Previous research has established that the degree to which nerve signals stimulate the circadian system determines how much melatonin is reduced. The model accounts for the sensitivity to the eye to different parts of the visible-light spectrum, and the level of light that triggers a signal to various parts of the brain. A brief flash of lightning does not trigger a signal to the circadian system, but a longer exposure to light would.
Applying this model to a hypothetical person under a white LED streetlight for one hour, Rea and colleagues calculate that one hour of exposure can reduce melatonin levels by about 3 to 8 percent.
"A 3 to 8 percent effect on modeling is small but maybe it does matter. I think the science just isn't there yet," he said.
Similar model-based calculations by his group show that light from computer screens and electronic devices such as tablets can reduce melatonin by 7 to 20 percent.
For perspective, Rea offered another example not involving white LED light. He fitted a colleague with a light measurement device when she attended a live hockey game. According to their model's calculations, this suppressed 20 to 25 percent of her melatonin production.
"But it's not like suppressing to zero," he said, "like you're getting rid of all your melatonin and you go out and look out the window and the streetlight and all of a sudden [you] don't have any melatonin for the rest of the night. That's just not true. So you really have to know the numbers."
Rea believes that looking at light at night alone won't answer fundamental questions about how the environment can affect the biological clock.
"You find you can't really talk about light at night without also knowing what you've got during the daytime. Taking into account the full 24 hour rhythm is essential," said Rea. "It's just too shallow to talk about melatonin suppression by 3 to 8 percent and draw any conclusions about what is going to be healthy or not healthy."
Haim calls for more detailed epidemiological studies that explore relationships between nighttime light and health problems, by following large populations over long periods of time.
Rea suggests moving beyond epidemiology. He suggests making detailed measurements of people's actual 24-hour light exposure and then designing experiments that create similar light-dark patterns for animals. Then, he said, you could test the hypothesis of whether the light-dark pattern causes health effects.
And there are technological solutions. Wendy Davis, a vision scientist at the National Institute of Standards and Technology, said that it is possible with LED technology to create "tunable" light, so that it would produce blue-rich light during the day and blue-poor at night. But whether this will be necessary remains to be seen.
"There is not enough research in circadian disruption to have a position other than we support good, intelligent, properly executed research, and when it's done, we'll review it and see if we need to change anything or what needs to be done," said Alex Boesenberg, manager of regulatory affairs at the National Electrical Manufacturers Association.
Haim said that decisions on indoor and outdoor lighting "must follow chronobiological ideas," and not ignore the biology of humans, wildlife, and other animals. He calls for more studies of animal models that would look for biological effects of light.
Absent definite answers at this point, Rea advised that people keep a fairly consistent 24-hour schedule when possible, which may be the safest way to keep the circadian system in a regular rhythm and not contribute to any possible adverse health effects.