An increased risk of coronary heart disease has been reported in a large number of studies of professional drivers, 1–8 especially among bus drivers. Unfavorable life-style habits, chemical or psychosocial factors in the work environment, or a combination of several risk factors have been suggested as causal factors, with varying degrees of support from epidemiologic data. High demands and low decision latitude (job strain) in the psychosocial work environment is associated with an increased risk of myocardial infarction (MI), 9 which might be an explanation for the excess risk noted among professional drivers. Motor exhaust exposure has been proposed as a possible risk factor for myocardial infarction. 10 The literature on myocardial infarction among professional drivers has been reviewed by Belkićet al. 11,12
The Stockholm heart epidemiology program (SHEEP) is a community-based case-control study of causes of myocardial infarction (first event) among men and women in Stockholm County. The study’s objective is to evaluate the following risk factors: chemical, physical and psychosocial work environment factors; general environment factors; social factors; and individual factors including tobacco-smoking and alcohol-drinking habits. 13 Previous analyses of this study showed that high exposure to combustion products from organic material was associated with an increased risk of myocardial infarction, but motor exhaust exposure was not associated with risk of myocardial infarction in a consistent way. 14 The aim of the present study was to investigate other potential causes for the increased risk of myocardial infarction among professional drivers, and to study the risk among bus, taxi and truck drivers separately. We restricted the study to men, as there were very few infarctions among the female drivers.
Identification of Cases and Controls
The study base comprised all Swedish citizens living in Stockholm County age 45–70 years and free of previous MI. Cases were men who had their first MI in 1992 or 1993. From January to October 1992 the upper age limit for cases and controls was 65; this limit was extended to age 70 thereafter. Cases comprised all men with a first-time MI, whether fatal (death within 28 days) or nonfatal, and diagnosed according to specified criteria using information on symptoms, ECG, enzymes and autopsy findings. 15 We identified cases from three sources: the medical care units at the 10 emergency hospitals within the Stockholm County (87% of the cases), other hospital units (1% of the cases, obtained from a computerized hospital discharge register) or death certificates from the Causes of Death Register at Statistics Sweden (12% of the cases). We selected controls randomly from the study base through a computerized population register, stratified for sex, 5-year age group, hospital catchment area and year of enrollment in the study (1992 or 1993). Detailed information about the procedure for identification of cases and controls has been reported elsewhere. 13 The number of identified and responding cases and controls, and vital status among included individuals, is presented in Table 1. The study was approved by the ethics committee of the Karolinska Institutet. All study subjects were included after providing informed consent.
We collected exposure information by postal questionnaires, supplemented by a telephone interview in case of incomplete answers. For fatal cases, the questionnaires were completed by next of kin. The questionnaires covered the physical and psychosocial work environment, social factors and personal life-style factors, as well as a lifetime occupational history including occupation, work tasks and company name and address for all jobs held for at least 1 year. We also invited nonfatal cases and a similar proportion of controls to a health examination at 3 months after either the onset of disease (cases) or inclusion (controls). The examination included blood samples and recording of blood pressure, height and weight. Data from the questionnaire/interview and health examination were combined, if applicable, when classifying subjects into exposure categories. We coded occupations according to the Nordic version of the international classification of occupations. 16 An occupational code was assigned to every job held for at least a year. There were 147 cases and 129 controls who had ever worked as a driver (for at least 1 year). These included 77 bus drivers, 78 taxi drivers and 179 truck drivers; 53 persons had worked in two of these categories, and 5 had worked in all three.
We classified smoking habits from age 15 in five categories: never-smokers, ex-smokers and current smokers smoking 1–10, 11–20 or >20 gm of tobacco/day. Persons who had stopped smoking less than 2 years before inclusion in the study were considered as current smokers. We classified alcohol habits in five categories: persons who never drink alcohol, and four groups according to average consumption during the period the person used alcohol: 1–10, 11–30, 31–50 and >50 gm of alcohol/day. We coded subjects who reported no regular exercise other than occasional walks during the last 5–10 years as physically inactive at leisure time. We calculated body mass index (BMI) from information on weight and height (BMI = weight [kg] per height squared [m2]); if the BMI exceeded 27, the man was classified as overweight. We coded subjects with a history of drug- or diet-treated diabetes as diabetics. We assessed the presence of hypertension from information on the use of antihypertensive drugs, or a systolic blood pressure exceeding 180 mmHg or a diastolic pressure exceeding 100 mmHg at the health examination. Blood laboratory data were available only for the nonfatal cases; because of a high proportion of missing data we did not include blood lipids in the standard set of confounding variables.
