Abstract

Conflicting findings in the literature have made the relation between job strain and coronary heart disease (CHD) controversial. The effect of high job strain on the 10-year incidence of CHD and total mortality was examined in men and women participating in the Framingham Offspring Study; 3,039 participants, 1,711 men and 1,328 women, aged 18–77 years, were examined between 1984 and 1987 and followed for 10 years. Measures of job strain, occupational characteristics, and risk factors for CHD were collected at the baseline examination. Before and after controlling for systolic blood pressure, body mass index, cigarette smoking, diabetes, and the total/high density lipoprotein cholesterol ratio in Cox proportional hazards models, the authors found that high job strain was not associated with mortality or incident CHD in either men or women over the follow-up period. Contrary to expectation, women with active job strain (high demands-high control) had a 2.8-fold increased risk of CHD (95% confidence interval: 1.1, 7.2) compared with women with high job strain (high demands-low control). For men, higher education, personal income, and occupational prestige were related to decreased risk of total mortality and CHD. These findings do not support high job strain as a significant risk factor for CHD or death in men or women.

Received for publication March 31, 2003; accepted for publication December 4, 2003.

There has been longstanding interest in the impact of job strain on health outcomes. In 1970, Karasek (1) proposed a two-dimensional model for measuring psychosocial work conditions in which the combination of high job demands and low freedom to make job decisions (low control or low decision latitude) was found to be related to depression, exhaustion, and job dissatisfaction. Since then, a number of studies have used this paradigm to define job strain in relation to coronary heart disease (CHD) or risk factors for CHD (217).

Another study used a complementary concept of effort-reward imbalance (18). Many of these studies, however, have serious limitations including reliance on prevalence data (3, 8), failure to carefully measure and control for CHD risk factors (2, 4, 6, 11), poor confirmation of CHD (2, 4, 8), measurement of occupational strain after the event (4, 10, 12, 13), and no inclusion of women (7, 9, 1214). Other problems have included inconsistent definitions of outcomes ranging from self-reported “heart weakness” to documented CHD death. In addition, “job strain” has been defined in various ways (9, 10, 12, 14, 19). These limitations have led to inconsistent findings.

When the literature is limited to the seven studies with prospective data that controlled for CHD risk factors, five studies (1418) (four from the same study population (1518)) found some form of job strain to be associated with the incidence of some form of CHD. Two studies found no association between job strain and incident CHD or CHD mortality (7, 9).

The Framingham Offspring Study can address these inconsistencies through a strong prospective cohort design, clinically verified disease incidence, the inclusion of both women and men, and direct measures of job strain based on Karasek’s model. Other variables that might modify or confound the relations between job strain and incident CHD and death were also collected routinely. These variables included occupational prestige, personal income, educational level, occupation, and the standard clinical risk factors for the development of CHD.

The purpose of this study was to test the hypothesis that high job strain is independently related to the 10-year incidence of CHD and total mortality in men and women in a well-defined cohort using standardized measures of occupational risk and disease.

MATERIALS AND METHODS

The Framingham Heart Study is a prospective longitudinal cohort study that began in 1948. In 1971–1975, the offspring (and spouses) of the original Framingham Heart Study cohort were enrolled in the Framingham Offspring Study (20). A total of 5,135 persons (75 percent of all invitees) volunteered for the first examination cycle. At the third examination cycle from 1984 through 1987, 3,873 participants aged 18–77 years were examined. Subjects were mailed the self-administered psychosocial assessment forms that were collected at the scheduled clinic examination. A total of 3,682 participants completed a psychosocial questionnaire, 1,769 men and 1,913 women, compromising a 95 percent response rate. Subjects were excluded from the present study for the following indications: incomplete questionnaire (n = 191) or prevalent CHD (n = 107). Included in the questionnaire were variables concerning job strain and occupational and job characteristics. Because this study was interested in job and occupational characteristics, we stratified the cohort into women (n = 1,328) and men (n = 1,711) who responded that they had been employed outside the home most of their adult years (18 years of age or older).

Data collected that related to job characteristics included Karasek’s questions to assess the following items: decision authority (three scale items), skill discretion (six scale items), and job demands (five scale items). Questions that defined decision authority included being able to make decisions about one’s own job, freedom to decide how to work, and having a lot to say about what happens on one’s job. Questions that defined skill discretion included learning new things, being creative, and developing one’s own specific abilities on the job. Questions that defined job demands included excessive work, working fast, working hard, and conflicting job demands. These scales were measured according to established scoring algorithms. Each variable that contributes to a scale was scored on a Likert scale from 1 (strongly disagree) to 4 (strongly agree) and summed over the particular items constituting the job scale. Because there are no normative data for these scales, we used continuous scale scores to define the mean for men and women separately.

A scale termed decision latitude was created by combining the two scales skill discretion and decision authority. This created scale of decision latitude in combination with job demands forms the basis to define job strain: 1) high strain, based on simultaneously high job demands (above the median score) and low job decision latitude (at or below the median score); 2) low strain, based on simultaneously low job demands and high job decision latitude; 3) passive, based on simultaneously low job demands and low decision latitude; or 4) active, based on simultaneously high job demands and high job decision latitude (2) (figure 1).

The Census Bureau’s six summary groups developed from the 1980 Standard Industrial Classification were used for classifying occupations (21). In addition, scores were assigned to each occupation using the occupational prestige ratings from the 1989 General Social Survey (22). Occupational prestige ratings ranged from a high score of 86 for physicians to 17 for “miscellaneous food preparation.”

Several variables were collected and examined that could potentially confound the relations between job strain and the health outcomes of interest (e.g., being related to both job strain and the outcome). These variables included occupation, occupational prestige, personal income, education, and housework responsibilities. The choice of potential confounders was based on findings from previous research suggesting that these characteristics may explain any observed relation between job strain and the outcomes of interest (23). The current research examined one’s personal occupational status, job attributes, prestige of one’s own occupation, and educational level, not a composite family score for standard of living. Therefore, total personal income was selected as a potential confounding variable, instead of total family income. The separate scales that defined job strain were also examined to see if they had a unique impact on health outcomes.

