Article Text
PDF
Abstract
Introduction The objective of this study is to determine the effects of night work, Arctic seasonal factors and cold working environments on human functions relevant to safety. The study aims to quantify the contribution of (1) several consecutive night shifts, (2) seasonal variation on sleepiness, alertness and circadian rhythm and (3) whether a computational model of sleep, circadian rhythms and cognitive performance can accurately predict the observed sleepiness and alertness.
Methods and analysis In an observational crossover study of outdoor and indoor workers (n=120) on a three-shift schedule from an industrial plant in Norway (70 °N), measurements will be conducted during the summer and winter. Sleep duration and quality will be measured daily by smartphone questionnaire, aided by actigraphy and heart rate measurements. Sleepiness and alertness will be assessed at regular intervals by the Karolinska Sleepiness Scale and the psychomotor vigilance test, respectively. Saliva samples will assess melatonin levels, and a blood sample will measure circadian time. Thermal exposures and responses will be measured by sensors and by thermography.
Ethics and dissemination All participants will give written informed consent to participate in the study, which will be conducted in accordance with the Declaration of Helsinki. The Norwegian Regional Committee for Medical Research Ethics South-East D waivered the need for ethics approval (reference 495816). Dissemination plans include academic and lay publications, and partnerships with national and regional policymakers.
- Health informatics
- Case-Control Studies
- Fatigue
- GENETICS
- OCCUPATIONAL & INDUSTRIAL MEDICINE
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Strengths and limitations of this study
Contributes to knowledge gaps on how Arctic seasonal factors affect sleepiness, alertness and circadian rhythm in a field study with real-world conditions.
Aims to develop a mathematical model that can predict sleepiness and alertness for Arctic conditions.
Aims to determine seasonal variation in circadian time with a single blood sample.
Limitations of the study is that no interventions will be performed.
Introduction
Work during nights and extended working hours are both occupational safety hazards for workers and for the environment.1 2 It has been estimated that 300 000 deaths annually can be attributed to occupational injuries worldwide.3 Being awake and active during the night perturbs circadian rhythms.4 Extended hours awake are associated with fatigue and increased homeostatic sleep drive.5 Both factors require that the employee activate psychophysiological response systems to maintain vigilance and attention, to overcome fatigue and sleep propensity, reducing the risk for performance failures.
Shift work can cause transient and long-term disruption of the circadian system, including ‘circadian misalignment’, which arises when sleep/wake behaviours are mistimed in relation to the endogenous circadian rhythm. Circadian misalignment increases risk for safety incidents.6 Little is known about the effects of circadian misalignment over multiple days.7 Sleep disruption, circadian misalignment and cognitive impairment have been linked to almost twofold increased risk for work-related injuries among shift workers compared with daytime-only workers8 and as well as those working long days or weeks.1 Particularly, working several consecutive long night shifts with insufficient breaks increases the risk.2
The strength and duration of natural light depends on season and geographical latitude. The Arctic, north of the polar circle (66 °N), has less than 5 hours of daylight from sunrise to sunset in the winter. At 70 °N the polar night lasts for approximately 2 months, a period for which the sun never rises above the horizon. Hence, darkness is a challenge for maintaining circadian rhythms, given that light is the most important time cue for the central circadian clock. Conversely, continuous light during the polar day may pose other problems for the circadian system. Entrainment characteristics of a circadian clock in the Arctic light environment deviates from a standard 12/12-hour light/dark regimen. In Arctic regions north of the 75 °N latitude, reduced efficiency, duration and quality of sleep and difficulties initiating sleep have been reported.9 Part of the explanation for these effects may be that lack of daylight weakens the entrainment of circadian rhythms.10 11 A more recent study during the Antarctic winter at 75 °S with long-term daylight deprivation, showed that circadian rhythm stability decreases and sleep-wake timing is delayed.12 However, at 69 °N latitude, the effects of prolonged darkness on sleep seem less severe.13 14 It is not evident that seasonal variation in daylight exposure will affect cognitive function. Studies from the Antarctic winter showed no significant deterioration in cognitive function.15 16 However, exposure to a dark cave environment was associated with increased reaction time in a systematic review.17 However, participants in both Antarctica and in the cave lived under very restricted and unusual conditions.18 Therefore, it is necessary to conduct ecologically valid studies on workers who typically reside in the Arctic to acquire knowledge regarding the impact of significant seasonal fluctuations in natural light exposure on circadian rhythms, sleepiness and alertness. As the present study will include both indoor and outdoor workers, seasonal variation in natural daylight exposure will vary between these groups of workers.
