The Association of Hobbies and Leisure Activities with Physician Burnout and Disengagement

exhaustion. Results: Significant differences were seen across age groups, genders, and physician specialties in the level of burnout ( p < 0.01, p < 0.01, p = 0.02, respectively) and job disengagement ( p < 0.01, p = 0.02, p < 0.01, respectively). Younger providers (age < 60) and women had higher levels of burnout. Trainees had higher levels of burnout than full time, part time or retired physicians. North American graduates reported a slightly higher rate of burnout and disengagement than international graduates. 93.9% of physicians viewed outside interests as a substantial mitigation factor for burnout and disengagement. Conclusion: Our study identified associations rather than causality. Nevertheless, emphasizing hobbies and non-medical outside interests might well prove useful to temper epidemic burnout among healthcare professionals. We especially encourage those hobbies with stronger social underpinnings. ABSTRACT


INTRODUCTION
Clinician burnout dramatically impairs physicians and secondarily extends to their patients, peers, students, staff, and organizations [2][3][4][5][6][7].The definition of burnout itself is problematic; it has been variably characterized as emotional exhaustion, cynicism and detachment from work, as well as a sense of low personal achievement [8,9].Burnout is now internationally recognized in the ICD-11 (International Classification of Diseases, Eleventh Revision) as a condition "resulting from chronic workplace stress" and encompassing a constellation of exhaustion, cynicism, and reduced effectiveness [10]."The cost of this epidemic is in the billions of dollars [11], with an overall prevalence reported between 40-75%, even before COVID-19 universally devastated hospitals and private practices.In China the incidence of burnout is even higher, with a total prevalence of over 80% [12], while a report from India places that number at > 90% [13].
In our study of the positive association of the effects of hobbies and leisure time on burnout, we found a subset of particular interest -the effect of socially oriented activities compared to more isolated leisure activities.
Although chronic stress is certainly a factor, burnout is not synonymous with stress.In a recent survey among 19,348 American College of Cardiology members, the number of cardiologists who reported feeling stressed actually declined from 49.5% in 2015 to 43.9% in 2019, but the number reporting burnout increased from 26.8% to 32% in that same period [14].Much of the etiology has resulted from accelerating workloads enabled by progressive technology-based demands (e.g., the electronic health record) and the resulting distancing from patients, in addition to ever increasing bureaucratic tasks, regulation, information overload, long hours, lack of respect from administrators, and so on [2,5,7,[15][16][17][18][19][20][21][22].
While the majority of burnout research and proposals is devoted to systemic change, some investigators have also written about development of hobbies and interests for mitigating burnout and enhancing engagement [2,[22][23][24][25][26][27][28][29][30][31]]-yet most have not focused on this potential form of relief.Sargent et al. [32], found that 58% of their respondents reported that devoting time at least weekly or more to hobbies and leisure activities resulted in less psychiatric morbidity, better marital relations, and, for faculty, less emotional exhaustion and improved levels of personal achievement.Oskrochi et al. [4] placed making time for hobbies on a list of "protective factors" for surgical residents.Sang Hee Kang et al. [21] reported that having hobbies significantly related to lower occupational stress for Korean surgeons, and Zwack and Schweitzer [33] characterized leisure-time activities in a study of 200 German physicians as among the major strategies that maintain resilience.These citations point to both the universality of burnout and the value of hobbies.
Extraordinary French philosopher, mathematician, physicist, inventor, and theologian, Blaise Pascal famously said: "All of humanity's problems stem from man's inability to sit quietly in a room alone.Pascal's uncannily prescient statement, written in the 1600s, surely applies to the isolation forced upon us by COVID-19.In pre-COVID 2018 Surgeon General Vivek Murthy had already described a "loneliness epidemic [34]," in which nearly a quarter of physicians reported suicidal thoughts or even attempts [35].During the COVID-19 era, with an intensely proliferated risk of seclusion to its psychosocial/medical/economic aspects [36][37][38][39][40][41][42], we ask: When in modern times has humanity ever been this globally isolated?
We are aware of only one major research work on the specific effect of a wide-ranging variety of hobbies and interests: in 2011 McManus et al. [43] conducted an exhaustive UK study.They concluded that avocation (leisure activities) positively correlated with engagement and vocation, but did not correlate with burnout.With the escalating crush of burnout, we are revisiting -a decade later and from across the pond -the relationship between burnout and engagement vs. hobbies/interests.We use the terms burnout and exhaustion interchangeably.Likewise, despite minor differences, we equate hobbies, non-work interests, and leisure time.

