Abstract
BACKGROUND:
Work-related stress can be defined as an individual’s reactions to work characteristics and indicates a poor relationship between coping abilities and work environment. If unmanaged, stress can impact mental and physical health (e.g., causing depression and cardiovascular disease). Many individuals use maladaptive stress-coping strategies, such as sedentary activities, unhealthy eating behaviors, and alcohol consumption, which do not contribute to long-term stress management. In contrast, stress reduction programs can help people manage and effectively reduce stress in the long term.
OBJECTIVE:
To gather the state of the art of work-related stress interventions, their efficacy and applications.
METHOD:
The PsycINFO and EBSCOHost databases were used. The search was carried out between January 28 and March 30, 2019. Inclusion criteria were full text available, text in English or Spanish and a study population comprising workers.
RESULTS:
Twenty-nine articles that included interventions involving aromatherapy, bibliotherapy, cognitive-behavioral therapy, exercise, alternative medicine, mindfulness, technology, stress management and sensory intervention were analyzed. The interventions showed significant reductions in stress, anxiety, depression and burnout; however, most of the studies were not based on specific stress models, and control groups often received no intervention whatsoever. As a result, it is challenging to draw conclusions regarding the success of the interventions, especially if they are novel.
CONCLUSION:
The results suggest that there is a broad portfolio of successful interventions regarding work-related stress. Most of the successful interventions were based on mindfulness; however, aerobic exercise and bibliotherapy may also be successful. The structure and level of evidence appear to be very relevant to the development of a successful intervention.
Introduction
Stress can be defined as a process that emerges from the relationship between an individual and his or her environment in which a demand, internal or external, is perceived as exceeding one’s resources for coping [1]. During this process, appraisal is fundamental; it refers to the way in which individuals conceive of their wellbeing, possible threats and coping resources for achieving wellbeing [2, 3]. Accordingly, a stress response occurs when the individual’s capacity for adaptation is exceeded. This response can be considered normal or adaptative when it is necessary for self-conservation; however, when chronic stress responses occur, they poses a threat to balance and homeostasis in the body, giving way to psychosomatic illnesses [4].
The transactional model, on which this review is based, conceives stress as a process in which stressor stimuli, appraisal, stress response and coping can be identified [5, 6]. Stressors can be categorized as psychosocial stressors and biogenic stressors. Work-related factors, which are the main focus of the interventions in this review, are situated in the first category [5]. Appraisal can be divided into two types: primary and secondary appraisal [3]. These two types function interdependently. Primary appraisal refers to whether the threat is relevant to the individual, whereas secondary appraisal refers to what the individual can do to cope. Coping refers to individuals’ efforts to respond to a demand that they have appraised as exceeding their resources.
According to McEwen [7, 8], the allostatic load theory, which posits that chronic exposure to stress, such as a stressful job, can weaken the allostatic process that returns the body to homeostasis after a stressful situation, can at least partially explain the connection between stress and health. Considering this, McEwen suggests that a sustained stress response can eventually result in detrimental health consequences, such as sleeping problems [9], burnout [10], increased risk of depression and anxiety [11], and coronary heart disease [12, 13].
Work-related stress can be defined as an individual’s reactions to characteristics of his or her work environment that indicate a poor relationship between the individual’s abilities to cope and the demands of the work environment [14]. According to Karasek and Theorell [15], work-related stress is produced by the interaction between demands and control. The effect of demands is moderated by the impact of decisions or control, resulting in an environment of high work-related tension if there is a high level of demands and a low level of control. The Effort-Reward imbalance model is another approach for describing the relationship between high effort and low reward and its relation to stress. This model introduces the concept of overcommitment, which leads individuals to exert great amounts of effort in exchange for little reward [16]. Additionally, this model is considered the most appropriate for measuring chronic stress in people engaged in work that requires a high degree of interaction with others, such as customer service, health care and teaching [16].
