Burnout is a stark metaphor that refers to the loss of mental energy. It is commonly used to describe a state or process of mental exhaustion, similar to the smothering of a fire, the extinguishing of a candle, or the draining of a battery. The Merriam-Webster dictionary describes it as ‘exhaustion of physical or emotional strength or motivation usually as a result of prolonged stress or frustration’ (noun) and ‘to cause to fail, wear out, or become exhausted especially from overwork or overuse’ (verb). Introduced in the 1970s and initially ridiculed in academia and ignored as pop psychology, burnout is currently a well-respected scholarly topic. This is evidenced by the fact that, according to PsycLit, over 13,000 scientific publications about burnout have appeared. Moreover, year after year more papers on burnout are published; for instance, between 1991 and 2010, publications rose by a factor of 7.7 (Heinemann & Heinemann, Reference Heinemann and Heinemann2017). So, it does not seem that burnout is a fad as initially believed; rather, it is here to stay. As we will see below, this is unsurprising as burnout can be seen as a modern, global affliction. First introduced in the United States, it readily expanded internationally. Tellingly, burnout originally appeared as a social problem, particularly in human services such as education, health care, and social work and was then taken up by academia. This means that initially it was a practical rather than a scholarly challenge because it comes with huge costs, for the individual as well as organizations. In addition, societal costs occur because of increased health care expenditure and income compensation for those suffering from burnout. In order to ameliorate these costs, an extensive burnout industry has emerged. Today, many professionals earn a living by treating burnout victims, offering workshops on burnout, or assisting organizations on how to prevent it. In this sense, burnout has boosted professional practice, thereby closing the vicious circle of supply and demand.

The current chapter sets out to present a state-of-the-art overview about our scientific knowledge on burnout, thereby also addressing some criticisms and ongoing debates. Clearly, it is beyond the scope of the chapter to review all scientific publications, but fortunately many useful meta-analyses have appeared that summarize research findings. This chapter is structured as follows. The second section explains the history and background of burnout, which illustrates that various forerunners exist. The third section focuses on the concept and definition and how these developed. The fourth section is about the assessment of burnout, discussing not only popular self-report questionnaires but also diagnostic criteria that are used by practitioners in clinical interviews. The fifth section addresses the prevalence of burnout and also includes a critical discussion about the criteria for proper epidemiological research that are often not met in burnout research. In the sixth section, a brief overview is presented about the antecedents, consequences, and correlates of burnout, thereby focusing on the individual as well as the organizational level. The seventh section is about various psychological explanations of burnout, whereas the eighth section discusses the effectiveness of individual, team, and organizational interventions to prevent and combat burnout. The chapter closes with a final section that summarizes the main conclusions and presents an outlook on the challenges of burnout research and practice that lie ahead.

History and Background

Although the term burnout as we know it today was first coined in the 1970s, the use of the metaphor is much older. Almost four hundred years ago William Shakespeare wrote: ‘She burnt with love, as straw with fire flameth. She burnt out love, as soon as straw out burneth’. Instead of hard work, Shakespeare associated burning love with an unfeeling of burnout. It was Herbert Freudenberger (Reference Freudenberger1974) who borrowed the term ‘burnout’ from the illicit drug scene, where it colloquially referred to the devastating effect of chronic drug abuse. He employed the metaphor to describe the gradual emotional depletion, loss of motivation, and reduced commitment among volunteers who worked with drug addicts. Not unimportantly, Freudenberger himself fell victim to burnout twice, which increased his credibility in spreading the message of burnout in the media, turning it soon into a buzzword. Independently and almost simultaneously, Christina Maslach (Reference Maslach1976) and her colleagues came across ‘burnout’ in California as well when interviewing a variety of human services workers. As a social psychological researcher, she was interested in how these workers coped with their emotional arousal using cognitive strategies such as detached concern. Through these interviews, she learned that these workers often felt emotionally exhausted, that they developed negative perceptions and feelings about their clients or patients, and that they experienced a crisis in professional competence as a result of the emotional turmoil. These practitioners referred to this syndrome as ‘burnout’. So, in fact, both Freudenberger and Maslach stumbled simultaneously and more or less accidently across the same colloquial term; apparently, burnout was in the air.

At a closer look, though, many cases of burnout avant la lettre have been documented, such as ‘acedia’ in Medieval monasteries and ‘melancholy’ among the nobility of the 15th and 16th century (Schaffner, Reference Schaffner, Neckel, Schaffner and Wagner2017). However, particularly interesting is the parallel with ‘neurasthenia’ – literally nervous weakness (Schaufeli, Reference Schaufeli, Neckel, Schaffner and Wagner2017). Neurasthenia emerged at the end of the 19th century and was seen as the result of modern, hectic life. It was believed that continuous overstimulation by artificial light, telephone and telegraph use, newspapers, ads, and steam trains weakens the nerves. This results in extreme fatigue, demotivation, and an inability to work, the reasoning went. Busy business people in particular used to fall victim to neurasthenia. Please note the striking similarity to the current public debate linking burnout with overstimulation due to the 24/7 economy, social media, mobile phones, and the Internet. In retrospect, one could say that the rise of neurasthenia was related to the transition from an agricultural to an industrial society, while burnout coincides with the transformation from an industrial to a postindustrial, service-oriented society. While in neurasthenia the businessman is the embodiment of the dynamic, overstimulated modern human being, in burnout, this is the emotionally overloaded human services professional. The businessman is the victim of hectic pace of life, the professional of psychosocial demanding life.