Study Population
From June to August 2018, an online questionnaire was sent to 23,316 practicing physicians and 2,065 residents and fellows; medical students were not included in this study.Participants were unpaid US-based members of the M3 Global Research healthcare practitioner panel.All panel members are validated yearly and have opted in to participate in market research studies.Participants were randomly selected, with recipients receiving a reminder email in the weeks following the original email.
Aside from hobby-related information, demographic data is presented in Table 1 (next page), including age, gender, practicing position, specialties, household settings, and job satisfaction.

Ethical Considerations
The MetroWest Medical Center IRB deemed this study exempt from IRB oversight.

Measurement of Burnout
Historically there have been multiple instruments developed for the assessment of burnout, with the Maslach Burnout Inventory the most widely used [44][45][46][47][48][49][50][51].It is based on three dimensions of burnout -namely, emotional exhaustion, depersonalization and (reduced) personal accomplishment.However, some have raised concerns about similarities in the subscales as being framed in the same direction, criticized as possibly inferior to scales with both positively and negatively worded items [44][45][46][49][50][51].The Copenhagen Burnout Inventory has been only used in small samples of physicians and other health care providers, primarily outside of the United States [44,51].The Mayo Clinic Well Being Index does not assess engagement [44].
We employed the Oldenburg Burnout Inventory (OLBI for our study (see Supplement Table 1, q8), which has been validated among diverse occupational groups in different continents as an alternative measurement of burnout [44][45][46][47][48][49][50][51].With 16 questions, it assesses the two core dimensions of burnoutexhaustion and disengagement (from work) [44][45][46][49][50][51] -utilizing both positively and negatively framed items.As exhaustion and engagement are two distinct aspects within the spectrum of burnout, a low level of exhaustion does not necessarily correlate with positive engagement at work.Thus, it is more advantageous to assess both aspects with the single instrument.The OLBI is also freely accessed online.

Hobbies and Leisure Activities
Based on previous relevant studies, newsletters, surveys and internal discussions, we grouped 117 hobbies (Supplement Table 3, q16-32) into 15 major "categories" of activities (Supplement Table 3, q14).Examples of "categories" included team sports, individual sports, playing musical instruments, reading for pleasure, writing, and visual arts.
Respondents were asked to describe their personal activity status for each of the 15 categories in the last six months as "currently active," "formerly active," or "rarely/never participated."Respondents then answered questions about individual hobbies within their selected categories.To minimize respondent wear-out and keep the survey length to a manageable duration, each respondent answered questions about individual hobbies in only up to five of their currently active categories.Topics included the frequency of participation for individual hobbies.We also examined "high frequency participation" (≥ a few times per week) vs. "low frequency participation (≤ a few times per month) to assess if participation frequency had an impact on work exhaustion or disengagement.Additionally, we asked those respondents who were no longer active in a hobby for the reasons for discontinued participation (Fig. 6).

Statistical analysis
The associations of demographic features and hobbies with exhaustion and disengagement were examined.The OLBI score was analyzed as a continuous variable, with the mean exhaustion and disengagement scores calculated respectively.To enable interpretation of data from the bi-directional items, all positively phrased OLBI questions were reverse-coded for analysis.Thus, in our study every question with a higher Likert Scale (1-4) score relates to more exhaustion or more disengagement.
The differences of mean exhaustion and disengagement scores among variables were conducted by one-way analysis of variance (ANOVA).The differences of mean exhaustion and disengagement scores between each hobby (currently active, formerly active, rarely/never participated) were examined by a t-test, with p < 0.05 considered statistically significant.All statistical analyses were performed with IBM SPSS statistics 22.0 software.

RESULTS
A total of 2,563 physicians (11%) and 512 trainees (24.8%) returned the questionnaire.Respondents were geographically distributed roughly evenly from the Northeast, South, Midwest and West regions of the United States.