We determined the average exposure to job strain during the last 5 years before recruitment or retirement by summering the scores of five questions on psychological demands at work and six questions on decision latitude according to the Karasek-Theorell model. 17 Job strain was considered present among those with both high demands (over 75th percentile) and a low decision latitude (below 25th percentile). 18 We coded socioeconomic status (in six groups) according to occupation 10 years before inclusion in the study. To keep the number of variables in the regression model as low as possible, these six groups were collapsed to three, based on the group-specific adjusted relative risk for MI: low risk (high-level employees and agricultural workers), intermediate risk (low- and middle-level employees and self-employed) and high risk (manual workers). We used indicator variables to represent these three groups. Age distribution and basic characteristics of the professional drivers are shown in Table 2.
We estimated the odds ratio (OR) of developing MI, and 95% confidence intervals (CIs), by unconditional logistic regression. The OR in each category of driver (bus, taxi and truck drivers) was calculated using all men who never worked as a driver at all as “unexposed.” We adjusted all analyses for the stratification factors used in the selection of controls: age (5-year age groups), hospital catchment area (10 areas), and year of enrollment in the study (1992 or 1993). The confounders were introduced in the regression model in steps to evaluate confounding from different types of exposures separately. First, we adjusted the ORs for socioeconomic status. Secondly, we adjusted for exposures not directly related to the work environment, including smoking habits, alcohol habits and physical inactivity at leisure time as well as the medical/metabolic factors of obesity, diabetes and hypertension. Finally, we adjusted the ORs for job strain. We used indicator variables corresponding to the categories described above for all exposures. We also calculated the OR of MI among drivers subdivided by duration of work (1–10 years, >10 years) and number of years since having stopped working as a driver (current, 2–20 years, >20 years). The cutoff for categories of work duration and time since having stopped driving was based on the number of drivers, aiming at an equal distribution among the categories.
The distribution of the potentially confounding factors in the various driver groups as well as among all men in the study are presented in Table 2. Smoking was more common among the drivers than in the general population; among the controls, 47% of the taxi drivers, 49% of the truck drivers and 32% of the bus drivers were current smokers, compared with about 30% in the general population. Being overweight was more common in bus drivers (52%) in particular, but also taxi (32%) and truck drivers (36%), compared with men in the general population (29%). Low physical activity at leisure time was characteristic for the controls in all three driver groups, especially for taxi drivers (62%), compared with 34% in the population. There was a tendency to lower alcohol consumption among bus drivers compared with other drivers and nondrivers. The exposure to high demands and low decision latitude (job strain) was slightly more common among drivers (bus drivers, 10%; taxi drivers, 6%; truck drivers, 7%) than among other men (4%).
The adjusted ORs for the variables listed in Table 2 are presented in Table 3. Smoking was the strongest risk factor for MI, followed by diabetes mellitus and overweight status, as well as job strain.
The crude OR for development of MI was more than doubled among bus drivers, about doubled among taxi drivers and slightly lower but still elevated among truck drivers, as shown in Table 4. The ORs decreased somewhat after adjustment for socioeconomic group but were still clearly elevated in all driver groups (“adjusted A” in Table 4). Additional adjustment for smoking, alcohol, physical inactivity at leisure time, overweight status, diabetes and hypertension further lowered the ORs (“adjusted B”). Among bus drivers and taxi drivers there was still a tendency to an increased risk, but among truck drivers the OR was close to unity after this adjustment. Including job strain in the regression model (“adjusted C”) generally lowered the risks only slightly.
To further investigate the effect of job strain, we included the job strain quotient (the ratio of the sum score of demand to the some score of control) as a continuous variable in the regression (data not shown). However, this variable had no additional effect on the risk estimates for the drivers. An extended model using a larger set of variables to adjust for smoking (including continuous variables for time since stopping smoking among ex-smokers, and the average consumption of tobacco per day among current smokers) showed no indication of residual confounding by smoking habits. Furthermore, there was no change in the risk estimates for the driver groups when duration of smoking was added to the regression model. Additional analyses with the original six groups of socioeconomic status did not change the results.
