfulness-based strategies were most effective for regulating behavioral outcomes, specifically pain tolerance. They were also more effective for regulating responses to stimuli with personal relevance (e.g., autobiographical recall, individualized worries) compared to standardized stimuli (e.g., IAPS). Effectiveness was not significantly moderated by hedonic direction (pro-hedonic versus contra-hedonic) or reactivity condition.

Exploratory follow-up on personal relevance, induction stimuli, and induction family indicated that conditions with the greatest effect size (personal stimuli, text stimuli, symptoms) also showed greater reactivity (see Supplement).

Sample Characteristics

Clinical status (overall status of healthy or clinical, superspectra, and subfactors) omnibus tests did not indicate significant moderation. However, the specific contrast of the fear subfactor showed a significantly greater effect compared to control. Only a limited number of subfactors could be evaluated and no clinical syndrome was represented with sufficient power for analysis. An analysis of spectra would have been redundant with superspectra and thus was not conducted. The proportion of individuals in a clinical group receiving treatment (including pharmacological or psychotherapy) did not moderate strategy effectiveness. Participant country (as a proxy for culture) showed significant moderation such that the effect of mindfulness-based strategies was greater in the Americas (North and South America) compared to European or other “Western” countries (including Australia). There were insufficient samples for more specific country-based contrasts.

Strategy Comparisons

There were significant omnibus effects of strategy and tactic (Table 3). Monitoring showed a significantly lower effect than overall mindfulness, specifically the tactic of emotional awareness. The pooled estimate for each strategy ranged from .17 and .43.

Table 3.

Effectiveness by strategy and tactic

Moderator	Studies (K)	Effects (k)	Hedges’ g [95% CI]	Test statistic, p	b (SE)
Strategy				F = 3.83, p = .002
MindfulnessI	14	72	.36 [.19, .52]	-	-
Equanimity	74	262	.3 [.19, .41]	t =  − 0.73, p = .468	− .06 (.08)
Monitoring	29	83	.17 [.02, .32]	t =  − 1.96, p = .05	− .19 (.1)
Reappraisal	46	174	.43 [.31, .55]	t = 0.98, p = .325	.07 (.07)
Distraction	18	73	.43 [.28, .58]	t = 0.87, p = .383	.07 (.08)
Suppression	27	79	.31 [.16, .46]	t =  − 0.55, p = .581	− .05 (.09)
Tactic				F = 3.17, p < .001
Mindfulness (H1)	14	72	.33 [.14, .52]
Equanimity—mixed (EM)	9	27	.36 [.07, .65]	t = 0.17, p = .861	.03 (.17)
Acceptance (E1)	34	133	.25 [.11, .39]	t =  − 0.8, p = .425	− .08 (.1)
De-fusion (E2)	25	64	.37 [.18, .56]	t = 0.3, p = .767	.04 (.13)
Self-compassion (E3)	10	38	.27 [.08, .47]	t =  − 0.5, p = .621	− .06 (.12)
Monitoring—mixed (MM)	9	18	.32 [.04, .59]	t =  − .08, p = .935	− .01 (.16)
Awareness—emotional (M2)	12	46	.02 [-.19, .22]	t =  − 2.46, p = .014	− .32 (.13)
Reappraisal—mixed (RM)	13	62	.39 [.22, .55]	t = 0.59, p = .558	.06 (.09)
Reappraise stimulus meaning (R1)	22	63	.33 [.17, .48]	t =  − .05, p = .957	− .01 (.1)
Reappraise via perspective taking (R2)	12	49	.56 [.37, .74]	t = 1.81, p = .071	.22 (.12)
Suppression—mixed (SM)	11	31	.38 [.16, .6]	t = 0.34, p = .732	.05 (.14)

^I = model intercept and reference category, ^E = exploratory analysis which was included despite being underpowered. See text for post hoc contrasts

Post hoc comparisons of strategy indicated equanimity was significantly less effective than reappraisal (F = 4.18, p = .016), while monitoring was significantly less effective than distraction (F = 4.13, p = .017).

Within equanimity, there was no significant difference among tactics (F = .46, p = .765). Within monitoring, emotional awareness was significantly less effective than mixed monitoring tactics (F = 3.38, p = .035). There was a significant difference in effect sizes among equanimity and reappraisal tactics (F = 2.84, p = .015) such that although acceptance did not significantly differ from reappraising stimulus meaning (F = .56, p = .573), it was significantly less effective than reappraising via perspective taking (F = 6.9, p = .001).

Facilitation

Facilitation analyses (Table 4) are considered exploratory due to the presence of fewer than 10 independent effects.

Table 4.

Models of facilitation

Moderator	Studies (K)	Effects (k)	b (SE)	Test statistic, p
Effectiveness
Overall	71	535	.41 (.07)	t = 5.58, p < .001
Mindfulness experience				F = 0.1, p = .902
GeneralI	18	80	.47 (.12)	t = 4.02, p < .001
Mixed	5	14	− .08 (.25)	t =  − 0.31, p = .759
Trained	6	13	.04 (.12)	t = 0.31, p = .754
Mindfulness	7	64	− .04 (.09)	t =  − 0.48, p = .63
Awareness	5	63	− .06 (.12)	t =  − 0.48, p = .631
Acceptance	8	70	− .04 (.09)	t =  − 0.42, p = .678
Reactivity
Overall	62	286	8.45 (.79)	t = 10.74, p < .001
Mindfulness experience				F = 0.79, p = .457
GeneralI	17	61	5.25 (.23)	t = 22.52, p < .001
Mixed	3	9	.55 (.6)	t = 0.93, p = .355
Trained	6	13	− .23 (.59)	t =  − 0.39, p = .695
Mindfulness	7	42	.0 (.31)	t = .01, p = .99
Awareness	5	39	.56 (3.18)	t = 0.18, p = .861
Acceptance	8	46	− .39 (.18)	t =  − 2.2, p = .033

