Is there a way to tell from a smoker's face that he or she is craving a cigarette? Psychologist Michael Sayette, PhD, thinks so. But, like so much with facial expression, it isn't a simple proposition. There may, in fact, be several facial expressions that smokers display, and there may be times when some people express nothing at all.

Sayette, associate professor of psychology at the University of Pittsburgh, began examining the link between facial expression and cigarette craving several years ago in the hope of identifying an objective way to measure craving since existing measures have been almost exclusively based on notoriously flawed self-reports.

He chose facial expression for its potential to help psychologists understand the role emotions play in craving. Many researchers contend that there is a direct link between facial expressions and people's feelings (see article, page 44). As a result, measuring facial expressions during craving may provide a window into the various emotions people feel during different phases of craving. That in turn may eventually allow psychologists to help people deal with their urges.

Based on preliminary data, Sayette isn't yet certain that facial expression is key to diagnosing craving or whether the expressions he's finding are specific to craving or that they generalize to types of craving other than that for cigarettes. But, he contends, using facial expression to measure cigarette craving may help researchers unravel the emotional components of the phenomenon.

For example, some researchers posit that there are several different craving states. Traditionally, researchers believed craving was invariably linked with frustration. Now, some wonder if craving can also be a type of positive anticipation when a person knows his or her waiting is almost over.

"No one has really established if this is true," says Sayette. "But facial coding may provide an approach to examine this possibility."


Indeed, in a study conducted in 1995 (Experimental and Clinical Psychopharmacology, Vol. 3, No. 4, p. 417-423), Sayette and his colleague Michael Hufford, PhD, now at the University of Montana, found that people who craved cigarettes exhibited different facial expressions depending on their expectations about being able to satisfy their craving.

For example, when smokers deprived of cigarettes for 12 hours believed that they would soon be able to smoke, they displayed facial expressions associated with positive emotion--showing signs of a smile as measured with the Facial Action Coding System (FACS), which codes facial expression. But when smokers were led to believe that they couldn't smoke, they displayed facial expressions associated with negative emotions.

Sayette is now analyzing the results of a large follow-up to that earlier study, funded by the National Institute on Drug Abuse. Although he's reluctant to make firm conclusions, it's clear that the vast majority of study participants show characteristic facial muscle movements when put in a craving situation, he says. Much of the movement revolves around the mouth, with participants pursing, tightening, pressing and sucking their lips.

"We see a lot of action around the mouth," says Sayette. "A lot more than we would see by chance."

As yet, the data are unclear on whether people display either negative or positive facial expressions depending on the particular craving situation.

Weighing measures

Even if facial expression can discriminate two types of craving, two questions arise when one is trying to validate any measure. How sensitive is the measure? And how specific is it?

In terms of sensitivity, researchers would want to see a majority of study participants showing the facial expressions during self-reported craving, which Sayette does find.

A good measure must also be specific to the construct a researcher is trying to evaluate. If, for example, people make the same facial expressions when they're craving as they do when they're getting dressed in the morning, the facial expression wouldn't be a specific measure for craving, says Sayette. So far, researchers have yet to discover if the expressions Sayette has related to craving are related to any other psychological state.

"I don't want to oversell what we've done," he says. "It's a fair question to ask about the specificity of this kind of measure. Facial coding may provide another helpful, sensitive measure of craving. But by no means is it, or any other craving measure for that matter, the gold standard."

It's more likely to be useful in studying the underlying emotional components of craving than craving itself, says psychologist Stephen Tiffany, PhD, of Purdue University, who is considered one of the fathers of drug craving research.

"I think it's implausible, theoretically, that craving has a face like anger or joy or sadness," he says. "But it's quite clear it has affective or emotional components to it. So it makes a lot of sense to study the emotion systems [that are] concurrently activated with craving. And from that perspective, the face is as good a measure if not better than any other one."