Risk indicators of ischemic heart disease among male professional drivers

Riskindicators of ischemicheart disease amongmaleprofessional driversin Possible risk indicators of ischemic heart disease relevant to the occupation of professional driving were identified in a cohort of 440 professionaldrivers and 1000referents from the Swedishcounties of Vasterbotten and Norrbotten. The subjects were randomly selected. Data on cardiovascular risk indicatorswerecollectedfrom questionnaires, bloodpressuremeasurements, serumlipid levels,height, and weight.The results showedthat significantlymoredrivers than referents were overweight,smok ers, and shift workers;were sedentaryin their leisure time; and had a work situationcharacterizedby high demands, low decision latitude, and low social support. There were no significant differences concerningblood pressure and serum lipid levels.The odds ratio for having a high score on a cardio vascular risk index was 3.18 (95% confidence interval 2.41-4.20) for the drivers when they were compared with the referents. When adjusted for age, heredity, shift work, educational level, marital status, and workingclass, the odds ratio was 2.34 (95%confidenceinterval 1.70-3.21).

Several investigations have indicated that cardiovascular diseases affect professional drivers more often than many other occupational groups. One of the earliest studies in this field was conducted by Morris et al (I). In an investigation of 667 middle-aged busmen (2) a higher incidence of coronary heart disease was found among bus drivers than among bus conductors. A study of the relationship between mortality and occupation in the Nordic countries in the lO-year period 1971-1980 (3) showed increased mortality from cardiovascular diseases among road transport workers in all Nordic countries when they were compared with all economically active men. A Danish study of bus drivers (4) showed that mortality due to coronary heart disease was significantly higher among bus drivers than among a reference group representing employed men in Copenhagen. An investigation into mortality from circulatory diseases, especially ischemic heart disease in a cohort of 1713 Swedish professional drivers (5) showed that the standardized mortality ratio (SMR) for circulatory diseases and ischemic heart disease were significantly higher in a group of professional drivers than in a reference group of Swedish men (SMR 127 and 138, respectively). A Norwegian study of coro-nary risk indicators in various occupational groups (6) showed that, among road transport workers, the coronary risk score was high because of the workers' high serum cholesterol levels and because many of the workers smoked. Taxi and bus drivers with a low income and a poor education had a particularly high risk score. In a study of coronary risk indicators by Hartvig & Midttun (7) bus and truck drivers were compared with industrial workers. The results showed higher mean values for serum cholesterol, serum triglyceride, and blood pressure in the driver group and therefore indicated a markedly higher risk of cardiovascular disease for that group.
The purpose of our investigation was to identify possible risk indicators of ischemic heart disease relevant to the occupation of professional drivers so that a preventive strategy can be devised against the development of ischemic heart disease in this occupational group.

Study base
The driver group consisted of male drivers from the Swedish counties of Vasterbotten and Norrbotten. In the two counties there were altogether 903 bus drivers and 2746 truck drivers. Of the drivers, 12% was randomly selected from lists of members of the national unions to which drivers belong. To be selected, the driver had to have been a member of the union for at least three years. The limit of three years was chosen for the study to exclude drivers with only short work experiences . The research group consisted of 110 male bus dri vers and 330 male truck dri vers between 25 and 64 years of age.
The reference group included all men in the MON-ICA (multinational monitoring of determinants and trends in cardiovascular diseases) study of the World Health Organization (WHO) in the northern Swedish MONICA center, which covered the counties of Vasterbotten and Norrbotten. In the MONICA study (8) determinants and trends in cardiovascular disease are being monit ored during a lO-year period in 40 centers in 27 countries. The study is exploring whether the differences in cardiovascular risk factor profiles in different geographic settings are related to the incidenc e of cardiovascul ar diseases and case fatality rates. The reference group consisted of 1000 male subjects from the count ies of Vasterbotten and Norrbotten. Stratified randomization from the total population between 25 and 64 years of age was used to select these subjects. The stratification included the following age groups: 25-34, 35-44, 45-54 and 55-64 years. In each stratum 250 men were randomly selected from continuously updated population registers.
The mean age of the dri vers was 43.8 (95% CI 42. 8-44.8) years. In figure I the age distributi on of the drivers is shown in relation to the age distribution of the male population in the counties of Vasterbotten and Norrbotten. The proportion of drivers was lower chiefly in the age group 60-64 years. Within the driver group 78% were married or cohabiting, while the figure was 82% for the reference group.
Regard ing educational background 80% of the drivers had only completed compulsory schooling (seven years for the oldest and nine years for the youngest groups), 19% had attended senior high school, and I % had attended a university. For the reference group the corresponding figure s were 60, 28, and 12%, respectivel y. The difference between the driver and reference group s was significant.
Seventy-five percent of the drivers and 46% of the referents belonged to the workin g class. Fourteen percent of the drivers and 10% of the reference group ran their own busine ss. The difference in the distribution according social class was signifi cant.
For all of the drivers the mean number of years as a professional driver was 18.2 (95% CI 17.0-19.4). Significantly more drivers (63%) than referents (32%) worked shift s or had variable workhours .

