Mechanisms of cardiac risk among professional drivers.

Mechanisms of cardiac risk among professional drivers. Scand J Work Environ Health 1994;20:73-86.This lit erature review indicates that professional drivers have excess cardiac risk that is not fully explained by standard risk factors. The contribution of occupation is suggested by two independent methods and by psychophysiological studies during on-the-job driving. Driving has been conceptualized as a threat-avoidance task. Stimuli encountered in traffic are not inherently aversive but become so by as sociation with driving experience, a formulation corroborated by laboratory studies in which stimuli such as car headlights elicit cardiovascular hyperreactivity and electroencephalographic signs of arousal in professional drivers. More-advanced neurophysiological methods (event-related potentials) show higher cortical electronegativity to imperative signals among professional drivers than among non driver referents. These data are viewed in light of reports of possible associations between event-re lated slow potentials and cardiac risk. A clinically and ecologically relevant neurocardiological mod el is proposed, and preventive strategies, including workplace interventions, are suggested.

Of the numerous studies performed in va rious countries on cardiovascular disease and professional driving over a sp an of almost three de cades, ne arly all have shown an exce ss ri sk of cardiovascular di sease among professional drivers. Despite rigorous se lection against these disorders at hiring and during periodic medical foll ow-up ( 1-4), a risk of hyp ertension, ventri cu lar arrhythmias, myocardial infarction, and other ischemic heart disease ha s been ob served in 28 of the 32 reports found to consider thi s question (1-32). (No ns ig nifica nt findings have be en reported in references 13,18,19, and 23.) Of particular note in many of these studies is the close relation between the number of work years as a driver and the untowardly yo ung ag e at which these ev e nts occur (l , 5, 7, 9 17, 26, 29, 32). For example, in studies of young myocardial infarct ion patients, fo r whom profession was determined, an unexpectedly high percentage (up to 40 %) comprised professional dri vers (9,17,26,29,32).  cm-1 (Il) a 11 = sign ificantly greater in drivers, 1 = . g; eater in drive rs bu t report ed dat a i ns uff ici ent for statistical calculation, 0 = no significant difference between drivers an d referen ts, I j = Significantly lower In driv ers , j = lo wer in driver s but repo rted data insufficient for st at isti cal calcu lat ion .
true today. In this section, we review the available literature on cardiovascula r and neuroendocri ne changes tha t occur during actual professional and amateur driving.

Blood p ressure
Recently, during a workday, Ugljesi c and her coworkers (39) measured the ambulatory blood pressure of 30 male city bus drivers who had been normotensive in their base-line examination and also the ambulatory blood pressure of 20 matched clerical worker referents. They found that both the systoli c and diastolic blood pres sures were significantly higher among the bus drivers immediately preceding the driving shift, during most of the driving shift, and just after it than among the referents at matched time s. During nonwork hours, the bloo d pressure of these two gro ups did not differ significantly. Littler et al (40) performed direct arterial measurement of blood pressu re in a small group of health y subjects, hypertensives, and coronary patient s durin g drivin g. Overa ll they found fairly constant blood pressure levels , although some persons in each group showed sudden dramatic rises .

74
The interaction between individual pred isposition towards high blood pressure and high "strain" occupations such as dri ving has been demonstrated in three investigation s by Theorell et al (4 1--43). In the first, it was found that subjects who were hypertensive at 18 years of age showed the greatest systolic blood pressure rise during high (compared to low) strain work 10 years later. The se subjects also had a higher systolic blood pressure dur ing work than previously normotensive subjects in either work category. The seco nd study was a longi tudinal followup of workin g men and women in six occupations. It showed that sys tolic blood pressure durin g workhours was significantly highe r during days of job strain . The third study revea led a significant correlation between job strain and dias tolic blood pressure during the work of borderline hypertensives. An association betwee n job strain and elevated systolic and diastolic blood pressure during work has also been reported for healt hy men and wome n (44--46).

