22nd International Symposium on Shiftwork and Working Time: Challenges and solutions for healthy working hours

Introduction

This decade has brought unprecedented progress in the documentation and understanding of the associations between working time, sleep schedules and circadian rhythms on the one hand and performance, safety and health outcomes on the other. Big epidemiologic studies and large prospective cohort studies strengthen the evidence base regarding these complex associations. Impactful changes in hours of service regulations and work policies across the globe, and fatigue risk management efforts associated with these changes, yield new insights into the neurobiological, logistical, psychosocial, behavioural, legal and financial issues to be considered when evaluating and optimizing work schedules. Basic research discoveries regarding peripheral circadian rhythms, intricate cascades of metabolic processes and critical interactions with the microbiome shed new light on the vulnerability of shift workers to long-term adverse health outcomes such as gastrointestinal and cardiovascular disease and increased risk of cancer. In terms of making a difference, these are exciting times for working time research, which has become more multidisciplinary than ever.

In June 2015, researchers from all around the world came together in Elsinore, Denmark, for the 22nd International Symposium on Shiftwork and Working Time. The overarching theme was “challenges and solutions for healthy working hours.” This special issue of Chronobiology International contains a collection of peer-reviewed papers stemming from the meeting, which together reflect the large diversity of disciplines and methods involved. Here, we give a brief overview of the articles that make up the special issue and provide some context for the main findings.

Methodological advances in shift work research

Arguably, the invention of electric light was the main catalyst for the prevalence of shift work in modern times. Stevens (2016) discusses the impact of electric light on endogenous circadian rhythmicity and proposes research to better understand circadian disruption from electric light and its effects. These effects need to be considered in view of the interplay of circadian rhythms and the homeostatic regulation of sleep and wakefulness.

Figure 1 illustrates this interplay with regard to fatigue. The timing of work periods determines the impact of the endogenous circadian pacemaker through the pressure for wakefulness it exerts as a function of time of day. The duration of work periods determines the impact of the homeostatic regulation of sleep and wakefulness through a build-up of pressure for sleep across time awake. The timing and duration of work periods also determine exposure to environmental light, which may cause circadian rhythms to shift in time.

Figure 1. Risk of errors, incidents and accidents in shift work is determined by interaction of the operational environment with the neurobiology of sleep and fatigue. Triangle: homeostatic drive for sleep (increasing during wakefulness, decreasing during sleep). Circle: circadian drive for wakefulness (oscillating across the 24 h of the day). Dotted-outline arrows: transient effects. Figure adapted from Satterfield and Van Dongen (2013) with permission.

The principal cause underlying operational risks in shift work is a misalignment of the endogenous circadian and homeostatic regulatory processes relative to night and day. As such, the timing and duration of work hours are by themselves, justifiably, a major topic of investigation in shift work research. Two papers in this special issue present advances in methodology that pave the way for future studies in this area.

Garde et al. (2016) addressed the question of how to best define and quantify exposure to night shift schedules in the context of behavioural and health outcomes. They explored this issue by comparing a number of possible definitions found in the literature on shift work and breast cancer – e.g. working any hour between 01:00 and 04:00 or starting work between 22:00 and 06:00. Garde and colleagues applied these definitions to the work schedules of 234 297 individuals in the Danish Working Hour Database, primarily drawn from the health care sector. They found much overlap in the quantification of exposure to night shift schedules among definitions based on whether or not an individual is working during a given time period at night. The quantification of exposure varied more considerably when definitions based on shift start and end times were used, suggesting that such definitions may be less optimal for studying whether shift timing has behavioural or biological impact.

Kosmadopoulos et al. (2016a) reported on a new algorithm for actigraphy (activity monitoring). The gold standard for measuring sleep in workplace settings is wrist actigraphy – in conjunction with a sleep/wake log that is used to document bedtimes and also to record periods when the actigraph is off-wrist. Kosmadopoulos and colleagues developed an actigraphic data processing algorithm for studies in which reliable records about off-wrist times are not available. The algorithm distinguishes periods of inactivity from non-wear times, based on data from the actigraph alone.