Study design
The subjec ts were invited by mail to participate in the investigation , which consi sted of the follo wing two parts: (i) to answer two que stionn aires at home (questionnaires 1 and 2) and (ii) to attend a screening in which measurements were taken and an additional questionnaire (number 3) was answered. If the subject did not keep his first appointment, he was given another by mail or by phone. Since the investigation was coordinated with the WHO MONICA study (8), the same procedure was used as in that investigation. Two team s, each consisting of two nurses and two medical technicians, performed the screenings, carried out between Janu ary and April 1990 in the counties of Vasterbotten and Norrbotten.

Questionnaires and measurements
Questionnaire 1 comprised items regarding possible medical history of cardiovascular disease, smokin g habits, physical activity during leisure time and work, eating and alcohol habits, and work conditions. Questionnaire 2 included items concerning occupation and tasks. Questionnaire 3 dealt with socia l support during leisure time. When items of information were missing, the questionn aires were completed by the respondent on the occasion of the exa mination.
Blood pressure levels were measured twice by the random zero method (9)   Per mille was calculated. Before the screening started, training sessions with the two teams were conducted to ensure conformity.
Serum samples were obta ined for the determin ation of total cholesterol and high-den sity lipoprotein (HDL ) after fasting for a minimum of 4 h. Serum samples were also obtain ed for the determination of triglyceride after fasting for a minimum of 12 h. Because of the long period of fasting, triglyceride was measured for only 65% of the subjects, randomly selected. Total cholesterol was measured by the enzymatic method and HDL cholesterol by phosphotungstate-Mfl " precipitation with the use of commercial kits (Boehringer Mannh eim).
Weight was measured with an electronic balance, which was calibrated daily. During the weighing, the subjec ts wore only a shirt and trou sers and no shoes; no corrections were made. Height was also measured without shoes. Body mass index was calculated as weight (kg) . length ? (m).

Analysis of nonrespon dents
A total of 89% of all drivers answered the postal questionnaires. The screening investigation was attended by 85%. Four percent could not participate in the screening, mainly due to their workhours. Most of the nonrespond ents in the screening belonged to the 25-to 34-year age group (16%). The dropout was 7% in the 35-to 44-year age group , 14% in the 45to 54-y ear age group, and 13% in the 55-to 64-year age group.
Seventy-eight percent of the reference group answered the questionnaires and participated in the screening. Most of the nonrespond ents were found in the younges t age group (38%). In the 35-to 44year age group the dropout rate was 2 1%, whereas in the 45-to 54-year and 55-to 64-year age groups it was 2 1 and 20%, respectively.