Heart rate
Increased heart rate has been con sistently reported during professional and amateur driving by healthy subjects and cardiac patients (4,38,(47)(48)(49). In some studies the increase was in the tachycardiac range and was particularly pronounced during "critical situations" such as sudden stops and passing (48)(49). Belkic et al (8) found that heart rate was maximum in professional drivers at the beginning and towards the end of a workshift, the latter most likely reflecting fatigue. Similar findings have been reported for other high-stress groups during work (46, [50][51][52][53] . Studies of heart-rate variability, as a reflection of autonomic outflow and risk of cardiac death (54,55), during driving are very sparse. Using spectral analysis, Egelund (56) reported that heart-rate variability increased among inexperienced (presumably nonprofessional) drivers. This finding was considered to be related to fatigue. However, no such effect was seen when the standard deviation of heart rate was assessed. Using the standard deviation of heart rate, Belkic (57) found minimum heart-rate variability at the beginning of the driving shift, when compared with the middle and end of the shift. However, this finding was not statistically significant due to the small numbers.

Ischemic electrocardiographic changes
Using ambulatory electrocardiographic monitoring, Taggart et al (49), in their investigation of 32 normal subjects and 24 coronary patients (all amateur drivers) during a 20-min drive in heavy London traffic, reported significant STT changes in three of the normal subjects and 13 of the coronary patients, two of whom experienced symptomatic angina pectoris. Similar results were also reported by Hoffmann (58) in an earlier study of patients with coronary heart disease. Lauwers et al (59) found ST depression in 4 of 13 postmyocardial infarction patients during a drive in peak hours of traffic. A flattening of the T wave was reported for three of four healthy amateur drivers behind the wheel in the study of Simonson et al (38), while no ischemic electrocardiographic changes were seen in 65 healthy young subjects examined during driving by Bellet et al (48) or in the 15 subjects studied by Littler et al (40). In the two available reports on the Holter monitoring of apparently healthy professional drivers (4, 60), 20 subjects in one study and 95 in the other, no significant STT changes were noted. Stressful occupational activity which bears some similarity to driving (eg, aerobatic flying and sea piloting) are reported to elicit ischemic changes in apparently healthy subjects (50,51). To our knowledge, no ambulatory electrocardiographic monitoring studies of professional drivers with ischemic heart disease have been published.

Cardiac rhythm
Cardiac rhythm disturbances during driving have received surprisingly little attention in the literature. In the study of Taggart et al (49), five of the coronary patients developed complex ventricular arrhyth-Scand J Work Environ Health 1994, vol 20, no 2 mias during the previously described drive. One of the 15 subjects in the paper of Littler et al (40) also developed multiple ventricular extrasystoles during driving. Belkic et al (8) found that nine of 20 (45%) young, apparently healthy professional drivers had ventricular arrhythmias during Holter monitoring, five (25%) had Lown grade 3 or more, and four (20%) had noteworthy bradyarrhythmias. The total number of ventricular extrasystoles was significantly greater when the recording included a day driving shift compared with nonwork days, and no other factor except driving itself distinguished the two periods. On the other hand a significant correlation was found, for workdays only, for the number of ventricular extrasystoles with smoking, coffee intake, and lack of sleep. Aerobatic flying and sea piloting have also been found to evoke ventricular arrhythmias in apparently healthy subjects (50,51). We have not encountered any studies comparing the incidence of arrhythmias during driving and nondriving periods among subjects who suffer from rhythm disturbances.

Catecholamines and corticosteroids
Catecholamine excretion is consistently reported to be elevated during exposure to driving among amateurs, professionals, and racers (61-65). Night driving and traffic congestion are particularly associated with increased urinary catecholamine (63, 65). Other occupations with high levels of mental stress, such as air traffic controllers, physicians and nurses, and railway dispatchers, are also associated with an elevated catecholamine excretion during work, especially at night. (66-69). Our group has a study in progress on plasma catecholamine levels immediately after the early morning driving shift of city bus and truck drivers.
Cortisol levels during driving have rarely been examined, Bellet et al (61) found elevated levels when amateurs drove. Theorell et al (70) very recently reported that the subway drivers who, subsequent to a "person under train" incident, experienced the highest level of depression and had a need for longterm sick leave showed high plasma cortisol levels. This finding is congruent with those showing that the cortisol excretion rate is directly proportional to the level of distress (71).