A recent systematic review indicated that exposure to cold air or water impaired cognitive performance in healthy participants.19 Cold exposure may increase the risk for safety incidents20–22 by potentiating effects on fatigue and sleepiness.23 Given the central circadian clock regulates daily rhythms in body temperature, circadian misalignment in shift workers could potentially alter thermal regulation and tolerance to cold exposure. The association may also go in the opposite direction, as circadian rhythms may synchronise to periodic changes in ambient temperature.24 Hence, exposure to cold temperatures could affect the circadian rhythm and lead to circadian misalignment by a phase shortening. However, the impact of cold exposure on circadian misalignment is unknown in humans living and working in the Arctic, which the present study will address.
There is support for environmental light conditions affecting sleepiness. However, there is a paucity of knowledge of the interaction effects of season factors, night work and cold exposure on sleepiness and alertness. Furthermore, effects of several consecutive night shifts should be elucidated, although a recent study among police officers indicates no effect on sleepiness.25
A variety of computational models have been developed for predicting worker performance as a function of sleep timing, circadian rhythms and other environmental factors.26 27 To date, however, these models have not been tested in extreme environmental conditions including seasonal extremes of photoperiod and harsh ambient temperature. Comparing model predictions with data obtained under these conditions will allow for further model development to enable accurate predictions under a wider range of workplace settings.
Overall aims and objectives
The overall aim of the present study is to elucidate effects of working hours, seasonal variations and exposure to cold on human factors of significance for reducing the risk level for safety incidents. The primary aim is to determine whether seasonal variations in temperature and light affect the detrimental effect of consecutive night shifts on sleep, circadian alignment, alertness and sleepiness among workers in the Arctic. The secondary aim is to test and expand a computational model of sleep, circadian rhythms and cognitive performance under the conditions of large seasonal variation associated with the Arctic.
The following research objectives will be addressed:
Determine the impacts of several consecutive night shifts versus several consecutive day shifts on.
Sleepiness and alertness.
Insomnia symptoms.
Circadian rhythm.
Relevant physiological and molecular biomarkers.
Determine whether seasonal variation in light and temperature in the Arctic moderates these effects.
Test whether a computational model of sleep, circadian rhythms and cognitive performance can accurately predict the observed sleepiness and alertness during night and day shifts in process operators in the Arctic regions.
Methods
Setting and study population
For the present observational crossover cohort study, we will recruit participants among 120 process operators of all genders employed at an industrial plant in Northern Norway at 70 oN. Most of the workers work outdoors on the plant, whereas a smaller portion work indoors in control rooms. Most of the workers live in the vicinity surrounding the plant, while some workers live in other parts of Norway and are fly-in/fly-out commuters. Workers will be informed about the study via the company’s intranet, by paper flyers in the canteen and by research staff from the National Institute of Occupational Health (STAMI) in a start-up meeting 6 weeks prior to study start.
Data will be collected during the first 25 days of the workers’ regular 6-week shift plan, based on the operational procedures of the employer. The period consists of seven day shifts (D), four evening shifts (E) and seven night shifts (N) with some days off in-between (figure 1). After workday E4 participants have 17 days off, during which no data collection takes place. Data collection will be repeated twice, once during the winter season and once during the summer season (figure 2). The workers will not receive any personal compensation for their participation, but they will have the opportunity to participate in a lottery where the prize is an electric bike.
Six-week shift plan and data collection. Sleepiness, fatigue and PVT will be measured on days E1, after day shift D1, D3 and D6, and after night shift N1, N3 and N6. Saliva will be sampled at home or at work during shift D6 and N6, when also a blood sample will be collected after. Between E1 and E4, a sleep diary is filled out, and an Oura ring measures sleep, heart rate and movement. Bold, shift length 12 hours; D, day shift; E, evening shift; N, night shift; PVT, psychomotor vigilance test.
After preparation and inclusion, data collection is repeated twice for the first half of the shift teams. Group 1 starts with summer data collection; group 2 starts with winter data collection. Dissemination to the workers and company will follow the study.
