The effects of mindfulness‐based interventions on nurses' anxiety and depression: A meta‐analysis

Abstract Objective This meta‐analysis aimed to determine the effect of mindfulness interventions on nurses' levels of depression and anxiety. Design Meta‐analysis of randomised controlled trials. Methods The following Chinese and English databases were searched: PubMed, Embase, Cochrane Library, Web of Science, and China National Knowledge Internet (CNKI). The retrieval period was from database construction to 30 March 2022. Two researchers screened the relevant literature and extracted the data. After a cross‐check, data were input into Stata version 16.0 for meta‐analysis. Results Twelve randomised controlled trials from 2017 to 2021 were included, which involved 807 subjects (405 and 402 in the intervention and control groups, respectively). Meta‐analysis results showed that nurses' anxiety reduced by mindfulness‐based interventions was significantly higher compared to that of the control group (SMD = 0.91, 95% CI: 0.27–1.55, p < 0.05). Furthermore, an 8‐week mindfulness‐based intervention (SMD = 1.43, 95% CI: 0.61–2.24) reduced the level of anxiety significantly more compared to a 4‐week intervention (SMD = 1.03, 95% CI: 0.36–1.71). Mindfulness‐based interventions were better compared to conventional intervention to reduce the level of depression (SMD = 1.02, 95% CI: 0.42–1.61, p < 0.05), and an 8‐week mindfulness intervention (SMD = 1.81, 95% CI: 0.78–2.84) reduced the level of depression significantly more compared to a 4‐week intervention (SMD = 0.82, 95% CI: 0.29–1.35). Since limited studies had interventions longer than 8 weeks, results on longer mindfulness interventions in reducing nurses' anxiety and depression are inconclusive. In conclusion, mindfulness intervention for 8 weeks or less can significantly reduce nurses' anxiety and depression levels. Patient or Public Contribution None.


| INTRODUC TI ON
Anxiety, a feeling of fear, occurs during a threatening or stressful situation, while depression usually involves a low mood, fatigue, and trouble sleeping in the face of stress (Aihong, 2017;Allen et al., 2006). Mild symptoms of anxiety or depression may seem vague; however, a rating scale can determine the severity of the illness. Once diagnosed, they require appropriate treatment (Beerse et al., 2020). Nursing staff have high responsibility, high risk, heavy workload, and often face patients' pain and death. Heavy pressure leaves nurses at a high risk of anxiety and depression (Botha et al., 2015). Anxiety and depression threaten nurses' physical and mental health and also increase the turnover rate and risk of nursing errors, resulting in a decline in nursing quality (Cheung & Yip, 2015).
Originally from the Eightfold Path of Buddhism, mindfulness, a Buddhist practice, emphasises the conscious awareness of the present without judgement. Mindability-based therapies is an umbrella term for various psychotherapies that focus on mindfulness. Mindfulnessbased interventions include mindfulness-based stress reduction (MBSR) and mindfulness-based cognitive therapy (MBCT). First developed by Kabat-Zinn in 1982, MBSR trains patients with chronic pain in self-regulation through a 10-week stress reduction and relaxation program (Chunmei & Chen, 2020). It has since been widely used in stress management. Originally developed by Teasdale JD, MBCT aims to prevent the recurrence of major depression by combining the elements of mindfulness training and cognitive therapy (Chunnan et al., 2021). Mindfulness-based interventions can effectively reduce a series of individual symptoms for anxiety and depression and their severity (Cumpston et al., 2019). The main implementation method included team teaching by mindfulness therapists to guide the objective of mindfulness training, training content including mindfulness diet, mindful breathing training, mindfulness meditation, mindfulness practice yoga, mindfulness sitting, and mindfulness walking. An intervention exercise record sheet was issued at the end of the mindfulness intervention to record negative emotions and pleasant or unpleasant events. Till date, most studies on the effect of mindfulness interventions on nurses focused on job burnout, and less attention has been paid to nurses' anxiety and depression. Many studies have reported on the systematic evaluations of mindfulness-based interventions to improve nurses' job burnout (Daigle et al., 2018), stress (Dean, 2016), and mental health (Deyang & Shoumin, 2018); however, there are no evidence-based reports for the effect of intervention on nurses' anxiety and depression. Therefore, this study conducted a meta-analysis based on previous literature to explore the effects of mindfulness interventions on nurses' anxiety and depression to provide evidence for its clinical application.

| MATERIAL S AND ME THODS
This meta-analysis was designed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines (Guendelman et al., 2017).