^I = model intercept and reference category. All facilitation analyses are considered exploratory due to low sample size. Bolded rows have an omnibus effect or specific contrast with p < .05

For effectiveness, mindfulness experience and study-level trait mindfulness, awareness, and acceptance did not demonstrate significant moderation. Further, mindfulness experience and study-level trait mindfulness or awareness did not significantly moderate reactivity. However, reactivity was significantly moderated by study-level acceptance such that greater acceptance was associated with lower reactivity.

Sensitivity

After removing outliers based on non-overlapping 95% confidence intervals, 102/110 studies and 346/417 effects were retained. While the effect estimate was reduced, it remained significant (Table 2); see Supplement for the forest plot. Influence analysis identified seven studies with undue influence (Goldin et al., 2021, Forsyth & Hayes 2014, Jackson et al., 2012, Hofman 2020; Horowitz, 2009, Vowles et al., 2007, Wolgast et al., 2011). After removing these studies, the effect estimate remained significant (Table 2). Leave-one-out analysis using the unnested random effects model indicated a range of Hedges’ g between .21 and .28. Those effects whose exclusion results in the lowest g value were generally from the same studies identified during influence analysis.

Bias

Overall, results from sensitivity and bias analyses suggest that the overall effect estimate is impacted by outliers and small-study effects (possibly reflecting publication bias, see Fig. 3). However, sensitivity analyses consistently observed significant effects between .15 and .3 (see Table 2 for effect size estimates and Supplement for more details on sensitivity analyses).

Fig. 3.

Funnel plot of within-subjects and between-subjects effects for mindfulness-based strategies. Note. Shaded area in the top half reflects 95% confidence range. Visual inspection of the funnel plot and box-and-whisker plot indicates asymmetry with more values to the right of the plot. See the main text and supplement for statistical tests of bias

Discussion

Consistent with the Dual-mode Model of Mindful Emotion Regulation (Raugh & Strauss, 2023), the present study evaluated the effective implementation of mindfulness-based strategies and the facilitation of emotion reactivity and strategy implementation by mindfulness (Fig. 3). Overall, results indicated strong support for effective implementation, some support for facilitation via lowered reactivity, and no support for facilitation via more effective implementation.

Implementation

As hypothesized, there was a robust effect of mindfulness-based strategies (overall g between .15 and .3). This is consistent with the effects of other strategies (Kohl et al., 2012; Webb et al., 2012) and supports that mindfulness-based strategies can be implemented to improve emotional experience (Raugh & Strauss, 2023). The effect was greater for within- than between-subjects designs, consistent with prior meta-analyses (Webb et al., 2012) and statistical theory that within-subjects designs reduce error to dis-attenuate effect size estimates (Greenwald, 1992).

Stimuli presented in text demonstrated a greater effect size than other stimuli. As they also showed the greatest reactivity, they may reduce floor effects that attenuate the ability to observe effective regulation. Personally relevant materials may prompt more cognitive fusion, identification, and rumination which mindfulness-based strategies are meant to target in therapeutic settings (Bernstein et al., 2015; Gustavson et al., 2018; Hayes et al., 2011). By reducing self-referential processing and identification with the target experience (especially symptoms), affective intensity may be reduced effectively. It is also notable that these three features appear to be independent (i.e., studies did not generally have a combination of personally relevant, symptom-evoking, and text-based stimuli); however, four studies that did use all three (Goldin & Gross, 2010; Goldin et al., 2013, 2019, 2021) had some of the greatest effect sizes included in this meta-analysis.

Consistent with hypotheses, there were significant differences in effectiveness for specific outcomes. In line with prior meta-analysis (Kohl et al., 2012), mindfulness-based strategies were more effective for pain tolerance than negative affect. This suggests that these strategies are effective for changing behavioral outcomes of unpleasant experiences; however, additional research is warranted given that other behavioral outcomes not involving pain (e.g., distress tolerance, task persistence) were not adequately represented.

Mindfulness-based strategies demonstrated effectiveness among clinical populations. While strategy effectiveness was not significantly moderated by clinical status, those with fear-based presentations (e.g., social anxiety, specific phobia, obsessive–compulsive disorder, trauma) may especially benefit from mindfulness-based strategies. This group demonstrated greater effectiveness than controls, which warrants further study as mindfulness-based skills may enhance exposure-based treatments (Hölzel et al., 2011; Kickuk & Austad, 2010; Uusberg et al., 2016).

Omnibus effects of strategy and tactic were significant; post hoc contrasts suggested that some mindfulness-based tactics (especially emotional awareness) can be less effective than comparisons (especially perspective taking reappraisal). Within the same regulation families (attentional control and cognitive change), monitoring was less effective than distraction and only certain contrasts between equanimity and reappraisal tactics were significant. Thus, although strategies are broadly comparable in effectiveness, those which emphasize only emotional awareness are likely less advantageous than those which include cognitive change. This is consistent with evidence and theory that emotional awareness without acceptance can worsen emotional reactivity and mood (Lindsay & Creswell, 2017; Ortner, 2015; Thompson et al., 2009). However, emotional awareness techniques are a vital component of mindfulness practice and their use should be encouraged despite lower efficacy than other strategies (Lindsay & Creswell, 2017). Emotional awareness can be an important step in the development and practice of emotional acceptance/equanimity, which was found to be effective in the present review.