Exposure variables
Work schedule
The three-shift schedule illustrated in figure 1 is practiced from September to May (figure 1). Monday through Thursday, the day shifts last 7 hours (from 07:00 to 14:00), the evening shifts last 9 hours (from 14:00 to 23:00) and the night shifts last 8 hours (from 23:00 to 07:00). Friday through Sunday, the day and night shifts last 12 hours (07:00 to 19:00 and 19:00 to 07:00, respectively). Each worker is part of one shift team of six. A two-shift/12 hours schedule is practiced in June, July and August (due to summer vacation). To keep the work schedule constant between seasons, we will not track the workers during the two-shift schedule. Objective working hours (daily start and end times) will be assessed from payroll data, to also capture overtime or other changes in the work schedule.
Seasonal variation in light and temperature
To compare seasons, measurements will be made in May (light ‘summer’ period) and November/December (dark ‘winter’ period) (figure 2). The same participants will participate in both seasons. It is estimated that the outdoor workers spend 5–6 hours outdoors each workday, depending on the operational demands. Weather data is recorded at the plant, capturing ambient temperature and daylight fluctuations on the days of data collection.
Data collection protocol
Data collection will take place during three periods: (1) summer 1 (May 2023), (2) winter (November 2023 to December 2023), (3) summer 2 (May 2024) (figure 2). Six-shift teams with 20–25 workers each, will be divided into two groups. This is to mitigate order-effect bias. Hence, each worker will participate in two 25-day data collection periods (table 1).
Data collection protocol. Columns show inclusion day and eight data collection days
After inclusion and signing of informed consent, a baseline questionnaire will be sent to the participants on smartphone. The questionnaire should be filled out before baseline (E1). At E1, an Oura ring is given to the participant to wear until after N6. At E1, the measurement of sleepiness, fatigue and alertness is obtained three times to limit learning effects. Then, at D1, D3, D6, N1, N3 and N6, sleepiness, fatigue, alertness will be measured in a 2-hour time window starting at 12:30 (end of day shift) and at 05:30 (end of night shift). As D3 is a weekend day shift lasting 12 hours, measurements will be performed at the same clock hour as during the regular 7-hour day shifts. A sleepiness measurement is also obtained on paper at the start and end of each shift throughout the whole 25-day period. Blood samples will be taken after the shift on D6 and N6 during the same 2-hour window. The sleep diary will be assessed daily at 18:00 by a smartphone questionnaire (E1–E4). On D6 and N6, six saliva will be sampled at 4-hour intervals (if awake). Skin temperature, heart rate and movement will be measured continuously with the Oura ring (except during working hours for outdoor workers, due to ATEX regulations).28
Dependent variables
Sleepiness, fatigue and alertness
Sleepiness and alertness are the primary dependent variables. Participants will rate their current sleepiness levels using the Karolinska Sleepiness Scale (KSS). The KSS is a widely used scale that assesses sleepiness ranging from 1 ‘very alert’ to 9 ‘very sleepy, fighting sleep, and effort to keep awake’.29 Mental and physical fatigue are measured on two 0–10 numerical rating scales,30 ranging from 0 ‘not at all’ to 10 ‘extremely’, as fatigue is a relevant risk factor for incidents or accidents.31 Alertness is measured by the psychomotor vigilance test (PVT), a reaction time test considered to be the gold standard measure for neurobehavioural effects of sleep loss and circadian misalignment, showing stable performance across repeated administrations.32 In the present study, a 3-min version33 of the test will be used. KSS, fatigue and PVT assessments will be performed using the Joggle Research platform (Pulsar Informatics, USA), designed for iPad. KSS measurements will also be made on paper.
Sleep diary
Sleep quality and quantity is measured by a simplified version of the standard sleep diary for insomnia34 and administered by a smartphone. The questions are: (1) ‘How many hours did you sleep over the last 24 hours?’ (drop-down menu with hours and minutes in 15 min intervals), (2) ‘Was your sleep divided into one or several periods?’ (alternatives ‘yes, one period’ or ‘no, several periods’) and (3) ‘What was the quality of your sleep?’ (rated on a 1–5 Likert scale). In addition to sleep duration, the sleep-wake ratio per 24 hours (amount of sleep/amount of wakefulness) will be calculated.35 The sleep diary will take <1 min to complete.
Tracking of sleep and heart rate
Skin temperature, movement, peripheral oxygen saturation (SpO2), heart rate and a proxy for sleep will be measured by the Oura Gen 3 health tracker ring (ŌURA, Oura Health Oy, Oulu, Finland)36 which is an accelerometer with additional sensors. The ring will only be worn outside work shifts, due to the ATEX directive. Data from this device will be used to determine levels of circadian misalignment by calculating ‘Composite Phase Deviations’.37
Saliva samples
Six saliva samples taken on D6 and on N6 will determine diurnal levels of relevant biomarkers, for example, melatonin.38 Samples will be taken at 4-hour intervals (07:00, 11:00, 15:00, 19:00, 23:00, 03:00) if awake. Participants will be instructed to avoid eating, drinking, smoking or brushing their teeth 30 min before sampling, then 3–5 mL saliva will be obtained by passive drool into tubes (Salimetrics, State College, Pennsylvania, USA).