Measures of occupational stress in professional drivers
Describing and quantitating the occupational factors that induce the aforementioned changes in professional drivers and potentially increase cardiovascular risk require appropriate methodological tools. One approach that provides easily obtainable, standardized numerical data is the psychosocial job-exposure matrix, whereby psychological demands, control over work conditions, support, physical work demands, and job hazards are considered (72). A jo b-strain index is then computed as the ratio between psychological demands and control (43). According to this system or an older, less sophisti cated occupational classification system, professional drivers are classified as a high-strain group (4 1). In several investigations the job-exposure matrix, or its precursor, has been shown to be a significant marker of cardiovascular disease morbidity and mortality (73-75) .
The theoretical constructs upon which the job -exposure matrix is based draw from a long tradition of international psychophysiological researc h centered heavily in Sweden. A fundamenta l observation is that a high degree of arousal, together with a high degree of distress, activates both the sympathoadre nomedullary and adrenocortical axes, the result being a maximally deleterious combination (7 1, 76). Integrating this formulation with Gardell's (77) dimension of "alie nation" and noting the association between work stress and risk of myocardial infarction, Karasek proposed that excessive load should be considered together with low decision latitude (78). Karasek and his co-workers (79) then developed the methodology for applying these hypotheses in epidemiologic, field , and clinical laboratory studies which focus strongly upon the cardiovascular system.
Belkic (57) incorporated these and other elements into a comprehensive occupa tional stress index, which includes a tota l of 58 equally weighted factors potentially affecting the cardiovascular system. These factors are arranged into a two-dime nsiona l matrix, as shown in table 2. The various aspects are placed along the horizon tal axis and include underload, high demand, strictness, extrinsic time pressure, aversivenes s, avoidance, and conflict-uncertai nty. The vertical axis is comprised of levels of information transmission, a concept presented by Welford (80) and used by Luczak (8 1) to simulate work conditions. The occupational stress index synthesizes fundamental ergometric concepts with the leading formulation s of how stress eve ntuates in cardiovascular pathology. These formulations include the compelling models of sudden cardiac death vulnerabil ity in experimental animals in response to aversive environments, avoidance task performance and approach-avoidance conflict reported by Lown' s (82-84) and Corley's (85,86) groups, as well as models based upon underload, high demand, strictness, and lack of autonomy presented by Karasek et al (79), Theorell et al (87), and Frankenhaeuser (88) . Descriptive studies of occupational stress in subjects at high cardiac risk, as well as investigatio ns of specific physica l and general factors implicated in cardiovascular disease , were also used to identify elements for inclusion in the occupational stress index. These elements include a need for high levels of vigilance, extreme time pressure , potentially disastrous consequences of an error or lapse of attent ion, exposure to hazards and physically noxious agents such as noise, glare, vibration, cold, heat, isometric stress  or heavy lifting, toxic chemical agents, long workhours, night shift work, lack of rest breaks, piecework, and holding down two or more job s (89)(90)(91)(92)(93)(94)(95)(96)(97)(98)(99)(100)(101)(102)(103)(104). Thus wide possibilities for hypothesis testing and the identification of key areas for preventive intervention trials are offered with this method, The occupational stress index also allows precise comparisons to be made among groups and indi viduals of any occup ational endeavor.
The occupational stress inde x is calculated from a self-administered questionnaire. It is based on concrete questions, which , to a great extent, are objectively verifiable. Subjecti ve responses are assessed separately as a job-adaptation index. Collaborativ e studies are underwa y to determine the cross-cultural validity of the occupational stress index and to compare the sensitivity and efficacy of the index with the psychosocial job-exposure matrix in predicting cardi ac risk.
The occupational stress index has been applied in Belgrade, Yugoslavia, to 258 professional drivers and 227 workers primarily in the building trade (referents), all of whom were actively working (57, 105). The mean total for the occupational stress index was more than twice as great for dri vers than for referents, the difference being highly significant. The extremely high total obtained for the occupational stress index suggests that professional driving bears the maj ority of features associated with cardiac risk. Briefly stated, professional driving is charac terized by decision-making underload coupled with input high demand, and it is prim arily avoidance activity performed under conditions of confl ict, physical and mental constraint, time pressure, and exposure to physically noxious agent s (eg, noise, glare, wholebody vibration, lead , carbon monoxide, other combusti on product s). Occupational stress is the most dramatic in profe ssional drivers at the input level [21.7 (SO 1.3) versus 6.6 (SO 1.7) for professional dri vers and referents, respecti vely]. The signals recei ved are primarily visual.