In practice, many individuals use maladaptive stress-coping strategies such as sedentary activities, unhealthy eating behaviors, and alcohol consumption, which do not contribute to long-term stress management [17]. Stress reduction programs can decrease the risk of several diseases [18]. There is great variability in intervention protocols for stress management; however, these interventions tend to be modal and are characterized by a predominance of group formats over individual formats, such as mindfulness-based stress reduction (MSBR) programs [19]. Stress reduction programs are vital for reliably reducing stress; however, training in coping strategies is a more behavioral approach that could significantly reduce both stress and the need for stress reduction programs.
To the best of the authors’ knowledge, there are few reviews regarding the state of the art of work-related stress interventions, such as the one carried out by Lamontagne et al. [20], who concluded, after a systematic review, that further study is needed to develop the job-stress intervention evidence base to guide policy and practice, and the meta-analysis on effects and durability of results by Richardson and Rothstein [18], which encourages researchers “to design quality experiments that incorporate random assignment to treatment and control groups and report the results of all outcomes, not just the statistically significant ones.” This highlights the need for useful, practical workplace interventions that could reduce stress and enhance work engagement is on the rise [21]. The lack of this evidence was the basis for the present systematic review, which has the aim of gathering the state of the art of work-related stress interventions, their efficacy and their application.
Method
A systematic review was conducted. This type of review was chosen because it allows the inclusion of qualitative studies, unlike meta-analyses. The review process was carried out with three rounds of screening, in which the authors identified studies through two database searches and proceeded to assess studies based on their abstracts and full text to determine whether they were eligible based on inclusion and exclusion criteria discussed below. The process is shown in Fig. 1.

PRISMA flow diagram for the review process and results. During the eligibility assessment, most articles were excluded for having low levels of evidence. Doi:10.1371/journal.pmed1000097.
Two databases were used for this review: PsycINFO, which was chosen for its focus on psychology and behavioral and social sciences research and literature, and EBSCOHost, which grants access to a wide set of databases with interdisciplinary scopes. Both databases were used to assess the state of the art of work-related stress interventions.
The search strategy was as follows: The terms Stress Relief AND Intervention AND Work-Related Stress were searched in papers published between 2013 and 2018. A five-year range was chosen due to the constant evolution of academic and scientific information and because this is also the range required by most journals that publish reviews.
Search
The database search was carried out between January 28 and March 30, 2019. PsychINFO yielded 2687 publications, whereas EBSCOHost yielded 1945.
The following procedure was used, as shown in Fig. 1: First screening of the titles. Second screening of abstracts. Third screening of the full articles.
Inclusion and exclusion criteria
Availability of full text to analyze content and pertinence Text in English or Spanish to ensure a full understanding of the articles Relevant authors cited, such as experts in the field of stress or in a specific area of intervention, such as mindfulness Study population comprised workers and excluded all other stress interventions performed with other populations
Results
Included studies
The selection process is presented in Fig. 1. Through this process, the authors identified 29 viable articles for the present review. PsycINFO yielded 15 articles suitable for the review, and EBSCOHost yielded 14 articles. All studies that did not include a clear intervention focusing on work-related stress and a working population were excluded.
Study characteristics
Supplementary Table 1 presents an overview of the 29 articles included in the review, including their sample characteristics, study design, instruments, procedure and outcomes. The studies were categorized according to the type of intervention, although some studies included various intervention types. For purposes of this review, those studies were categorized according to the type of intervention that was the focus of the study. The types of interventions included aromatherapy [22, 23], bibliotherapy [24], cognitive-behavioral therapy or CBT [25], exercise/physical activity [26, 27], alternative medicine [28, 29], mindfulness [30–39], technology-based interventions [40–46], stress management interventions or SMI [47–49] and sensory intervention [50]. Aromatherapy was set as a separate category from the sensory intervention because there were more studies on aromatherapy alone and because sensory interventions [50] evaluated the effects of both monosensory and multisensory stimuli on stress.