Although the overall effect was broadly consistent across other moderators, this may be due to the low power of moderated meta-analysis rather than an absence of true differences (Cuijpers et al., 2021). The nonsignificant effect of instruction clarity may highlight an important methodological consideration. Instructions provided to participants may not reflect the strategy intended by researchers, be insufficient to represent the “active ingredients” of a given strategy, or blend strategies (Hernández-Gómez & Hervas, 2022; Kohl et al., 2012; Webb et al., 2012; Wojnarowska et al., 2020). Researchers seek brief instructions for how participants should regulate their emotions; however, brevity should be carefully balanced to maintain validity. Examples of equanimity instructions which were brief (i.e., delivered in less than 5 min) but rated as “clear” can be found in the Appendix. The moderation effect of participant country supports the importance of additional research into cultural factors which may impact emotion regulation. Cultures in North and South America may endorse more harmful emotion beliefs such as a negative orientation towards unpleasant or negatively valenced emotions (Miyamoto et al., 2017) which may be uniquely targeted by mindfulness-based strategies.

Facilitation

Facilitation analyses generally did not support hypotheses. Average self-reported equanimity within a study was associated with lower average emotional reactivity of that study, while similar relations among mindfulness traits and reactivity or regulation were not observed. This provides an estimate of what effects may be present, but cannot test individual-level predictions such as those made by D-MER. It cannot be said from this study that individual differences in equanimity are associated with individual differences in reactivity, nor that individual differences in other mindfulness traits are not associated with individual differences in emotion regulation effectiveness. Thus, facilitation analyses should be considered preliminary and explicitly evaluated in future research.

Limitations

While power, cell size for some moderators or interactions, and lack of correlations to test facilitation have been discussed previously, other limitations should be considered. First, studies evaluated the instructed use of mindfulness, but designs including implicit emotion regulation or spontaneous regulation were not included. Similarly, it is unclear how the effects found in this meta-analysis may generalize to more ecologically valid designs or those which allow for polyregulation. Results around personally relevant and experiential stimuli suggest that mindfulness strategies are effective in response to situations either drawn from or reminiscent of daily life, especially regarding clinical symptoms. Some studies show that our results hold for ecological designs as well (Ford et al., 2018; Vilardaga et al., 2013; Zetsche et al., 2023). As for polyregulation, there is evidence that self-compassion (a tactic within equanimity) can improve cognitive reappraisal (Diedrich et al., 2016) and that acceptance can be a frequently selected strategy (Szasz et al., 2018), but extensive research on polyregulation is lacking. Although promising, generalizability of the results is inherently limited as they are drawn from laboratory settings and samples that are primarily young, White, European, women (Henrich et al., 2010). While analyses were conducted in order to minimize the influence of publication bias, bias cannot be precisely quantified and thus should be considered in the interpretation of results.

Conclusions

This meta-analysis was undertaken to test two propositions of D-MER: (1) effective implementation of mindfulness-based strategies and (2) facilitation of emotion regulation by trait-level mindfulness. Even with the limitations listed, the present study presents a large, well-powered, theoretically rigorous, preregistered investigation of the effectiveness of mindfulness-based strategies (monitoring and equanimity) across clinical presentations and outcomes. The first of the propositions (implementation) was supported; mindfulness-based strategies can be effectively used across populations to improve emotional experience and adaptive behaviors, although monitoring alone may be ineffective. Overall, results indicated a small, significant effect of mindfulness-based strategies between .15 and .3 which is consistent across moderators and maintained after efforts to remove the effects of outliers and small studies. The second of the propositions (facilitation) was not clearly supported due the nature of the data available; however, there is some preliminary indication that trait-level equanimity reduces emotional reactivity. No emotion regulation strategy will be “best” for all people at all times under all circumstances; however, mindfulness is a valuable and effective set of tools which can help people manage their emotions and symptoms.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Appendix

Tables 5 and 6.

Table 5.

