How much do insomnia estimates vary across studies?
A systematic review of five studies covering over 10,000 participants found that insomnia prevalence among shift workers ranged from 12.8% to 76.4% [1]. That is a six-fold difference—far too wide to be explained by random sampling error alone. The review noted that this range is higher than estimates for the general population, but it did not clearly separate the contributions of schedule type or diagnostic method [1].
A large multi-center study of 33,669 workers in Korea reported an overall insomnia prevalence of 38.7% [2]. This figure sits in the middle of the range from the systematic review, suggesting that the specific mix of schedules and the diagnostic tool used in that study produced a moderate estimate. The wide spread across studies means that any single prevalence number should be interpreted with caution unless the schedule category and diagnostic criteria are specified.
Do consecutive night shifts change the odds of insomnia?
Yes—but not in a simple linear way. In the multi-center study, compared with workers who did not work consecutive night shifts, the odds of insomnia were highest for those working three consecutive nights (odds ratio 2.65), followed by two nights (1.81), five nights (1.78), and four nights (1.68) [2]. An odds ratio of 2.65 means that workers on three consecutive nights had about 2.6 times the odds of insomnia compared with the reference group.
The pattern is not monotonic: three nights carried the highest odds, while four and five nights were lower than three but still elevated. The study authors suggest this could inform limits on consecutive night shift work, but the cross-sectional design cannot establish causation [2]. The key takeaway is that schedule category—specifically the number of consecutive night shifts—is associated with meaningful differences in insomnia odds, even if the exact shape of the relationship needs more research.
Are irregular shifts worse than regular shifts?
Irregular shift workers—those whose schedules change frequently without a fixed pattern—reported significantly worse sleep outcomes than both non-shift workers and regular shift workers on multiple measures, including time in bed, total sleep time, sleep efficiency, sleep quality, sleep latency, and number of awakenings [4]. However, their values still fell within the average range and did not indicate clinical insomnia [4]. This is an important distinction: worse sleep symptoms do not necessarily translate into a clinical diagnosis.
Regular shift workers, by contrast, reported the best sleep quality and longest sleep duration among shift workers, and the least nocturnal awakenings [4]. The study authors suggest this could be partly due to personality traits—higher conscientiousness and lower neuroticism in that group—but the cross-sectional design and self-report measures limit causal inference [4]. For a reader trying to understand prevalence, the implication is that lumping all shift workers together obscures real differences: irregular schedules appear to produce more sleep disruption than regular ones, even if clinical insomnia rates are not clearly higher.
How much does the diagnostic instrument matter?
The diagnostic tool matters a great deal. A study using standardized clinical criteria (ICSD-3) for shift work disorder reported an insomnia prevalence of 19.9% among shift workers in China [5]. That is far below the 38.7% overall prevalence in the Korean multi-center study [2] and near the bottom of the 12.8%–76.4% range from the systematic review [1]. The Chinese study used a clinical diagnosis of shift work disorder, which requires specific criteria beyond just insomnia symptoms, whereas other studies may have used symptom questionnaires or broader definitions.
The systematic review itself did not analyze how prevalence varied by diagnostic instrument, but its wide range is consistent with the possibility that different tools capture different things [1]. A symptom-based questionnaire might count anyone who reports poor sleep, while a clinical diagnosis requires duration, severity, and functional impairment. The result is that prevalence estimates are not directly comparable across studies unless the instrument is the same. For a general reader, the practical answer is: when you see a prevalence number for insomnia in shift workers, check what diagnostic criteria were used—it can easily change the estimate by a factor of two or more.
Sources used in this answer
Prevalence of insomnia in shift workers: a systematic review.
A systematic review of five studies found insomnia prevalence in shift workers ranged from 12.8% to 76.4%, higher than general population estimates, with higher prevalence among women and singles but no clear variation by age or profession; the relationship with shift schedule remained controversial.
The Association Between the Number of Consecutive Night Shifts and Insomnia Among Shift Workers: A Multi-Center Study
In a multi-center study of 33,669 workers, overall insomnia prevalence was 38.7%, and compared with no consecutive night shifts, the odds of insomnia were highest for three consecutive nights (OR 2.65), followed by two nights (1.81), five nights (1.78), and four nights (1.68).
Shift work sleep disorder: prevalence and consequences beyond that of symptomatic day workers
This study examined shift work sleep disorder and insomnia prevalence using standardized criteria in a diverse working population, but did not explicitly analyze variations across schedule categories or diagnostic instruments.
Personality Traits and Insomnia Symptoms in Shift Workers
Irregular shift workers had significantly worse sleep outcomes than non-shift and regular shift workers on multiple measures, but their values remained in the average range and did not indicate clinical insomnia; regular shift workers reported the best sleep quality and longest sleep duration.
Shift work disorder and related influential factors among shift workers in China
Using ICSD-3 criteria for shift work disorder, the prevalence of shift work disorder among shift workers in China was 19.9%, with associations to various scales and related factors.