The aromatherapy studies differed in approach. While the first study [22] performed the intervention using essential oils hung in bottles from nurses’ necks during their workday, the second study [23] exposed the subjects to aroma diffusers while they performed a specific web-based task. The bibliotherapy intervention consisted of reading a book about stress management and resilience training [24]. CBT consisted of 16 one-hour sessions with a therapist to assess stress and develop coping strategies. A workplace intervention was included [25]. There were three different studies regarding exercise/physical activity. The first study assessed the difference in the effects of outdoor and indoor exercise sessions [26]. The second study of this type of intervention examined aerobic exercise as recommended by the American College of Sport Medicine [27]. Regarding alternative medicine interventions, both related studies provided a combination of CBT and alternative medicine [28, 29].
Regarding mindfulness interventions, there was somewhat more variation among studies. One study [30] was based on the MSBR program and loving kindness meditation. Another article [31] was unclear regarding its application of mindfulness but stated that it was, in fact, a mindfulness-based intervention. Other studies [32–34] provided an adapted version of MBSR. Only one of the studies used the full MSBR program as an intervention [37]. One study used a mindfulness and yoga intervention [34], and two studies involved only yoga interventions [36, 38]. Regarding technology-based interventions, four studies [42–45] were web-based interventions; of these, one study [43] used a CBT approach, another study [45] used a mindfulness approach, and the other two [42, 46] used different techniques. Some studies [40, 41] were app based, and one study [44] involved virtual reality and role-playing. Regarding SMIs, one study [47] consisted of a group-oriented stress intervention seminar, another [48] consisted of 10 separate stress education sessions, and the other SMI study included in this review consisted of a 5-component multimodal stress reduction intervention. Finally, regarding sensory interventions, the included study [50] consisted of four groups that received increased sensory input.
Of the 29 studies, 14 were RCTs [23, 50]. The other 15 were pretest-posttest studies. The grouping used in the as were as follows: two studies [45, 50] had three treatment groups and a control group that received no intervention; one study [39] had two treatment groups and a control group that received no intervention; another study [47] had a treatment group, a wait-listed control group and an external control group from a parallel study; one study [44] had a treatment group, a control group and a wait-listed control group. Studies [26, 28] provided interventions for their control groups, [29, 46] had wait-listed control groups, and studies [22, 49] provided no intervention for their control groups. Finally, studies [24, 48] had no control group. Of the 29 studies, only one study [30] had a mixed design, and the rest were quantitative studies. Sample sizes and composition also varied. The sample sizes in 22 studies [22, 41–50] were large, between 53 and 551 participants. The other seven studies [23, 40] had relatively small sample sizes of between 12 and 42.
Quality of the studies
Of the 29 studies reviewed, only 14 were RCTs, and the remaining 15 presented low levels of evidence and quality according to the PRISMA guidelines due to lack of a control group or a nonspecified duration [24, 49]. These quality issues weaken the conclusions that can be drawn from both the studies and the resulting review and makes them less generalizable.
Effects of interventions
The results presented in the following section refer to improvements in the treatment groups compared to the control groups, unless otherwise specified.
Stress and burnout
The studies showed different effects on various variables related to stress. First, stress was significantly reduced by bibliotherapy, mindfulness, aerobic exercise and online structured interventions [24, 48–50]. Specifically, bibliotherapy, mindfulness and tailored stress reduction programs [24, 49] reduced perceived stress; studies based on yoga [33, 38] decreased work-related stress; web-based mindfulness interventions [45] showed a significant relation between group support and stress reduction; and a multisensory intervention study [50] found that bisensory interventions were the most successful sensory interventions for reducing stress. Moreover, aromatherapy, mindfulness and multisensory interventions [22, 50] significantly reduced stress symptoms, and aromatherapy and some mindfulness studies and stress-reduction programs [29, 49] reduced psychosomatic symptoms.
