Related Experiment Video
Updated: Mar 17, 2026

Collecting Sleep, Circadian, Fatigue, and Performance Data in Complex Operational Environments
Published on: August 8, 2019
Higher risks when working unusual times? A cross-validation of the effects on safety, health, and work-life balance
Jana Greubel1, Anna Arlinghaus2,3, Friedhelm Nachreiner2
1GAWO (Gesellschaft für Arbeits-, Wirtschafts- und Organisationspsychologische Forschung e.V.), Achterdiek 50, 26131, Oldenburg, Germany. jana.greubel@gawo-ev.de.
Purpose:
Replication and cross-validation of results on health and safety risks of work at unusual times.
Methods:
Data from two independent surveys (European Working Conditions Surveys 2005 and 2010; EU 2005: n = 23,934 and EU 2010: n = 35,187) were used to examine the relative risks of working at unusual times (evenings, Saturdays, and Sundays) on work-life balance, work-related health complaints, and occupational accidents using logistic regression while controlling for potential confounders such as demographics, work load, and shift work.
Results:
For the EU 2005 survey, evening work was significantly associated with an increased risk of poor work-life balance (OR 1.69) and work-related health complaints (OR 1.14), Saturday work with poor work-life balance (OR 1.49) and occupational accidents (OR 1.34), and Sunday work with poor work-life balance (OR 1.15) and work-related health complaints (OR 1.17). For EU 2010, evening work was associated with poor work-life balance (OR 1.51) and work-related health complaints (OR 1.12), Saturday work with poor work-life balance (OR 1.60) and occupational accidents (OR 1.19) but a decrease in risk for work-related health complaints (OR 0.86) and Sunday work with work-related health complaints (OR 1.13). Risk estimates in both samples yielded largely similar results with comparable ORs and overlapping confidence intervals.
Conclusions:
Work at unusual times constitutes a considerable risk to social participation and health and showed structurally consistent effects over time and across samples.

