Dispositional mindfulness and mental health in Chinese emerging adults: A multilevel model with emotion dysregulation as a mediator.

Using a multilevel model, this study examined emotion dysregulation as a mediator between dispositional mindfulness and mental health among Chinese emerging adults. Participants were 191 Chinese emerging adults (female = 172) between 18 and 27 years old (M = 21.06 years, SD = 2.01 years), who completed a questionnaire that assessed their dispositional mindfulness, emotion dysregulation, and mental health outcomes for three times over 12 months, with a three-month lag between each time point. Within-person analysis revealed that emotion dysregulation mediated between dispositional mindfulness and mental health outcomes, including subjective well-being and symptoms of depression and anxiety. Time was positively associated with emotion dysregulation and negatively associated with symptoms of depression and anxiety. Between-person analysis revealed that emotion dysregulation negatively mediated between dispositional mindfulness and symptoms of depression and anxiety, but not subjective well-being. These findings call attention to within-person versus between-person effects of emotion dysregulation as a mediator between dispositional mindfulness and psychological outcomes, particularly of symptoms of depression and anxiety. Attesting to the relations established in western societies, the relations are also applicable to emerging adults in the Chinese context. Evidence was thus advanced to inform translational research efforts that promote mindfulness and emotion regulation as assets of mental health.

Introduction

Mindfulness is a mental state whereby individuals attend to their cognitive, physical, and emotional experience in the present moment nonjudgmentally [1, 2]. Previous research suggested that mindfulness is associated with mental health outcomes, such as a low level of psychological symptoms and better subjective well-being [3-6]. Moving beyond the simple association between mindfulness and mental health, recent studies have identified several mediating mechanisms, including greater self-esteem [7], greater positive affect, hope, and optimism [8], lower rumination [9], and greater emotion regulation [10, 11], thereby suggesting possible chains of processes between mindfulness and mental health. Among these mediators, emotion regulation has received much theoretical and empirical attention in recent years [12, 13].

Emotion regulation is defined as a process in modulating emotions and emotional responses [14, 15]. According to Gratz and Roemer [16], adaptive emotion regulation involves such aspects as acceptance of emotional responses, engagement in goal directed behavior, maintenance of emotional clarity and awareness, and access to adaptive emotion regulation strategies, such as cognitive reappraisal and an ability to savor positive experiences. On the contrary, emotion dysregulation involves difficulties in acquiring regulatory skills or a frequent use of maladaptive strategies, such as rumination, avoidance, and expressive suppression [17, 18]. Previous research suggested that our capability to regulate emotions is central to subjective well-being [18]. Moreover, emotion dysregulation not only undermines well-being, but also gives rise to clinical depression, anxiety, psychological distress, and elevated depressive and anxiety symptoms [10, 17, 19–21].

Theoretical tenets of emotion regulation as a mediator between mindfulness and mental health outcomes

As a proximal correlate of psychological well-being, people’s capability to regulate emotions is associated with mindfulness [6, 22, 23]. More specifically, Teper et al. [13] theorized that mindfulness is linked to emotion regulation through heightened metacognitive awareness, non-judgmental acceptance, and executive control, all of which are crucial to well-being. Likewise, in mindfulness-to-meaning theory, Garland et al. [12] postulated that mindfulness promotes purpose in life by virtue of greater metacognitive awareness, broadened attention to context, adaptive emotion regulation, ability to savor hedonic experience, and prosocial actions. Both theoretical frameworks suggest close connections between mindfulness, emotion regulation, and psychological well-being.

Supporting the frameworks, cross-sectional studies revealed that emotion regulation mediated between dispositional mindfulness and lower levels of depression and anxiety [11, 23–25] as well as better life satisfaction [26]. Studies based on a clinically depressed and anxious sample similarly suggested that emotion regulation strategies, including cognitive reappraisal and rumination, mediated between dispositional mindfulness and symptoms of depression and anxiety [25]. These findings resonated with research conducted with non-clinical emerging adult samples, in that emotion regulation mediated the inverse relation between dispositional mindfulness and depressive symptoms, in that dispositional mindfulness was related to better emotion regulation and fewer depressive symptoms [23]. In a similar vein, Coffey and Hartman [11] and Coffey et al. [24] found that emerging adults’ regulation of negative emotions, impulse control, and goal engagement mediated between dispositional mindfulness and well-being. Altogether, evidence to-date converged to indicate emotion regulation as a vital process between dispositional mindfulness and well-being.

Processes of mental health in the Chinese context

Among the handful of studies conducted in the Chinese context, dispositional mindfulness predicted better impulse control, better emotion regulation, less procrastination, and less psychological distress [10, 27]. Moreover, mindful awareness was associated with subsequent changes of stress response, emotion regulation, and anxiety symptoms [28, 29]. Changes in emotion regulation strategies were also associated with changes in depressive symptoms, life satisfaction, and general health over time [30]. Consequently, timing and changes were crucial in linking between mindfulness, emotion regulation, and mental health outcomes. Beyond dispositional mindfulness, mindfulness training also increased Chinese adolescents’ and adults’ well-being and psychological adjustment in Hong Kong [31, 32]. In a randomized controlled trial involving a Chinese sample with clinical anxiety, Wong et al. [33] revealed that participants who took an 8-week mindfulness-based cognitive therapy experienced increased dispositional mindfulness and reduced anxiety. In addition, such an increase was found to be more significant in these participants than did participants who received psychoeducation based on cognitive behavioral therapy principles. Similar findings were reported in another randomized controlled trial involving a Chinese sample with recurrent symptoms of depression and anxiety [34]. In the study, participants who received an 8-week compassion–mindfulness therapy showed more mental health improvements than did participants from the waitlist control group. These findings reveal that the mechanisms associated with mindfulness are integral to mental health in the Chinese context.

Within- versus between-person effects

A majority of studies to-date focus on between-person associations of mindfulness, emotion regulation or dysregulation, and mental health [e.g., 5–11]. Only a handful of studies have partitioned within-person from between-person processes. Between- and within-person effects are conceptually and statistically independent. Within-person effect aids the understanding of intraindividual processes underlying well-being. For instance, when mindfulness of person i increases, then his or her own mental health is expected to increase over time. Between-person effect accounts for interindividual processes underlying well-being. That is, when person i has a higher score in mindfulness than person j, then person i is expected to have a higher score in mental health than is person j. Of note, partitioning among within- and between-person associations is important, because otherwise the results could sometimes be biased [35] or uninterpretable [36].

Several studies to-date have demonstrated the utility of within-person analyses in mindfulness and well-being. For example, Galla [37] indicated that a five-day mindfulness training was associated with significant within-person reductions in rumination and depressive symptoms, as well as increases in life satisfaction in healthy adolescents. In addition, many of the improvements were maintained at the three-month follow-up assessment. In another study, an 8-week mindfulness training gave rise to an inverse within-person association between mindfulness and relationship stress [38]. Apart from stress, psychological symptoms, and subjective well-being, another study also showed within-person associations between dispositional mindfulness and relationship satisfaction [39]. By investigating within-person effects, these studies methodologically precluded between-person effects of mindfulness on psychological outcomes.

This study

Aside from the studies described in the previous section, past research predominantly examined between-person effects among mindfulness, emotion regulation, and mental health [5-11]. Although some studies have begun to examine within-person effects [37-39], few, if any, have teased apart within-person from between-person effects. Although within- and between-person findings converged to suggest the mental health benefits of mindfulness, it remains unclear as to whether its significance on emotion regulation and various mental health outcomes differ.

The present study investigated longitudinal within- versus between-person mediating effects of dispositional mindfulness and mental health outcomes via emotion dysregulation. Through this analytic technique, the intraindividual and interindividual processes could be identified. Based on the literature [23-25], we hypothesized that greater dispositional mindfulness would be related to lower emotion dysregulation. Lower emotion dysregulation would, then, be related to fewer symptoms of depression and anxiety, as well as greater subjective well-being. We further hypothesized that the relation between dispositional mindfulness and mental health outcomes would be mediated by emotion dysregulation at both within- and between-person levels. The strength of associations may vary as a function of specific levels. As stated earlier, time was an important measure in assessing changes within and between variables [28, 30]. Consequently, time was included as a within-person covariate to control for dispositional mindfulness, emotion dysregulation, and mental health outcomes. Based on previous research showing the links between gender and depressive symptoms [40], anxiety symptoms [41], and subjective well-being [42], gender was included as a between-person covariate for mental health outcomes.

Method

Participants

Participants were 191 Chinese emerging adults (female = 172; 90.05%) recruited online through two mass emails at a major public university in Hong Kong. Inclusion criteria included university-enrolled emerging adults who were proficient in Chinese, within the age range between 18 and 29 years old [43] and agreed to participate for three time points over the course of a year. Participants were between 18 and 27 years old, with a mean age of 21.06 years at Time 1 (SD = 2.01 years). As for the retention rate, 93.72% (n = 179) of the participants from Time 1 (T1) were retained at Time 2 (T2); 94.41% (n = 169) of the participants from T2 were retained at Time 3 (T3).

Procedures

The study was approved by the Human Research Ethics Committee of The Education University of Hong Kong (Approval #: 2015-2016-0352) prior to its implementation. All procedures performed were in accordance with the ethical standards of the institutional research committee and the 1964 Helsinki declaration and its later amendments. A written consent was obtained from each participant prior to the beginning of the study. At the first time point, participants completed the baseline measures. They were then invited to complete the follow-up questionnaires twice, with a three-month lag between time points, over the course of a year. Each packet took approximately 30 minutes to complete. Upon completion, participants received a supermarket coupon at each time point (totaling HK$200, i.e., ~US$25.71).

Measures

Dispositional mindfulness

The 12-item Cognitive and Affective Mindfulness Scale–Revised (CAMS-R) [44] was used to assess dispositional mindfulness on a 4-point scale from 1 (rarely/ not at all) to 4 (almost always). The CAMS-R assessed three dimensions of mindfulness including awareness, acceptance, and attention. A sample item included, “I can usually describe how I feel at the moment in considerable detail.” The item scores were averaged, such that higher scores indicated a greater level of mindfulness. The measure was validated previously in a Chinese college sample with adequate convergent validity, reliability, and factor structure [45]. Specifically, participants from this study had a similar mean level of dispositional mindfulness (see Table 1). In this study, the test-retest correlations of the measure between T1, T2, and T3 were moderate at .53—.58, ps < .001. Cronbach’s alpha of the measure = .91—.94 between T1 and T3. More specifically, between T1 and T3, Cronbach’s alpha of the awareness subscale = .63—.80, the acceptance subscale = .62—.79, and the attention subscale = .77—.81.

