Assessment of stress and associated factors in employees of a public higher education institution

Introduction Occupational stress is considered as the negative imbalance between work demands and resources, and it can generate consequences to an individual’s health and interfere with his or her quality of life. Objectives To investigate stress and its associated factors among employees of a higher education institution through a cross-sectional study (at the baseline of a longitudinal study) including 176 individuals aged 18 years or older. Sociodemographic characteristics related to physical surroundings, lifestyle, working conditions, and health and illness were tested as explanatory variables. Methods Stress was estimated using prevalence rate, prevalence ratio (PR), and a 95% confidence interval. For a multivariate analysis, we employed a Poisson regression model with robust variance, where a p-value ≤ 0.05 was considered significant. Results The prevalence of stress was 22.7% (16.48-28.98). This study noticed that depressive individuals, professors, and those who self-assessed their health as poor or very poor had a positive association with stress within the studied population. Conclusions Studies of this type are important for identifying characteristics in this population that could contribute to the planning of public policies in order to improve the quality of life of employees of public institutions.


INTRODUCTION
In the last decades, the technological innovations and fast transformations that took place in work environments contributed to significant changes in the profile of work, which started requiring important abilities of physical, mental, and social adaptation; these, many times, end up frequently exposing the working population to situations involving stress, anxiety, distress, and emotional destabilization. 1,2 Due to this paradoxical situation between positive and negative aspects, the stress related to the work environment has been thoroughly studied in the sense of identifying its participation in the etiology of worker illness. 3 The word "stress" derives from Latin, being popularly employed in the 17th century with the meaning of "fatigue" and "tiredness." According to Selye (1936), stress is a syndrome characterized by an ensemble of reactions developed by the body when subjected to a conflicting situation. 4 Therefore, everything that causes a disruption of internal homeostasis, that is, requires some adaptation from the individual, can be called a stressor. 5 In this sense, stress can be caused not only by psychological or psychic disturbances, but also by a set of daily events (external events) that can lead to psychological distress. 6 Stress is a condition that can affect people of all ages. However, some population subgroups seem to be more vulnerable, such as women, older adults, people at higher vulnerability, individuals with a low education level, and workers. 7 In this context, stress is considered a social and public health problem in the 21st century, being regarded as negative because it leads to damages and social issues such as mood changes, reduced productivity, cognitive and psychomotor alterations, and loss of initiative, among others. 8,9 The intensity of the experienced stress is related to stressor severity and to the individual's psychosocial coping resources. The World Health Organization (WHO) states that the stress load at the workplace and mental disorders within these places indicate a need to promote healthy work environments where physical health, safety, and wellbeing are achieved. 10 Global health authorities recognize stress as the world's largest epidemic of this century. In 2019, Gallup's Global Emotions Report disclosed a study performed in 142 countries where one in every three interviewees experienced a lot of worry or stress, and at least one in every five people reported feeling sad or angry. 3 The consequences of stress in the work domain are extensive, including depression, lack of initiative, lack of commitment, lack of motivation, irritability, impatience, difficulties with interpersonal relationships, loss of productivity, absenteeism and frequent tardiness, excess visits to the medical department, and medication dependence. 5 Occupational stress is normally considered as the negative imbalance between work demands and resources experienced by workers and it can be generated by precarious working conditions, compromising an individual's mental and physical health. 11 A study by Cacciari et al indicates that chronic stress is an independent risk factor for the development of cardiovascular diseases, especially stroke. 12 Within the context of occupational health problems, stress is seen as a worrisome aspect due to its serious consequences that can interfere with people's quality of life. Therefore, this study proposes to assess stress levels among workers of a public higher education institution, as well as associated factors.

STUDY DESIGN
This is a cross-sectional study (at the baseline of a longitudinal study) that aimed to investigate stress and its associated factors in a public higher education institution located in the southeast region of the state of Bahia (BA); the study was performed between January and June 2016.

STUDY POPULATION
Out of a total of 204 employees (including professors, technicians, and outsourced employees), 191 individuals met the eligibility criteria. The study was performed at a public higher institution in Vitória da Conquista, BA. The following eligibility criteria were adopted: individuals aged 18 years or older, who were not on leave at the moment, and who consented to participate in a longitudinal followup by signing an informed consent form.

PILOT STUDY
For verifying recruitment dynamics, testing the data collection instruments, and confirming the viability of our investigation, we performed a pilot study with employees of another university, who were comparable to the sample of this study and whose number comprised 8% of the final study population.