Coronary-prone behavior of professional drivers
Behavioral indices of cardiac risk have not been extensively studied among professional drivers. Comparing 258 drivers and 227 building worker referents, Belkic et al ( I) found that three features of interview type A behavior (hyperalertness, tight facial musculature and explosive speech) were significantly more pronounced among dri vers, as was obsessiv e punctuality. Using a more-extensive interview along the lines of Friedman & Rosenman' s ( 106), Hartvig & Midttun ( 15) also assessed type A behavior in a group of 52 bus and truck drivers. They similarly found a much higher prevalence of coronary-prone behavior among drivers than among matched industrial worker referents. Evans et al ( 107) noted that self-reports of occupat ional stress were more com-Scand J Work Environ Health 1994. vol 20. no 2 mon among male bus drivers exhibiting type A behavior than among their type B counterparts. In addition heavily smoking professional drivers (>4 0 cigarettes a day) have shown a particularl y high prevalence of coronary-prone behavior ( I). This high prevalence suggests an interrelation among occupational, behavioral, and standard cardiac risk factor s among professional drivers.
The question logically arises of whether individuals who enter this profession are behav iorally or otherwise innately predisposed to cardiovascular disease . The answer to this question is far from resolved. However, two large independent studies provide intriguing data. In 1990 Belkic and her co-workers (7) examined the prevalence of hypertension among 293 Belgrade city mass transit drivers, who were compared with 194 maintenance workers employed at the same company. A signifi cantl y greater number of drivers aged 20 to 40 years had high blood pressure when compared with their agematched referents. Drivers with 6 to 10 years of experienc e also had a significantly higher prevalence of hypertension than the referents matched for work exper ience . However , the drivers and refer ents with one to five years of experience in their respecti ve job s did not differ significantly in this respect. In the other study of 258 professional drivers of various profiles, compared with 227 build ing worker referents ( I), all aged 40 years or younger, blood pressure was again significantly greater in drivers with 6 to 10 years of ex perience, and not in those with one to five years of experience, in compa rison with referent s matched for work experience. The conclusion which emerges from these two studies is that a particular vulnerability exists in young professional drivers with respec t to blood pressure and that a threshold period of dri ving exposure is necessary for these effects to be seen in terms of resting blood pressure.

Potential role of cognitively relevant visual signals in untoward cardiovascular changes in professional drivers
The glar e pressor test As previou sly discu ssed, driver s face maximal occupation al stress at the input level , and the vast majority of signals they recei ve are visual. As Fuller ( 108) has pointed out, these stimuli are not intrinsically aversive, but they become so by association with driving experience. Driving is thereby conceptualized as a threat-avoidance task, since the dri ver must avoid aversive stimuli or situations such as driving off the roadway, losing control of the vehicle, or colliding with another road user. Thu s, for example, an inherentl y neutral stimulus like a car headlight would attain specific cognitiv e significance because of its assoc iation with impendin g danger from an approaching vehicle during night driving and the Scand J Wo rk Environ Health 1994, vol 20, no 2 need for accurate and timel y responses to avoid a collision.
In 1984 Belkic (109) performed a comparative study on the blood pressure responses of 22 healthy young subjects to a large battery of diverse procedures, including several standard mental stress tests, aerobic and isometric stress, sensory stimuli such as 90 dB noise, the cold pressor test, and exposure to impulses of glare from an ordinary car headli ght. Unexpectedly, the blood pres sure rise in response to glare was highly significant, and the dia stolic response was among the greatest when comp ared with the corresponding response to the other maneu vers. Of parti cular note was a diastoli c blood pressure increase of 30 mm Hg (3.99 kPa ) in one subj ect, an amateur dri ver, who later told of his colli sion with an oncoming vehicle after having faced glare from a headlight while driving at night.
On the basis of these findings, a new reactive procedure, the "glare pressor test" was introdu ced into the literature (90) . During this test, light impulses from an ordinary car headli ght were delivered at 55-s inter vals through a dark cylinder, I m in length , into which the subject peered. A total of five light flashes were presented so that the entire glare pressor test lasted 5 min .