Data p rocessing and statistics
We used the Swedish socioeconomic classification (\0) in determining the distribution accord ing to social class.
A measure of fat consumption was constructed by multipl ying the daily number of glasses of milk by their aver age fat content.
The responses to the five questions dealing with social support at work were used to calculate an index for this variable ( I I). The higher the index, the better the social support at work.
An index concerning the qual ity and quantity of social support during leisure time was calculated accordin g to a model by Henderson et al ( I I) from the respon ses to the 13 questions dealing with this variable . The higher the index, the better the social support.
Ten questions dealt with the psychosocial work environment according to a model by Karasek & Theorell (12). The model describe s the jo b situation of a worker by means of the demands made and the decision latitud e given durin g work. From the answers to these questions indices were calculated with which to group the subjects into the following four different job situations: active (high demands and high decision latitude), low strain (low demand s and high decision latitude), high strain (high demands and low decision latitude), and passive (low demand s and low decisi on latitude ).
The significa nce of the difference between two distribution s was tested with the chi-square test or the analysis of variance for continuous variables, and for the difference between the mean values the t-test was used. The 5% level was chosen as the critical level of significance . The 95% confidence inter vals (95% CI) have also been reported. To eliminate differences between the driver and reference groups due to different age structures within the group s, we standardized the data for age with the age structure of the drivers as reference.
A cardiovasc ular risk index was calculated to sum up the different risk indicators (risk indicators concerning medical aspects, life-style, and job situation).
The score +1 was given to each of the following five risk indicators: hyperten sion [medical treatment or diastolic blood pressure~90 mm Hg (~1 1 . 97 kPa)], high fat consumpt ion (over median), daily tobacco use, low leisure-time physical activity, and fin ally a j ob situation characterized by high demand s and low decision latitud e (high strain) ( I I). The score could thus vary between 0 and 5. The risk score was then dichotomized so that those with a risk score above the median had a score of 1 and all others had a score of O. The odds ratio (OR) was calculated. Adjustment for potenti al confounders was made with the logistic regression. Heredit y (a close relative having died of cardiovascular disease), marital status (single or widow/wi dower ), age (~4 5 years old), shift work, educational level (only compulsory schooling), and working class membership were regarded as confounders.
The study was approved by the regional Ethics Research Committ ee of the University of Urnea. The data proce ssing procedures were approved by the National Compute r Data Inspection Board .

Risk indicators
The prevalence of ischemic heart disease, stroke, and diabetes did not differ significantly between the two groups. Significantly more drivers than referents had close relativ es who had died of cardiac infarction before the age of 65 years. Concerning mortality from stroke, there was no significa nt diffe rence between the two groups.
For the drivers the mean value was 130 mm Hg ( 17 kPa) (95% CI 128-1 32) for systolic blood pressure and 83 mm Hg (95% CI= 82-84) for diastolic Table 1. Srnokinq among th e profession al drivers and referents . Table 2. Levels of occupat ion al physi cal activity repor ted by the professional drivers and referents .
The dri vers averaged 6.3 (95% CI 5.9-6.6) cups of coffee per day, and the referents had 5.5 (95% CI 5.3-5.8) cups . The differ enc e was significant. The drivers also co nsumed a significantly higher proportion of milk fat per day [mean 7 .69 (95% CI 7.0-8.4) g) than the referents [mean 5.73 (95% CI 5.3-6.2) g). The rem ain ing dieta ry data will be presented in a spec ial article . There were no differe nces between the drivers and the referents concerni ng alcohoi consumptio n.
Th e mean value of the body mass index was 26. 3 (95 % CI 26.0-26.7) for the drivers and 25 .9 (95% CI = 25.6 -26.1) for the reference group. The difference bet ween the two groups was sig nificant. For the dr ivers the mean bod y mass index increased with advancing age up to 55-64 years of age (figurc 3). Level of physica l activity blood pressure. There were no significant differences bet ween the drivers and refe rents regard ing the mean values of blood pre ssure or regarding treatment for hyperten sion. Th e drivers had a mean value of 6.32 (95 % CI 6.19 -6.45) mmol . I-I for total cholestero l in ser um, and the corresponding valu e for the referents was 6.30 (95% CI 6.22-6.39) mmol . [-I. The mean values increas ed with age up to 54 yea rs of age. Th e mean value for the dri vers decre ased in the 55-to 64-year age group ( figure 2).
Th e mean value for HDL cholesterol was 1.26 (95 % CI 1. 23 A comparison between the drivers and the referent s showed a significantly higher share of smokers amon g the drivers (table I ). There was no significant difference between the two groups with respect to the use of snuff. Table 2 shows the se lf-reported physical activity at work. Marc dri vers than referents had seden tary and heavy work . The opp osite was true for light and some flexible work. The difference in the distribution of the responses wa s sig nificant.
The fre quency of dri vers reg ularly engaging in moderate ph ysical activity during leisure time increased with age. Th ere was a significa nt difference bet ween the drivers and the referents rega rding leisure-time physical activ ity. More men in the reference group had a higher degree of phy sical act ivity (