PRISMA checklist

Section and Topic	Item #	Checklist item	Location where item is reported
TITLE
Title	1	Identify the report as a systematic review	1
ABSTRACT
Abstract	2	See the PRISMA 2020 for Abstracts checklist	2
INTRODUCTION
Rationale	3	Describe the rationale for the review in the context of existing knowledge	3–7
Objectives	4	Provide an explicit statement of the objective(s) or question(s) the review addresses	7–8
METHODS
Eligibility criteria	5	Specify the inclusion and exclusion criteria for the review and how studies were grouped for the syntheses	8
Information sources	6	Specify all databases, registers, websites, organisations, reference lists and other sources searched or consulted to identify studies. Specify the date when each source was last searched or consulted	8
Search strategy	7	Present the full search strategies for all databases, registers and websites, including any filters and limits used	8
Selection process	8	Specify the methods used to decide whether a study met the inclusion criteria of the review, including how many reviewers screened each record and each report retrieved, whether they worked independently, and if applicable, details of automation tools used in the process	9
Data collection process	9	Specify the methods used to collect data from reports, including how many reviewers collected data from each report, whether they worked independently, any processes for obtaining or confirming data from study investigators, and if applicable, details of automation tools used in the process	9–10
Data items	10a	List and define all outcomes for which data were sought. Specify whether all results that were compatible with each outcome domain in each study were sought (e.g., for all measures, time points, analyses), and if not, the methods used to decide which results to collect	9–11
	10b	List and define all other variables for which data were sought (e.g., participant and intervention characteristics, funding sources). Describe any assumptions made about any missing or unclear information	9–11
Study risk of bias assessment	11	Specify the methods used to assess risk of bias in the included studies, including details of the tool(s) used, how many reviewers assessed each study and whether they worked independently, and if applicable, details of automation tools used in the process	11–12
Effect measures	12	Specify for each outcome the effect measure(s) (e.g., risk ratio, mean difference) used in the synthesis or presentation of results	11
Synthesis methods	13a	Describe the processes used to decide which studies were eligible for each synthesis (e.g., tabulating the study intervention characteristics and comparing against the planned groups for each synthesis (item #5))	9
	13b	Describe any methods required to prepare the data for presentation or synthesis, such as handling of missing summary statistics, or data conversions	10–12
	13c	Describe any methods used to tabulate or visually display results of individual studies and syntheses	10–12
	13d	Describe any methods used to synthesize results and provide a rationale for the choice(s). If meta-analysis was performed, describe the model(s), method(s) to identify the presence and extent of statistical heterogeneity, and software package(s) used	10–12
	13e	Describe any methods used to explore possible causes of heterogeneity among study results (e.g., subgroup analysis, meta-regression)	10–12
	13f	Describe any sensitivity analyses conducted to assess robustness of the synthesized results	10–12
Reporting bias assessment	14	Describe any methods used to assess risk of bias due to missing results in a synthesis (arising from reporting biases)	10–12
Certainty assessment	15	Describe any methods used to assess certainty (or confidence) in the body of evidence for an outcome	10–12
RESULTS
Study selection	16a	Describe the results of the search and selection process, from the number of records identified in the search to the number of studies included in the review, ideally using a flow diagram	12
	16b	Cite studies that might appear to meet the inclusion criteria, but which were excluded, and explain why they were excluded	Supplement
Study characteristics	17	Cite each included study and present its characteristics	Appendix
Risk of bias in studies	18	Present assessments of risk of bias for each included study	Appendix
Results of individual studies	19	For all outcomes, present, for each study: (a) summary statistics for each group (where appropriate) and (b) an effect estimate and its precision (e.g., confidence/credible interval), ideally using structured tables or plots	Figure 2
Results of syntheses	20a	For each synthesis, briefly summarise the characteristics and risk of bias among contributing studies	15–16
	20b	Present results of all statistical syntheses conducted. If meta-analysis was done, present for each the summary estimate and its precision (e.g., confidence/credible interval) and measures of statistical heterogeneity. If comparing groups, describe the direction of the effect	12–16
	20c	Present results of all investigations of possible causes of heterogeneity among study results	13–16
	20d	Present results of all sensitivity analyses conducted to assess the robustness of the synthesized results	13–16
Reporting biases	21	Present assessments of risk of bias due to missing results (arising from reporting biases) for each synthesis assessed	13–16
Certainty of evidence	22	Present assessments of certainty (or confidence) in the body of evidence for each outcome assessed	13–16
DISCUSSION
Discussion	23a	Provide a general interpretation of the results in the context of other evidence	16–20
	23b	Discuss any limitations of the evidence included in the review	19–20
	23c	Discuss any limitations of the review processes used	19–20
	23d	Discuss implications of the results for practice, policy, and future research	16–20
OTHER INFORMATION
Registration and protocol	24a	Provide registration information for the review, including register name and registration number, or state that the review was not registered	8
	24b	Indicate where the review protocol can be accessed, or state that a protocol was not prepared	8
	24c	Describe and explain any amendments to information provided at registration or in the protocol	8–12 & Supplement
Support	25	Describe sources of financial or non-financial support for the review, and the role of the funders or sponsors in the review	1
Competing interests	26	Declare any competing interests of review authors	1
Availability of data, code and other materials	27	Report which of the following are publicly available and where they can be found: template data collection forms; data extracted from included studies; data used for all analyses; analytic code; any other materials used in the review	8

From: Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. https://doi.org/10.1136/bmj.n71

Table 6.

