There is a growing body of evidence from studies in experimental laboratory settings (Kahn & Quinn, 1970) and in the workplace (Margolis, Kroes & Quinn, 1974) to suggest that occupational stress is a causal factor in these diseases. By occupational stress is meant negative environmental factors or stressors (e.g. work overload, role conflict/ambiguity, poor working conditions) associated with a particular job. In addition to the environmental precursors, inherent characteristics of the individual and his behaviours may also contribute to occupational ill health. In fact, as Jenkins (1971a) has suggested, the early clinical studies of psychosomatically oriented General Practitioners and psychiatrists led to a number of theories about a predisposing state of neuroticism being confronted by environmental stressors leading to reaction of anxiety, changes in cardiovascular function and, in time, to coronary heart disease or mental ill health.

There are, therefore, two central features of stress at work, the interaction of which determines either coping or maladaptive behaviour and stress-related disease (Cooper & Marshall, 1975): (1) the dimensions or characteristics of the person and (2) the potential sources of stress in the work environment; or as Lofquist & Dawis (1969) have labelled this interaction The Person–Environment Fit. There is, however, a third set of extra-organizational variables which can also be sources of stress. These are not linked directly to the individual’s characteristics or the work environment but are related to outside relationships and events, for example, family problems, financial difficulties, life crises (death in the family), etc., which have an impact in the workplace. In Figure 1.1 we provide a diagram which will highlight a model of stress at work which incorporates these concepts.

It is our intention here to examine the stress research literature linking environmental and individual sources of stress to physical and mental disease or illness. We are attempting this in the hope that we will be able to reveal where much of the work has been done and where the gaps lie. In addition, since much of the research has been carried out ‘within’ particular disciplines (i.e. psychology, management, sociology, medicine) and not ‘between’ disciplines, it is hoped that we may indicate the potential of interdisciplinary work in this field and begin to encourage this development.

Before embarking on this review, it might be helpful to note that many of the stress studies over the last 10–15 years have utilized two primary indices of occupational disease, coronary heart disease (CHD) and mental ill health (MIH). A more limited number of studies have focused on other physical illnesses thought to be stress-related, such as peptic ulcers, respiratory diseases and allergies, but these have not been as thoroughly researched in respect to social and psychologic stressors as heart disease or mental ill health. We have concentrated on the CHD and MIH studies primarily but have included the others, where they have contributed a new dimension or perspective to this topic. It might be added here that our literature review covered all the journals incorporated in the MEDLARS literature-retrieval system of the US National Library of Medicine, which includes most of the relevant medical and social science journals.

We should like to preface this review with a brief summary of some of the methodological problems encountered in some of the studies to follow. In the main text detailed criticism of individual studies has been kept to a minimum to allow breadth of coverage; we hope here, however, to forewarn the reader of the difficulties of interpretation of the research reported. Some of the methodological shortcomings are:

On the other hand, research into work and overload has been given substantial empirical attention. French & Caplan (1973) have differentiated overload in terms of quantitative and qualitative overload. Quantitative refers to having ‘too much to do’, while qualitative means work that is ‘too difficult’. (The complementary phenomena of quantitative and qualitative underload are also hypothesized as potential sources of stress but with little or no supportive research evidence.) Miller (1960) has theorized, and Terryberry (1968) has found, that ‘overload’ in most systems leads to breakdown, whether we are dealing with single biological cells or individuals in organizations. In an early study, French & Caplan (1970) found that objective quantitative overload was strongly linked to cigarette smoking (an important risk factor or symptom of CHD). Persons with more phone calls, office visits, and meetings per given unit of work time were found to smoke significantly more cigarettes than persons with fewer such engagements. In a study of 100 young coronary patients, Russek & Zohman (1958) found that 25 per cent had been working at two jobs and an additional 45 per cent had worked at jobs which required (due to work overload) 60 or more hours per week. They add that although prolonged emotional strain preceded the attack in 91 per cent of the cases, similar stress was only observed in 20 per cent of the controls.

Breslow & Buell (1960) have also reported findings which support a relationship between hours of work and death from coronary disease. In an investigation of mortality rates of men in California, they observed that workers in light industry under the age of 45, who are on the job more than 48 hours a week, have twice the risk of death from CHD compared with similar workers working 40 or under hours a week. Another substantial investigation on quantitative workload was carried out by Margolis et al. (1974) on a representative national sample of 1496 employed persons, 16 years of age or older. They found that overload was significantly related to a number of symptoms or indicators of stress: escapist drinking (r = 0·06), absenteeism from work (r = 0·06), low motivation to work (r = 0·26), lowered self-esteem (r = 0·10), and an absence of suggestions to employers (r = 0·17). Although these coefficients are significant, they are very small indeed, contributing between less than 1 per cent to a maximum of 5 per cent of the variance. Quantitative overload is obviously a potential source of occupational stress, as other studies (Quinn, Seashore & Mangione, 1971; Porter & Lawler, 1965) also indicate, but on the evidence available it may not, by itself, be a main factor in occupational ill health.

There is also some evidence that (for some occupations) ‘qualitative’ overload is a source of stress. French, Tupper & Mueller (1965) looked at qualitative and quantitative work overload in a large university. They used questionnaires, interviews and medical examinations to obtain data on risk factors associated with CHD for 122 university administrators and professors. They found that one symptom of stress, low self-esteem, was related to work overload but that this was different for the two occupational groupings. Qualitative overload was not significantly linked to low self-esteem among the administrators but was significantly correlated for the professors. The greater the ‘quality’ of work expected of the professor the lower the self-esteem. They also found that qualitative and quantitative overload were correlated (r = 0·25 and r = 0·42 respectively) to achievement orientation. And more interestingly, in a follow up study, that achievement orientation correlated very strongly with serum uric acid (r = 0·68) (Brooks & Mueller, 1966). Several other studies have reported an association of qualitative work overload with cholesterol level; a tax deadline for accountants (Friedman, Rosenman & Carroll, 1958), medical students performing a medical examination under observation (Dreyfuss & Czackes, 1959), etc.

French & Caplan (1973) summarize this research by suggesting that both qualitative and quantitative overload produce at least nine different symptoms of psychological and physical strain; job dissatisfaction, job tension, lower self-esteem, threat, embarrassment, high cholesterol levels, increased heart rate, skin resistance, and more smoking. In analysing these data, however, one cannot ignore the vital interactive relationship of the job and employee; objective work overload, for example, should not be viewed in isolation but relative to the individual’s capacities and personality. It should also be noted that since qualitative and quantitative work overload measures stem from perceptions of the subject, they may be influenced by the subject’s personality predispositions (e.g. nAch people).