Sick and Tired of Being Sick and Tired: Scientific Evidence, Methods, and Research Implications for Racial and Ethnic Disparities in Occupational Health

Abstract

The extent of racial/ethnic disparities in occupational health have not been well studied. The author reviews the evidence about workers of color and occupational injuries and disease.

Patterns of employment in the U.S. workforce according to education, gender, and race/ethnicity are discussed, and how these patterns might cause disproportionate exposure leading to disproportionate disease and injury. Methodological issues are explored that have hampered research about occupational health disparities, and future research needs are identified.

OCCUPATIONAL DISEASE and injury are significant, if underappreciated, public health problems in the United States and worldwide. Far too many US workers die on the job, despite the fact that the rate for occupational fatalities has fallen from 18 per 100 000 workers in 1970, when the Occupational Safety and Health Act was passed, to 4.3 per 100 000 workers in 2000.1 The Bureau of Labor Statistics estimates that there were 5915 traumatic occupational fatalities in 2000, and over 6 million work-related injuries and illnesses were reported in the private and public sectors.1 This means that for each day of 2000 more than 16 workers died on the job, and some 15 600 were injured or became ill in the private sector. There were 362 500 newly reported cases of occupational illnesses in private industry in 2000.2 Estimates for annual deaths from work-related diseases is between 50 000 and 60 000 workers per year. The risk of injury and death is not evenly distributed. For example, although less than 3% of the labor force works on farms, farm workers have one of highest fatality rates, at 20 per 100 000 workers in 2000.1

Do occupational health disparities exist by class, race, and ethnicity? If so, what is their cause? Are there differential exposures to occupational and environmental toxicants by class, race, and ethnicity? Are there differences in genetic susceptibility to occupational exposures? To what extent do other factors such as stress, lack of access to medical care, and poor nutrition result in disparate work-related health outcomes?

PEOPLE OF COLOR IN THE US WORKFORCE

The US workplace is a product of our social, economic, and cultural history, and therefore is littered with gender, racial, and ethnic “job ghettos.” Workers of color generally are underrepresented in professional categories and overrepresented in bluecollar and service jobs, especially in certain occupations. For example, in 1996 50% of all garbage collectors, over 33% of all elevator operators, and 33% of all nursing aides and orderlies were Black. Similarly, more that three-fourths of all miscellaneous woodworkers, 68% of all farm product graders and sorters, 37% of all farmworkers, and 34% of all fabric machine operators were Latino. Meanwhile, 97% of all dental hygienists were White.3

In 1999 almost 12% of the population (32.3 million people) lived at or below the official poverty level; of these, some 6.8 million were classified as “working poor,” people who spent at least 27 weeks working or looking for work and whose incomes are below the poverty level. The majority (64%) of the working poor were full-time workers. Blacks (10.2%) and Latinos (10.7%) were more likely than Whites (4.3%) to be classified as working poor. The rate of working poor was highest among Black women (13.6%) and was almost twice that of Black men (6.2%). Occupations with high proportions of workers in poverty included industries with seasonal patterns (forestry, fishing, farming) and operators, fabricators, and laborers.4 Although labor trends vary with economic cycles, the fact remains that Black and Latino workers lag behind White workers in the desirability of and prestige and compensation from their work.5

CLASSIC CASE STUDIES

Many classic case studies have documented differential exposure to work-related toxicants resulting in disproportionately high rates of occupational diseases among miners, steelworkers, chemical-industry workers, rubber and textile workers, and others. These studies have been reviewed elsewhere.3,6–8

During the Great Depression, Gauley Bridge in West Virginia was the site of the worst industrial disaster, with the greatest death toll, in the history of the United States.9 Although Union Carbide was well aware of the possible health consequences of digging through a mountain of almost pure silica, they ignored the safer method of wet-drilling, which was widely used in Europe, and instead performed dry-drilling operations. Despite the desperation to find work during the Depression and the fact that the local population surrounding the tunnel was 80% White, it is remarkable that the workers hired to dig inside the tunnel—where is was dustiest—were 80% Black.10 Hundreds died of acute silicosis within weeks to months of working in the tunnel. Of the estimated 700 workers who died between 1930 and 1935, some 76% were Black. In this case the cause seems obvious—Black workers were placed in the dirtiest, least desirable part of the work process.

