Artigos

Conteúdo (article):

PHYSICAL AND PSYCHOSOCIAL DEMAND AT WORK: INEQUITIES RELATED TO RACE/SKIN COLOR

Abstract
This cross-sectional study investigated the association between self-reported race/skin color and two outcomes –psychosocial demand and physical demand at work– in 1,032 workers in an urban cleaning services company and two footwear manufacturers, located in the State of Bahia, Brazil. Psychosocial demand was measured through the Job Content Questionnaire and physical demand was measured through questions about postures and cargo handling. A Cox regression analysis provided prevalence ratios (PR) adjusted by age, gender, and educational level. Among blacks, there is a higher proportion of garbage collectors and a lower proportion of supervisory positions. Black workers are more subject to high psychological demand and low control and, consequently, to higher demand at work (PR=1.65). Also, they are more exposed to work with arms above shoulder level (PR=1.93), and load handling (PR=1.62), compared to white workers. Mixed workers are more exposed to low control (PR=1.36), work with arms above shoulder level (PR=1.48), and load handling (PR=1.25), also compared with whites. Social support is lower among blacks and mixed. The study demonstrated inequities in psychosocial and physical exposures at work that are in line with the structural conception of racism. This evidence can contribute to the adoption of practices that increase equity in the world of work.
Keywords: Psychosocial factors; Physical effort; Inequities; Racism; Worker’s health.

Introduction
The world of work is marked by inequalities and inequities. While the former arises from the variability of human characteristics, inequities consist of differences that are unjustly produced, avoidable, and liable to extinction. They make social groups vulnerable and are generally maintained by power relations1-3.
Structural racism, present in society and thus in institutions, contributes to the perpetuation of disadvantages among historically subjugated ethnic groups, sustaining inequities4-6.
The concentration of the worst socioeconomic status among blacks or mixed –often understood in the literature as ”Afrodescendants”, due to the common history of vulnerability7,8,9– is one of the effects of structural racism, and may have repercussions on occupational exposures4,6,9-10. Being African-American has been associated with work in occupations that have a high risk of nonfatal injuries and sickness, in the United States11. Moreover, occupations in which there is a higher proportion of Afrodescendants are associated with the worst self-reported health status by Afrodescendant and non-Afrodescendant workers12.
In the world of work, tasks with the least autonomy and social recognition, and the highest physical demand are frequent among Afrodescendant workers. Domestic work was highlighted among Afrodescendant female workers (21.8% among Afrodescendant women versus 12.6% among white women)8, as well as informal, more precarious work was superior among Afrodescendants (47.3% versus 34.6%)9, while whites predominate among magistrates (82.8%)14 and in managerial positions (68.6%)9. This panorama shows the establishment of a racial hierarchy in the world of work6,14.
In this perspective, a race is understood as a product of historical, socio-cultural, and ancestral relations that contribute to the constitution of identity, as opposed to identities based on genetic differentiations15-19. In epidemiological studies, the race is used for the operationalization of the "ethnic group" construct.
This study is based on the hypothesis that there are –scarcely investigated– racial inequities related to occupational exposure, that is, the highest physical and psychosocial overload is found in tasks performed by black and mixed, expressing the adverse conditions in the world of work.
The study of psychosocial demand at work has been based on theoretical models, among which the Job Demand-Control (JDC) model20, which advocates that psychosocial aspects can act as stressors, constituting risk factors for physical and psychological disorders. This model defines two dimensions: psychological, which consists of the psychological demand required in the course of a task, and control at work, which involves the worker\'s use of skills and creativity, in addition to autonomy for executing work. The model describes four experiences at work, according to levels of exposure to psychological demand and control. High-demand work, which results from high psychological demand combined with low control, is related to lack of motivation, low self-esteem, and loss of skills at work20, and is the experience most associated with health problems when compared to other work experiences: low-demand work, passive work, and active work20.
The JDC model was expanded to incorporate the social support dimension, coming from coworkers and supervisors, and this is understood as the dimension that can change the high-demand work experience, protecting workers who benefit from high social support21.
There is evidence that Blacks and mixed are more exposed, mainly, to low control and high demand, but the scientific literature on the subject is scarce10, 22-24.
In addition to higher psychosocial demand, black and mixed workers may be more subject to physical demand at work, which involves painstaking whole-body or body segments postures, in addition to load handling and use of force25, associated, among other harms, with musculoskeletal disorders and, particularly, with occupational back pain26-30. This situation was verified among African-American workers, who perform more tasks involving squatting and load handling, contributing to a higher perception of pain among them, when compared to White workers31.
Thus, this study aimed at analyzing the association between race/skin color and two work outcomes –the psychosocial and physical demands to which the workers are subject.