Demographics
Exhaustion and disengagement levels showed significant differences among different age groups, genders and practicing positions (p < 0.01, p < 0.01, p = 0.02, respectively).By decade, trainees and physicians aged 20-59 had similar levels of exhaustion, higher than physicians 60-69, and in turn higher than those ≥ 70.Female physicians had higher levels of exhaustion and minimally higher disengagement than males.International graduates reported slightly lower burnout and disengagement than US/Canadian graduates.Exhaustion and disengagement levels were similar among various cultural backgrounds.Religious/spiritual groups and volunteer activities, either medically or non-medically related, were associated with reduced exhaustion and disengagement.
Physicians in various practice settings reported similar exhaustion and disengagement levels, but they differed significantly among specialties (p = 0.02).The five specialties associated with highest exhaustion level in our survey were Radiology, Nephrology, Rheumatology, Internal Medicine, and Physiatry/Sports Medicine; the five specialties associated with highest disengagement level were Physiatry/Sports Medicine, Radiology, Urology, Internal Medicine and Family/General Practice.Subgroups having less than 10 respondents were excluded from analysis (Table 2 ).
Physicians' Job-related Attitudes vs. Burnout and Disengagement (Fig. 1) Physicians who had more frequent positive perspective of their jobs reported less exhaustion and disengagement, while 61.0% would likely choose medicine if they were to re-choose a career.The majority of physicians (93.9%) considered it as important or extremely important to have outside interests; 84.3% felt that outside interests helped relieve stress and prevent exhaustion at least to some extent, while 65.6% felt that outside interests enabled better engagement in their career.

Time Spent With Family (Fig. 2)
Of the total respondents, 1,585 (51.5%) were able to spend daily quality time with family in the household and 1,323 (43.0%) reported spending quality time with friends and family outside the household a few times a month.The more quality time physicians spent with their family and friends, the less exhaustion and disengagement they reported.

Hobbies
A. Status of hobbies associated with exhaustion and disengagement level (Fig. 3) Across the 15 major categories of hobbies, exhaustion was significantly lower in physicians who had active hobbies compared with those who had no hobbies or had given up certain hobbies (p ≤ 0.02, p ≤ 0.03, respectively).A similar trend was found in disengagement levels.

B. Favorite hobbies and exhaustion level (Fig. 4)
Despite being among favorite hobbies by the majority of respondents, listening to music (C4) and watching TV, movies, internet, video games or other media at home (C11) are associated with the highest level of exhaustion.On the other hand, participating in team sports (C1) and group games/role play (C12) are associated with the lowest level of exhaustion.

C. Numbers of active hobbies associated with exhaustion and disengagement level (Fig. 5)
With increasing numbers of active hobbies, respondents reported decreasing levels of exhaustion and disengagement.Conversely, the more hobbies given up, the higher levels of exhaustion and disengagement were reported.Similar trends were found in all demographic groups.

D. Time spent on hobbies
Among 117 individual hobbies, more participants practiced their interests/hobbies a few times per month or less.An average of 64% practiced individual hobbies at low frequency (≤ a few times per month), while 32% practiced individual hobbies at high frequency (≥ a few times per week).The frequency of practicing hobbies did not impact on burnout or disengagement.