In a subsequent study (110) , four group s of subjects [profe ssional drivers, amateur drivers, nondrivers who were profe ssionall y exposed to glare from other sources (welding), and nondriver s with no other professional glare expo sure] performed the glare pressor test. A continuous electro card iogram and blood pressure were recorded. The profe ssional and amateur drivers showed a signi ficantly higher diastolic resp onse to the test than did the nondriver referents. The group expo sed to welding glare did not differ from the referent s in their response to the test. The test elicited ventricular extrasystoles only in profess ional drivers who drove at night. These findings appeared to corroborate Fuller's contention, suggesting that headlight glare, as a symbolically avers ive visual stimulus for the driver, was indeed capable of evoking a powerful autonomic response.
Continuing with these investigations, the glare pressor test was appl ied during electroencephalographic (EEG) and polygraphic recordin g in another group of young profession al dri vers with night driving experience, and the results were co mpared with those of nondriver referents (57, III ). Of the 19 tested drivers, 15 showed persistent blockade of spontaneous alpha acti vity or complete EEG desynchronization after the first glare impulse. Thi s reaction was accompanied by a signific ant fall in digital pulse amplitude and, as previously seen , a significant rise in diastoli c blood pressure, most likely refleetin g vasoconstrictio n, which is reported to occur in response to condit ional aversive stimuli ( 112). The electromyogram (EMG) reveal ed a signi ficant increase in facial muscle tone. Thu s the professional dri vers exhibited electrocortical, somatic, and auto-78 nomic arousal indicati ve of a defen se reaction in response to expo sure to an automobile headlight, whereas no significant effects of the glare pressor test were found in the reference group.
The response of dri vers to the glare pres sor test typifi es the type A reaction to a specific chall enge. Hyperalertness is objectively confirmed by the finding of EEG desynchronization. Tight facial musculature is reflected in the increased ampl itude of the EMG. Zeier (113) found a significantly greater frontal EMG amplitude in driv ers than in persons with passenger status. This finding was said to reflect " readiness for performing motor acti vity as well as general tension in co gnitive tasks [p 800]." Most of the drivers reco vered their base-line levels of alpha acti vity and showed milder diastolic blood pressure , as well as digital pulse changes, after the final (fifth) glare impulse. In other words, with repeated exposure to the glare impulse s, the subje cts appeared to distinguish the laboratory exposure from the real exigencies of driving.
However, in two driv ers, this last glare stimulu s elicited the most pronounced changes. For a 31-yearold truck driver, for instance, it was the fifth glare impulse which most rapidly elicited ventricular arrhythmi as. One city bus driver, with extreme type A behavi or, demonstrated complete EEG desynchronization during the entire test and a marked rise in diastolic blood pressure at the final, compared with the first, glare impulse. These responses were considered malad aptive and therefore sugges tive of hyper sensitivit y to glare expos ure, with a break down of protective mechani sms. Both of these drivers, although clinically healthy according to standard screening criteria, as well as to an exerci se stres s test, showed notewor thy ventri cular arrhythmias during Holter monitoring in the aftermath of a drivin g shift, and they had a very high standard risk factor status. They both worked very long hour s (Ill , I 14).
To date, the glare pressor test with electrocardiographi c recording has been applied to 74 subj ects . A total of 25 professional drivers with night driving experience were included. For three of these drivers, the glare pressor test elicited ventricular arrhythmias, whereas arrhythmi as were not evoked in any of the 49 tested subjects who were not profe ssional drivers with night driving experience. This finding is statistic ally significant and suggests a possible link between the occurrence of ventricular irritability in professional drivers and exposure to cognitively relevant input stress that is transmitted via the visual system.

Event-related potentials in prof essional dri vers
The results presented in the previous section sugge st that cog nitive aspects of headlight glare are primarily responsible for the central, somatic, and autonomic arousal seen in professional drivers during the glare pressor test. In order to define more precisely the central stress mechani sms which become ope ra-tive when drivers are faced with relevant visual signals, more sophisticated neurophysiological methods, event-related poten tials , have recently been used by Belkic et al (1 15). The pattern of response was examined for young, healthy Belgrad e city bus drivers and nondriver referents using a GOINO GO visua l cont ingen t negative variation (CNV) paradigm . A tacitly neutral task was compa red with one which explicit ly mimicked traffic conditio ns.