Method
A cross-sectional study was carried out with workers from an urban cleaning services company and with workers from two footwear manufacturing companies, in the State of Bahia, Brazil. A survey pool was conceived in the context of an international cooperation project32,33. The strategy of aggregating surveys –conducted with the same procedures and techniques– aimed at ensuring the participation of as many workers as possible and favoring the variability of occupational exposure, both aspects of interest from an analytical perspective.
The survey with urban cleaning services workers –all of them male– was a census (n=624) conducted in Salvador, the capital of the State, in the year 2010. The second survey, with workers from two footwear manufacturing companies, both located outside Salvador, used a random sample stratified by gender and company (n=446) and was conducted in the year 2012. The urban cleaning services workers were allocated to operation and maintenance occupations, while the footwear manufacturing workers were in administrative, operational, and maintenance occupations. The most frequent occupations were those of collectors, cleaning agents, and those related to footwear manufacturing operational activities. In both surveys together, the response rate was 97%. The surveys were conducted based on the premises of Worker\'s Health Surveillance –which indicated their opportunity–, in the context of interventions in working conditions, in cooperation with the State’s official agencies related to Workers\' Health.

Procedures and data collection tool
The questionnaire was applied in the workplace, during work hours, by a team of trained interviewers composed by five master’s program students (three physical therapists, one safety engineer, and one ergonomist) and four physical therapy undergraduate students. All the workers were informed about the objectives of the study. The confidentiality of the collected data was guaranteed and the workers’ privacy was ensured.
From the original questionnaire, the following types of questions were used in this study: questions regarding socio-demographic data –age, gender, and educational level–, questions about occupations –related to the Job Content Questionnaire (JCQ)34,35–, and questions regarding physical demand.
To minimize the healthy worker survival effect (that may result from the exclusion of those temporarily incapacitated for work), all workers that were employed for 12 or more months at the time of the study were considered eligible. Once drawn, individuals that were on leave of absence for reasons that supposedly weren’t related to work –such as maternity leave– were replaced by the next individual in the nominal list of workers provided by the respective company. There was a home interview with a worker from the footwear manufacturing company who was on medical leave. There was no long-term leave among the respondents.
Dependent Variables
1. Psychosocial demand at work
According to the Job Demand-Control-Support (JDCS) model20,21, the psychosocial aspects of work that characterize the psychosocial demand are psychological demand, control over work, social support, and high-demand work experience, measured using the JCQ34,35. The JDC model quadrants describe distinct forms of psychosocial exposure. High control and high demand constitute active work; low control and low demand configure passive work; low demand and high control result in low-demand work; finally, high demand and low control characterize high-demand work, which is considered the most deleterious experience20.
Social support from supervisors and coworkers, together, composed the dimension of social support.
The JCQ uses a 4-point Likert scale, ranging from "strongly disagree" (1) to "strongly agree" (4).
The validity and reproducibility of the JCQ were evaluated and considered good in a study conducted with formal and informal workers35. Santos et al.36 confirmed the dimensionality of the JCQ, according to the JDCS model.
2. Physical demand at work
Physical demand at work was investigated utilizing six questions regarding general postures or postures of body segments (standing; walking; squatting; rotating trunk; arms above shoulder level; exert pressure with hands on an object); and three questions regarding load handling (lifting; pushing; pulling). A six-point numeric response scale (0 to 5) with duration qualifiers (0-"never" and 5-"all the time") was used, except for the item regarding exerting pressure with hands on an object, whose response scale was of intensity (0-"very weak" and 5-"very strong"). The validity and reliability of these items of physical demand were tested by Fernandes et al.37 in the population of this study. The results show dimensionality and reliability ranging from excellent to moderate for most items, which recommends the use of the questions in epidemiological studies.
The utilization of questions regarding physical demand aims at overcoming the measurement of exposure based on the job title, allowing for incorporating the variability of exposure among subjects that have the same job title, by considering the tasks performed. Therefore, self-reporting by workers has been considered the most valid measure in the evaluation of physical exposure37.