E. Reasons for giving up hobbies (Fig. 6)
Among the 15 major categories of hobbies, participating in team sports (n = 683), playing a musical instrument (n = 518), outdoor activities (n = 399) and playing individual sports or exercise (n = 390) were the top hobbies previously enjoyed the most, but had given up.No longer having time was the dominant reason to give up hobbies.
We have examined the association of a set of physician hobby categories and subcategories with burnout and engagement.These two areas differ from each other, and indeed, McManus al. [43] found that leisure activities correlated positively with engagement, although not so with burnout.As opposed to that work, however, our own findings show that both engagement and burnout are improved in association with hobbies and interests.But the variables are complex, and hobby categories are dissimilar across studies [13,19,63].We found that women reported higher levels of burnout and disengagement than men, consistent with many reports globally [2-4, 6, 23].We found, too, that trainees and physicians aged 20-59 experienced similar exhaustion, more pronounced than in physicians aged 60-69, and higher still than those ≥ 70.But here the literature becomes much murkier, with descriptions of the greatest burnout among the youngest physicians ≤ 40 [21] or those 45-54 [5], or the lowest burnout (among women, at least) at age 50 [6].An extensive metanalysis of 182 studies of nearly 110,000 individuals in 45 countries concluded burnout by age simply cannot be reliably determined [9].
Additionally, we found the five highest levels of burnout in Radiology, Nephrology, Rheumatology, Internal Medicine, and Physiatry/Sports Medicine.Three of those same specialties (Physiatry/Sports Medicine, Radiology, Internal Medicine), were also among the highest levels of disengagement (Table 2).As for predictors of burnout, decreased leisure time and fewer years of practice were among the strongest factors among 209 transplant surgeons [27], while orthopedic surgeons reported hobbies were associated with less burnout [17].It is unknown if there are specialty-specific factors that lead to higher rates of disengagement or burnout.
Our study results are consistent with many self-help articles that focus on positive attitudes as a major influence on happiness and energy.We found that a more positive work outlook correlates with both lower burnout and disengagement.Additionally, nearly 94% of physicians felt it was important or extremely important to have outside interests, 84% identified them as a stress reliever, and 66% as associated with better work-based engagement.Time spent with family and friends similarly lessens burnout and disengagement.
Among our 15 activity categories, physicians with ongoing active hobbies reported the least stress, with more stress associated with an absence of hobbies, and the most stress among those with discontinued hobbies.We found that the number of active hobbies is inversely associated with exhaustion and disengagement, while increasingly giving up hobbies is directly associated -findings similar in all demographic groups (Fig 5).We also examined the amount of time spent on individual hobbies.Surprisingly, we found no difference in the association of hobby activity frequency with burnout and disengagement.The reasons for giving up hobbies (Fig. 6) were what one would expect-lack of time and loss of interest were the most frequent reasons cited.The time component would be consistent with busy professionals, and loss of interest may be due to time constraints or higher priority demands.
In addition to the previously described association of hobbies and leisure time with less burnout and less disengagement, when we assigned each of the 117 hobbies into categories of "social," "isolated," and "indeterminate," we found that the average burnout with social hobbies was lower than either isolated or indeterminate hobbies, regardless of frequency.Music listening and home media entertainment (e.g., TV, Video games, internet, etc.) registered highest on the burnout scale of associated hobbies, but perhaps these findings reflect more of a compensatory mechanism secondary to levels of stress that initially led to the pursuits.Recent articles have addressed the potentially isolating nature of these passive digital-based activities [64].Socialization in hobbies was not associated with disengagement.
Socialization is a remarkably powerful factor in mental and physical health, with loneliness and isolation linked to heart disease, cancer, depression, diabetes and suicide [53,55,62,[64][65][66].In the supplement we have provided considerable background commentary and data regarding the importance of the socialization component of leisure interests in the mitigation of burnout [22, 33, 56, 58-60, 62, 67-70].We discuss the history of burnout and fundamental concepts [71] and mechanisms of caring for the provider as well as the complex relationships between depression and burnout [1,24,[72][73][74][75].We caution administrators who might excuse systemic causes of burnout as due to lack of focus on the kinds of self-care that hobbies and leisure interests can enhance.Finally, we point out some potential downsides of leisure interests, when excess pursuit of "balance" in life and intense involvement in hobbies can itself become a contributing factor to burnout [76].
Regarding physician burnout, Drs.Fralick and Flegel ask, "Who will protect us from ourselves [77]?"With the recommendation from Fancourt et al. for "social prescribing [62]," it seems time to seriously reflect on the benefits of social involvement and constructive hobbies.

LIMITATIONS
The usual limitations imposed by cohort studies apply to this work, so the issue of causality vs. association is something we cannot avoid.Likewise, the determination of clinical significance in light of modest statistical significance is a frequent source of ambiguity.While on an individual basis our data may not necessarily seem clinically meaningful, considering that burnout and disengagement are epidemic in medicine, conclusions applied to this large population should be considered seriously, but may be restricted in application.Although some of the lay population studies require extrapolation to physicians, we think there is no reason to regard physicians as significantly different in psychosocial needs and risks.An obvious concern in attempting to extract the effect of socialization and burnout is the determination/definition of social vs. isolated activities and any crossover.Another is the striking disparity of burnout in different geographic and ethnic groups.A relatively low response rate to the survey raises concern for generalizability, as do potential biases in online surveys, but we think that the relatively large absolute number of respondents confers support to our conclusions.Finally, limiting the itemizing of hobbies to a maximum of five, done to avoid the risk of a wearying survey, could potentially dilute the strength of conclusions.