Standard CHD risk factors were collected at the clinic examination and included systolic blood pressure, body mass index (weight (kg)/height (m)2), current cigarette smoking, the presence of diabetes (defined as a fasting blood glucose level of at least 126 mg/dl or on treatment), and the ratio of total to high density lipoprotein cholesterol.

The two outcomes of interest included 10-year total mortality and incident CHD. The definition of CHD has been published previously (24, 25); the manifestations of interest in these analyses included myocardial infarction (recognized and unrecognized), coronary insufficiency, and CHD death (both sudden and not sudden).

Regarding statistical analyses, descriptive statistics were generated in the first phase of the analysis for all study variables for the total sample and then for men and women, considered separately. Means, standard deviations, medians, and interquartile ranges were investigated for continuous variables, and relative frequencies were examined for discrete variables. Tests for differences in distributions of discrete variables were performed using the chi-square test. Tests were performed only for those persons who were employed outside the home most of their adult life.

In the next phase of the analysis, total mortality and CHD events were tallied for men and women separately according to specific levels of the predictor variables. Age-adjusted rates of total mortality and CHD over 10 years were then computed for each level of each variable using Cox proportional hazards regression. Finally, using a sex-specific Cox proportional hazards regression analysis, we estimated the relative risks of total mortality and CHD for each job-related and demographic variable in the age-adjusted analyses. These analyses were adjusted for standard CHD risk factors and were performed on the individuals employed outside the home most of their adult life. Relative risks for continuous measures were computed relative to a 1 standard deviation change in the continuous measure. It should be noted that high strain is the referent category, since the hypothesis being tested is that people scoring in this quadrant will have the greatest risk compared with others in the other three quadrants.

RESULTS

Among those reporting having been employed outside the home most of their adult life, most men (31.1 percent) were in the active category of job strain, and the greatest percentage of women (35.6 percent) were in the passive job strain category (table 1; figure 1). With regard to occupation, the highest proportion of men was in management/professional occupations (42.6 percent), and the highest proportion of women was in technical/sales positions (42.9 percent).

Tables 2 and 3 present the relations between measures of job strain and the separate scales used to define job strain as risk factors for CHD and total morality for men and women, respectively. Men (table 2) reporting passive job strain were significantly older and had higher systolic blood pressure. Men reporting active strain were younger and had the lowest systolic blood pressure. When the scales that comprise job strain were examined, decision latitude, skill discretion, decision authority, and job demands were inversely and significantly related to age. With the exception of decision authority, these scales were also significantly and inversely related to systolic blood pressure.

Women (table 3) reporting active job strain were significantly younger and had lower systolic blood pressure and ratio of total to high density lipoprotein cholesterol levels. Women with active job strain were less likely to be smokers or to have diabetes. When the scales that comprise job strain were examined, decision latitude, skill discretion, and decision authority were all significantly and inversely related to age, systolic blood pressure, and the ratio of total to high density lipoprotein cholesterol.

Table 4 presents the age-adjusted 10-year incidence rates for total mortality and CHD by job strain variables and potential confounding variables (i.e., those variables related to both job strain and the outcome of interest). For men, job strain was not associated with either incident CHD or total mortality over the 10 years of follow-up. A higher educational level and a higher personal income were related to lower risk of both the development of CHD and total mortality. Men employed as operators/laborers had the highest rates of both death and CHD, and those employed in management/professional or farming/forestry occupations had the lowest rates. Women in service or construction occupations had the highest total mortality rates, and those in management/professional or technical/sales occupations had the highest 10-year CHD rates. Although the association of occupation with both outcomes was significant in men and women, there were no events in some of the categories (e.g., farming/forestry).

In age-adjusted models, with regard to the job characteristics (table 5), higher occupational prestige was associated with significantly lower total mortality (relative risk (RR) = 0.73, 95 percent confidence interval (CI): 0.62, 0.86) and 10-year incidence of CHD (RR = 0.81, 95 percent CI: 0.68, 0.97) in men. None of the other separate job characteristic scales that define job strain was significantly associated with total mortality or CHD in men. In women, the composite scale of decision latitude (made up of both skill discretion and decision authority) (RR = 1.62, 95 percent CI: 1.09, 2.42) and the single scale of high decision authority (RR = 1.55, 95 percent CI: 1.05, 2.29) were associated with significantly higher age-adjusted risk of CHD. Job demands were not associated with either total mortality or the 10-year incidence of CHD in women. Occupational prestige did not reach statistical significance with either the development of CHD or total mortality in women in this sample.

Table 6 presents the multivariable analyses for the 10-year follow-up for total mortality and incident CHD. Each job-related or demographic variable was entered into a Cox proportional hazards model with age, systolic blood pressure, the ratio of total cholesterol to high density lipoprotein cholesterol, body mass index, cigarette smoking, and diabetes.

Job strain was significantly associated with the 10-year incidence of CHD in women. Contrary to our hypothesis, however, women with increased job demands and increased decision latitude (those in the active job strain quadrant in figure 1) had significantly higher risk of developing CHD over the 10 years of follow-up compared with women reporting decreased decision latitude and increased job demands (those in the high job strain quadrant in figure 1). When the individual scales that define job strain were examined separately in the multivariable analyses, decision latitude and the subscales of decision authority and skill discretion were significantly and positively related to the 10-year incidence of CHD in women. To determine if the effect of active job strain on the 10-year incidence of CHD in women could be confounded by occupational level, occupation was included in the multivariable equation with the standard risk factors. The relative risks for each category compared with high strain after adjustment for standard risk factors and occupation were as follows: active job strain (RR = 2.97, p = 0.02), low job strain (RR = 1.93, p = 0.23), and passive job strain (RR = 0.43, p = 0.15). Family income and education were also entered into the model with job strain in women, and the results were not changed.

It is possible that the association between high decision authority/high decision latitude and CHD in women might be explained by the strain of household responsibilities. However, this hypothesis does not seem to be the case, as the number of children a woman had, housework responsibilities, and housework as a big strain were not related to the age-adjusted 10-year incidence of CHD in women employed most of their adult years (data not shown).