Main independent variable
The variable "race/skin color" was obtained through self-reporting, according to the categories proposed by the Instituto Brasileiro de Geografia e Estatística – IBGE: white, black, yellow, mixed, and indigenous7.
The co-variables of interest were: age, gender, and educational level.
The statistical approach to data
1. Descriptive Approach
The minimum and maximum values, means, and medians of quantitative variables were obtained. The absolute and relative frequencies of the qualitative variables were also obtained.
Arising from the notion of homogeneous exposure groups, five occupational groups were defined according to similarities between the tasks performed and also based on the positions occupied by workers in their respective organizational structures. The occupational groups are collectors/cleaning agents; urban cleaning drivers; coordinator/supervisor/administrative position; workers in operations and footwear manufacturing workers; and maintenance workers. The distribution of workers according to race/skin color in each of the occupational groups was presented. The description of the tasks performed in each occupational group is not the object of this article32.
The scores obtained from the JDCS model dimensions were described by median, maximum, and minimum values. Then, "psychological demand" and "control over work" were dichotomized in high or low exposure, from the median. The same occurred to the score obtained in the "social support" dimension, which encompasses the support of coworkers and supervisors.
The categorical variables of the JDCS model –demand, control, and social support– were described separately. Then, the JDC model quadrants were used; the high-demand work was the risk stratum to be compared with the absence of high demand, which aggregated the other quadrants: passive work, active work, and low-demand work, being described according to the race/skin color variable.
The answers to the questionnaire on physical demand at work were categorized. Answers ranging from zero to two in the numeric scale corresponded to "low exposure" and answers ranging from three to five corresponded to "high exposure". Exposure to physical demand was described for each item, according to race/skin color.
The variable race/skin color was categorized into "white", "black", and "mixed". Workers who self-reported as yellow (n=15) or indigenous (n=20) were excluded from the study population. They existed in small numbers, which rendered analyzing the main independent variable in separate strata not viable. The inclusion of indigenous to the black stratum and of yellows to the white stratum –in order not to exclude them from the analysis– was explored by the authors. The results obtained in these analyses were similar to those obtained without the aforementioned incorporations, as expected, given the low number of individuals in each of those groups. However, the decision for exclusion was based on the criticism of a classification that did not rely on self-reporting.

