Abstract

The objective of this article is to evaluate the interrelationships of factors associated with depressive symptoms (DS) in teachers, considering dissatisfaction with the teaching job as a possible mediator. This was a cross-sectional study using data from 700 teachers from the public school system of a Brazilian municipality. The outcome of interest was DS, as assessed using the Beck Depression Inventory (BDI). Direct and indirect interrelationships between the outcome and dissatisfaction with work, age, income, lifestyle and adiposity were tested. These variables composed the operational model tested by structural equation modelling. Older age (β=0.12) and greater dissatisfaction with work (β=0.12) were directly associated with DS. A more favourable lifestyle (β=-0.60) and adiposity (β=-0.10) were associated with a lower occurrence of DS. The variables lifestyle (β=-0.06) and adiposity (β=-0.02) also had negative indirect effects on DS, mediated by job dissatisfaction. The structural equation model tested identified interrelationships that influenced DS. Dissatisfaction with teaching work was associated with DS and mediated the relationship of other factors with such symptoms.

Key words:
Professors; Depressive symptoms; Epidemiology; Occupational health

Introduction

Among teachers, depression is the most prevalent mental disorder and has been identified as the cause of more than half of absences from work among teachers of basic education¹1 Batista JBV, Carlotto MS, Oliveira MN, Zaccara AAL, Barros EO, Duarte MCS. Transtornos mentais em professores universitários: estudo em um serviço de perícia médica. J Res Fundam Care Online 2016; 8(2):4538-4548.. National studies have found prevalence rates of depressive symptoms (DSs) among public school teachers varying between 21.6%²2 Scandolara TB, Wietzikoski EC, Gerbasi ARV, Sato SW. Avaliação dos níveis de estresse e depressão em professores da rede pública do município de Francisco Beltrão - PR. Arq Cienc Saúde UNIPAR 2015; 19(1):31-38. and 50%³3 Ferreira-Costa RQ, Pedro-Silva N. Níveis de ansiedade e depressão entre professores do Ensino Infantil e Fundamental. Pro-Posições 2019; 30:e20160143., values considerably higher than those observed in other groups⁴4 Julio RS, Lourenção LG, Oliveira SM, Farias DHR, Gazetta CE. Prevalência de ansiedade e depressão em trabalhadores da Atenção Primária à Saúde. Cad Bras Ter Ocup 2022; 30:e2997.

5 Monteiro MAM, Orfanó IS, Castro RIV. Prevalência de transtornos mentais comuns em trabalhadores de restaurantes. Res Soc Dev 2021; 10(3):e214103133212.^-66 Lima EP, Assunção AÁB, Sandhi M. Prevalência de depressão em bombeiros. Cad Saude Publica 2015; 31(4):733-743.. The international literature on this subject also indicates a high prevalence of DSs among professors⁷7 Shen X, Yang YL, Wang Y, Liu L, Wang S, Wang L. The association between occupational stress and depressive symptoms and the mediating role of psychological capital among Chinese university teachers: a cross-sectional study. BMC Psychiatry 2014; 14:329.^,88 Soria-Saucedo R, LopezRidaura R, Lajous M, Wirtz VJ. The prevalence and correlates of severe depression in a cohort of Mexican teachers. J Affect Disord 2018; 234:109-116.. In basic education, especially in public education, there are significant challenges due to the high demands⁹9 Assunção AA. Educatel Brasil 2015/16: estimativas da frequência e distribuição dos principais condicionantes de saúde e de faltas ao trabalho na população de professores da educação básica no Brasil. Belo Horizonte: Ed do autor; 2016., such as precarious working and infrastructure conditions, a lack of interest/engagement of students and their families, pressure to meet deadlines and tasks, an excessive workload, and open positions due to low wages, among other factors¹⁰10 Mahan PL, Mahan MP, Park NJ, Shelton C, Brown KC, Weaver MT. Work environment stressors, social support, anxiety, and depression among secondary school teachers. AAOHN J 2010; 58(5):197-205..

Depression is a multifactorial mental disorder caused by complex interrelationships among genetic, biological, environmental and psychological factors and negatively affects the health, well-being and behaviour of individuals¹¹11 Oliveira AM. Relação da Satisfação no Trabalho, Depressão e Síndrome de Burnout na Cultura de Segurança do Paciente: Estudo Transversal com Análise de Modelagem por Equações Estruturais [tese]. Araraquara: Universidade Estadual Paulista "Júlio de Mesquita Filho"; 2018.

