The very thing that helps make well-timed bright light therapeutic may make ill-timed exposure unhealthy: melatonin. Not enough melatonin, that is.

In response to bright light, the body's "third eye" - the pineal gland - shuts off its production of the time-sensitive hormone melatonin, which it otherwise pumps out freely in the dark of night.

Dark of night. But how many of us experience really dark nights in this age of Edison, except when we're asleep?

A recent review by writer Janet Raloff in Science News suggests that an unimpeded evening rise in melatonin may be as beneficial to our health as a daytime drop. And it's not just because melatonin - a popular over-the-counter sleeping pill in the US ( in Canada, the Health Protection Branch has banned it) - mellows us out for bedtime. Scientists are finding it has other vital functions too. Most notably, melatonin is an antioxidant (like vitamins E and C), it's an anticarcinogen, and it temporarily dampens the production and release of estrogen. Given that too much estrogen can promote certain cancers - especially breast cancer - and that antioxidants help prevent cancer, it's no surprise that nearly a decade of studies implicate the light/melatonin connection in the etiology (causation) of breast and other cancers.

Most provocative have been studies comparing the health of blind with sighted people. Consistently, profoundly blind female subjects have proven much less vulnerable to breast cancer than sighted ones - some 30 to 50 percent less. In one study (though not in another), the former were also less prone to ovarian and stomach cancer.

Among profoundly blind men, researchers in Stockholm have found a lower incidence of prostate, stomach, colon, rectal, lung, and skin cancer. Interestingly, blind people who still see light have either had a normal or only a somewhat reduced incidence of these cancers, just as the melatonin/cancer hypothesis, advanced by an American Energy Department scientist, would predict.

Raloff notes that test tube studies "indicate that a reduction of melatonin can alter the production of other hormones, may suppress the immune system's ability to recognize and respond to newly emerging cancers, and appears to spur the growth of at least some tumor tissues." This "growing body of data on melatonin, light, and cancer suggests that certain populations, such as shift workers or others who regularly work in bright light at night, could face unusual risks."

Should we then start living by candle light or oil lamp after sunset? The idea may have romantic appeal, but we may not have to go that far. Green and, to a lesser extent, blue light are much, much more suppressive of melatonin than warm-hued light. While this adds to the case for warm candle light at night, it also suggests that dim incandescent light (also warm hued) would be less disruptive than cool fluorescent. But it may have to be really dim: one study found that even average indoor light levels prevented people's melatonin levels from rising at night.

There may be another alternative. Just as a weak dose of green light can suppress melatonin therapeutically, as with Murray Waldman's Sunnex units, a complementary light source devoid of green and blue wavelengths might give us enough light at night to function normally without blocking the desirable evening rise in melatonin. Murray Waldman, are you listening?

There is another alternative. Leave the lights on at night and take a low dose of slow-release melatonin. Health Protection Branch, are you listening?