2. Analytical approach
The category "white" of the independent variable race/skin color was considered the reference stratum (not exposed) in the analyses.
The variable "age" was dichotomized from the median (31 years) and the "educational level" was stratified into less than high school and high school graduate.
Associations were presented through prevalence ratios (PR) provided by the Cox regression analysis for cross-sectional studies38,39.
For the modeling, besides the main independent variable (race/skin color), the variables gender, age, and educational level were inserted, selected based on evidence in the literature. The first model presents the gross prevalence ratio (PR) for each of the two outcomes, a result of the univariate analysis. In the second model, PR is adjusted by gender and age (women and youngsters were the exposure strata)40. In the third model, the variable "educational level" was inserted –in addition to the three variables previously inserted– and the PR adjusted by the three covariables was obtained, constituting the final models.
Since the population of this study is not a random sample, the inferential approach was avoided. Such an approach is inappropriate –despite being usual– when a census or a nonrandom sample is involved. This is an analytical study in which procedures compatible with the nonrandom nature of the population investigated were adopted, as recommended by the specialized literature on the subject41-43. Thus, the multivariate analysis presents the variables with their respective measures of association (prevalence ratios), which characterize an analytical study.
The data analysis was performed using the Statistical Package for Social Sciences (SPSS) software, version 21.0.
Ethical aspects
The research project from which the study is derived was approved by the Comitê de Ética em Pesquisa of the Faculdade de Medicina da Universidade Federal da Bahia and registered under number 1621917, in accordance with Resolution number 466/12 of the Conselho Nacional de Saúde. All the survey participants provided written informed consent.
Results
A total of 1,032 workers were studied, among whom blacks predominate (47.4%; n=489), followed by mixed (42.7%; n=441) and whites (9.9%; n=102). There was lost data for race/skin color of three individuals.
The median age was 31 years and the male gender represented 78.7% (842). A significant portion of the workers did not graduate from high school (46.0%; n=490). Whites had higher educational level (62.7%), i.e. high school graduate or completed undergraduate course, followed by mixed (58.0%) and blacks (48.6%).
Table 1 shows the distribution of occupational groups according to race/skin color. Among blacks, there is a higher proportion of “urban garbage collector" (57.5%) and a lower proportion in positions that involve supervision, coordination, and administrative work (3.5%), while among whites, 23.5% are collectors and 20.6% have supervisory and administrative functions. While 17.6% of whites are urban cleaning drivers, this percentage is approximately 8.0% among blacks.
Table 2 shows the variables related to psychosocial exposure at work according to the JDCS model, stratified by race/skin color. Black individuals had higher psychological demand and lower control over work; whites had the most favorable measures –higher control and lower psychological demand. The prevalence of high demand at work was higher among blacks. Among these, the highest frequency of low social support from supervisors was observed, however, the highest frequency of high support from coworkers.
The multivariate analysis showed that compared to whites, blacks had 65% more prevalence of high-demand work and 60% more prevalence of low control at work, in models adjusted by age, gender, and educational level. High psychological demand and low social support were also more prevalent among blacks, but these associations had smaller magnitudes. Among mixed, the prevalence of low control was 36% higher than the prevalence of low control among whites (Table 3).
Table 4 shows the distribution of the items of physical demand at work, according to race/skin color. Exposure to standing work was more frequent among blacks. In the other items, there are higher frequencies of exposure among blacks, followed by mixed, and the lower frequencies are for whites.
A positive association between physical demand and race/skin color black was verified for all items. The associations were maintained after adjustment by gender, age, and educational level. Work with arms above shoulder level and in squatting posture were highlighted, besides load handling, with high exposure prevalences, at least 44% higher among blacks, when compared to whites (Table 5).
When compared to whites, mixed had higher exposures in all physical demand items, even after adjustment, however, the associations had smaller magnitudes than when comparing blacks to whites. Pushing load and working with arms above shoulder level were 25% and 48% more frequent, respectively, among blacks (Table 5).
Discussion
The findings are in line with the study hypothesis about racial inequities in occupational exposure. Black workers are subject to higher physical and psychosocial demands at work. Mainly, blacks had a lower degree of control over their work, besides being more subject to high psychological demand –and consequently to high-demand work– when compared to whites. They are also more exposed to anomalous body postures, such as working with arms above shoulder level and squatting, besides performing more load handling. Among the mixed, there have been, in general, associations of smaller magnitudes with the demands at work, when compared to whites.
Therefore, the physical work that imposes higher demands on the body, also designated as heavy physical work, has been destined mainly to blacks, who, moreover, are subject to stressful situations, with a higher time pressure and less autonomy to execute tasks.
Black workers predominated in operational occupations but were much less represented in coordination, supervision, and administrative positions, compared to whites –3.5% and 20.6%, respectively. Besides, in garbage collection teams, blacks were predominantly collectors rather than urban cleaning drivers. Although the population of this study is relatively homogeneous in terms of low social status, there are differences in the levels of control over the work according to the occupational group, as is the case of urban cleaning drivers –with a higher proportion among whites–, who have a higher degree of control than collectors/cleaners, the latter with a higher proportion among blacks28. As for this aspect, a study on the work of garbage collection teams, based on Ergonomic Work Analysis, revealed that drivers led the collection team and were responsible for determining the work pace and making decisions for the team44. Therefore, it was observed a level of control among drivers that surpasses that of the collectors.
Meyer et al.10, verified the highest scores of the JCQ for psychological demand and the lowest scores for control over work among Black workers, compared to Whites. However, the authors used fewer JCQ items than this study, making comparison difficult. There is a gap in scientific production about the ethnic distribution of occupational exposures, which limits the discussion of the results obtained based on other evidence.
However, there are findings in the literature that converge with those of the present study, including greater magnitudes of association between race/skin color and psychosocial demand, obtained among nursing professionals22. Blacks had twice as much low control and almost three times high-demand work, after adjustment by socio-demographic and occupational variables. When stratifying by performance categories, the frequency of high demand was increased among Black nursing assistants, but there were no disparities by race/skin color among the most qualified professionals.
A study found a higher prevalence of high psychological demand among yellows (54.8%), followed by whites (51.2%), mixed (49.9%), blacks (48.8%) and indigenous (41.8%) and showed that the proportion of high level of control at work was observed in the same order23. Therefore, the frequencies of high demand and high control between yellows and whites observed by the authors correspond to active work –the most favorable experience according to the JCD model–, which is related to motivation, the elevation of self-esteem, and development of new skills for work20. High psychological demand and low control at work, in turn, more frequent among blacks, configure high-demand work, consistent with our findings. Similarly, other authors24 found that the high demand was higher among non-Whites (28.1%) when compared with Whites (21.6%).