Table 1

Descriptive statistics of the variables under study (N = 191).

Variable	M	SD	Min	Max	Range of Scale	Intraclass Correlation	Design Effect
(1) Gender (0 = male; 1 = female)	-	-
Time 1
(2) Dispositional mindfulness	2.85	.43	1.80	4.00	1.00–4.00
(3) Emotion dysregulation	2.03	.49	1.15	3.66	1.00–5.00
(4) Depressive symptoms	.61	.50	.00	3.00	.00–3.00
(5) Anxiety symptoms	.61	.58	.00	3.00	.00–3.00
(6) Subjective well-being	4.49	.86	.79	5.00	.00–5.00
Time 2
(7) Dispositional mindfulness	2.80	.48	1.60	4.00	1.00–4.00
(8) Emotion dysregulation	2.12	.51	1.17	3.47	1.00–5.00
(9) Depressive symptoms	.68	.56	.00	3.00	.00–3.00
(10) Anxiety symptoms	.65	.64	.00	3.00	.00–3.00
(11) Subjective well-being	4.28	.86	.50	5.00	.00–5.00
Time 3
(12) Dispositional mindfulness	2.78	.52	1.20	4.00	1.00–4.00
(13) Emotion dysregulation	2.12	.54	1.16	3.93	1.00–5.00
(14) Depressive symptoms	.74	.61	.00	2.89	.00–3.00
(15) Anxiety symptoms	.81	.68	.00	3.00	.00–3.00
(16) Subjective well-being	4.15	.92	.14	5.00	.00–5.00
Across Time Points
Dispositional mindfulness	2.81	.48			1.00–4.00	.56***	2.01
Emotion dysregulation	2.09	.51			1.00–5.00	.73***	2.31
Depressive symptoms	.68	.56			.00–3.00	.64***	2.13
Anxiety symptoms	.68	.63			.00–3.00	.51***	1.93
Subjective well-being	4.32	.89			.00–5.00	.63***	2.14

Note:

***p < .001. Design effect was defined as 1 + (average cluster size—1) × intraclass correlation.

Emotion dysregulation

The 36-item Difficulties in Emotional Regulation Scale (DERS) [16] was used to assess emotional regulation on a 5-point scale from 1 (almost never) to 5 (almost always). The measure included 6 subscales including non-acceptance of emotions, difficulties in engaging in goal-directed behavior, impulse control difficulties, lack of emotional awareness, limited access to emotion regulation strategies, and lack of emotional clarity. Sample items included, “I am clear about my feelings” and “When I’m upset, I believe that my feelings are valid and important.” The final scores were then averaged, such that higher scores indicated a greater emotion dysregulation. The measure had been validated in a Chinese sample [46]. Compared to another Chinese adult sample [47], participants from this study had a similar mean level of emotion dysregulation (see Table 1). In this study, the test-retest correlations of the measure between T1, T2, and T3 were .70—.76, ps < .001. Cronbach’s alpha of the measure = .83—.87 between T1 and T3. More specifically, between T1 and T3, Cronbach’s alpha of the non-acceptance of emotions subscale = .83—.87, the difficulties in engaging in goal-directed behavior subscale = .56—.75, the impulse control difficulties subscale = .77—.85, the lack of emotional awareness subscale = .72—.75, the limited access to emotion regulation strategies subscales = .83—.87, and the lack of emotional clarity subscale = .74—.77.

Depressive symptoms

The 9-item Patient Health Questionnaire-9 (PHQ-9) [48] was used to measure depressive symptoms in the past two weeks on a 4-point scale from 0 (not at all) to 3 (nearly every day). A sample item included, “Feeling bad about yourself—or that you are a failure or have let yourself or your family down?” The item scores were averaged to form a score of depressive symptoms, with higher scores indicating more symptoms. The measure has previously been validated in a Chinese community sample and demonstrated an adequate factor structure, construct validity, internal consistency, and test-retest reliability [49]. Compared to a cutoff score of 11 for the detection of depressive disorders [50], upon rescaling our mean scores to summed scores, participants’ scores of 5.44–6.67 between T1 and T3 were below the clinical cutoff. Compared to another Chinese college sample [51] participants from this study had a comparable level of depressive symptoms (see Table 1). However, a total of 20, 21, and 28 participants at T1, T2, and T3, respectively, reported a summed score of above 11 (i.e., above the cutoff). In this study, the test-retest correlations of the measure between T1, T2, and T3 were .62—.67, ps < .001. Cronbach’s alpha of the measure = .87—.91 between T1 and T3.

Anxiety symptoms

The 7-item Generalized Anxiety Disorder-7 measure (GAD-7) [52] was used to measure anxiety symptoms in the past two weeks on a 4-point scale from 0 (not at all) to 3 (nearly every day). A sample item included, “Feeling nervous, anxious or on edge.” The item scores were averaged to form a score of anxiety symptoms, with higher scores indicating more symptoms. The measure has been validated in a Chinese sample of hospital outpatients [53] and a sample of Chinese individuals with epilepsy [54]. Compared to a cutoff score of 10 for the detection of generalized anxiety disorder [52], upon rescaling our mean scores to summed scores, participants’ scores of 4.22–5.67 between T1 and T3 were below the clinical cutoff. Compared to another Chinese college sample [51] participants from this study reported a similar level of anxiety symptoms (see Table 1). However, a total of 14, 20, and 26 participants at T1, T2, and T3, respectively, reported a total score of above 10 (i.e., above the cutoff). In this study, the test-retest correlations of the measure between T1, T2, and T3 were .52—.56, ps < .001. Cronbach’s alpha of the measure = 93—.95 between T1 and T3.

Subjective well-being

The 14-item Mental Health Continuum Short Form (MHC-SF) [55] was used as a measure for well-being over the past four weeks. A 6-point scale ranging from 0 (never) to 5 (every day) was used, with sample items including, “how often did you feel satisfied with life” (emotional well-being), “how often did you feel that that you had something important to contribute to society (social well-being) and “how often did you feel that your life has a sense of direction or meaning to it” (psychological well-being). The item scores were averaged to form a mean score, with higher scores indicating better well-being. MHC-SF had been previously validated in a Chinese adolescent sample and yielded good validity and reliability [56]. Compared to another Chinese college sample [57], participants from this study had a greater mean level of subjective well-being (see Table 1). In this study, the test-retest correlations of the measure between T1, T2, and T3 were .59—.69, ps < .001. Cronbach’s alpha of the measure = .79—.94 between T1 and T3. More specifically, between T1 and T3, Cronbach’s alpha of the emotional well-being subscale = .90—.94, the social well-being subscale = .79—.86, and the psychological well-being subscale = .92—.93.

Data analysis

Multilevel mediation analyses were conducted using Mplus 8.0 [58] to examine the mediating effect of emotion dysregulation on the relationship between dispositional mindfulness and mental health outcomes, including depressive symptoms, anxiety symptoms, and subjective well-being, at both within- and between-person levels. To illustrate the necessity of multilevel modeling, we computed the intraclass correlations (ICC), which quantified the proportion of variance of the variables attributable to individual differences, as well as the design effect, defined as 1 + (average cluster size—1) × ICC [59]. Multilevel modeling has been used in previous research for panel data (e.g., [60]), with a minimum of three waves of observations to partition between within- and between-person associations, which, as introduced earlier, have distinctive practical implications. Compared to the traditional multilevel modeling approach to multilevel mediation, the multilevel structural equation modeling (MSEM) approach has the advantage of having a smaller bias.

In this study, we relied on the MSEM approach. Covariates at both levels were considered. At the within-person level, a time variable was included to control for the potential changes over time [61]. At the between-person level, gender was included to control for possible gender differences in symptoms of anxiety and depression. Regarding whether random slopes should be considered, with only three measurement occasions at the within-person level, the model allowing all mediation paths (i.e., “a” path from mindfulness to emotion dysregulation, “b” paths from mindfulness to mental health outcomes, and “c” paths from emotion dysregulation to mental health outcomes) and the slope of time to be random could not be identified. Therefore, we tested possible random mediation paths sequentially. Results showed that none of the “a”, “b”, and “c′” paths showed substantial individual differences, except for the “c′” path for depression. Additionally, we tested possible random slopes of time (i.e., change between two consecutive measurement occasions) and found no significant individual differences. Consequently, all mediation paths and slopes of time were fixed to be invariant across participants, except for the “c′” path for depression in the subsequent analyses.

In sum, the multilevel mediation model under consideration was as follows. The independent variable, the mediator, and the three dependent variables were partitioned into two latent parts:

More specifically, the level-1 model defined the within-person mediation model. Between the independent variable and the mediator (the fixed “a” path), the equation was:

From the mediator to the dependent variables (the fixed “b” paths), the equations were:

Here, η was the latent within-person component of variable for person i at time t. The coefficient α was the within-person association between the independent variable and the mediator. The coefficients b, b, and b quantified the within-person relationships between the mediator and the three mental health outcome variables, i.e., depression symptoms, anxiety symptoms, and subjective well-being, respectively. The coefficients , and represented the relationships between the independent variable and the three dependent variables, respectively, after controlling for the effect of the mediator. The random slope was assumed to follow a normal distribution with a mean of and a variance of .

The level-2 model was the between-person mediation model. Specifically, from the independent variable to the mediator (the “a” path at the between-person level), the equation was:

From the mediator to the dependent variables (the “b” paths at the between-person level), the equations were:

Similar to the level-1 model, η was the latent between-person component of a variable for person i. The coefficient α was the between-person association between the independent variable and the mediator. The coefficients b, b, and b quantified the between-person relationships between the mediator and the three mental health outcome variables, respectively. The coefficients , and represented the between-person relationships between the independent variable and the three dependent variables, respectively, after controlling for the effect of the mediator. Following the traditional steps of mediation analysis [62], we fitted a multilevel mediation model without the mediator to study the overall effect of mindfulness on the outcomes. The model was estimated under missing data theory using all available data [58, ch.9]. Given the limitations listed by Hayes [63], the ratio of the indirect effect to the total effect was not calculated in determining the strength of the indirect effect. That is, this ratio might be out of the expected range between 0 and 1. Also, the estimate of the ratio might be highly unstable from sample to sample. Unless the sample size was fairly large, Hayes [63, p.189] recommended not “having much faith” in this measure.

