Effectiveness of mindfulness-based interventions on burnout, resilience and sleep quality among nurses: a systematic review and meta-analysis of randomized controlled trials

Abstract

Background

Due to the occupational characteristics of clinical work, nurses often face many challenges related to physical and psychological health. Mindfulness-based interventions are psychological intervention therapies based on mindfulness, which has been proven to effectively improve burnout, resilience and sleep quality among nurses. But there is not sufficient evidence to examine the effectiveness of mindfulness-based interventions on burnout, resilience and sleep quality among nurses.

Objective

To evaluate the effectiveness of mindfulness-based interventions in reducing burnout, resilience and sleep quality among nurses and enhance the quality of nursing medical service.

Methods

Seven electronic databases, including PubMed, EMBASE, Scopus, Web of Science, PsycINFO, CINAHL and Chinese National Knowledge Infrastructure were searched from their inception to November, 2024. Randomized controlled trials investigated the effects of mindfulness-based interventions for improving burnout, resilience and sleep quality among nurses. The quality of the included studies was assessed by the Cochrane Risk of Bias Tool v2. Meta-analyses were conducted using Review Manager 5.4 and the Grading of Recommendations, Assessment, Development, and Evaluation approach was used to assess the certainty of evidence.

Results

A total of 4622 studies were initially retrieved and 16 articles (n = 1384 individuals) were included. The meta-analysis showed that mindfulness-based interventions could effectively improve burnout (SMD = -1.43, 95% CI: -1.94 to − 0.92, P < 0.001), resilience (MD = 9.78, 95% CI: 0.38 to 19.17, P = 0.04) and sleep quality (SMD = -1.1, 95% CI: -1.79 to -0.41, P = 0.002). However, due to some moderate risk of bias and high level of heterogeneity, the overall quality of the evidence was not high.

Conclusion

Mindfulness-based interventions could improve burnout, resilience and sleep quality among nurses. This suggests that mindfulness-based interventions can provide preliminary clinical practice support for addressing mental health and well-being and work quality of nurses. But because of moderate to low certainty of evidence and some concerns of bias, more rigorous studies are needed to confirm the effectiveness of mindfulness-based interventions for improving the outcomes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12912-025-03101-0.

Introduction

Nurses are key forces in providing comprehensive and continuous health services to patients. However, the demanding nature of clinical work, characterized by high mental tension, heavy workloads, emotional labor, and irregular schedules, contributes to a high prevalence of physical and psychological health challenges [1, 2]. These problems include occupational stress, burnout, poor sleep quality, and reduced resilience, which severely affect quality of life and work efficiency among nurses [3].

Burnout manifests as a range of negative emotional experiences, often encompassing emotional exhaustion, fatigue, decreased work enthusiasm and satisfaction, and indifference to service providers [4]. Studies indicate a disproportionately high occurrence of burnout among frontline nurses. A meta-analysis of global burnout among general nurses conducted by experts reveals that 11.23% of nurses experience symptoms of burnout. This data indicates that 1 in 10 nurses worldwide suffers from high levels of burnout symptoms, urgently needing attention and intervention [5]. Burnout is closely linked to various psychological disorders, including fatigue, anxiety, depression, and sleep disorders [6]. It also negatively affects job performance and satisfaction, impeding the healthy development of the nursing workforce within hospitals.

Resilience is a process of adaptation and development that occurs in response to stress, adversity, or trauma highly associated with psychological problems [7, 8]. The concept of resilience was introduced into the nursing practice in the 21st century, with research focusing on nurses in oncology, intensive care, and psychiatry [9–11]. Resilience has long served as a protective factor for nurses coping with stressors. This quality reflects resilience to stress and is closely associated with happiness and professional quality of life among nurses [12]. A relatively high level of resilience can reduce work stress in high-pressure environments, alleviate tension and anxiety, improve work efficiency, and ultimately enhance their quality of life [13].

Studies indicate that 33–78% of nurses struggle with sleep disorders, with shift work as the primary contributing factor [14–17]. Sleep disorder among nurses leads to adverse outcomes, including chronic illness and physical fatigue. These conditions can contribute to burnout, reduced work efficiency, and a decline in job satisfaction and quality of nursing care, ultimately affecting quality of life [18, 19].

Various interventions have recently been adopted to lower the occurrence of burnout, diminished resilience, and poor sleep quality among nurses. Commonly used methods include emotion regulation training, Williams LifeSkills training, and communication skills training, relaxation techniques, resilience training, and mindfulness training [20–24]. Mindfulness-based interventions are well-established programs that have drawn increasing research interest in recent years.