The means of psychological demand and decision authority found by Thomas et al.45 through the JCQ were similar among the racial groups studied. Other authors46 also found no differences in the psychosocial demands at work according to race/skin color, in a study with workers aged 50 years or older, however, these authors did not use the JDCS model or the JCQ, which restricts comparison of results.
The social support coming from supervisors was lower among blacks. It is known that social support coming from supervisors tends to be more protective in the face of work situations with high psychological demand, given the role of supervisors in managing and organizing work47. Thus, the low social support of supervisors can contribute to conditions of psychological and physical suffering, including musculoskeletal pain48.
On the other hand, the higher social support from coworkers among blacks may indicate greater solidarity among these workers, probably as a strategy to cope with higher exposure to adverse conditions at work. Therefore, social support is also referred to as collective control22,47. However, low social support, combining coworkers and supervisor support, predominated among blacks and mixed, compared with whites. In situations where there is high psychological demand, low control to regulate work pace and workload, and low social support, the adoption of breaks during the working day and the relieving of body postures –intended to avoid physical and mental overload– are compromised47.
Racial inequities have also been found in physical demand at work, revealing an overlap of disadvantages among Afrodescendant individuals. Work with arms above shoulder level was more frequent among them. This posture reduces the subacromial space in the shoulder and can cause disabling pain27,30,49,50. Frequently the arms above shoulder level posture and load handling occur simultaneously27,49, as when collectors throw loads into garbage trucks44,51.
Cargo handling and squatting, which have been associated with musculoskeletal symptoms30,31,50, occurred in greater proportions among Afrodescendants –mixed and blacks– and may be related to tasks performed by urban garbage collectors.
While lifting materials is mainly associated with back pain, tasks involving pushing and pulling are often related to pain in the cervical region and also to shoulder pain30,50. By simulating in the laboratory tasks performed by urban garbage collectors, Schibye et al.26 found increased compressive and shearing forces on the vertebral column when lifting materials and when pushing and pulling containers on two wheels.
High exposure to the handling of materials or to anomalous postures at work was associated with long term absenteeism52 and a 50% increase in pain at multiple body sites in the previous seven days33. This higher morbidity is related to a higher prevalence of limitations to work33.
There was a lower proportion of blacks in coordination, supervision, or administrative positions, which generally favor the adoption of diversified body postures and little cargo handling. On the other hand, high exposure of blacks to sustained standing work –which is strenuous– can result in lower back pain and lower limb edema29. The requirement of adopting a standing posture to perform tasks is described in the sectors of cutting, assembly, and finishing of footwear manufacturing, sometimes associated with rotation of the trunk53.
The higher frequencies of cargo handling and anomalous postures between blacks and mixed raise the discussion about the slavery period, when bodies of Africans and Afrodescendants were used for cargo transportation. Thus, heavy physical work can represent the perpetuation of racism that is made explicit by inequality of exposure at work6,14.
In line with the literature54, the exposure gradients found among blacks and mixed suggest that individuals with more pronounced Afrodescendant characteristics –such as darker skin tone–, probably self-reported as blacks, are more vulnerable than those with lighter skin colors. These, however, remain at a disadvantage compared to whites. Therefore, the gradients expressed in this study for the prevalence ratios could be higher if it were not for miscegenation, which is characteristic among the studied population, especially in the city of Salvador. At the same time, the narrative of miscegenation as a mechanism of racial democratization has contributed to the concealment of inequities6.
The lower educational level among Afrodescendants, observed in the characterization of the study population, is possibly one of the manifestations of structural racism, which includes the naturalization of multigenerational inequities of economic status and reflects the past of slavery, although it is not the exclusive and irreversible consequence of this history6,14. A low educational level was found in the studied population. The association between race/skin color and demand at work –physical or psychosocial– was adjusted by the educational level, as can be seen in the analysis models. Adjustments were also made by gender and age. The findings show that black workers, therefore, are destined to jobs that involve higher physical and psychological overload, compared to white workers of the same educational level, gender, and age, ratifying the role of race/skin color in occupational exposure.
This study contributes to the scientific literature by presenting a little-explored approach: the racial inequities in the distribution of exposure to physical and psychosocial demands, which are important modalities of occupational exposure.
In the world of work in the private sector, studying 1,032 workers from the industrial and services sectors performing in-person data collection represents the overcoming of a barrier in terms of the access of researchers to this population. Epidemiological studies in Workers\' Health have often been restricted to populations of public servants or the general public, in which an attempt is made to specify "occupation" and the contributions of work, with all the limitations of validity that are known in such circumstances. Moreover, the variability in the study population regarding occupations has favored the variability of race/skin color.
The procedures adopted may represent strong points of the study: guaranteeing privacy and confidentiality to collect data in the workplace, considering the possibility of conflicts of interest between workers and employers; the independence of the researchers to companies; and data collection through interviews conducted by researchers affiliated to a public education institution, which made it possible to minimize information bias and increase the response rate. Also, the selection strategy of the participants contributed to reducing the possibility of healthy worker survival bias.
The race/skin color variable may present classification errors and transience over time. Besides, it may convey conceptual inaccuracy, for health studies17-20. However, the epidemiological analysis was performed in line with the use of this variable that is recommended in the literature. The methodological decision of excluding yellows and indigenous was justified based on the number of individuals from these ethnic groups in the studied population. However, it is worth highlighting the gap in the literature regarding occupational exposure in these ethnic groups and the precarious participation of indigenous people, especially in the labor market9.
Conclusion
The race/skin color of workers is a marker of inequities in the world of work, evidenced in this study by the greater exposure to psychosocial and physical demand among blacks and mixed, with a gradient of exposure between them, when compared to whites.
Among black workers, were highlighted the highest exposure to high psychosocial demand, as well as to heavy physical work requiring the adoption of anomalous postures, besides load handling. The mixed workers also presented low control over the work and physical exposure to anomalous postures, mainly.
The evidence presented, added to the available literature, may contribute to the social debate about the repercussions of structural racism in the world of work, as well as to the formulation of antiracist strategies by companies focusing on the restructuring of work environments and processes, aiming at racial equity. Interventions to minimize the exposure of all workers to psychosocial and physical demands and those directed to selection processes and mobility mechanisms in the organization are some possibilities.

