METHODS

The proportion of the civilian, employed labour force in selected white collar occupations (that is, managerial, professional speciality, and technical occupations) was used to represent the occupational structure of counties in New York state.⁴ Information on the proportion of each county labour force in these selected white collar occupations was obtained from the US census in 1990. The 62 counties were ranked by the percentage of the labour force in white collar occupations, and the range between the first and the 99th centiles of the distribution was divided into three occupational structure categories of equal ranges of percentages.^1,3,4 Counties with percentages of white collar workers below the first and above the 99th centiles of the distribution were included in categories I and III, respectively. Construction of the occupational structure categories based on equal divisions of the range of percentage of managerial/ professional workers permits extrapolation of an observed pattern to the continuous variation in the occupational structure variable. The correlation between the 1980 and 1990 distributions is 0.90, which indicates that the relative position of counties in the distribution was quite stable.

Two social classes, white collar and blue collar, were defined based on the “usual lifetime” occupation listed on the death certificate of decedents. As occupation on the death certificate was missing or described as “housewife” for 49% of women decedents, only men were included in this analysis, which is a limitation of this approach described previously.¹² Furthermore, because occupation narratives from the death certificate of residents of New York City are not coded or computerised, this analysis was restricted to New York residents other than the five counties of New York City, which is commonly referred to as “upstate New York.” The percentage of black decedents who were missing occupation information was 7% in category I, 18% in category II, and 15% in category III. The percentages for white decedents were 10%, 12%, 18%, for categories I, II, and III, respectively.

Mortality rates were calculated for black men and white men, aged 35–64 years; this age group represents premature mortality.¹³ Deaths with underlying causes assigned to ICD-9 codes¹⁴ 410–414, 402, 429.2 were included as CHD deaths. Denominator data describing white collar and blue collar population counts at the county level were obtained by special request from the Bureau of Census, using the Public Use Microdata Sample (PUMS 5%) for New York state in 1990. County level counts permitted the calculation of mortality rates simultaneously by social class and occupational structure category. Census data describe “current” or “most recent” occupation compared with “usual lifetime” occupation on the death certificate. However, relative stability over the course of the lifetime for the broad categories of white collar and blue collar was reported by a previous study.¹² Therefore, census population counts may be used as estimates of the population at risk for calculation of mortality rates among white collar and blue collar workers.

CHD deaths and population counts for five year age groups, from 35–64 years, were summed within social class, race, and occupational structure categories, for 1988–92. Average annual age adjusted CHD mortality rates were calculated by social class, race, and occupational structure category for 1988–92, by the direct method using the 1970 US population as the standard. A mortality rate is not presented for white collar black men in occupational structure category I because of the small population size for this group (n=1264) and too few deaths (n=4) to generate a reliable mortality rate.¹⁵ As data for the calculation of rates represent complete death counts and as upstate New York counties were categorised rather than randomised into occupational structure categories, there is no basis for observed associations to be attributable to chance resulting from randomisation. Therefore, statistical testing of differences between categories was not conducted.^16–18

Data on urbanicity, income, unemployment, house values, education, and labour force participation were obtained from the 1990 US Census of Population and Housing.¹⁹ Population weighted averages of county data were calculated for each category of occupational structure. Information on earnings by industry were obtained from the 1990 Census City and County Data Book.

Income inequality was calculated, for each category of occupational structure using the “Robin Hood Index”.²⁰ The income inequality index represents the share of total income that would need to be transferred from those above the mean level of income to those below the mean, and therefore achieve equality in the distribution of income within each occupational structure category.

Residential racial segregation was calculated by a previously published method,²¹ and this same approach was used to calculate residential class segregation. County residential race segregation indices were population weighted by the black population and combined into an overall residential race segregation index for each occupational structure category. The residential race segregation index represents the percentage of the black population that would need to change their census tract of residence so that all tracts reflect the overall percentage of black population in the county. Residential class segregation was calculated based on the percentage of the population in selected white collar occupations (that is, managerial, professional speciality, and technical occupations). County residential class segregation indices were population weighted by the county population and combined into an overall residential class segregation index for each occupational structure category. The residential class segregation index represents the percentage of “upper” white collar workers that would need to change their census tract of residence so that all tracts reflect the overall percentage of “upper” white collar workers in the county.