Analyses restricted to nonfatal cases showed that the adjusted ORs (in a model corresponding to “adjusted B” in Table 4) for MI in the various driver groups were slightly lower for bus drivers (OR = 1.38; CI = 0.79–2.39) and for taxi drivers (1.24; 0.71–2.16), but slightly higher for truck drivers (1.15; 0.79–1.66).
The association of MI risk among professional drivers with work duration, used as a proxy for cumulative exposure, is shown in Table 5. An exposure-response pattern was evident for bus drivers, with an almost two-fold adjusted OR among those who had been working more than 10 years as a bus driver but only slightly increased OR among those who had been working 1 to 10 years. A similar but less pronounced trend was present among taxi drivers. Among truck drivers there was no evidence of increased risk with longer exposure.
If professional driving is associated with an increased risk of MI and the effect is reversible, then the risk would be expected to decline after stopping working as a driver. Table 6 shows the estimated relative risk among drivers subdivided by number of years since having stopped working as a driver. Among bus drivers, the risk was highest in the intermediate class (stopped driving 2–20 years ago). There was no apparent excess risk among bus drivers who currently were driving, or among those who had stopped more than 20 years ago. Current taxi drivers had an increased risk of developing MI, but no increased risk was noted for taxi drivers who had stopped working more than 2 years ago. The risk was slightly increased among truck drivers who had stopped driving 2–20 years ago, but no excess risk was noted among those who currently were truck drivers.
Because some of the drivers had been driving more than one vehicle, we also investigated the risk among those who had been bus, taxi or truck drivers only. This reduced the group sizes to 48 bus drivers, 39 taxi drivers and 138 truck drivers, and gave slightly reduced risks in all driver groups. A model corresponding to “adjusted B” in Table 4 gave ORs of 1.43 (CI = 0.76–2.67) for bus drivers, 1.21 (0.62–2.35) for taxi drivers and 0.97 (0.67–1.40) for truck drivers. However, an exposure-response pattern was still evident for bus drivers (OR = 0.94 [CI = 0.39–2.29] for 1–10 years and 2.21 [0.87–5.64] for >10 years) and to a lesser extent for taxi drivers (0.92 [0.33–2.59] for 1–10 years and 1.46 [0.61–3.48] for >10 years).
The present study showed a high risk of MI among male professional drivers in Stockholm 1992–1993. The risk excess was most pronounced among bus and taxi drivers, both with and without adjustment for individual risk factors. Because an occupational factor such as stress may exert its effect through mechanisms such as hypertension or metabolic changes, the adjustment for hypertension, diabetes and BMI may introduce an “overadjustment” if the purpose is to evaluate the influence of occupational factors alone. Thus, both the crude risk and the various adjusted risks should be considered when evaluating the effect of occupational factors. An advantage with this study was that the occupational information covered the whole lifetime up to the time of inclusion.
Low socioeconomic status is associated with an increased risk of MI, 19 and adjustment for socioeconomic status partially reduced the risks among professional drivers. Uncontrolled and residual effects of socioeconomic status seem unlikely because socioeconomic status did not differ much among the driver groups, and therefore can not explain the higher risk of MI among bus and taxi drivers than among truck drivers. The drivers smoked more than the general population, but the excess risk of MI that was observed among bus and taxi drivers was only partially explained by their tobacco smoking habits. An extended regression model incorporating time since stopping smoking and smoking intensity, as well as duration of smoking, did not alter the risk estimates, indicating that residual confounding from tobacco smoking is unlikely. Drivers were more often overweight than the general population, but the cases among bus drivers were less obese than the cases in general. We have no explanation for this pattern, but the numbers were small and so it may be a chance finding. The proportion of drivers with diabetes and hypertension did not differ much from the general population, although again numbers were small. Thus, the risk remained high among bus and taxi drivers but not among truck drivers when adjusting for socioeconomic, life-style and metabolic risk factors. We conclude that the increased risk of MI among bus and taxi drivers may be explained only partially by an overrepresentation of individuals with these risk factors.
Earlier epidemiologic studies show that stress affects the development of coronary heart disease, and occupational stress has been suggested as an explanation for the high cardiovascular morbidity and mortality in drivers. 4,20,21 In the present study, perceived job strain was more common among bus, taxi and truck drivers than among other men, although the proportion of exposed persons was small. Job strain is thought to affect the risk of cardiovascular disease through physiologic changes such as increases in blood pressure and blood levels of neuroendocrine hormones (adrenaline, noradrenaline, cortisol). 20 Applying job strain at the final step in the regression may have underestimated the effect slightly. However, the specific mechanisms by which job strain contributes to cardiovascular disease are still not certain. Job strain in terms of demand/control explained only a small part of the risk among bus and taxi drivers after adjustment for the other risk factors. It is possible, however, that this way of measuring stress does not adequately reflect the aspects of job stress that are relevant for the drivers.