Briefly, the GO warning signa l (GO S1 ) was a flash of red light, and the imperative signa l (GO S2 ) was a yell ow light 1.5 s later, to which the subj ect was instructed to press a butt on. When the NOGO warning signal (NOGO s, ) appeared (pseudo randoml y), the subjec t was told not to press the button upon the appearance of the imperative (GO S2 ) signal. During the neutral task, these instructions were given without any additio nal remarks (ie, no reference was made to traffic conditions). For the second task (traffic paradigm), the subject was told to imagine himself as a driver in heavy traffic and that the red light represented a warning and, when the yellow light appeared, he shou ld press the button as if it were a gas pedal. Fail ure to react on time would be the equiva lent of a serious traffic accident caused by him. He was also told to co nsider inappro priately reacting to the yellow light that followe d a green light, as the equivalent of a serious accident caused by him.
Thirt y GO sequences and all NOGO sequences were averaged for the neu tral and for the traffic CNV paradigms. The following three CNV amplitude measures were then dev ised : (i ) 5 00-to SOO-ms post -GO S1 (CNV !), 1200-to ISOO-ms post-GO S1 (CNV 2 ) and 500-to SOO-ms post-GOS 2 (CNV]).
The CNV results for both paradigms are gra phically displayed in fig ure I. As can be sce n, the GO CNV ! and CNV 2 amplitudes were significantly more negative than the corresponding values of the NOGO were, an expec ted finding for normal subjects (116)(117). This GO-NOaO difference was accentuated for drivers, a finding suggesting an enhanced capacity for the differentiation betwee n relevant and nonrelevant visual signals. The mean CNV 2 was also significantly greater for drivers than for referents for the GO condition durin g both the neutral and the traffic paradig m. This late CNV, which has been ca lled the "readiness potent ial," is associa ted with preparation for the appearance of an imperati ve sign al, which may have an affective character (l IS-120). Overall, a high amplitude (negative) GO CNV reflects subjective and objective task difficu lty, as well as involvement and tension durin g task perform ance (121)(122). Anxious subjects, once having learned the association between the warning GO stimulus and the imperative stimulus, show higher GO CNV amplitudes than low anxiety subjects (123). On the other hand, the NOGO CNV becomes more positive with alertness (12l). The driver s reacted in essentially the same manner when performing the tacitly neutral CNV task and that which explicitly mimicked traffic conditions. Thus the pattern of respon se of these drivers during the two paradigms suggests a generalized heightened sensiti vity to the cognitive significance of visual stimuli , with a greater degree of task involvement , tension, and alertnes s than for nondriver referents. These CNV findings provide an insight therefore into how type A behavior develops in relation to the demands made upon drivers by their profession.
This electrocortical respon se pattern may also be more directly related to the increased cardiac vulnerability of professional drivers. The importance of this connection has been stated by Skinner (124) as follows: "the study of the set of event-related electrochemical responses of the cerebral cortex may eventually provide an understanding of the complete cerebral mechanism by which a stressful stimulus event is transformed into increas ed cardi ac vulnerabilit y [p 92B]. " Skinner and his co-workers (125 ) have demonstrated a strong positive correlation between the amplitude of the event-related slow potential and ventricular arrhythmia rate in coron ary patients in response to antiarrhythmic agents. The CNV, or cerebral event-related slow potential, as a reflection of frontal lobe function, is attenuated after damage to that region ( 117). In addition to losing the ability to experience stress, these patients fail to show any autonomic changes to cogn itively relevant stimuli (124). Simons (126) has also recently emphasized the interrelation between slow cortical potenti als and autonomic responses, particularly midinterval heartrate acceleration associated with displeasure and arousal. Experimental data, explored very briefly in the next section, provide powerful corroboration of the role of this higher neural traffi c in cardiac vulnerability.
Brain mechanisms and stress-related cardiovasc ular changes A general theoretical model of how brain structures mediate the effect of stress upon the cardiovascular system has been presented by Skinner (124). His group demonstrated experimentally that, during stress, neural traffic from the frontal lobes (neocortex), as well as from the amygdala (limbic systempaleocortex), contributes significantly to the development of hyperten sion, as well as to life-threatening ventricular arrhythmias in the setting of myocardi al ischemia (127)(128)(129). Repeated stimulation of the lateral hypothalamus can also produce sustained hypertension (130). The posterior hypothalamus, when electrically stimulated, elicits an alerting response, and, according to some reports, the defense reaction, along with increased cortisol outflow and ventricular arrhythmias (including ventricular tachycardia) , 80 and the threshold for ventricular fibrillation are lowered (83,(131)(132). In the face of experimentally induced myocardial ischemia, posterior hypothalamic stimulation proves to be the critical factor in cardiac stability. It has induced ventr icular fibrillation in 62.5 % of coronary occluded dogs, whereas only 6.3 % of the dogs with ischemia but no central nervous intervention developed ventricular fibrillation (83) . Stimul ating the midbr ain reticular formation also lowers the ventricular fibrillation threshold and raises blood pressure and heart rate (133)(134).