Regarding burnout, this review found that aromatherapy and some mindfulness interventions and stress-reduction programs [27, 49] significantly reduced burnout symptoms. Burnout-related interventions [43] specifically reduced cynicism, while meditation and mindfulness mediated improvements resulting from web-based mindfulness interventions [45].
Emotional symptoms and quality of life
Aromatherapy, bibliotherapy, CBT and CAM, and mindfulness interventions [23, 44] were associated with a significant decrease in anxiety traits. Depression was significantly reduced by aerobic exercise, CBT and CAM, mindfulness and internet-based interventions for burnout [27, 43]. An increase in quality of life was reported as a result of bibliotherapy [24]; although a significant increase in the mental and physical components of QoL was achieved, a decrease in the physical component was reported after six months, while the improvement in the mental component was maintained. Psychological and emotional wellbeing were increased by web-based mindfulness and internet-based burnout interventions [43, 45], but the improvements resulting from web-based mindfulness [45] were not statistically significant. Improvements in vitality were correlated with mindfulness in web-based mindfulness interventions [45].
Coping strategies
Self-help stress management training [40] was associated with the acquisition of two coping strategies, a decrease in denial and an increase in active coping, in nurses. Computer-based stress management training [42] was associated with an increase in knowledge-regarding stress coping, and virtual reality training for stress management [44] was associated with an overall increase in coping skills, particularly emotional support skills.
Mindfulness, self-sompassion, self-esteem, self-transcendence and self-acceptance
Bibliotherapy [24] increased mindfulness after 12 weeks, and MBSR [32] had a significant increase in variables related to mindfulness, such as observation, consciousness, nonjudgement and nonreaction. Mindfulness interventions [34, 35] increased self-reported mindfulness. MBSR and mindfulness interventions [32, 35] improved self-compassion, and MBSR [32] also increased self-kindness and decreased self-judgment. An improvement in self-esteem was reported with MBSR [33]. Studies of multisensory interventions [50] also reported an increase in self-esteem in the multisensory group. An online mindfulness intervention [39] led to significant improvement in self-transcendence and significant improvement in self-acceptance.
Work performance and workplace wellbeing
A slight increase in workplace performance was reported with the use of stress-reduction programs [49], while web-based mindfulness [45] led to improvements in role functioning and productivity that were correlated with mindfulness and meditation. Additionally, there was a significant increase in workplace wellbeing with MBSR [37] and stress-reduction programs [49] compared to baseline. The stress management intervention [47] was associated with a slight decrease in over commitment in a population of German managers.
Health
Regarding health, the stress-reduction program [49] reported better health in the treatment group (TG) and worse health in the control group (CG). Overall, green exercise, mindfulness and yoga [25, 36] improved quality of sleep (QoS). Green exercise [25] improved QoS from 0 to 4 months but showed no significant improvement from 4 to 10 months. Mindfulness [32] was associated with better sleep quality, deeper sleep, fewer daytime inconveniences due to lack of sleep and a significant decrease in the use of medicine as a sleep aid.
Regarding physiological variables, on the one hand, green exercise, stress-management programs and multisensory interventions [26, 50] improved blood pressure (BP) and the cortisol awakening response (CAR), with green exercise [26] improving overall BP levels, stress-management programs [34] decreasing both BP and CAR, and multisensory interventions [50] resulting in a decrease in cortisol in the multisensory group after 30 days. On the other hand, aromatherapy [23] was associated with an improvement in HRV performance.
Discussion
This review included 29 articles published between 2013 and January 2019, and it sought to assess the state of the art in work-related stress interventions. Considering that demanding workplace challenges can produce stress and symptoms of burnout [10] and that teachers and healthcare providers are situated within the most stressful and demanding professions [46], this review considered a broad portfolio of interventions to further analyze their reach and impact. The included studies mainly included healthcare professionals, teachers and managers as the intervention populations.
