Association between burnout syndrome and obesity: A cross-sectional population-based study

1 Introduction

Burnout Syndrome (BS) significantly affects health professionals, and its occurrence varies from 4.8% to 39.3% [1]. Despite the quality of care provided, the syndrome has negative impacts on the health and quality of life. BS is characterized by chronic physical and mental fatigue produced from prolonged exposure to a stressful situation, where the professional requirements go beyond the personal capacity [2].

Due to the growing occurrence of BS and the impacts it exerts, a significant number of studies have been developed in recent years to explain the phenomenon. As a syndrome, it has specific dimensions and a series of psychological reactions to chronic work stress. This situation is common among professions with intense interpersonal contact [3, 4], namely: Emotional exhaustion (EE) is the experience of being emotionally exhausted by the demands of work; Depersonalization (DP) is a distancing process from people to whom care is provided; and Reduction of Professional Fulfillment (RPF), which is assessed by feelings of self-efficacy and accomplishment at work [5].

The World Health Organization (WHO) included BS in the International Classification of Diseases, 11th Revision (ICD-11), among the factors that influence health status as a work-related problem. In this ICD issue, the syndrome is described as an occupational phenomenon specifically related to professional context experiences [6], which evidences there are exposure factors at work that lead to sickening.

Burnout is evident among health professionals due to the characteristics related to their work. Burnout specificities impose uncertainty in relation to the intended results. These characteristics are associated with the precarious conditions of work and employment nowadays and they determine a high degree of stress in the health care provided. This subject has been very relevant for scientific investigations and indicates significant prevalence of burnout among physicians [7, 8], health managers [9], medical students [10], physiotherapists [11], and nurses [12–16]. This list of workers includes Primary Health Care (PHC) professionals, which is the preferred gateway of patients to the Brazilian health system.

Sometimes PHC work requires living with conflicting and stressful situations, inherent to work issues, in addition to those involving the social and family life of healthcare users due to emotional bonds established. Often the physical working environment of the healthcare units is inadequate, usually units have scarce human resources, insufficient equipment, and supplies. In addition, workers have low salaries and are exposed to different occupational risk factors and physical violence. These factors can be the source of a wide spectrum of diseases, including BS, which may lead to weight gain and obesity [17, 18]. A study carried out by Lu et al. reaffirms the effect of such labor pressure and the physical and mental health of busy workers [19].

The relationship between BS and obesity is not clearly defined in the literature, however, the biological plausibility for such findings is based on the possible relationship between BS and alterations in the Hypothalamic-Puitary-Adrenal (HPA) axis. Physiological responses to chronic stress through hypercortisolism may be related to the general severity of burnout symptoms as they involve changes in basal levels of cortisol [19, 20]. In order to increase the amount of energy available to deal with situations of stress and insulin resistance [21–23], therefore, people with high levels of stress and BS are more likely to gain weight and obesity over time [24, 25]. In addition, less healthy lifestyle habits of individuals with BS may increase the outcome in question [26, 27].

In view of the above, this study aims to estimate the association between burnout and obesity among primary health care nursing professionals in the state of Bahia, Brazil.

2 Material and methods

2.2 Sample and eligibility criteria

The study includes a representative population sample of PHC nurses in Bahia, Brazil, who were chosen by cluster sampling.

The selection took place by means of a drawing on Microsoft Office Excel 2010, which sampled 10% of municipalities (n = 43) at each mesoregion (stratum) of Bahia (conglomerates). All the PHC nursing professionals (nursing technicians and nurses) of the sampled conglomerates were included, totaling 1,195 individuals [15]. It is noteworthy that nursing in Brazil includes not only workers with higher education training (nurse), but also secondary education (nursing technician).

A pilot study was developed due to the lack of studies that could be taken as a reference for the sample calculation, thus, the frequency of MS in the unexposed group were 20%, while the exposed ones were 33.3%, α error of 0.05, 90% power (1 - β error), 1:1 ratio, reaching a sample n of 464. The design effect of 2.0 (cluster sampling) was also considered, the sample was doubled to 928, adding 20% for possible losses and refusals. In total, 1,114 PHC nursing professionals were obtained. The sample calculation was performed by using Epi Info software, version 7.0 (Centers for Disease Control and Prevention, Atlanta, United States) [15, 20].