Results

Table 1 summarizes the variable means, SDs, ranges of the scales, minima, maxima, ICCs, and design effects. Table 2 summarizes the between-person correlations among the variables under study. Specifically, dispositional mindfulness, emotion dysregulation, and mental health outcomes were correlated with each other at ps < .05 over time, ranging widely from small to large effect sizes [64]. Gender was associated with T2 and T3 dispositional mindfulness, as well as T3 depressive symptoms, anxiety symptoms, and subjective well-being, ps < .05. The variables showed substantial within-person correlations, which were captured by the random intercepts in the multilevel model. Changes in means over time were observed, suggesting the necessity of including time as a covariate.

Table 2

Zero-order correlations of the variables under study (N = 191).

Variable

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

(14)

(15)

(16)

(1) Gender (0 = male; 1 = female)

-

Time 1

(2) Dispositional mindfulness

.07

-

(3) Emotion dysregulation

-.09

-.55***

-

(4) Depressive symptoms

-.02

-.48***

.69***

-

(5) Anxiety symptoms

-.11

-.42***

.65***

.77***

-

(6) Subjective well-being

.11

.64***

-.58***

-.65***

-.56***

-

Time 2

(7) Dispositional mindfulness

.19*

.58***

-.54***

-.41***

-.37***

.57***

-

(8) Emotion dysregulation

-.03

-.42***

.76***

.63***

.59***

-.49***

-.62***

-

(9) Depressive symptoms

-.05

-.23**

.49***

.67***

.57***

-.41***

-.47***

.63***

-

(10) Anxiety symptoms

-.09

-.21**

.39***

.51***

.52***

-.43***

-.47***

.52***

.73***

-

(11) Subjective well-being

.13

.46***

-.54***

-.51***

-.40***

.68***

.70***

-.62***

-.52***

-.52***

-

Time 3

(12) Dispositional mindfulness

.17*

.53***

-.47***

-.42***

-.38***

.47***

.58***

-.54***

-.44***

-.38***

.51***

-

(13) Emotion dysregulation

-.12

-.42***

.70***

.55***

.48***

-.44***

-.49***

.75***

.52**

.38**

-.53***

-.65***

-

(14) Depressive symptoms

-.16*

-.45***

.53***

.62***

.57***

-.44***

-.44***

.56***

.67***

.54***

-.47***

-.61**

.68***

-

(15) Anxiety symptoms

-.16*

-.39***

.44***

.53***

.54***

-.42***

-.36***

.48***

.56***

.56**

-.44***

-.49***

.59**

.80**

-

(16) Subjective well-being

.20*

.50***

-.43***

-.45***

-.38***

.59***

.55***

-.47***

-.46***

-.48***

.69***

.71***

-.56***

-.64***

-.60***

-

*p < .05,

**p < .01,

***p < .001.

Multilevel structural equation modeling

The ICCs of the variables under study ranged from .51 (anxiety symptoms) to .73 (emotion dysregulation), ps < .001. This indicated substantial between-person variances, which should be considered through multilevel modeling. The design effect was larger than two for any of the studied variables, except for anxiety symptoms, suggesting that the hierarchical structure of the data should be considered [59]. Based on the R-squared measures for multilevel models proposed by Rights and Sterba [65], we computed the proportion of within-cluster outcome variance explained by level-1 predictors via fixed slopes and random slope variation (i.e., in [65]) and the proportion of between-cluster outcome variance explained by level-2 predictors via fixed slopes (i.e., in [65]) for each of the three outcome variables. Specifically, dispositional mindfulness, emotion dysregulation, and time explained 23% of the within-person variances via fixed slopes and random slope variation for depressive symptoms, and 18% and 30% of the within-person variances via fixed slopes for anxiety symptoms and subjective well-being, respectively. At the between-person level, dispositional mindfulness, emotion dysregulation and gender explained 31% and 42% of the between-person variances for depression and anxiety symptoms, respectively. Dispositional mindfulness and emotion dysregulation explained 67% of the between-person variances for subjective well-being (see Fig 1 for details on path coefficients).

Fig 1

Multilevel mediation model of dispositional mindfulness, emotion dysregulation, and mental health outcomes.

*p = /< .05, **p = /< .01, ***p = /< .001. Unstandardized parameter estimates and standard errors in parentheses are presented. For simplicity, the random slope of depressive symptoms regressed on dispositional mindfulness is not depicted.

Within-person indirect effect of dispositional mindfulness

The effect of dispositional mindfulness on depressive symptoms was negatively mediated by emotion regulation, , p < .001, 95% CI = [-.18, -.06], with a significant mean direct effect, , p < .01, 95% CI = [-.30, -.05]. Specifically, dispositional mindfulness predicted lower emotion dysregulation, , p < .001, 95% CI = [-.41, -.21]. Emotion dysregulation further predicted greater depressive symptoms, , p < .001, 95% CI = [.26, .54]. The results regarding anxiety symptoms showed similar patterns (see Table 3). Specifically, emotion dysregulation inversely predicted by greater dispositional mindfulness was associated with greater anxiety symptoms, , p < .001, 95% CI = [-.23, -.07], suggesting a partial mediation effect, , p = .01, 95% CI = [-.34, -.04].

Table 3

Unstandardized parameter estimates and standard errors of the multilevel model.

Parameter	Unstandardized B (SE)
Within-Person Effect
Time
→ Dispositional mindfulness	-.03 (.02)
→ Emotion dysregulation	.04 (.01)**
→ Depressive symptoms	.04 (.02)*
→ Anxiety symptoms	.07 (.02)**
→ Subjective Well-being	-.12 (.03)***
Dispositional mindfulness
→ Emotion dysregulation	-.31 (.05)***
→ Depressive symptoms (mean, cw10′)	-.17 (.06)**
→ Anxiety symptoms	-.19 (.08)*
→ Subjective Well-being	.60 (.10)***
Emotion dysregulation
→ Depressive symptoms	.40 (.07)***
→ Anxiety symptoms	.49 (.10)***
→ Subjective well-being	-.43 (.12)***
Depressive symptoms ←→ Anxiety symptoms	.06 (.01)***
Depressive symptoms ←→ Subjective Well-being	-.04 (.01)***
Anxiety symptoms ←→ Subjective Well-being	-.04 (.01)**
Between-Person Effect
Gender (0 = male; 1 = female)
→ Depressive symptoms	.03 (.09)
→ Anxiety symptoms	-.07 (.11)
Dispositional mindfulness
→ Emotion dysregulation	-.91 (.09)***
→ Depressive symptoms	.46 (.55)
→ Anxiety symptoms	-.23 (.16)
→ Subjective Well-being	1.44 (.19)***
Emotion dysregulation
→ Depressive symptoms	1.31 (.48)**
→ Anxiety symptoms	.62 (.14)***
→ Subjective Well-being	-.24 (.14)
Depressive symptoms ←→ Anxiety symptoms	.11 (.05)*
Depressive symptoms ←→ Subjective well-being	.01 (.05)
Anxiety symptoms ←→ Subjective well-being	-.05 (.02)**
Depressive symptoms ←→ cw1i′	-.14 (.06)*
Anxiety symptoms ←→ cw1i′	-.02 (.02)
Subjective well-being ←→ cw1i′	-.01 (.02)
Dispositional mindfulness ←→ cw1i′	.01 (.01)
Emotion dysregulation ←→ cw1i′	-.02 (.02)
cw1i′ ←→ cw1i′	.05 (.02)*

*p < .05,

**p < .01,

***p < .001. The random coefficient represents the within-person relationship between dispositional mindfulness and depressive symptoms, after controlling for the effect of emotion dysregulation. is the mean of across participants.

Regarding the indirect effect of dispositional mindfulness on subjective well-being through emotion dysregulation, we found a partial positive mediation effect, , p = .002, 95% CI = [.05, .22]; , p < .001, 95% CI = [.41, .79]. Specifically, dispositional mindfulness was negatively associated with emotion dysregulation, which in turn predicted worse subjective well-being, , p < .001, 95% CI = [-0.20, -.66] (see Table 4).

Table 4

Results of multilevel mediation analyses concerning the mediating effect of emotion dysregulation between dispositional mindfulness and mental health outcomes.

Outcome	a	b	c′	ab	c
			Within-Person
Depression		.40[.26,.54] *	-.17[-.30,-.05] *	-.12[-.18,-.06] *	-.29[-.43,-.16] *
Anxiety	-.31[-.41,-.21]*	.49[.29,.69] *	-.19[-.34,-.04] *	-.15[-.23,-.07] *	-.34[-.49,-.19] *
Subjective well-being		-.43[-.66,-.20] *	.60[.41,.79] *	.13[.05,.22] *	.73[.52,.94] *
			Between-Person
Depression		1.31[.36,2.26] *	.46[-.63,1.54]	-1.19[-2.01,-.38] *	-.74[-1.35,-.13] *
Anxiety	-.91[-1.09,.74] *	.62[.35,.89] *	-.23[-.54,.08]	-.57[-.85,-.29] *	-.80[-1.01,-.59] *
Subjective well-being		-.24[-.50,.03]	1.44[1.07,1.82] *	.22[-.02,.45]	1.66[1.40,1.93] *

The numbers with asterisk (*) indicate significant results. Numbers within brackets are the lower and upper limits of confidence intervals. The coefficients a, b, c, and c′ represent the paths from dispositional mindfulness to emotion dysregulation, from emotion dysregulation to mental health outcomes, and from dispositional mindfulness to mental health outcomes before and after controlling for the effect of emotion dysregulation. The coefficient ab represents the indirect effect of emotion dysregulation. Note that the within-person c′ and c paths for depression was the average of c′ and c paths across participants. The upper panel reports the results regarding the within-person level whereas the lower displays those for the between-person level.