It has been argued that a particular social, cultural, and economic constellation existed in the United States that led to the ‘discovery’ of burnout the 1970s (Schaufeli et al., Reference Schaufeli, Leiter and Maslach2009). For example, President Lyndon B. Johnson’s ‘War on Poverty’ caused a large influx of idealistically motivated young people into human services professions. However, after struggling to reduce poverty for over a decade, they found themselves increasingly disillusioned. Their frustrated idealism was critical to the concept’s momentum: the experience of burnout was not merely an inconvenience or an occupational hazard but rather a damaging attack on their professional identity. Second, starting in the late 1950s, human services rapidly professionalized and bureaucratized because of increasing government and state influence. Particularly, small-scale, traditional agencies where work was considered a calling changed into largescale formalized organizations. The frustration and disillusionment resulting from the clash of utilitarian organizational values with providers’ professional values further contributed to the spread of burnout. Third, the cultural revolution of the 1960s and 1970s weakened professionals’ authority so that their traditional prestige was no longer self-evident. At the same time, entitled and empowered recipients expected much more than ever before. Together, these two trends increased the emotional demands of professional work considerably, hence contributing to burnout as well.

The three factors mentioned above are more or less specific for human services, the occupational group where burnout was observed first. However, other sociocultural developments also seem to have contributed to the emergence and proliferation of burnout. Notably, since the World War II, a process of social fragmentation unfolded whereby traditional social communities and networks such as the church, neighborhood, community, and family have gradually eroded. This resulted in increased individualization and decreased community support, thereby fostering the ‘corrosion of character’ (Sennett, Reference Sennett1998), a notion similar to burnout. In parallel, a ‘narcissistic culture’ (Lasch, Reference Lasch1979) developed, characterized by transient, unrewarding social relationships which produce self-absorbed, manipulative individuals who remain perpetually unsatisfied. As Farber (Reference Farber and Farber1983, p. 11) noted, the combination of these trends toward social fragmentation, individualization, and narcissism produces ‘a perfect recipe for burnout’.

After its initial emergence in the United States, the concept of burnout rapidly spread across the globe. First it was introduced in Europe, followed by all other continents. Roughly speaking, the order in which the interest in burnout seems to have spread corresponds with the country’s socioeconomic development. For instance, currently, the economies of and China and India are booming, and burnout attracts attention in these countries as well. It has been suggested that globalization, privatization, and liberalization cause rapid changes in modern working life, such as learning new skills, new types of work, pressure of higher productivity and quality of work, and hectic, time-pressured jobs, which, in their turn, may produce burnout – particularly in rapidly developing countries (Kulkarni, Reference Kulkarni2006). In addition, social fragmentation, individualization, empowerment of recipients and customers, and the rise of ‘me-culture’ are pervasive in not only the United States and Europe but also other countries in the world.

Yet, ‘burnout’ does not necessarily mean the same thing in different countries. For instance, in North America it signifies a crisis in the relationship of employee with their jobs, whereas in welfare states in continental Europe, burnout is considered a psychological disorder. In other words, in North America burnout is a nonmedical, psychological term that is colloquially used, but in many European countries it is also used as a medical diagnosis that serves as an entry ticket which legitimizes the use of public social and health services. In some European countries, burnout is used as an officially recognized occupational disease (Lastovkova et al., Reference Lastovkova, Carder, Rasmussen, Sjoberg, Groene, Sauni, Vevoda, Vevodoca, Lasfargues, Svartengren, Varga, Colosio and Pelclova2018). That means that workers suffer from it are eligible for treatment and financial compensation in case of sickness absence or work.

Concept and Definition

Initially, in the pioneering phase of burnout research, attempts were made to characterize burnout by means of careful but unstandardized observations and case studies. This resulted in a laundry list of 132 symptoms associated with burnout that ranged from A (anxiety) to Z (loss of zeal) (Schaufeli & Enzmann, Reference Schaufeli and Enzmann1998). These symptoms can be grouped into affective (e.g., anxiety, depressed mood), cognitive (e.g., forgetfulness, intellectualization), physical (e.g., headaches, muscle pain), behavioral (e.g., impulsivity, procrastination), and motivational (e.g., loss of zeal and interest) symptoms. Moreover, burnout symptoms were identified at not only the individual level but also the interpersonal (e.g., stereotyping recipients, blaming the victim) and organizational level (e.g., poor work performance, absenteeism). This laundry list approach is not very helpful because it is overinclusive; by listing all conceivable symptoms, the concept itself becomes meaningless. Yet, it gives an impression of the kind of symptoms that are involved.