Among men, educational level and personal income were inversely and significantly related to death and CHD (table 6). Compared with men in management/professional positions, men who were operators/laborers had over twice the risk of death or CHD over the 10 years of follow-up. With regard to the job characteristic scales, occupational prestige was significantly protective in men against both death and CHD over the 10 years of follow-up. To further explore the relation between occupational prestige and total mortality, we entered both educational level and occupational prestige into the same multivariable equation with the standard risk factors. A man’s educational level or income level did not account for the relation observed between occupational prestige and total mortality (data not shown).

DISCUSSION

These findings from the Framingham Offspring Study failed to support the notion that high job strain, as defined by high job demands and low decision latitude, was a significant risk factor for the development of CHD or total mortality over a 10-year period in either men or women. Contrary to our hypothesis, women reporting active job strain, as defined by high job demands and high decision latitude, were significantly more likely (2.8 times) to develop CHD compared with those reporting high job strain in their work setting (figure 1). This was true after adjustment for the standard risk factors associated with CHD. The adverse effect of active job strain on the 10-year incidence of CHD in women seems to be not the result of high job demands per se but the result of higher levels of decision latitude and/or decision authority or skill discretion. It is of interest that women (as well as men) in the active job strain category actually have lower levels on most of the standard risk factors measured in this study (tables 2 and 3). The potential mechanisms for this increased risk may be through characteristics not measured in this study, such as hemostasis and thrombosis, coronary vasoconstriction, platelet aggregation, or plaque rupture.

The question arises as to why we observed a counterintuitive relation in our data in which increased authority and latitude in one’s job were related to increased risk of developing CHD in women. One speculative explanation involves historical shifts in social status and roles. In explaining the shift in the relation between social status and CHD in men over time, Morgenstern (26) observes that there are new demands from a new social environment where individuals need to develop new values, attitudes, and patterns of behaviors. In the past, a woman’s social status was most likely a reflection of the income and occupation of her husband, not of her own role in society. In the current study, we see women with jobs associated with autonomy, authority, and control at higher risk for the development of CHD. For women, the adoption of new social roles may be particularly difficult because of cultural resistance to the change in women’s roles (27). It could be argued that women in the mid-1980s who were in positions of authority and control were at the cutting edge of a social transition. As women broke into positions of authority, the social pressures and reactions to women in a “man’s role” may have had deleterious effects on women’s risk of CHD. For example, discrimination and disparities with regard to wages compared with men may have had an effect. As recently as 2001, the General Accounting Office reported that full-time female managers earned less than their male counterparts in all 10 industries studied in 1995 and 2000. In 2000, women managers in the communications industry made 73 cents for every dollar earned by a man (28). It is important for future research to explore issues and characteristics of working women and the work environment that may produce a high-risk situation for the development of CHD.

There have been seven studies (four from the same population) of job strain and risk of CHD that were prospective in design and controlled for CHD risk factors (7, 9, 1418). One study found high job strain to be marginally protective in a population of men of Japanese ancestry in Hawaii (7). Another study found that the combined effect on incidence of CHD of high job demands and “low resources” was mitigated by income level (14). When decision latitude (job control) was examined separately in the Western Electric Study (9), it was not a significant predictor of CHD death but was more pronounced (protective) for white-collar than for blue-collar workers. Job demands were also not associated with CHD death in the Western Electric Study. Contrary to these findings, the Whitehall Study found low job control to be significantly associated with an increased risk of CHD disease in both men and women regardless of employment grade and CHD risk factors (15). Job demands and social support were examined in this report from the Whitehall Study and found not to be related to CHD. Two more recent publications from the Whitehall II Study examined the combined effects of high job demands and low job control (job strain) (17) and the imbalance between effort and reward (18) on the incidence of CHD. The former study found that job strain was significantly related to the incidence of all CHD (hazard ratio = 1.38, 95 percent CI: 1.10, 1.75) but not to the “hard” endpoints of fatal CHD/nonfatal myocardial infarction (hazard ratio = 1.16, 95 percent CI: 0.78, 1.71). In the study of effort-reward imbalance, a similar finding was reported: A high ratio of effort to rewards was related to all CHD but not to fatal CHD/nonfatal myocardial infarction. The outcome of all CHD included “potential cases of myocardial infarction and angina”(18, p. 778).

An important similarity between the Whitehall Study and the Framingham Offspring Study is that, when definite CHD was the outcome of interest, job strain was not a significant risk factor after potential confounders were controlled in either men or women. We know from previous research of psychosocial risk factors in the Framingham data that inclusion of definite angina pectoris as an outcome for coronary heart disease can result in significant findings that are not observed when only “hard” endpoints are observed (29). The importance of the distinction between findings related to definite or hard endpoints and those related to potential CHD endpoints cannot be overemphasized. In a study of the validation of the London School of Hygiene (Rose) Questionnaire, Garber et al. (30) found a poor relation between “Rose Questionnaire angina” and exercise thallium-201 scintigraphy, an objective measure of myocardial ischemia. In this study, the Rose Questionnaire resulted in a false positive rate of 75 percent in females and 27 percent in males.

Among men, the work characteristics that were related to their health in a positive way were increased occupational prestige and management/professional occupational level. Perhaps men are socially and psychologically rewarded for being in occupational positions of higher prestige. Women in similar positions may not be afforded the same benefits as men. In addition, a work environment that encourages and rewards authority and control may be counter to an environment of cooperation and that may be more conducive to women’s health. Similar findings have been reported by Bassuk et al. (31), who found that education, household income, and occupational prestige were associated with lower mortality in men, but only household income was protective in women.

The strengths of the Framingham Offspring Study include direct measures of occupational characteristics, prospective design, inclusion of both men and women, a stable cohort, carefully assessed endpoints, and information on standard risk factors. A limitation of this particular study is that the Framingham cohort is predominantly Caucasian and was predominantly middle aged at the time of the study; the findings may not be generalizable to other ethnicities and the elderly. In addition, the present study had a greater power to detect predictors of mortality and CHD in men than in women. For instance, when comparing two similarly sized employed groups, the study had power exceeding 80 percent to detect a relative risk for total mortality of at least 2.0 in employed women and a relative risk of 1.6 in employed men. Hence, one should be cautious in interpreting our negative findings in women. It should be noted, however, that this is one of the larger data sets with prospective data on women.