Between-person indirect effect of dispositional mindfulness

The effects of dispositional mindfulness on symptoms of depression and anxiety were found to be negatively mediated by emotion dysregulation (see Table 3). Notably, full mediation effects were found at the between-person level, as indicated by the nonsignificant direct effects of dispositional mindfulness on symptoms of depression, , p = .41, 95% CI = [-.63, 1.54] and anxiety, , p = .15, 95% CI = [-.54,.08]. Greater dispositional mindfulness was associated with lower emotion dysregulation. Emotion dysregulation, in turn, was associated with greater symptoms of depression and anxiety.

The between-person indirect effect of dispositional mindfulness on subjective well-being was not significant, , p = .07, 95% CI = [-.02, .45]. Although the “a” path between mindfulness and emotion dysregulation was significant, the “b” path between emotion dysregulation and well-being was not, , p = .08, 95% CI = [-.50, .03] (see Table 4).

Discussion

Building on existing theories on mindfulness, emotion regulation, and well-being [12, 13], this study supported emotion dysregulation as a mediator between dispositional mindfulness and mental health outcomes in university-enrolled Chinese emerging adults. Findings based on multi-level modeling further suggested within- and between-person nuances for the associations (see Fig 1). At the within-person level, emotion dysregulation mediated between dispositional mindfulness and all mental health outcomes. Surprisingly, time was associated with worse emotion dysregulation and well-being, suggesting that psychological functioning worsened as a function of time, potentially due to an increasing level of academic stress over the school year. At the between-person level, dispositional mindfulness negatively predicted emotion dysregulation and positively predicted subjective well-being, whereas emotion dysregulation positively predicted symptoms of depression and anxiety, but not subjective well-being. These findings advanced the field by establishing within- and between-person relations between dispositional mindfulness and mental health outcomes via emotion dysregulation.

The present study used multilevel modeling by partitioning between-person relations from within-person relations. Importantly, the variables were related in the hypothesized directions within-person, suggesting emotion dysregulation as a central intraindividual process linking dispositional mindfulness and mental health outcomes. By orienting to the present moment and paying attention on purpose and non-judgmentally [66], individuals became more mindful and were more capable of decentering themselves from subjective emotional experience [67], less preoccupied with invalidating, avoiding, or rejecting emotional experiences [68], more likely to disengage themselves from autopilot [69], and more likely to engage in adaptive emotion regulation [12, 13]. Based on these findings, practitioners could tailor their practices and interventions to the progress of each client. For example, to enhance subjective well-being and reduce psychological symptoms, practitioners can enhance within-person improvements in mindfulness and emotion regulation, as mindfulness affects mental health outcomes both directly and through adaptive emotion regulation.

Turning to the between-person findings, dispositional mindfulness was directly associated with emotion dysregulation and subjective well-being, but not with symptoms of depression and anxiety. These findings signified that when person i had a higher score in dispositional mindfulness than person j, then person i was also more likely to have a lower score in emotion dysregulation and a higher score in subjective well-being than person j. Despite a lack of a direct relation between dispositional mindfulness and psychological symptoms, as an explanatory variable emotion dysregulation accounted for the between-person relation. The findings resonated with recent findings conducted in the Chinese context, in that emotion dysregulation mediated between dispositional mindfulness and symptoms of depression and anxiety [10]. Surprisingly, emotion dysregulation did not hold up as a between-person mediator between dispositional mindfulness and subjective well-being. In fact, emotion dysregulation was not related to subjective well-being at all, after controlling for the effect of dispositional mindfulness. These findings contradicted recent cross-sectional studies [30, 70] showing the link between Chinese college students’ emotion regulation strategies and life satisfaction (i.e., a major component of subjective well-being [71]). Instead of a general dysregulation of emotions, perhaps nuances such as emotion regulation strategies (e.g., cognitive reappraisal) [30], regulatory flexibility [72], and affective states, frequency, and intensity [73] are more proximal predictors of subjective well-being. Alternatively, perhaps over time, emotion dysregulation is more strongly associated with negative outcomes (e.g., psychological symptoms) than with positive outcomes (e.g., positive well-being and satisfaction with life), after controlling for the effect of dispositional mindfulness. Furthermore, stage-salient variables in emerging adulthood, such as achieving financial independence, having greater responsibility, and having greater commitment in romantic relationships [74], might be more strongly associated with positive well-being. Consequently, future studies should replicate the present findings and investigate other longitudinal predictors of subjective well-being in the Chinese context. In terms of clinical implications, practitioners should be made aware of their clients’ levels of mindfulness, emotion dysregulation, and mental health relative to other people. For example, returning to Fig 1, in relation to person j, person i’s greater level of dispositional mindfulness was associated with a lower level of emotion dysregulation and better well-being than person j, but not directly with fewer symptoms of depression and anxiety. Understanding these associations could help practitioners gauge the between-person importance or relevance of mindfulness in various mental health outcomes. Finally, gender was not related to psychological symptoms in the model. However, our small sample of men (i.e., 9.95%) precluded us from drawing meaningful conclusions about gender as a correlate of psychological symptoms.

Time was related to greater emotion dysregulation, greater symptoms of depression and anxiety, and worse subjective well-being. Contrary to previous research suggesting that individuals were more capable of emotion regulation as they mature [75], our findings suggested the opposite. It might be that our short-term longitudinal data were insufficient to capture a positive development in emotion regulation. Hence, a long-term longitudinal approach spanning over several years may be necessary to verify whether this is the case. Alternatively, given the participants were emerging adults enrolled in a university, they may be facing increasing stress as they approached the end of the semester or graduation. Future work is needed to rule out confounding variables, such as academic stress and sample characteristics, associated with time and mental health [76].

Moving onto the mean levels, the current university participants reported a similar average of dispositional mindfulness, emotion dysregulation, and symptoms of depression and anxiety compared to others Chinese samples [45, 47, 51]. Paradoxically, our participants also reported a greater level of emotional, psychological, and social well-being than did another college sample [57]. Again, future studies should replicate the mean levels as well as the strength of associations in other Chinese samples.

Limitations and future directions

The study has several limitations. First, we utilized self-report measures. Future researchers may adopt a multi-method approach by incorporating physiological and neural measures. Next, given this sample comprised mainly of emerging adults at a university, it is uncertain whether the findings can be generalized to other contexts. Future studies should examine the relation in a more representative sample from the community. In addition, the timeframes of the mental health questionnaires differ. For example, the PHQ-9 assessed depressive symptoms in the past two weeks, whereas the MHC-SF assessed emotional, social, and psychological well-being in the past four weeks. Future research may consider standardizing the timeframes to increase precision of the variables. Furthermore, our participants were mainly female. Future studies with gender-balanced samples are necessary to draw meaningful conclusions for the effects of gender. Finally, researchers should translate the present findings and design targeted interventions to improve mental health outcomes.

Conclusion

This study calls attention to the relation between dispositional mindfulness and mental health outcomes through emotion dysregulation. Taking into account the effect of time, findings based on multilevel modeling demonstrated differential effects at the within- and the between-person levels, thereby suggesting a need to partition these levels in future research. In addition, our findings converged to underscore the association between dispositional mindfulness and different aspects of mental health in emerging adulthood, including symptoms of depression and anxiety, as well as emotional, social, and psychological well-being. Psychological intervention programs and public health campaigns geared toward enhancing mindfulness and emotion regulation merit future research investigations.

(DAT)

Click here for additional data file.

30 Apr 2020

PONE-D-20-06030

Mindful Awareness and Mental Health: A Multilevel Model with Emotion Regulation as a Mediator

PLOS ONE

Dear Dr Cheung,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

We would appreciate receiving your revised manuscript by Jun 14 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Therese van Amelsvoort

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

Additional Editor Comments (if provided):

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: Yes

Reviewer #3: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you for the interesting manuscript, studying emotion regulation as a mediator between mindfulness and mental health in the Chinese young adult population. I am happy to read that both the within- and between subject factors are taken into consideration in this study design.

A general remark relates to the umbrella term ‘mental health’ used in the research question, this remains rather broad. The introduction includes a description, but throughout it is not always clear how mental health is defined with respect to design and outcomes. The separate measures (depression, anxiety, and subjective wellbeing) are sometimes interchanged with the broader mental health concept. Does this then refer to a composite measure of these questionnaires?

A very high percentage of the study population is female (90%), this is correctly mentioned as a limitation at the end of the manuscript, but briefly. Gender is taken as a covariate in the analyses and conclusions are drawn on these analyses in the beginning of the discussion. This seems to me a big leap and I would advise to be careful with this statement. Reflecting on this earlier in the discussion would help. Another suggestion is to change the number into percentage for easy reading.

In the discussion, more attention could be payed to the implications of the findings for the wellbeing of the Chinese population. Some elements of the introduction, methods and results do not resonate in this last section.

Introduction:

The introduction provides clear evidence for the link between mindfulness, emotion regulation and mental health. The importance of considering all three factors is furthermore evident. I did have some difficulties keeping track of the order in which the evidence is discussed. The three concepts are sometimes mixed within a paragraph aiming to focus on one of the concepts. A clearer distinction would aid readers in the process. The rational for examining both within- and between subject findings simultaneously (an important part of this manuscript) could use some elaboration.

In the first paragraph, I would avoid using the words adjustment outcomes and change these to mental health outcomes to make the link explicit.

Section 'processes of mental health in the Chinese context':

- This section contains several long sentences with unclear structures.

- ‘similar findings were suggested’ needs to be ‘were found’.

- I would highlight a study into emotion regulation in this context to balance the section (e.g., take out a mindfulness example and replace it with one for emotion regulation).

- ‘These findings highlight’ in present tense.

Section 'within- vs. between-subject effects':

- What about emotion regulation in the first sentence?

- ’Partitioning between within- and between-person’: consider using a synonym (e.g. among).

- A clinical interpretation of the importance to distinguish the within-person results and between-person results is lacking. Why is it relevant to separate them, apart from possible biases? What are the differences in interpretation for between-person results and within-person results? Examples are given for within-person results but these are not compared to between-person differences. How can separating both approaches help advance guidelines for future treatment options?

- The transition between the two paragraphs in this section is not very clear.

'This study':

- Is the analysis technique novel or the used approach?

- Why is this unique? Elaborate in the paragraph above.

- The information ‘may be linked to covariates including time and gender’ is new information. A motivation for these choices could be added to the introduction.

