Effects of online mindfulness-based stress reduction training on depression and anxiety symptoms among psychiatric healthcare workers in a randomized controlled trial: the mediating role of emotional suppression

Abstract

Background

Psychiatric healthcare workers experience prolonged exposure to psychological trauma and negative emotions, which ultimately predisposes them to mental health issues. The effectiveness of online mindfulness-based stress reduction training on depression and anxiety symptoms among psychiatric healthcare workers (HCWs) and the possible mechanisms that benefit these outcomes remain unclear.

Methods

This randomized controlled trial evaluated the effect of an 8-week online MBSR among psychiatric HCWs (165 participants) at the mental health center in Shenyang, China from August to November 2022, compared with the waitlist control group. The effective sample size that actually completed the intervention was 135 (74 participants in the intervention group and 61 in the control group), and the loss-to-follow-up rate was approximately 18%. Participants completed assessments of depression and anxiety symptoms (Hospital Anxiety and Depression Scale, HADS), and emotional suppression (Courtauld Emotional Control Scale, CECS) at baseline and week 8 online via the “Wenjuanxing” Platform. Intervention outcomes were estimated via analysis of covariance (ANCOVA). Multiple linear regression and mediation models with bias-corrected bootstrapping were performed to explore the mediating role of emotional suppression.

Results

Compared with those in the control group, the participants in the online MBSR group presented lower emotional suppression (P < 0.001), symptoms of depression (P < 0.001) and anxiety (P = 0.002). The decrease in emotional suppression mediated the effects of online MBSR on 8-week symptoms of depression (indirect effect = − 1.106, 95% CI [− 1.683, − 0.600]) and anxiety (indirect effect = − 1.182, 95% CI [− 1.731, − 0.706]).

Conclusions

Online MBSR training could be utilized to alleviate depression and anxiety symptoms in psychiatric HCWs. It might weaken depression and anxiety symptoms by reducing emotional suppression. Emotional suppression might be the key change process of the effects of online MBSR on depression and anxiety symptoms in psychiatric HCWs.

Trial registration

The clinical trial registration code was obtained from the Chinese Clinical Trial Registry (date of registration: 02/08/2022, registration number: ChiCTR2200062347).

Supplementary Information

The online version contains supplementary material available at 10.1186/s12888-025-06967-1.

Introduction

Background

Depression and anxiety symptoms are common mental health problems worldwide. The average prevalence of depression worldwide is approximately 4.4% [1]. In 2019, an estimated 301 million people worldwide suffered from anxiety [2]. In 2021, depression was the second highest cause of years lived with disability (YLDs) worldwide (56.3 million YLDs), and anxiety was sixth (42.5 million YLDs) [3]. A meta-analysis revealed that 31% of healthcare workers reported depressive symptoms, and 31% reported anxiety symptoms, significantly higher than those reported by the general population (depression: 25%; anxiety: 27%) [4]. Psychiatric HCWs were more likely to experience depression and anxiety symptoms than HCWs from other specialties [5]. The incidence of depression symptoms (17.74%) in Chinese psychiatric HCWs was higher in comparison with the Chinese general population [6]. Anxiety and depression symptoms among psychiatric HCWs would negatively impact their mental health, the quality of mental health care services for patients, and healthcare systems (e.g., high turnover rates, absenteeism, reduced productivity, and increased recruitment costs) [7].

Psychiatric HCWs face prolonged high-pressure environments and frequent verbal and physical aggression from patients [8]. Prolonged exposure to psychological trauma and negative emotions increases the risk of mental health problems such as anxiety and depression symptoms [8]. Psychiatric HCWs often ignore their mental well-being, as societal expectations assume their proficiency in managing stress and emotional challenges, whereas stigma prevents them from seeking help [9, 10]. Given that it takes years to qualify as a psychiatric HCW, there is no indication that the supply of psychiatric HCWs will increase. Therefore, psychological training for psychiatric HCWs to alleviate anxiety and depression symptoms is warranted, enabling them to deliver quality care that meets increasing patient needs.

Mindfulness and MBSR training

Mindfulness-Based Stress Reduction (MBSR) is a mindfulness-based program designed to reduce stress and improve health and quality of life by teaching individuals to focus on the present moment in a non-judgmental way [11, 12]. Owing to the development of mHealth, online delivery, which expands access to MBSR, is emerging [13]. A meta-analysis revealed that mindfulness-based training such as stress reduction programs, apps, meditation and training, had small to large effects on anxiety symptoms, burnout, stress, depression symptoms, psychological distress and job strain outcomes in the healthcare population [14]. The cost of individualized, one-on-one treatment is still a deterrent for many. In this context, mindfulness-based stress reduction training is a low-cost technique that can be delivered online, without requiring a patient role, to proactively manage symptoms of depression and anxiety within the population [15, 16]. Psychiatric HCWs often face heavy workloads and inflexible schedules, and stigma deters them from seeking help. "Self-care" may be a valuable supplement to the clinical training of psychiatric HCWs [9]. Online mindfulness-based stress reduction training provides occupational groups such as psychiatric HCWs with a feasible and practical opportunity to engage in self-care, with significant public health implications. However, research on the efficacy of online mindfulness-based stress reduction training for this population remains limited.