12 Cavalheiro G, Tolfo SR. Trabalho e depressão: um estudo com profissionais afastados do ambiente laboral. Psico USF 2011; 16(2):241-249.^-1313 Ferreira-Costa RQ, Pedro-Silva N. Ansiedade e depressão: o mundo da prática docente e o adoecimento psíquico. Estud Psicol 2018; 23(4):357-368.. Factors such as age¹⁴14 Caetano LM, Souza JM, Costa RQF, Silva D, Dell'Agli BAV. A saúde mental dos professores: a espiritualidade como estratégia protetiva em tempos de pandemia. Saud Pesq 2022; 15(2):e10334.^,1515 McCall WV, Cohen W, Reboussin B, Lawton P. Effects of mood and age on quality of life in depressed inpatients. Affect Disord 1999; 55(2-3):107-114., income¹⁴14 Caetano LM, Souza JM, Costa RQF, Silva D, Dell'Agli BAV. A saúde mental dos professores: a espiritualidade como estratégia protetiva em tempos de pandemia. Saud Pesq 2022; 15(2):e10334.^,1616 Santos LB, Nascimento KG, Fernandes AGO, Silva TCR. Prevalência, severidade e fatores associados à depressão em estudantes universitários. SMAD 2021; 17(1):92-100.^,1717 Souza JM, Dell'Agli BAV, Costa RQF, Caetano LM. Docência na pandemia: saúde mental e percepções sobre o trabalho on-line. Teor Prat Educ 2021; 24(2):142-159., lifestyle¹⁵15 McCall WV, Cohen W, Reboussin B, Lawton P. Effects of mood and age on quality of life in depressed inpatients. Affect Disord 1999; 55(2-3):107-114.^,1818 Ravindran AV, Matheson K, Griffiths J, Merali Z, Anisman H. Stress, coping, uplifts, and quality of life in subtypes of depression: a conceptual frame and emerging data. J Affect Disord 2002; 71(1-3):121-130. and adiposity¹⁹19 Milaneschi Y, Lamers F, Peyrot WJ, Baune BT, Breen G, Dehghan A, Forstner AJ, Grabe HJ, Homuth G, Kan C, Lewis C, Mullins N, Nauck M, Pistis G, Preisig M, Rivera M, Rietschel M, Streit F, Strohmaier J, Teumer A, Van der Auwera S, Wray NR, Boomsma DI, Penninx BWJH. Genetic Association of Major Depression With Atypical Features and Obesity-Related Immunometabolic Dysregulations. JAMA Psychiatry 2017; 74(12):1214-1225.^,2020 Casselli DDN, Silva ESM, Figueira GM, Demarchi ME, Souza JC. Comorbidade entre depressão, ansiedade e obesidade e complicações no tratamento. Res Soc Dev 2021; 10(1):e16210111489. influence the occurrence of DSs. Such illness has also been associated with the possible relationship an individual has with his or her work¹¹11 Oliveira AM. Relação da Satisfação no Trabalho, Depressão e Síndrome de Burnout na Cultura de Segurança do Paciente: Estudo Transversal com Análise de Modelagem por Equações Estruturais [tese]. Araraquara: Universidade Estadual Paulista "Júlio de Mesquita Filho"; 2018.

12 Cavalheiro G, Tolfo SR. Trabalho e depressão: um estudo com profissionais afastados do ambiente laboral. Psico USF 2011; 16(2):241-249.^-1313 Ferreira-Costa RQ, Pedro-Silva N. Ansiedade e depressão: o mundo da prática docente e o adoecimento psíquico. Estud Psicol 2018; 23(4):357-368..

Satisfaction with work, i.e., the perception that workers have regarding their work or the realization of their value through work²¹21 Locke EA. What is job satisfaction? Organ Behav Hum Perform 1969; 4(4):309-336.^,2222 Hora GPR, Ribas Junior R, Souza MA. Estado da Arte das medidas em satisfação no trabalho: uma revisão sistemática. Trends Psychol 2018; 26(2):971-986., is an individual and subjective feeling that reflects how much a worker enjoys his or her work²²22 Hora GPR, Ribas Junior R, Souza MA. Estado da Arte das medidas em satisfação no trabalho: uma revisão sistemática. Trends Psychol 2018; 26(2):971-986.

23 Krueger P, White D, Meaney C, Kwong J, Antao V, Kim F. Predictors of job satisfaction among academic family medicine faculty: Findings from a faculty work-life and leadership survey. Can Fam Physician 2017; 63(3):177-185.^-2424 Pepe A, Addimando L, Veronese G. Measuring Teacher Job Satisfaction: Assessing Invariance in the Teacher Job Satisfaction Scale (TJSS) Across Six Countries. Eur J Psychol Assess 2017; 13(3):396-416.. It is a complex concept that is influenced by various aspects related to work that interact with living and health conditions²⁵25 Sampaio RJ. A influência da (in)satisfação laboral em indicadores de Ansiedade e Depressão [dissertação]. Covilhã: Universidade da Beira Interior; 2020.^,2626 Marqueze EC, Moreno CRC. Satisfação no trabalho - uma breve revisão. Rev Bras Saude Ocup 2005; 30(112):69-79.. Satisfaction with work can be positive for health, just as dissatisfaction can harm the health of workers and the organization/institution where they work²²22 Hora GPR, Ribas Junior R, Souza MA. Estado da Arte das medidas em satisfação no trabalho: uma revisão sistemática. Trends Psychol 2018; 26(2):971-986.. Task overload, without proper management, can cause workers to perceive work as a burden, leading to loss of achievement, burnout and dissatisfaction¹¹11 Oliveira AM. Relação da Satisfação no Trabalho, Depressão e Síndrome de Burnout na Cultura de Segurança do Paciente: Estudo Transversal com Análise de Modelagem por Equações Estruturais [tese]. Araraquara: Universidade Estadual Paulista "Júlio de Mesquita Filho"; 2018..