Method:

- Participants were at 18-27 years old: needs to be between. This information seems part of the result section (same as the retention rate).

- What were the inclusion and exclusion criteria? Which number of participants was aimed at? How many emails were send out? Some basic Information is lacking.

- The begin structure is unclear; some information from the procedure part belongs to a part about participants and vice versa.

- What sort of informed consent was given?

- How much time did participants investigate in the study?

Measures section:

- This section could benefit from a description of the reliability and validity of the used scales.

- The timeframes of the mental health questionnaires differ (one assesses the past week, another the past month). Could this have influenced the findings in any way?

- Anxiety symptoms: was used to measure anxiety, not depression.

Data analysis section:

- The general writing style of this section is suggestive, which results in unclearity when reading.

- Why are there no details provided on the multilevel mediation model used here (now upon request)? Was growth curve modelling used? Some information would better fit the introduction.

- Please specify the used levels.

- Separate results from methods.

- The section on the mediation path is difficult to follow.

- The last part of this section is especially unclear, does this mean that the used analysis require not much faith? Why is this not mentioned in the discussion?

Results:

- When reading this section I would first expect general sample information, including cores on all outcome measures. Then correlational information.

- Table 1: This table shows many results at the same time. While it seems correct, it might be easier for readers to separate the correlations from the means and standard deviations of the outcomes (make two tables). Before considering correlations, I am curious about how the population scored on the measures. To aid the interpretation, it would be good to mention some reference points for ranges. For example, to interpret the 2.85 score on emotion regulation, I would like to know what this means. What score reflects serious difficulties with emotion regulation?

- Gender does not make sense as M and SD.

- The within-person correlations are not part of the analyses section, a clear rational for these results does not follow naturally.

Section 'multilevel structural equation modeling':

- Are the design effect results reported somewhere?

- Some information in this section would fit the analyses section better.

- Results are better presented numerically, an interpretation of the variances explained suites the discussion (interesting information). They are part of figure 1, might be good to mention this earlier in the section.

Section 'within-person indirect effect of mindfulness awareness (and between)':

- How do the results mentioned here relate to the heading title of indirect effects (they mention direct effects)?

- Table 2: Please explain the arrows and the c’w1i notations in the notes.

- Table 3: The explanation helps. Some of these elements could be taken into the table, providing a heading for within and between analyses results. Consider keeping the format similar to the other tables (e.g., using * instead of boldfaced numbers for significance).

Discussion:

- The first section clearly summarizes the results. This part could benefit from an integrated interpretation of the findings. For example, what could be the reason for the surprising finding of time?

- The statements ‘findings advance the field by using a process-oriented approach’ and ‘demonstrates the utility of multilevel modeling’ could use more explanation.

- The within-person and between-person definitions (when mindfulness of person i increases etc.) can be left out; this is sufficiently clear from the introduction.

- The section on between-person findings with emotion regulation as an explanatory variable is confusing because of the sentence structures.

- Please elaborate on ‘stage-salient variables’.

Reviewer #2: This paper investigates whether emotion regulation is a mediator between mindful awareness and depression, anxiety, and subjective wellbeing in Chinese young adults. Findings suggest that emotion regulation mediates the association between mindful awareness and wellbeing both at the within-person and the between-person level, with the exception that emotion regulation did not mediate the between-person association between mindfulness and subjective wellbeing.

This is an interesting and well-written paper. I only have a few minor comments.

Minor comments

1. I would prefer the term ‘sum score’ or ‘total score’ rather than ‘composite score’ (p. 14) to reflect the total score on a single questionnaire, given that composite score is usually reserved for a score that reflects a combination of scales.

2. “The item scores were summed to form a composite score of depressive symptoms”. However, Table 1 appears to report mean scores rather than sum scores (on a scale from 0 to 3 instead of the original scale from 1 to 4?).

3. For subjective wellbeing, unlike the other scales, it is not explicitly stated how the score was calculated (mean score of the 14 items?).

4. In Table 1 the mean for emotion regulation is 2.97 but in the supplementary data set it appears to be 3.97 if I am not mistaken.

5. Please explain ‘stage-salient variables’ (p. 24). Do have some suggestions as to which stage-salient variables may be relevant to include in future research?

6. The manuscript contains a few typos:

- p. 10 stwell-being

- p. 14 “how often did you feel that that you had ….”

- p. 14 “A sample items included” “

- p. 14 The 7-item Generalized Anxiety Disorder‐7 … was used to measure depressive symptoms”. Depressive symptoms should be anxiety symptoms.

- p. 13 “The 15-item Mindful Attention Awareness Scale …. on a 6-point scale from 1 (almost always) to 5 (almost never)”. This should say ‘a scale from 1 to 6’?

Reviewer #3: The study does not encompass the Emotion Regulation topic but does emotional dysregulation. The implemented measures suggest another title (see attached file). MAAS measures dispositional mindful awareness and DERS assesses emotional dysregulation.

The sample is not enough, gender was not balanced, there was no intervention, the undergraduate sample is too biased for the conclusions (should be tentative rather than so affirmative).

IMuch data were collected and the information about them were scarcy. You collected much data and there was no information about subscales, cutoff (GAD and CES-D). I need more information and explanation of why the addressed these statistical analyses made by the authors. Table 1 is not an ICC (intraclass correlations), it is intercorrelations.

Discussion can be more tentative than affirmative due to data did not allow those conclusions.

see attached file to see more comments in the document

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes: José M Mestre

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: PONE-D-20-06030_reviewer.pdf

Click here for additional data file.

17 Jul 2020

We would like to thank the three reviewers for the thoughtful and constructive comments. Below, we outline the revisions that we have made to address these comments. The revisions have also been highlighted in yellow in the revised manuscript for your convenience.

Reviewer #1

General Comments

Comment:

Thank you for the interesting manuscript, studying emotion regulation as a mediator between mindfulness and mental health in the Chinese young adult population. I am happy to read that both the within- and between subject factors are taken into consideration in this study design.

Our response:

We thank Reviewer #1 for the positive comments.

Comment:

A general remark relates to the umbrella term ‘mental health’ used in the research question, this remains rather broad. The introduction includes a description, but throughout it is not always clear how mental health is defined with respect to design and outcomes. The separate measures (depression, anxiety, and subjective wellbeing) are sometimes interchanged with the broader mental health concept. Does this then refer to a composite measure of these questionnaires?

Our response:

Thank you for the comment. Our definition is consistent with the definition of health provided by the World Health Organization [3]. Specifically, we refer mental health to better mental well-being and low levels of psychological symptoms.

Comment:

A very high percentage of the study population is female (90%), this is correctly mentioned as a limitation at the end of the manuscript, but briefly. Gender is taken as a covariate in the analyses and conclusions are drawn on these analyses in the beginning of the discussion. This seems to me a big leap and I would advise to be careful with this statement. Reflecting on this earlier in the discussion would help. Another suggestion is to change the number into percentage for easy reading.

Our response:

Thank you for the comment. We have now used percentage and number for easy reading (p. 7). In the Discussion, we have removed, “[g]ender was not related to symptoms of depression and anxiety.” We have instead stated (p. 23), “gender was not related to psychological symptoms in the model. However, our small sample of men (i.e., 9.95%) precluded us from drawing meaningful conclusions about gender as a correlate of psychological distress.”

Comment:

In the discussion, more attention could be payed to the implications of the findings for the wellbeing of the Chinese population. Some elements of the introduction, methods and results do not resonate in this last section.

Our response:

Thank you for the comment. We have now incorporated existing findings of Chinese samples in the Discussion to make the manuscript more coherent. For example, we have added (p. 22), “[t]he findings resonated with recent findings conducted in the Chinese context, in that emotion dysregulation mediated between dispositional mindfulness and symptoms of depression and anxiety [10].” We have also noted (p. 22), “[t]hese findings contradicted recent cross-sectional studies [29,69] showing the link between Chinese college students’ emotion regulation strategies and life satisfaction (i.e., a major component of subjective well-being [70]).” By situating our findings to the Chinese context, we hope our readers are better informed.

Introduction

Comment:

The introduction provides clear evidence for the link between mindfulness, emotion regulation and mental health. The importance of considering all three factors is furthermore evident. I did have some difficulties keeping track of the order in which the evidence is discussed. The three concepts are sometimes mixed within a paragraph aiming to focus on one of the concepts. A clearer distinction would aid readers in the process. The rational for examining both within- and between subject findings simultaneously (an important part of this manuscript) could use some elaboration.

Our response:

In the revised introduction, we broadly discussed the role of mindfulness in mental health (p. 3) and then moved onto emotion regulation as a predictor of mental health (p. 3). In the next section, we briefly discussed Teper et al.’s theoretical framework [13] and Garland et al.’s framework [21] concerning the mediating role of emotion regulation (p. 4), which was followed by empirical findings (p. 4-5). To situate the study in the Chinese context, we further discussed the findings in Chinese samples (p. 5), with mindfulness as a predictor of emotion regulation and mental health. To ensure that our audience can fully grasp the meanings of within- vs. between-person effects, we discussed the differences, with examples in the context of mindfulness, emotion regulation, and mental health correlates. Of note, partitioning among within- and between-person associations is important, because otherwise the results could sometimes be biased [34] or uninterpretable [35] (p. 6).

We hope that the flow has now provided a clearer picture in supporting the present study.

Comment:

In the first paragraph, I would avoid using the words adjustment outcomes and change these to mental health outcomes to make the link explicit.

Our response:

We thank the reviewer for the comment. We have revised the introduction by changing “adjustment outcomes” to “mental health outcomes.”

Introduction: Section 'processes of mental health in the Chinese context'

Comment:

Section 'processes of mental health in the Chinese context': This section contains several long sentences with unclear structures.

Our response:

The long sentences have now been shortened to increase the readability (p. 5).

Comment:

‘similar findings were suggested’ needs to be ‘were found’.

Our response:

We have now updated the sentence on page 5.

Comment:

I would highlight a study into emotion regulation in this context to balance the section (e.g., take out a mindfulness example and replace it with one for emotion regulation).

Our response:

Thank you for the comment. We have now highlighted a study (p. 5) showing that changes in emotion regulation strategies were associated with changes in changes in depressive symptoms, life satisfaction, and general health over time [29].

Comment:

‘These findings highlight’ in present tense.