However, our understanding of the mechanisms by which MBSR conferred benefits on depression and anxiety symptoms is lacking. The term mechanism refers to underlying psychological, social, and neurophysiological processes or mediators through which therapeutic change occurred [17, 18]. Some HCWs had low acceptance of MBSR, in part due to lack of clear understanding of the process by which MBSR training took effect [19]. Furthermore, while MBSR has been shown to reduce depression and anxiety symptoms in HCWs, results have been mixed [20]. Several studies showed that HCWs did not experience significant changes in mood disorders such as depression and anxiety symptoms after MBSR practice [20]. Research on the mechanisms of change underlying the effects of MBSR on psychological functioning is necessary. Understanding how interventions operate could help identify those facets that might be essential in driving psychological outcomes, which is beneficial for developing online training that is more effective and accessible to psychiatric HCWs [17].

Theoretical framework

The Conditional Process Model (CPM) of mindfulness and emotional regulation proposed that individuals who engaged in MBI provided buffer space for rapidly generated emotions by breaking the automatic reaction process, thereby engaging in emotional regulation [21]. A meta-analysis based on 76 studies of the MBI for youth suggested that emotional regulation might be a mediating variable for the effect of MBI on mental health outcomes [22]. Emotional suppression is a behaviorally oriented, response-focused emotional regulation strategy involving the conscious inhibition of negative emotional expression [23, 24]. Emotional suppression reduced emotional expression behavior, but did not reduce the psychological experience of emotion, and even exacerbated adverse psychological and physiological reactions to emotion [24]. The individual’s inhibition and non-acceptance of negative emotions might lead to the persistence of negative emotions, eventually leading to emotional disorders. Emotional suppression has, for example, been linked with higher levels of depression and anxiety symptoms [25]. In light of the CPM, MBSR might influence symptoms of individual anxiety and depression disorders by dampening emotional suppression, similar to previous studies. An experimental study conducted by Fung et al. (2019) revealed that the effects of school-based mindfulness interventions on perceived stress were mediated by reductions in expressive suppression and rumination [23]. A cross-sectional study among 1067 Chinese adolescents revealed that mindfulness could indirectly affect their depression symptoms through emotional suppression [26]. Emotional suppression seems to be a mechanism through which MBSR influences psychological outcomes.

Research objectives

Although the mental health of psychiatric HCWs has received increasing attention, few previous studies have investigated the effectiveness of online MBSR training among psychiatric HCWs and the mechanisms of change. Therefore, this study aims to explore the effectiveness of online MBSR training on depression and anxiety symptoms as well as emotional suppression in psychiatric HCWs, and further to test whether a possible mechanism—emotional suppression—plays a mediating role in the relationship between online MBSR training and depression and anxiety symptoms, providing a scientific basis for developing psychological training for psychiatric HCWs and achieving targeted mental health management.

Methods

Trial design

This single-center, parallel-group randomized controlled trial was conducted at Shenyang Mental Health Center, China, between August and November 2022. The study was registered with ClinicalTrials.gov (ChiCTR2200062347) and received ethical approval from the Research Ethics Committee of China Medical University (LJKMR20221215). The trial assessed the efficacy of an online MBSR training for depression and anxiety symptoms in psychiatric HCWs compared to the waitlist control group. Eligible participants were screened and enrolled through interviews conducted by a research assistant. Using SPSS v21.0, 165 qualified participants were randomly allocated at a 1:1 ratio to the online MBSR group or the waitlist control group through a computer-generated random number sequence (uniform distribution, range 0–1). An independent statistician generated the randomization sequence, which was sealed in sequentially numbered, opaque envelopes. Group assignments remained fixed after allocation. Data were collected via an electronic platform and managed by the research team. The research assistant (research implementer) was blinded to participants’ outcomes to prevent potential bias. Assessment personnel and statisticians remained blinded to group assignments throughout the study.

Participants

Settings and locations

Participants were recruited in Shenyang Mental Health Center, a regional tertiary psychiatric hospital in Northeast China. Participants were recruited through a research assistant (a psychiatrist from the mental health center laboratory) by distributing electronic recruitment posters via hospital WeChat groups (a social media group) and posting recruitment posters in staff lounges and break rooms. Potentially eligible respondents were enrolled in a dedicated WeChat group for study coordination.

Eligibility criteria

Participants eligibility criteria included [27]: (1) had a Certificate of Licensed Practicing Qualification Certificate; (2) had worked at mental health center for least one year; (3) had been working in roles that involved direct contact with patients for a minimum of 1 day per week; (4) possessed the ability to use an Apple or Android smartphone or tablet or computer with the WeChat app downloaded; (5) had no mindfulness-based training or other similar psychological interventions within the past 6 months; (6) participants who had clinical depression or anxiety disorders were excluded; and (7) provided informed consent to participate in the study.

