Statistics from Altmetric.com
With a recent study in this journal, Hansen and Lassen1 followed a promising path to answer pressing questions with regard to postulated causal links between shift work and cancer.2 By employing diurnal preference to characterise the individual's chronobiological propensity to work and sleep or rest, the authors may have used a marker of susceptibility to stress and strain experienced by individuals who work at chronobiologically unusual times. This novel approach may have captured information relevant to the undefined link, namely ‘circadian disruption’, in the ‘probable’ chain of cancer causation in shift workers which experts of the International Agency for Research on Cancer (IARC) identified in 2007.3 Here is why and how we could diligently follow-up on this thought-provoking case-control study.
First, some remarkable observations: In this study of night-shift work among women in the Danish military, diurnal preference was associated with differential breast cancer risks. For those with 884 or more cumulative night shifts, the OR for morning types was 3.9 (95% CI 1.6 to 9.5) and 2.0 (95% CI 0.7 to 5.8) for similarly exposed evening types; in the so-called intermediate types no increased risks were detected. The information on night work was collected before IARC's aforementioned classification, which lends methodological weight to the observations of this case-control study.
Second, some necessary information: When Hansen and Lassen refer to ‘morning type’ or ‘evening type’ preference, this corresponds colloquially to ‘larks’ and ‘owls’ to denote in what time window over 24 h an individual's body functions (including physiology, endocrinology, metabolism, behaviour) render him/her most active or passive or make him/her want to be awake or asleep. It seems intuitive that individuals who wake up very early in the morning (larks) and are most alert in the 8–16 h that follow would also work in this time window with relative ease. Conversely, individuals who prefer to sleep in (owls) are most alert in late evening hours and can be expected to perform with greater ease in much later time windows than early types. Looking at the ‘negative’ side of the chronotype: larks and owls likely experience more stress and strain when working outside their preferred time windows of activity, that is, at night or early in the morning, respectively, than their intermediate counterparts.
That individuals’ sleep–wake preferences vary is common sense. Scientifically, chronotype or the internal propensity to wake and sleep has been shown to depend on genes4 ,5 and on sex, age6–8 and environment8 (wake–sleep behaviour can be significantly confused by the strong Zeitgeber light when experienced at chronobiologically unusual times). That extreme chronotypes shall experience more stress and strain when they rotate into—or work permanently in—time windows which are chronobiologically unusual for them is supported by initial data9 and may be mirrored in selection processes. In the latter vein, Grundy et al10 found fewer workers with extreme chronotypes in their sample of nurses who worked both at day and night than was to be expected when compared with—albeit very limited—observations in middle-aged working adults.11 ,12
Interestingly, from a causal framework point of view, the Danish study provided first support for what was predicted in 2010:13 ‘So-called ‘owls’ are likely to suffer considerably less from night-shift work than the ‘lark-types’ in human populations. Ultimately, therefore, the chronotype may explain why among workers, who share the same shift work exposures, some may develop cancer and others not or why some develop it earlier and others later’.
Hansen and Lassen emphasise that ‘it will be important to test our results in future studies’.1 To appropriately capture chronotype information in such studies, epidemiological research should at least include a single question like Hansen and Lassen did because it may lead far. In their Discussion section, Hansen and Lassen justify this seemingly ‘(over?)simplistic’ approach by a reference to work by Roenneberg et al.14 The assessment of diurnal preference from a single question ‘may have resulted in imprecision, although it has been shown that answers to a similar single question on diurnal preference correlated well with answers on more comprehensive questionnaires’.14 Indeed, when chronobiological propensity assessments were systematically compared in 2007, study participants were asked to self-rate their individual chronotype out of seven categories. The conclusion then—to which Hansen and Lassen refer with the aforementioned quote—was: ‘It is remarkable that an introduction combined with a single question of self-assessing one's chronotype gives almost the same results as a questionnaire consisting of 19 items’.14 We must bear in mind, though, that answers to a single question may differ—at the conceivable cost of being able to compare and merge data across studies—depending on the culture, country and language in which the study takes place. Where feasible, therefore, studies may consider using the validated two-page Munich Chronotype Questionnaire (MCTQ);8 for individuals who work in rotating shifts, a modified version called the MCTQshift was envisaged in 2010.15 Finally, we should include age-associated chronotype changes in our research. To exemplify, what shift workers may have tolerated in certain time windows when they were younger may not be advisable in the same time windows as they age and vice versa. To take note of such chronobiological propensity changes over a lifetime, prospectively followed cohorts such as The Nurses’ Health Studies could offer intraindividual assessment of chronotypes, and of their possible variation, in the course of regular follow-up questionnaires. Case-control studies may work out sensible questions for relevant age ranges similar to the MCTQ chronotype categories: ‘As a child, I was …; As a teenager, I was …; In case you are older than 65: In the middle of my life, I was …’.8
Ultimately, we may want to define shift work specifically for an individual as ‘work at chronobiologically unusual times’, that is, the time window of the work is not readily compatible or ‘in phase’ with the individual's ‘time-of-day-type’. To determine when this is and for whom, we may use the chronotype as it reveals the genetically determined propensity of when individuals prefer to time their sleep and work in reference to local time. Such a paradigm shift in what we label as shift work for whom and when, implying the need for clarification of when individuals are actually ‘exposed’ or ‘not exposed’ to shift work involving circadian disruption, may be(come) a necessary means to achieve two ends: to reliably research shift workers’ possible health risks, including cancer, and to effectively prevent disease in working populations who differ chronobiologically as evinced by their chronotypes.
Overall, Hansen and Lassen provided a promising precedence of using diurnal preference or chronotype as a key candidate to zero in on the possible susceptibility of female night-shift and rotating-shift workers when working at chronobiologically unusual times. The case for following up chronobiology-driven facets in the Danish study may be even stronger after Parent et al16 reported increased cancer risks with regard to a whole series of endpoints (lung, colon, bladder, prostate, rectum and pancreas) in male night-shift workers.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.