Thus, job satisfaction and mental health complement and influence each other to the extent that if an individual is not satisfied with his or her work conditions, his or her mental health will be affected, and vice versa²⁵25 Sampaio RJ. A influência da (in)satisfação laboral em indicadores de Ansiedade e Depressão [dissertação]. Covilhã: Universidade da Beira Interior; 2020.. The mediation of job satisfaction in the occurrence of DSs has already been evaluated among health professionals¹¹11 Oliveira AM. Relação da Satisfação no Trabalho, Depressão e Síndrome de Burnout na Cultura de Segurança do Paciente: Estudo Transversal com Análise de Modelagem por Equações Estruturais [tese]. Araraquara: Universidade Estadual Paulista "Júlio de Mesquita Filho"; 2018. and seems plausible among professors³3 Ferreira-Costa RQ, Pedro-Silva N. Níveis de ansiedade e depressão entre professores do Ensino Infantil e Fundamental. Pro-Posições 2019; 30:e20160143.^,1313 Ferreira-Costa RQ, Pedro-Silva N. Ansiedade e depressão: o mundo da prática docente e o adoecimento psíquico. Estud Psicol 2018; 23(4):357-368.. Although the identified relationships have been recognized, no models have been employed to investigate the subject among professors and elucidate the complex possible interrelationships that influence DSs, with job dissatisfaction as a mediator, thus requiring further clarification^3.13. In this context, understanding the interrelationships that influence the occurrence of DSs among teaches may shed light on the mental illness that results from these relationships and contribute to the generation of public strategies that value teachers and their health. Thus, the aim of the present study was to evaluate the interrelationships of factors associated with depressive symptoms in teachers, considering dissatisfaction with teaching as a possible mediator.

Method

In this present study, data from the ProfSMoc project “Chronic health conditions and associated factors among teachers at the public state school of Montes Claros: A population-based study” were used. This was an epidemiological, cross-sectional study conducted in the municipality of Montes Claros, which is an urban centre in north Minas Gerais that had an estimated population of 404,804 inhabitants in 2018²⁷27 Whelton PK, Carey RM, Aronow WS, Casey DE, Collins KJ, Dennison Himmelfarb C, DePalma S, Gidding S, Jamerson KA, Jones D, MacLaughlin EJ, Muntner P, Ovbiagele B, Smith Jr SC, Spencer CC, Stafford RS, Taler Sj, Thomas RJ, Williams Sr K, Williamson JD, Wright Jr JT. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension 2018; 71(6):1269-1324.. Data collection occurred from March to December 2016.

The study participants were basic education teachers (elementary and high school) of the state public network. Probability sampling was performed using cluster sampling in a single stage, with the school as the primary sampling unit (PSU). Of the 49 state schools in the urban area of the municipality, 35 were randomly selected (PSU) by probability proportional to size (PPS), and the number of teachers in the schools was the reference parameter for the draw. A formula for a finite population was used (N=1,851) considering a prevalence of 50% of the event of interest, a confidence level of 95%, a tolerable error of 5%, an additional 10% to compensate for possible losses, and deff=2.0. The estimated sample size was 700 teachers. All teachers working in the classroom for at least one year at the schools included were invited to participate. Individuals who deviated from their teaching roles or were on leave from work were excluded because the intention of the study was to observe the influence of classroom performance; additionally, the exclusion of these individuals led to a more homogeneous sample.

A pilot study was conducted to adjust the instrument and data collection strategy. Data collection took place between March and December 2016 and consisted of three stages in each school. The first stage involved the school administrators. The project was presented, consent was requested from the administrators to conduct the study on site, and the second stage was scheduled (meeting with teachers to raise awareness and invite them to participate). During the second stage, at Module II meetings (meetings that are part of each teacher’s workload), self-administered questionnaires were delivered to the teachers who agreed to participate in the study, consent forms were collected, and the third stage was scheduled. The third stage consisted of the teachers returning the completed questionnaires, with verification by the field team to avoid loss of/inconsistencies in information, and physical assessments were also carried out (weight, height, bioimpedance, waist circumference (WC), hip circumference (HC) and blood pressure (BP). These evaluations were conducted in accordance with World Health Organization (WHO) guidelines using standardized procedures²⁸28 World Health Organization (WHO). Depression and other common mental disorders: global health estimates [Internet]. 2017 [cited 2022 jan 14]. Available from: https://apps.who.int/iris/handle/10665/254610.
https://apps.who.int/iris/handle/10665/2... and respect for biosafety standards. The measurements were performed in duplicate, and the mean was recorded. The evaluators received prior training and calibration, resulting in satisfactory intra- and inter-examiner agreement (weighted kappa greater than 81% and intraclass correlation coefficient (ICC) greater than 0.85). The collected data were also input in duplicate, followed by an analysis to check for/correct inconsistencies.

The outcome of interest in this study was DSs. DSs were evaluated using the Beck Depression Inventory (BDI)²⁹29 Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961; 4:561-571.. The Brazilian version was validated among university students, was cross-culturally adapted and was shown to be valid and reliable. The internal consistency of the BDI was high (0.81), and the general pattern of results confirmed construct validity³⁰30 Gorenstein C, Andrade L. Inventário de depressão de Beck: propriedades psicométricas da versão em português. Rev Psiquiatr Clín 1998; 25(5):245-250.. The instrument has 21 questions, covering feelings and attitudes. Each question is scored from 0 to 3, with a maximum score of 63 points³¹31 Cunha JA. Manual em português das Escalas Beck. São Paulo: Casa do Psicólogo; 2001.. Higher scores reflect greater involvement of DSs.