DATA COLLECTION
Data were collected through individual inperson interviews, using a portable computer (HP Pocket Rx5710) for recording and storing data. The instrument adopted for the interviews was the National Health Survey (Pesquisa Nacional de Saúde [PNS]) semi-structured questionnaire, adapted to the study population. All interviewers were undergraduate students at the institution; they were organized in teams that were duly trained for data collection.

SEMI-STRUCTURED INTERVIEWS
Individual interviews were performed using portable computers. The collected data were transferred to a dedicated database.
The instrument adopted for the interviews was based on instruments from national and international studies, such as: the questionnaire developed by the Belo Horizonte Observatory for Urban Health (

ARTERIAL PRESSURE (AP) MEASUREMENT
AP was measured through the oscillometric method, using an internationally validated digital sphygmomanometer (Omron HEM-742). Three AP measurements were performed with 1-minute intervals. Data collection was performed after the in-person interview in order to ensure that the interviewee did not eat or smoke and remained seated, at rest, for at least 10 minutes before measurements.

NUTRITIONAL STATUS ASSESSMENT
Nutritional status was assessed according to the body mass index (BMI), calculated by dividing the individual's weight (in kg) by height squared (in m). Weight was measured by a previously calibrated portable digital scale with capacity for 150 kg and a precision of 50 g. The individual should stand on the center of the scale, barefoot and wearing light clothes. Stature was verified with a mobile stadiometer that reached measurements of up to 213 cm, with a precision of 0.35 cm. The individual should stand with the back touching the stadiometer, feet together, with an erect posture and looking forward, and the reading was done at the closest millimeter when the head-plate touched the individual's head. For the BMI classification, we adopted the criteria established by the WHO and adapted by the National Institutes of Health (NIH): underweight (BMI < 18.50 kg/m 2 ); normal weight (BMI ≥ 18.50 kg/m 2 and ≤ 24.99 kg/m 2 ); overweight (BMI ≥ 25.0 kg/m 2 and ≤ 29.99 kg/m 2 ); and obesity (BMI ≥ 30.00 kg/m 2 ). 15

FAT PERCENTAGE ASSESSMENT
The assessment of body fat percentage was performed through an electric bioimpedance analysis with a 0.5% grading scale for body fat; this analysis assessed the lower segment of the body. The participant laid on a stretcher in the dorsal decubitus position, and electrodes were placed on predefined sites: the first electrode was placed directly below the joint of the middle finger; the second one was placed on the wrist joint; the third was placed between the second and third toes; and the fourth was placed between the ankle and the lower leg bone. The fat percentage was classified according to Lohman 16 as: risk of diseases associated with malnutrition: ≤ 5% in men and ≤ 8% in women; below average: 6 to 14% in men and 9 to 22% in women; on average: 15% in men and 23% in women; above average: 16 a 24% in men and 24 to 31% in women; and risk of diseases associated with obesity: ≥ 25% in men and ≥ 32% in women.

STUDIED VARIABLES
The dependent variable in this study was stress, constructed from an analysis of the literature on organizational stressors of psychosocial nature that cause adverse psychological reactions to workers. Based on principles by Paschoal & Tamayo, 17 the outcome variable was created by combining two variables in this study: nervousness and tiredness. For constructing this proxy variable for stress, we used two questions: "Do you feel nervous, tense, or worried?" and "Do you get tired easily?" The answer options were "yes" or "no" for both questions; this way, individuals who answered "yes" to both questions were classified as stressed.
A conceptual analysis model was built based on the literature for determinant factors of stress ( Figure 1). The sociodemographic block was constructed using variables such as sex (male and female), age (22 to 39 years; 40 years or older), marital status (married/civil union; not married/single, divorced, or widowed), family income, education level (incomplete higher education: from the first year of primary education to incomplete higher education; higher education: from college education to post-doctoral education), and occupation (dichotomized between professor or technician/outsourced worker).
Regarding the "physical surroundings" block, we assessed safety and violence perceived in the neighborhood, both with the following answer options: agree (completely or partially) and disagree (neither agree nor disagree, partially disagree, or completely disagree). The questions were extracted from the ELSA-Brasil neighborhood scale, which is a Likert-type scale.
As for the "lifestyle" block, we assessed "body satisfaction" (satisfied; dissatisfied, underweight; dissatisfied, overweight); for "physical activity (PA)", we considered PA in its four domains (leisure, work, household, and commuting), and multiplied the weekly frequency of PA (in days) by the mean duration (in minutes) of walking performance or other moderate and vigorous PA. When considering vigorous PA, the duration was multiplied by two. Individuals who practiced 150 minutes or more of PA per week in at least one of the assessed domains were considered physically active. 10  Regarding lifestyle habits, alcohol consumption was calculated using the daily doses of alcohol reported by participants and risky alcohol use was considered when an individual consumed more than 30 g/day of alcohol (for men) or 15 g/day (for women). For smoking habits, two categories were considered according to the individuals' answers: current smokes or does not currently smoke, being categorized as smokes/does not smoke.
For the "working conditions" block, we defined the following variables: job tenure (up to 2 years; more than 2 and less than 5 years; more than 5 and less than 8 years; or 8 years or more) and workweek (20 to 30 hours a week; 31 to 40 hours a week; or more than 40 hours a week).
The health/illness block consisted in fat percentage (below average/average; above average; or obesity), BMI, categorized into underweight/ normal weight or overweight/obesity, and selfreported health status (very good/good, regular, or poor/very poor). Depression and diabetes were selfreported. As to hypertension, we used the mean value between the last two measurements, and individuals were considered hypertense when having a systolic arterial pressure ≥ 140 mmHg and/or a diastolic arterial pressure ≥ 90mmHg and/or using antihypertensive drugs.