Some of the examined interventions offered higher levels of evidence than others, suggesting that MBSR, aerobic exercise and structured stress-reduction programs are effective while other approaches require further evidence to support their use. On the one hand, bibliotherapy, mindfulness, MBSR and multisensory interventions [24, 48–50] showed high levels of evidence, indicating that these interventions could be more successful than the others considered in this review. In the bibliotherapy study [24], the book “Train Your Brain, Engage Your Heart, Transform Your Life” was read over 12 weeks. Both of the included alternative medicine studies [28, 29] proved to be effective in terms of stress and symptoms. Both studies [28, 29] consisted of a combination of Chinese alternative medicine (CAM) and CBT. While one study [29] compared the CAM-CBT intervention to a wait-listed control group, the other study [28] compared it to a CBT-only control group, which allowed for better conclusions to be drawn concerning CAM-CBT interventions. Among the studies regarding mindfulness, the most successful interventions were reported in studies [32–39]. None of these interventions used the traditional MBSR program; instead, studies [32, 39] used adapted versions that mostly modified the time span (less than 8 weeks) or included other factors (such as yoga). Studies [34, 38] consisted of mainly yoga classes.
Although promising results were achieved in several studies regarding stress, burnout, anxiety and quality of life, there is a lack of foundation from stress models among the studies. Of the 29 studies, only 7 based their studies on or made reference to stress models, such as Lazarus’s transactional model of stress [25, 41], Karasek’s job strain model [40], Siegrist E–R imbalance model [31, 47], Cohen’s unifying model of stress [28, 29], Van Der Doef and Maes’ demand-support-control model [31], and Bakker’s job demand-resources model [31]. The remaining studies made no reference whatsoever to any model [22–24, 48–50]. However, only 4 out of the 7 studies that referred to these models reported the consequent measurements and results and made reference to the theoretical model through the discussion [31, 47]. Furthermore, it is important for models to be appropriate to the study population; as mentioned previously, some models are better suited for certain populations, such as the use of the ERI model for interaction-centered professions such as teaching, customer support and health care. Although 15 of the articles included health care workers or teachers, only one based its research on the ERI model and mentioned overcommitment as an important variable in work-related stress. Further research should be based on specific models that suit the study population. Additionally, future research should assess the importance of overcommitment and possible interventions for it.
Methodological issues and limitations
Despite the promising findings of the studies in our review, the following limitations should be considered when drawing conclusions about these interventions.
Conducting a systematic review
As previously mentioned, the authors chose to conduct a systematic review instead of a meta-analysis. This decision was based on the availability and quality of current studies on work-related stress interventions. Meta-analyses, although more thorough, would have significantly reduced the number of articles that could be included, making this process nonviable.
Short- and long-term effects
Most research has measured medium- to long-term effects [24–30, 49], with some studies providing measurements over six-months [25, 47]. Two studies did not report the duration of measurement [31, 35]. The remaining studies measured short-term effects [22, 50]. Work-related stress interventions would be more attractive if the effects were sustained over longer periods of time. Future research focused on measurement that aims to examine long-term effects could further understanding of these effects in the long-term. Duration and long-term effects are especially significant in studies of stress because of the allostatic charge, which is what interventions aim to reduce and maintain over time to achieve better health and quality of life [7, 8].
Length of treatment program
Seven studies had short interventions involving less than 4 weeks of treatment [22, 50]. The other 12 studies had treatment periods between 5 and 8 weeks [28–30, 41–45]. Two studies did not report the length of treatment [31, 35]. Eight studies had treatment periods longer than 12 weeks and up to 26 weeks [24, 49]. The duration and intensity of the program is particularly impactful in work-related stress since employees often report having a lack of time outside of work. Hence, programs can be more accessible when delivered during working hours in a concise and time-efficient format.
Small sample sizes
Twenty-two of the studies [22, 41–50] had large samples of between 53 and 551 participants. The other seven studies [23, 40] had relatively small sample sizes of between 12 and 42 participants. Smaller samples should be avoided if possible in further interventions to provide more reliable results.
