The researchers said even veteran night-shift workers who drove after working all night exhibited reactions similar to those seen in drivers with elevated blood alcohol.
The study, led by Brigham and Women's Hospital (BWH) of Boston, MA, is published in the Proceedings of the National Academy of Sciences.
The researchers compared the daytime driving performance of 16 night-shift workers after a night of work with their performance after a night of sleep.
The results showed that 37.5% of the drivers had a near-crash event when they drove after a night of work, compared with none of the same drivers having a near-crash event when they drove after a normal night of sleep.
Most drivers admit they have driven a motor vehicle while drowsy. Within the past year, 28% of American drivers have reported falling asleep at the wheel, say the researchers.
Corresponding author Charles A. Czeisler, a professor of sleep medicine at Harvard Medical School and chief of BWH's Division of Sleep and Circadian Disorders, says drowsy driving is a major and preventable public health hazard. He notes:
"These findings help to explain why night-shift workers have so many more motor vehicle crashes than day workers, particularly during the commute home."
The study highlights the risks of driving after working all night, particularly for the 9.5 million Americans (15% of the US workforce) who work overnight or rotating shifts.
Participants drove real vehicles, not simulator
Previous studies have assessed the effect of night-shift work on driving using driving simulators.
But Prof. Czeisler and colleagues believe their results - obtained with subjects driving real vehicles on a test circuit and not in a driving simulator - are the first to show a link between drowsy driving and higher risk of motor vehicle crashes.
They are also the first to show an increase in self-reported and biological measures of drowsiness when driving a real motor vehicle during the day after working a night shift.
For their study, the team, including researchers from the Liberty Mutual Research Institute for Safety in Hopkinton, MA, assessed 16 night-shift workers as they completed two driving sessions, lasting 2 hours each, at the Institute's closed driving track.
Before one of the sessions, the workers slept a normal night - averaging 7.6 hours - with no night-shift work. Before the other session, the same workers worked a night shift.
The two sessions took place at around the same time of day for each participant, and all participants were accompanied by a driving safety observer who had secondary controls.
The researchers collected measures of drowsiness and driving performance during the driving sessions. The drowsiness measures included electroencephalograms (EEGs) to assess episodes of micro-sleeps and partial eyelid closures with slow eye movements - which indicate transition from wakefulness to sleep.
The researchers assessed each participant's driving performance by counting near-crash events, drives that had to stop because of failure to control the vehicle, and how often the vehicle weaved in and out of the driving lane.
For the purpose of the study, the team defined near-crash events as those times when the safety observer had to brake because the vehicle was leaving the roadway and the driver had not taken corrective action.
'Reactions similar to that of elevated blood alcohol'
When they compared the results of the post-sleep and the post-night-shift driving sessions, the researchers found that during the post-night-shift session, the participants showed increased drowsiness, gradually deteriorating performance and a higher risk of near-crashes.
When driving after a night of sleep, there were no near crashes, but after a night of work, there were 11 near-crashes in 6 of the 16 drives (37.5%), and all of them occurred after at least 45 minutes of driving.
The authors also note that: "7 of 16 post-night-shift drives (43.8%) were terminated early for safety reasons, compared with zero near-crashes or early drive terminations during 16 post-sleep drives."
Plus, the results showed that sleep-related impairment was evident within the first 15 minutes of driving following a night shift. Also, participants had longer blink durations and a higher number of slow eye movements in the driving sessions that followed a night of work, compared with a night of sleep.
The risk of micro-sleep periods - where the EEG pattern of electrical brain activity showed the driver was asleep for up to 3 seconds - also increased after driving for more than 30 minutes following a night of work.
Lead author Dr. Michael L. Lee, a research fellow in the Division of Sleep and Circadian Disorders at BWH, notes that a short commute for these drivers following a night of work appears to be potentially dangerous and the longer the drive, the higher the risk. He says:
"Even veteran night shift workers were vulnerable to the risks associated with drowsy driving, and exhibited reactions similar to behaviors observed in drivers with elevated blood alcohol concentrations."
Dr. Lee suggests educating shift workers about the hazards of drowsy driving could reduce these risks - perhaps by prompting them to either eliminate or reduce the need to drive after working nights, or to stop driving when their performance drops because of drowsiness.
He and his colleagues are aware that the limitations of their study are very different to a real-world commute - not least the fact the tests took place on a closed track, with researchers inside the vehicle, and included frequent interruptions to obtain measurements.
Nevertheless, project co-investigator Dr. William J. Horrey, a principal research scientist at the Liberty Institute, concludes:
"The number of near-crash events that occurred during the study starkly emphasized the statistics of nearly half a million crashes and 6,500 fatalities annually that directly result from driver fatigue. While we are all generally aware of the risks associated with drowsy driving, these outcomes really underscored just how dangerous a homeward commute can be for this working population."
In January 2014, Medical News Today learned of a detailed study that exposed real-life dangers of driver distraction, particularly for newly qualified teenagers, who the researchers found were the drivers most likely to be involved in an accident or near-miss incident because of doing distracting things like texting at the wheel.