RESULTS

In 1990, the proportion of the labour force in managerial/professional occupations in upstate New York counties ranged from 20% to 50% (table 1). Category I represented counties with the smallest proportions of managerial/professional workers, 20%–29%, and category III represented counties with the largest proportion of managerial/professional workers, 40%–50%. However, percentages of black workers in managerial/ professional jobs ranged only between 20% and 28% in upstate New York in 1990. Category I represented 25% of the population of upstate New York and 14% of residents in these counties lived in urban areas, compared with 92% of category III residents. The proportion of the population 25 years and older holding a (bachelors or higher) college degree was two times greater in category III (35%) compared with category I counties (14%) (data not shown). Furthermore, there were over two times the percentage of college educated among white compared with black men in all occupational structure categories. Labour force statistics reveal decreasing unemployment with increasing occupational structure. However, at least two times the unemployment among black men compared with white men in all occupational structure categories is shown.

Home values are an indicator of community economic resources, as a local tax base, which supports services such as schools and the local health department, is primarily based on property values. Median home values were 4.5 times the value in category III compared with category I. This is also related to average annual per capita income, which was two times the level in category III compared with category I, and per capita income from capital investment resources (stocks, dividends, rent) was three times greater in category III compared with category I counties.

Figure 1 shows an inverse association between CHD mortality and occupational structure among blue collar and white collar workers, among black men and white men. Therefore, the lowest CHD mortality rates were observed in occupational structure category III, which had the greatest community resources. Blue collar mortality in category I compared with category III was 2.1 times higher among black workers and 1.5 times higher among white workers. White collar mortality in category I compared with category III was 1.7 times higher among white workers. In addition, blue collar workers had substantially higher mortality compared with white collar workers in all occupational structure categories; about two times higher among blacks and about 1.8 times higher among whites.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Age adjusted CHD mortality among upstate New York men, ages 35–64, 1988–92, by occupational structure, class, race.

Mortality was substantially higher among black compared with white men among blue collar workers in each occupational structure category, but was similar among white collar workers (table 2). Mortality ratios for black compared with white men in blue collar jobs, are inversely associated with occupational structure category; 1.8, 1.5, and 1.3 in category I, II, and III, respectively.

DISCUSSION

Occupational structure was chosen in this study as the primary characteristic for analysis as income is primarily distributed by means of work^22,23; therefore, distributions of jobs across communities represents a fundamental mechanism for distributions of income and other related social and economic resources.^24–26 However, a community ecology does not represent a linear or hierarchical (causal) sequence, but should instead be viewed as a circular set of relations that influence and perpetuate each of the other characteristics; and these characteristics operate in relation to state, regional, national, and international economies.²⁷ For example, the skills and education levels of a labour force reflects the current set of jobs in a community (table 1) and is an important factor in competing for relocating businesses and capital resources to initiate new business development in a community.^28–32 Also, the diversity and availability of consumer services reflects occupational structure (that is, via local incomes) and is a factor in attracting additional skilled workers to relocate into a community.^32,33 Thus, the ecological model conceptualises risk factors and CHD rates, occupational and industrial structure, consumer and producer services, and other aspects of community infrastructure to be highly interdependent. Therefore, we wish to describe how commonly these conditions coexist in communities and to better understand the relations between all of these community level characteristics and health outcomes including CHD mortality levels.

Higher CHD mortality among blue collar compared with white collar workers was shown in all occupational structure categories. Differential access and utilisation of beneficial community conditions among social class groups may reflect several advantages conferred by higher social class position, including income and wealth, occupational prestige and working conditions, social networks, and education levels.^11,34 Such advantages influence coronary health³⁵ through knowledge of coronary risk factors and motivation and resources to change practices^36–39; ability to afford services, especially routine, health promotion services (for example, gym costs, preventive medical care); ability to afford comfortable, safe living conditions (for example, housing costs⁴⁰ and healthy foods^41–43); amount of leisure time and time flexibility (for example, for exercise and the preparation of healthy foods)⁴⁴; higher job control and/or lower job demand (that is, lowered stress)^45,46; adequate economic wealth (for example, savings) to provide for temporary periods of insufficient income.²³ The inverse association between CHD mortality and occupational structure among all workers is consistent with increasingly beneficial community conditions for coronary health. However, white collar more than blue collar workers may have the necessary personal, social and economic resources to utilise available community services.