In summary, this research did not support the hypothesis that high job strain is associated with the incidence of CHD or total mortality in men and women. Women in occupational positions with high decision latitude and/or high authority over their work were adversely affected for the occurrence of definite CHD. For men, low education, low occupational prestige, and low income adversely affected their health. The mechanisms by which these variables operate are complex and require additional investigation. Similar studies in other multiethnic cohorts are also warranted.

ACKNOWLEDGMENTS

This research was supported by the following grants from the National Heart, Lung, and Blood Institute: 1 R03 HL 67426-01 and N01-HC-25195.

Reprint requests to Dr. Elaine D. Eaker, Eaker Epidemiology Enterprises, LLC, 8975 Country Road V, Chili, WI 54420 (e-mail: eakerepi@tznet.com).

FIGURE 1. Definition of job strain.

FIGURE 1. Definition of job strain.

TABLE 1.

Demographic and psychosocial characteristics of participants employed outside the home most of their adult life, Framingham Offspring Study, 19841987

 Men (%)(n = 1,711)  Women (%)(n = 1,328)  
Characteristic     
Education (years)     
≤12  40.1  46.1  
13–15 22.3  27.4  
≥16 37.7  26.5  
Job strain     
Active 31.1  21.5  
High strain 16.4  29.0  
Low strain 24.4  13.9  
Passive 28.1  35.6  
Personal income ($)     
0–9,000 2.9  27.5  
10,000–19,000 10.7  35.0  
20,000–29,000 26.3  25.0  
30,000–49,000 39.3  10.2  
≥50,000 20.8  2.2  
Occupation     
Management/professional 42.6  37.5  
Technical/sales 19.6  42.9  
Service 8.7  11.5  
Farming/forestry 0.8  0.4  
Production/craft 5.4  0.2  
Construction 13.5  2.0  
Operators/labors 9.4  5.4  
     
 Mean (SD*) Median Mean (SD) Median 
Occupational scales     
Occupational prestige 48.9 (12.6) 50.0 46.9 (12.9) 46.0 
Decision latitude 75.0 (12.1) 74.0 69.0 (12.6) 70.0 
Skill discretion 18.0 (3.0) 18.0 16.7 (3.3) 17.0 
Decision authority 19.4 (3.6) 20.0 17.8 (3.7) 18.0 
Job demands 31.5 (6.0) 31.0 31.9 (6.0) 32.0 
 Men (%)(n = 1,711)  Women (%)(n = 1,328)  
Characteristic     
Education (years)     
≤12  40.1  46.1  
13–15 22.3  27.4  
≥16 37.7  26.5  
Job strain     
Active 31.1  21.5  
High strain 16.4  29.0  
Low strain 24.4  13.9  
Passive 28.1  35.6  
Personal income ($)     
0–9,000 2.9  27.5  
10,000–19,000 10.7  35.0  
20,000–29,000 26.3  25.0  
30,000–49,000 39.3  10.2  
≥50,000 20.8  2.2  
Occupation     
Management/professional 42.6  37.5  
Technical/sales 19.6  42.9  
Service 8.7  11.5  
Farming/forestry 0.8  0.4  
Production/craft 5.4  0.2  
Construction 13.5  2.0  
Operators/labors 9.4  5.4  
     
 Mean (SD*) Median Mean (SD) Median 
Occupational scales     
Occupational prestige 48.9 (12.6) 50.0 46.9 (12.9) 46.0 
Decision latitude 75.0 (12.1) 74.0 69.0 (12.6) 70.0 
Skill discretion 18.0 (3.0) 18.0 16.7 (3.3) 17.0 
Decision authority 19.4 (3.6) 20.0 17.8 (3.7) 18.0 
Job demands 31.5 (6.0) 31.0 31.9 (6.0) 32.0 

* SD, standard deviation.

TABLE 2.

Relation between occupational characteristics and risk factors related to coronary heart disease and total mortality: men employed outside the home most of their adult life, Framingham Offspring Study, 19841987

 Age (mean years) p value SBP* (mean mmHg) p value BMI* (mean kg/m2p value Total/HDL* cholesterol ratio (mean) p value  Cigarette smoking (%) p value Diabetes mellitus (%) p value 
Job strain  <0.01  0.02  0.85  0.39   0.08  0.68 
High strain 47.8  126.4  27.2  5.1   30.4  5.8  
Active 46.3  124.2  27.2  5.0   29.7  3.9  
Passive 50.1  127.2  27.2  5.2   30.1  4.8  
Low strain 48.9  125.6  27.0  5.1   23.2  4.9  
Decision latitude†  <0.01  0.01  0.73  0.20   0.10  0.35 
Below median 49.3  126.8  27.2  5.2   30.0  5.2  
Above median 47.5  124.9  27.1  5.0   26.3  4.2  
Skill discretion  <0.01  0.02  0.13  0.09   0.14  0.44 
Below median 49.1  126.6  27.3  5.2   29.5  5.0  
Above median 47.4  124.8  27.0  5.0   26.2  4.2  
Decision authority  <0.01  0.12  0.38  0.88   0.10  0.40 
Below median 49.4  126.5  27.1  5.1   30.0  5.1  
Above median 47.7  125.3  27.2  5.1   26.3  4.3  
Job demands  <0.01  0.05  0.62  0.22   0.20  0.29 
Below median 49.5  126.4  27.1  5.1   26.8  4.8  
Above median 46.8  124.9  27.2  5.0   30.0  4.5  
 Age (mean years) p value SBP* (mean mmHg) p value BMI* (mean kg/m2p value Total/HDL* cholesterol ratio (mean) p value  Cigarette smoking (%) p value Diabetes mellitus (%) p value 
Job strain  <0.01  0.02  0.85  0.39   0.08  0.68 
High strain 47.8  126.4  27.2  5.1   30.4  5.8  
Active 46.3  124.2  27.2  5.0   29.7  3.9  
Passive 50.1  127.2  27.2  5.2   30.1  4.8  
Low strain 48.9  125.6  27.0  5.1   23.2  4.9  
Decision latitude†  <0.01  0.01  0.73  0.20   0.10  0.35 
Below median 49.3  126.8  27.2  5.2   30.0  5.2  
Above median 47.5  124.9  27.1  5.0   26.3  4.2  
Skill discretion  <0.01  0.02  0.13  0.09   0.14  0.44 
Below median 49.1  126.6  27.3  5.2   29.5  5.0  
Above median 47.4  124.8  27.0  5.0   26.2  4.2  
Decision authority  <0.01  0.12  0.38  0.88   0.10  0.40 
Below median 49.4  126.5  27.1  5.1   30.0  5.1  
Above median 47.7  125.3  27.2  5.1   26.3  4.3  
Job demands  <0.01  0.05  0.62  0.22   0.20  0.29 
Below median 49.5  126.4  27.1  5.1   26.8  4.8  
Above median 46.8  124.9  27.2  5.0   30.0  4.5  