Sample size

The sample size was calculated using effect sizes from previous related publications [28], with G*Power statistical software version 3.1 [29, 30]. The power analysis yielded an approximation for a total of 128 participants when we expected a medium effect size of 0.25 (α = 0.05; power, 80%) in a 2-group comparison. Assuming an anticipated dropout rate of 20%, we increased our sample size to 165 participants, with 82 participants in the intervention group and 83 participants in the control group.

Online MBSR group

Before the intervention, the implementers of the intervention received unified training and had a certain grasp of the intervention content, form, method, data collection and ethical requirements of this research. The project was assisted and guided by instructors holding national-level psychologist certifications and a licensed laboratory psychiatrist. The training framework was mainly based on the Mindfulness-Based Stress Reduction Theory (MBSR) created by Professor Kabat-Zinn [11]. The content was derived from the works of Professor Linda Carlson (translated by Dr. Yujing Sun), encompassing text, image, audio, and video resources for teaching mindfulness-based techniques [31, 32]. The intervention content included: first awareness mindfulness, body scan, mindfulness yoga, sitting meditation, mindfulness walking, mindfulness meditation, loving-kindness meditation, etc. (see Table 1 for a brief description of each session’s content and activities). The research assistant (a Licensed Laboratory Psychiatrist) pushed weekly MBSR training session to the participants in WeChat group every Monday, and reminded the participants to click to watch. Participants were allowed to practice repeatedly according to their personal needs. Participants gave feedback in the WeChat group after completing the course each week. Research assistants conducted WeChat check-in statistics for each participant who took the course. Considering the irregular working hours of psychiatric HCWs, the training completion criterion was set as: participants attended at least one audio and video-guided mindfulness-based stress reduction training session per week.

Table 1.

A brief description of each session’s content and activities of online MBSR

Module	Themes	Practice
1	Introduction to mindfulness and first experience of mindfulness	Introduce the basic concepts, theory, benefits and recent researches of mindfulness and MBSR; the online MBSR training program preparation and orientation (aims, contents, and structure) Practice mindful breathing meditation (awareness of breath)
2	Pay attention to your body	Introduce and practice body scan technique to help participants feel the moment and relax
3	Awareness in sports	Introduce the origin and characteristics of mindfulness yoga (mindfulness yoga teaching videos and mindfulness yoga action diagrams) Practice mindfulness yoga
4	Thought is not reality	Explain the issues that need attention in mindfulness yoga practice last week Introduce sitting meditation and how to enter meditation mode. Help participants interrupt behavioral habits, consciously respond to stress instead of reactivity Practice sitting meditation, and keep a record of negative emotions over the past week
5	Accept difficulties	Introduction to walking meditation and its role in relieving negative emotion Practice walking meditation to focus on the feeling of walking when taking a walk after a meal, and record pleasurable feelings
6	Emotional management by mindfulness	Introduce and practice mindfulness meditation Listen to your emotions, understand the relationship between the body and emotions. Encourage to deal with negative emotions in life in a mindful way
7	Being compassionate to ourselves and our patients	Introduce and practice loving kindness meditation (attention to any stressful events during a week and being kindness toward it instead of any efforts to change it) Record the negative emotions you overcome every day and encourage yourself
8	Continued mindfulness practice	Review and summarize the entire 8-week MBSR training program Encourage to continue MBSR practice in daily life and provide online learning and practice resources

Waitlist control group

Participants in the control group received no intervention during the intervention period. They will receive the full 8-session online MBSR intervention upon completion of all intervention procedures by the intervention group after the follow-up survey if they are still interested.

Measures

The questionnaire was uploaded to the online professional survey platform what was named “Wenjuanxing” (www.wjx.cn) for data collection questionnaire surveys. The survey link online was distributed to the participants via the social media platform WeChat. After obtaining informed consent, participants completed self-reported assessments of sociodemographic/work-related characteristics, depression, anxiety symptoms (HADS), and emotional suppression levels (CECS) at baseline and after the end of the training.

Sociodemographic and work-related characteristics

Psychiatric HCWs reported their age, gender (Male, Female), marital status (Married/cohabited, Single/divorced/widowed/separated), education background (Junior college or lower, College, Graduate or higher), working experience (≤ 10 years, > 10 years), profession (Doctor, Nurse), professional title (Junior, Intermediate, Senior), monthly income (≤ 5000 yuan, > 5000 yuan), weekly working time (≤ 40 h, > 40 h), night duty (No, Yes).

Depression and anxiety symptoms

The 14-item Hospital Anxiety and Depression Scale (HADS) was used to measure depression and anxiety symptoms [33]. HADS consisted of a 7-item anxiety subscale and a 7-item depression subscale. Scores ranged from 0 to 21 in each subscale. A score of 8 or higher indicated anxiety or depression symptoms [34]. It has been used in research to identify symptoms of anxiety and depression in HCWs [35]. The HADS was a reliable tool with a Cronbach’s α by subscale of 0.832 for anxiety and 0.845 for depression in this study.