The two monoaminergic system s, the catecholaminergic (dopaminergic and noradrenergic ) and serotonergic systems, appear to be important mediator s of the connection between the processing of relevant sensory signals and cardio vascular response. Dopaminergic neurons are needed for motivated behavior and orientation to sensory stimuli (135)(136). Neuroleptics, which block dopamine receptors, impair conditional avoidance responses and hamper vigilance. These medication s also lower blood pressure (137). Cortical noradrenergic and beta-receptor functions are required for the detection of a change in visual input , and they appear to be involved in local event-related slow potential formation (124). Central beta-adren ergic mechanisms have been recentl y implicated in stress-related lowering of the ventricular fibrillation threshold ( 138). Serotonergic neurons, on the other hand, appear to attenuat e efferent sympathetic neutral activity to the heart and vessels, such that, for example , elevated levels of brain serotonin raise the threshold for ventricular fibrillation (139 ).

Model f or increased cardiac vulnerability caused by visually mediated stress in professional drivers
We have recently proposed one model to show how neural mechanisms might increase cardiac vulnerability, the focus of the model being night driving and exposure to headlight glare (36). The essential elements include: (i) avoidance task perform ance, (ii) physically aversive exposure (to glare), (iii) uncertainty (due to dimin ished visibility), (iv) conflicting forces acting upon the level of arousal (darkness plus circadian factors diminish arousal, glare exposure plus driving task demands increase arousal), (v) abrupt changes in oculomotor autonom ic reflexes which involve sympathetic and parasympathetic branches and conflict between the central and refle x tendencies. The first three elements have been associated with a lowered threshold for ventricular fibrillation and with sudden cardiac death in experimental animals (84,86,140). The abrupt change s mentioned in the fifth element could disrupt neuroregulation of the autonomic nervous system and, as seen in clinical and experimental studies, increase cardiac electrical instability (141)(142)(143).
As discussed, an oncoming headli ght represents danger to the driver. It remands him to respond appropriately to avoid a colli sion. Headlight glare, es-pecially from a misaimed beam , is unpleasant and can also impair vision for several seconds. It is thereby a specific night driving hazard. It should be recalled that glare is an incongruity between an object in the visual field and the level of light to which the eye is adapted . It is a recurrent stress that occurs impulsively (ie, suddenly) during night driving .
Abrupt changes between dark adaptation and glare exposure occur during night driving, particularly on poorly lit roads. Adaptation to dark and light involve, respectively , sympathetic and parasympathetic reflex pupillary responses, which synapse near the midbrain reticular formation.
Pupillary tone is controlled by reflex, as well as by higher control mechanisms. Central arousal promotes pupillary dilatation (dark reflex). On the other hand , glare elicits a sensory light reflex (pupillary constriction), but it also causes central arousal which favors mydriasis. This a conflicting situation with strong opposing tendencies for the autonomic nervous system.
The ensuing "chaos" can be played out in the midbrain reticular formation (figure 2), which aside from mediating pupillary reflexes, transmitting information to the visual cortex, and controlling cortical arousal, also mediates autonomic outflow from the hypothalamus. Stimulation of the midbrain reticular formation lowers the threshold for ventricular fibrillation and raises blood pressure and heart rate (134,144).
This model may help explain the untoward cardiovascular changes that occur in professional drivers in response to the glare pressor test, and it suggests mechanisms which may particularly increase cardiac vulnerability during night driving. Detailed exploration of other factors burdening the nervous system of professional drivers, particularly driving in heavy traffic, which is also implicated in cardiac risk (3), is warranted as well.