CONCLUSION
Observationally we have demonstrated that hobbies are proportionately associated with physician satisfaction, desire to stay in medicine, and the propensity to re-choose medicine as a career.In fact, the more hobbies the better, as our study shows an inverse relationship between number of activities and burnout/disengagement, and a direct relationship with number of hobbies given up.
In this study we further compartmentalized hobbies and leisure activities, so as to independently examine the association of those with social emphasis.We found that whether these activities are practiced at low or high frequency, social activity is associated with less burnout than more isolated pursuits.We did not find a significant additive socialization effect on disengagement, except for quality time spent with family and friends, which is inversely related to both burnout and disengagement.
Development of new hobbies is expected to be more difficult with aging, so we advocate exploring any potentially intriguing interests at the very first opportunity, as challenging as that might be.We believe that a prospective study of the effect of leisure activities among healthcare professionals is worthy of consideration and may show similar improvements in burnout As to our comments about the importance of hobbies and interests, we add this recommendation: Share them with family and friends.75.Alfonso F 3rd.The long lost hobbies people around the world are revisiting during the coronavirus pandemic.CNN 2020 Apr 5 [cited 2020 Apr 5] Dr. Vivek Murthy reports that loneliness and social isolation are "associated with a reduction in lifespan similar to that caused by smoking 15 cigarettes a day and even greater than that associated with obesity [34]."It should not be surprising, then, that mortality is 29% greater for those who feel socially isolated [64].Much is known about the neurobiology of burnout as well, confirming that the loss of connectedness is directly connected to burnout [66].It is unlikely coincidental that the systemic factors previously discussed -e.g., the universally present EHR, increasing work hours, productivity demands, colleague and patient meetings in digital format -are related to ever more isolation.Dafne et al. reported that moderate-intensity dancing reduced cardiovascular mortality more than light-intensity dancing or walking [53] but could not separate the psychosocial aspects.Schnohr et al. have found that various sports are associated with strikingly different associations with longevity, with the highest lifespans in sports with more social interactions [55].
These results, of course, require active participation and involvement.In the NEJM Catalyst event Leadership: Translating Challenge to Success, Tait Shanafelt talks about building firewalls to protect the needs in one's personal life, including hobbies."You have to have those spaces in the personal life where, no matter what, work will not intrude on it [22]."This was, in fact, addressed in the earlier noted study from Germany, in which 200 physicians were interviewed about hobbies as resilience strategies.In what could be a reply to the many physicians who report having no time for hobbies: "Respondents did not simply pursue hobbies when they had time to do so.Rather, they made sure to find the time they needed to pursue the hobbies that were important to them [33]." Kanamori et al. prospectively studied 12,951 individuals ≥age 65 for 4 years, whose social participation was sorted into 8 categories [56].They confirmed past studies that demonstrated the risk of functional disability declined significantly with social participation, and in their study, particularly in hobby, community, and sports groups, the risk declined proportionally to the number of groups in which the subjects took part.
From the Japanese literature we have also taken the liberty to utilize noteworthy measures of happiness among the lay public as surrogates for burnout in physicians.In a prospective study of "active aging," Hirosaki et al. examined in the elderly population the effects of hobbies on ADLs, depression, and quality of life [58].They found that the 43% of the elderly Japanese population without hobbies had significantly higher scores in all categories of depression and lower scores in every category of ADLs, as well as frequency of laughter.Tomioka et al. prospectively confirmed prior work determining that the absence of hobbies and purpose in life was strikingly associated with increased mortality (hazard ratio 2.08) and declines in ADLs (OR 1.89) [59].Having no hobbies may be an even more potent risk factor than the absence of purpose in life.
Prospective studies have shown that regular participation in cultural activities in the non-medical world may be protective against worsening depression and could reduce morbidity and mortality [60].Theorell and Nyberg also write that depression in its own right can cause decreased cultural activity, and speculated that if this leads to lower levels of hobby activity, it could signal a self-perpetuating and interwoven problem.Indeed, Løkken has demonstrated that music, singing and theater participation increases self-rated health in women and decreases all-cause mortality for men [67].It seems reasonable to surmise that physicians would achieve similar benefits from nonwork-related activities.