STATISTICAL ANALYSIS
We initially estimated the prevalence of stress among workers with a 95% confidence interval (95%CI). Independent variables were presented as absolute and relative frequencies. For verifying which factors were associated with stress, we performed a univariate analysis estimating PRs and calculating the respective CIs. A p-value ≤ 0.05 was considered significant. Data analysis was performed using Stata software, version 12.0.

RESULTS
Out of 191 eligible individuals, 176 participated in this study. The population of this study comprised individuals employed by a higher education institution in the southeast region of BA. Among the observed losses, 7.85% happened due to refusal to participate in the longitudinal study.
Most (51.1%) participants were male, aged between 22 and 39 years (73.3%), married (67.6%), with a monthly family income (57.1%) of more than five times the minimum wage, and 69.3% had a higher education degree. As for the physical surroundings block, 58.5% of the participants agreed they had safety in their neighborhood; however, 49.4% stated they had suffered some type of violence in the neighborhood. (Table 1).
Regarding the lifestyle block, most of the population (75%) in this study revealed they were not satisfied with their own bodies. Among all participants, 79.4% were classified as physically active in at least one of the assessed domains of PA, 29.7% presented risky alcohol use, and 4.6% of them were smokers. Considering the working conditions block, 26.7% of the population had a job tenure of 2 to 5 years. Most of them (42.1%) had a workweek of 31 to 40 hours and did not have other sources of income (82.4%) ( Table 1).
As for the health/illness block, 23.9% were classified as having arterial hypertension; self-reported diabetes and depression were observed in 2.3% and 15.9% of individuals, respectively. Regarding their body fat percentage, 42.3% were classified as above average and 52.2%, as overweight/obese. Among all individuals, 65.9% self-assessed their health as "very good/good" (Table 1).
Higher prevalence rates were observed for stress among women (34.9%) in relation to men, with statistical significance (p = 0.001); this was also observed in individuals aged between 22 and 39 years (24.8%) and in unmarried individuals (24.6%). However, no statistical significance was observed for these variables. Moreover, a higher prevalence of stress was observed in employees presenting a family income of more than five times the minimum wage (28%). We observed that the lower the family income was, the lower the prevalence of stress in this study; however, this aspect did not have statistical significance. Individuals with incomplete higher education and belonging to the technicians/outsourced employees group presented lower prevalence rates for stress: 13% and 15.2%, respectively -only the latter presented statistical significance (Table 2). We observed a higher prevalence (44.4%) of stress among individuals who were dissatisfied with their bodies due to low weight, with statistical significance (p = 0.025). Physical inactivity was also associated with this outcome, that is, individuals who practiced less than 150 minutes of PA a week presented a Continued on next page  higher prevalence of stress (36.1%), with statistical significance (p = 0.028). Individuals who displayed risky alcohol use and those who reported being smokers presented lower prevalence rates for stress (13.2% and 12.5%, respectively) when compared to the reference categories, but this finding did not have statistical significance (Table 2). When it comes to working conditions, the prevalence rates observed for stress were higher among individuals who worked more than 40 hours a week (26.9%) and had a job tenure of 8 years or more (42.9%); the latter was statistically associated with the outcome (p = 0.004) ( Table 2).
Higher prevalence rates of stress were observed among workers with depression (53.6%), diabetes (75%), and in individuals who assessed their own health as poor or very poor (75%), being positively and significantly associated with the outcome (p = 0.000) in individuals with a fat percentage below average/on average (28.6%), with a nutritional status of underweight/normal weight (29.3%), and who did not have hypertension (25.4%); these were not statistically associated with the outcome ( Table 2).