All PHC nursing professionals were considered eligible and invited to participate in the study. The following nursing professionals were excluded from the study: those on leave, with less than 6 months of experience in PHC, who works at exclusively administrative activity, within puerperal pregnancy cycle and menstrual period, with diagnosis of depression, anxiety, and burnout before occupying the current position, liver cirrhosis, as well as alcohol and drug addiction. 48 professionals were excluded and 22 refused to participate in the research, and a final sample of nursing professionals totaled 1,125 participants, with a response rate of 94.1% [15, 20]. It is noteworthy that only nurses who made up the final sample of 455 analyzed were considered for this study.

OPEN IN VIEWER

Image 1

Descriptive flowchart of study selection.

3 Results

Within the 455 participants, 391 (85.9%) were female and 320 (70.3%) were aged up to 35 years, mean of 33.3 years (DP±6.8). Regarding race/ethnicity, 321 (71.3%) declared themselves black. Regarding marital status, 260 (57.1%) did not have a partner. Regarding lifestyle and health, 152 (33.4%) participants declared not to practice regular physical exercise; 46 (10.1%) smoking; 18 (4.0%) consuming alcoholic beverages routinely; 222 (48.8%) reported having an unhealthy diet, and 237 (52.1%) did not have a satisfactory sleep pattern (Table 1).

Regarding working conditions, 53 (11.6%) were dissatisfied with their occupation; 165 (36.3%) worked in PHC≥5 years; 144 (31.6%) worked outside PHC; 90 (19.8%) had night shifts; 167 (36.7%) and 136 (29.9%) reported that PHC working conditions and employment relationship were precarious, respectively; 145 (31.9%) have already suffered from any type of aggression at work. In the MBI analysis, 138 (30.5%) scored a high EE; 190 (41.9%) scored a high DP, and 285 (62.8%) had a high RPF. Considering the criterion of Ramirez et al. (1986), burnout was indicated by the prevalence of 80 (17.7%), which points out considerable frequencies in the three dimensions of the MBI (Table 1).

The prevalence of obesity was 58 (12.7%). Table 1 also presents the bivariate analysis of sociodemographic, labor, lifestyle, and health characteristics associated with the outcome (obesity). There was a statistically significant association between the variables: gender (PR = 2.13; 95% CI = 1.26–3.60; p-value < 0.01); age (PR = 1.92; 95% CI = 1.19–3.10; p-value < 0.01); smoking (PR = 2.08; 95% CI = 1.16–3.72; p-value 0.01); sleep pattern (PR = 1.74; 95% CI = 1.05–2.90; p-value 0.02); work outside PHC (PR = 1.75; 95% CI = 1.08–2.83; p-value 0.02); PHC time (PR = 1.88; 95% CI = 1.16–3.03; p-value < 0.01). Burnout was associated with obesity (BMI≥30) (PR = 2.08; 95% CI = 1.26–3.44; p-value < 0.01) as well as the three dimensions: EE (PR = 1.98; 95% CI = 1.23–3.19; p-value < 0.01); DP (PR = 2.11; 95% CI = 1.29–3.45; p-value < 0.01); RPA (RP = 1.86; 95% CI = 1.05–3.29; p-value 0.02). According to the pre-established methodological criteria, no modifying or confounding effect variables were identified (Table 2).

The variables gender, age, regular physical exercise, healthy eating habits, occupational satisfaction, another job, night shift, PHC time and working conditions were maintained in the multivariate model for adjustment, given their relevance and according to knowledge of their influence on both the exposure factor and the outcome. Thus, burnout is still associated with obesity, adjusted PR of 1.85 and statistical significance (95% CI 1.11–3.06).

The Hosmer and Lemeshow’s tests showed p > 0.05, which leads to the hypothesis that the model was duly adapted to the data. The ROC curve showed an area of 0.72 and indicated that the final model has an adequate discriminative power and it is adjusted to the data (Table 3).

For the first time in the literature, the direct association between BS and obesity in professional nurses working in PHC is evidenced. It is already known that BS is linked to comorbidities [36], however, specific studies of the effects of BS on obesity in this specific population were not identified.

This association is even more relevant, given the high prevalence of BS observed in this population, as well as its dimensions. Considering the criterion of Ramirez et al. [31], the prevalence found was 17.7%. Prevalences higher than those found in other studies: in higher education professionals in the PHC network in a Brazilian city, prevalences of the syndrome were observed between 6.7% and 10.8% [37], and in PHC nursing professionals the BS prevalence was 16.7% [38]. A prevalence similar to our study (18.3%) was observed in a multicenter study carried out in Bahia [15].