Given the absence of a specific theoretical model, an operational model was proposed to identify interrelationships that influence DSs among teachers (Figure 1). This model included variables already indicated in the literature as exerting some influence on DSs, for example, age¹⁴14 Caetano LM, Souza JM, Costa RQF, Silva D, Dell'Agli BAV. A saúde mental dos professores: a espiritualidade como estratégia protetiva em tempos de pandemia. Saud Pesq 2022; 15(2):e10334.^,1515 McCall WV, Cohen W, Reboussin B, Lawton P. Effects of mood and age on quality of life in depressed inpatients. Affect Disord 1999; 55(2-3):107-114., income¹⁴14 Caetano LM, Souza JM, Costa RQF, Silva D, Dell'Agli BAV. A saúde mental dos professores: a espiritualidade como estratégia protetiva em tempos de pandemia. Saud Pesq 2022; 15(2):e10334.^,1616 Santos LB, Nascimento KG, Fernandes AGO, Silva TCR. Prevalência, severidade e fatores associados à depressão em estudantes universitários. SMAD 2021; 17(1):92-100.^,1717 Souza JM, Dell'Agli BAV, Costa RQF, Caetano LM. Docência na pandemia: saúde mental e percepções sobre o trabalho on-line. Teor Prat Educ 2021; 24(2):142-159., lifestyle¹⁵15 McCall WV, Cohen W, Reboussin B, Lawton P. Effects of mood and age on quality of life in depressed inpatients. Affect Disord 1999; 55(2-3):107-114.^,1818 Ravindran AV, Matheson K, Griffiths J, Merali Z, Anisman H. Stress, coping, uplifts, and quality of life in subtypes of depression: a conceptual frame and emerging data. J Affect Disord 2002; 71(1-3):121-130. and adiposity¹⁹19 Milaneschi Y, Lamers F, Peyrot WJ, Baune BT, Breen G, Dehghan A, Forstner AJ, Grabe HJ, Homuth G, Kan C, Lewis C, Mullins N, Nauck M, Pistis G, Preisig M, Rivera M, Rietschel M, Streit F, Strohmaier J, Teumer A, Van der Auwera S, Wray NR, Boomsma DI, Penninx BWJH. Genetic Association of Major Depression With Atypical Features and Obesity-Related Immunometabolic Dysregulations. JAMA Psychiatry 2017; 74(12):1214-1225.^,2020 Casselli DDN, Silva ESM, Figueira GM, Demarchi ME, Souza JC. Comorbidade entre depressão, ansiedade e obesidade e complicações no tratamento. Res Soc Dev 2021; 10(1):e16210111489.. The influence of satisfaction with teaching on DSs has been investigated¹²12 Cavalheiro G, Tolfo SR. Trabalho e depressão: um estudo com profissionais afastados do ambiente laboral. Psico USF 2011; 16(2):241-249.^,1313 Ferreira-Costa RQ, Pedro-Silva N. Ansiedade e depressão: o mundo da prática docente e o adoecimento psíquico. Estud Psicol 2018; 23(4):357-368.. However, the interrelationships that affect DSs mediated by job satisfaction was investigated this study; and the study design was based on a study conducted among health professionals¹⁵15 McCall WV, Cohen W, Reboussin B, Lawton P. Effects of mood and age on quality of life in depressed inpatients. Affect Disord 1999; 55(2-3):107-114.^,1818 Ravindran AV, Matheson K, Griffiths J, Merali Z, Anisman H. Stress, coping, uplifts, and quality of life in subtypes of depression: a conceptual frame and emerging data. J Affect Disord 2002; 71(1-3):121-130..

Figure 1 presents the hypothesized multivariate operating model for testing. The direct and indirect interrelationships between DSs (outcome of interest) and the other variables, mediated by job dissatisfaction, were evaluated. The observed variables are in the rectangles, and the latent variable is in the ellipse. The interrelationships are illustrated using arrows directed from the independent to dependent variable.

The study variables were job dissatisfaction, age, income, lifestyle and adiposity (Figure 1). Job dissatisfaction was assessed using a separate question (“Overall, how do you feel about your job as a teacher?”) scored with a five-category Likert scale (very satisfied, moderately satisfied, neither satisfied nor dissatisfied, dissatisfied and very dissatisfied), with scores ranging from 1 to 5 and higher values indicating greater dissatisfaction. This variable mediates the relationships between DS and the other study variables. Locke²¹21 Locke EA. What is job satisfaction? Organ Behav Hum Perform 1969; 4(4):309-336. defines job satisfaction as the result of an evaluation by a worker about his or her work or feeling of value through work, being a positive emotion of well-being. For Martinez³²32 Martinez MC. As relações entre a satisfação com aspectos psicossociais no trabalho e a saúde do trabalhador [dissertação]. São Paulo: Faculdade de Saúde Pública; 2002., satisfaction and dissatisfaction are on a continuum, in which satisfaction is at one extreme and dissatisfaction at the other. This concept of job satisfaction allows for an evaluation using a Likert scale, an approach that has been used in other studies³³33 Alves FC. A (in)satisfação dos professores: estudo de opiniões dos professores do ensino secundário do distrito de Bragança. In: Estrela MT, organizadora. Viver e Construir a Profissão Docente. Porto: Porto Editora; 1997. p. 81-115.^,3434 Ferreira LCM. Crenças de autoeficácia docente, satisfação com o trabalho e adoecimento. Psicol Ensino Form 2014; 5(2):19-37..