* SBP, systolic blood pressure; BMI, body mass index; HDL, high density lipoprotein.

† Decision latitude is a combination of skill discretion and decision authority.

TABLE 3.

Relation between occupational characteristics and risk factors related to coronary heart disease and total mortality: women employed outside the home most of their adult life, Framingham Offspring Study, 19841987

 Age (mean years) p value SBP* (mean mmHg) p value BMI* (mean kg/m2p value Total/HDL* cholesterol ratio (mean) p value  Cigarette smoking (%) p value Diabetes mellitus (%) p value 
Job strain  <0.01  0.05  0.66  <0.01   0.07  0.03 
High strain 46.0  120.8  25.4  4.2   33.4  3.9  
Active 44.1  117.8  25.0  3.7   23.3  0.4  
Passive 48.0  121.0  25.5  4.1   29.6  1.9  
Low strain 47.6  118.3  25.1  3.9   29.2  3.7  
Decision latitude†  0.01  <0.01  0.44  0.02   0.08  0.62 
Below median 47.3  121.5  25.5  4.1   32.2  2.6  
Above median 45.9  118.6  25.2  3.9   27.6  2.2  
Skill discretion  0.04  <0.01  0.05  <0.01   0.25  0.30 
Below median 47.2  121.6  25.7  4.2   31.7  2.9  
Above median 46.1  118.6  25.1  3.9   28.6  2.0  
Decision authority  0.05  0.04  0.78  0.05   0.07  0.30 
Below median 47.4  121.5  25.3  4.1   32.5  3.0  
Above median 46.3  119.4  25.4  3.9   27.6  2.1  
Job demands  <0.01  0.30  0.51  0.31   0.96  0.90 
Below median 48.0  120.6  25.4  4.0   29.4  2.5  
Above median 45.3  119.6  24.2  3.9   29.5  2.7  
 Age (mean years) p value SBP* (mean mmHg) p value BMI* (mean kg/m2p value Total/HDL* cholesterol ratio (mean) p value  Cigarette smoking (%) p value Diabetes mellitus (%) p value 
Job strain  <0.01  0.05  0.66  <0.01   0.07  0.03 
High strain 46.0  120.8  25.4  4.2   33.4  3.9  
Active 44.1  117.8  25.0  3.7   23.3  0.4  
Passive 48.0  121.0  25.5  4.1   29.6  1.9  
Low strain 47.6  118.3  25.1  3.9   29.2  3.7  
Decision latitude†  0.01  <0.01  0.44  0.02   0.08  0.62 
Below median 47.3  121.5  25.5  4.1   32.2  2.6  
Above median 45.9  118.6  25.2  3.9   27.6  2.2  
Skill discretion  0.04  <0.01  0.05  <0.01   0.25  0.30 
Below median 47.2  121.6  25.7  4.2   31.7  2.9  
Above median 46.1  118.6  25.1  3.9   28.6  2.0  
Decision authority  0.05  0.04  0.78  0.05   0.07  0.30 
Below median 47.4  121.5  25.3  4.1   32.5  3.0  
Above median 46.3  119.4  25.4  3.9   27.6  2.1  
Job demands  <0.01  0.30  0.51  0.31   0.96  0.90 
Below median 48.0  120.6  25.4  4.0   29.4  2.5  
Above median 45.3  119.6  24.2  3.9   29.5  2.7  

* SBP, systolic blood pressure; BMI, body mass index; HDL, high density lipoprotein.

† Decision latitude is a combination of skill discretion and decision authority.

TABLE 4.

Age-adjusted 10-year rates of total mortality and coronary heart disease among people employed at least half of their adult years, Framingham Offspring Study, from 19841987 to 1994–1999