Kabat-Zinn introduced mindfulness to Western society and developed a mindfulness curriculum [25]. Mindfulness is the purposeful, nonjudgmental awareness of paying attention to the present moment [26]. Over the past few decades, studies have increasingly validated the effectiveness of mindfulness-based interventions. These interventions typically consist of various approaches, including mindfulness-based stress reduction (MBSR) [27]; mindfulness-based cognitive therapy (MBCT) [28]; mindfulness-based stretching and deep breathing exercises [29]; mindfulness meditation exercises, and other modified mindfulness practices [30]. In addition, mindfulness exercises constitute five components: mindfulness meditation, body scanning, mindfulness yoga, mindfulness dieting, and mindfulness walking [31]. These mindfulness-based therapies or exercises can alleviate anxiety, depression, and insomnia due to work pressure. Moreover, they promote mental health, reduce burnout, enhance occupational well-being, and foster resilience among nurses, allowing them to conduct clinical work with calmness [32–34].

Previous systematic reviews have shown that mindfulness-based interventions can relieve psychological distress, depression, stress, and burnout among nurses; however, these studies are limited by inadequate sample sizes, insufficient follow-up, and lack of robust evidence [35–37]. O’Brien et al. [38] and Yang et al. [39] reported that the effects of MBIs could reduce the psychological distress in nurses. Several studies found evidence in support of MBSR in reducing anxiety and depression among nurses [35, 39, 40]. Two studies showed a significant decrease in depression and stress after mindfulness-based interventions [40, 41]. No systematic review and meta-analysis has comprehensively assessed the effectiveness of mindfulness-based interventions on burnout, resilience, and sleep quality among nurses. Therefore, the current study integrates relevant research to evaluate the effects of mindfulness-based interventions on burnout, resilience, and sleep quality among nurses. It also provides a foundation for supporting clinical practice and nursing management, leading to an improved quality of life for nurses.

Methods

We implemented this study on the basis of the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines [42]. We have registered the protocol in the International Prospective Register of Systematic Reviews (PROSPERO) with the registration number CRD42024610425.

Data sources and search strategies

We conducted a comprehensive search of seven electronic databases (PubMed, EMBASE, Scopus, Web of Science, PsycINFO, CINAHL and Chinese National Knowledge Infrastructure) from their inception to November, 2024. We applied the Medical Subject Headings terms and free terms for obtaining studies on the impact of mindfulness-based interventions on nurses. The search terms are as follows: “Nurses”, “Registered nurses”, “Nursing staff”, “Nursing personnel”, “Mindfulness”, “Mindfulness meditation”, “Mindfulness therapy”, “MBSR”, “Mindfulness-based stress reduction”, “Mindfulness walking”, “Mindfulness eating”, “Mindfulness-based cognitive therapy”, “MBCT”. To collect relevant literature extensively and comprehensively, we did not limit the search to the outcome variables of “burnout” “resilience” and “sleep quality”. More details about the search strategies are provided in the Supplementary Material 1 Table S1.

Inclusion criteria

We applied the PICOS framework to confirm the eligibility inclusion criteria for the studies:

P (Population): Registered nurses with continuous working ≥ 1 year.

I (Intervention): Applied mindfulness-based interventions: MBSR, MBCT, ACT, or other main mindfulness exercises (mindfulness training, mindfulness meditation, mindfulness yoga as well as mindfulness eating and mindfulness walking, loving-kindness meditation).

C (Comparison): Nurses adopted conventional psychological education care or blank intervention in the control group.

O (Outcomes): Studies that reported outcomes: burnout, resilience, or sleep quality.

S (Study Design): Randomized controlled trials were published in English or Chinese.

Exclusion criteria

• Nursing students and other healthcare professionals;

• Nurses who have previously participated in mindfulness-based interventions;

• Studies with no elements of mindfulness exercises, such as music therapy, or combination therapy with mindfulness-based interventions;

• Conference abstracts, master and doctoral theses, non-RCT study design, protocols, systematic reviews and meta-analyses.

Screening and data extraction

All searched documents were imported into a document management program Zotero, and the duplicates were eliminated. Two researchers then independently screened the retrieved records for titles and abstracts and re-screened the full text according to the inclusion and exclusion criteria. When there was disagreement, the two researchers worked together to resolve it. If a consensus could not be reached, a third researcher was consulted to confirm that the inclusion criteria were met.

The data were extracted by two researchers according to a pre-established Excel sheet template and submitted to a third researcher for review. The extracted details included: study information (name of the first author, publication year, country), participants’ information (sample size, (mean ± sd) age, age range), intervention characteristics (program, length, frequency, delivery mode, duration/follow-up) and outcomes (measurement tools). Two researchers extracted the data separately and tested the accuracy, and if there was any disagreement, they discussed it with a third researcher and reached an agreement.