Trends in the organization of working time

New Ways of Working (NWW) is a new work schedule and work organization approach characterized by flexibility in working time and work location, aided by information and communication technologies and performance-based management. Using a quasi-experimental approach in a large financial company in the Netherlands, Nijp et al. (2016) studied a group of 361 individuals using the NWW approach to working. They were assessed 1–2 months before, 4 months after and 10 months after implementation of NWW and were compared over time and relative to a non-NWW control group of 80 individuals. The NWW group showed a considerable shift from hours worked at the office to hours worked at home after the NWW implementation, with accompanying reductions in commute time. Working time continued to be daytime oriented and primarily during week days, not weekends. Psychosocial factors, perceived stress, fatigue, organizational commitment, work performance and job satisfaction were largely unaffected. Self-reported health declined after NWW implementation, suggesting a possible need for health monitoring when introducing NWW-like flexibility into the workplace.

Using data from a nationally representative survey conducted in Sweden, Tucker et al. (2016) investigated a 6-item scale of work time control, which they explored as a potential predictor of accidents at work or during leisure time. Self-reported sleep disturbance and frequency of short sleeps were examined as potential mediators of associations between work time control and accidents. Work time control was inversely associated with accidents, with greater control over time off work predicting fewer accidents in particular. Sleep disturbance was found to be a weak mediator of this relationship, whereas frequency of short sleeps was not a significant mediator. As such, while work time control was associated with reduced risk of subsequent involvement in an accident, whether or not sleep underlies this association requires further investigation.

Work and sleep scheduling tends to be more complex in individuals who work multiple jobs. Using American Time Use Survey data, Marucci-Wellman et al. (2016) compared sleep duration per 24 h between workers holding multiple jobs and workers holding a single job. Workers with multiple jobs obtained less sleep due to a number of dissociable factors, such as working longer hours and working late at night more frequently. Working both primary and secondary jobs on a weekend day interfered with recuperative sleep on the weekend, which was especially evident in male workers in the sample. Night work had a detrimental effect on sleep duration particularly in female workers. The reasons for these differences remain unclear.

An under-investigated issue in working time research is whether or not shift work interferes with being a skilled worker. In a sample of 285 night and rotating shift workers in Norway, Foldal et al. (2016) investigated whether there is an association between shift workers’ self-perceived mastery of work and select personality traits. They found that neuroticism and conscientiousness predicted perceived mastery of work, but dispositional resistance to change – a possible correlate of shift work tolerance – did not.

Shift work and sleep

Night and shift workers face a dual challenge: they are asked to work at circadian times not optimal for alertness and to rest at times not conducive for sleep. Roach et al. (2016) examined sleep strategies used in breaks between consecutive shifts in 253 shift workers. They found that if a break included an entire night, workers tended to have a single sleep period. If a break occurred predominantly during the daytime, then more than one sleep period during the break was common (41%–50% of occasions). These data shed light on how shift workers schedule and split their sleep – which is important information for, e.g. predicting sleep/wake behaviour in model-based fatigue risk management (Darwent et al., 2010; Van Dongen, 2010).

Vincent et al. (2016) measured sleep by means of wrist actigraphy in a sample of 33 salaried firefighters in Australia, who were involved in planned (preventive) wildland burn operations. The authors found that self-reported fatigue was greater on burn days compared to non-burn days, but there were no differences in sleep measures (regardless of work shift time or sleeping location). Sleep was curtailed by approximately 0.5 h, however, when planned burn shifts were greater than 12 h.

A similar effect of long days was observed in elite athletes by Sargent and Roach (2016). The sleep of Australian Rules Football competition players was less after the evening game than after a day game. Whether the reduction in sleep after evening games impaired physical recovery remains to be determined.