Article details

Reference	Type	N	Samples	k	Sample description	Induction stimulus	Induction type	Outcome(s)	Reactivity	Strategy training	Regulation (instruction clarity)	Quality
Alberts et al., 2012	Between	60	2	2	Students	Video	NegAffect	NegAffect	NR1	Brief	M2 (ambiguous), SM (clear)	1.50
Aldao & Mennin, 2012	Between	67	2	24	GAD and CN	Video	Anxiety and disgust and sadness	Anxiety, disgust, HRV, sadness	NR1	Brief	E1 (unclear), R2 (clear)	1.75
Ascone et al., 2017	Between, brief intervention	50	1	4	SZ, inpatient	Experiential	NegAffect	EDA, happiness, NegAffect, Sx intensity	O	Extended	E3 (clear)	1.58
Baquedano et al., 2017	Within	50	2	4	Healthy, expert and general mindfulness	Image	Food	Craving, reaction time	NR3	Brief	E2 (clear)	1.58
Barrett-Model, 2010	Between, dissertation	46	2	6	Trauma exposed students	Experiential	Fear	NegAffect, risk detection	NR1	Moderate	E1 (clear), SM (clear)	1.67
Beblo et al., 2018	Within	65	3	12	General	Video	Anxiety and joy	Arousal, NegAffect, PA	NR4	Brief	M1 (clear), M2 (ambiguous), O (clear)	1.67
Bigliassi et al., 2022	Within	24	1	8	Healthy	Experiential	Somatic	Exertion, HR, HRV, pain	NR1	NA	C1 (clear), H1 (vague)	1.64
Boehme et al., 2019	Within	29	1	18	Healthy	Image	Anxiety and sadness	Anxiety, arousal, EDA, sadness	NR2	Brief	E1 (vague), R2 (clear), S2 (clear)	1.82
Boon et al., 2023	Within	55	1	4	Students	Video	Sadness	NegAffect	NR4	Brief	D3 (clear), E1 (vague), R1 (clear), S1 (clear)	1.67
Braams et al., 2012	Within	64	2	6	General	Experiential	Pain	HR, NegAffect, pain	NR1	Brief	SM (clear), E1 (clear)	1.67
Brady et al., 2021	Within	71	2	24	Healthy students	Video	Amusement and fear	Amusement, EDA, EMG, fear	NR4	Brief	H1 (clear), RM (unclear)	1.71
Branstetter-Rost et al., 2009	Between	99	2	6	Students	Experiential	Pain	Pain, pain tolerance	NR1	Extended	EM (vague)	1.58
Burklund, 2010	Within, dissertation	30	2	4	SAD and CN	Video	Disapproval	Sx intensity	NR4	Moderate	MM (clear), R1 (clear)	1.75
Burnham & Kocovski, 2024	Between	458	4	4	Students	Experiential	Anxiety	NegAffect	NR1	Brief	E3 (clear), R1 (clear)	1.83
Byrne & Kangas, 2022	Between	228	2	4	Healthy students	Video	NegAffect	NegAffect, PA	NR1	Brief	E1 (clear), O (clear)	1.83
Campbell-Sills et al., 2006	Within	58	2	8	Mixed clinical	Video	NegAffect	EDA, HR, NegAffect, RSA	NR1	Brief	E1 (clear), S2 (clear)	1.73
Dan-Glauser & Gross, 2013	Within	37	1	12	Students	Image	NegAffect and PA	ECG, EMG, NegAffect, PA	NR4	Brief	M2 (ambiguous), SM (clear)	1.82
Degen, 2007	Between, dissertation	114	1	3	Students	Experiential	Fear	NegAffect, Sx intensity	NR1	Brief	E2 (vague)	1.67
Diedrich et al., 2014	Within	48	1	3	MDD	Mixed	Sadness	Sadness	NR4	Brief	E1 (clear), E3 (clear), R1 (clear)	1.75
Ding et al., 2015	Between	40	1	3	Healthy students	Experiential	Frustration	HR, NegAffect, PA	NR1	Brief	EM (ambiguous)	1.50
Dixon et al., 2020	Within	148	2	4	SAD and CN	Audio	Negative beliefs	NegAffect	NR4	Brief	E1 (ambiguous), R1 (vague)	2.00
Dunn et al., 2009	Between	60	1	4	Healthy	Video	NegAffect	EDA, HR, happiness, NegAffect	NR1	Brief	M2 (ambiguous)	1.50
Ehret et al., 2018	Within	90	3	9	MDD and CN	Mixed	NegAffect	Sadness	NR1	Brief	E3 (clear), E1 (clear), R1 (clear)	1.75
Eifert & Heffner, 2003	Between, brief intervention	40	1	4	Panic symptoms in students	Experiential	Fear	EDA, HR, NegAffect, Sx intensity	NR1	Moderate	E1 (vague)	1.42
Ellard, 2013	Within, dissertation, intervention	21	2	8	GAD, mindfulness trained	Text	Worry	NegAffect	NR4	Brief, intervention	E1 (clear), SM (clear)	1.71
Erisman & Roemer, 2010	Within, brief intervention	15	1	7	Emotion dysregulation	Video	Mixed and NegAffect and amusement	EDA, HR, NegAffect, PA	NR1	Moderate	H1 (clear)	1.83
Evans et al., 2013	Within	42	1	6	BPD, mixed	Video	NegAffect	EDA, NegAffect, PA	NR1	Brief	E1 (vague), SM (vague)	1.82
Fitzpatrick et al., 2020	Within	120	3	18	BPD and GAD and CN	Audio	Rejection	EDA, NegAffect, RSA	NR4	Brief	D3 (clear), H1 (vague)	1.91
Forsyth & Hayes, 2014	Within	37	2	4	General	Experiential	Pain	Pain, pain tolerance	NR1	Moderate	E1 (clear), M1 (clear)	1.67