Emotional suppression

The Courtauld Scale of Emotional Control (CECS) [36] with 21 items was used to measure individuals’ emotional suppression. The scale used a 4-point Likert scale, ranging from “1 = almost never” to “4 = almost always.” Five items (5, 9, 15, 16, 18) presented emotional expressions that needed to be reverse scored. Higher total scores correspond to greater emotional suppression. Li et al. (2015) revised the Chinese version of the CECS with good internal consistency [37]. It has been well applied in populations such as patients and adolescents, etc. [37, 38]. In this study, Cronbach’s alpha was 0.910 for the whole scale.

Adverse events

The research assistant assessed for safety and adverse events by informing participants that if any trauma resurfaced or other difficult emotions occurred during the course, the participant should contact the research assistant (who is also a licensed psychiatrist). The participant and research assistant would determine together the safest course of action or terminate the intervention [39].

Statistical analysis

All statistical analyses were performed using SPSS 21.0 software (IBM, Asia Analytics Shanghai). Independent samples t-tests and chi-square tests of homogeneity were used to determine whether there were differences between groups at baseline. Analysis of covariance (ANCOVA) models controlling for baseline criterion scores were used to examine the effect of online MBSR training on emotional suppression, depression and anxiety symptoms at post-intervention between groups. We calculated linear regressions and implemented the Preacher and Hayes method using the PROCESS macro with 5,000-iteration bias-corrected bootstrapping [40] to test whether change in emotional suppression from baseline to week 8 mediated the effects of online MBSR on depression and anxiety symptoms at week 8. Using PROCESS, we tested a mediation model with online MBSR (Online MBSR vs. Control) as the independent variable, change in emotional suppression as a mediator variable, and either depression symptoms or anxiety symptoms as the dependent variable. We adjusted these models for baseline emotional suppression and baseline depression symptom or anxiety symptom. Unstandardized regression coefficients were evaluated as a measure of effect size [41]. Bootstrapped estimates of the indirect effects were interpreted with bias-corrected 95% confidence intervals (CI), in which any CIs not containing zero indicated a significant mediation effect. A two-sided P-value < 0.05 was considered statistically significant.

Results

Baseline characteristic

A total of 165 psychiatric HCWs were enrolled, of which 82 participated in the online MBSR training group and 83 participated in the waitlist control group. 8 participants in the intervention group and 22 participants in the control group withdrew the program due to work time conflicts, lack of interest, and the belief that this exercise was not of great use to themselves, etc. Ultimately 135 participants who completed the program were included in the analysis, including 74 participants in the intervention group and 61 participants in the control group (Fig. 1). Participants were recruited in August 2022. The final assessment was completed in November 2022. The trial was stopped as planned after reaching the target sample size and completing all follow-up assessments. Participants in the intervention and control groups who completed the program did not statistically differ in any baseline outcome variable or demographic characteristics (P > 0.05) (Table 2). No adverse events were reported to the research assistant. The study included participants with a mean age of 37.56 years in the waitlist control group and 36.36 years in the online mindfulness-based stress reduction (MBSR) group. Both groups were predominantly female (65.6% control, 75.7% MBSR) and married or cohabiting (75.4% control, 75.7% MBSR). The MBSR group included 33.8% of the participants with intermediate titles and 23.0% with senior titles, and the control group included 21.3% with intermediate titles and 18.0% with senior titles. A total of 16.2% of the participants in the MBSR group and 6.6% in the control group held graduate or higher education degrees. In the MBSR group, 40.5% of the participants were doctors, and 27.9% were in the control group. In the MBSR group, 67.6% of the participants reported night duties, and 62.3% of the participants in the control group reported night duties. Both groups had similar proportions of participants with over 10 years of work experience (54.1% MBSR, 59.0% control) and monthly incomes exceeding 5000 yuan (55.4% MBSR, 62.3% control). However, the control group had a slightly greater percentage of workers working > 40 h per week (50.8% vs. 44.6%). The baseline emotional suppression scores were greater in the MBSR group (mean = 50.54) than in the control group (mean = 47.57), whereas the anxiety (MBSR = 11.36, control = 11.26) and depression (MBSR = 11.68, control = 11.43) scores were comparable.

Fig. 1.

Consort diagram

Table 2.