Lammers-van der Holst et al. (2016) reported on sleep data from a longitudinal study of 42 police officers, measured first while in training at the Dutch Police Force Academy and then approximately 4, 12 and 20 months after entering the work force and commencing shift work. In agreement with initial findings reported when the study was just begun (Lammers-van der Holst et al., 2006), there were systematic inter-individual differences in actigraphically estimated sleep duration and subjective sleep quality. Self-reported flexibility of sleeping habits predicted the duration and quality of daytime sleep. Perhaps not surprising, subjective quality of nighttime sleep also predicted subjective quality of daytime sleep. Furthermore, women exhibited greater actigraphically assessed daytime sleep durations than men. This finding was unexpected, but the authors point out that the women in the sample were on average younger than the men. They may not have had child-rearing and other domestic duties that may otherwise contribute to increased sleep curtailment in female shift workers.

Split sleep and napping as countermeasure strategies

On-call work has the potential to disrupt sleep, but it may also allow for napping between calls. However, apprehension about receiving (or missing) a call may reduce the recuperative potential of nap sleep. Also, sleep inertia after a nap (i.e. transient impairment immediately after awakening) may interfere with on-call job performance. Reviewing the relevant literature, Ferguson et al. (2016) conclude that it is generally a good strategy to get some sleep while on call, even if the sleep is interrupted by calls or otherwise reduced in quality. However, the potential for sleep inertia needs to be managed, for example by means of caffeine intake (Van Dongen et al., 2001; Newman et al., 2014).

Sleep inertia may also put workers at risk in split sleep scenarios. Hilditch et al. (2016) investigated sleep inertia associated with split sleep in split duty schedules. Sixteen healthy adults participated in a laboratory study with two baseline nights followed by 4 days on a 6- on/6-h off split duty schedule. Two alternating schedules were evaluated – one with wake-up times at 02:00 and 14:00 and one with wake-up times at 08:00 and 20:00. Performance was impaired and subjective sleepiness was increased during the first hour after awakening, providing evidence of sleep inertia. Importantly, morning wake-ups (02:00 and 08:00) showed the greatest impact of sleep inertia.

Napping has the potential to help improve long-term health. In a cross-sectional study of 2588 female nurses in Rio de Janeiro, Brazil, conducted by Rotenberg et al. (2016), nurses were informally allowed to nap for up to three consecutive hours during night shifts. Workers provided self-reported information on physician diagnosis of hypertension, napping and sleep complaints. Both current and former night workers showed increased probability of hypertension compared to workers with no previous experience in night work. Among the night workers, however, nappers were less likely to report hypertension than non-nappers. This was particularly evident among those who self-identified as poor sleepers.

A cautionary note about the benefits of napping is provided by Centofanti et al. (2016), who investigated the duration of performance benefits gained from short nap opportunities during simulated night shifts. A group of 31 healthy subjects was randomly assigned to a 10-min or a 30-min nap opportunity ending at 04:00 or a no-nap condition. Performance and subjective sleepiness were measured 1.5 and 2.5 h after the nap. The 30-min nap reduced post-nap subjective sleepiness, but neither of the short nap conditions had a significant effect on performance. These data are not consistent with the idea that brief naps (“power naps”) are effective at mitigating performance impairment during night shifts.

The results of Centofanti and colleagues are congruent with another study in this special issue of Chronobiology International, conducted by Kosmadopoulos et al. (2016b). When extended wakefulness precedes the rotation to a night shift schedule, performance impairment may be greater on the first of a series of night shifts than on subsequent night shifts, and Kosmadopoulos and colleagues investigated whether a brief nap prior to the first night shift could undo this effect. In a laboratory simulation of two 12-h night shifts, they observed no difference in performance measures and subjective sleepiness between the two night shifts when a 1-h afternoon nap preceded the first shift.

Morningness/eveningness and working time

While morningness/eveningness, or circadian phase preference, has been a focus of investigation in working time research for many decades (e.g. Kerkhof, 1985), there has been renewed interest in this topic in recent years. In a study of day and shift workers who participated in a health check program (Yong et al., 2016), 2474 individuals filled out a modified version of the Munich chronotype questionnaire (MCQ; Roenneberg et al., 2003) as well as the work ability index (WAI; Tuomi et al., 1994). Based on the MCQ data from this study, Yong and colleagues estimated workers’ mid-sleep times on shift days and free days, from which they derived an index of circadian misalignment during work days (“social jet lag”). The authors also estimated workers’ sleep durations and their self-perceived health. Greater circadian misalignment and shorter sleep duration, but not circadian phase preference per se, were associated with decreasing self-perceived health, while longer sleep duration counteracted the effect of circadian misalignment.