The results indicate an increased risk of developing MI among bus drivers who stayed more than 10 years in the profession of driving. A similar but less pronounced tendency was present among taxi drivers. This exposure-response pattern would not emerge if the risk excess was caused by a selection of individuals with unfavorable life-style into the driving occupation. Among taxi drivers, there was a pronounced excess in risk only among current drivers and drivers who had stopped working less than 2 years ago, pointing towards an effect of a very short duration. The reason why bus drivers who stopped driving 2–20 years ago had a higher risk than the group including current drivers is unclear. These apparently contradictory patterns may be caused by random variation because numbers are small, but may possibly be explained by differences in the type of employment contract for bus and taxi drivers. Bus drivers are employees of large companies whereas taxi drivers often are self-employed. The bus drivers’ working hours are fixed, and drivers have few opportunities to adjust the job to the health situation. MI is often preceded, sometimes for many years, by symptoms of coronary insufficiency (angina pectoris). Bus drivers may more often have to leave their job if they have symptoms of preliminary stages of coronary heart disease. Taxi drivers, on the other hand, might instead adapt their working hours to these symptoms and stay in service.
Our findings are in accordance with results from earlier studies. In the Oslo study 22 bus, taxi and tram drivers had higher total and coronary heart disease mortality rates compared with other occupational groups; these results were partly attributed to an unfavorable risk factor profile. Rosengren et al. 4 found an increased risk of coronary heart disease among 103 middle-age bus and tram drivers in Gothenburg, compared with men from other occupational groups, with the risk being independent of standard risk factor status. After a mean of 11.8 years of follow-up, the adjusted odds ratio among bus and tram drivers was 3.0 (CI = 1.8–5.2) for incident coronary heart disease. In the same study taxi drivers tended to have an increased risk, but no excess risk was found among truck drivers. In Sweden, occupations in transport work, and among them professional drivers, constitute one of the occupational groups with the highest incidence of MI. 23 In a cohort study on the whole working population of Denmark, bus drivers and taxi drivers had excess risks of being admitted to hospital with ischemic heart disease. 7 In a recent study of professional male drivers in Denmark the standardized hospital admission ratios for diseases of the circulatory system were higher among professional drivers than in the male working population and were higher for drivers of passenger transport than for drivers of goods vehicles. 24 In both driver group comparisons, the standardized hospital admission ratio for acute MI increased with time.
In conclusion, an excess risk of MI was found among male professional drivers in Stockholm County in 1992–1993. Adjustment for unfavorable life-style factors and social factors reduced the odds ratios, although bus and taxi drivers, but not truck drivers, still tended to have an increased risk. An exposure-response pattern (in terms of duration of work) was found for bus and taxi drivers. Job strain assessed according to the demand/control model seemed to explain only a small part of the increased risk among bus and taxi drivers. Previous analyses of this study indicated that motor exhaust exposure was not associated with risk of MI. 14 It seems possible that factors in the work environment for urban bus and taxi drivers contribute to their increased risk of MI, but among truck drivers individual risk factors seem to explain most part of the elevated risk. Further investigations of the psychological demands in various driver groups are warranted, especially if the transportation of passengers is more stressful than the transportation of goods.
We thank Anders Ahlbom for coordinating the SHEEP study, Annika Gustavsson and Irene Samuelsson for management of data collection and databases, and the physicians and nurses at the medical departments for identifying and enrolling cases and performing health examinations.
1. Morris JN, Heady JA, Raffle PAB, Roberts CG, Parks JW. Coronary heart-disease and physical activity at work. Lancet
2. Hedberg G, Jacobsson KA, Langendoen S, Nyström L. Mortality in circulatory diseases, especially ischemic heart disease, among Swedish professional drivers: a retrospective cohort study. J Hum Ergol (Tokyo) 1991; 20: 1–5.
3. Alfredsson L, Hammar N, Hogstedt C. Incidence of myocardial infarction
and mortality from specific causes among bus drivers in Sweden. Int J Epidemiol 1993; 22: 57–61.