Depression as a surrogate indicator of burnout has long evoked controversy.There are many difficulties in even diagnosing depression and burnout among physicians, and the many barriers in treatment of physician mental health include the stigmata particularly dreaded by this group [68].Overlap with depression has been debated since the construct of burnout was introduced in the 70s by Freudenberger [61,69], and although some authorities feel that depression and burnout are the same, it now seems clear that overlap is indeed the case [70] .There is a more practical aspect to considering the overlap between depression and burnout as well.Undoubtedly, physicians will likely be more willing to request help to handle an institutional problem if they are not to be stigmatized with the label of mental illness.[70] Granting validity of a crossover between depression and burnout as well as basic psychosocial similarities between the population at large and physicians in particular, Fancourt et al. provide us prospective data for our recommendation.They have recently published a fascinating and uniquely dynamic study of hobbies and depression of 8,780 adults aged 52-99 [62].Their work details the effects of changes of hobby pursuits over time, by having created an intervention simulation of participants who did not have a hobby.For those free of baseline depression, taking up a hobby was associated with 32% lower odds (OR 0.68, 95% CI 0.56-0.83) of developing it.For those with baseline depression, taking up a hobby was associated with 272% higher odds of recovering (OR 2.72, 95% CI 2.09-3.53),although their data did not include use of antidepressants and, unfortunately, was limited to nearly 98% whites [63].
Are there potential downsides to taking up hobbies and interests?Tim Wu's 2018 intriguing piece in The New York Times discusses maiming by leisure activities because of ever increasing pressures to achieve.[72] "Our 'hobbies,' if that's even the word for them anymore, have become too serious, too demanding, too much an occasion to become anxious about whether you are really the person you claim to be."To which we might add, who but physicians could be more susceptible to misconstruing the very motive for hobbies?
Likewise, Brad Stulberg addresses the sometimes seemingly endless pursuits to be "balanced [73]," reminding us that "devoting equal proportions of time and energy to other areas of [our] life [could] have detracted from the formative experiences." We are cautioned by The Copenhagen City Heart Study [76] that too much of a good thing may be detrimental.A prospective investigation followed 1098 healthy joggers and 3950 healthy non-joggers and found the lowest mortality among light joggers and a higher mortality of moderate and strenuous joggers.In our study, at least, we did not find more time spent on hobbies to be associated with higher burnout or disengagement rates.
In a creative look at caring for the provider, Silverman describes a passion for storytelling: "The greater risk, however, is for the health care professional to appear superhuman by pretending to not feel grief, suffer from moral distress, laugh at work, or need rest.These emotions should be explored, not con-cealed… In this environment of physician burnout, storytelling may actually help to humanize the physician [74]." Music has long been part of the caring process.From an anonymous article in 1908: "Billroth was a superb pianist; Strümpell is a clever violinist; many among our colleagues are excellent performers on musical instruments, and are all the better surgeons and physicians, for their genial and humanizing accomplishments, all the better qualified to comprehend the sufferings they must alleviate.[1]" Finally, it should be stated that administrators need to be especially mindful to avoid placing blame on providers for "inadequate" coping skills that invite burnout.Dean et al. [71] so wisely point out that this pseudo-culpability may be likened to "gaslighting" -i.e., system-induced self-doubt, rather than the system-inflicted "moral injury" that is truly responsible for these crises.In other words, hobbies and social interactions may help, but are not the solutions.

Figure 4 :
Figure 4: Association of Exhaustion and Active Hobby Categories *The bubble colors correspond to hobby categories, while the height represents burnout level (mean exhaustion score), and the size of the bubbles represents reported hobby frequency.

Figure 5 :
Figure 5: Association of Number of Active and Former Hobbies With Exhaustion and Disengagement * 5a and 5b: the burnout level and disengagement levels were negatively associated with the number of active hobbies.* 5c and 5d: the burnout level and disengagement levels were positively associated with the number of former (given-up) hobbies.

Figure 6 :
Figure 6: Reasons for Giving Up Hobbies and Leisure Activities

Table 1 :
Demographics * P value represents variation among all subgroups of each demographic feature

Table 2 :
Exhaustion and Disengagement Score Among Different Specialties