Clinical Question Are pharmacological interventions associated with better-quality sleep and alertness in shift workers?
Bottom Line Low-quality evidence shows that melatonin is associated with 24 minutes longer daytime sleep after the shift but not with faster falling asleep compared with placebo. There is no association between hypnotics, such as zopiclone, and sleep outcomes, alertness, or harms. The alertness-promoting medications armodafinil and modafinil are associated with improved alertness during shift work but are also associated with headache and nausea.
In the United States, 29% of workers do not work regular daytime shifts.1 Shift work is associated with reduced sleep duration, impaired daytime sleep quality, and reduced alertness during night shifts.2,3 Workers frequently use pharmacological products to ameliorate the adverse effects of shift work.4,5 This JAMA Clinical Evidence Synopsis summarizes data from a Cochrane review6 of pharmacological interventions for sleepiness and sleep disturbances caused by shift work.
Juha Liira, MD, PhD1; Jos Verbeek, MD, PhD1; Jani Ruotsalainen, MSc1
No. of randomized clinical trials: 15
Study years: Published, 1982-2008
Median follow-up: 7 nights of shift work or days after night shifts
No. of participants: 1240
Men: 664 Women: 576
Age, median (range): 33.5 years (25-42)
Race/ethnicity: Not reported
Settings: Workplace: health care, police, oil rig, various occupations
Countries: France, Germany, Iran, Korea, Norway, United States, United Kingdom
Comparisons: All trials compared pharmacological interventions against placebo; evaluations included melatonin (9 trials), hypnotics for improving sleep problems (2 trials), modafinil (1 trial), armodafinil (2 trials), and caffeine plus naps to decrease sleepiness or to increase alertness (1 trial)
Primary outcomes: Sleep duration, time to fall asleep, and alertness or sleepiness measured with a validated scale
Secondary outcomes: Adverse outcomes, economic outcomes, resource use, and injuries
Melatonin (1 mg-10 mg) after the night shift is associated with increased daytime sleep duration (mean difference [MD], 24 minutes [95% CI, 9.8 to 38.9]; 7 trials, 263 participants; low-quality evidence) and increased nighttime sleep duration (MD, 17 minutes [95% CI, 3.7 to 30.2]; 3 trials, 234 participants; low-quality evidence) compared with placebo (Figure). There was no dose-response association. There was no association between melatonin and sleep latency time (ie, the time required to fall asleep) compared with placebo (MD, 0.4 minutes [95% CI, −1.6 to 2.3]; 5 trials, 74 participants; low-quality evidence). Exclusion of low-quality studies resulted in lack of an association with daytime sleep duration.
Figure. The Association of Melatonin With Self-Reported and Objectively Measured Sleep Duration During Days After the Night Shift Source: Data have been adapted with permission from Wiley.6 Two studies only reported night sleep after the night shift period and could not be included. The size of the data markers indicate the weight of the study.
The hypnotic medication zopiclone (7.5 mg) was not associated with longer daytime sleep compared with placebo (MD, 44.0 minutes [95% CI, −22.7 to 110.7]; 1 trial, 28 participants; low-quality evidence).
Armodafinil (150 mg) before the night shift was associated with less sleepiness by 1 point on the Karolinska Sleepiness Scale (scale from 1 to 9) (MD, −0.99 [95% CI, −1.32 to −0.67]; 2 trials, 572 participants; moderate-quality evidence) and an increase in alertness in a simple reaction time test (MD, −50.0 milliseconds [95% CI, −85.5 to −15.5]) at 3 months’ follow-up in patients with shift work sleep disorder and compared with placebo. Modafinil was associated with a reduction in sleepiness on the Karolinska Sleepiness Scale (MD, −0.90 [95% CI, −1.45 to −0.35]; 1 trial, 183 participants; moderate-quality evidence) and an increase in alertness measured with the psychomotor vigilance test among patients with shift work sleep disorder. Modafinil (200 mg) was also associated with headache (34% incidence for modafinil vs 23% incidence for placebo) and nausea (11% incidence for modafinil vs 3% incidence for placebo). Postmarketing surveys reported Stevens-Johnson syndrome in association with modafinil. There were no trials in workers without shift work sleep disorder.
Based on 1 trial, caffeine (300 mg) plus napping before the night shift were associated with less sleepiness on the Karolinska Sleepiness Scale (MD, −0.6 [95% CI, −1.1 to −0.2]; 1 trial, 39 participants; low-quality evidence).
There is low-quality evidence that melatonin is associated with increased sleep length after a night shift but not with improved sleep latency time. Both modafinil and armodafinil were associated with increased alertness and reduced sleepiness in employees who experience shift work sleep disorder, but those medications were also associated with headache and nausea. Low-quality evidence shows that caffeine plus naps are associated with reduced sleepiness during the night shift. Hypnotics are not associated with sleep quality after a night shift or with adverse effects.
An update of the search of the original review up to December 1, 2014, yielded 155 references of which 6 were fully assessed. One trial of the herbal treatment Rhodiola rosea fulfilled our inclusion criteria but was not associated with improved sleep outcomes.7 This additional study would not have changed our conclusions.
Most trials on melatonin were small with a duration of 1 week or less. Few data were available for modafinil and armodafinil and follow-up for these drugs was limited to 3 months.
Comparison of Findings With Current Practice Guidelines
The American Academy of Sleep Medicine recommends melatonin, hypnotic medications, and modafinil for shift workers but is not definitive about the clinical use of caffeine. The American Academy of Sleep Medicine states that benefits of modafinil outweigh risks.3 We found that modafinil was associated with a small benefit but with frequent adverse outcomes. The European Medicines Agency has withdrawn the license for modafinil for shift work sleep disorder because associations of modafinil with improved sleep outcomes were small and because postmarketing surveillance revealed Stevens-Johnsons syndrome.
Areas in Need of Future Study
High-quality trials are needed with a longer follow-up of the benefits and adverse effects and costs of all pharmacological agents that induce sleep or promote alertness in shift workers with and without a diagnosis of shift work sleep disorder.