Similar cases existed in the chromate industry. The dustiest “dry end” of the process with the highest exposure to chromate carcinogens employed 41% of the Black workers in the plant and only 16% of the White workers. Although chromate workers had 29 times the rate of respiratory cancers as the general population, Black chromate workers were shown to have a significantly higher 80-fold excess.11

Lloyd’s classic epidemiological study of US steelworkers12 clearly showed that a disproportionate number of Black workers worked in the coke oven department, where exposure to the powerful coke oven fumes and carcinogens was the highest. Furthermore, some 21% of all Black steelworkers worked on top of the ovens—where the heat could burn through a pair of boots in a few days—compared to only 8% of White steelworkers. This meant that Blacks made up 74% of the workers who toiled on top of the coke ovens. Consequently, the higher rates of lung cancer among Black steelworkers were best explained not by genetic differences, as many suggested, but rather by the fact that Black steelworkers were 5 times more likely than Whites to work on the top side of coke ovens, and full-time top-side coke oven workers with at least 5 years’ exposure had a 10-fold increased risk for lung cancer.12,13

Other subpopulations also face disproportionate occupational health risks. A small East Coast manufacturing plant that produced urethane-coated waterproof fabric employed predominately monolingual Spanish-speaking Puerto Ricans. The plant was small, with dangerous working conditions, and the workers—who were not fluent in English and therefore unaware of their rights and pertinent safety information—were encouraged to work many hours of overtime. Physicians were able to document an outbreak of occupational liver disease from exposure to the solvent dimethylformamide.14 When the union was able to implement recommendations from health and safety experts, hepatotoxicity among the workers stopped.14

In the uranium mines of New Mexico, American Indians were more likely than Whites or Mexican Americans to work in underground jobs. Native Americans had higher rates of x-ray changes and restrictive lung disease than Whites or Mexican Americans.15 Whereas agricultural workers represent only a small fraction of Latino workers in the United States, Latinos make up 71% of all seasonal farmworkers and 95% of all migrant farmworkers; 56% of all Latino agricultural workers were involved in migrant farm work.16,17

A registry of adult blood lead in California showed that Latinos made up 46% of those with levels above 60 μg/dL,18 the level at which OSHA regulations require that workers must be removed from the workplace (although biological effects occur at much lower levels). The industries most frequently reported in the registry and at highest risk for lead exposure were small industries employing a disproportionate number of Latino workers (battery reclamation plants, radiator repair shops, brass and copper foundries).18

Newly immigrated Asian women make up 85% of garment workers in San Francisco, and are disproportionately represented in garment sweatshops in New York, Los Angeles, and other cities.8

DISPROPORTIONATE EXPOSURE, DISPROPORTIONATE DISEASE

Unemployment rates fluctuate with economic cycles, but have historically been higher for workers of color, particularly Blacks. Even as unemployment rates fall, the rate for Blacks remains twice as high as that for Whites. People in poor health are more likely to be unemployed. However, unemployment (or the threat of unemployment) itself creates stressors that pose health risks including elevated blood pressure and excess morbidity and mortality. The absence of employment turns out to be an important work-related hazard that should not be forgotten.3,19

Workers of color do not uniformly show increased rates for all occupational diseases. For example, Blacks have lower mortality rates for asbestosis and coal worker’s pneumoconiosis but higher mortality rates for byssinosis and silicosis.20 One study looking at responses to the National Health Interview Survey from 1983 to 1987 found similar rates of self-reported occupational injury between Blacks and Whites, and that Blacks reported fewer injuries that restricted regular activities or required medical attention.21

There have been a few attempts to look at the probabilities of exposure to occupational hazards over the entire workforce. Lucas, using the 1967 Survey of Economic Opportunity and an estimated probability of exposure to hazard for US Census occupational classifications, found that Black men had a 25% greater chance of exposure to at least 1 hazard, and Black women a 93% greater chance of exposure to at least 1 hazard, than White men and women, respectively.22 By assuming that white-collar workers in all industries have the same injury rate as in the finance industry and by considering the distribution of workers in white- and blue-collar occupations, Kottlechuck estimated that Blacks have a 37% greater risk of injury and a 24% greater risk of occupational death than Whites.23 Robinson also looked at a number of large economic databases to estimate exposure of workers to occupational hazards.24,25 Using the Current Population Survey for 1968 to 1986 and the Panel Study of Income Dynamics for 1971 to 1984, he found racial differences in rates of disabling occupational injuries. Robinson demonstrated a narrowing of disparities in exposure to occupational hazards since the 1960s for Black men compared to White men; however, he did not find a similar decrease for Black women. The excess risk for injury for Black men relative to White men declined by 50%, but increased by 20% for Black women between 1968 and 1986.24,25 Using California data, he demonstrated that Black and Hispanic workers face higher risks of occupational illness and injury even after controlling for education and work experience.25 A study of North Carolina occupational fatalities showed that Blacks were more likely to die on the job, with much of the difference being attributed to differential exposure to hazards because of job ghettos.26

Some smaller studies also looked at differential rates of occupational disease and injury. The New Jersey Department of Health’s analysis of their fatal occupational injury surveillance registry from 1983 to 1989 found differences in fatality rates among construction workers.7 The annual fatality rate for US-born White workers was 10.6 per 100 000 employees, compared to 34.8 per 100 000 employees for non-US-born Latino workers and 10.6 per 100 000 employees for Black workers (no rates were reported for US-born Latinos).