Baseline characteristics of psychiatric HCWs across intervention group and control group

Characteristic	Waitlist control group (N = 61)	Online MBSR (N = 74)	t/χ2	P
Age (years), mean (SD)	37.56 (8.93)	36.36 (7.14)	0.862	0.390
Gender, n (%)			1.661	0.197
Male	21 (34.4)	18 (24.3)
Female	40 (65.6)	56 (75.7)
Marital status, n (%)			0.001	0.971
Married/cohabited	46 (75.4)	56 (75.7)
Single/divorced/widowed/separated	15 (24.6)	18 (24.3)
Education background, n (%)			4.706	0.095
Junior college or lower	9 (14.8)	5 (6.8)
College	48 (78.7)	57 (77.0)
Graduate or higher	4 (6.6)	12 (16.2)
Working experience, n (%)			0.335	0.563
≤ 10 years	25 (41.0)	34 (45.9)
> 10 years	36 (59.0)	40 (54.1)
Profession, n (%)			2.366	0.124
Doctor	17 (27.9)	30 (40.5)
Nurse	44 (72.1)	44 (59.5)
Professional title, n (%)			4.225	0.121
Junior	37 (60.7)	32 (43.2)
Intermediate	13 (21.3)	25 (33.8)
Senior	11 (18.0)	17 (23.0)
Monthly income (RMB), n (%)			0.654	0.419
≤ 5000 yuan	23 (37.7)	33 (44.6)
> 5000 yuan	38 (62.3)	41 (55.4)
Weekly working time, n (%)			0.520	0.471
≤ 40 h	30 (49.2)	41 (55.4)
> 40 h	31 (50.8)	33 (44.6)
Night duty, n (%)			0.410	0.522
No	23 (37.7)	24 (32.4)
Yes	38 (62.3)	50 (67.6)
CECS, mean (SD)	47.57 (11.45)	50.54 (8.47)	−1.728	0.086
HADS-D, mean (SD)	11.43 (3.88)	11.68 (3.39)	−0.399	0.691
HADS-A, mean (SD)	11.26 (3.53)	11.36 (3.35)	−0.173	0.863

Abbreviations MBSR mindfulness-based stress reduction, HCWs healthcare workers, HADS-A Hospital Anxiety and Depression Scale-Anxiety subscale, HADS-D Hospital Anxiety and Depression Scale-Depression subscale, CECS Courtauld Emotional Control Scale, SD standard deviation

Intervention effects on outcomes

Intervention effects on psychiatric HCWs’ outcomes at week 8 was showed in Table 3. Compared to those in the control group, psychiatric HCWs in the intervention group reported fewer anxiety symptoms (adjusted means: 10.36 [95%CI 9.83, 10.90] vs. 11.66 [95%CI 11.07, 12.25], η² = 0.073, P = 0.002), fewer depression symptoms (adjusted means: 9.99 [95%CI 9.43, 10.56] vs. 12.04 [95%CI 11.42, 12.67], η² = 0.148, P < 0.001), and lower emotional suppression (adjusted means: 43.42 [95%CI 42.01, 44.82] vs. 51.18 [95%CI 49.63, 52.72], η² = 0.292, P < 0.001). Partial ƞ² was used as an index of effect size. η_p² > 0.14 represented a large effect; 0.06–0.14 represented a medium effect, and 0.01–0.06 represented a small effect [30].

Table 3.

Effects of online MBSR on emotional suppression, anxiety, and depression symptoms

	Waitlist control group (N = 61), Adjusted Mean (95% CI)	Online MBSR group (N = 74), Adjusted Mean (95% CI)	Adjusted Mean Difference (SE)	95% CI	P	ηp2
HADS-A	11.66 (11.07, 12.25)	10.36 (9.83, 10.90)	−1.30 (0.40)	(−2.10, −0.50)	0.002	0.073
HADS-D	12.04 (11.42, 12.67)	9.99 (9.43, 10.56)	−2.05 (0.43)	(−2.89, −1.20)	< 0.001	0.148
CECS	51.18 (49.63, 52.72)	43.42 (42.01, 44.82)	−7.76 (1.06)	(−9.84, −5.67)	< 0.001	0.292

Results of ANCOVA models controlling for baseline anxiety, depression symptoms and emotional suppression level

Abbreviations MBSR mindfulness-based stress reduction, CI Confidence Interval, SE standard deviation, HADS-A Hospital Anxiety and Depression Scale—Anxiety Symptoms, HADS-D Hospital Anxiety and Depression Scale—Depression Symptoms, CECS Courtauld Emotional Control Scale

Depression symptom effects mediated through change in emotional suppression

Compared to the control group, online MBSR training was associated with the significant reduction in 8-week depression symptoms (B = − 2.047, SE = 0.427, 95% CI [− 2.892, − 1.202], P < 0.001), controlling for baseline depression symptoms. Decreases in emotional suppression (B = 0.142, SE = 0.034, 95% CI [0.076, 0.209], P < 0.001) were significantly associated with fewer depression symptoms at week 8, controlling for baseline emotional suppression and depression symptoms. With both online MBSR (Online MBSR vs. Control) and changes in emotional suppression simultaneously entered in the regression model, the effect of online MBSR training on depression symptoms became non-significant (B = − 0.929, SE = 0.485, 95% CI [− 1.887, 0.030], P = 0.057), indicating a potential mediation effect. The mediation regression model assessing the effect of the online MBSR training on depression symptoms revealed a significant indirect effect through change in emotional suppression. Participants in the online MBSR training group experienced a 1.106-point reduction in depression symptoms indirectly through decreased emotional suppression, with 95% bias corrected confidence interval of (−1.683, −0.600) (Table 4). The full model accounted for 57.8% of the variance in depression symptoms, and 54.5% of the total effect was attributed to the indirect effect. Statistical power calculated by G*Power software version 3.1 suggests that a sample size of 135 participants in the multiple linear regression predicting depression symptom with 5 independent variables will have more than 90% power to detect an effect size ƒ² of 0.50, which is considered a large-sized effect according to Cohen’s (2013) guidelines [30].