References

1. Whitehead M. The concepts and principles of equity and health. Health Promot Int. 1991;6(3):217–29.

2. Silva JB, Barros MBA. Epidemiologia e desigualdade: notas sobre a teoria e a história. Rev Panam Salud Pública [Internet]. 2002;12(6):375–83.

3. Barreto ML. Desigualdades em Saúde: uma perspectiva global. Ciênc. Saúde Coletiva. 2017;22(07):2097–108.
4. Bailey Z, Krieger N, Agénor M, Graves J, Linos N, Bassett MT. Structural racism and health inequities in the USA: evidence and interventions. Lancet [Internet]. 2017; 389:1453–63.

5. Williams DR, Lawrence JA, Davis BA. Racism and Health: Evidence and Needed Research. Annu. Rev. Public Health 2019. 40:14.1–14.21.

6. Almeida S. O que é racismo estrutural. Letramento. Belo Horizonte; 2018. 203p. ISBN:9788595300972

7. Osório R. O Sistema classificatório de “cor ou raça” do IBGE [Internet]. Brasília-DF: Instituto de Pesquisa Econômica Aplicada-IPEA; 2003. p. 50. Disponível em: http://www.ipea.gov.br/portal/images/stories/PDFs/Tds/td_0996.pdf. Acesso em 06/12/2018.
8. Brasil. Instituto de Pesquisa e Economia Aplicada - IPEA. Retrato das desigualdades de gênero e raça. 4 ed. [Internet]. 2011. 39p. Disponível em: http://www.ipea.gov.br/retrato/. Acesso em 09/02/2019.
9. IBGE. Desigualdades sociais por cor ou raça no Brasil. 323.12(81)-Complemento 2:v.41. Rio de Janeiro, 2019. 12p.ISBN: 9788524045134.

10. Meyer JD. Race-based job discrimination, disparities in job control, and their joint effects on health. Am J Ind Med. 2014; 57:587–95.

11. Steege AL, Baron SL, Marsh SM, Menéndez CC; Myers JR. Examining Occupational Health and Safety Disparities Using National Data: A Cause for Continuing Concern. Am J Ind Med. 2014;57(5):527–38.
12. Chung-Bridges K, Muntaner C, Fleming LE, Lee DJ, Arheart KL, LeBlanc WG et al. Occupational Segregation as a Determinant of US Worker Health. Am J Ind Med. 2008; 51:555–67.