Age and income were analysed in their original numerical form. Lifestyle was assessed using the “Fantastic Lifestyle” tool, which includes the main aspects of a healthy lifestyle. This instrument was developed in Canada and later translated into Portuguese and validated among young adult students at the Federal University of Santa Catarina; the translated version met internal and external consistency criteria and construct validity, with psychometric properties appropriate for the assessment of lifestyle in epidemiological studies³⁵35 Añez CRR, Rodrigo SR, Petroski EL. Versão brasileira do questionário "estilo de vida fantástico". Arq Bras Cardiol 2008; 91(2):102-109.. Fantastic is an acronym that represents 9 domains: F = family and friends; A = physical activity; N = nutrition; T = tobacco/cigarettes and drugs; A = alcohol; S = sleep, seat belt, stress and safer sex; T = type/pattern of behaviour; I = introspection; and C = career/job. The instrument has 25 questions divided into 9 domains, whose response options are in the form of a Likert scale. The total maximum score is 100 points³⁵35 Añez CRR, Rodrigo SR, Petroski EL. Versão brasileira do questionário "estilo de vida fantástico". Arq Bras Cardiol 2008; 91(2):102-109.; the lower the score, the greater the need for change in the respondents’ lifestyle, and the higher the score, the better the lifestyle adopted by the respondent. This scoring system was consistent that used most studies that have employed this instrument; the measurement model adjusted, using CFA, for the lifestyle construct did not show adequate fit: χ²/df=5.68; GFI=0.843; CFI=0.615; TLI=0.578 and RMSEA=0.08 (p=0.000). Adiposity was treated as a latent variable, defined by three observed variables: waist circumference (WC), body mass index (BMI) and waist-to-height ratio (WHR); the numerical values were used for these variables (Figure 2).

BMI is an indicator of nutritional status and is used to assess the distribution of fat throughout the body; WC and WHR assess the accumulation of fat in the abdominal region. For this reason, we chose to study adiposity, a manifestation of abdominal fat accumulation and body fat distribution, as a latent variable. Furthermore, this model for measuring adiposity has already been adopted in previous studies³⁶36 Freire RS, Reis VMCP, Brito AB, Brito MFSF, Pinho L, Silva RRV, Silveira MF. Analysis of the interrelationships between factors influencing blood pressure in adults. Rev Saude Publ 2020; 54:147.^,3737 Barbosa IA, Silveira MF, Fonseca ADG, Brito MFSF, Pinho L, Silva CSO. Quality of life and adiposity in adolescents: direct and indirect effects. Cien Saude Colet 2021; 26(11):5661-5670..

In the descriptive analyses, the ordinal categorical variable (job satisfaction) is presented as simple and relative frequencies. The other variables, all numeric, are presented as the mean, standard deviation, median, minimum and maximum values, coefficients of skewness (sk) and kurtosis (ku). For inclusion in the model, numerical variables must have met the assumption of normality (sk<3 or ku<10)³⁸38 Marôco J. Análise de equações estruturais: fundamentos teóricos, software e aplicações. Lisboa: Report Number; 2014.. Only the variable income violated this assumption, and therefore, values were logarithmically transformed before being introduced into the model, correcting the problem.

The model fit was performed in two stages. First, the latent variable (adiposity) was established through confirmatory factor analysis (Figure 2). A good fit of this variable was considered based on significant factorial weights (p<0.05) and standardized factorial weights greater than or equal to 0.50 (λ≥0.50)³⁸38 Marôco J. Análise de equações estruturais: fundamentos teóricos, software e aplicações. Lisboa: Report Number; 2014..

Next, the multivariate model was adjusted using structural equation modelling, with estimation of direct and indirect effects that influence the presence of DSs as well as potential interrelationships between the variables considered in the analysis and job dissatisfaction (Figure 3).

Standardized coefficients were estimated, and the critical ratio (CR), at the 5% level, was used to assess significance¹²12 Cavalheiro G, Tolfo SR. Trabalho e depressão: um estudo com profissionais afastados do ambiente laboral. Psico USF 2011; 16(2):241-249.. For interpretation, the possibility of a small effect (values close to 0.10), medium effect (close to 0.30) or large effect (greater than 0.50) was considered. Bentler’s comparative fit index (CFI), goodness of fit index (GFI) and Tucker-Lewis index (TLI) were used to assess the quality of the fit, with values >0.90 indicating a good fit. In addition, the root mean square error of approximation (RMSEA) was used, with values <0.10 indicating an acceptable fit, and the absolute index (X²/df), with values <5 indicating an acceptable fit³⁹39 Kline RB. Principles and Practice of Structural Equation Modeling. New York: Guilford Press; 2011..

The descriptive data were analysed using the Statistical Package for Social Sciences (SPSS) software, version 18.0. For the analysis (latent variable analysis and structural equation modelling) of the models, IBM SPSS AMOS (version 23.0) was used.

This study was approved by the Research Ethics Committee (CEP) of the State University of Montes Claros - Unimontes (Opinion 1,293,458 of 10/23/2015) and was conducted in accordance with the ethical precepts of the National Health Council (CNS)/Resolution No. 466/2012 and the Declaration of Helsinki of the World Medical Association.

Results

In this study, data from 700 basic education teachers were analysed. There was a predominance of females (n=582; 83%). Regarding, 18% of the participants were 30 years or younger, 66% were between 31 and 50 years, 14.4% were between 51 and 60 years, and only 1.7% were 61 years or older. Regarding race, 58.2% self-declared as mixed race, 29.6% self-declared as white, 9.3% self-declared as black, and 3% self-declared as indigenous or yellow. Regarding education, 44.7% reported having only undergraduate degrees, and 55% reported having latu sensu graduate degrees. There was a predominance of teachers with a stable union (61.6%). The descriptive measures of the variables of the proposed model are presented in Table 1; the majority of teachers were dissatisfied with their work (34.9%), followed by those who were dissatisfied often (28.7%), sometimes (28.6%), rarely (6.1%) and almost never (1.7%).

Figure 2 displays the results of the confirmatory factor analysis of the latent variable adiposity; all the observed variables that composed this variable had adequate factor weights (≥0.5) and were statistically significant (p<0.001). The fitted model was saturated (determined); that is, the number of estimated parameters was equal to the number of nonredundant elements of the covariance matrix, and df=0. Thus, this model had a single solution and, therefore, a perfect fit.

