Health consequences of shift work and insufficient sleep
A semistructured review
“Health consequences of shift work and insufficient sleep” by Göran Kecklund and John Axelsson (BMJ 2016;355:i5210).
Why do the study?
Most doctors work shift patterns that are outside conventional daytime hours. Many other professions do so as well; almost 30% of workers in the United States work outside the hours of 0600-1800, and one in five employees in the European Union work nights at least once a month.
In the United Kingdom, changes to the junior doctor contract mean that doctors who work shifts are likely to see an increase in weekend work and twilight shifts.
But what are the health effects of working in a shift pattern? In the short term, we know that sleep cycles get disrupted, which can lead to fatigue and cognitive changes. But is shift work bad for your long term health? The study authors sought to answer this question by looking at the relation between shift work, its effect on sleep, and chronic health problems.
What did the authors do?
The authors performed a semistructured review, which examined published literature that looked at the exposure variables of shift work and short and/or poor quality sleep on various health outcomes, including heart disease, stroke, type 2 diabetes, obesity, depression, and death.
The authors had clear search criteria: they searched three major databases (PubMed, PsycINFO, and Cochrane Library) and included mainly systematic reviews and meta-analyses of case-control, prospective, or randomised controlled designs published between 2006 and 2016. Only articles written in English were included.
Exclusion criteria included systematic reviews and meta-analyses that included cross sectional or simulated studies, or data from non-working populations (such as adolescents). Apart from studies related to disturbed sleep—because data published on this exposure variable were limited—all articles not related to disturbed sleep met the inclusion criteria.
Box 1: Undertaking a literature review—to be systematic or not? (online only)
A systematic literature review and a semistructured review have different aims. A systematic review attempts to comprehensively answer a research question by looking at all published and unpublished literature that meet defined methodological criteria within a certain timeframe. It then draws comparisons between cohorts using statistics (meta-analyses). This differs from a semistructured review, which aims to answer a clear clinical question but does not usually have the same methodological rigour.
When deciding on which studies to include in a systematic review, authors must be clear on the selection criteria. For example: what happened if authors disagreed on which studies to include and how did they keep a checklist for the studies they included. Authors also need to consider whether the studies chosen have been subject to publication bias, which is a tendency among researchers and journal editors to publish results that are statistically significant, interesting, from large well funded studies, or of higher quality, which are more likely to be published, than work without such characteristics.
Lastly, systematic reviews usually include a formal process to rate the quality of scientific evidence included in order to develop recommendations that are as evidence based as possible using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach.
Semistructured reviews are more commonly written up in the form of narrative reviews, clinical reviews, or basic literature reviews. They do not usually have such a comprehensive search strategy as systematic reviews, as seen in this published review, but are still clear in what the authors did, and usually have inclusion/exclusion criteria. However, they rarely consider differences in study methods or study quality, although in this published review they did. Higher quality semistructured reviews comment on the soundness of the methodology of included studies, but usually do not have assessment of publication bias or utilise GRADE.
What did the authors find?
Shift work and sleep
The authors looked at the effect of shift work on sleep episodes that end close to the circadian trough (around 4 am); the effect on sleep according to the speed of rotation of shifts (slowly rotating schedules comprising four to seven consecutive shifts versus rapidly rotating schedules involving one to three consecutive shifts), and degree to which shift workers are able to sleep and recover between work shifts.
Seven prospective studies showed a trend towards an increased risk for shift workers to develop chronic sleep disturbances, such as insomnia (risk ratio 1.16, 95% confidence interval 0.97 to 1.38). The authors also concluded that typical sleep obtained before a night shift was 5-5.5 hours, as assessed by polysomnography.
Effect on health
The effect of shift night work, short sleep, and disturbed sleep can be seen in table 1 (Health risks associated with shift work, short sleep (4-7 hours), and disturbed sleep (insomnia symptoms)) in the paper (BMJ 2016;355::i5210).
Data from 34 observational studies, including over two million people, concluded that shift work was associated with myocardial infarction (pooled relative risk 1.23, 95% confidence interval 1.15 to 1.31) and ischaemic stroke (1.05, 1.01 to 1.09).
A meta-analysis including 12 cohort studies with 226 652 participants found that shift work increases the risk of developing type 2 diabetes. Pooled adjusted odds ratio of 1.12 (95% confidence interval 1.06 to 1.19) was found in shift workers compared with non–shift workers, which is in line with previously published cohort studies.