Occupational structure is an indicator of overall economic resources and services of communities but does not represent race and class residential patterns and differential access to services and living conditions within communities. Several studies provide evidence of differential opportunities for healthful living among neighbourhoods. Higher costs and fewer selections of foods, including fresh fruits and vegetables and low fat milk, have been reported for poorer compared with more affluent neighbourhoods.^47–51 Targeted advertising of cigarettes and alcohol consumption to lower income neighbourhoods has also been reported.^52,53

Table 3 shows indices of income inequality and class and race residential segregation for upstate NY counties in 1990. The scores for residential social class segregation reveal relatively low segregation in general and only modest differences in class segregation between occupational structure categories. In contrast, residential racial segregation seems high in all occupational structure categories. This score represents the percentage of the black population that would need to relocate so that census tracts reflect the percentage of county populations that are black, on average for counties in each occupational structure category. Therefore, at least 59% of black populations would need to relocate to achieve residential racial integration in each occupational structure category, which suggests differential access to living conditions based on race.

Black, blue collar workers had the highest CHD mortality rates in all occupational structure categories, which may reflect greater vulnerability to effects of segregation in access to services and living conditions than among black, white collar workers. Racial residential segregation has been associated with adverse population health and racial differences in population health in previous studies.^54,55 In addition, in more rural areas such as occupational structure category I (table 1), the black population represented only 3% of the total population (table 2). Therefore, awareness of racism may be heightened and acts of discrimination more commonly experienced given the small black population in these areas. Also, occupational experiences of racism and discrepancies between expectations and actual experiences may be more apparent.

Income inequality scores were relatively low within each occupational structure category, ranging from 10 to 22 (table 3). The income inequality score represents the share of total income that would need to be transferred from those above the mean income level to those below the mean level to achieve equality in the distribution of income. In 1990 the score for the United States was 30.2%; state scores ranged from 27 to 34, with a score of 33 for NY state overall.²⁰ Occupational structure may represent the underlying cause of income inequality, which may be the reason for observing relatively low income inequality within categories, after stratifying on occupational structure.

The main limitations of this study were missing occupation information in the death certificate data and uncertainties about population denominator counts. In a sensitivity analysis, when decedents who were missing occupation information were analysed as blue collar workers, resulting CHD mortality rates showed very similar patterns to those shown in figure 1. Lower labour force participation among black compared with white men was observed in all occupational structure categories, however, substantially lower participation among black men (46%) compared with white men (83%) was observed in category I (table 1). However, workers in areas of chronic high unemployment, who are “discouraged workers” and have given up looking for a job are not considered to be unemployed and are not counted as part of the labour force. This is one explanation for the lower labour force participation among black people, especially among black men in occupational structure category I. Also, the incarcerated are included in county population counts, which is the denominator for the labour force participation calculation, but are not considered to be in the labour force (numerator value). When incarcerated populations are added to the numerator for the calculation of labour force participation, black labour force participation increases to 75%, with little change for white men in occupational structure category I.

Higher premature mortality among blue collar compared with white collar workers and among black men compared with white men are consistent with previous studies.^6,12 Also, results showing effects of area resources on mortality levels and of social class and area resources^{10,56 57} are consistent with the literature. However, in previous studies there has been an emphasis on isolating effects of individual class or area of residence,⁵⁸ rather than on quantifying joint effects of class, race, and community conditions as presented in this study.

Results from large community based heart disease prevention trials, current theories of health education, as well as studies using occupational structure as a measure of community resources, indicate the need for a better understanding of how community social and physical environments impede or improve the coronary health of residents. To our knowledge and based on a search of the PubMed database, this study is the first to analyse joint effects of individual social class position, race, and community occupational structure on CHD mortality. Results from this study showed consistent beneficial effects of better community conditions on premature CHD mortality, but also revealed persistent excess mortality for less advantaged subgroups within communities (that is, by class and race). However, this study represented a limited geographical area (upstate New York). This approach needs to be replicated to ascertain the consistency of these results in other geographical localities and, therefore, to better understand policy implications suggested by these findings.