4. Rosengren A, Anderson K, Wilhelmsen L. Risk of coronary heart disease in middle-aged male bus and tram drivers compared to men in other occupations: a prospective study. Int J Epidemiol 1991; 20: 82–87.
5. Michaels D, Zoloth SR. Mortality among urban bus drivers. Int J Epidemiol 1991; 20: 399–404.
6. Netterstrøm B, Suadicani P. Self-assessed job satisfaction and ischaemic heart disease mortality: a 10-year follow-up of urban bus drivers. Int J Epidemiol 1993; 22: 51–56.
7. Tüchsen F, Bach E, Marmot M. Occupation and hospitalization with ischaemic heart diseases: a new nationwide surveillance system based on hospital admissions. Int J Epidemiol 1992; 21: 450–459.
8. Gustavsson P, Alfredsson L, Brunnberg H, et al
. Myocardial infarction
among male bus, taxi, and lorry drivers in middle Sweden. Occup Environ Med 1996; 53: 235–240.
9. Belkić K, Landsbergis P, Schnall P, et al
. Psychosocial factors: review of the empirical data among men. In: Schnall PL, Belkić; K, Landsbergis P, Baker D, eds. The Workplace and Cardiovascular Disease.
Philadelphia: Hanley & Belfus, Inc., 2000;24–46.
10. Stern FB, Halperin WE, Hornung RW, Ringenburg VL, McCammon CS. Heart disease mortality among bridge and tunnel officers exposed to carbon monoxide. Am J Epidemiol 1988; 128: 1276–1288.
11. Belkić; K, Savić; C, Theorell T, Rakić; L, Ercegovac D, Djordjević; M. Mechanisms of cardiac risk among professional drivers. Scand J Work Environ Health 1994; 20: 73–86.
12. Belkić; K, Emdad R, Theorell T. Occupational profile and cardiac risk: possible mechanisms and implications for professional drivers. Int J Occup Med Environ Health 1998; 11: 37–57.
13. Reuterwall C, Hallqvist J, Ahlbom A, et al.
Higher relative, but lower absolute risks of myocardial infarction
in women than in men: analysis of some major risk factors in the SHEEP study. The SHEEP Study Group. J Intern Med 1999; 246: 161–174.
14. Gustavsson P, Plato N, Hallqvist J, et al.
A population-based case-referent study of myocardial infarction
and occupational exposure
to motor exhaust, other combustion products, organic solvents, lead, and dynamite. Stockholm Heart Epidemiology Program (SHEEP) Study Group. Epidemiology 2001; 12: 222–228.
15. Möller J, Hallqvist J, Diderichsen F, Theorell T, Reuterwall C, Ahlbom A. Do episodes of anger trigger myocardial infarction
? A case-crossover analysis in the Stockholm Heart Epidemiology Program (SHEEP). Psychosom Med 1999; 61: 842–849.
16. Nordic Standard Classification of Occupations (in Swedish).
Stockholm, Sweden: Swedish National Labour Market Board, 1983.
17. Karasek R, Theorell T. Healthy Work. New York: Basic Books, 1990.
18. Hallqvist J, Diderichsen F, Theorell T, Reuterwall C, Ahlbom A. Is the effect of job strain on myocardial infarction
risk due to interaction between high psychological demands and low decision latitude? Results from Stockholm Heart Epidemiology Program (SHEEP). Soc Sci Med 1998; 46: 1405–1415.
19. Marmot M, Theorell T. Social class and cardiovascular disease: the contribution of work. Int J Health Serv 1988; 18: 659–674.
20. Evans GW. Working on the hot seat: urban bus operators. Accid Anal Prev 1994; 26: 181–193.
21. Schnall PL, Landsbergis PA. Job strain and cardiovascular disease. Ann Rev Public Health 1994; 15: 381–411.
22. Holme I, Helgeland A, Hjermann I, Leren P, Lund-Larsen PG. Coronary risk factors in various occupational groups: the Oslo study. Br J Prev Soc Med 1977; 31: 96–100.
23. Hammar N, Alfredsson L, Smedberg M, Ahlbom A. Differences in the incidence of myocardial infarction
among occupational groups. Scand J Work Environ Health 1992; 18: 178–185.
24. Hannerz H, Tüchsen F. Hospital admissions among male drivers in Denmark. Occup Environ Med 2001; 58: 253–260.