Whether shift work has an effect on mortality has rarely been studied compared with other outcomes. One meta-analysis found no association with cardiovascular mortality (risk ratio 1.04), whereas another showed a small, but significantly increased risk for cardiovascular death (1.03, 1.00 to 1.05). Non-significant differences were noted for cancer related death and all cause death.
Duration of sleep and sleep quality
Short sleep, defined as lasting four to seven hours, increased the risk for coronary heart disease, stroke, type 2 diabetes, obesity/weight gain, workplace based injuries, depression, and mortality.     Although evidence showing the association between short sleep and weight gain remains conflicted.
Poor sleep quality was associated with occupational injuries, coronary heart disease, type 2 diabetes, depression, and all cause mortality, including coronary disease mortality.     A meta-analysis looking at workers who found it difficult to initiate sleep found an association with type 2 diabetes, all cause mortality, and coronary disease mortality.  However, little evidence was published overall on the effect of poor sleep quality on health outcomes.
Strengths and limitations
One of the difficulties when interpreting semistructured reviews is the plethora of information that can be included, which makes it difficult to tease out what’s important and what’s not. Authors of semistructured reviews are compelled to show that the articles they have included are non-biased and comparable. In a semistructured review, however, authors are able to “cherry pick” evidence to support their claims and are not committed to comment on the quality of data they include. They can therefore insert references of incomparable cohorts. This allows for discussion on a greater inclusion of articles but puts more onus on the reader to scrutinise the quality of the references.
Well done semistructured reviews should briefly comment on the studies they select and include the type of study, enrolment of patients (prospective versus retrospective), adjustment for known confounders where relevant, and full descriptions of the statistical measure—that is, risk ratio along with confidence intervals.
In this review, the authors were careful to comment on what studies were included and why. For example, simulated shift work was excluded as this would not reflect real time working scenarios. They were also clear about their search strategy and the databases they used.
As semistructured reviews are not classified as original research, they can discuss other topics related to the research question. For example, nearly half of this review article focuses on discussion about the mechanisms linking shift work with adverse health outcomes and individual differences in resilience to shift work. This gives the reader more of an opportunity to consider the wider aspects of the clinical question and not just the evidence supporting or denying the association between shift work and adverse health outcomes.
What does the study mean?
Shift work might be linked to adverse health outcomes, including cardiovascular health and type 2 diabetes. Night work, short sleep, and poor sleep quality are risk factors that may put employees at higher risk of adverse health outcomes compared with other shift patterns, such as day work and twilight shifts.
So should we all opt out of specialties that are not nine to five? Not just yet. Much of the evidence included in this semistructured review showed an association between adverse health outcomes and shift work, but not necessarily that one causes the other.
The authors admit that the number of studies for several of the outcomes is small. Also, assessment of workers’ shift pattern differed across studies, and misclassification bias might have been introduced if a shift pattern was put into one category—that is, day shift—when it was meant to be in another—that is, twilight. Evaluation of exposure and whether it was uniform across studies should be considered in any future reviews of the evidence. Objective assessments of working times, sleep, physiological mechanisms, and health outcomes should be the focus of future, prospective, longitudinally designed studies.
As healthcare becomes increasingly led by consultants, with more senior cover in the evenings, nights, and at weekends, we might find that this is not just a problem for junior doctors, but for the rest of our careers. As with any medical research, more data are needed, but in this case, it’s doctors’ health that is at risk, not patients’.Neil Chanchlani, specialist trainee year 1 in paediatrics
University College London Hospitals NHS Foundation Trust, UK
Competing interests: None declared.
Provenance and peer review: Commissioned; not externally peer reviewed.
- Alterman T, Luckhaupt SE, Dahlhamer JM, Ward BW, Calvert GM. Prevalence rates of work organization characteristics among workers in the US.: data from the 2010 National Health Interview Survey. Am J Ind Med 2013;56:647-59. doi:10.1002/ajim.22108 pmid:22911666.
- Eurofound. Sixth European Working Conditions Survey 2015. www.eurofound.europa.eu/surveys/data-visualisation/sixth-european-working-conditions-survey-2015.
- BMA, NHS Employers, Department of Health. Junior Doctors Contract Agreement. 2016. www.acas.org.uk/media/pdf/l/p/ACAS_FINAL_AGREED_NHS_Emp_BMA_DoH_Package_180516.pdf.
- Wright KP Jr, , Bogan RK, Wyatt JK. Shift work and the assessment and management of shift work disorder (SWD). Sleep Med Rev 2013;17:41-54. doi:10.1016/j.smrv.2012.02.002 pmid:22560640.