Quality assessment

Two independent researchers used the Revised Cochrane Risk-of-Bias tool for randomized trials (RoB 2) to evaluate the study quality of randomized controlled trials [43]. This RoB2 assessment tool consists of five main components: randomization process、deviations from the intended interventions、missing outcome data、measurement of the outcome and selection of the reported result. All included studies were rated as “low”, “some concerns” or “high” risk of bias.

Data synthesis

The meta-analysis was performed with Review Manager 5.4 and Stata 16.0. Different measurement tools were used for the outcomes in the study. Therefore, we used the the standardized mean difference (SMD) or mean difference (MD) and 95% confidence interval (CI) to confirm the effect size. The heterogeneity of each study was assessed via the I² statistic; I² > 50% indicated high heterogeneity [44]. If I² ≥ 50%, a random effects model was conducted; I² < 50%, we employed the fixed effects model [45].

We performed a sensitivity analysis to verify the robustness and reliability of the results. When the study heterogeneity was high, we explored sources of heterogeneity by conducting subgroup analyses and meta-regression [46]. Given that more than 10 studies reported the same outcome, we used funnel plots or Egger’s regression to test publication bias and funnel plot symmetry [47]. Stata 16.0 was used to conduct the meta-regression analysis and Egger’s regression test.

Quality of evidence

The Grading of Recommendation, Assessment, Development and Evaluation system was employed through GRADEPro software to evaluate the overall evidence quality for the outcomes [48, 49]. The system contains five domains: risk of bias, inconsistency, indirectness, imprecision and reporting bias [50]. The quality of evidence included four aspects: very low, low, moderate and high.

Results

Search results

From the PRISMA flow diagram in Fig. 1, we can get a total of 4622 studies retrieved from seven databases. After removing 1155 duplicates, 3467 studies remained for the title and abstract screening. Then 3271 studies were excluded that did not meet the inclusion criteria. One study was retrieved from the manual reference list, but eventually deleted due to irrelevant outcomes. In total, 197 full-text articles were assessed for eligibility. After the full-text screening, 181 studies were excluded due to the irrelevant outcomes, lack of data, non-mindfulness-based interventions, non-RCTs, systematic reviews and meta-analyses and master and doctoral theses. Finally, 16 studies were included in this meta-analysis.

Fig. 1.

PRISMA flow diagram

Study characteristics

Participants characteristics

According to Table 1, The 16 included studies were published from 2018 to 2024. These studies were conducted in China [51–60], Jordan [61], Spain [62], US [63, 64], Iran [65–66]. This meta-analysis included a total of 1384 participants, with 715 in the intervention group and 669 in the control group. In addition, the sample sizes of the included studies ranged from 40 to 151. As for the age, the mean age in the intervention group ranged from 27.4 years (SD: 3.5 years) [55] to 34 years (SD: 7 years) [65], while the control group mean age ranged from 27 years (SD: 3.3 years) [55] to 33.17 years (SD: 7.98 years) [53]. Yuan et al. [60] did not report the mean age in the intervention and control groups and Pérez et al. [62] reported only the mean age of the total sample sizes.

Table 1.