Table 4.

Effects of online MBSR on depression symptom mediated by Emotional suppression

Variable	B	SE	95%CI		P	VIF
Regression 1: effects on depression symptom at 8 weeks
Online MBSR	−2.047	0.427	(−2.892, −1.202)		< 0.001	1.001
Baseline depression symptom	0.657	0.059	(0.540, 0.774)		< 0.001	1.001
Model summary: depression symptom	R2 = 0.519; F (2,132) = 71.304; P < 0.001
Regression 2: effects on mediator (emotional suppression Δ)
Online MBSR	−7.763	1.058	(−9.855, −5.670)		< 0.001	1.024
Baseline depression symptom	0.033	0.162	(−0.287, 0.354)		0.837	1.252
Baseline emotional suppression	−0.126	0.059	(−0.243, −0.009)		0.035	1.279
Model summary: emotional suppression Δ	R2 = 0.334; F (3,131) = 21.901; P < 0.001
Regression 3: effects on depression symptom accounting for mediator (emotional suppression Δ)
Emotional suppression Δ (baseline to 8 weeks)	0.142	0.034	(0.076, 0.209)		< 0.001	1.502
Online MBSR	−0.929	0.485	(−1.887, 0.030)		0.057	1.445
Baseline depression symptom	0.658	0.063	(0.534, 0.781)		< 0.001	1.253
Baseline emotional suppression	0.013	0.023	(−0.033, 0.059)		0.567	1.323
Model summary: depression symptom	R2 = 0.578; F (4,130) = 44.434; P < 0.001
	Effect	Bootstrapped SE		Bias-corrected 95%CI
Indirect effect of online MBSR on depression symptom accounting for emotional suppression Δ	−1.106	0.278		(−1.683, −0.600)

Abbreviations Δ change, MBSR mindfulness-based stress reduction, CI Confidence Interval, SE standard deviation, VIF variance inflation factor

Anxiety symptom effects mediated through change in emotional suppression

Compared to the control group, online MBSR training was associated with a significant reduction in 8-week anxiety symptoms (B = − 1.299, SE = 0.403, 95% CI [− 2.097, − 0.502], P = 0.002), controlling for baseline anxiety symptoms. Decreases in emotional suppression (B = 0.153, SE = 0.031, 95% CI [0.091, 0.215], P < 0.001) were significantly associated with fewer anxiety symptoms at week 8, controlling for baseline emotional suppression and anxiety symptoms. With both online MBSR (Online MBSR vs. Control) and changes in emotional suppression simultaneously entered in the regression model, the effect of online MBSR training on anxiety symptoms became non-significant (B = − 0.107, SE = 0.449, 95% CI [− 0.995, 0.780], P = 0.811), indicating a potential mediation effect. The mediation regression model assessing the effect of the online MBSR training on anxiety symptoms revealed a significant indirect effect through change in emotional suppression. Participants in the online MBSR training group experienced a 1.182-point reduction in anxiety symptoms indirectly through decreased emotional suppression, with the 95% bias corrected confidence interval of (−1.731, −0.706) (Table 5). The full model accounted for 58.5% of the variance in anxiety symptoms, and 91.0% of the total effect was attributed to the indirect effect. Also, a sample size of 135 participants in the multiple linear regression predicting anxiety symptoms with 5 independent variables will have more than 90% statistical power to detect an effect size ƒ² of 0.51.

Table 5.

Effects of online MBSR on Anxiety Symptom Mediated by Emotional suppression

Variable	B	SE	95%CI		P	VIF
Regression 1: effects on anxiety symptom at 8 weeks
Online MBSR	−1.299	0.403	(−2.097, −0.502)		0.002	1.000
Baseline anxiety symptom	0.664	0.059	(0.548, 0.781)		< 0.001	1.000
Model summary: anxiety symptom	R2 = 0.509; F (2,132) = 68.390; P < 0.001
Regression 2: effects on mediator (emotional suppression Δ)
Online MBSR	−7.721	1.057	(−9.811, −5.630)		< 0.001	1.026
Baseline anxiety symptom	0.134	0.172	(−0.206, 0.474)		0.437	1.275
Baseline emotional suppression	−0.142	0.059	(−0.260, −0.024)		0.018	1.303
Model summary: emotional suppression Δ	R2 = 0.337; F (3,131) = 22.184; P < 0.001
Regression 3: effects on anxiety symptom accounting for mediator (emotional suppression Δ)
Emotional suppression Δ (baseline to 8 weeks)	0.153	0.031	(0.091, 0.215)		< 0.001	1.508
Online MBSR	−0.107	0.449	(−0.995, 0.780)		0.811	1.444
Baseline anxiety symptom	0.649	0.062	(0.527, 0.771)		< 0.001	1.281
Baseline emotional suppression	0.018	0.022	(−0.025, 0.061)		0.404	1.360
Model summary: anxiety symptom	R2 = 0.585; F (4,130) = 45.896; P < 0.001
	Effect	Bootstrapped SE		Bias-corrected 95%CI
Indirect effect of online MBSR on anxiety symptom accounting for emotional suppression Δ	−1.182	0.264		(−1.731, −0.706)

