When does drowsy turn dangerous?

Judging and measuring sleepiness is tricky business. It’s totally subjective and personal — you may feel sleepy and perform poorly with the same hours of shut-eye that leave someone else completely refreshed.

So, how little sleep is too little when you’re behind the wheel of a car? An 18-wheeler? A military jet? There are no standards, though people have been convicted of reckless driving for car accidents they caused after pulling an all-nighter.

About 150 academic and industry researchers spent two days last month holed up in a conference room off Longwood Avenue, drinking lots of coffee and talking about how to measure the impact of sleep deprivation. They wanted to begin to think about ways to treat excessive sleepiness, and to identify people whose performance is about to nose-dive because of a lack of sleep.

Keynote speaker Dr. Charles Czeisler, chief of the Division of Sleep Medicine at Brigham and Women’s Hospital, said he became interested in the subject after a reportedly sleep-deprived driver killed US Army Major Robert Raneri of Nashua, days before his wedding in 2002. Czeisler said State Senator Richard T. Moore suggested a test like a breathalyzer was needed, to measure when it’s legally unsafe to drive.

One in five motor vehicle accidents is related to “drowsy’’ driving, Czeisler said. The problem is that, like drunk drivers, most of us don’t recognize when we’re too drowsy to drive well.

“The average American is sleepy all the time and doesn’t know it because that’s their baseline,’’ said Robert Stickgold, an associate professor of psychiatry at Harvard Medical School and director of the Center for Sleep and Cognition at Beth Israel Deaconess Medical Center. “We know how much food a person needs. We don’t know how much sleep a person needs.’’

Measuring the consequences of sleep deprivation is important for courts that want to assign blame after an accident, for transportation companies that need to make sure their drivers are safe to be on the road, for hospitals that require residents to be alert, and for the military, which wants to maximize performance and minimize risk to lives and equipment.

And, of course, there are companies that would love to develop devices to measure dangerous sleepiness and remedies to help people compensate for staying up to watch late-night ballgames on TV.

(Cephalon Inc., which makes the anti-sleepiness drug Modafinil, and Philips Healthcare, which makes sleep-monitoring and therapy devices for conditions such as sleep apnea, co-sponsored the conference along with the National Heart, Lung and Blood Institute. Several other companies, including drugmakers Eli Lilly and Johnson & Johnson, had representatives on the participant list.)

There’s no question that sleep deprivation can impair performance as much as or more than alcohol, the speakers agreed. After about 18 hours of awake time, it becomes harder to sustain performance levels. Performance also slips when you accumulate a sleep deficit over several short nights. Sleep 5 to 6 hours a night for a week and you have the same level of impairment as if you had pulled an all-nighter, Czeisler said. Circadian rhythms also make it harder for us to perform optimally at certain times of day.

Sleep loss leads to deactivation of parts of the brain, Thomas Balkin, a sleep researcher at the Walter Reed Army Institute of Research, told the crowd. The prefrontal cortex, which is responsible for higher-level thought, and the thalamus, which processes and relays information from the senses, are particularly affected, he said.

Sleepiness is personal in that there is significant variation among people in terms of how much sleep they need, how much their performance is impaired by less-than-optimal amounts, and how quickly they bounce back from a bad night’s sleep. (It generally takes longer, the speakers said, to recover from a sleep deficit of several nights than one sleepless night.)

Can’t we just down a latte or pop a pill to maintain our performance? No, said Czeisler, one of the few scientists in the room not nursing a cup of coffee. Caffeine can make us feel less sleepy, but lack of sleep still takes its toll on ability to function.

“You can have stayed up all night, but if you’ve drunk enough Starbucks coffee you may not be sleepy, even though you are deficient,’’ Czeisler told the researchers. “Caffeine is not the solution. It simply gums up the thermostat that allows the brain to sense how much [sleep] it needs.’’

Drugs also fall short of compensating for sleep deprivation, he said. In a recent study, patients who were sleep deprived felt less sleepy when they took the drug Modafinil, but physical measures of the adverse health effects of sleep deficiency, such as the body’s ability to process sugar, were the same for the subjects who took the drug and those who took a placebo. “The drug is not changing the adverse physiological affects associated with sleep deficiency,’’ Czeisler said.

Likewise, older adults generally sleep less than younger and middle-aged ones, but don’t necessarily need less sleep, he said.

Over the course of the two-day conference, a handful of speakers suggested different methods for measuring alertness — some way to predict when a person’s alertness level is about to nose-dive — but then shot down their own ideas:

Feelings of sleepiness? Too subjective, and not necessarily associated with performance, Balkin said.

Brainwaves? They don’t correlate well to sleepiness, according to James Krueger, a professor of neuroscience and sleep researcher at Washington State University.

Measuring certain gene activity? Maybe, but researchers haven’t found the right genes yet.

Signs of an immune response? Perhaps, but a lot of other things could trigger an immune response, too.

Infrared sensors, or proteins in blood, urine, saliva, or exhaled air? All cool ideas, but pretty far from reality.

Giving someone a simple test like standing on a balance plate designed to measure steadiness? That one has some promise, said Dr. Stuart F. Quan, the conference’s organizer and a professor of medicine at Harvard Medical School and Brigham and Women’s division of sleep medicine. But “there probably isn’t going to be a single measure,’’ he added.

Although the conference did not reach any conclusions, Quan said it laid the foundation for another gathering, which participants said they want to hold in about a year. It’s too early, he said, to set a timetable for solving the problem of how to measure and treat the effects of sleep deprivation.

So, you will still just have to judge for yourself whether today’s a good day to take the train instead of driving to work.

Karen Weintraub can be reached at karen@karenweintraub.com.