Brief summary of the included studies

Study	Country	Sample size (IG/ CG)	Age Mean ± SD/Range (Mean)((IG/CG)	Intervention		Duration/follow-up	Outcomes
				IG (program, length, frequency, delivery mode)	CG
Wang et al. [51]	China	50/50	32.22 ± 3.45/32.62 ± 3.62	MBSR; 1 h offline theoretical and practical guidance + online guidance (not reported length) for Once a week; group, offline + online	Conventional psychological intervention	3months/baseline- 3 months	Resilience/burnout
Wei [52]	China	43/43	33.00 ± 7.24/31.90 ± 6.08	MBSR; 1 h offline theoretical and 1 h exercise guidance + online supervision (not reported length) for once a week; group, offline + online	Conventional nursing education	2months/baseline-2-3-6months	Burnout
Yan [53]	China	33/32	31.22 ± 8.74/33.17 ± 7.98	Mindfulness-based therapy (MBSR + MBCT); once a week for 90–120 min; group, offline	Conventional psychological education	10weeks/baseline- 10weeks	Resilience/burnout
Gan et al. [54]	China	20/20	28.69 ± 2.98/28.78 ± 3.02	MBSR; 1 h offline guidance + online supervision (not reported length) for once a week; group, offline + online	Conventional nursing education	2months/baseline- 2 months	Resilience/burnout
Lin et al. [55]	China	40/40	27.4 ± 3.5/27.0 ± 3.3	MBSR; once a week for 45 min; group, offline	Conventional psychological education	2months/baseline- 2 months	Burnout
Luan et al. [56]	China	32/32	32.44 ± 5.98/32.00 ± 4.90	MT; once a week for 60–120 min; group, offline	Conventional psychological education	2months/baseline-1-2 months	Resilience/burnout
Al-Hammouri et al. [61]	Jordan	60/63	30.43 ± 4.0/31.41 ± 4.9	MBSR; two sessions a week for 60 min + family exercise; group, offline	Blank intervention	4 weeks/baseline- 4 weeks	Sleep quality
Xie et al. [57]	China	53/53	27.96 ± 4.9/27.4 ± 3.9	MBIs (MBSR + MBCT + ACT); once a week for 150 min + family exercise; group, offline	Education related to burnout	2months/baseline- post-test-1-3months	Burnout
Hilcove et al. [63]	US	41/37	24–69 (42.4)/24–64 (42.5)	Mindfulness-Based yoga; once a week (not reported length) + home practice; group, offline	Blank intervention	6weeks/baseline- 6weeks	Burnout/sleep quality
Pérez et al. [62]	Spain	39/35	37 ± 9.13 (total sample)	MBIs (MBSR + MBCT); once a week for 60 min + family exercise; personal, online	Blank intervention	6weeks/baseline- 6weeks-3months	Burnout
Pratt et al. [64]	US	69/33	28.06 ± 5.30/28.85 ± 7.75	MI; Once a day(not reported length) + home practice; personal, online	Blank intervention	4weeks/baseline- 4weeks	Burnout
Lin et al. [58]	China	44/46	32.86 ± 7.49/30.20 ± 6.09	MBSR; once a week for 120 min + home practice; group, offline + online	Blank intervention	8weeks/baseline- 8weeks-12weeks	Resilience
Asadollah et al. [65]	Iran	33/33	34 ± 7/32 ± 7	LKM; three times a week for 20 min + home practice; personal, online	Mental health education	4weeks/baseline- 4weeks	Burnout
Talebiazar et al. [66]	Iran	30/30	30.73 ± 5.98/29.03 ± 4.13	MBSR; once a week for 120 min + home practice; group, offline	Usual routine care	8weeks/baseline- 8weeks	Burnout
Wang et al. [59]	China	52/47	32.25 ± 3.56/31.45 ± 4.00	MBIs (MBCT + MBSR); 5 days a week(not reported length) + family exercise; personal, online	Psycho-education	8weeks/baseline- 4weeks-8weeks	Resilience/burnout
Yuan et al. [60]	China	76/75	22–56/23–58	MBSR; once a week for 60 min; group, offline	Blank intervention	8weeks/baseline- 8weeks	Sleep quality

Notes: IG: Intervention Group; CG: Control Group; MBSR: Mindfulness-based Stress Reduction; MBCT: Mindfulness-based Cognitive Therapy; ACT: Acceptance of Commitment Therapy; CD-RISC: Connor–Davidson Resilience Scale; MBI: Maslach Burnout Inventory; MBI-HSS: Maslach Burnout Inventory-Human Services Survey; MT: Mindfulness Training; MBIs: Mindfulness-based Interventions; PSQI: Pittsburgh Sleep Quality Index; GSQ: Global Sleep Quality; ProQoL: Professional Quality of Life Scale; MI: Mindfulness Intervention; LKM: Loving-kindness Meditation

Intervention group characteristics

Table 1 have summarized the intervention characteristics, including program, length, frequency, delivery mode and duration. According to different programs of mindfulness-based interventions in this meta-analysis, five adopted mindfulness-based interventions or therapies studies (n = 5) [53, 57, 59, 62, 64], three mindfulness exercises studies (n = 3) [56, 63, 65], the rest of studies were MBSR [51, 52, 54, 55, 58, 60, 61, 66]. The length of the intervention group ranged from 20 min [65] to 150 min [57], the most common length was 60 min [51, 54, 60–62], others lengths of intervention such as 45 min [55], 120 min [52, 58, 66], 60–120 min [56], 90–120 min [53], but three did not report the intervention study length (n = 3) [59, 63, 64]. The intervention frequency was conducted once a week (n = 12) [51–58, 60, 62, 63, 66], two sessions a week [61], three times a week [65], five days a week [59], once a day [64]. According to the mode, there were three modes: offline, online + offline and online. Offline (n = 8) [53, 55–57, 60, 61, 63, 66]were the main mode; four studies were online [59, 62, 64, 65]; four studies were offline + online [51, 52, 54, 58]. As for the duration, interventions ranged from four weeks to twelve weeks.