Abbreviations Δ change, MBSR mindfulness-based stress reduction, CI Confidence Interval, SE standard deviation, VIF variance inflation factor

Discussion

The findings supported and extended the evidence for the effectiveness of online MBSR. 81.8% of participants completed the program. 18.2% withdrew from the program due to work time conflicts, lack of interest, and the belief that this exercise was not of great use to them, etc. Online MBSR training demonstrated significant efficacy in reducing depression (moderate-to-large effect sizes) and anxiety symptoms (moderate-to-large effect sizes), as well as emotional suppression (large effect sizes) among psychiatric HCWs compared with the control group [30]. A previous study revealed that after an 8-week online MBI, nurses’ perceived stress levels, depression, and anxiety symptoms were significantly improved, with a compliance rate of 70.12% [42, 43]. Watson et al. (2021) reported that the MBI program has the potential to benefit psychiatric HCWs, personally by reducing stress and emotional reactivity, and professionally by promoting deep listening and communication [44]. MBSR training was shown to influence emotion regulation including avoidance, thought suppression, and emotional suppression, with small effect size of d = 0.28, and cognitive reappraisal in non-clinical populations [45]. MBSR was useful in reducing emotional suppression, possibly because mindfulness focuses on the feeling of the moment, accepting emotions without judgment rather than suppressing them [46]. Through the process of mindfulness, one was able to disidentify with the contents of consciousness (i.e., one’s thoughts) and view his or her moment-by-moment experience with greater clarity and objectivity. This process was termed reperceiving by Shapiro et al. (2006) as it involved a fundamental shift in perspective [47]. Through reperceiving, MBI reinforced non-evaluation of the experience and weakened the emotional bias toward stimulus perception [47]. With regard to psychiatric HCWs, MBSR might help facilitate the process of ‘reperceiving’ in the workplace to reduce the suppression of negative emotions faced at work.

Mediation analyses revealed that changes in emotional suppression indirectly facilitated the alleviation of severity of anxiety/depression symptoms during the online MBSR training. The indirect effects of emotional suppression on depression and anxiety symptoms were robust, accounting for 54.4% and 91.0% of the total effect of online MBSR training on these outcomes, respectively. The findings offered a preliminary explanation for how MBSR training operated, with a focus on the extinction of maladaptive habits and disengagement from negative states of emotion, similar to previous studies. A previous study found that the effects of MBSR on reduction in perceived stress of ethnic minority youth were mediated by reductions in expressive suppression [23]. Changes in emotional suppression were associated with depression and anxiety symptoms. Similarly, Dawel et al. (2021) conducted a three-time-point prospective longitudinal study of 704 adults in Australia, indicating that high levels of emotional suppression might lead to more serious depression and anxiety symptoms [48]. The use of avoidant-oriented coping or emotional suppression appeared to preempt more effective coping efforts and/or induce intrusion of maladaptive thoughts and emotions [49]. In addition, inhibition of emotion required psychophysiological effort, and this additional cognitive effort negatively impacted psychological functioning [50]. In addition to emotional suppression, other unexamined emotion regulation mechanisms in this study also warrant attention. Unlike emotion suppression, a late-stage reaction-focused strategy [24], cognitive reappraisal is an early-stage in the emotion process, cause-focused strategy that alters emotional experiences by reinterpreting events [51]. A cross-sectional study suggested that positive reappraisal may be a key process linking mindfulness to reduced depression, and anxiety symptoms [52]. Following mindfulness-based techniques, levels of cognitive reappraisal have been shown to increase in patients with major depressive disorder [53]. Self-compassion is an emotion-regulation strategy in which the individual engages in self-kindness, particularly after a mistake, failure, and/or rejection [54]. Among oncology nurses, changes in self-compassion mediated the impact of mindfulness-based intervention on burnout, depression, anxiety, stress and satisfaction with life [55]. Future research should incorporate more outcomes such as stress responses and sleep quality, and examine additional emotion regulation mechanisms such as cognitive reappraisal and self-compassion to better understand and interpret the findings.