Our response:

We have revised the introduction by changing “These findings highlighted” to “These findings highlight” on page 5.

Introduction: Section 'within- vs. between-person effects'

Comment:

What about emotion regulation in the first sentence?

Our response:

Thank you for the comment. The first sentence has now been updated to, “[a] majority of research to-date focuses on between-person associations of mindfulness, emotion dysregulation, and mental health.”

Comment:

’Partitioning between within- and between-person’: consider using a synonym (e.g. among).

Our response:

We have changed “Partitioning between within- and between-person associations” to “Partitioning among within- and between-person associations” on page 6.

Comment:

A clinical interpretation of the importance to distinguish the within-person results and between-person results is lacking. Why is it relevant to separate them, apart from possible biases? What are the differences in interpretation for between-person results and within-person results? Examples are given for within-person results but these are not compared to between-person differences. How can separating both approaches help advance guidelines for future treatment options?

Our response:

Thank you for the comment. In the revised introduction, we have highlighted the statistical significance for distinguishing among the within-person and between-person findings (p. 5-6), in that “partitioning between among within- and between-person associations is important, because otherwise the results could be biased [34] or uninterpretable [35].”

Given that the literature has primarily shown between-person effects (p. 5-6), conclusions could only be drawn, specifically, in that when person i has a higher score in mindfulness than person j, then person i would have a higher score in emotion regulation and mental health than would person j. In this study, we argued that it is theoretically important to also address whether increases in mindfulness in person i are linked to his or her own better emotion regulation and mental health over time.

The present findings have important implications, as addressed in the Discussion. At the within-person level (p. 21-22), practitioners could tailor their practices and interventions to the progress of each client. For example, to enhance subjective well-being and reduce psychological symptoms, practitioners can enhance within-person improvements in mindfulness and emotion regulation, as mindfulness affects mental health outcomes both directly and through adaptive emotion regulation. At the between-person level (p. 23), practitioners should be made aware of their clients’ levels of mindfulness, emotion dysregulation, and mental health relative to other people. For example, returning to Figure 1, in relation to person j, person i’s greater level of mindfulness was associated with a lower level of emotion dysregulation and better well-being than person j, but not directly with fewer symptoms of depression and anxiety. Understanding these associations could help practitioners gauge the between-person importance or relevance of mindfulness in various mental health outcomes.”

Comment:

The transition between the two paragraphs in this section is not very clear.

Our response:

Under the section 'Within- vs. Between-Person Effects, we have now strengthened the transition by adding the sentence, “[s]everal studies to-date have demonstrated the utility of within-person analyses in mindfulness and well-being.”

Introduction: 'This study'

Comment:

Is the analysis technique novel or the used approach?

Our response:

We have deleted “novel” from “This Study” given that it has been used in previous research.

Comment:

Why is this unique? Elaborate in the paragraph above.

Our response:

In the revised manuscript, we have added (p. 6-7), “[a]s stated earlier, previous research predominantly examined between-person effects among mindfulness, emotion regulation, and mental health [5-11]. Although some studies have begun to examine within-person effects [36-38], few, if any, have teased apart within-person from between-person effects.”

Comment:

The information ‘may be linked to covariates including time and gender’ is new information. A motivation for these choices could be added to the introduction.

Our response:

Thank you for the comment. On page 5, we have strengthened our arguments concerned with time by adding, “changes in mindful awareness were associated with subsequent changes of stress response and anxiety symptoms [28]. Similarly, changes in emotion regulation strategies were associated with changes in changes in depressive symptoms, life satisfaction, and general health over time [29]. As such, timing and changes were crucial in linking mindfulness, emotion regulation, and mental health outcomes.”

As for gender, previous research conducted with Chinese samples revealed that gender was associated with depression [39], anxiety [40], and subjective well-being [41]. As such, gender was included in this study as a covariate (p. 7).

Method

Comment:

Participants were at 18-27 years old: needs to be between. This information seems part of the result section (same as the retention rate).

Our response:

We have updated “[p]articipants were at 18-27 years old” to “[p]articipants were between 18 and 27 years old.” Since age and retention were not the study focus or part of the hypothesis, we have only included the information under “Participants” of the Method section (p. 7-8).

Comment:

What were the inclusion and exclusion criteria?

Our response:

We have added the following sentence (p.8), “Inclusion criteria included college-enrolled emerging adults who were proficient in Chinese, within the age range between 18 and 29 years old [42], who agreed to participate for three time points over the course of a year.”

Comment:

Which number of participants was aimed at?

Our response:

Thank you for the comment. We have conducted a post hoc power analysis based on Monte Carlo simulations [76]. Based on the estimated parameter values from the multilevel mediation analysis, we repeatedly simulated 1000 new data sets and fitted the same multilevel mediation model to these simulated data sets. By counting the proportion of significant results, we obtained the empirical post hoc power. The results showed that on average, the post hoc power for the six indirect effects considered in the study (three at the within-person level and three at the between-person level) was 80.2%, suggesting that it was adequate to aim for 191 participants.

Reference:

76. Gelman A, Hill J. Data analysis using regression and multilevel/hierarchical models. Cambridge: Cambridge University Press; 2006.

Comment:

How many emails were send out?

Our response:

Two mass emails were sent out (p. 7-8).

Comment:

The begin structure is unclear; some information from the procedure part belongs to a part about participants and vice versa.

Our response:

We have worked to ensure that the Participants and Procedures sections were independent from each other (p. 7-8). Specifically, the Participants section involves recruitment method, inclusion criteria, retention rate, and demographics. The Procedures section involves ethics approval information, informed consent details, and data collection procedures.

Comment:

What sort of informed consent was given?

Our response:

We have revised the Procedures section by stating the following (p. 8), “The study was approved by the Human Research Ethics Committee of the first author’s university prior to its implementation. All procedures performed were in accordance with the ethical standards of the institutional research committee and the 1964 Helsinki declaration and its later amendments. Informed consent to participate in the present study was obtained prior to the administration of the questionnaire.”

Comment:

How much time did participants investigate in the study?

Our response:

We have now included the following information (p. 8), “[e]ach packet took approximately 30 minutes to complete.”

Method: Measures

Comment:

This section could benefit from a description of the reliability and validity of the used scales.

Our response:

Based on previous research, we have now included the reliability and validity of the scales used in the present study (p. 8-12).

Comment:

The timeframes of the mental health questionnaires differ (one assesses the past week, another the past month). Could this have influenced the findings in any way?

Our response:

We agree with Reviewer #1 that the timeframes of the mental health questionnaire differ. As a limitation, we have now stated (p.24), “the timeframes of the mental health questionnaires differ. For example, the PHQ-9 assessed depressive symptoms in the past two weeks, whereas the Mental Health Continuum Short Form assessed emotional, social, and psychological well-being in the past four weeks, whereas the MHC-SF assessed emotional, social, and psychological well-being in the past four weeks. Future research may consider standardizing the timeframes to increase precision of the variables.”

Comment:

Anxiety symptoms: was used to measure anxiety, not depression.

Our response:

Thank you for the comment. We have made relevant changes in the revised manuscript (p. 11).

Method: Data analysis

Comment:

The general writing style of this section is suggestive, which results in unclearity when reading.

Our response:

We have expanded and revised the Data Analysis section. A more affirmative writing style has been used.

Comment:

Why are there no details provided on the multilevel mediation model used here (now upon request)? Was growth curve modelling used? Some information would better fit the introduction.

Our response:

Thank you for the comment. We have now provided a statistical description of the multilevel mediation model (p. 12-15). The multilevel mediation model used in the analysis was not a growth curve model. To avoid confusion, we have removed sentences related to growth curve modeling in the revised manuscript. Two major differences between multilevel mediation model and growth curve model are as follows:

First, in our analysis, the relationship between the independent variables and the mediator (i.e., reflected by the “a” path) and the relationship between the mediator and the dependent variables (i.e., reflected by the “b” paths) were estimated simultaneously. In growth curve modeling with time-varying covariates, however, these relationships were estimated separately. The estimated “a” path and “b” paths could be correlated when data are nonnormal and when robust methods are used (i.e., when MLR is used in Mplus; see the discussion in http://www.statmodel.com/discussion/messages/11/9365.html?1396831771). In our analysis, MLR was used and the possible association between the estimated “a” path and “b” path could be captured. In contrast, the association could not be captured using growth curve modeling. This might have an impact on the inferences on the indirect effect, which was defined as the product of the “a” path and the “b” path (e.g., see [77; p. 92]; and the explanation regarding the difference between the delta method used in Mplus and Sobel test in http://www.statmodel.com/discussion/messages/11/9365.html?1396831771).

Second, we relied on the multilevel SEM approach [34] to separate the within- and between-person components of variables. In contrast, as a typical multilevel modeling, growth curve modeling relies on group centering to fulfill this objective. As discussed in [34], the multilevel approach was relatively more biased.

Reference:

77. MacKinnon DP. Introduction to statistical mediation analysis (Multivariate applications series). Taylor & Francis Group/Lawrence Erlbaum Associates; 2008.

Comment:

Please specify the used levels.

Our response:

We have now specified the used levels in greater detail.

Comment:

Separate results from methods.

Our response:

Thank you for the comment. The results have been separated from methods.

Comment:

The section on the mediation path is difficult to follow.

Our response:

Thank you for the comment. We have revised the Data Analysis section. Now the logic of this section flows more naturally and it is easier to understand. Specifically, we first described the statistical approach to multilevel mediation used in this study (p. 12). We then explained what covariates and why they were considered in the model (p. 12). Next, we explained why some paths were fixed and some was allowed to be random across participants (p.13). We also described in detail the multilevel mediation model used in the analysis and explained the practical meanings of relevant model parameters (p. 13-15). Finally, we explained why the proportion of indirect effect was not calculated (p. 15).

Comment:

The last part of this section is especially unclear, does this mean that the used analysis require not much faith? Why is this not mentioned in the discussion?

Our response:

Thank you for the comment. The last sentence of the Data Analysis section (p. 15), “[u]nless the sample size was fairly large, Hayes [60, p.189] recommended not ‘having much faith’ in this measure” explains the reason why we did not compute the ratio of indirect effect to the total effect. It was not related to the method used in the analysis.