Control group characteristics

Table 1 have shown that the control group received conventional psychological intervention (mental health education, usual routine care, psycho-education or education related to outcome) and blank intervention.

Intervention fidelity

Santacroce et al. [67] identified key fidelity determinants in intervention studies, including adherence, implementer competence, and participant completion rates. Among the 16 studies included in this study, three studies reported the high adherence [52, 56, 57]; 11 studies documented the implementer qualifications [51–55, 57, 58, 60, 61, 65, 66]; five studies provided participant high completion or attendance [58, 59, 62–64].

Summary of outcomes and measurements

In total, 13 studies reported burnout. Burnout was measured with Chinese version of the Maslach Burnout Inventory (MBI) (n = 2), Maslach Burnout Inventory-Human Services Survey (MBI-HSS) (n = 5), Maslach Burnout Inventory (MBI) (n = 5) and Professional Quality of Life Scale (ProQoL) (n = 1). Six studies reported the resilience and it was measured with Connor–Davidson Resilience Scale (CD-RISC) (n = 3), Chinese version CD-RISC (n = 3). In addition, three studies reported the sleep quality with the Pittsburgh Sleep Quality Index (PSQI) (n = 2) and Global Sleep Quality (GSQ) (n = 1).

Assessment of risk of bias

From Fig. 2, we can summarize the assessment risk of bias of 16 RCT studies. In the randomization process domain, 12 studies (75%) were rated as low risk, four studies (25%) were classified as some concerns, since they had no sufficient evidence about the allocation sequence concealment. In the deviations from intended interventions domain, six studies (37.5%) were rated as low risk, 10 studies (62.5%) were classified as having some concerns, because they lack sufficient information on blinding to the participants by the delivering the intervention people. In addition, due to the report incomplete data, one study was considered as high risk. Because one study had bias risk in measurement of the outcome, it was rated as high risk. Besides, six studies lack of research protocols, so they had some concerns in the selection of the reported result domain. Overall, two studies (12.5%) were classified as low risk, 12 studies (75%) were classified as having have some concerns, two studies (12.5%) were judged high risk.

Fig. 2.

Risk of bias assessment of the included studies

Quality assessment of evidence

We evaluated the summarized results through GRADEPro software. In these studies, burnout, resilience and sleep quality were classified as low, moderate and low respectively. More details of quality assessment can be found in Table 2.

Table 2.

Summary of quality of evidence appraisal

Certainty assessment								No of participants		Effect		Certainty	Importance
Outcome	No of studies	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	Experimental	Control	Relative (95% CI)	Absolute (95%CI)
Burnout	13	RCT	Not Serious	Serious2	Not serious	Not serious	Reporting bias3	535	485	-	SMD 1.43 lower (1.94 to 0.92 lower)	⊕⊕OO Low	Critical
Resilience	6	RCT	Not Serious	Serious2	Not serious	Not serious	None	231	227	-	SMD 1.33 higher (0.09 to 2.56 higher)	⊕⊕⊕O Moderate	Critical
Sleep quality	3	RCT	Serious1	Serious2	Not serious	Not serious	None	177	175	-	SMD 1.1 lower (1.79 to 0.41 lower)	⊕⊕OO Low	Critical

Notes: RCT: Randomized Controlled Trial; CI: Confidence Interval; SMD: Standardized Mean Difference

1 Almost half of the included studies were assessed as having some concerns or high risk if bias

2 The evidence was downgraded because of a high level of heterogeneity

3 The evidence was downgraded due to the funnel plot had asymmetrical

Effects of mindfulness-based interventions on nurses

Burnout

Thirteen studies on burnout were included in our meta-analysis, including 1020 participants and pointed out that mindfulness-based interventions greatly improved burnout [SMD = -1.43, 95% CI (-1.94, -0.92), Z = 5.52, P < 0.001], with substantial heterogeneity (I² = 92%, P < 0.001). See Fig. 3(a).

Fig. 3.

Forest plot of the effectiveness of mindfulness-based interventions on (a) burnout (b) resilience (c) sleep quality

Resilience

This meta-analysis enrolled six studies on resilience, including 458 participants and showed that mindfulness-based interventions could promote nurses’ resilience ability [MD = 9.78, 95% CI: 0.38 to 19.17, P = 0.04], with substantial heterogeneity (I² = 98%, P < 0.001). See Fig. 3(b).