Overall, among the psychologically knowledgeable group such as psychiatric HCWs, online MBSR training may be a low-cost, feasible, and effective method for alleviating mental health issues. At present, there is few mature and reliable psychological training for psychiatric HCWs. In addition to online MBSR training, interventions that promote emotional expression, such as Emotion-Focused Therapy (EFT), art therapy, and group peer support, could be integrated into mental health training for psychiatric HCWs. EFT facilitates the healthy identification, experience, and expression of emotions, effectively alleviating anxiety and depression symptoms [56]. Art therapy, using non-verbal mediums like painting and music to express emotion, has proven effective in reducing anxiety and depression symptoms among cancer patients [57]. Group peer support interventions enhance emotional expression and coping [58], improving mental health outcomes and aiding in the recovery of individuals with mental illnesses [59]. Understanding the operation of the intervention could help develop better strategies to trigger key change processes [17, 18, 60]. Experimental dismantling studies could be conducted in the future to determine which components of MBSR training are driving these mechanism effects [61], in order to develop practical and feasible intervention strategies for psychiatric HCWs.

Several limitations of the study should be considered. Firstly, the single-site trial was conducted at a tertiary psychiatric hospital in Liaoning Province with a sample of psychiatric HCWs. The program may help other healthcare workers with similar stressors. The generalizability of the findings in primary care was limited due to variations in resource availability, service content, and working conditions between urban mental health centers and rural or private clinics. Secondly, self-report measures may introduce recall and response biases. Our study attempted to minimize bias by using the scales well-validated in the Chinese population, such as CECS and HADS, and training the investigators uniformly before investigation. Thirdly, Participant attrition might reduce the sample size and compromise the ability to detect significant effects. However, with an anticipated dropout rate of 20%, a post hoc power analysis conducted using G*Power indicated that the statistical power exceeded 80%, suggesting that the study was likely adequately powered to detect the expected effects. Although attrition did not introduce bias in primary outcomes (see Supplementary File 1), the significantly higher dropout rate among older participants may limit the generalizability of findings to this subgroup. While sensitivity analyses confirmed the robustness of our results, future replication studies in older populations are warranted. Fourthly, intervention effects were assessed solely based on mental health outcomes measured at the post-intervention time point. Future research should incorporate long-term follow-ups to assess the sustained impact of online MBSR training among psychiatric HCWs. Fifthly, we could not rule out the possibility that changes in the depression and anxiety symptoms (outcomes) led to changes in emotional suppression (mediator) because temporal order of mediator and outcome variables could not be determined [17]. However, this explanation seemed less likely in light of some evidence showing that changes in emotional suppression preceded changes in psychological symptoms [48]. Future work could combine multiple time points and cross-lag models to determine the directionality of these changing mechanisms of MBSR training. Finally, in future research, we will further explore the fidelity of mindfulness-based stress reduction training based on the fidelity framework [62].

Conclusions

Online MBSR training alleviated depression and anxiety symptoms in psychiatric HCWs and may be a valuable addition to mental health training programs. Furthermore, it indirectly reduced these symptoms by decreasing emotional suppression. Future enhancements to mental health training strategies for psychiatric HCWs could focus on interventions alleviating emotional suppression.

Implications

Online mindfulness-based stress reduction (MBSR) training, along with interventions that promote emotional expression, could be systematically integrated into routine mental health training programs for psychiatric HCWs to mitigate symptoms of depression and anxiety.

Future mechanistic studies could be combined with active components studies through dismantling, additive, and direct component comparisons to clarify how specific factors interact with behavior change mechanisms to influence outcomes [61].

Abbreviations

MBSR

Mindfulness-based Stress Reduction

CECS

Courtauld Emotional Control Scale

HADS

Hospital Anxiety and Depression Scale

CPM

The Conditional Process Model

ANCOVA

Analysis of Covariance

VIF

Variance Inflation Factor

SE

Standard Error

Authors’ contributions

Yuekun Wu participated in implementation of the intervention, data collection, data analyzation, making tables, writing and revising the manuscript. Yue Ban was responsible for the design and adaptation of the online intervention content. Guoliang Pan and Mingyang Yao were involved in recruiting participants, on-site supervision, data collection. Li Liu participated in data analyzation, provided guidance in study design, and supervised the investigation. Ting Chen and Hui Wu contributed to the training of the project implementers, provided guidance in study design, and supervised the investigation. All authors read and approved the final manuscript.

Funding

This work was financially supported by Basic Scientific Research project of Higher education Department of Liaoning Province (LJKMR20221215) and National Natural Science Foundation of China (71673300).

Data availability

The data that support the findings of this study are available on request from the corresponding author.

Declarations

Ethics approval and consent to participate

The trial was registered with Chinese Clinical Trial Registry (registration number: ChiCTR2200062347). The study was approved by the Research Ethics Committee of China Medical University [LJKMR20221215] and performed in accordance with the Declaration of Helsinki. Written informed consent was obtained from each participant. Information collected from all participants was kept confidential and anonymous. The authors declare that they have no competing interests.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.