| Inclusion criteria
The inclusion criteria for this meta-analysis were: (1) Participants: Nurses who had obtained their nurse qualification certificate, worked in clinical practice, and were at least aged 18 years; (2) Intervention: Mindfulness-based interventions, which included mindful meditation, mindful yoga, mindful cognitive training, and mindfulnessbased stress reduction (MBSR), etc; (3) Comparison: Blank control or routine intervention; (4) Outcome: Changes in nurses' anxiety and depression levels before and after the intervention (unrestricted evaluation scale); and (5) Study: Randomised controlled trials (RCT).

| Search strategy
All published articles from the establishment of the database till 30 March 2022 were via the PubMed, Cochrane Library, Embase, Web of Science, and China National Knowledge Internet (CNKI) databases. In addition, the search scope was expanded manually by searching for references to other articles and going to the library to find journals. Using PubMed as an example, the retrieval formula is shown in Table 1. After literature were retrieved, all records were imported into Endnote X9 for classification.  (3) conditions required to assess the risk of bias.

| Quality evaluation and statistical analysis
Two researchers used the bias risk assessment tool of RCTs recommended in the Cochrane Handbook 5.1.0 (Guojie et al., 2021)  exist between studies, and a random-effects model was adopted.
In contrast, a fixed-effects model was used. Continuous data were represented by the standardised mean difference (SMD) and 95% confidence interval (CI). The results of the meta-analysis were presented with a forest map. Subgroup and sensitivity analyses were performed to explore the source of heterogeneity in the results with high heterogeneity. Egger's test and funnel plots were used to evaluate publication bias. Statistical significance was set at p < 0.05.

| Study characteristics
A total of 831 studies in English or Chinese were retrieved, which included 830 studies in from the databases and one study from another source. Of these, 144 duplicate references were excluded using EndNote X9. After the titles and abstracts were read, another 662 studies were excluded. Furthermore, 13 other studies were excluded after the full text was read. Finally, 12 RCTs were included in this meta-analysis. The literature screening and exclusion process is presented in Figure 1.

| Quality assessment
The results of the quality evaluation of the 12 included studies are shown in Figure 2. Random sequence generation methods were reported in seven studies (Hofmann & Gómez, 2017;Kabat-Zinn, 1982;Kang & Myung, 2022;Liberati et al., 2009;Maharaj et al., 2018;Qiu, 2019;Shapiro et al., 2006), which included the random number table method and random number generator. One study (Teasdale et al., 2000) was grouped according to the order of entry into the study. Four studies (Kang & Myung, 2022;Liberati et al., 2009;Qiu, 2019;Shapiro et al., 2006) reported allocation concealment of a random scheme. Due to the particularity of the study, all mindfulness-based interventions required the active cooperation of the participants, and the blind principle could not be performed.
Therefore, the principle of blind method was not explained in all the studies; however, informed consent from the participants was obtained. Data from all the studies were complete without selective reports, and there were no other deviation risks.

Anxiety scores meta-analysis results
The 12 included studies reported changes in participants' anxiety levels. The heterogeneity test showed I 2 = 94.2% and p = 0.000, which indicated significant heterogeneity between the studies. A meta-analysis was conducted using random effects model (as shown in Figure 3), and the combined SMD was 0.91 (95% CI: 0.27-1.55).
Nurses' anxiety level reduced by mindfulness interventions was significantly higher compared to that of the control group. Furthermore, the difference was statistically significant (p = 0.005). A sensitivity analysis was conducted on the included studies to explore the sources of heterogeneity ( Figure 4). Regardless of the studies that were excluded, heterogeneity did not decrease significantly, nor did the combined results change significantly, which suggested that the results of this study were relatively robust.

Subgroup analysis
The intervention duration in the included studies ranged from 4 to 13 weeks, with five that lasted for 8 weeks, five for 4 weeks, one for 12 weeks, and one for 13 weeks. A subgroup analysis was performed based on the intervention time ( Figure 5). Heterogeneity was I 2 = 92.0% (p = 0.000) for the 8-week of intervention (combined SMD = 1.43 (95% CI: 0.61-2.24)). It was I 2 = 84.6% and p = 0.000 for the 4-week intervention, and the combined SMD of the random effects model was 1.03, 95% CI: 0.36-1.71. Since there was only one study for the intervention duration for 12 weeks and 13 weeks, it could not be combined; therefore, it was not included in the subgroup. The results of the subgroup analysis of intervention duration showed that the longer the intervention time, the more the anxiety levels were reduced.
Six scales were used to evaluate anxiety levels, among which seven studies used the Self-rating Anxiety Scale (SAS) and one used each of the other scales. The measurement scale was divided into subgroups for analysis ( Figure 6). Since there was only one study that used the other scales, subgroups were not included, and only studies that used the SAS were analysed. The heterogeneity of the SAS subgroups was I 2 = 90.7%, p = 0.000, and the random effect model combined SMD = 1.51, 95% CI: 0.84-2.18, p = 0.000. The SAS was used to assess the reduction in anxiety by mindfulness intervention more compared to the other scales.