METHODOLOGICAL ISSUES

Occupational disease and injury are poorly understood for a variety of reasons. For one, there has historically been a reluctance to dedicate appropriate resources to document occupational disease and injury, because the discovery of hazards and proposed remedies have the potential to adversely affect the profit margins of business. For another, although the Bureau of Labor Statistics sends out an annual survey to a sample of workplaces to derive estimates of the number of traumatic injuries, there are no reliable data for the racial and ethnic makeup of the workforce. The United States does not have an adequate national system to monitor occupational disease and injury, and therefore lacks adequate data sources for tracking occupational diseases among workers in the United States.

Additionally there are a number of scientific challenges. US physicians and other health care professionals receive little or no training in environmental and occupational health. Occupational disease and injury do not (except in rare instances) differ in their biological presentation from disease and injury from nonoccupational bases. Occupational disease often has a long latency period spanning decades between exposure and appearance of clinical disease. We still lack the clinical tools with appropriate sensitivity and specificity to diagnose early biological changes from most occupational exposures.

There is also a wide variability among individuals in biological susceptibility to occupational exposures. The exact toxicants and doses to which workers are exposed are usually unknown and poorly estimated. The human toxicity of chemicals has not been well studied. Of at least 60 000 commonly used commercial chemicals, only about 10 000 have been tested in animals.1 One toxicant may affect the body in multiple ways, causing several different disease processes. At the same time, workers are usually exposed to multiple toxic agents in a single workplace, with interactions of multiple exposures that are poorly understood.

STRATEGIES TO DECREASE INJURY AND DISEASE

The literature shows that minority workers, particularly Blacks, have in the past suffered a disproportionate burden of occupational disease and injury in key industries. What is less clear is the extent to which these disparities continue today. Despite a lack of clarity on the extent of present disparities, there are a number of strategies that can be used to decrease occupational injuries and disease among workers of color.

Unfortunately, the observation of J. Robinson remains true today: “The findings suggest that black workers with the same levels of education and experience as whites will, on average, find themselves in substantially more dangerous occupations.”27 The elimination of job discrimination and job ghettos based on gender, race, and ethnicity is critical to the broad issues of social justice and will clearly impact the problem of disproportionate exposure.

The best strategy for eliminating risks would be simply to eliminate the most hazardous substances from the workplace. Where substitution with a less hazardous substance is not possible, then workers must be kept from the exposure, preferably through engineering controls. The least desirable (and usually the least effective) method is personal protective equipment to keep workers away from hazardous materials.

Most physicians and other health professionals know little about occupational diseases and injuries; primary care providers need education in these areas. In addition, more occupational safety and health professionals (industrial hygienists, nurses, physicians, ergonomists) are needed. Most important, as far as education goes, workers of all types need education and training as well. Workers should know about potential hazards in their workplace and how to avoid injury and disease. These educational programs should be carried with culturally sensitive and linguistically specific methods. The settings for such programs must expand beyond the usual workplace to include churches, community settings, popular radio, and television programs.28,29 Workers should also have easy access to clinical occupational health services, something most workers do not currently enjoy.

Federal and particularly state and local public health agencies do not have adequate resources to inspect and monitor workplaces. A national system of regional occupational and environmental health centers would coordinate the efforts of all 3 levels of government. Such centers would enable the close monitoring of new processes and industries, and research on hazards and prevention strategies could be more quickly implemented. These centers could also coordinate efforts within the public sector and across private industry.

Workplaces and occupations that are the most hazardous should be given highest priority for hazard control, training, and clinical services. Such prioritization might help minority workers in the most dangerous occupations reduce the risk of injury.

Finally, and most importantly, the best way to guarantee safer workplaces is by ensuring that workers have the right to organize unions. The fact that less than 15% of all US workers over age 16 belong to unions hampers effective legislation and regulations needed to guarantee safe workplaces.

RESEARCH AND SURVEILLANCE AGENDA

More research is needed on many issues in occupational health and safety, and the health status of workers of color should be given a high priority. This research should use qualitative and ethnographic methods to examine perceived risk, evaluate the effectiveness of interventions, and conduct quantitative data collection and statistical analysis.