Results

Comment:

When reading this section I would first expect general sample information, including cores on all outcome measures. Then correlational information... Before considering correlations, I am curious about how the population scored on the measures. To aid the interpretation, it would be good to mention some reference points for ranges. For example, to interpret the 2.85 score on emotion regulation, I would like to know what this means. What score reflects serious difficulties with emotion regulation?

Our response:

Thank you for the comment. We have now briefly described the correlational findings at the beginning of the results section (p. 15).

As for how the current sample scored on the measures compared to other study samples, we have now included the information under the measures section (p. 8-12). Specifically, compared to other Chinese samples, participants from this study had similar mean levels of dispositional mindfulness [44], emotion dysregulation [46], and symptoms of depression and anxiety [50]. Nevertheless, they had greater mean levels of subjective well-being [56]. This information has also been included in the discussion section (p. 24).

Comment:

Table 1: This table shows many results at the same time. While it seems correct, it might be easier for readers to separate the correlations from the means and standard deviations of the outcomes (make two tables).

Our response:

We have separated Table 1 into two tables according to Reviewer #1’s recommendation. Table 1 shows the means, standard deviations, ranges, ICCs, and design effects, whereas Table 2 shows the zero-order correlations.

Comment:

Gender does not make sense as M and SD.

Our response:

We have removed the M and SD of gender in Table 1.

Comment:

The within-person correlations are not part of the analyses section, a clear rational for these results does not follow naturally.

Our response:

Thank you for the comment. We have revised the Data Analysis section (p. 12-15). Now the subsection contains a detailed description on the multilevel mediation model. The analyses regarding the within- and between-person associations have also been described.

Results: 'Multilevel Structural Equation Modeling'

Comment:

Are the design effect results reported somewhere?

Our response:

The design effect results have now been reported in Table 1.

Comment:

Some information in this section would fit the analyses section better.

Our response:

To enhance clarity and organization, we have restructured the Data Analysis and the Results sections (p. 12-20).

Comment:

Results are better presented numerically, an interpretation of the variances explained suites the discussion (interesting information). They are part of figure 1, might be good to mention this earlier in the section.

Our response:

Thank you for the comment. As we were unable to denote the % of variance in Figure 1, we did not describe the figure early on. However, at the end of the paragraph, we have recommended the readers to refer to Figure 1 for specific path coefficients.

Results: 'Within-person indirect Effect of dispositional mindfulness (and between)'

Comment:

How do the results mentioned here relate to the heading title of indirect effects (they mention direct effects)?

Our response:

Thank you for the comment. The section refers to indirect effects.

Comment:

Table 3: Please explain the arrows and the c’w1i notations in the notes.

Our response:

Thank you for the comment. We have added an explanation in note on Table 3 to explain the notations.

Comment:

Table 4: The explanation helps. Some of these elements could be taken into the table, providing a heading for within and between analyses results. Consider keeping the format similar to the other tables (e.g., using * instead of boldfaced numbers for significance).

Our response:

We have revised Table 4 by using * instead of boldfaced numbers for significant numbers.

Discussion

Comment:

The first section clearly summarizes the results. This part could benefit from an integrated interpretation of the findings. For example, what could be the reason for the surprising finding of time?

Our response:

Thank you for the comment. At the beginning of the Discussion, we have included a potential reason for the surprising finding of time (p. 21) by stating, “potentially due to an increasing level of academic stress over the school year.” To ensure that the opening paragraph is clear and succinct, we have explained the findings in greater detail in the subsequent paragraphs.

Comment:

The statements ‘findings advance the field by using a process-oriented approach’ and ‘demonstrates the utility of multilevel modeling’ could use more explanation.

Our response:

Thank you for the comment. To ensure that our discussion is clear and informative, the statement, “findings advance the field by using a process-oriented approach,” have been revised to “[t]hese findings advanced the field by establishing within- and between-person relations between dispositional mindfulness and mental health via emotion dysregulation.” (p.21)

The statement, “demonstrates the utility of multilevel modeling” have been revised to “[t]his study used multilevel modeling by partitioning between-person relations from within-person relations.”(p.21)

Comment:

The within-person and between-person definitions (when mindfulness of person i increases etc.) can be left out; this is sufficiently clear from the introduction.

Our response:

We thank Reviewer #1 for the recommendation. We have removed the within-person definition in the discussion. To concretely explain the between-person findings and their clinical significance, however, we have decided to keep the notations (e.g., person i vs person j; p. 22-23).

Comment:

The section on between-person findings with emotion regulation as an explanatory variable is confusing because of the sentence structures.

Our response:

We have now revised the section (p.22-23).

Comment:

Please elaborate on ‘stage-salient variables’.

Our response:

Thank you for the comment. Stage-salient variables refer to variables specific to a certain developmental period, i.e., in our case, emerging adulthood. In the discussion, we have added, “other stage-salient variables in emerging adulthood, such as having increasing responsibility, becoming more financial independent, and having a greater commitment in romantic relationships [73], may be more strongly associated with positive well-being.” (p.23)

Reviewer #2

Comment:

This paper investigates whether emotion regulation is a mediator between mindful awareness and depression, anxiety, and subjective wellbeing in Chinese young adults. Findings suggest that emotion regulation mediates the association between mindful awareness and wellbeing both at the within-person and the between-person level, with the exception that emotion regulation did not mediate the between-person association between mindfulness and subjective wellbeing.

This is an interesting and well-written paper. I only have a few minor comments.

Our response:

We thank Reviewer #2 for the positive comments.

Minor Comments

Comment:

I would prefer the term ‘sum score’ or ‘total score’ rather than ‘composite score’ (p. 14) to reflect the total score on a single questionnaire, given that composite score is usually reserved for a score that reflects a combination of scales.

Our response:

We have revised the method by changing “composite scores” to “mean scores.” (p.8-12)

Comment:

“The item scores were summed to form a composite score of depressive symptoms”. However, Table 1 appears to report mean scores rather than sum scores (on a scale from 0 to 3 instead of the original scale from 1 to 4?).

Our response:

Thank you for the comment. We apologize for the confusion. In this study, participants rated on a scale from 0 to 3 for each item of the PHQ-9. Afterwards, the item scores were averaged to form a score. The description has now been updated in the manuscript (p. 10).

Comment:

For subjective wellbeing, unlike the other scales, it is not explicitly stated how the score was calculated (mean score of the 14 items?).

Our response:

We have now stated how the subjective well-being score was calculated (p. 12), “The item scores were averaged to form a mean score, with higher scores indicating better well-being.”

Comment:

In Table 1 the mean for emotion regulation is 2.97 but in the supplementary data set it appears to be 3.97 if I am not mistaken.

Our response:

Thank you for pointing this out. We have corrected the means and standard deviations for emotion dysregulation.

Comment:

Please explain ‘stage-salient variables’ (p. 24). Do have some suggestions as to which stage-salient variables may be relevant to include in future research?

Our response:

Thank you for the comment. Stage-salient variables refer to variables specific to a certain developmental period, i.e., in our case, emerging adulthood. In the discussion, we have added (p.23), “other stage-salient variables in emerging adulthood, such as having increasing responsibility, becoming more financial independent, and having a greater commitment in romantic relationships [73], may be more strongly associated with positive well-being.”

6. The manuscript contains a few typos:

- p. 10 stwell-being

- p. 14 “how often did you feel that that you had ….”

- p. 14 “A sample items included” “

- p. 14 The 7-item Generalized Anxiety Disorder‐7 … was used to measure depressive symptoms”. Depressive symptoms should be anxiety symptoms.

- p. 13 “The 15-item Mindful Attention Awareness Scale …. on a 6-point scale from 1 (almost always) to 5 (almost never)”. This should say ‘a scale from 1 to 6’?

Our response:

Thank you for the comment. The typos have been removed in the revised manuscript.

Reviewer #3:

Title Page

Comment:

Title arose some expectations that were not accomplished later.

Our response:

Thank you for the comment. We have updated the title to, “Dispositional Mindfulness and Mental Health in Chinese Emerging Adults: A Multilevel Model with Emotion Dysregulation as a Mediator.”

Comment:

DERS: Difficulties in ER Scale measures emotional dysregulation rather than emotion regulation. Dysregulation: A failure to regulate properly.

Our response:

We agree with Reviewer #3’s comment. In our original manuscript, the scores of the negative items of DERS were reversed, such that the final scores reflected a greater ability in emotion regulation.

Based on Reviewer #3’s comment, we have unreversed the scores, such that higher scores reflected greater emotion dysregulation. Where appropriate, we have now used “emotion dysregulation” in the revised manuscript.

Comment:

Many authors have pointed out that these mindful scales are not measuring mindful activities, they assess dispositional mindful activities.

Our response:

We agree with Reviewer #3’s comment and have changed our wordings to “dispositional mindfulness” in describing the variable under study.

Abstract

Comment:

Provide here M and SD for participants’ age.

Our response:

We have revised the abstract by adding M and SD of the participants’ age.

Comment:

A questionnaire about what?

Our response:

We have added that the questionnaire assessed their dispositional mindfulness, emotion dysregulation, and mental health outcomes.

Introduction

Comment:

Avoid using regulating emotions in the definition of emotion regulation.

Our response:

We have revised the revised the sentence to the following, “[e]motion regulation is defined as the process in modulating emotions and emotional responses [14-15].” (p.3)

Introduction: Within- vs. Between-Person Effects

Comment:

Agree but this section above is a little bit tangential to your research. You did not implement a RCT MBI.

Our response:

We apologize about the confusion and agree with Reviewer #3 that we did not implement a mindfulness-based intervention. Given that very few, if any, studies assessed within- vs. between-person effects of dispositional mindfulness on mental health, we have reviewed mindfulness-based intervention studies using a multilevel modeling approach.

Method

Comment:

Does this Committee a code?

Our response:

We have revised the Method by stating (p. 8), “The study was approved by the Human Research Ethics Committee of the first author’s university institution prior to its implementation. All procedures performed were in accordance with the ethical standards of the institutional research committee and the 1964 Helsinki declaration and its later amendments.”

Comment:

The sample is not enough, gender was not balanced, there was no intervention…

Our response:

We agree with Reviewer #3 that the sample size could have been larger and gender could be more balanced.