Blood samples
Blood will be collected from an antecubital vein using 21G BD Vacutainer Safety-Lok Blood Collection Set, BD Vacutainer Plasma Preparation Tubes (BD PPT, 9.0 mg K2EDTA, BD, Franklin Lakes, USA).
A method using machine learning has been used to determine circadian timing from a single blood sample taken at any given time of the day.39 Peripheral monocyte blood cell will be isolated from EDTA stabilised whole blood samples and incubated with CD14 antibody coated magnetic beads for positive selection using magnetic columns. Purity of cells is validated using flow cytometry (CD14, CD15, CD45), another aliquot is spun down and stored at −70°C. Isolated RNA from the cells is analysed with RNA sequencing and NanoString nCounter technology to quantify messenger RNA (mRNA) expression levels, in particular specific panels of clock genes reflecting circadian time. Alternative mRNA quantification methods in our own laboratory are Digital Droplet Polymerase Chain Reaction (ddPCR) and Quantitative Polymerase Chain Reaction (qPCR). Monocyte expression profiles will be used to estimate circadian misalignment at different phases of the shift cycles and seasons. Comparisons will be made with some of the methods mentioned, actigraphy37 and salivary melatonin.40
For blood plasma isolation, tubes will be inverted immediately on a tilt tray and centrifuged for 10 min at 1500×g. Plasma will be immediately pipetted into cryo tubes and aliquots will be stored at −80°C until analysis. Plasma samples will be analysed using Luminex (Luminex Corporation, Austin, USA). Assays will be analysed on a MAGPIX Multiplex Reader using Bio-Plex Manager MP Software, and inflammation marker concentrations were obtained from analysis of raw data using Bio-Plex Data Pro software (Bio-Rad Laboratories, Texas, USA). Magnetic beads are pre-coupled with antibodies for the detection of inflammation markers like C-reactive protein (CRP), fractalkine (CX3CL1), interleukin (IL)-8 (CXCL8), monocyte chemoattractant protein 1 (MCP1) (CCL2) and tumor necrosis factor (TNF). Inflammatory markers can also indicate poor sleep quality.41 42 We will also measure metabolic factors like fibroblast growth factor 21 (FGF-21), growth differentiation factor 15 (GDF-15), and markers of sympathetic activation like norepinephrine to complement thermal exposure data. These plasma markers will be used to indicate cold responses and possibly cold acclimation.
Baseline questionnaire
A smartphone-based questionnaire at baseline (based on validated scales) will assess age, gender, height, weight, shift work seniority, medication, caffeine, tobacco and alcohol use, insomnia,43 Munich Chronotype Questionnaire for shift workers,44 psychosocial factors at work, heavy physical load at work, fatigue, subjective pain issues and subjectively cold or warm at work. The baseline questionnaire is sent to the worker’s phone during the second workday (E2 in figure 1).
Mathematical model
A validated mathematical model of the adenosine system, the Phillips, Klerman, Butler (PKB)-model,26 coupled with a validated computational model of the human circadian system,45 will be used to predict PVT performance on different shift types (day work, night work) (figure 3). Input to the model is work schedule and seasonal variation. We first test the model predictions with default parameters against the collected data. We then attempt to refine the model by refitting the model parameters against the data (ie, training) and potentially by incorporating other explanatory factors such as ambient light and temperature.
Conceptual model. The mathematical model will estimate sleep and psychomotor vigilance test scores based on the work schedule and season.
Exclusion criteria for participation
Workers not understanding written and spoken Norwegian, will be excluded.
Data management and storage
The STAMI is the data controller. Alertness data, data collected by smartphone and activity sensory data will be cloud-stored temporarily by Pulsar Informatics, Walr Group, and by Oura Health Oy. Once each of the three data collection periods are finished, data will be downloaded and stored in STAMI’s safe zone for data storage (limited access and two-factor authentication).