Sleep quality

Three studies were selected on outcome sleep quality in this study, including 352 participants and demonstrated that mindfulness-based interventions significantly raised sleep quality [SMD = -1.1, 95% CI (-1.79, -0.41), Z = 3.14, P = 0.002], with substantial heterogeneity (I² = 89%, P < 0.001). See Fig. 3(c).

Subgroup analyses

Country

In Supplementary Material 1 Figure S1, this study was divided into two types of countries: developing country and developed country. The results had shown that developing country [766 participants, SMD = -1.78, 95% CI (-2.43, -1.12), P < 0.001], developed country [254 participants, SMD = -0.64, 95% CI (-1.18, -0.10), P = 0.02], significantly reduced burnout (Z = 5.56, P < 0.001). However, the results reported a high heterogeneity between the two groups of countries (P < 0.001, I² = 92%).

Sample

As shown in Supplementary Material 1 Figure S2, samples (n < 100) [712 participants, SMD = -1.60, 95% CI (-2.21, -0.99), P < 0.001] significantly improved burnout, whereas samples (n ≥ 100) [308 participants, SMD = -1.08, 95% CI (-2.24, -0.07), P = 0.07] didn’t significantly reduce burnout. In addition, there was high heterogeneity between the different samples ((P < 0.001, I² = 92%).

Types of intervention

In the Supplementary Material 1 Figure S3, MBSR [366 participants, SMD = -3.28, 95% CI (-4.74, -1.82), P < 0.001], mindfulness therapy and interventions [446 participants, SMD = -0.72, 95% CI (-1.05, -0.40), P < 0.001, I² = 63%], mindfulness exercises [208 participants, SMD = -0.76, 95% CI (-1.04, -0.48), P < 0.001, I² = 0%], significantly improved burnout (P < 0.001, I² = 92%). The subgroup analysis showed that a statistically significant difference between the mindfulness exercises group and the control group (I² = 0%). However, high heterogeneity remained across the different intervention types.

Lengths of intervention

As shown in Supplementary Material 1 Figure S4, the interventions included more than eight weeks [700 participants, SMD = -1.91, 95% CI (-2.64, -1.18), P < 0.001] had larger impacts on alleviating burnout than with less than eight weeks [320 participants, SMD = -0.71, 95% CI (-1.14, -0.28), P = 0.001].

Modes of intervention

From the Supplementary Material 1 Figure S5, online + offline [226 participants, SMD = -2.19, 95% CI (-3.13, -1.26), P < 0.001] had more effects on relieving burnout than online [341 participants, SMD = -0.78, 95% CI (-1.22, -0.33), P < 0.001] and offline [453 participants, SMD = -1.72, 95% CI (-2.65, -0.80), P < 0.001].

Resilience

Types of intervention

As shown in Supplementary Material 1 Figure S6, MBSR [230 participants, MD = 16.71, 95% CI (8.29, 25.14), P < 0.001] and mindfulness exercises [64 participants, MD = 8.75, 95% CI (3.73, 13.77), P < 0.001] improved resilience of nurses, while mindfulness therapy and interventions [164 participants, MD = 0.34, 95% CI (-14.61, 15.28), P = 0.96] had no effects in promoting resilience scores. High heterogeneity was reported in different types of intervention (P < 0.001, I² = 98%).

Modes of intervention

In the Supplementary Material 1 Figure S7, online + offline [230 participants, MD = 16.71, 95% CI (8.29, 25.14), P < 0.001] was the most effective in raising the resilience ability in different delivery modes of intervention program, the other was offline[129 participants, MD = 8.11, 95% CI (5.81, 10.41), P < 0.001] (I² = 0%).

It still remained high heterogeneity in the subgroup analyses. So we conducted a meta-regression analysis to explore the heterogeneity of burnout. The meta-regression analysis indicated that we found no sources of heterogeneity. The covariates in the meta-analysis did not explain the variability of effect sizes. More details were in the Supplementary Material 1 Table S2. This suggests that some undetected factors may cause the high heterogeneity in the meta-regression. The reasons may include unreported study characteristics or results, study design or method and other unconsidered factors. Therefore, we should be cautious about the interpretation of study results.

Sensitivity analysis and publication bias

Sensitivity analysis plots indicated the study results were robust (See Supplementary Material 1 Figure S8).

Supplementary Material 1 Figure S9 suggested an asymmetry funnel plot for burnout scores (n = 13) and a high risk of publication bias. In addition, Egger’s tests also demonstrated the risk of publication bias in the included studies (Egger, P < 0.05) (Supplementary Material 1 Figure S10).