The relationship between morningness/eveningness and food intake was examined by Mota et al. (2016). Seventy-two resident physicians kept a food diary for three non-successive days. They also reported physical activity, sleepiness and sleep duration and quality. Eveningness was associated with more cholesterol and with increased consumption of sweets and vegetables. Eveningness was also associated with less physical activity and shorter sleep duration. Thus, although the evening-types ate more vegetables than morning-types, the evening-types generally showed more unhealthy behaviours.

Silva et al. (2016) also reported relationships between morningness/eveningness and food intake. Among 204 Brazilian university students, morningness/eveningness in interaction with perceived sleep debt affected meal timing and food types consumed. A difference in endogenous circadian phase position underlies the morningness/eveningness trait (Kerkhof & Van Dongen, 1996), suggesting that biological clock differences may affect health outcomes.

Shift work and physical and mental well-being

Well-being is a multidimensional concept that can be hard to quantify. Yet, it stands to reason that well-being is an important factor for worker motivation and retention. Several papers in this special issue explored the aspects of well-being in shift workers.

Moreno et al. (2016) investigated insomnia and musculoskeletal pain symptoms in airline pilots, rural workers and factory workers. Rural and night work were found to predict insomnia and pain outcomes. Musculoskeletal pain was also found to be a predictor of insomnia, but the relationship did not appear to be bidirectional (i.e. insomnia did not predict musculoskeletal pain).

Associations between work stressors and work ability – being able to perform one’s job given the physical and mental demands – were assessed by Martinez et al. (2016) in a cohort of 498 hospital workers. It was found that work stressors associated with perceived lack of job control, limited social support, effort-reward imbalance, over-commitment and work-related pain or injury had a negative effect on work ability over a 4-year time span.

In a Swedish study by Dahlgren et al. (2016), a sample of 1459 nursing students completed a questionnaire prior to graduation and 3 years after graduation. The authors explored the effects of quick returns, defined as intervals of less than 11 h between the end of one shift and the start of the next, on a number of self-reported symptoms. Although quick returns did not predict self-reported recovery after rest days, more frequent quick returns were found to be associated with poorer sleep quality, more frequent short sleeps, problems unwinding, exhaustion, less satisfaction with work hours and more work–family interference.

Focusing on specific shift schedules, Karhula et al. (2016) compared a 12- shift system to 8-h fast and 8-h slow forward-rotating shift schedules in a group of 599 predominantly male Finnish industrial employees. The 12-h shift system compared favourably to the 8-h shift systems in terms of shift schedule satisfaction, self-reported sleep duration and quality, self-reported alertness, perceived general health and work–life balance. The 8-h slow forward-rotating shift system compared least favourably on these survey outcomes. Shift schedule satisfaction did not vary between younger and older employees in the study. While a number of study limitations preclude firm conclusions about the implications of 12-h versus 8-h shift scheduling for sleep, alertness, productivity, safety and health, the survey results highlight multiple reasons why, in Europe, 12-h shift systems are gaining popularity (Estryn-Behar & Van der Heijden, 2012).

These results add to mounting evidence of a link between working time and well-being. However, there appear to be large individual differences in the effects of working shifts or irregular hours (Härmä, 1993). In the laboratory, symptoms of poor physical and emotional well-being have been connected with sleep loss (Haack & Mullington, 2005). It is conceivable, therefore, that the extent to which workers experience physical or mental symptoms may be related to their individual vulnerability to sleep loss (Van Dongen, 2006).

Working time and health

In order to give recommendations regarding the optimal organization of work hours, it is important to compare work scheduling practices and find out which are associated with better health outcomes and which are associated with worse health outcomes.

Siqueira et al. (2016) conducted a panel study of 372 nursing professionals from a hospital in Rio de Janeiro, Brazil, each participating in two cross-sectional surveys approximately 7 years apart. For those who changed from a day shift schedule to a night shift schedule during the time between the two surveys, the odds of a weight gain of more than 5 kg were doubled compared to those who changed from night to day shift or stayed on day shift or night shift. This effect may be due to a combination of shorter sleep, circadian misalignment, light exposure at night and unhealthy food intake often found to be associated with night work.