Statistical analysis
For research objectives 1 and 2, the primary dependent behavioural variables are sleepiness (measured by KSS) and alertness (measured by four PVT metrics: mean 1/reaction time (RT), number of lapses, slowest 10% 1/RT, median RT).46 Secondary variables are sleep/wake behaviour (time in bed, sleep duration, sleep/wake ratio—assisted by actigraphy), sleep quality, fatigue, chronotype data and subjective issues. Primary biological variables are salivary melatonin concentration (pg/mL during D6 and N6). Secondary biological variables are circadian time and blood levels of CRP, fractalkine, IL-8, MCP1, TNF after D6 and N6.
To determine whether dependent variables vary between season (summer vs winter), shift type (day vs night), time of day (before vs after work bout) and shift number (D1−D6 or N1−N6) linear mixed models will be used. Separate models investigating different main effects and interaction effects will be tested. Based on Akaike information criterion random slope for season, shift type, time of day or shift number will be added if it improves the model fit. In all analyses, participants will be included as a random effect. Statistical analyses will be performed in Stata (StatCorp, Texas, USA) or R (www.r-project.org).
For research objective 3, it will be determined whether the model can account for PVT data collected in participants undergoing work-related sleep restriction, as collected in work package 1 (WP1). The PBK-model outputs an objective assessment of sleepiness, as well as the predicted number of lapses on the PVT. In the first stage (prediction), model predictions will be made a priori, before data collection. In this stage, the model will be provided with the work shift times as input, as well as lighting assumptions that correspond to the simulated season (summer vs winter). From the model output, we will report the average predicted PVT lapses for each shift type by season. We will additionally report the model’s predicted circadian phase, as this may help to provide an interpretation of the model’s predictions. In the second stage (testing and refinement), we will compare the model’s predicted PVT lapses with the collected data. We will report the model’s goodness of fit including R2 and root mean square error. We will report whether the model’s accuracy can be improved by refitting model parameters. Finally, we will report whether the model predictions are improved by incorporating other explanatory variables including season and ambient temperature as additional predictors, since activation of adenosine receptors within the central nervous system (CNS) has been shown to blunt normal thermoregulatory responses to cold exposure.47 The team is already developing algorithms for improving model accuracy and applicability, which we could potentially use in this project. For more details of the model equations and parameters, see.26 We also have a preliminary agreement on access to real data on safety incidents. It would be highly interesting to set the real incident data, sleepiness ratings and predicted sleepiness on the same timeline.
Loss to follow‐up, compliance and missing data
Based on previous experiences, we expect between 60% and 80% of the total workforce at the plant to volunteer for participation in the project. The project will be anchored with the plant’s leadership and with the managers of the shift teams. Participants will be closely followed by the Occupational Health Service after inclusion, and we expect 10–20% loss to follow-up for the whole 3-year period. Outcome will be censored in case of people leaving the employment at the plants due to unemployment, disability pension or rehabilitate pension. Missing outcome values will be considered by the mixed model approach, and no prior imputation is necessary. For the adjustment variables, we do not expect much missingness.
Study power
Study power has been estimated based on output of the PVT test, more specifically measurements of mean inverse reaction time (1/RT) and lapses, which has been shown to be the most sensitive measurements of total sleep deprivation and partial sleep deprivation, respectively.46 With an effect size of 0.5, it has been estimated that 32 participants are sufficient (power 80%, α=0.05), whereas an effect size of 0.4 requires 52 participants. In the present study, we plan to invite n=120 participants, which should yield between 72 and 96 participants (60–80%) and sufficient power.
Ethics and safety considerations
All participants will give written informed consent to participate in the study, which will be conducted in accordance with the Declaration of Helsinki. The Norwegian Regional Committee for Medical Research Ethics South-East D waivered the need for ethics approval (reference 495816). The Norwegian Agency for Shared Services in Education and Research approved the project (reference 438031). Participation does not imply increased exposure or risk of injuries.
Dissemination
Results will be published in peer-reviewed scientific journals with focus on occupational safety and health, sleep, working hours, etc, with high impact factor. At least five original articles and one doctoral thesis are planned. Direct communications with researchers will take place at national and international congresses. All scientific findings are presented in their original form at the STAMI webpages and through the STAMI open scientific database. All scientific findings are, as well, in their original form linked to the more popular dissemination by abstract or, in some cases, full text access to the original article.
Popularised results will be disseminated through STAMI’s different communication channels in close cooperation with STAMI’s Department of Communication. Direct meetings with the authorities, as well as internationally through the Arctic Council, are of utmost importance.
Once the models are developed, they will be made available. Implementation will depend on their overarching judgement, taking a range of other factors into account. However, knowledge and tools produced in the current project, may be applied directly for designing work schedules to reduce the risk of accidents and enhance the health benefit of the working environment.
