TL;DR
This study leverages wearable sensors to objectively measure physical activity and sleep patterns in day and night shift nurses over 10 weeks. It reveals that night shift nurses are more sedentary, experience poorer sleep quality, and report lower life satisfaction, providing a data-driven understanding of the physiological and behavioral impacts of shift work.
What Problem Does This Paper Address?
The increasing demand for 24/7 services across various industries, including healthcare, has led to a significant portion of the workforce operating on non-daytime shifts. This shift work is associated with numerous negative health outcomes, including circadian misalignment, burnout, decreased work performance, and increased risk of chronic illnesses. While self-reported data has provided some insights, there's a need for objective, long-term measurements of physical activity, sleep, and circadian rhythm disruptions in real-world settings to better understand and mitigate these effects. This paper aims to bridge this gap by using wearable sensors to objectively quantify these factors in nurses, a profession known for its demanding shift schedules.
Methods
The study employed a multimodal approach, combining objective data from commercial wearable sensors (Fitbit Charge 2) with subjective self-reported data.
- Wearable Sensors: Participants wore Fitbit Charge 2 devices continuously for 10 weeks to collect data on heart rate, step count, and sleep patterns (sleep onset, wake-up time, duration, and efficiency).
- Physical Activity Analysis: Heart rate data was used to categorize activity intensity (rest, moderate, vigorous). A "walk activity ratio" was calculated based on step counts. These metrics were analyzed across different time periods of the day and night, both on workdays and off-days.
- Sleep Pattern Analysis: Processed sleep data from the Fitbit was aggregated. A key metric introduced was "social jet lag" (AMS), calculated as the difference between mid-sleep time on workdays and off-days, to quantify circadian misalignment.
- Self-Reported Data: Participants completed baseline assessments for demographics, sleep quality (Pittsburgh Sleep Quality Index - PSQI), affect (Positive and Negative Affect Schedule - PANAS), anxiety (State-Trait Anxiety Inventory - STAI), life satisfaction (Satisfaction with Life Scale - SWLS), and personality (Big Five Inventory-2 - BFI-2). Ecological Momentary Assessments (EMAs) were also used daily to capture real-time affect, anxiety, and stress levels.
- Statistical Analysis: Independent t-tests and three-way ANOVAs were used to compare demographic and behavioral variables between day and night shift nurses, controlling for age and gender. Linear mixed models were employed to analyze EMA data. Linear regression models were used to assess the relationship between sensor-derived physical activity and self-report variables.
Datasets
The study utilized data collected from 113 nurses at a large hospital setting over a 10-week period. This dataset, referred to as the "TILES: Tracking Individual Performance with Sensors" study, is a newly introduced benchmark for longitudinal, multimodal sensing in healthcare professionals.
Key Insights and Ideas
- Objective vs. Subjective: The study highlights the power of wearable sensors in providing objective, granular data that complements and sometimes challenges self-reported measures of well-being.
- Quantifying Sedentariness: The objective measurement of rest-activity and walk-activity ratios clearly demonstrates that night shift nurses are more sedentary, even outside of their work hours.
- Social Jet Lag as a Marker: The quantification of social jet lag (AMS) provides a concrete measure of circadian misalignment, showing significantly larger differences in sleep timing for night shift nurses on off-days.
- Holistic Impact: The findings suggest that shift work's impact extends beyond work hours, affecting sleep patterns and overall life satisfaction even during rest periods.
Key Results
- Life Satisfaction & Sleep Quality: Night shift nurses reported significantly lower life satisfaction (SWLS) and poorer sleep quality (higher PSQI scores) compared to day shift nurses.
- Physical Activity Differences:
- Night shift nurses exhibited a higher rest-activity ratio and a lower walk-activity ratio on both workdays and off-days.
- Night shift nurses spent less time engaged in vigorous activity on off-days.
- Diurnal analysis showed night shift nurses had higher rest-activity ratios during typical daytime hours (7-19h) and lower during nighttime hours (23-7h) on workdays, suggesting circadian misalignment.
- Sleep Patterns:
- Night shift nurses slept significantly less on workdays (approximately 80 minutes less) than day shift nurses.
- Night shift nurses exhibited significantly larger social jet lag (AMS), with a mean difference of 425 minutes between workday and off-day sleep patterns, compared to 52.5 minutes for day shift nurses.
- While night shift nurses had more irregular sleep onset and wake-up times on off-days, they showed more regular sleep onset on workdays.
- Affect and Anxiety: Primary shift pattern was not significantly associated with anxiety, affect, or stress levels, though night shift nurses reported slightly higher negative affect on workdays.
Limitations and Future Work
- Shift Pattern Assessment: The study captured the primary shift schedule but did not directly assess the frequency or rotation of shifts, which could influence results. Future research could investigate the impact of rotating shifts.
- Work Environment Nuances: The study did not account for specific work environments within the hospital (e.g., ICU vs. non-ICU), which could differentially affect nurses. Future work could explore these subgroups.
- Sample Size: While 113 nurses is a substantial sample, a larger cohort could further strengthen the findings.
- Inferred Workdays: Workday status was inferred rather than directly obtained, which could introduce minor inaccuracies.
- Compliance Decline: The daily burden of compliance with sensors and surveys led to a gradual decline, potentially impacting data quality. Future studies could explore methods to reduce this burden.
- Causality: The study identifies associations, not causal relationships. Future research could explore interventions to mitigate negative effects.
- Future Directions: The authors suggest exploring how wearable measures can capture unique patterns across different nursing subgroups and investigating the impact of specific shift patterns.
Summary updated