Publication bias
Funnel plots were drawn for the included studies to evaluate publication bias (Figure 7). All research points were scattered, with some discrete points, and there might have been publication bias in the subjective evaluation. Egger's test was used for quantitative evaluation (t = 1.08, p = 0.307), which suggested that there was no publication bias among the studies.

| Depression scores
Depression scores meta-analysis results Ten of the included articles reported changes in participants' depression levels. The heterogeneity test showed I 2 = 92.4%, p = 0.000, which suggested significant heterogeneity between the studies.
A meta-analysis was conducted using the random effects model (Figure 8), and the combined SMD was 1.02, 95% CI: 0.42-1.61. The F I G U R E 1 Literature screening flow chart mindfulness intervention was better compared to the control group in reducing the level of depression, and the difference was statistically significant (p = 0.001). Sensitivity analysis was conducted on the included studies to explore the source of heterogeneity ( Figure 9). Regardless of the studies that were excluded, heterogeneity did not decrease significantly, and the combined results were all within the credibility range, which suggested that the results of this study were relatively robust.

Subgroup analysis
In the included literature, four studies had an intervention duration of 8 weeks, four with 4 weeks, and one each for 12 and 13 weeks.

Publication bias
Funnel plots were drawn for the included studies to evaluate publication bias (Figure 12). The distribution of each study point was relatively uniform; however, there were two discrete points, which may have resulted in publication bias. Egger's test was used for quantitative evaluation (t = 2.18, p = 0.060), which suggested that there was no publication bias between studies.

| DISCUSS ION
In this study, 12 high-quality randomised controlled trials (RCTs), which involved 807 clinical nurses from three countries, were obtained and included in the meta-analysis from the Chinese and  Mindfulness-based interventions, the practice of mindfulness (for example, through mindfulness-based stress reduction, sit-down meditation, yoga, or other mindfulness practices), causes individuals to be less reactive to unpleasant internal phenomena and more reflective of themselves, resulting in positive psychological responses (Tiller, 2013). Previous studies confirmed that mindfulness-based interventions reduced the severity of anxiety and depressive symptoms in individuals who had extensively sought treatment (Chunmei & Chen, 2020). A meta-analysis review based on 39 studies with 1140 participants also found that mindability-based therapies were moderately effective in improving anxiety and mood disorders (Watanabe et al., 2019). Furthermore, this study found that mindability-based interventions were equally effective for nurses' anxiety and depressive symptoms. Based on neurobiology, mindfulness intervention may produce biological mediators with positive effects, which can improve sleep quality, reduce salivary cortisol, and change brain structure and function. This can improve mood regulation, self-control, and stress reduction in intervention subjects (Xiumei & Yanbin, 2018). Psychologically, Shapiro (Yang et al., 2018) proposed a mindfulness model to explain the underlying mechanism of how mindfulness had a positive effect, which suggested that mindfulness worked by changing attention, intention, and attitude. Guendelman (Yaxue et al., 2021) proposed that the nature and mechanism of mindfulness interventions were very complex, and they contained various psycho-neurocognitive models.
According to the subgroup analysis of the intervention duration, the 8-week mindfulness intervention reduced anxiety and depression levels better compared to the 4-week intervention. However, the intervention durations of the included studies ranged from 4 to 13 weeks. Since only one study each had an intervention duration of      level of anxiety or depression. However, the intervention study for a longer time in this study was less, and there were limited research interventions for more than 8 weeks. Hence, mindfulness interventions to reduce nurses' anxiety and depression in the long-term should be further discussed.

| CON CLUS ION
Mindfulness-based intervention can effectively reduce nurses' level of anxiety and depression. Furthermore, the effect of an 8-week intervention was better compared to that of a 4-week intervention.

AUTH O R CO NTR I B UTI O N S
H.L. and L.K. wrote the first draft of the paper and had equivalent contribution; L.K. and X.M. designed the research; Q.S. and L.K. provided statistical guidance and revised the manuscript. L.K. acquisition and interpretation of data. H.L., X.M. and Q.S. performed the statistical analysis; All authors approved the final manuscript.

ACK N OWLED G M ENTS
The authors thank Wiley Editing Services (http://wiley editi ngser vices.com) for English language editing and review services.

FU N D I N G I N FO R M ATI O N
This subject was supported by Outstanding Youth Project of Hunan Province (No. 20B070).

CO N FLI C T O F I NTE R E S T
This paper has no financial interest in any person or organisation.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

E TH I C S S TATEM ENT
This study is a meta-analysis and does not involve personal samples, so no Ethical approval is required. That is, it is not applicable to this study.