Patient and public involvement
The occupational health service and representatives for the workers were involved in the design phase and during pilot testing, and recruitment. Several physical and video meetings were held for this purpose. This affected the final research questions. After pilot testing, an evaluation formula was sent to the participants, in which they were able to give feedback to the researchers on design elements of the study. Dissemination will be in the form of courses and seminars for the company’s management and employees, as well as knowledge of any risk-reducing measures and discussion of how these can best be implemented in the company.
Discussion
Potential impact of the proposed research
A key outcome of the present project is the mathematical model for the Arctic that could predict the risk of accidents and the work environment impact, based on the work schedule characteristics. This project would benefit from advances in modelling that are already being developed. In turn, the outcomes could feed into existing applications and relationships, where the goal is to develop better individualised predictions of sleep, circadian phase and performance in shift workers. Members of the project group have several existing collaborative relationships with industry for the deployment of these types of models. Another key outcome is the variation of sleepiness and alertness across seven consecutive worknights versus workdays, and between seasons. It is an aim that the modelling tool should be made available for all workplaces in the North.
Knowledge from this project will be available to Petroleum Safety Authority Norway and will enable more can use on when supervising petroleum installations and plants. This knowledge is also vital for the employer, in designing sustainable work schedules. It is the aim that new knowledge from the project will gain shift scheduling in the Arctic region at large.
Given that the project location in the Barents region is likely to end up as a virtually ice-free area by the year of 2050, availability of natural resources can spur increased human activities. Thus, a model bringing together the combined exposure of work schedule and the seasonal challenges in the North, will also have an impact for sectors outside petroleum, such as fishing and shipping.
Potential limitations of study design
A limitation is that day shifts are always preceded by night shifts, as the study was obliged to follow the employer’s work schedule. Therefore, order effects may affect the comparison between measurements taken on day shifts and night shifts. Another limitation is that the day and night shifts on Fridays, Saturdays and Sundays were of 12 hours duration, compared with 8 hours the remaining days. This affected PVT and KSS measurements taken on D3, which were not taken at the end of the shift, but at the same clock hour as the end of the 8-hour day shifts. It also affected PVT and KSS measurements taken after at the end of N6, during which workers had been to work for 12 hours instead of 8 hours after N1 and N3. Furthermore, measurements taken in a field study will usually have higher risk of being confounded by unknown or unmeasurable factors, for example, noise from communication radios during PVT testing, forgetting to store saliva measurements taken at home in refrigerator. A final limitation of the study is that no information of diet was obtained, although this may influence the results.
Outcomes and impact
Research-based knowledge on seasonal variation and night work is vital for the employer/employee, in designing sustainable work schedules. The aim of reducing the number of deaths due to occupational injuries is evident. Another key outcome will be a mathematical model that may use work schedule to predict sleepiness as an occupational hazard for safety incidence in the Arctic regions. This outcome can be used to mitigate the health hazards for workers in Arctic environments. A potential future outcome of the mathematical model is to make a smartphone application that can give the petroleum worker insight into the risk of safety incidents at any time point. Human activity in the Arctic will increase. Thus, a model bringing together the combined exposure of work schedule with the seasonal challenges in the North will have effects for many occupations in the Arctic.
Ethics statements
Patient consent for publication
Acknowledgments
We are grateful for crucial project coordination from Kathrine Holm and helpful methodological input from Anne-Mari Gjestvang Moe, Kathrine Holm and Tiril Schjølberg at STAMI. We are also grateful for comments and hints from our STAMI colleagues Suzanne Merkus and Jenny-Anne S Lie, as well as Stein Knardahl, who has reviewed the protocol to ensure that the project is in line with STAMI’s overall aims and strategy.
References
Footnotes
Contributors FH and DM designed the study, wrote the grant applications and are chief investigators of the study. AJKP, KBN, LVM and MS participated in designing the study and will participate in interpretation of data. All authors contributed to refining the protocol. All authors approved the final manuscript and will be accountable for all aspects of the work.
Funding This work was supported by the Research Council of Norway grant number 326291/INNO (program Petromaks2).
Competing interests AJKP has received research funding from Versalux and Delos, and he is the co-founder and co-director of Circadian Health Innovations PTY LTD. The remaining authors declare no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous 3 years; no other relationships or activities that could appear to have influenced the submitted work.
Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.
Provenance and peer review Not commissioned; externally peer reviewed.