Clinical relevance and implications of the neurocardiological model
In 1985, Benjamin Natelson (145) first introduced the term "neurocardiology" as an important, new multidisciplinary area of research. He aptly stated that "cardiologists have focused their attention preponderantly on their end organ (i.e. the heart) and neurologists have focused on their organ (i.e. the brain) [p 1831." While breakthroughs in this "interface" have been made in the laboratory, "full appreciation of the importance of the nervous system in the pathogenesis of cardiac dysfunction does not exist at the bedside [p 183]." A crucial barrier to progress in this area has been the lack of appropriate clinical models.
Steptoe & Vogele (146) recently analyzed the methodological problems concerning mental stress tests in cardiovascular research and outlined essential criteria for their validity. These criteria include methodological, ecological, diagnostic, prognostic, and therapeutic considerations. It can be noted that it is difficult to combine a satisfactory degree of methodological rigor with "ecological" relevance (ie, applicability to real life stressors).
In studying professional drivers , we have had a unique opportunity to develop standardizable methods that are extremely relevant to this high-risk group and that consistently elicit strong cardiovascular responses, with concomitant signs of central arousal. The prognostic significance of these responses remains to be determined. However, it is plausible to consider that cardiovascular hypemeactivity to stressors such as the glare pressor test may precede the development of essential hypertension or other acquired cardiovascular disorders, as outlined in figure 3. The predictive power of other reactive tests suggests , by implication, that this may actually be so (147)(148).
Thus far, the glare pressor test, the CNV model , Holter monitoring during driving, and the occupational stress index have been applied to apparently healthy professional drivers. These results need replication in other centers in various countries. They could, however, subsequently form the basis for comparison with responses in drivers with prepathological cardiovascular changes, as well as with responses in young drivers who have already suffered fullblown disease.
Together with mainline cardiological methods, these neurocardiological tests might , in the future , be used to assess cardiovascular driver fitness more accurately. This more accurate assessment may help answer the question of return to work, which is of vital importance, since these drivers often suffer from cardiovascular disease at untowardly young ages. These methods could also be of value in the periodic screening of professional drivers .
However, the most urgent task, as recently stated by Michaels & Zoloth (2) is "to apply specific in-  terventions designed to reduce j ob strain and other risk facto rs for heart disease among ... driver s [p 403]." Since there is a lack of data concerning workplace intervention and the attenuation of cardiovascular risk, this work will require considerable investigation in orde r to determ ine which of the numerou s, potential factors affec ting the dri ver' s cardiovascular system are most crucial. Some guidelines for this process of reconstructing worklife have been presented by Karasek & Theorell ( 149). As emphasized there in, job-related monitoring feedback of health is esse ntial in estimating the effectiveness of any intervention. A particularly important area for investigation is that of driving in heavy traffic and night driving, as well as the role of road conditions, sedentary work in a fixed posit ion, and expos ure to toxic and physically noxious agents such as lead , noise, glare, whole-body vibration, and carbon monoxide . Transcultu ral factors , such as norms of behavior for participants in traffi c, crowding in mass transit vehicles, and the like, would also be of interest for study with regard to burden upon the driver.
In addition, however, we recommend that certain basic measures be implemented. (Many of these, in fact, are often already established policy but may not be enforc ed in practice.) The following measures are included in these recommendations: I. The number of driving hours should be carefully controlled. Overtime work should be the solution of last resort and should be strictly limited. Drivers should be discouraged from accepting overtime for economic reasons, and dispatchers should be discouraged from relying upon the responsible, good dri v-82 ers to compensate for driver shortages due to absenteeism and the like. Educational work on the hazards of overtime and driver fatigue should be aimed not only at the drivers, but also at the dispatcher s.
2. There should be obligatory rest breaks.
3. Any other skill the driver may have should be used for task alternation to minimi ze the number of hours behind the wheel. 4. Laws should be enforced about disorderly or otherwise disturbin g conduct on mass transit, especia lly for passenger s who disturb the driver. 5. Separate traffi c lanes should be used, whenever possible, for public transportation and heavy trucks. 6. Traffic laws should be more strictly enforced for private vehicles and, especially, for pedestrian s (ie, j ay-walk ing, cros sing on a red light, and the like).
Overall, a three-pronged approach aimed at modifiable occupational, behavioral, and standard risk factors is esse ntial to effec tive primary cardiac prevention among professional drivers. This approach might eve ntually be combined with pharmacologic measures, such as the use of a beta-blockade, for selected drivers, since the se agents appear promi sing for this group (150)(151)(152)(153). Such an approach is being developed and will be the subject of future reports.