Characteristics Total mortality*  Coronary heart disease† 
Men (n = 1,711)  Women (n = 1,328)  Men (n = 1,624)  Women (n = 1,316) 
Age-adjusted rate p value  Age-adjusted rate p value  Age-adjusted rate p value  Age-adjusted rate p value 
Job strain  0.46   0.58   0.87   0.10 
Active 5.3   2.7   5.4   2.5  
High strain 5.4   2.0   6.3   1.2  
Low strain 4.1   1.6   5.0   1.6  
Passive 5.7   2.8   5.3   0.8  
Education (years)  0.002   0.09   0.002   0.48 
≤12 6.2   3.2   6.5   1.1  
13–15 6.4   2.0   7.5   1.6  
≥16  3.3   1.4   3.1   1.5  
Occupation  <0.001   <0.001   0.03   <0.001 
Management/professional 4.2   1.7   4.3   1.7  
Technical/sales 6.2   2.9   5.4   1.6  
Service 4.2   3.3   8.1    
Farming/forestry 0.0   0.0      
Production/craft 6.9   0.0   7.6    
Construction 4.9   4.7   5.1   1.1  
Operators/laborers 8.4   2.7   8.2   0.6  
Personal income ($)  0.01   0.80   0.001   0.19 
0–9,000 11.0   2.7   9.4   1.1  
10,000–19,000 7.8   1.8   6.2   1.2  
20,000–29,000 4.0   2.0   5.3   1.6  
30,000–49,000 4.6   4.5   6.2   1.4  
≥50,000 5.9     4.2   1.4  
Characteristics Total mortality*  Coronary heart disease† 
Men (n = 1,711)  Women (n = 1,328)  Men (n = 1,624)  Women (n = 1,316) 
Age-adjusted rate p value  Age-adjusted rate p value  Age-adjusted rate p value  Age-adjusted rate p value 
Job strain  0.46   0.58   0.87   0.10 
Active 5.3   2.7   5.4   2.5  
High strain 5.4   2.0   6.3   1.2  
Low strain 4.1   1.6   5.0   1.6  
Passive 5.7   2.8   5.3   0.8  
Education (years)  0.002   0.09   0.002   0.48 
≤12 6.2   3.2   6.5   1.1  
13–15 6.4   2.0   7.5   1.6  
≥16  3.3   1.4   3.1   1.5  
Occupation  <0.001   <0.001   0.03   <0.001 
Management/professional 4.2   1.7   4.3   1.7  
Technical/sales 6.2   2.9   5.4   1.6  
Service 4.2   3.3   8.1    
Farming/forestry 0.0   0.0      
Production/craft 6.9   0.0   7.6    
Construction 4.9   4.7   5.1   1.1  
Operators/laborers 8.4   2.7   8.2   0.6  
Personal income ($)  0.01   0.80   0.001   0.19 
0–9,000 11.0   2.7   9.4   1.1  
10,000–19,000 7.8   1.8   6.2   1.2  
20,000–29,000 4.0   2.0   5.3   1.6  
30,000–49,000 4.6   4.5   6.2   1.4  
≥50,000 5.9     4.2   1.4  

* Total mortality: no. of events among men = 160 and among women = 54.

† Coronary heart disease: no. of events among men = 118 and among women = 31.

TABLE 5.

Age-adjusted relative risks per 1 standard deviation change in each scale for total mortality and incident coronary heart disease by job characteristics for men and women, Framingham Offspring Study, from 19841987 to 1994–1999

Job characteristics Total mortality  Coronary heart disease 
Men  Women  Men  Women 
Age-adjusted relative risk 95% confidence interval  Age-adjusted relative risk 95% confidence interval  Age-adjusted relative risk 95% confidence interval  Age-adjusted relative risk 95% confidence interval 
Occupational prestige 0.73* 0.62, 0.86*  0.88 0.65, 1.18  0.81* 0.68, 0.97*  1.23 0.87, 1.74 
Decision latitude 0.91 0.76, 1.08  0.86 0.63, 1.18  1.03 0.85, 1.24  1.62* 1.09, 2.42* 
Skill discretion 0.94 0.79, 1.11  0.81 0.58, 1.12  0.96 0.80, 1.16  1.47 0.99, 2.19 
Decision authority 0.89 0.75, 1.05  0.93 0.68, 1.26  1.06 0.88, 1.27  1.55* 1.05, 2.29* 
Job demands 1.01 0.85, 1.21  0.75 0.55, 1.02  1.00 0.82, 1.21  0.94 0.66, 1.35 
Job characteristics Total mortality  Coronary heart disease 
Men  Women  Men  Women 
Age-adjusted relative risk 95% confidence interval  Age-adjusted relative risk 95% confidence interval  Age-adjusted relative risk 95% confidence interval  Age-adjusted relative risk 95% confidence interval 
Occupational prestige 0.73* 0.62, 0.86*  0.88 0.65, 1.18  0.81* 0.68, 0.97*  1.23 0.87, 1.74 
Decision latitude 0.91 0.76, 1.08  0.86 0.63, 1.18  1.03 0.85, 1.24  1.62* 1.09, 2.42* 
Skill discretion 0.94 0.79, 1.11  0.81 0.58, 1.12  0.96 0.80, 1.16  1.47 0.99, 2.19 
Decision authority 0.89 0.75, 1.05  0.93 0.68, 1.26  1.06 0.88, 1.27  1.55* 1.05, 2.29* 
Job demands 1.01 0.85, 1.21  0.75 0.55, 1.02  1.00 0.82, 1.21  0.94 0.66, 1.35 

* Significant (p ≤ 0.05).

TABLE 6.

Multivariable-adjusted relative risks† for total mortality and incident coronary heart disease over 10 years among people employed at least half of their adult years, men and women, Framingham Offspring Study, from 19841987 to 1994–1999