In terms of sample size, we have conducted a post hoc power analysis based on Monte Carlo simulations [76]. Based on the estimated parameter values from the multilevel mediation analysis, we repeatedly simulated 1000 new data sets and fitted the same multilevel mediation model to these simulated data sets. By counting the proportion of significant results, we obtained the empirical post hoc power. The results showed that on average, the post hoc power for the six indirect effects considered in the study (three at the within-person level and three at the between-person level) was 80.2%, suggesting that it was adequate to aim for 191 participants.

As for gender, we have stated in the Discussion (p. 23), “our small sample of men (i.e., 9.95%) precluded us from drawing meaningful conclusions about gender as a correlate of psychological distress.” We have also acknowledged, (p. 24), “our participants were mainly female. Future studies with gender-balanced samples are necessary to draw meaningful conclusions for the effects of gender.”

We understand that intervention studies are necessary to improve mental health. However, it is beyond the scope of the present study. As a future direction, we have added (p. 24), “researchers should translate the present findings and design targeted interventions to improve mental health outcomes.”

Reference:

76. Gelman A, Hill J. Data analysis using regression and multilevel/hierarchical models. Cambridge: Cambridge University Press; 2006.

Comment:

This tool measures dispositional mindful awareness.

Our response:

We have now updated the description to the following (p.8), “[t]he 12-item Cognitive and Affective Mindfulness Scale – Revised (CAMS-R) [43] was used to assess dispositional mindfulness on a 4-point scale from 1 (rarely/ not at all) to 4 (almost always).” The measure, CAMS-R, broadly assessed dispositional mindfulness, in addition to dispositional mindful awareness.

Comment:

Cronbach's alpha of subscales should be also reported. You can use intercorrelations at T1, T2 and T3 as well.

Our response:

Thank you for the comment. We have now indicated the Cronbach's alpha of subscales for the Cognitive and Affective Mindfulness Scale – Revised (p. 9), the Difficulties in Emotional Regulation Scale (p. 9), and the Mental Health Continuum Short Form (p. 12). The intercorrelations between T1, T2 and T3 have been reported for all measures as well (p. 8-12).

Comment:

Is there a cut-off in the scoring to indicate depression?

Our response:

Thank you for the question. The cut-off scores for the Patient Health Questionnaire-9 and the Generalized Anxiety Disorder‐7 measures have now been reported on page 10.

Results

Comment:

Gender: 0 male and 1 female?

Our response:

We have now denoted “0=male; 1=female” on Tables 1-3.

Comment:

Why are the correlations of mindfulness moderate / low at T1, T2, T3?

Our response:

The correlations of dispositional mindfulness between T1 and T3 were .53 - .58 over time, ps < .001, suggesting moderate effect sizes [62]. Even though the scale was intended to measure stable and dispositional mindfulness, our data suggest potential changes over time.

Comment:

Based on the correlations, gender could be a moderator.

Our response:

We agree with Reviewer #3’s comment. However, our small sample of men (i.e., 9.95%) precluded us from drawing meaningful conclusions about gender as a moderator. As a limitation, we have stated (p. 24), “our participants were mainly female. Future studies with gender-balanced samples are necessary to draw meaningful conclusions for the effects of gender.”

Comment:

M and SD seem to be averaged, this should be notified in the table and put in the first column the range of every single instrument.

Our response:

We have now stated (p. 15), “Table 1 summarizes the means, SDs, ranges, ICCs, and design effects.”

Comment:

ICC is more appropriate to assess the level of agreement among raters. Where is reported the ICC, Table 1 is not a ICC.

Our response:

Thank you for the comment. We have added ICCs in Table 1 for all the variables under study.

Discussion

Comment:

You did not implement an intervention... The discussion needs to follow a tentative argument… The undergraduate sample is too biased for the conclusions (should be tentative rather than so affirmative). Discussion can be more tentative than affirmative due to data did not allow those conclusions.

Our response:

Thank you for the comment. We agree with Reviewer #3 that the present study involved no experiments and interventions. Thus, in the discussion, we have avoided wordings such as “causes”, “affects”, or “influences.” Based on the present findings, tentative arguments have been made, without biases or exaggerations. As a limitation, we have added, “researchers should translate the present findings and design targeted interventions to improve mental health outcomes.”

Submitted filename: Response Letter 13-7-2020.docx

Click here for additional data file.

11 Aug 2020

PONE-D-20-06030R1

Dispositional Mindfulness and Mental Health in Chinese Emerging Adults: A Multilevel Model with Emotion Dysregulation as a Mediator

PLOS ONE

Dear Dr. Ke

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Sep 25 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Therese van Amelsvoort

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: No

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: No

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: The authors have addressed all my comments adequately. However, there are a few issues with the revised version that should be addressed:

1. Table 1. Means and standard deviations of the study variables (N = 191): The table not only includes means and standard deviations but also range, ICC and design effect. Under the column range, however, only a single number is given rather than a range. Furthermore, the abbreviation ICC should be explained in the caption of the table and it should also explain what 'design effect' is.

2. "Based on a cutoff score of 11 for the detection of depressive disorders [49], upon rescaling, our participants’ summed scores of 5.44 - 6.67 between T1 and T3 were below the clinical cutoff": the rescaling is not explained. How were the the PHQ-9 scores rescaled? The same applies for the anxiety symptoms measures with the GAD-7 ("Based on a cutoff score of 10 for the detection of generalized anxiety disorder [51], upon rescaling our participants’ summed scores between 4.22 and 5.67 between T1 and T3 were below the clinical cutoff").

3. "The design effect, defined as 1+ (average cluster size-1)×ICC". This should be explained in the Methods (and in the caption of Table 1 if you include the design effect in this table) rather than introducing it in the Results. In the current manuscript the text first refers to the design effect (first paragraph results, Table 1) and only explains it later.

4. “.... Hayes [60, p.189] recommended ...’”. The reference should be reference 61 rather than 60.

5. The manuscript needs some language editing. E.g.:

- ".... changes in emotion regulation strategies were associated with changes in changes in depressive symptoms ...";

- "Such an increase was found to be more significantly than did participants who ...";

- ".... confounding between-person factors from the within-person associations ..."

Reviewer #3: Congratulations on the implemented efforts to improve the paper. I left some comments throughout the paper to amend some issues that I found (even the reference list)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

Reviewer #3: Yes: Jose M Mestre PhD. Department of Psychology, Universidad de Cádiz

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: PONE-D-20-06030_R1 (1).pdf

Click here for additional data file.

13 Aug 2020

Reviewer #2

Comment:

Table 1. Means and standard deviations of the study variables (N = 191): (a) The table not only includes means and standard deviations but also range, ICC and design effect. (b) Under the column range, however, only a single number is given rather than a range. (c) Furthermore, the abbreviation ICC should be explained in the caption of the table and it should also explain what 'design effect' is.

Our response:

Thank you for the comment.

(a) We have updated Table 1’s caption (p. 10) to “Table 1. Descriptive statistics of the variables under study.”

(b) The range was originally calculated by the difference between lower and higher scores in every scale. To improve clarity, we have relabeled the column to “Range of Scale” and provided the range. To include further descriptive statistics, we have also added “Minimum” and “Maximum” in Table 1.

(c) The abbreviation ICC has now been replaced by “Intraclass Correlation.” A note has been added to the bottom of Table 1 to define “design effect.”

Comment:

"Based on a cutoff score of 11 for the detection of depressive disorders [49], upon rescaling, our participants’ summed scores of 5.44 - 6.67 between T1 and T3 were below the clinical cutoff": the rescaling is not explained. How were the the PHQ-9 scores rescaled? The same applies for the anxiety symptoms measures with the GAD-7 ("Based on a cutoff score of 10 for the detection of generalized anxiety disorder [51], upon rescaling our participants’ summed scores between 4.22 and 5.67 between T1 and T3 were below the clinical cutoff").

Our response:

Thank you for the comment. Under the PHQ-9 and GAD-7, we have added “upon rescaling the mean scores to summed scores” to clarify how we rescaled the scores.

Comment:

"The design effect, defined as 1+ (average cluster size-1)×ICC". This should be explained in the Methods (and in the caption of Table 1 if you include the design effect in this table) rather than introducing it in the Results. In the current manuscript the text first refers to the design effect (first paragraph results, Table 1) and only explains it later.

Our response:

Thank you for the comment. We have revised Table 1 (p. 10) and the Methods accordingly (p. 12).

Comment:

“.... Hayes [60, p.189] recommended ...’”. The reference should be reference 61 rather than 60.

Our response:

The relevant correction has now been made (p. 15 and 34).

Comment:

The manuscript needs some language editing. E.g.:

- ".... changes in emotion regulation strategies were associated with changes in changes in depressive symptoms ...";

- "Such an increase was found to be more significantly than did participants who ...";

- ".... confounding between-person factors from the within-person associations ..."Our response:

Thank you for the comment. We have conducted language editing throughout the manuscript (e.g., p. 5, 6).

Reviewer #3

Comment:

Table 1 should include descriptive measures of the study, not just M and SD.

Our response:

Thank you for the comment. We have now included the means, standard deviations, maxima, minima, ranges of the scales, intraclass correlations, and design effects in Table 1 (p. 10).

Comment:

In Table 1, “*” should be added to describe ICC significance (e.g., *p < .05, **p < .01, ***p < .001.). Also, add info regarding how you calculated design effect.

Our response:

Thank you for the comment. We have added “*” to indicate the significance of the ICC (p. 10 and 18). We have also added a note at the bottom of Table 1 to define design effect (p. 10).

Comment:

Under “Data Analysis,” replace the word “mindfulness” by “dispositional mindfulness.”

Our response:

Thank you for the comment. The correction has been made (p. 12).

Comment:

For reference #54, include the retrieving day. Sometimes, the link is no longer available.

Our response:

Thank you for the comment. We have updated the link and added a retrieving day (p. 33).

10 Sep 2020

Dispositional Mindfulness and Mental Health in Chinese Emerging Adults: A Multilevel Model with Emotion Dysregulation as a Mediator

PONE-D-20-06030R2

Dear Dr. Ke,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Therese van Amelsvoort

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

10 Nov 2020

PONE-D-20-06030R2

Dispositional Mindfulness and Mental Health in Chinese Emerging Adults: A Multilevel Model with Emotion Dysregulation as a Mediator

Dear Dr. Ke:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Therese van Amelsvoort

Academic Editor

PLOS ONE