Figure 3 presents the results of the multivariate analysis using structural equation modelling. The fit indices were considered satisfactory: X²/df=2.65; CFI=0.996; GFI=0.992; TLI=0.982; and RMSEA=0.049 (90%CI 0.025-0.073). A significant positive direct effect was observed between age and DSs (β=0.12) and between job dissatisfaction and DSs (β=0.12). A significant negative direct effect was observed between lifestyle and job dissatisfaction (β=-0.51), between lifestyle and DSs (β=-0.60), between lifestyle and adiposity (β=-0.17), between adiposity and job dissatisfaction (β=-0.19), and between adiposity and DSs (β=-0.10). Thus, older age and greater dissatisfaction with teaching were directly associated with DSs, and a more favourable lifestyle and adiposity were associated with a lower occurrence of DSs.

In the mediation analysis (Table 2), a significant indirect effect (mediated by job dissatisfaction) was observed for the variables lifestyle (β=-0.612) and adiposity (β=-0.0228), both with negative effects. In other words, a healthier lifestyle and adiposity were associated with a lower occurrence of DSs, even when mediated by dissatisfaction with teaching.

Discussion

The operational model tested identified interrelationships that influence DSs, directly and indirectly, mediated by job dissatisfaction. This model proved to be adequate because it has a single solution, and all degrees of freedom of its covariance matrix were consumed to estimate the relationships of the proposed model. These measures indicate a very good fit and adequacy of the model. The model revealed that DSs were influenced by direct and indirect effects, mediated by job dissatisfaction, confirming the hypothesis investigated. Regarding the direct effects, age, job dissatisfaction, lifestyle and adiposity were associated with the occurrence of DSs.

The direct effect of age on DSs was positive (β=0.12), revealing that an increase in the age of teachers is associated with the presence of more DSs. Previous studies have identified that increasing age is related to a higher occurrence of common mental disorders among professors, for example, depression and anxiety, especially among women³3 Ferreira-Costa RQ, Pedro-Silva N. Níveis de ansiedade e depressão entre professores do Ensino Infantil e Fundamental. Pro-Posições 2019; 30:e20160143.^,4040 Freitas RF, Ramos DS, Freitas TF, Souza GR, Pereira EJ, Lessa AC. Prevalência e fatores associados aos sintomas de depressão, ansiedade e estresse em professores universsitários durante a pandemia da COVID-19. J Bras Psiquiatr 2021; 70(4):283-292.. A study by Souza and Costa⁴¹41 Souza JC, Costa DS. Qualidade de vida de uma amostra de profissionais de educação física. J Bras Psiq 2011; 60(1):23-27. suggested that younger age is linked to better physical quality of life and functional capacity of professors. Notably, the universe of teachers is predominantly female, and in the sample investigated, the mean age of females was 41±9.5 years (data not shown). For women, this age group is rightly considered borderline because the person, by ceasing to perceive themselves as young, may judge themselves incapable of performing certain actions and attaining certain dreams, leading to increased anxiety and/or depression, which affects professional practice³3 Ferreira-Costa RQ, Pedro-Silva N. Níveis de ansiedade e depressão entre professores do Ensino Infantil e Fundamental. Pro-Posições 2019; 30:e20160143..

The direct effect of job dissatisfaction on DSs was also positive (β=0.12), indicating that greater job dissatisfaction is associated with the occurrence of DSs and confirming the main hypothesis of this study. Exposure to stressors at work can cause dissatisfaction with work and thus impact mental health⁴²42 Ribeiro RBN, Assunção AA, Araújo TM. Factors associated with job satisfaction among public-sector physicians in Belo Horizonte, Brazil. Int J Health Serv 2014; 44:787-804.. Previous studies have identified a relationship between worse perception of work and the impact on the mental health of professors³3 Ferreira-Costa RQ, Pedro-Silva N. Níveis de ansiedade e depressão entre professores do Ensino Infantil e Fundamental. Pro-Posições 2019; 30:e20160143.^,4343 Carraro MM. Condições de trabalho e transtornos mentais comuns em professores da rede básica municipal de encino de Bauru-SP [dissertação]. Botucatu: Universidade Estadual Paulista Júlio de Mesquita Filho; 2015.

44 Martinez MC, Paraguay AIBB. Satisfação e saúde no trabalho - aspectos conceituais e metodológicos. Cad Psicol Soc Trab 2003; 6:59-78.^-4545 Valle LELR. Estresse e distúrbios do sono no desempenho de professores: saúde mental no trabalho [tese]. São Paulo: Universidade de São Paulo; 2011.. Among university professors, a direct relationship was also observed between DSs and the Vulnerability to Work Stress Scale, which scores items related to work nuisance that measure dissatisfaction with work⁴⁶46 Baptista MN, Soares TFP, Raad AJ, Santos LM. Burnout, estresse, depressão e suporte laboral em professores universitários. Rev Psicol Organ Trab 2019; 19(1):564-570.. Among Italian teachers, in 2016⁴⁷47 Zurlo MC, Pes D, Capasso R. Personality Characteristics, Job Stressors, and Job Satisfaction: Main and Interaction Effects on Psychological and Physical Health Conditions of Italian Schoolteachers. Psychol Rep 2016; 119(1):27-38., a lower occurrence of mental health problems was observed regarding the perception of satisfaction with work relationships (OR=0.27). The study also observed that the superposition of negative questions related to teaching increased the possibility of mental health problems (OR=17.44)⁴⁷47 Zurlo MC, Pes D, Capasso R. Personality Characteristics, Job Stressors, and Job Satisfaction: Main and Interaction Effects on Psychological and Physical Health Conditions of Italian Schoolteachers. Psychol Rep 2016; 119(1):27-38..