Gao et al., 2022	Within, intervention post	85	2	8	Healthy, mindfulness trained	Image	NegAffect	Arousal, NegAffect	NR1	Intervention, NA	E3 (clear), E2 (clear)	1.75
Germain & Kangas, 2015	Within	102	3	6	Anger	Experiential	Anger	Anger, BP	NR1	Brief	RM (ambiguous), RM (vague), E1 (vague), E1 (ambiguous), S2 (ambiguous), S2 (vague)	1.78
Godbee & Kangas, 2021	Within	71	2	4	Students, mixed mindfulness	Experiential	Stress	NegAffect, PA	NR1	Extended	MM (vague), RM (vague)	1.83
Goldin & Gross, 2010	Within, intervention	15	1	4	SAD, mindfulness trained	Text	Negative beliefs	NegAffect	NR1	Brief, intervention	D4 (clear), M1 (clear)	1.58
Goldin et al., 2013	Within, intervention	49	2	4	SAD, mindfulness trained	Text	Negative beliefs	NegAffect	NR4	Brief, intervention	M2 (vague)	1.62
Goldin et al., 2019	Within	35	1	8	Healthy	Text	Negative beliefs	EDA, HR, NegAffect, respiration	NR4	Moderate	H1 (clear), R1 (clear)	1.75
Goldin et al., 2021	Within, intervention	68	2	8	SAD, mindfulness trained	Text	Negative beliefs	NegAffect	NR4	Brief, intervention	M2 (ambiguous), R1 (vague)	1.73
Guendelman et al., 2022	Within, intervention	58	2	8	Healthy, mindfulness trained	Image	NegAffect	NegAffect	NR2	Moderate, intervention	E1 (vague), R2 (vague)	1.75
Gutiérrez et al., 2004	Within, brief intervention	40	2	4	Students	Experiential	Pain	Pain, pain tolerance	NR1	Extended	E2 (clear), DM (clear)	1.58
Harrison, 2011	Within, dissertation	24	2	8	OCD	Experiential	Intrusive thoughts	NegAffect, Sx intensity, Sx frequency	NR1	Moderate	E2 (clear), S3 (clear)	1.75
Haspert et al., 2020	Within	30	1	2	General	Experiential	Pain	NegAffect, pain	NR2	Brief	EM (vague)	1.64
Hayes et al., 1999	Within, brief intervention	21	2	4	Students	Experiential	Pain	NegAffect, pain tolerance	NR1	Extended	E2 (vague), DM (vague)	1.46
Hehr et al., 2022	Within	12	1	3	Cancer	Video	NegAffect	NegAffect	NR4	Brief	D3 (clear), E1 (vague), M1 (vague)	1.91
Hessler-Kaufmann et al., 2020	Within	48	1	12	Eating pathology, inpatient	Mixed	NegAffect	Sadness, self-esteem, Urges	NR1	Brief	E3 (clear), MM (clear), R1 (clear)	1.58
Hofman, 2020	Within, dissertation	101	2	4	Students	Experiential	Anger	Anger, Sx intensity	NR1	Brief	R1 (clear), E1 (clear)	1.67
Horowitz, 2009	Within, dissertation, brief intervention	39	2	4	Phobia, students	Experiential	Fear	HR, NegAffect	NR1	Extended	E1 (clear), SM (clear)	1.67
In et al., 2021	Within	61	3	9	NSSI history	Video	NegAffect	Anxiety, NegAffect, Urges	NR1	Brief	D3 (vague), R2 (clear), M2 (unclear)	1.67
Jackson et al., 2012	Within	102	4	8	Students	Experiential	Pain	Pain, pain tolerance	NR1	Brief	DM (vague), E2 (clear)	1.67
Jaén et al., 2021	Within	39	1	6	Healthy students	Experiential	Pain	Anxiety, EDA, HR	NR4	Brief	E1 (vague), R1 (vague)	1.67
James, 2022	Between, dissertation	126	4	12	High and low anxiety students	Experiential	NegAffect	Anxiety, NegAffect, PA	NR1	Brief	E2 (clear), R1 (clear)	1.58
Karl et al., 2018	Between	46	1	2	Remitted MDD	Mixed	Sadness	NegAffect, PA	NR1	Brief	MM (clear)	1.50
Keenan, 2013	Between, dissertation	112	3	3	Students	Video	Sadness	NegAffect	NR1	Brief	S1 (clear), MM (unclear), R2 (clear)	1.67
Kehoe et al., 2014	Within	148	5	5	Healthy students	Experiential	Pain	Pain tolerance	NR1	Moderate	E2 (clear), D1 (clear)	1.45
Keng et al., 2013	Within	66	2	2	Depressive symptoms, mixed mindfulness	Mixed	Sadness	Sadness	NR1	Moderate	H1 (clear), RM (clear)	1.67
Keng & Tan, 2017	Between, brief intervention	71	2	2	BPD traits, mixed mindfulness	Experiential	Shame	Shame	NR1	Moderate	M1 (clear), E3 (clear)	1.75
Keng et al., 2017	Within	125	3	3	Students, mixed mindfulness	Mixed	Sadness	NegAffect	NR4	Moderate	H1 (vague), RM (vague), SM (clear)	1.50
Keng & Ang, 2019	Between, Brief intervention	55	1	3	Eating pathology, mixed mindfulness	Mixed	NegAffect	NegAffect, Sx intensity, Urges	NR1	Moderate	M1 (vague)	1.58
Kleinstäuber et al., 2019	Within	96	2	32	Somatic disorders and CN	Audio	Somatic	NegAffect, strength, Sx intensity, worry	NR1	Brief	DM (clear), E3 (clear), H1 (clear), RM (clear)	1.73
Knight & Emery, 2022	Within	56	2	16	High and low anxiety	Image	NegAffect and PA	Arousal, NegAffect, PA	NR4	Moderate	H1 (clear), S1 (clear)	1.64