13. Conselho Nacional de Justiça. Censo do Poder Judiciário: Vetores iniciais e dados estatísticos. Conselho Nacional de Justiça; Brasília, 2014, 213p. Disponível em:http://cnj.jus.br/images/dpj/CensoJudiciario.final.pdf. Acesso em 09/02/2019.

14. Batista WM, Mastrodi J. Dos fundamentos extraeconômicos do racismo no Brasil. Rev Direito e Práxis. 2018;9(4):2332–59.

15. Cooper R. A note on the biologic concept of race and its application in epidemiologic research. Am Heart J. 1984;108(3 PART 2):715–23.

16. Chor D, Lima CRA. Aspectos epidemiológicos das desigualdades raciais em saúde no Brasil. Cad. Saúde Pública. 2005;21(5):1586–94.

17. Araújo EM, Nascimento C, Hogan VK. A utilização da variável raça/cor em Saúde Pública: possibilidades e limites. Interface-Comunic., Saúde, Educ. 2009;13(31):383–94.

18. Kabad JF, Bastos JL, Santos RV. Raça, cor e etnia em estudos epidemiológicos sobre populações brasileiras: revisão sistemática na base PubMed. PhysisRev Saúde Coletiva. 2012;22(3):895–918.

19. Hicken MT. Measurement and modeling of race and health in Brazil: continuing the discussion. Cad SaúdePública. 2017;33(suppl 1):1–5.

20. Karasek R. Job demands, job decision latitude, and mental strain. Implications for job redesign. AdmSci Q. 1979; 24:285–308.

21. Johnson JV, Hall EM, Theorell T. Combined effects of job strain and social isolation on cardiovascular disease morbidity and mortality in a random sample of the Swedish male working population. Scand J WorkEnviron Health. 1989;15(4):271–9.

22. Hurtado DA, Sabbath EL, Ertel KA, Buxton OM, Berkman LF. Racial disparities in job strain among American and immigrant long-term care workers. IntNurs Rev. 2012;59(2):237–44.

23. Juvanhol LL, Melo ECP, Chor D, Fonseca MJM, Rotenberg L, Bastos LS et al. Association between demand-control model components and blood pressure in the ELSA-Brasil study: exploring heterogeneity using quantile regression analyses. Scand J Work Environ Health. 2018;44(6):601–12.

24. Lopes SV, Silva MC. Estresse ocupacional e fatores associados em servidores públicos de uma universidade federal do sul do Brasil. Ciênc Saúde Colet. 2018;23(11):3869–80.

25. Stock SR, Fernandes R, Delisle A, Vézina N. Reproducibility and validity of workers’ self-reports of physical work demands. Scand J WorkEnviron Health. 2005;31(06):409–37.

26. Schibye B, Sùgaard K, Martinsen D, Klausen K. Mechanical load on the low back and shoulders during pushing and pulling of two-wheeled waste containers compared with lifting and carrying of bags and bins. ClinicalBiomechanics. 2001;16:549–59.

27. Pope DP, Silman AJ, Cherry NM, Pritchard C, Macfarlane GJ. Association of occupational physical demands and psychosocial working environment with disabling shoulder pain. Ann RheumDis. 2001; 60:852–8.

28. Pataro SMS, Fernandes RCP. Trabalho físico pesado e dor lombar: a realidade na limpeza urbana. RevBrasEpidemiol. 2014;17(1):17-30.

29. Antle DM, Cormier L, Findlay M, Miller LL, Côté JN. Lower limb blood flow and mean arterial pressure during standing and seated work: Implications for workplace posture recommendations. Prev Med Reports. 2018; 10:117–22.
30. Jakobsen M, Sundstrup E, Brandt M, Persson R, Andersen L. Estimation of physical workload of the low-back based on exposure variation analysis during a full working day among male blue-collar workers. Cross-sectionalworkplacestudy. AplliedErgon. 2018; 70:127–33.
31. Allen K, Chen J, Callahan L, Golightly Y, Helmick C, Renner J et al. Racial Differences in Knee Osteoarthritis Pain: Potential Contribution of Occupational and Household Tasks. J Rheumatol. 2014;39(2):337–44.

32. Fernandes RCP, Pataro SMS, Carvalho RB, Burdorf A. The concurrence of musculoskeletal pain and associated work-related factors: A cross sectional study. BMC Public Health. 2016;16(1):1-9.