In addition to identifying its symptoms in this pioneering phase, burnout was also considered as a process that unfolds across time. For instance, Edelwich and Brodsky (Reference Edelwich and Brodsky1980) described burnout as a process of increasing disillusionment; that is, as a ‘…progressive loss of idealism, energy and purpose experienced by people in the helping professions as a result of conditions in their work’ (p. 14). More specifically, they distinguished four progressive stages: enthusiasm, stagnation, frustration, and apathy. In the final stage, a full-blown burnout developed that includes symptoms like those listed above. Although such stage models have an intuitive appeal and therefore tend to be popular, they have not been confirmed empirically and are only based on anecdotal evidence.

The most influential definition of burnout originates from Maslach and Jackson (Reference Maslach and Jackson1981), which marks the start of a next empirical phase: ‘Burnout is a syndrome of emotional exhaustion, depersonalization, and reduced personal accomplishment that occurs frequently among individuals who do “people work” or some kind’ (p. 99). In addition to emotional exhaustion and depersonalization (i.e., an impersonal attitude toward recipients), a third factor was distinguished: reduced personal accomplishment, or the tendency to judge oneself negatively in contact with recipients with whom one works professionally. The basic assumption was that working with people requires a lot of energy, effort, attention, and empathy and is therefore emotionally demanding. It was argued that high emotional demands are the root cause of burnout (see also ‘Explanations’).

Soon after the introduction of the Maslach Burnout Inventory (MBI) – a short, easy-to-administer self-report questionnaire – burnout research started to boom. Please note that initially – by definition – burnout could only occur among those who work with other people. So, basically, the massive use MBI fueled the self-fulfilling prophecy that burnout is limited to the human services and cannot be found in other occupational groups. This changed in the mid-1990s with the introduction of the general version of the MBI (Maslach et al., Reference Maslach, Jackson and Leiter1996). From then on, burnout could be assessed in any occupational group. In the process the original three constituting components of burnout were relabeled: emotional exhaustion turned into exhaustion, depersonalization into cynicism, and reduced personal accomplishment into lack of professional efficacy. The exhaustion component represents the stress dimension of burnout and refers to feelings of being overextended and depleted of one’s mental and physical resources. The cynicism component represents the attitudinal dimension of burnout and refers to a negative or excessively detached attitude to various aspects of the job. Finally, reduced efficacy represents the self-evaluation dimension of burnout, referring to feelings of incompetence and a lack of achievement and productivity at work.

Soon after its introduction in the early 1980s, the MBI became the gold standard to assess burnout. It was used in 93% of all scientific papers on burnout that appeared until 1996 (Schaufeli & Enzmann, Reference Schaufeli and Enzmann1998, p. 71). Almost two decades later, the MBI continues to dominate the field; it still is the instrument of choice in 88% of all scientific papers (Boudreau et al., Reference Boudreau, Boudreau and Mauthe-Kaddoura2015). This supremacy has the undesirable side effect that the concept of burnout corresponds with the way in which it is assessed. In other words, burnout is what the MBI measures, and vice versa. Needless to say, that this circularity is undesirable because it impedes new and innovative research that leads to a better understanding of burnout. On a positive note, the supremacy the MBI is conducive for conducting meta-analyses.

The conceptualization of burnout from Maslach and her colleagues is not completely uncontested. A discussion raged about its dimensionality from the outset. For instance, Pines et al. (Reference Pines, Aronson and Kafry1981) described burnout as ‘a state of physical, emotional and mental exhaustion caused by long-term involvement in situations that are emotionally demanding’ (p. 9). In other words, burnout was reduced to mere exhaustion, and consequently their Burnout Measure yields a single score. In a similar vein, Shirom and Melamed (Reference Shirom and Melamed2006) distinguished between physical fatigue, emotional exhaustion, and cognitive weariness, and Kristensen et al. (Reference Kristensen, Borritz, Villadsen and Christensen2005) between physical and psychological exhaustion. Also, their ‘burnout’ measures produce a single, overall exhaustion score. All authors above argue that the correlations among the three dimensions of the MBI are not strong enough to justify calling them a syndrome. In other words, they consider the dimensions of cynicism and inefficacy as incidental and therefore unnecessary for understanding the presumed unitary nature of burnout. Exhaustion is indeed a hallmark of burnout. However, although it is a necessary criterion, it is not sufficient. If it were, then it should be recognized for what it is and plainly labeled ‘exhaustion’. In short, it would not make much sense to use a novel label (‘burnout’) for an already existing construct (‘exhaustion’).

Also, Schaufeli and Taris (Reference Schaufeli and Taris2005) criticized the approach of Maslach and her colleagues on conceptual grounds. In their view, the essence of fatigue is both the inability and the unwillingness to spend effort, which is reflected by the energetic and motivational components of burnout, respectively. The unwillingness to perform manifests itself by increased resistance, reduced commitment, lack of interest, disengagement, and so on – in short, mental distancing. Thus, inability (exhaustion) and unwillingness (mental distancing) constitute two sides of the same burnout coin. Moreover, they argued that in the case of normal, occasional occupational fatigue, the unwillingness to spend effort is functional because it fosters taking a break to recuperate or switching to another task and will therefore reduce fatigue. In chronic fatigue (burnout), however, the protective distancing mechanism is dysfunctional because it has habituated into relatively permanent impaired motivation. Mental distancing has become part and parcel of occupational life: instead of a solution, it is now part of the problem.