The National Institute for Occupational Safety and Health has established research priority areas under its National Occupational Research Agenda (NORA). NORA includes a Special Populations at Risk Team that focuses on workers of color, older workers, disabled workers, agricultural workers, and child labor. NORA has resulted in increased funding in focused areas of concern.

Table 1 ▶ summarizes the availability of data for occupational fatalities. Even though these are the easiest data to collect, the gaps are remarkable. Oversampling of minority workers and pooling of data are just 2 approaches that can be used to address the instability of small numbers in studies.

TABLE 1.

—Data Availability for Occupational Fatalities: Total Deaths per 100 000 Workers Over Age 16 Years, United States, 1998.

	All Industries	Mining	Construction	Transportation	Agriculture
	4.5	23.6	14.6	11.8	24.1
Race/ethnicity
American Indian	DSU	DSU	DSU	DSU	DSU
Asian	DNC	DNC	DNC	DNC	DNC
Native Hawaiian/other Pacific Islanders	DNC	DNC	DNC	DNC	DNC
Black	3.9	DNA	DNA	DNA	DNA
White	4.5	DNA	DNA	DNA	DNA
Latino	5.2	DNA	DNA	DNA	DNA
Non-Hispanic Black	DNA	DNA	DNA	DNA	DNA
Non-Hispanic White	DNA	DNA	DNA	DNA	DNA
Sex
Women	0.8	DNA	DNA	DNA	DNA
Men	7.7	DNA	DNA	DNA	DNA
Family income levela	DNC	DNC	DNC	DNC	DNC
Poor	DNC	DNC	DNC	DNC	DNC
Near poor	DNC	DNC	DNC	DNC	DNC
Middle/high income	DNC	DNC	DNC	DNC	DNC
Disability status
With disability	DNC	DNC	DNC	DNC	DNC
Without disability	DNC	DNC	DNC	DNC	DNC

Note. DNA = data not analyzed; DNC = data not collected; DSU = data statistically unreliable.

^aPoor = < 100% Federal Poverty Level (FPL); Near poor = 100–199% FPL; Middle/high income = ≥ 200% FPL.

Source: Healthy People 2010: Understanding and Improving Health. Washington, DC: US Department of Health and Human Services; 2001.

We must improve our data infrastructure by routinely collecting several kinds of data. Race and ethnicity data should be collected with occupational titles. This would allow the calculation of race- and ethnicity-specific rates of injury and disease across occupations. The size of plants, number of shifts, and whether a plant is unionized also should be tracked, because these have been shown to be risk factors for certain types of exposures.

Exposure data should be collected and tracked. This would help reduce the difficulty of appropriately diagnosing occupational and environmental diseases. We need to support the development of systematic and reliable data sources regarding occupational disease and injury. The establishment of true active surveillance systems in occupational health should be a national priority. Such systems must include interventions to alter exposures and decrease or eliminate hazards. It is not acceptable to wait decades while the body count rises in order to recognize a new occupational hazard.

National surveys and surveillance systems should include information about class and occupation as well as race and ethnicity. This would allow the development of hypotheses about the interactions between class, race, and ethnicity. Occupation and known workplace exposures should become a routine component of databases and disease registries for conditions such as cancer, birth defects, and other measures of health outcomes. We must resist engaging in research in separate silos (e.g., class, race, exposure-related attributes) and look at interactions between pathways. The following question should guide all health research: What might be the role of occupation and occupational hazards?

Sexton and colleagues30 proposed a conceptual model for generating hypotheses about race and ethnicity and environmental exposures (Figure 1 ▶). Much of the research in occupational health has historically been centered around the question of how exposure and individual variability affects health risk. By focusing our analysis only at the individual level we will remain unable to understand the complex social issues involved. Race and ethnicity are frequently surrogates for complicated social categories. By looking carefully at class, race, and ethnicity, we can begin to generate possible hypotheses about their effects on health risk. The effects of class, race, and ethnicity can be mediated through 2 broad categories: (1) exposure-related attributes and (2) individual-related attributes. The individual attributes might include differences in diet, gender, and genetic makeup. They might also include increased health risks because of underlying diseases that may or may not be associated with working conditions (hypertension, diabetes). The exposure-related attributes involve discrimination in the workplace that may result in job ghettos and disproportionate exposure to certain hazards. Studies should be designed to address the questions of how class, race, and ethnicity differentially effect occupational health risks, and how class and race/ethnicity affect exposure- and susceptibility-related attributes. Such studies might help us unravel how socioeconomic factors, race, and ethnicity contribute to occupational health injury and disease.

FIGURE 1.

—Conceptual model for generating testable hypotheses about causal relationships between demographic variables and environmental/occupational exposures.