33. Fernandes RCP, Burdorf A. Associations of multisite pain with healthcare utilization, sickness absence and restrictions at work. IntArchOccupEnviron Health. 2016; 89:1039–46.

34. Karasek R., Gordon G, Piotrowski C. The Job Content Instrument: Questionnaire and User’s Guide. 1986.

35. Araújo TM, Karasek R. Validity and reliability of the job content questionnaire in formal and informal jobs in Brazil. Scand J WorkEnviron Health. 2008; Suppl. (6):52–9.

36. Santos KOB, Araújo TM, Carvalho FM, Karasek R. The job content questionnaire in various occupational contexts: Applying a latent class model. BMJ Open. 2017;7(5):1–8.

37. Fernandes RCP, Cunha LP, Lima, VC; Santos KOB. Mensurando a demanda física no trabalho: estrutura fatorial e confiabilidade de itens sobre posturas, manuseio de carga e repetitividade. Cad Saúde Pública. 2019;35(1):1–14.

38. Barros AJD, Hirakata VN. Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio. BMC Med Res Methodol [Internet]. 2003;3(21):1–13.

39. Coutinho LMS, Scazufca M, Menezes PR. Métodos para estimar razão de prevalência em estudos de corte transversal. Rev Saúde Pública. 2008;42(6):992–8.

40. LaMontagne AD, Krnjacki L, Kavanagh AM, Bentley R. Psychosocial working conditions in a representative sample of working Australians 2001-2008: An analysis of changes in inequalities over time. OccupEnviron Med. 2013;70(9):639–47.

41. Greenland S. Randomization, Statistics, and Causal Inference. Epidemiology. 1990;6(1):421–9.

42. Hahn GJ, Meeker WQ. Assumptions for Statistical Inference. The American Statistician.1993;47(1):1-11.
43. Daniel, WW, Cross, CL. Biostatistics : a foundation for analysis in the health sciences / Tenth edition.2013. John Wiley: USA.

44. Camada IMDO, Pataro SMS, Fernandes RCP. Heavy physical work under time pressure: The garbage collection service-a case study. Work. 2012;41(SUPPL.1):462–9.

45. Thomas KMS, Nelesen RA, Ziegler MG, Bardwell WA, Dimsdale JE. Job strain, ethnicity, and sympathetic nervous system activity. Hypertension. 2004;44(6):891-6.

46. Mezuk B, Ratliff S. Job Strain, workplace discrimination, and hypertension among older workers: the health and retirement study.RaceSoc Probl. 2011;3(1):38–50.

47. Luchman JN, González-Morales MG. Demands, control, and support: A meta-analytic review of work characteristics interrelationships. J Occup Health Psychol. 2013;18(1):37–52.

48. Carvalho RLRB, Fernandes RCP. Demandas psicológicas, baixo apoio social e repetitividade: fatores ocupacionais associados à dor musculoesquelética de trabalhadores da indústria de calçados. RevBrasSaudeOcup. 2019;44(6):1-11.

49. Kratzenstein S, Wanstrath M, Behrenbruch K. Height adjustments on backpack-carrying systems and muscle activity. ApplErgon. 2019;74:172–6.

50. Argubi-Wollensen A, Wollensen B, Leitner M, MattesK. Human Body Mechanics of Pusshing pulling: Analyzing the factors of task-related Strain on the musculoeskeletal System. Saf Health Work. 2017:8(1):11-18.

51. Ziaei M, Choobineh A, Abdoli-Eramaki M, Ghaem H, Jaberi O. Psychological and physical job demands, decision latitude, and work-related social support among Iranian waste collectors. Waste Management. 2019; 95:377–87.

52. Andersen LL, Fallentin N, Thorsen SV, Holtermann A. Physical workload and risk of long-term sickness absence in the general working population and among blue-collar workers: Prospective cohort study with register follow-up. OccupEnviron Med. 2016;73(4):246–53.

53. Lourinho MG, Negreiros GR, Almeida, LB, Vieira ER, Quemelo PRV. Riscos de lesão musculoesquelética em diferentes setores de uma empresa calçadista. Fisioterapia e Pesquisa, São Paulo,2011, 18(3): p. 252-7, jul/set. 2011.
54. Dixon AR, Telles EE. Skin Color and Colorism: Global Research, Concepts, and Measurement. Annu. Rev. Sociol. 2017. 43:405–24.