Furthermore, Schaufeli and Taris (Reference Schaufeli and Taris2005) argue that instead of a constituting component of burnout, inefficacy should be considered a consequence. As such, personal accomplishment or professional efficacy can be seen as a consequence of being unable and unwilling to no longer spend effort at work. After all, because people cannot and no longer want to perform at work, they feel inefficacious. This view is supported by two longitudinal studies that showed that exhaustion leads to depersonalization and depersonalization leads, in turn, to reduced personal accomplishment (Taris et al., Reference Taris, Le Blanc, Schaufeli and Schreurs2005).

Most recently, building on (1) the theorizing of Schaufeli and Taris (Reference Schaufeli and Taris2005), (2) a review of current burnout instruments, and (3) in-depth interviews with professionals who deal with burned-out employees on a daily basis, burnout was reconceptualized as ‘a work-related state of exhaustion that occurs among employees, which is characterized by extreme tiredness, reduced ability to regulate cognitive and emotional processes, and mental distancing’ (Schaufeli, De Witte, & Desart, Reference Schaufeli, Desart and De Witte2020a, p. 4). In addition to the inability (exhaustion) and unwillingness (distancing) to perform, cognitive and emotional impairment are added as distinct components that, for instance, refer to poor concentration and overreacting emotionally, respectively. As a result of exhaustion, the person’s functional capacity to regulate cognitive and emotional processes is deteriorated. Seen from that perspective, cognitive and emotional impairment are specific types of exhaustion that manifest themselves in the cognitive and emotional domain, respectively. Clearly, there is a certain parallel with the approaches of Pines et al. (Reference Pines, Aronson and Kafry1981) and Shirom and Melamed (Reference Shirom and Melamed2006), who also distinguished between various forms of exhaustion. As in the definition of Maslach and her colleagues, exhaustion and cynicism (mental distancing) are included, but reduced professional efficacy is excluded for conceptual reasons (see above). Besides, it was also not mentioned in the in-depth interviews with professionals (Schaufeli et al., Reference Schaufeli, Desart and De Witte2020a).

A recurrent theme in the discussion whether or not burnout is ‘old wine in new bottles’ is its relationship with depression. A seminal review and meta-analyses including 92 studies on the burnout–depression overlap (Bianchi et al., Reference Bianchi, Schonfeld and Laurent2015) concluded that: (1) most burned-out employees also exhibit depressive symptoms; (2) burnout and depression scores are moderately to highly correlated, particularly as far as the exhaustion component is concerned; (3) results regarding the causal link between burnout and depression are heterogeneous; (4) somatic and biological levels of analysis seem to suggest some degree of distinctiveness; and (5) burnout and depression associate differently with both job-specific (burnout) and generic (depression) factors. Hence, it seems that the distinction between burnout and depression is – at least to some extent – supported by empirical research. This is also confirmed by a more recent meta-analysis that revealed no conclusive overlap between burnout and depression, ‘…indicating that they are different and robust constructs’ (Koutsimani et al., Reference Koutsimani, Montgomery and Georganta2019, p. 1).

Assessment

Basically, burnout can be assessed in two ways: by using self-report questionaries and by means of a structured, clinical interview. Most psychological researchers use the former, considering burnout a continuous variable, whereas occupational physicians and occupational health psychologists prefer the latter, considering it a dichotomous variable. Psychological research focusing on working and relatively healthy employees with only mild burnout symptoms is abundant. In contrast, research is relatively scarce among those with a burnout disorder who suffer from severe symptoms and are not able to work. Below, the assessment of burnout with self-reports is discussed first, followed by diagnostic guidelines that are employed to assess burnout as a mental disorder.

Prevalence

For proper epidemiological research to estimate the prevalence of burnout, two requirements need to be met: (1) the use of a valid and reliable questionnaire with appropriate cut-off values; and (2) a representative sample. Although the MBI can be considered a valid and reliable burnout measure, cut-off values to discriminate burnout cases from noncases are lacking. The current MBI test-manual even explicitly warns against using cut-off values by including a note of caution that reads: ‘It is important to understand that there is no definitive score that “proves” a person is “burned out”’ (Maslach, Jackson & Leiter, 2016, p. 23). In other words, the MBI should not be used to estimate the prevalence of burnout. Nevertheless, previous versions of the test-manual (Maslach & Jackson, Reference Maslach and Jackson1986; Maslach et al., Reference Maslach, Jackson and Leiter1996) specified cut-off values that have been used extensively for that very purpose. Those cut-off values corresponded with the upper third of the score distribution of particular reference groups from previous test-manuals such as teachers and medical and social services professionals. Unfortunately, these were nonrepresentative samples and data were gathered in the late 1970s, which, in fact, makes the MBI cut-offs useless and outdated.