Regarding the specific dimensions of BS assessed by the MBI, there is a high EE prevalence at 30.5% among nurses, high DP prevalence at 41.9% and high RPF prevalence at 62.8%. Similar and high prevalences of DP (48.3%) and RPF (56.6%) were also observed in PHC nursing professionals [38].

Also noteworthy is the high prevalence of obesity, 12.7%, which was also associated with other demographic, lifestyle, and occupational factors, namely: gender (male), age (≥36 years), smoking, unsatisfactory sleep pattern, work outside the PHC, time≥to 5 years of PHC.

The hypothesis is that stress can lead to weight gain/obesity in virtue of behavioral components. Thus, professionals more exposed to work stress may have insufficient time to prepare healthy meals, which contributes to weight gain [39]. Other studies point to a positive association between greater workload and obesity [40, 41], a relationship intertwined with professional exhaustion (BS).

Age might be established as a factor since it declines physical functioning and metabolism, which leads to weight gain. Thus, older ages are more related to higher BMI [42]. This fact was confirmed by our findings when observing a higher probability of obesity among older nurses (36 years or more).

The positive association between BS and obesity was maintained even after adjustment (adjusted PR 1.85-CI 95% 1.11–3.06). A study prospectively tested the hypothesis if obesity predicts burnout or vice-versa and found lack of support to the hypothesis that burnout predicts obesity. However, it is noteworthy that the participants in such study were not health professionals and the specific characteristics used in the evaluation of burnout used were different from our study. In addition, the measurement of obesity showed differences from our analysis and goes beyond the BMI [43]. It is noteworthy that BMI was used in our assessment, and this measure is widely used in large-scale studies due to its easy applicability, low costs, and validation by the World Health Organization [44].

The relationship between burnout and obesity was evidenced in a population sample from Finland [39]. Since there is association between burnout, eating behavior and weight, the ones who experienced burnout may be more vulnerable to eating in an emotional and uncontrolled way and have an impaired ability to make changes in their eating behavior. It is advisable that burnout and eating behavior be evaluated to control obesity [27], as well as considering that obesity and overweight also require additional factors, such as genetics and personal eating habits [45, 46]. In the meantime, there is a greater vulnerability to emotionally and uncontrolled eating and burnout [27].

Having more than one job and longer stressful exposure are factors that lead to stress at an exacerbated work level. In this study, additional work shifts and longer working hours were associated with the statistically significant outcome. It confirms the association between a stressful job and a higher BMI [47], and mental health and BS [48].

In turn, a national study conducted with PHC nursing professionals to estimate the association between burnout and increased abdominal adiposity obtained a positive result (PR: 1.63; 95% CI, 1.29 to 2.06). In the study mentioned above, it is noteworthy that waist circumference measurement was used and not BMI [49].

The findings presented and discussed in this study come from robust statistical analysis, such as evaluated and well-adjusted models, with a strong theoretical basis and a representative sample of the population. This allows for good predictive ability and statistical inference. However, some limitations should be pointed out: the cross-sectional design, which makes it impossible to establish a cause-and-effect relationship; use of some self-reported variables; the use of BMI to assess obesity, which uses weight as a measure of risk, and not the percentage of body fat, even though this is a widely used measure.

Despite the limitations described above, our study showed association between burnout and obesity and the relevance to research on workers’ health, with emphasis on healthcare professionals, so that occupational health programs can be created to implement strategies for preventing and coping with burnout and its health consequences.

5 Conclusion

The study results point out to a statistically significant association between burnout and the following variables: gender, age, smoking, sleep pattern, work outside PHC, and PHC time. In addition, burnout was associated with obesity (BMI≥30), even after adjustment and having statistical significance. In fact, the literature suggests a path to the association identified here, through the hyperactivity of the HPA axis and the continuous production of cortisol resulting from chronic work stress. New studies on this subject are relevant and they require more robust methodologies so that the causality path is better clarified. In addition, occupational health programs become an alternative to control stress and therefore, its complications, such as burnout and other health aggravations.

Ethics statement

The present study was approved by the Research Ethics Committee involving human beings at the State University of Bahia (UNEB), Brazil, under protocol number 872.365/2014. The Declaration of Helsinki of the World Medical Association and Resolution 466/2012 of Brazil were fully respected.

Conflict of interest

Not applicable.