Kober et al., 2019	Within	17	1	2	Healthy	Image	NegAffect and pain	NegAffect, pain	NR4	Brief	E1 (clear)	1.67
Kohl et al., 2013	Within	109	3	6	Students	Experiential	Pain	Pain, pain tolerance	NR1	Moderate	E2 (clear), D2 (clear), R1 (clear)	1.75
Konstantinou et al., 2023	Within	44	2	10	Students	Experiential	Pain	EDA, HRV, NegAffect, pain, pain tolerance	NR1	Moderate	E1 (clear), DM (vague)	1.83
Kross et al., 2009	Within	24	1	2	General	Text	NegAffect	NegAffect	NR4	Brief	E2 (clear), R2 (vague)	1.55
Kuo et al., 2016	Within	55	2	16	BPD and CN	Image	NegAffect	EDA, HR, NegAffect, PA	NR4	Brief	D3 (vague), H1 (vague)	1.73
Lalot et al., 2014	Within	45	1	3	General	Video	PA	PA	NR1	Brief	E2 (clear), R2 (clear), S1 (clear)	1.50
Lebois et al., 2015	Within	30	1	1	Students	Text	NegAffect	NegAffect	NR4	Extended	E2 (clear)	1.45
Levitt et al., 2004	Between, brief intervention	56	2	8	Panic symptoms	Experiential	Fear	Anxiety, HR, Sx intensity, temperature	NR1	Moderate	E2 (vague), SM (clear)	1.42
Lohani, 2014	Within, dissertation	120	2	12	Healthy students	Video	Sadness	EMG, HR, sadness	NR4	Brief	M2 (unclear), S1 (clear)	1.73
Low et al., 2008	Between, brief intervention	55	1	3	Students	Experiential	NegAffect	HR, NegAffect, PA	NR1	Brief	E1 (clear)	1.50
Martinez, 2022	Within, dissertation	54	2	8	Healthy	Image	NegAffect	Arousal, HR, NegAffect	NR4	Moderate	R2 (vague), E1 (ambiguous)	1.67
Masedo & Rosa Esteve, 2007	Within	146	2	6	Students	Experiential	Pain	NegAffect, pain, pain tolerance	NR1	Extended	EM (vague), S3 (clear)	1.67
McMullen et al., 2008	Within	64	4	8	Students	Experiential	Pain	Pain, pain tolerance	NR1	Extended	E2 (clear), D1 (clear), E2 (unclear), D1 (vague)	1.75
Moses, 2011	Between, dissertation	81	2	4	Anxiety	Experiential	Anxiety	EDA, NegAffect	NR1	Moderate	EM (clear), R1 (clear)	1.54
Murakami et al., 2015	Within	21	1	2	Healthy	Image	NegAffect	NegAffect	NR4	Brief	MM (clear), S2 (clear)	1.50
Najmi et al., 2009	Within	20	1	6	OCD	Experiential	Intrusive thoughts	NegAffect, Sx frequency	NR1	Brief	D1 (clear), E2 (clear), S3 (clear)	1.64
Opoka et al., 2021	Within	119	2	36	SZ and CN	Image	Anxiety and sadness	EDA, HR, sadness	NR2	Moderate	DM (clear), E1 (ambiguous), RM (clear)	1.92
Ortner, 2015	Between	120	2	4	Healthy students	Image	NegAffect	EDA, NegAffect	NR1	Brief	M2 (clear), R1 (clear)	1.58
Páez-Blarrina et al., 2008a, 2008b	Between	30	2	8	Students	Experiential	Pain	Pain, pain tolerance	NR1	Moderate, extended	EM (vague), E2 (clear), R1 (vague), SM (clear)	1.42
Páez-Blarrina et al., 2008a, 2008b	Within	20	2	4	Students	Experiential	Pain	Pain, pain tolerance	NR1	Extended	E2 (clear), S3 (clear)	1.33
Perry et al., 2012	Within	49	2	6	SZ and CN	Video	Sadness	Sadness	NR1	Brief	M2 (unclear), R2 (clear), S1 (vague)	1.64
Polizzi & Lynn, 2023	Between	100	3	9	Students	Video	Fear	Anxiety, NegAffect, PA	NR1	Brief	E1 (clear)	1.67
Predatu et al., 2020	Between	161	2	8	Students	Video	NegAffect	EDA, HR, NegAffect, respiration	NR1	Moderate	E1 (ambiguous), R1 (vague)	1.67
Prefit et al., 2020	Between	105	2	6	Healthy students	Image	Body dissatisfaction	NegAffect, PA, Sx intensity	NR1	Brief	E2 (unclear), R2 (vague)	1.58
Roche et al., 2007	Within	20	4	12	Students	Experiential	Pain	NegAffect, pain, pain tolerance	NR1	Brief	E1 (vague), O (vague)	1.33
Rompilla et al., 2022	Within	129	1	9	Healthy	Video	Grief	Arousal, NegAffect	NR4	Brief	E1 (clear), R2 (clear), R1 (clear)	1.92
Rosenbaum et al., 2020	Between, brief intervention	69	1	1	High and low rumination	Video	NegAffect	NegAffect	NR1	Moderate	H1 (clear)	1.73
Roys et al., 2023	Between	50	1	2	Tobacco use students	Image	NegAffect	NegAffect, Urges	NR1	Moderate	MM (clear)	1.83
Saeedi et al., 2022	Within	57	2	4	Musculoskeletal pain	Experiential	Pain	NegAffect, pain	NR4	Brief	SM (unclear), MM (vague)	1.50
Santos, 2007	Between, dissertation	108	2	10	Students	Experiential	Frustration	EDA, HR, NegAffect, PA, temperature	NR1	Brief	EM (clear), RM (vague)	1.58
See et al., 2023	Within	180	1	2	Students	Text	NegAffect	NegAffect, Urges	NR1	NA	E1 (vague)	1.67