Given the problematic nature of the MBI cut-offs, it is not surprising that estimates of the prevalence of burnout are uninformative. For instance, a systematic review that includes 131 studies among physicians concludes that the prevalence of burnout ranges from 0% to 87% (Rotenstein et al., Reference Rotenstein, Torre, Ramos, Rosales, Guille, Sen and Mata2018). Hence, the authors conclude: ‘These findings preclude definitive conclusions about the prevalence of burnout’ (p. 1145). In a similar vein, Adriaenssens et al. (Reference Adriaenssens, De Gucht and Maes2015) found that the prevalence of burnout among emergency nurses ranges across 17 studies from 9% to 76%. Another systematic review including 113 studies reports an overall pooled prevalence of 11% among nurses from 49 different countries (Woo et al., Reference Woo, Ho, Tang and Tam2020). However, significant heterogeneity exists in the estimated burnout rates so that the pooled prevalence is not very informative, as the authors note.

Proper epidemiological research was carried out in Sweden’s two most northern counties using a large random sample (Norlund et al., Reference Norlund, Reuterwall, Höög, Lindahl, Janlert and Birgander2010). To assess burnout, the SMBM was employed with validated cut-offs that were based on a reference group that meets the Swedish diagnostic criteria for exhaustion disorder (see above). This study revealed that 13% had a high level of burnout – or, more precisely, show elevated exhaustion levels. Moreover, exhaustion is more prevalent among women (16%) compared to men (10%), is observed most frequently in the age group between 35 and 44 years, and decreases with age in both sexes. The latter might point to the so-called healthy worker effect: a selection bias whereby the least healthy workers have left their jobs, thus leaving their healthier colleagues.

The results of this Swedish study are remarkably similar to that of the Dutch Annual Working Conditions Survey, which is carried out among a large, representative sample of the working population. This survey uses the exhaustion scale of the MBI with cut-offs that are based on a reference group that meets the Dutch diagnostic criteria for burnout (see above). It was estimated in 2019 that 17% experience severe exhaustion, 18% among women and 16% among men (Houtman, Reference Houtman, Schaufeli and Bakker2020). Exhaustion is most prevalent among those aged between 25 and 34 years and rates for both sexes decrease with age. Moreover, exhaustion is most prevalent in teaching (23%), followed by health care (19%) and industry (18%). Finally, the prevalence of severe work-related exhaustion appears to be more or less stable across time, initially fluctuating around 11% from 1997 till 2012, after which it increased gradually to 17% in 2019.

The only study so far that compares national representative samples from various countries (De Beer et al., Reference De Beer, Schaufeli, De Witte, Hakanen, Shimazu, Glaser, Seubert, Bosak, Sinval and Rudnev2020) showed that, using the BAT, burnout levels in Japan are significantly higher compared to six European countries (Austria, Belgium, Finland, Germany Ireland, the Netherlands).

In conclusion, the MBI is not suitable for epidemiological research, and when it is nonetheless used, the prevalence of burnout varies too much across studies to draw any meaningful conclusions. Employing questionnaires that tap exhaustion only, it was estimated in Sweden and the Netherlands that between 13% and 17% of the working population suffer from this hallmark of burnout. In both countries, exhaustion was more prevalent in women and younger age groups and decreased with age.

Antecedents, Consequences, and Correlates

The main antecedents, consequences, and correlates of burnout, which emerged mostly from systematic reviews and meta-analyses, are summarized below. The Job Demands-Resources (JD-R) model is used to structure this section (Bakker & Demerouti, Reference Bakker and Demerouti2016). Before discussing the main empirical findings, two notes should be made.

First, the vast majority of empirical burnout studies is cross-sectional in nature, which precludes drawing any conclusions about its causes and effects. Using the JDR-model, Lesener et al. (Reference Lesener, Gusy and Wolter2019) analyzed the results of 74 longitudinal samples and found that job characteristics (i.e., job demands and job resources) are reciprocally related to burnout. This means, for instance, that work overload leads to burnout, but also the other way around, that employees who feel burned out at a certain point in time are likely to perceive a higher future workload. Hence, perceived work overload may act as an antecedent as well as a consequence of burnout.

Second, various antecedents may interact with each other. Negative effects of job demands, such as work overload or role problems, may be compensated or buffered by job resources, such as job control and social support (Xanthopoulou et al., Reference Xanthopoulou, Bakker, Dollard, Demerouti, Schaufeli, Taris and Schreurs2007). These may mitigate these effects and thus protect employees from burning out. Recently, a meta-analysis of 48 longitudinal studies found that the ‘stressor-effect’ (job demands ➔ burnout) is smaller than the reverse ‘strain-effect’ (burnout ➔ job demands) and that the latter is moderated (i.e., reduced) by job control and social support (Guthier et al., Reference Guthier, Dormann and Voelkle2020). Supplementary analyses suggest that the reciprocal relations are mainly driven by exhaustion and that exhaustion and depersonalization/cynicism are reciprocally related across time.

Hence, it can be concluded that the longitudinal relationships between job characteristics and burnout are more complex and go beyond the simple idea that job stressors and lack of job resources ‘cause’ burnout. Hence, the phrase potential antecedents and consequences is used below. First, potential antecedents and consequences of burnout are discussed, followed by its correlates, such as person and sociographic factors. Next, the potential consequences and correlates of burnout will be examined, respectively.