Characteristic Total mortality  Coronary heart disease 
Men  Women  Men  Women 
Adjusted relative risk 95% confidence interval  Adjusted relative risk 95% confidence interval  Adjusted relative risk 95% confidence interval  Adjusted relative risk 95% confidence interval 
Job strain            
Active 1.10 0.66, 2.00  1.00 0.37, 2.75  0.87 0.51, 1.50  2.80* 1.09, 7.17* 
High strain Referent   Referent   Referent   Referent  
Low strain 0.85 0.48, 1.50  0.76 0.24, 2.42  0.85 0.50, 1.45  1.63 0.57, 4.67 
Passive 0.99 0.59, 1.68  1.37 0.63, 2.97  0.90 0.55, 1.48  0.45 0.15, 1.33 
Education (years)            
≤12  Referent   Referent   Referent   Referent  
13–15 0.92 0.61, 1.38  0.55 0.25, 1.19  1.12 0.72, 1.76  1.44 0.62, 3.36 
≥16 0.62* 0.41, 0.94*  0.49 0.19, 1.29  0.57* 0.35, 0.94*  1.62 0.61, 4.33 
Occupation            
Management/professional Referent   Referent   Referent   Referent  
Technical/sales 1.51 0.98, 2.32  1.76 0.81, 3.84  1.25 0.74, 2.12  0.88 0.40, 1.91 
Service 0.72 0.35, 1.48  2.26 0.84, 6.11  1.68 0.90, 3.16    
Farming/forestry            
Production/craft 1.79 0.92, 3.46     1.55 0.72, 3.34    
Construction 1.15 0.69, 1.91  2.78 0.73, 10.63  1.18 0.64, 2.16  0.68 0.09, 5.44 
Operators/laborers 2.02* 1.24, 3.29*  0.42 0.05, 3.41  2.11 1.17, 3.78    
Personal income ($)            
0–9,000 Referent   Referent   Referent   Referent  
10,000–19,000 0.71 0.43, 1.17  0.78 0.38, 1.60  0.75 0.39, 1.47  1.08 0.44, 2.64 
20,000–29,000 0.43* 0.26, 0.71*  0.97 0.43, 2.20  0.67 0.37, 1.18  1.45 0.54, 3.84 
30,000–49,000 0.59* 0.38, 0.91*  2.0 0.76, 5.50  0.86 0.51, 1.44  1.60 0.35, 7.30 
≥50,000 0.42* 0.24, 0.75*     0.34* 0.16, 0.73*  7.71 0.89, 66.81 
Occupational prestige 0.77* 0.65, 0.91*  0.99 0.97, 1.02  0.80* 0.66, 0.98*  1.02 0.99, 1.05 
Decision latitude 0.97 0.81, 1.16  0.98 0.96, 1.02  0.99 0.98, 1.02  1.98* 1.26, 3.10* 
Skill discretion 0.99 0.83, 1.19  0.94 0.85, 1.04  0.98 0.92, 1.05  1.20* 1.03, 1.37* 
Decision authority 0.95 0.80, 1.14  0.98 0.90, 1.07  1.00 0.95, 1.06  1.85* 1.21, 2.85* 
Job demands 1.05 0.78, 1.41  0.96 0.91, 1.01  1.00 0.97, 1.04  0.99 0.93, 1.06 
Characteristic Total mortality  Coronary heart disease 
Men  Women  Men  Women 
Adjusted relative risk 95% confidence interval  Adjusted relative risk 95% confidence interval  Adjusted relative risk 95% confidence interval  Adjusted relative risk 95% confidence interval 
Job strain            
Active 1.10 0.66, 2.00  1.00 0.37, 2.75  0.87 0.51, 1.50  2.80* 1.09, 7.17* 
High strain Referent   Referent   Referent   Referent  
Low strain 0.85 0.48, 1.50  0.76 0.24, 2.42  0.85 0.50, 1.45  1.63 0.57, 4.67 
Passive 0.99 0.59, 1.68  1.37 0.63, 2.97  0.90 0.55, 1.48  0.45 0.15, 1.33 
Education (years)            
≤12  Referent   Referent   Referent   Referent  
13–15 0.92 0.61, 1.38  0.55 0.25, 1.19  1.12 0.72, 1.76  1.44 0.62, 3.36 
≥16 0.62* 0.41, 0.94*  0.49 0.19, 1.29  0.57* 0.35, 0.94*  1.62 0.61, 4.33 
Occupation            
Management/professional Referent   Referent   Referent   Referent  
Technical/sales 1.51 0.98, 2.32  1.76 0.81, 3.84  1.25 0.74, 2.12  0.88 0.40, 1.91 
Service 0.72 0.35, 1.48  2.26 0.84, 6.11  1.68 0.90, 3.16    
Farming/forestry            
Production/craft 1.79 0.92, 3.46     1.55 0.72, 3.34    
Construction 1.15 0.69, 1.91  2.78 0.73, 10.63  1.18 0.64, 2.16  0.68 0.09, 5.44 
Operators/laborers 2.02* 1.24, 3.29*  0.42 0.05, 3.41  2.11 1.17, 3.78    
Personal income ($)            
0–9,000 Referent   Referent   Referent   Referent  
10,000–19,000 0.71 0.43, 1.17  0.78 0.38, 1.60  0.75 0.39, 1.47  1.08 0.44, 2.64 
20,000–29,000 0.43* 0.26, 0.71*  0.97 0.43, 2.20  0.67 0.37, 1.18  1.45 0.54, 3.84 
30,000–49,000 0.59* 0.38, 0.91*  2.0 0.76, 5.50  0.86 0.51, 1.44  1.60 0.35, 7.30 
≥50,000 0.42* 0.24, 0.75*     0.34* 0.16, 0.73*  7.71 0.89, 66.81 
Occupational prestige 0.77* 0.65, 0.91*  0.99 0.97, 1.02  0.80* 0.66, 0.98*  1.02 0.99, 1.05 
Decision latitude 0.97 0.81, 1.16  0.98 0.96, 1.02  0.99 0.98, 1.02  1.98* 1.26, 3.10* 
Skill discretion 0.99 0.83, 1.19  0.94 0.85, 1.04  0.98 0.92, 1.05  1.20* 1.03, 1.37* 
Decision authority 0.95 0.80, 1.14  0.98 0.90, 1.07  1.00 0.95, 1.06  1.85* 1.21, 2.85* 
Job demands 1.05 0.78, 1.41  0.96 0.91, 1.01  1.00 0.97, 1.04  0.99 0.93, 1.06 

* Significant (p ≤ 0.05).

† Adjusted for age, systolic blood pressure, the ratio of total cholesterol to high density lipoprotein cholesterol, body mass index, cigarette smoking, and diabetes.