Discussion

This study is the first to conduct a systematic review and meta-analysis of randomized controlled trials to evaluate the effects of mindfulness-based interventions on burnout, resilience, and sleep quality among nurses. Previous systematic reviews have examined the effects of mindfulness-based interventions on psychological distress, depression, stress, and burnout but not on resilience and sleep quality among nurses [35–37]. Liu et al. [68] and Watanabe et al. [69] indicated that mindfulness-based interventions significantly reduced anxiety among nurses. Kang et al. [70] and Norouzinia et al. [71] revealed that MBIs were beneficial in reducing stress, anxiety, and depression among nurses. Penque [72] and Motaghedi et al. [73] showed that mindfulness-based programmes had an impact on improving nurses’ burnout. Ramachandran et al. [37]found that MBIs could only promote the personal accomplishment of burnout among nurses. Our meta-analysis of 16 randomized controlled trials demonstrated that mindfulness-based interventions could improve burnout and enhance resilience and sleep quality among nurses. This finding provides practical support for using mindfulness-based interventions to improve physical and mental health and quality of life among nurses. The study identified several research constraints, including (i) substantial heterogeneity, (ii) some concerns of bias due to the blinding of personnel, (iii) high risk of bias associated with missing outcome data and outcome measurement, and (iv) publication bias. Our analysis revealed significant publication bias through Egger’s test (P < 0.05) and funnel plot asymmetry. This may stem from two key factors: (i) Selective publication bias: studies with statistically positive outcomes are more likely to be published and retrieved [74]. (ii) Language bias: Restricting inclusion to English and Chinese publications may have excluded relevant studies in other languages. These biases likely contributed to the elevated heterogeneity observed in our findings (I² = 92%). Given the limitations and the low certainty of evidence, we should interpret the findings of this study with caution and explore higher-quality studies to corroborate the present findings.

Maslach described burnout as a three-dimensional syndrome characterized by emotional exhaustion, depersonalization, and a diminished sense of personal accomplishment [75]. The current meta-analysis revealed that mindfulness-based interventions can significantly alleviate burnout among nurses, consistent with Motaghedi et al. [73], whose findings support the use of mindfulness-based interventions for burnout reduction among nurses. Mindfulness-based interventions reduce burnout primarily through physiological and psychological mechanisms. For instance, a physiological study demonstrated that mindfulness training can lessen pain, ease emotional exhaustion, and alleviate job burnout among nurses by altering the activity of the anterior cingulate cortex and insula [76]. From a psychological perspective, an individual’s thinking shifts from a “Doing Mode Of Mind” to a “Being Mode Of Mind.” Mindfulness training enables practitioners to objectively perceive and accept their current experience rather than resorting to actions aimed at changing an uncomfortable environment. This shift in perspective eliminates the connection between themselves and negative emotions and reduces burnout [77].

Considering the substantial heterogeneity in the effects of mindfulness-based interventions on burnout among nurses, we conducted subgroup analyses based on countries, sample size, intervention types, lengths, and modes. Interventions implemented in developing country had a greater effect on burnout, compared with those in developed country. For instance, two studies reported a high prevalence of burnout among ICU nurses—81% in the United States and 66.1% in Spain [78, 79]. By contrast, study from Iran reported prevalence rates of 42% [80]. Burnout among nurses was higher in developed country than in developing country. In the current study, short-term mindfulness-based interventions were less effective for individuals in developed country than those in developing country. We concluded that mindfulness-based interventions can reduce burnout among nurses when implemented in smaller groups (fewer than 100 participants) but have no significant effect when conducted in larger groups (more than 100 participants). The reason may be that the smaller group size (fewer than 100 participants) may allow for more centralized management of participants. No study has thus far suggested that the sample size in mindfulness-based interventions affects burnout among nurses. Of the three intervention types, MBSR was most effective in relieving burnout among nurses, likely because it incorporates core mindfulness techniques, resulting in a more comprehensive and potent intervention [34]. Ghawadra et al. [35] showed that implementing MBSR can significantly reduce burnout among nurses. The present study revealed that interventions lasting more than eight weeks more effectively reduced burnout than those lasting less than eight weeks, consistent with the meta-analysis by Maricuţoiu et al. [81]. With regard to intervention modes, online + offline mindfulness-based interventions could be more effective in improving burnout among nurses. Relevant research currently focuses on the offline delivery of mindfulness-based interventions. Offline intervention can strengthen the relationship between participants and interventionists, allowing problem identification and prompt correction during implementation, significantly increasing the effectiveness of intervention. However, online + offline interventions integrate online continuous practice with supervision based on offline interventions, enhancing aspects of mindfulness-based interventions and comprehensively addressing burnout among nurses. Previous studies have not established the most effective intervention mode, necessitating further research. With the low certainty of evidence and potential publication bias considered, the results regarding burnout should be interpreted with caution.