Explanations

To date, a specific overall psychological theory that explains burnout does not exist. Instead, the picture is rather scattered when it comes to explaining the underlying psychological processes that play a role in burnout. Early explanations focused on the individual, interpersonal, or organization factors as the main causal agents (for an overview, see Schaufeli & Enzmann, Reference Schaufeli and Enzmann1998, pp. 100–142). But most of these theoretical explanations are speculative and lack sufficient empirical support. This section discusses three theoretical explanations that can draw on empirical evidence and in which emotional overload, lack of reciprocity, and emotional contagion play a key role, respectively. Two descriptive frameworks are considered that are helpful for integrating the results from burnout research discussed in the previous section.

Interventions

A curious discrepancy exists regarding interventions to reduce burnout. On the one hand, a true burnout industry has emerged over the years with countless providers of just as many different, sometimes even very exotic, interventions. Yet, on the other hand, relatively few solid, well-designed studies on their effectiveness have been carried out. Based on the systematic reviews and meta-analyses discussed below, it is estimated that only .004% of all publications evaluate the effectiveness of burnout interventions.

In principle, burnout interventions may target either the individual or the organization. Most interventions, by far, focus on the individual, probably because organizational interventions are more difficult to implement as they may interfere with work and interpersonal processes at the shop floor. Examples of individual approaches are cognitive–behavioral interventions (e.g., stress inoculation training, rational emotive therapy), time-management training, and relaxation, mindfulness, and interpersonal skills training, whereas examples of organizational interventions include the introduction of new human resources (HR) practices, job redesign (e.g., task restructuring), time rescheduling, and management development (for more detailed descriptions, see Schaufeli & Enzmann, Reference Schaufeli and Enzmann1998, pp. 143–183).

The picture that emerges from various meta-analyses is rather unambiguous and consistent: burnout interventions have a statistically significant, small positive effect with effect sizes (Cohen’s d about .20) This is illustrated by a meta-analysis of 47 controlled studies among different occupational groups, of which 96% focus on individual interventions (Marţicuroiu, Reference Maricutoţiu, Sava and Butta2014). Results show significant but small effects on general burnout (Cohen’s d = .22) and exhaustion (d = .17), but not on cynicism/depersonalization (d = .04) or reduced professional efficacy (d = −.02). The strongest positive effect is found for relaxation, followed by learning new role behavior and cognitive–behavioral interventions, respectively.

More or less similar results are obtained by various meta-analyses that focused on specific occupational groups. For instance, a meta-analysis of 23 controlled intervention studies among teachers (Iancu et al., Reference Iancu, Rusu, Măroiu, Păcurar and Maricuțoiu2018) shows small yet statistically significant positive effects on overall burnout (d = .18), exhaustion (d = .18), and reduced efficacy (d = .14), but not on depersonalization (d = .03). The strongest effects are found for cognitive–behavioral approaches, mindfulness and stimulating professional development, and, to a somewhat lesser extent, learning how to organize social support. A meta-analysis of 27 controlled intervention studies among mental health professionals performed by Dreison et al. (Reference Dreison, Luther, Bonflis, Sliter, McGrew and Salyers2018) shows small, positive effects on overall burnout (Hedges’ g = .13), exhaustion (g = .20), and depersonalization (g = .15), but not on reduced efficacy (g = .08). Subsequent moderator analyses shows that the effect of person-oriented interventions is stronger than that of organization-oriented interventions.

Physicians are the most often studied occupational group. A meta-analysis identified 15 randomized clinical trials (RCTs) and 37 controlled cohort studies carried out among physicians and find significant intervention effects on overall burnout, emotional exhaustion, and depersonalization, with absolute symptom levels decreasing on average 10%, 14%, and 4%, respectively (West et al., Reference West, Dyrbye, Erwin and Shanafelt2016). Lack of efficacy is not considered in this study. Most interventions focus on individual physicians (e.g., mindfulness, stress management and self-care training, communication skills training). Another meta-analysis including 20 interventions analyzed interventions to reduce physician burnout (Panagioti et al., Reference Panagioti, Panagopoulou, Bower, Lewith, Kontopantelis, Chew-Graham, Dawson, van Marwijk, Geraghty and Esmail2017). The database of this analysis partly overlapped with the previous meta-analysis but only focused on emotional exhaustion, and, most importantly, additional subgroup analyses were carried out. First, and not surprising, a small but significant positive intervention effect was found (d = .29). Second, this effect appeared to be stronger for organization-oriented interventions (d = .45), such as rescheduling hourly shifts, reducing workload, discussion meetings to enhance teamwork, leadership, and structural changes than for physician-oriented interventions (d = .18), such as mindfulness, stress reduction techniques, exercise, and educational interventions targeting physicians’ self-confidence and communication skills. Third, no difference in effectiveness of interventions was shown for more vs. less experienced physicians or for those working in primary vs. secondary care settings.