- Lo JC, Groeger JA, Cheng GH, Dijk DJ, Chee MW. Self-reported sleep duration and cognitive performance in older adults: a systematic review and meta-analysis. Sleep Med 2016;17:87-98. doi:10.1016/j.sleep.2015.08.021 pmid:26847980.
- Linton SJ, Kecklund G, Franklin KA, et al. The effect of the work environment on future sleep disturbances: a systematic review. Sleep Med Rev 2015;23:10-9. doi:10.1016/j.smrv.2014.10.010 pmid:25645126.
- Åkerstedt T. Shift work and disturbed sleep/wakefulness. Occup Med (Lond) 2003;53:89-94. doi:10.1093/occmed/kqg046 pmid:12637592.
- Vyas MV, Garg AX, Iansavichus AV, et al. Shift work and vascular events: systematic review and meta-analysis. BMJ 2012;345:e4800. doi:10.1136/bmj.e4800 pmid:22835925.
- Gan Y, Yang C, Tong X, et al. Shift work and diabetes mellitus: a meta-analysis of observational studies. Occup Environ Med 2015;72:72-8. doi:10.1136/oemed-2014-102150 pmid:25030030.
- Lin X, Chen W, Wei F, Ying M, Wei W, Xie X. Night-shift work increases morbidity of breast cancer and all-cause mortality: a meta-analysis of 16 prospective cohort studies. Sleep Med 2015;16:1381-7. doi:10.1016/j.sleep.2015.02.543 pmid:26498240.
- Cappuccio FP, Cooper D, D’Elia L, Strazzullo P, Miller MA. Sleep duration predicts cardiovascular outcomes: a systematic review and meta-analysis of prospective studies. Eur Heart J 2011;32:1484-92. doi:10.1093/eurheartj/ehr007 pmid:21300732.
- Ge B, Guo X. Short and long sleep durations are both associated with increased risk of stroke: a meta-analysis of observational studies. Int J Stroke 2015;10:177-84. doi:10.1111/ijs.12398 pmid:25366107.
- Cappuccio FP, D’Elia L, Strazzullo P, Miller MA. Quantity and quality of sleep and incidence of type 2 diabetes: a systematic review and meta-analysis. Diabetes Care 2010;33:414-20. doi:10.2337/dc09-1124 pmid:19910503.
- Patel SR, Hu FB. Short sleep duration and weight gain: a systematic review. Obesity (Silver Spring) 2008;16:643-53. doi:10.1038/oby.2007.118 pmid:18239586.
- Cappuccio FP, D’Elia L, Strazzullo P, Miller MA. Sleep duration and all-cause mortality: a systematic review and meta-analysis of prospective studies. Sleep 2010;33:585-92.pmid:20469800.
- Magee L, Hale L. Longitudinal associations between sleep duration and subsequent weight gain: a systematic review. Sleep Med Rev 2012;16:231-41. doi:10.1016/j.smrv.2011.05.005 pmid:21784678.
- Kucharczyk ER, Morgan K, Hall AP. The occupational impact of sleep quality and insomnia symptoms. Sleep Med Rev 2012;16:547-59. doi:10.1016/j.smrv.2012.01.005 pmid:22401983.
- Sofi F, Cesari F, Casini A, Macchi C, Abbate R, Gensini GF. Insomnia and risk of cardiovascular disease: a meta-analysis. Eur J Prev Cardiol 2014;21:57-64. doi:10.1177/2047487312460020 pmid:22942213.
- Baglioni C, Battagliese G, Feige B, et al. Insomnia as a predictor of depression: a meta-analytic evaluation of longitudinal epidemiological studies. J Affect Disord 2011;135:10-9. doi:10.1016/j.jad.2011.01.011 pmid:21300408.
- Li Y, Zhang X, Winkelman JW, et al. Association between insomnia symptoms and mortality: a prospective study of US men. Circulation 2014;129:737-46. doi:10.1161/CIRCULATIONAHA.113.004500 pmid:24226807.
- Sutton AJ, Duval SJ, Tweedie RL, Abrams KR, Jones DR. Empirical assessment of effect of publication bias on meta-analyses. BMJ 2000;320:1574-7. doi:10.1136/bmj.320.7249.1574 pmid:10845965.
- Cochrane. The GRADE approach. 2016. http://handbook.cochrane.org/chapter_12/12_2_1_the_grade_approach.htm