References

1.
Karasek R. Job demands, job decision latitude, and mental strain: implications for job redesign.
Admin Sci Q
 
1979
;
24
:
285
–307.
2.
Karasek R, Baker D, Marxer F, et al. Job decision latitude, job demands, and cardiovascular disease: a prospective study of Swedish men.
Am J Public Health
 
1981
;
71
:
694
–705.
3.
Karasek RA, Theorell T, Schwartz JE, et al. Job characteristics in relation to the prevalence of myocardial infarction in the US Health Examination Survey (HES) and the Health and Nutrition Examination Survey (HANES).
Am J Public Health
 
1988
;
78
:
910
–18.
4.
Hammar N, Alfredsson L, Johnson JV. Job strain, social support at work, and incidence of myocardial infarction.
Occup Environ Med
 
1998
;
55
:
548
–53.
5.
Theorell T. Stress at work and risk of myocardial infarction.
Postgrad Med J
 
1986
;
62
:
791
–5.
6.
Falk A, Hanson BS, Isacsson SO, et al. Job strain and mortality in elderly men: social network, support, and influence as buffers.
Am J Public Health
 
1992
;
82
:
1136
–9.
7.
Reed DM, LaCroix AZ, Karasek RA, et al. Occupational strain and the incidence of coronary heart disease.
Am J Epidemiol
 
1989
;
129
:
495
–502.
8.
Pieper C, LaCroix AZ, Karasek RA. The relation of psychosocial dimensions of work with coronary heart disease risk factors: a meta-analysis of five United States data bases.
Am J Epidemiol
 
1989
;
129
:
483
–94.
9.
Alterman T, Shekelle RB, Vernon SW, et al. Decision latitude, psychologic demand, job strain, and coronary heart disease in the Western Electric Study.
Am J Epidemiol
 
1994
;
139
:
620
–7.
10.
Schnall PL, Pieper C, Schwartz JE, et al. The relationship between ‘job strain,’ workplace diastolic blood pressure, and left ventricular mass index. Results of a case-control study.
JAMA
 
1990
;
263
:
1929
–35.
11.
Alfredsson L, Spetz CL, Theorell T. Type of occupation and near-future hospitalization for myocardial infarction and some other diagnoses.
Int J Epidemiol
 
1985
;
14
:
378
–88.
12.
Bobak M, Hertzman C, Skodova Z, et al. Association between psychosocial factors at work and nonfatal myocardial infarction in a population-based case-control study in Czech men.
Epidemiology
 
1998
;
9
:
43
–7.
13.
Netterstrom B, Nielsen FE, Kristensen TS, et al. Relation between job strain and myocardial infarction: a case-control study.
Occup Environ Med
 
1999
;
56
:
339
–42.
14.
Lynch J, Krause N, Kaplan GA, et al. Workplace conditions, socioeconomic status, and the risk of mortality and acute myocardial infarction: the Kuopio Ischemic Heart Disease Risk Factor Study.
Am J Public Health
 
1997
;
87
:
617
–22.
15.
Bosma H, Marmot MG, Hemingway H, et al. Low job control and risk of coronary heart disease in Whitehall II (prospective cohort) study.
BMJ
 
1997
;
314
:
558
–65.
16.
Marmot MG, Bosma H, Hemingway H, et al. Contribution of job control and other risk factors to social variations in coronary heart disease incidence.
Lancet
 
1997
;
350
:
235
–9.
17.
Kuper H, Marmot M. Job strain, job demands, decision latitude, and risk of coronary heart disease within the Whitehall II Study.
J Epidemiol Community Health
 
2003
;
57
:
147
–53.
18.
Kuper H, Singh-Manoux A, Siegrist J, et al. When reciprocity fails: effort-reward imbalance in relation to coronary heart disease and health functioning within the Whitehall II Study.
Occup Environ Med
 
2002
;
59
:
777
–84.
19.
Schwartz JE, Pieper CF, Karasek RA. A procedure for linking psychosocial job characteristics data to health surveys.
Am J Public Health
 
1988
;
78
:
904
–9.
20.
Kannel WB, Feinleib M, McNamara PM, et al. An investigation of coronary heart disease in families. The Framingham Offspring Study.
Am J Epidemiol
 
1979
;
110
:
281
–90.
21.
New York City housing and vacancy survey—longitudinal file: 1999 technical documentation. (Machine-readable data file). Washington, DC: Housing and Household Economic Statistics Division, Bureau of the Census, 2000. (http://www.census.gov/hhes/www/housing/nychvs/1999/nychvs99.html).
22.
Davis JA, Smith TW, Hodge RW, et al. Occupational prestige ratings from the 1989 General Social Survey. (Computer file). National Opinion Research Center (Chicago, IL) (producer) and Inter-university Consortium for Political and Social Research (Ann Arbor, MI) (distributor), 1991. (http://webapp.icpsr.umich.edu/cocoon/ICPSR-STUDY/09593.xml).
23.
Eaker ED, Pinsky J, Castelli WP. Myocardial infarction and coronary death among women: psychosocial predictors from a 20-year follow-up of women in the Framingham Study.
Am J Epidemiol
 
1992
;
135
:
854
–64.
24.
Sorlie P. Cardiovascular disease and death following myocardial infarction and angina pectoris: Framingham Study, 20-year follow-up. Washington, DC: US Government Printing Office, 1977.
25.
Shurtleff D. Some characteristics related to the incidence of cardiovascular disease and death: 18-year follow-up. In: Kannel WB, Gordon T, eds. The Framingham Study: an epidemiological investigation of cardiovascular disease. Washington, DC: US Department of Health, Education, and Welfare, Public Health Service, National Institutes of Health, 1974. (DHEW publication no. (NIH) 74-599).
26.
Morgenstern H. Socioeconomic factors: concepts, measurement and health effects. In: Ostfeld A, Eaker ED, eds. Measuring psychosocial variables in epidemiologic studies of cardiovascular disease. Washington, DC: US Department of Health and Human Services, Public Health Service, National Institutes of Health, 1985.
27.
Faludi S. Backlash: the undeclared war against American women. New York, NY: Anchor Books Doubleday, 1991.
28.
Women in management: analysis of selected data from the Current Population Survey. Washington, DC: Government Accounting Office, 2001. (GAO-02-156).
29.
Eaker ED, Abbott RD, Kannel WB. Frequency of uncomplicated angina pectoris in type A compared with type B persons (the Framingham Study).
Am J Cardiol
 
1989
;
63
:
1042
–5.
30.
Garber CE, Carleton RA, Heller GV. Comparison of “Rose Questionnaire angina” to exercise thallium scintigraphy: different findings in males and females.
J Clin Epidemiol
 
1992
;
45
:
715
–20.
31.
Bassuk SS, Berkman LF, Amick BC 3rd. Socioeconomic status and mortality among the elderly: findings from four US communities.
Am J Epidemiol
 
2002
;
155
:
520
–33.