Typically, all meta-analyses discussed above include preventive interventions among those without or with mild burnout complaints who are still working. But what about the effectiveness of treating those with severe burnout symptoms? Ahola et al. (Reference Ahola, Toppinen-Tanner and Seppänen2017) identified 18 studies, of which four were RCTs, that evaluated the effectiveness of burnout treatment programs. Mostly this involved individual cognitive–behavioral therapy, but also group interventions based on psycho- and sociodrama methods. Unfortunately, their meta-analysis did not show a significant decrease in burnout complaints across the intervention groups compared to the waitlist control groups. A disappointing result indeed, in terms of both the number of studies and their effect. A Swedish follow-up study showed that seven years after burnout treatment finished, only 16% fully recovered, whereas 73% reported reduced stress tolerance, 46% extreme fatigue, and 43% problems with memory (Glise et al., Reference Glise, Wiegner and Jonsdottir2020). So, it seems that residual symptoms are very persistent in burnout disorder.

In addition to symptom reduction, return to work is another important outcome of burnout interventions. Although both are related, they do not completely overlap. For instance, even if burnout symptoms have not completely disappeared, gradual return to work is possible. Indeed, modern disability management approaches assume that return to work fosters symptom reduction (Waddell & Burton, Reference Waddell and Burton2006). This view contradicts the traditional approach, in which return to work should only be considered after all symptoms have disappeared.

How effective are interventions that foster return to work after burnout? A meta-analysis of eight interventions that are targeted at return to work among those with clinically significant burnout do not find significant results compared to a waitlist control group (Perski et al., Reference Perski, Grossi, Perski and Niemi2017). The extent to which burned-out employees fully returned to work does not differ whether they had received treatment, were on the waiting list, or received ‘treatment as usual’. However, compared to the control group, the intervention group returns to work partially more quickly. Remarkably, burnout, anxiety, and depression symptoms do not decrease significantly as a result of return-to-work interventions. So, despite the fact that symptom levels do not decrease, interventions lead to faster partial but not full return to work. This illustrates that symptom reduction and return to work are indeed two relatively independent outcomes.

In conclusion, it seems that burnout symptoms can be decreased, particularly as a result of person-oriented interventions, such as relaxation, mindfulness, and cognitive–behavioral approaches. This is most convincingly shown for general burnout and exhaustion, and less so for cynicism/depersonalization and personal efficacy. This pattern is similar across various occupational groups, such as teachers, mental health care workers, and physicians. One meta-analysis found that organization-oriented interventions were more effective than interventions that were targeted to individual physicians. It is important to note that these positive intervention effects apply to those with relatively mild complaints. It seems that interventions for those with severe burnout symptoms are less effective, though. In addition to symptom reduction, some interventions also explicitly aim to foster return to work. These interventions are partly successful because they lead to partial work resumption but not to symptom reduction or full work resumption.

Conclusions and Outlook

Burnout – a metaphor that refers to a severe loss of mental energy – can be seen as a modern affliction that emerged in the 1970s due the transformation from an industrial into a postindustrial society. In a way, this parallels the rise of neurasthenia – a syndrome akin to burnout – at the end of the 19th century when the agricultural society transformed into an industrial society. Originally, it seemed that burnout occurred exclusively among human services professionals who work with other people, but this resulted from the fact that the questionnaire for measuring burnout was designed to be used in this occupational group only. Once a general version of the MBI – the leading questionnaire that serves as the gold standard to assess burnout – was developed, burnout was also observed in other occupational groups. The MBI, which boasts good psychometric qualities, has been challenged by other questionnaires that reduce burnout to mere exhaustion. However, this boils down to throwing the baby out with the bath water because it denies the specificity of burnout. After all, conceptually speaking, burnout is a chronic, work-related state of fatigue that is characterized by the inability and the unwillingness to spend effort, representing its energetic and motivational component, respectively. Hence, inability (exhaustion) and unwillingness (mental distancing) are two sides of the same burnout coin.

Throughout this chapter, it became clear that mild burnout symptoms have to be distinguished from severe symptoms that are indicative for a burnout disorder. On the one hand, burnout may be considered a multidimensional phenomenon that can be measured using continuous scales, while on the other hand, it may also be considered dichotomously in terms of a diagnosis that is based on a clinical interview. Typically, most psychological research focuses on mild burnout symptoms – as assessed by self-reports – among those who are working. This means that our current knowledge of burnout is mainly based on workers with mild symptom levels. This knowledge can be used to prevent the development of burnout. Indeed, various successful interventions have been developed, at both the individual and the organizational level. In contrast, research among those who suffer from severe burnout and therefore dropped out of work is relatively scarce. Hence, more research is needed on severe burnout. An important prerequisite is that agreement exists about the diagnostic guidelines for burnout disorder that goes beyond the mere description that is included in the ICD-11, which is essentially based on the original definition of Maslach and Jackson (Reference Maslach and Jackson1981). In order to boost research on severe burnout, clinically validated cut-off values should be available, as for instance, in the BAT; Schaufeli et al., Reference Schaufeli, De Witte and Desart2020b).