Lifestyle had a direct negative effect on DSs (β=-0.60), revealing that a healthier lifestyle was associated with a lower occurrence of DSs. The relationship between mental health problems, including DSs, and lifestyle is also recognized in the scientific literature⁴⁸48 Kataoka S, Kitamura K, Kataoka Y, Yamamoto Y. Depressive symptoms and their associated factors in nursery school teachers: a multicenter cross- sectional study. Cureus 2021; 13(7):e16545.

49 Zhou J, Yuan X, Huang H, Li Y, Yu H, Chen X, Luo J. The Prevalence and Correlative Factors of Depression Among Chinese Teachers During the COVID-19 Outbreak. Front Psychiatry 2021; 12:644276.^-5050 Martins RJ, Silva MB, Ávalos YRB, Belila NM, Garbin CAS. Nível de ansiedade e depressão e problemas relacionados ao uso de álcool em funcionários de uma universidade pública. Res Soc Develop 2021; 10(4):e53610414243.. Among university professors, lifestyle, as measured using the same scale used in our study, was also associated with stress levels⁵¹51 Amim EF, Daher DV, Brito IS, Cursino EG, Teixeira ER, Koopmans FF, Faria MGA. Lifestyle and health promotion of university professors. Res Soc Develop 2020; 9(9):e593997507.. Among the elderly in Santa Catarina in 2016, factors such as not having a healthy diet, a compromised sleep pattern and lower volumes of physical activity had impacts on the occurrence of DSs⁵²52 Silva MR, Ferretti F, Pinto SS, Tombini Filho OF. Depressive symptoms in the elderly and its relationship with chronic pain, chronic diseases, sleep quality and physical activity level. Br J Pain 2018; 1(4):293-298.. Physical exercise, specifically, has been strongly indicated as a protective factor against the occurrence of DSs⁵³53 Cebrino J, Portero de la Cruz S. Diet Quality and Sociodemographic, Lifestyle, and Health-Related Determinants among People with Depression in Spain: New Evidence from a Cross-Sectional Population-Based Study (2011-2017). Nutrients 2020; 13(1):106.^,5454 Costa RA, Soares HLR, Teixeira JAC. Benefícios da atividade física e do exercício físico na depressão. Rev Dep Psicol UFF 2007; 19(1):269-276., and among teachers from Barretos, a high prevalence of physical inactivity was observed, with a sedentary lifestyle associated with worse quality of life⁵⁵55 Cirilo JC, Oliveira DM, Fernandes EV, Macedo AG, Santos D. Influence of teaching work on individual well-being, quality of life, and (in) physical activity of elementary teachers. Res Soc Develop 2022; 11(1):e1511123919..

Adiposity also had a direct negative effect on DSs (β=-0.10); that is, higher adiposity had a protective effect against the occurrence of DS. The relationship between weight and DSs is evidenced in the literature; however, its intensity is related to the stage in which these symptoms are found⁵⁶56 Peixoto HE, Vasconcelos IA, Sampaio ACM, Ito MK. Antidepressivos e alterações no peso corporal. Rev Nutr 2008; 21(3):341-348.

57 Kazes M, Danion JM, Grangé D, Pradignac A, Simon Ch, Burrus-Mehl F, Schlienger JL, Singer L. Eating behaviour and depression before and after antidepressant treatment: a prospective, naturalistic study. J Affect Disord 1994; 30(3):193-207.^-5858 Flórez KR, Dubowitz T, Ghosh-Dastidar MB, Beckman R, Collins RL. Associations between depressive symptomatology, diet, and body mass index among participants in the supplemental nutrition assistance program. J Acad Nutr Diet 2015; 115(7):1102-1108.. Lack of appetite is listed as an important complaint in individuals with depression⁵⁵55 Cirilo JC, Oliveira DM, Fernandes EV, Macedo AG, Santos D. Influence of teaching work on individual well-being, quality of life, and (in) physical activity of elementary teachers. Res Soc Develop 2022; 11(1):e1511123919.^,5656 Peixoto HE, Vasconcelos IA, Sampaio ACM, Ito MK. Antidepressivos e alterações no peso corporal. Rev Nutr 2008; 21(3):341-348.. A study investigated the eating behaviour of severely depressed patients before and after treatment with antidepressants. Before treatment, the patients had lower caloric intake than did individuals in the control group; however, with treatment, there was an increase in the energy intake of the patients, who began to ingest the same amount of energy as the control group. At the end of treatment, there was an average gain of 2.5 kg, and at the end of the maintenance period, there was a gain of 5.2 kg compared to pretreatment weight. The recovery of weight lost before treatment due to inappetence was a positive response to antidepressant treatment⁵⁷57 Kazes M, Danion JM, Grangé D, Pradignac A, Simon Ch, Burrus-Mehl F, Schlienger JL, Singer L. Eating behaviour and depression before and after antidepressant treatment: a prospective, naturalistic study. J Affect Disord 1994; 30(3):193-207.. In contrast, overweight and obesity have been associated with the occurrence of DSs⁵⁸58 Flórez KR, Dubowitz T, Ghosh-Dastidar MB, Beckman R, Collins RL. Associations between depressive symptomatology, diet, and body mass index among participants in the supplemental nutrition assistance program. J Acad Nutr Diet 2015; 115(7):1102-1108..