Body Mass Index
The questions concerning social support at work referred to the current j ob. The results of the calculated index showed that the drivers had significant-Iy less social support at work than the referents (table 4) . The results of the calc ulated index concerning social support during leisure time showed a significant differenc e between the drive rs and the referents as regards quantitati ve but not qualitati ve social support (table 4) .
Significantl y more drivers (15%) than referents (7 %) reported that they lifted, carried, or dragged heavy items several times an hour. Significantly more drivers than referent s also thought that the job involved physical strain and mental stress.
The results of the analysis of the relationship between the psychological demands and decision latitude in current job showed that more professional drivers (40%) than referents (18%) had a work situation characterized by high strain (high demands and low decision latitude) (figure 4) . The difference between the distri bution of the driver s and the referents was significant for the four job situations. Table 5 shows the occurrence of the risk indicators included in the cardiovascu lar risk index . The odds ratio for having a high score on the cardiovascular

Discussion
The drivers represented professional dr ivers who had been members of the national union s for at least three years . T here may have been a few dr ivers who were not members of any union. Howe ver, it see ms plausible to assume that in such cases the se men work as dri vers only on a casual basis. Since the sele ction wa s car ried out acc ording to the cri terion of thr ee years of experie nce, determ ined from union membership , the numb er of missing dri vers was probably low. Some of the driv ers ow n of their own trucks as one-m an fir ms. These drivers are members of the union of haulage contrac tors , which was a part of the basi s for the selection. The two groups of subj ect s were obtained by random se lection. Both gro ups were inves tigated with the same meth od s and by the sa me research tea m. The result s obt ained therefore ca nnot be expl ained by differences in the method s of selec tion or measuremen t. Th ere were few nonrespond ent s. Th ere were also onl y small differences in the numb ers of nonrespondents among bu s and tru ck drivers and amon g the driv ers from the two counties and among the age gro ups . Th erefore there is no reason to believe that nonr espond en ts have affec ted the results in any parti cul ar direction .
Th e proportion of dri vers in rel ation to the male popul at ion decreases chiefly in the age group 60-64 years (figure I). This decrease co uld be an effect of a health-based dropout from the job of professional driver ("healthy worker effect" ) as has been shown in an ear lier study of health among professional drivers ( 13). In a study of factors influencing turn over among Swedish professional drivers (14), older drivers ofte n gave health factors such as car diovascular diseases and musculoskeletal co mplain ts as reasons for leaving the trade. In this study the reference group was a sample from the tot al population and thu s included both actively wor king men and men ou t of wo rk. Th erefore the fig ures given may be an underestimation of the risk indic ators in a popul ation co nsisting of only actively working men.

