Editorials

Shift work and cancer

BMJ 2009; 339 doi: http://dx.doi.org/10.1136/bmj.b2653 (Published 15 July 2009) Cite this as: BMJ 2009;339:b2653
  1. Lin Fritschi, head of cancer epidemiology
  1. 1Western Australian Institute for Medical Research, Nedlands, WA, 6012, Australia
  1. fritschi{at}waimr.uwa.edu.au

    Short and long term effects provide compelling reasons to act now

    The effect of shift work on cancer, particularly breast cancer, has received increasing interest from the lay media since a panel of the International Agency for Research on Cancer declared in 2007 that “shift work that involves circadian disruption is probably carcinogenic to humans.”1 This conclusion was based on sufficient evidence from animal studies and limited evidence from human studies. The evidence is strongest for breast cancer, although the risk of prostate and colorectal cancer may also be increased by shift work. Recently, 38 women with breast cancer who had previously worked night shifts for at least 20 years were compensated by the Danish national board of industrial injuries.2 This has implications for employers, compensation boards, and employees internationally.

    Two recent meta-analyses have suggested that the risk of breast cancer is increased by about 50% in night workers and by about 70% in flight personnel.3 4 If this association is true, shift work might have its effect by several different mechanisms (box).

    Possible mechanisms by which shift work increases the risk of cancer

    • Light at night suppresses the production of melatonin, which has direct and indirect anticancer effects5

    • Sleep disruption stimulates the hypothalmic-pituitary axis to release glucocorticoids, which results in depression of immune function6

    • Phase shift, in which the peripheral rhythms of functions such as digestion are out of phase with central sleep and wake rhythms. This may result in changes in the control of cell and tissue proliferation7

    • Shift work may result in changes in lifestyle factors such as smoking, diet, alcohol use, or exercise8

    • Decreased production of vitamin D9

    All these mechanisms are biologically plausible and are supported by experimental evidence. However, the human evidence lags behind because of problems in accurately defining and assessing the relevant exposure and the difficulty in controlling for confounding factors, especially in cohort studies in one profession.

    A range of health effects other than cancer may also be associated with shift work. Short term effects are relatively easy to study, and we have fairly consistent evidence that shift work increases fatigue, decreases quality of life, and increases the risk of injury. The possible longer term effects—which may include adverse pregnancy outcomes, coronary heart disease, gastrointestinal disturbances, and mental health disorders—are less clear.

    We now have a 24 hour society, and about one in five people in the Western world work shifts of some sort. We have no alternative but to find ways to mitigate the short and long term health effects of shift work on workers. Considerable effort has been invested in finding “better” shift systems. To decrease injuries, recommendations suggest having shorter length shifts, fewer shifts before a rest day, stable rather than rotating shifts, and frequent rest breaks or scheduled napping within a shift.10 11

    The aviation industry has an intense interest in research in this area, and the Aerospace Medical Association recommends frequent rests during shifts, including napping.12 Unfortunately, it also concludes that the major differences in individual responses to sleep loss, sleep disruption, and time zone transitions make it impossible to develop a “one size fits all” shift schedule, and it is aiming to develop biomathematical models of fatigue to predict performance and alertness at an individual level.

    Other researchers have concentrated on countermeasures that increase sleep duration, promote quick adaptation to night work, or improve subjective wellbeing at work. One possibility is to use our increasing understanding of the physiological control of the sleep-wake cycle to time our exposure to light and darkness for maximum adaptation. Using phototherapy lamps (especially those producing blue light, which is most efficient in resetting melatonin release time), wearing goggles, wearing sunglasses when driving home, and darkening bedrooms or wearing sleeping masks are being tried. Medications that are stimulants, hypnotics, or chronobiotics (substances that control the body clock) are also being used.11 It may also be possible to screen workers to select those with factors that seem to be associated with better tolerance of shift work, such as being an “evening person,” having better family support, and having fewer responsibilities at home.11

    Most of this evidence on shift work relates to mitigating short term outcomes, and it is not known whether any of these measures help prevent the potential longer term effects of shift work. The only evidence on mitigation of long term effects is the suggestion that only longer term shift work (20-30 years) increases the risk of breast cancer.3

    So where does that leave us in our attempt to answer the questions posed by workers, employees, and unions about shift work and breast cancer? Should employers warn longer term shift workers about the risks, offer mammogram screening programmes, or reorganise shift schedules to minimise short term effects? From a public health perspective we probably have enough evidence that shift work is harmful in the short term for us to make several recommendations. Shift lengths should be shorter; rest breaks should be included; and researchers should educate shift workers and employers as to how sleep-wake cycles are controlled and how this knowledge can be used to maximise sleep quality, sleep duration, and alertness at work.

    Notes

    Cite this as: BMJ 2009;339:b2653

    Footnotes

    • Competing interests: None declared.

    • Provenance and peer review: Commissioned; not externally peer reviewed.

    References