In a study by Reinhardt et al. (2016), the inflammatory marker interleukin-6 (IL-6) was assessed from saliva in permanent morning shift workers. The workers were stratified, on the basis of actigraphically assessed sleep duration, as workers with longer sleep duration (6 subjects) and shorter sleep duration (15 subjects). The longer sleepers showed a drop in IL-6 levels in the afternoon and evening, which was absent in the shorter sleepers. This result suggests that sleep restriction in morning shift workers may impact on systemic inflammatory processes, although the validity of salivary IL-6 as a marker thereof requires further investigation.

Consistent with the classification of shift work with circadian disruption as a probable human carcinogen by the International Agency for Research on Cancer of the World Health Organization (Straif et al., 2007), the evidence of a link between night work and increased risk of breast cancer is increasingly strong (Lin et al., 2015). Using data from a French case-control study including 975 cases and 1317 controls, Cordina-Duverger et al. (2016) explored the association of night work with breast cancer subtypes defined by tumour status for oestrogen receptor, progesterone receptor and human epidermal growth factor receptor 2. The authors found increased odds ratios for associations of night work with receptor-positive subtypes of breast cancer, particularly in premenopausal women.

Conclusion

The health consequences of night and shift work are wide-ranging and diverse. Although many factors contribute, circadian misalignment and sleep loss play a critical role, as illustrated in Figure 2.

Figure 2. Circadian misalignment and sleep loss contribute to poor health outcomes. The diagram illustrates several key pathways involved (see, e.g. Kryger et al., 2016), although the underlying mechanisms are more multifactorial and complex. Adipokines: cytokines (cell signalling proteins) secreted by fat tissue and involved in immune responses. Endothelial dysfunction: imbalance between vasodilating and vasoconstricting substances acting on the endothelium, the inner lining of blood vessels. Estrogen, testosterone: sex hormones. HPA axis: hypothalamic–pituitary–adrenal axis – complex interactions between the hypothalamus, pituitary gland and adrenal glands implicated in stress response. Insulin resistance: impaired cell response to insulin, a hormone regulating metabolism and utilization of energy from food, especially glucose. Leptin, ghrelin: hormones signalling satiety and hunger, respectively. Melatonin: hormone, secreted at night, that anticipates the daily period of darkness and protects against cell damage from free radicals. Metabolic syndrome: a cluster of symptoms that may lead to diabetes and cardiovascular disease. ↑ increased or activated; ↓ decreased.

Understanding the neurobiology of circadian rhythms and sleep and their interaction with working time and the environment (see Figure 1) has been instrumental in the development and implementation of fatigue countermeasures. We envision that a better understanding of the mechanisms by which working time may lead to adverse health outcomes, through circadian misalignment and sleep loss as well as social and environmental factors, will similarly lead to the development and implementation of health-related countermeasures. As reflected in this special issue of Chronobiology International, researchers from around the globe are working hard to achieve that goal.

Declaration of interest

We would like to acknowledge the sponsorship of institutions and associations that gave substantial support to the 22nd International Symposium on Shiftwork and Working Time: CO-industri, Confederation of Danish Industry, FOA, Danish Regions and The National Research Centre for the Working Environment.

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Acknowledgements

The 22nd International Symposium on Shiftwork and Working Time was held by the Working Time Society (WTS) and the International Commission on Occupational Health (ICOH) subcommittee of Shiftwork and Working Time. The meeting was organized by The National Research Centre for the Working Environment in collaboration with colleagues from the Department of Public Health, University of Copenhagen, The Danish Cancer Society Research Center and the Department of Occupational Medicine, Danish Ramazini Centre, Aarhus University Hospital. We thank the publisher, Taylor & Francis, and the editor, Prof. Francesco Portaluppi, of Chronobiology International for their support to the Working Time Society in publishing this special issue. We are grateful to the reviewers of the submitted manuscripts for contributing their valuable time and expertise.