Singer & Dobson, 2007	Between	60	2	2	Current and remitted MDD	Mixed	NegAffect	NegAffect	NR1	Moderate	E2 (unclear), DM (clear)	1.58
Szasz et al., 2016	Between	105	4	4	Students	Experiential	Anger and sadness	Anger, sadness	NR1	Brief	RM (vague), M2 (unclear)	1.50
Troy et al., 2018	Within	134	2	16	High and low stress students	Video	Sadness	EDA, NegAffect, PA, sadness	NR1	Brief	M2 (unclear), RM (clear)	1.64
Tsujimoto et al., 2023	Within	35	1	2	Healthy	Text	Anxiety	Anxiety	NR4	NA	EM (ambiguous), RM (ambiguous)	1.92
Vinci et al., 2014	Between	112	2	4	Alcohol use students	Image	NegAffect	NegAffect, Urges	NR1	Moderate	H1 (clear), O (clear)	1.67
Vishkin et al., 2020	Within	66	1	4	Students	Text	Fear and sadness	NegAffect	NR1	Brief	E1 (vague), R1 (clear)	1.91
Volkaert et al., 2020	Between, brief intervention	217	4	16	Students	Video	NegAffect	Anger, anxiety, happiness, sadness	NR3	Extended	O (clear), DM (vague), E1 (vague), RM (vague)	1.50
Volkaert et al., 2022	Between, brief intervention	52	1	5	High stress	Experiential	NegAffect	Boredom, frustration, HR, HRV, happiness	NR4	Extended	E3 (clear)	1.50
Vowles et al., 2007	Within	49	2	4	Pain	Experiential	Pain	Impairment, pain	NR1	Brief	E2 (vague), O (unclear)	1.50
S.-M. Wang et al., 2016	Between	48	1	3	Panic symptoms	Image	NegAffect	HRV, NegAffect, PA	NR1	Brief	H1 (clear)	1.58
Wang, Vantieghem et al., 2022	Within, intervention post	19	1	2	Healthy	Image	NegAffect	NegAffect	NR2	Intervention	E2 (clear), M1 (vague)	1.73
Wang, Vlemincx et al., 2022	Between	42	2	4	Healthy	Experiential	NegAffect	NegAffect, sleep	NR1	Brief	R1 (clear), M1 (vague)	1.67
Westbrook et al., 2013	Within	54	1	2	Tobacco use	Image	Craving	Craving, NegAffect	NR4	Brief	MM (clear)	1.73
Wilson et al., 2014	Within	50	3	9	Students	Experiential	Stress	Anxiety, EDA, relaxation	NR1	Brief	E2 (unclear), M3 (ambiguous), S3 (clear)	1.50
Wolfe et al., 2023	Between	84	2	8	Healthy students	Experiential	Stress	Anxiety, performance	NR4	NA	H1 (clear), R1 (clear)	1.54
Wolgast et al., 2011	Between	94	2	18	Students	Video	Disgust and fear and sadness	EDA, EMG, NegAffect	NR1	Brief	R2 (clear), E1 (unclear)	1.58
Wooten et al., 2022	Within	200	2	8	History and absence of childhood maltreatment, MTurk	Image	NegAffect and PA	NegAffect, PA	NR4	Brief	E1 (clear), RM (clear)	1.91
Yuan et al., 2014	Between	64	2	4	Healthy students	Experiential	Frustration	Depressed, EDA	NR1	Brief	E2 (vague), S1 (clear)	1.50
Yuan et al., 2019	Within	17	1	2	Students	Experiential	NegAffect and PA	NegAffect	NR1	Brief	E2 (vague)	1.58
Zangri et al., 2024	Between	112	2	12	Students	Experiential	NegAffect	Anxiety, guilt, happiness, sadness, shame, tension	NR4	Moderate	E3 (clear), R1 (clear)	1.83
Zhang et al., 2023	Between	66	1	3	Healthy students	Experiential	NegAffect	Aggression, NegAffect, PA	NR1	Brief	EM (ambiguous)	1.83

CN healthy control, SAD social anxiety disorder, SZ schizophrenia spectrum, MDD major depressive disorder, GAD generalized anxiety disorder, BPD borderline personality disorder, OCD obsessive–compulsive disorder, NSSI non-suicidal self-injury, Mturk Amazon Mechanical Turk platform, HRV heart rate variability, EDA electrodermal activity, Sx syndrome specific symptoms, HR heart rate, EMG electromyography, ECG electrocardiogram (composite of multiple measures), RSA respiratory sinus arrhythmia, BP blood pressure; see main text Table 1 for reactivity and regulation codes

Additional Information

Funding

Research was supported by National Institute of Mental Health grants R61-MH121560 and R21-MH119438 to Dr. Strauss.

Competing interests

The authors declare no competing interests.

Data Availability

Data, codebooks, and analytic code are provided on OSF (osf.io/qv43d/).

Code Availability

Analytic code available on OSF: osf.io/qv43d/.

Author Contribution

IMR conceptualized the project, conducted the search, ran all analyses, and prepared tables and figures. IMR and AMB coded articles. IMR wrote the main manuscript text with substantive edits by GPS and AMB. All authors reviewed the manuscript.

Ethical Approval

Not applicable.

Informed Consent

Not applicable.