Unfortunately, our current knowledge on the prevalence of burnout is rather limited because proper epidemiological studies are lacking. It seems that about 15% of the working population shows elevated levels of exhaustion. But, as was noticed above, burnout goes beyond mere exhaustion. Estimates of burnout prevalence in various occupational groups vary for 0% to almost 90%, which, of course, is rather uninformative (see ‘Prevalence’). The reason for this exceptionally wide range is that virtually all studies use neither representative samples nor validated cut-off values so that results cannot be compared. Representativity and the use of similar, valid cut-offs are necessary for a reliable estimation of the prevalence of burnout. Unfortunately, the MBI cannot be used for epidemiological research because it was not designed for that purpose; it yields three scores that should not be combined into a single, overall burnout score (Maslach et al., Reference Maslach, Leiter and Jackson2017). In addition, the MBI lacks validated cut-off scores to discriminate between burnout ‘cases’ and ‘noncases’. Although the MBI can be used to separately assess and compare mean levels of exhaustion, cynicism/depersonalization, and reduced efficacy between groups, it cannot establish level of burnout as such. This points to a fundamental ambiguity of the MBI: burnout is defined as a ‘syndrome’, yet only its components can be measured separately. The recently developed BAT overcomes this problem as it produces a single burnout score (Hadzibajramovic et al., Reference Hadzibajramovic, Schaufeli and De Witte2020).

A host of potential antecedents, consequences, and correlates of burnout has been uncovered (see ‘Interventions’), but it also appears that the idea that burnout mediates the effects of unfavorable job characteristics on negative personal and organizational outcomes is an oversimplification. Instead, it seems that reciprocal causation plays a role as well. For instance, particular job demands, such as work overload, may act as not only antecedents but also consequences of burnout, thus pointing at a dynamic and bidirectional rather than a one-directional process. To further complicate the picture, person characteristics may play a moderating role in this process, for instance by buffering or exacerbating the negative impact of job characteristics on burnout. Taken together, it seems that the burnout process is dynamic and rather complex. Although the different elements of that process have been identified (job demands, lacking job resources, person factors, individual and organizational outcomes), future research should uncover how these elements are linked and how the process unfolds across time.

It seems that, unfortunately, the quest for biomarkers that can be linked to burnout has been in vain so far. Perhaps this might be due to a lack of sensitivity of the physiological measures applied or to poor sampling of burnout cases, so that it makes sense to continue research in this area with better measures and samples (Danhof-Pont et al., Reference Danhof-Pont, Van Veen and Zitman2011). But it might also be that we are simply on the wrong track by studying peripheral biomarkers and a better way forward would be to explore other biological pathways, such as brain activity or epigenetics. For instance, Van Luijtelaar et al. (Reference van Luijtelaar, Verbraak, van den Bunt, Keijsers and Arns2010) found different electrophysiological brain activity in burnout patients compared to health controls that also differs from patterns found in the literature for depression and chronic fatigue syndrome. Moreover, it is found that burnout severity of medical professionals can be explained by reduced empathy-related brain activity (Tei et al., Reference Tei, Becker, Kawada, Fujino, Jankowski, Sugihara, Muria and Takahashi2014) and that burnout severity is related to epigenetic changes (Bakusic et al., Reference Bakusic, Manosij, Polli, Bekaert, Schaufeli, Claes and Godderis2020). Hence, it seems worthwhile to further explore these avenues.

As far as the psychological explanations for burnout are concerned, the picture is somewhat scattered. No overarching burnout theory exists and probably never will, given the complexity of the phenomenon. However, two handy heuristic frameworks exist that allow for the integration of most burnout research; the Job-Demands Resources model and the Six Areas of Working Life approach (see ‘Explanations’). Although useful for practitioners who look for possible drivers of burnout, these frameworks do not explain the underlying psychological mechanisms of burnout. Three theoretical approaches supported by empirical evidence have been discussed in ‘Explanations’, in which emotional overload, lack of reciprocity in social exchange relations, and emotional contagion play a key role. However important these processes may be, given the nature of burnout as energetic and motivational dysfunction, it makes sense to also consider theories of work effort (Hockey, Reference Hockey2013) and motivation, such as conservation of resources theory (Hobfoll, Reference Hobfoll and Folkman2011) or self-determination theory (Deci & Ryan, Reference Deci, Ryan, Van Lange, Kruglanski and Higgins2012), for explaining burnout.

There is considerable empirical evidence for the effectiveness of burnout interventions, at the individual as well as the organizational level. Although the effects are small, they are consistent across various occupational groups and work settings and are relevant for preventing burnout. However, most interventions are targeted at the individual and carried out among relatively healthy workers with only mild burnout symptoms. Research on organizational interventions is scarce, as is research on the effectiveness of interventions for severe burnout. Clearly, more intervention studies are needed, particularly focusing on the organization and on those with severe burnout.

So, taken together, in order to increase our understanding of burnout, future research should focus on (1) severe burnout, using clinical interviews as well as clinically validated cut-offs; (2) the prevalence of burnout across various occupational groups, industries, and countries; (3) the dynamic nature of burnout and its development across time; (4) alternative biological pathways that go beyond peripheral biomarkers; (5) explanatory frameworks using theories on work effort and motivation; and (6) burnout interventions, particularly regarding severe burnout. Burnout is here to stay, as is burnout research.