Homogeneous and heterogeneous samples
Most of the research consisted of only one population group per study, such as teachers [28, 34], health care workers [22, 46,48–50], police officers [31], or managers [33, 47]. Only one study included two groups: teachers and nurses [44]. The remaining studies had samples of employees in general [23–27, 45]. Although specific groups such as nurses, police officers and teachers have specific needs that need to be analyzed to provide tailored interventions, research in general could benefit from more diverse samples or a larger quantity of groups, including overlooked groups such as workers in finance or customer service. As previously mentioned, most of the studies included samples that consisted mostly of healthcare professionals, teachers and office employees in general. These samples are not representative of other groups of workers, and research should strive to assess interventions in different occupational sectors to ensure that their conclusions can be generalized.
Variables measured
Over 83 different instruments were used in 29 articles to measure 22 variables and the impact of the interventions on work-related stress. This variety imposes a great limitation for both conducting reviews and allowing generalization since differences in variables and instruments make it difficult to draw conclusions based on upon comparisons of the results achieved by different studies. The most commonly used instruments were the State–Trait Anxiety Inventory (STAI), used in 4 different studies [23, 43]; the Perceived Stress Scale (PSS), used in eleven [22, 45]; and the Maslach Burnout Inventory (MBI), used in eight [27, 49]. The remaining instruments were used fewer than 3 times, and 62 instruments were used in only one study. Furthermore, it is important to highlight that although environment and performance measures are very important in these types of interventions, they were not present to a desirable degree. The studies that measured these variables were as follows: studies [37, 41], which measured work engagement; study [42], which measured work performance; study [49], which measured productivity; and study [31], which measured workload, organizational constraints and workplace incivility.
Quantitative and qualitative data
Future research may benefit from mixed-methods approaches that combine quantitative and quantitative data. In the present review, only one study had a mixed method design [30], and the remaining studies were quantitative. To acquire a deeper understanding of the particular needs of certain groups and populations, more qualitative research should be conducted in work-related stress intervention research.
Coping skills and strategies
Only three studies assessed the impact of their intervention on the coping skills and strategies acquired by the employees [40, 44]. This poses a limitation on the reach and the evaluation of the results of the interventions.
Additional organizational intervention
As previously mentioned, stress is the result of the interaction between environmental stimuli and an individual’s capability for response. In this review, most of the studies targeted either environmental factors (nature, aroma) or personal coping skills (CBT, mindfulness). Person-centered interventions have been shown to be only partly effective for improving well-being [34]. Secondary prevention interventions can be carried out at either the individual or organizational level in the form of stress management courses that include a range of information about stressor elements, cognitive strategies, and relaxation strategies. However, it is important to highlight that there is little evidence regarding which elements within these courses are key [52]. A review by Lamontagne [20] showed that most individual interventions had an impact on the individual but not on the organization; however, organizational interventions had an impact on both individuals and organizations. An integrated approach that includes both environmental and individual factors is needed to result in improvements [21]. Only one of the studies presented this combined approach [25]. Further research should include combined multilevel interventions to assess their effects.
Conclusions
This systematic review intended to report the state of the art of work-related stress interventions and yielded 29 studies. From the review, three conclusions were drawn. First, there is a broad portfolio of interventions available for work-related stress; however, most interventions were compared to wait-listed control groups, a situation that only allows conclusions about the validity of the given intervention and cannot permit comparisons to other interventions. Multilevel interventions may offer better results in this context, since the results showed that individual interventions fell short effect-wise in several cases. It is important that research be based on theoretical models and assess variables appropriately. Such efforts can be valuable due to enhanced consistency within the study thereby ensuring that the same variables are followed from the theoretical model to the experimental application. A theoretical foundation will allow studies to be more thorough and permit better comparisons among interventions in specific contexts.