Notably, although the literature has reported greater trends in DS occurrence with decreasing levels of education and income¹⁶16 Santos LB, Nascimento KG, Fernandes AGO, Silva TCR. Prevalência, severidade e fatores associados à depressão em estudantes universitários. SMAD 2021; 17(1):92-100.^,5959 Cunha RV, Bastos GAN, Duca GF. Prevalência de depressão e fatores associados em comunidade de baixa renda de Porto Alegre, Rio Grande do Sul. Rev Bras Epidemiol 2012; 15(2):346-354., in our study, income had a direct effect on DSs. However, this relationship was not statistically significant.

DSs were also influenced by indirect effects (mediated by job dissatisfaction). In this case, lifestyle (β=-0.0612) and adiposity (β=-0.0228) both had a negative indirect effects on DSs. Dissatisfaction with work potentiated the effect that a less healthy lifestyle had on a higher occurrence of DSs. Likewise, higher adiposity, mediated by job dissatisfaction, had a protective effect against the occurrence of DSs. These findings reinforce the hypothesis tested that such dissatisfaction, in addition to having a direct effect on DSs, may also potentiate the effect of other existing factors that also predispose individuals to this condition. Thus, our findings corroborate the notion that dissatisfaction with work goes beyond the limits of the school environment, being influenced by and exerting an influence on life outside of work²⁶26 Marqueze EC, Moreno CRC. Satisfação no trabalho - uma breve revisão. Rev Bras Saude Ocup 2005; 30(112):69-79.. No other studies addressing this issue among teachers were found, making comparisons of the results difficult. However, this relationship has already been observed among health professionals¹¹11 Oliveira AM. Relação da Satisfação no Trabalho, Depressão e Síndrome de Burnout na Cultura de Segurança do Paciente: Estudo Transversal com Análise de Modelagem por Equações Estruturais [tese]. Araraquara: Universidade Estadual Paulista "Júlio de Mesquita Filho"; 2018..

The prevalence of dissatisfaction with teaching observed in this study (35%) is worrying and deserves more attention. Among teachers in public schools in Passo Fundo, 37.3% were dissatisfied⁶⁰60 Viana CV, Maronezi LFC, Tribino UM, Acrani GO, Lindemann IL, Silva SG. Satisfação no Trabalho Docente: Uma Análise Epidemiológica de Professores da Rede Pública de Ensino de Passo Fundo, RS [Internet]. 2021 [acessado 2022 jun 15]. Disponível em: https://publicacoeseventos.unijui.edu.br/index.php/conintsau/article/view/19407.
https://publicacoeseventos.unijui.edu.br... , a percentage similar to that found in this study and higher than that found among teachers of higher education (18.2%)⁶¹61 Cardoso CG, Costa NM. Factors connected with professional satisfaction and dissatisfaction among nutrition teacher. Cien Saude Colet 2016; 21(8):2357-2364..

This study has limitations: it is cross-sectional and limited to causal inference. In addition, the nonparticipation of teachers who were not working in the classroom (teachers on sick leave or who performing tasks other than teaching) may have led to an underestimation of DSs. Notably, the diagnosis of depression exceeds the power of the instrument used in the study, given that it is a screening scale for epidemiological studies to identify possible cases of DS; therefore, it is necessary to perform clinical diagnoses to rule out possible symptoms that may interfere with teaching activity. It was also difficult to establish comparisons with other studies on DSs due to the different approaches and different instruments and cut-off points found in the literature, in addition to populations of nonteachers.

The positive aspects of the study were the representative size of the sample, the rigor in conducting the study, the training and calibration of the team, the adoption of validated instruments, and the entry of data into the database in duplicate and an audit of the database, which confirmed internal validity to the study. In addition, the operational model tested was coherent, with satisfactory fit indices, and structural equation modelling allowed the evaluation of the set of possible relationships between the study variables, estimating the interrelationships.

Notably, it is necessary to implement motivational factors at work that psychologically empower teachers to provide them with tools to better deal with work issues. It is also necessary to ensure decent working conditions for teachers. The absence of these factors can generate feelings of frustration, discouragement, and lack of competence, in addition to lower commitment to work, more absenteeism and lower satisfaction⁶²62 Llorente-Alonso M, Topa G. Prevention of Occupational Strain: Can Psychological Empowerment and Organizational Commitment Decrease Dissatisfaction and Intention to Quit? J Clin Med 2018; 7(11):450., which can negatively affect the educational system and cause personal losses. Thus, it is necessary to create strategies that act as protective factors.

Conclusion

The structural equation model tested identified interrelationships that influence DSs. The increase in DSs was directly associated with increasing age, dissatisfaction with work, a less healthy lifestyle and lower adiposity. DSs also increased from indirect effects, mediated by dissatisfaction with work, from a less healthy lifestyle and lower adiposity. This evidence, considering DSs, teaching and job dissatisfaction, unveils mental illness derived from these relationships and situations, showing that attitudes that value teachers and their health are necessary, as are the needs for society to recognize this situation and attempt to provide subsidies to avoid the precariousness of teaching work.

Acknowledgements

The authors are grateful for the support of the Fundação de Amparo à Pesquisa do Estado de Minas Gerais - FAPEMIG (Universal Demand). DS Haikal and AM Batista de Paula are productivity fellows of the Conselho Nacional de Desenvolvimento Científico e Tecnológico. We also thank the participants of this study.

References

Publication Dates

History