Intended for healthcare professionals


Overwork can kill

BMJ 1996; 312 doi: (Published 13 April 1996) Cite this as: BMJ 1996;312:921
  1. Susan Michie,
  2. Anne Cockcroft
  1. Senior research fellow in clinical health psychology Senior lecturer in occupational medicine Occupational Health and Safety Unit, Royal Free Hospital and School of Medicine, London NW3 2QG

    Especially if combined with high demand, low control, and poor social support

    The death of a junior doctor in Britain last year, after working excessive hours and sleeping little, brought new relevance to the question, can overwork kill? It seems reasonable to suggest that excessive workload could be harmful. In Japan, there is even a recognised syndrome of “death from overwork”; the family of a Japanese man who killed himself after working for 17 months without a day off has recently won compensation from his employer.1 But the British government disagrees with the European Union that a proposal to limit the working week to 48 hours should be regarded as a health and safety issue. Indeed, there is surprisingly little hard evidence about whether mortality can be increased by physical or psychological overwork. A recent review noted a lack of research on the health effects of the interaction of physical and psychosocial factors at work.2

    The empirical research that exists suggests that higher workloads do increase disease and death rates. A Danish study, which followed up 2465 bus drivers over seven years, found that objective workload, as measured by the intensity of traffic on the drivers' routes, was the factor most strongly associated with death or admission to hospital with acute myocardial infarction.3 The incidence of both death and hospitalisation in those with higher workloads was more than twice that in the group with low workloads. Increased pace of work over the preceding five years and a lack of social contact with colleagues during spare time were also associated with an increased risk of myocardial infarction. A seven year follow up of 500 retired Swedish men found that, after adjustment for social class and recognised cardiovascular risk factors, job strain predicted mortality.4 Job strain was defined as the combination of high demand and low control. Mortality was significantly reduced in those who had good social networks and support, suggesting a buffering effect. A study of 99029 Italian railway men aged 40-59 years also found interactions between different aspects of workload and mortality.5 The men were classified into three levels of physical activity and job responsibility and followed up for five years. Low physical activity at work and high job responsibility were associated with increased risk of myocardial infarction, while high physical activity and low or medium job responsibility were associated with greater risk of death from chronic bronchitis or violence.

    These results are consistent with Karasek's job strain model,6 as well as studies showing that the combination of high job demands (workload and pace) and low decision latitude (individual control at work, influence on decisions, and opportunities for skill development) contributes to an increased risk of ischaemic heart disease and increased mortality.7 8 In combination with other factors, high workload has also been associated with other forms of ill health. Musculoskeletal disease, for example, has been found to be associated with high perceived workload, time pressure, low control on the job, monotonous work, and lack of social support by colleagues.9

    By what mechanisms might work overload influence health? Some researchers have suggested that, for heart disease, neurogenic and hormonal factors could operate, with cardiac electrical changes and increased secretion of catecholamines and cortisol.10 A longitudinal study of Swedish nurses found that those with the highest workload, defined by objective criteria, had the highest levels of plasma cortisol, the highest systolic blood pressure during working hours, and the most sleep disturbance.11 A British study found that bus drivers facing changes in traffic congestion experienced changes in job strain, which, in turn, could account for fluctuations in neuroendocrine activity.12

    Cognitive mediators may also play a part. For example, one could postulate that high perceived job demand may impair performance due to poorer memory and concentration and increased clumsiness and disorganisation, leading to increased errors, accidents, and safety hazards. Because of the time and energy commitment, high job demand may also interfere with access to the buffering effects of social support and to information and instruction on avoiding accidents. High workloads may also be associated with environments that are not conducive to learning how to perform tasks safely, nor to receiving help if there is an accident.1

    The actual physical strain of work may also pose a risk to life. This still occurs in many countries where there are extremes of climate and a poorly nourished, poorly organised labour force (sometimes including children) weakened by disease, and working long hours in physically demanding jobs. Such conditions lead to a high mortality from accidents and disease.13 The detailed circumstances and the contribution of high physical workload and other factors are rarely documented.

    It seems that overwork can kill, but that we know precious little about when, who, and how. Of course there have been big improvements in developed countries since the Industrial Revolution, but sometimes physical overwork has been replaced by psychological overload. And as unemployment has increased over the past decade, those in work have experienced increased workload, work pressure, and hours of work. If this is not to reap its predicted toll, we need much greater understanding about effects, mechanisms, and, most importantly, preventive strategies. We also need government strategies and legislation to increase employment, reduce the working week, and monitor and intervene to prevent health and safety hazards at work, which include overwork. This should involve making employers responsible for preventing work overload and stress, as well as providing help for individuals with work stress related illness.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    View Abstract

    Log in

    Log in through your institution


    * For online subscription