Risk indicators
Th e result s of this study showe d that man y professional drivers did sedenta ry or hea vy work and were not physicall y active during their leisure time. Researc h results are contradictory co ncerning the effect of physical activity at work on the pre vention of cardiovascular diseases. Th e study by Morris & Crawford ( 15) pointed out that a high level of physical activity at work is a factor whic h preven ts cardi ovasc ular disease. A cla ssic study of the incid ence of ischemic he art disease in ma le bus drivers and co nductors by Morris et al (2) showed a higher incidence among bus drivers. Th e authors thought that this was partly a con sequ ence of less phy sical activity at work among the dri vers. A study by Holme et al (16) has shown an increa se in mortality from coronary heart disease with increasing physical activ ity during work. On the other hand the coro nary heart disease mortality decre ased as the degree of leisure-time phys ica l activity increased. One reason for this differenc e may be the type of phy sical acti vity involved. Leisure-time phy sical activity often includes aerobic train ing, while physical activ ity at wor k -especi ally the work of tru ck drivers ( 17, 18) -ca n invo lve much heavy lifting and draggin g co mbined with per iod s of sede ntary wor k. In thi s study about one-th ird of the truck drivers rep orted that they lifted , carried, or dragged heavy objects once or several times an hour. Heavy lifts and hea vy dragging work involve both static and dyn amic elem ent s, for exa mple, when driv ers carry or drag weights. A significa nt increase in both systolic and diastolic blo od pressure can be related to static muscle co ntractions, especially if the cont ractions are force d to exhaustio n. If static work is co mbined wit h dynamic work, the increase in the blood pressure will occur from a higher level ( 19). According to Folkow (20) frequently repeated pressor bouts every day over a long period ca n give rise to cardi ovas-cular structural adaptations that can lead to permanent high blood pressure. However, in thi s study, there was no difference between the drivers and the referents concerning blood pressure.
Many drivers were overweight. Investigations have shown that obesity can be related to the development of cardiovascular disease either as a precursor to the development of other coronary risk indicators or as a risk factor in itself (21,22) . The exce ss proportion of those overweight among the dri vers could have been a con sequence of their eating habits and their low level of physical activity both at work and during their lei sure time . Many drivers (63%) worked shifts or had variable workhours and were often forced to eat out.
Several investigations have shown that a poor social network and a lack of social support increases the risk of cardiovascular disease (23 , 24). A 10year follow-up of middle-aged men showed that social isolation is as important a risk indicator as traditional clinical risk indicators (25). Ischemic electrocardiographic changes were four time s more frequent among men with poor emotional support than amon g men with good emotional support. This relationship was independent of other risk indicators and earlier pre valence of cardiovascular disease. Among the drivers in our study the social support at work was significantly lower than in the reference group. On the other hand there were only small differences between the groups concerning social support during leisure time. In this case the quantitative support, which is of less importance for cardiovascular diseases, was somewhat better in the dri ver group.
In this investigation the work of many of the drivers was characterized by a combination of high demand s and low deci sion latitude (high strain). According to a model by Kara sek & Theorell ( 12) such a work situation leads to a high er risk of coronary heart disea se. In an evaluation of the model in a study on 1461 Swedish men, it was shown that symptoms of and mortality from coronary heart disease were more common among the men who had earlier reported that their work was characterized by high strain (26). The association existed after a check for age, edu cation , smoking, and overwe ight.
Significantly more drivers than referent s worked shift s or had variable workhours . According to Petters son (27) about 27 % of the labor force in Sweden works shifts or has variable workhours. In this investigation 63% of the drivers had such workhours. Alfredsson et al (28) have shown that shift work and monotony were associated with an excess risk of myocardi al infarction, especially in combination with work characterized by low decision latitude or few possib ilities for promotion.
Significantly more dri vers than referents thought that their job involved mental stress . For bus drivers the combination of busy traffic, a tight schedule, and impatient pas sengers can cause stress at work .

332
In an investigation among urban bus driver s (29) high work load (driving in heavy traffic ) was significantly associated with the occurrence of myocardial infarction . The work of truck drivers also invol ves mental stress, especi ally for tanker truck drivers (18) and truck drivers eng aged in the distribution of goods ( 17). The motoring in itself can involve mental stress. Bellet et al (30) found a significant increase in the excretion of catecholamines and l l-hydrocorticosteroid in both normal subjects and patients with coronary heart disease when driv ing . According to the authors the se results suggest that automobile dri ving represents a mental stre ss. It is also known that car dri ving can increase the heart rate con siderably and cause electrocardiographic changes (3 1). Our study focused on a rural and smaller city population. In a Danish study (29) the results showed that urban bus drivers with routes in the center of town have a higher risk of developing ischemic heart disease than their colle agues who have routes in the suburbs. It is possible that the findings of our study would have been clearer in a population representing large citi es.
Thi s investig ation and an earlier study of Swedish professional dri vers (14) have shown that many profe ssional drivers have a low level of educ ation . In a stud y by Holme et al (32) the lowe st social class exhibited a much higher total mortality than the other classes. This difference was pronounced for, among others, coronary heart disease. A Swedish study (33) also showed that social class, defin ed by occupation, was clearly related to the incidence of coronary heart disease . According to Marmot & Th eorell (34) part of the association between social class and cardiovascular diseases may be due to differences in psycho social work conditions. The se conditions can affect risk through either neuroendocrine mechanisms or life-style . The occupation of professional driving can lead to poor eating and smoking habits and a low level of physical activity during leisure time.

Card iovascular risk index
Th e cardiovascular risk inde x showed a significa ntly higher value for the drivers than for the referents. Of the different variables controlled for in the estimation of the odds ratio for having a high cardiovascular risk inde x, working class membership was the one which affected the odd s ratio the most. Educationallevel and shift work also had a substantial influence on the odds ratio, while heredity, mar ital status, and age had a minor effect. When all of these potential confounders were taken into account, the odds ratio was reduced from 3. 18 to 2.34, but it was still significantly high. Thus there seems to be factor s other than age , heredity, marit al status, shift work, educational level, and workin g class member ship that account for the elevated risk amon g dri ver s.