DISCUSSION
In this work, the variables who remained positively associated with stress were professorship as occupation, a diagnosis of depression, and a health self-assessment as poor/very poor.
The prevalence of stress (22.7%) in this work was lower than that found in a study performed among professors of a public institution in Teresina, in the state of Piauí, where 42.9% of the participants presented mild stress. 8 The observed difference may have been due to the fact that this study had, as its target, not only professors, but also technicians and outsourced employees. In a study performed with 1,000 Brazilian workers in the cities of Porto Alegre and São Paulo, published by the International Stress Management Association (Brazil), 70% of the participants suffered with stress, and its largest motivation was related to the professional domain. This study encompassed various professions. Most of the interviewees stated that stress was associated with work, such as long hours and excess tasks. 19 In this study, occupation was a factor that remained associated with stress. Individuals in the group of professors were considered more stressed than the group of technicians/outsourced employees. Teaching is one of the professions that are more prone to developing stress and other psychological disorders. This is due to the fact that this occupation is related with a high workload, activity organization, planning and overseeing classes, and a continuous search for improving one's curriculum, in addition to extracurricular activity management, which generates a psychological environment. 8,9 In addition, professors are responsible for helping students with their healthy progress, considering individual differences, and respecting their learning abilities. Therefore, matters related to the work characteristics also contribute to a high stress level. Factors such as interactions between the environment and organizational conditions, the content of work, and the tasks, efforts, and individual characteristics of workers can significantly interfere with occupational health, possibly being beneficial or damaging to health. When considering professors, one should consider fast decision-making, balancing demands, and relationships between colleagues, higher-ranking employees, and students. 20,21 Although not statistically significant, a higher prevalence of stress was seen in the age group of individuals aged 22 to 39 years. According to the literature, professional experience and theoretical knowledge for facing challenges at work can provide greater confidence for decision-making and thus be allies against stress in older employees when compared to the younger ones. 11 In this study, workers classified as physically active presented a lower prevalence of stress in the bivariate analysis. Although this did not remain associated in the multivariate analysis, PA has shown considerable efficacy in results related to work, such as absence of disease, higher satisfaction, lower occupational stress, increased productivity, among others. 22 Moreover, recent studies state that the lack of PA can contribute to important physiological and psychological problems, such as increased obesity, cancer, metabolic syndrome, cardiovascular diseases, worse body satisfaction, and lower self-esteem. Instead of PA being used as a tool for coping with stress, individuals sometimes resort to unhealthy behaviors as a way of facing and expressing their emotions. 23 Depression was a comorbidity that remained associated with the studied event. In agreement with our findings, Donato et al. 24 reported that individuals diagnosed with depression presented high prevalence rates of stress. According to Meyer et al., 25 if a stressed individual does not reveal what he or she feels, this absence of demonstration of feelings and emotions slowly generates internal conflicts and, consequently, depression. Moreover, it can have a negative impact on quality of life, work satisfaction, and relationships with colleagues, causing more stress and worsening of the individual's depressive state. 26 Studies propose that mental health interventions such as cognitive-behavioral therapy, when performed at the workplace, could decrease the levels of depression symptoms among workers. 27,28 Shen et al. 20 state that professors are not only responsible for transmitting intellectual and moral knowledge to students, but they need to keep up with modern technological evolution, which requires the understanding of new learnings. Therefore, some professionals may feel unable to cope with the tension, resulting in exhaustion, which can contribute to the development of depression. 20 The PR for stress was 3.79 times higher among those who self-assessed their health as poor/very poor in comparison with those who self-assessed their health as good/very good. Corroborating these findings, Reis et al. 29 consider that factors affecting the psychological domain, such as stress, can influence the health perception of individuals.
The self-assessment of health is a measure that encompasses various aspects such as demographic, socioeconomic, cultural, and lifestyle factors, in addition to issues related to the work environment and health status. As to the work characteristics, it is known that the exposure to divergent, exhaustive, and stressful conditions is a risk factor for the psychological health of individuals. 30,31 Considering our study design, this study has limitations referring to causal inferences. The population size may have prevented some associations from being observed. Therefore, the results of this study will serve as subsidy for the elaboration of policy interventions aimed at the occupational environment, considering that individuals are prone to stress in this scenario.

Author contributions
ASF was responsible for conceptualization, formal analysis, investigation, and writing -review & editing. NTF was responsible for data curation, formal analysis, and writing -review & editing. RMAS was responsible for conceptualization and writing -review & editing. LRCSS was responsible for conceptualization and writing -review & editing. DSR was responsible for conceptualization and writing -review & editing. VMB was responsible for conceptualization, data curation, formal analysis, supervision, and writing -review & editing. All authors have read and approved the final version submitted and take public responsibility for all aspects of the work.