Currently, no meta-analysis has summarized the effect of mindfulness-based interventions on resilience among nurses. However, the present study provides moderate-certainty evidence supporting the positive effect of mindfulness-based interventions on resilience. Moreover, one randomized controlled trial concluded that mindfulness-based interventions significantly increased resilience in doctoral candidates. Similarly, Barry et al. [82] demonstrated that mindfulness training significantly improved positive psychological capital. Luthans et al. [83] defined psychological capital as “a positive psychological state expressed by individuals during growth and development,” characterized by four core components: self-efficacy, optimism, resilience, and hope. Therefore, we can infer that mindfulness-based interventions can improve resilience among nurses. Given the high heterogeneity in resilience, we conducted subgroup analyses. These analyses, by intervention type and mode, showed that MBSR and combined online–offline mindfulness-based interventions positively affected resilience among nurses. This effect could be attributed to the comprehensive and consistent implementation of the interventions, maximizing their impact. However, further research is needed to confirm these findings.

Nurses experience high stress levels [84] due to frequent shifts, heavy workloads, and emotional strain, which affect sleep quality [85]. The current study showed that mindfulness-based interventions markedly improved sleep quality among nurses. This finding is consistent with previous research indicating that mindfulness positively correlates with sleep quality: higher mindfulness is associated with better sleep [86]. Mindfulness can transform negative emotions into positive experiences by shifting attention to the present moment, thus reducing pre-sleep arousal; meanwhile, it can lower sympathetic activity in the brain, promoting sleep [87, 88]. However, given the low certainty of evidence in the included studies, the results of the current research should be interpreted with caution.

Strengths and limitations

This meta-analysis has several advantages. First, we strictly adhered to the Revised Cochrane Risk-of-Bias assessment tool to evaluate the quality of studies and followed the PRISMA flow diagram for literature screening, ensuring a standardized approach. Second, previous studies have mostly focused on the effects of mindfulness-based interventions on psychological distress, depression, stress, and burnout among nurses. However, this research synthesized studies on resilience and sleep quality among nurses, providing a more comprehensive overview of their physical and mental well-being and a basis for improving their quality of life. Finally, we conducted subgroup, meta-regression, and sensitivity analyses to explore sources of heterogeneity and ensure the robustness of the results.

This study has several limitations. First, studies have shown that 99.1% of mindfulness-based interventions RCTs are published in English, with Chinese-language publications being the second most prevalent [89, 90]. we included only studies published in English and Chinese and excluded those in other languages. The inclusion of studies in other languages would strengthen the evidence of the meta-analysis. Second, overall heterogeneity was high, and the sources of heterogeneity could not be determined through subgroup, meta-regression, and sensitivity analyses. Thus, further research is needed to identify these undetected factors, such as differences in cultural background, medical policy or work environment. Finally, this study has a high risk of bias due to incomplete data, outcome measurement, and publication bias; in addition, the certainty of evidence is moderate to low. Therefore, the findings of this study should be interpreted cautiously.

Conclusion

This systematic review and meta-analysis evaluated the effect of mindfulness-based interventions on burnout, resilience, and sleep quality among nurses. The findings suggest that these interventions effectively improve all three. Therefore, mindfulness-based interventions offer important clinical practice support for enhancing physical and mental well-being, quality of life, and the effectiveness of nursing management among nurses. The moderate-to-low quality of the evidence, coupled with risk bias to a certain extent, necessitates rigorous, high-quality studies to validate these findings.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

SMD

Standardized mean difference

MD

Mean difference

CNKI

China National Knowledge Infrastructure

GRADEpro

Development and Evaluation profler

RoB 2

Risk-of-bias tool for randomized trials

Author contributions

Jin Dou: Writing-review & editing, Methodology, Investigation, Software, Resources, Formal analysis, Data curation, Conceptualization. Yujia Lian: Visualization, Methodology, Formal analysis, Data curation, Conceptualization. Lili Lin: Validation, Software, Resources, Investigation, Data curation. Siti Noraini Binti Asmuri: Visualization, Validation, Supervision, Project administration, Conceptualization. Peixi Wang: Visualization, Supervision, Conceptualization.Ruthpackiavathy A/P Rajen Durai: Writing-review & editing, Visualization, Methodology, Supervision, Project administration, Conceptualization.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

No applicable.

Consent for publication

No applicable.

Competing interests

The authors declare no competing interests.