Robert M. Sapolsky

The pantheon of science includes individuals who have made enormous contributions to human health -- the likes of Pasteur and Salk. A pedestal in that temple awaits the scientist who solves the following mystery: Why do we eat junk food when we feel unloved?

This isn't a silly question, certainly not during September, which happens to be National Childhood Obesity Awareness Month. There's an epidemic of obesity-related health problems, with adult-onset diabetes leading the way throughout the world. The fact that we eat when we're not actually hungry contributes a lot to this problem.

So why do we do it? It can be because everyone around us is eating. Or because food ads can be so persuasive. Or because we want to bankrupt a hated party host by eating all his Cheetos.

One of the best-understood examples of non-nutritive eating is the fact that stress tends to make us eat more. It makes sense psychologically, in that the people most prone to stress eating are those most actively restricting food intake the rest of the time: When the going gets tough and they need to be nice to themselves, this is how they ease up. They prefer to eat fats and carbs. If the boss is a creep, why not run wild on the chocolate-covered walrus blubber?

But we can't trace these habits merely to the complexities of the human psyche, because it's not just humans who exhibit them. Stress a lab rat by, let's say, putting an unknown rat in its cage, and it will eat more and show a stronger preference for high-fat/high-carb options than usual.

This phenomenon's occurrence in many species makes evolutionary sense. For 99% of animals, stress involves a major burst of energy use as they, say, run for their lives. Afterward, the body stimulates appetite, especially for high-density calories, to rebuild depleted energy stores. But we smart, neurotic humans keep turning the stress-response on for purelypsychological reasons, putting our bodies repeatedly into the restocking mode.

Scientists are beginning to understand how this stress-related junk-food craving works. Stress increases the release of 'endogenous opioids' in some brain regions. These neurotransmitters resemble opiates in their structure and addictive properties (and opiates work by stimulating the receptors that evolved for responding to the brain's opioids). This helps to account for the hugely reinforcing properties of junk food at such times.

Stress also activates the 'endocannabinoid' system in the brain. Yes, there's a class of chemicals in the brain that resemble the ingredient in cannabis that famously links pot to getting the munchies. And stress activates another brain chemical called neuropeptide Y that can stimulate the craving for fat and sugar.

The most fundamentalmechanism to explain this stress effect is that comfort food is, well, comforting. As first demonstrated by Mary Dallman and colleagues at the University of California, San Francisco, working with lab rats, fat and carbs stimulatereward systems in the brain, thereby turning off the body's hormonal stress-response.

It may seem unlikely that one type of pleasure works to offset the effects of a very different source of displeasure. Why should fat-laced rat chow lessen angst about a new cage mate? Yet we regularly make much bigger leaps. Burdened with unrequited love? Shopping often helps. Roiled with existential despair? Bach might do the trick. The common currency of reward in the brain makes for all sorts of unlikely ports in a storm.

But despite the varied possibilities of sources of comfort, some exert particularly strong primal pulls -- to the detriment of our health. It is a sign of our evolutionary legacy that, at the end of a stressful day, far fewer of us will seek solace in the poetry of Robert Frost than in a pint of double fudge brownie ice cream.