How Scientists Study Food Addiction

"Their egg rolls are addictive." "That ice cream is like crack." "I need my chocolate fix." Food as a drug is a common trope in the ongoing cultural conversation about health, nutrition, and eating habits: It's apparent in advertising, with Frito-Lay's potato chip slogan "Betcha can't eat just one," and it's deeply coded into the way diners talk about restaurant food. Addiction psychology has long been woven into conversations about eating, but in the past decade, an increasing body of research has shown that there may be something to this analogy. Even if the socio-economic, environmental, and sociological factors associated with the global obesity epidemic could be resolved, what if some people really are just addicted to food?

Several studies have demonstrated some behavioral and neurologic similarities between substance addiction and food addiction (FA). Although the current edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM V) does not recognize FA as a psychological disorder, it does include binge eating disorder, which some believe to be a symptom of food addiction. Whether the eventual diagnosis is of "food addiction" or eating addiction, it's clear that addictive behavior has a growing role in discussions about eating behaviors and obesity.

Initial studies on food addiction focused on self-identified addicts, untested assessment tools, and the assumption that obesity is a reliable indicator of food addiction. But in 2009, researchers developed the Yale Food Addiction Scale (YFAS), a scientifically vetted, self-reported survey to identify people who may exhibit addictive behaviors towards food. Its applications are wide-reaching — it's been used to assess everything from weight-loss interventions to binge eating to psychological approaches to food. But in the seven years since its development, YFAS was mainly discussed in academic and medical circles — until a widely reported study last month applied it to establish the addictive properties of specific foods. Using the scale, researchers concluded that people may, in fact, respond to pizza like a drug: The results went viral, and the YFAS was abruptly launched into prime time, mentioned in the dozens of articles written about "The Most Addictive Foods, According to Science." But is that what the survey really says?

Already translated into three languages, YFAS is the current standard defining how researchers and healthcare professionals measure addictive food behaviors; it will likely be a fixture of studies going forward. Dr. Ashley Gearhardt, one of the survey's co-founders, emphasizes that "[scientists] still do not know if 'food addiction' is a real thing, but the YFAS allows researchers and clinicians to use standard diagnostic criteria to try and identify people who may be eating in an addictive-like way." Gearhardt and her colleagues built the survey based on more than a century of research on substance abuse, and the YFAS was patterned on the criteria for substance dependence described in the DSM-IV:

1. Food often taken in larger amounts or over a longer period than was intended (3 questions)
2. Persistent desire or unsuccessful efforts to quit (4 questions)
3. Great deal of time spent in activities necessary to obtain food or recover from eating (3 questions)
4. Important social, occupational, or recreational activities given up because of food consumption (4 questions)
5. Use continued despite knowledge of adverse consequences (1 question)
6. Tolerance (2 questions)
7. Withdrawal (3 questions)
8. Use causes clinically significant impairment or distress (2 questions)

When taking the YFAS survey, individuals are asked to focus on their behaviors within the last 12 months. In particular, they are instructed to consider consumption of high fat, sugary, or carbohydrate-heavy foods. The calorie-laden foods that make up the questionnaire correlate with brain responses associated with substance addiction: In their preliminary validation of the YFAS scale, Gearhardt and her fellow researchers note that "high fat sweets" have been scientifically shown to release opiates, while dopamine "has been associated with the perceived value of reward of both food and psychoactive substances."

The 25 questions on the YFAS are all answered with either a frequency (never, multiple times a week, daily, etc.) or a yes/no response. When the survey is scored, symptom counts ranges from zero to seven — if a patient indicates symptoms on at least one question of a given criterion, that criterion is met. A patient is considered to have a food addiction when they test positive for at least three symptoms and meet the criterion for significant distress or impairment.

Not surprisingly, higher scores on the YFAS are associated with cravings for palatable, processed, and fatty foods. As many published reports from last month suggested, addictive behaviors may be more likely with french fries than carrots; pizza landed the "#1" spot in the recent University of Michigan study that claimed to rank the "Most Addictive Foods." Other applications of the YFAS study have shown that rates of food addiction are higher in women, consistent with studies indicating that women are more likely to use the language of addiction when discussing food.

Don't blame addiction for your pizza intake just yet: Several issues emerge when evaluating 'addiction' based on a self-reported survey.

But there are several issues when evaluating "addiction" by self-reported survey. While the scale is intended to assess food addiction, it may misidentify individuals without food addiction who answer positively to questions. A person's interpretation of a question and his/her own behavior can influence the results in myriad ways: Those struggling to lose weight for bathing suit season may qualify as "food addicts" based on their efforts. An individual who perceives themselves to have an addiction to chocolate may technically test positive for the disorder, even when it's not clinically relevant. And research indicates that the most common symptom of food addiction is consistently reported as "persistent desire or unsuccessful efforts to cut down or control eating." While some studies have found that nearly 100 percent of obese individuals meet this criterion, the same may be true in more than 70 percent of respondents from the general population. These findings may significantly skew researchers' calculation of the prevalence of the food addiction.

There are further issues with self-reported surveys on the individual level. Participants' feelings at the time of taking the survey may influence answers. Some people will simply have better recall regarding their eating habits. Respondents also may intentionally alter answers to hide problems. And, most crucially, these surveys turn qualitative data (stories, descriptions) into quantitative data (numbers), asking participants to rate themselves on a scale with no opportunity for further explanation. This leaves answers up to personal interpretation. What one person perceives as a problem, another will view as a normal activity. For these reasons, health-care professionals use surveys to assess patients for addictive symptoms and consider them in the overall clinical picture — this is a screening test, not a diagnostic one.

When the "most addictive foods" study was released, it was no surprise that it blew through popular media: Who doesn't want to blame their pizza cravings on its inherent addictive properties rather than their own tastebuds? ("It's literally addictive!") That's a catchy line, but the study of food addiction and the application of the Yale Food Addiction Survey is still in its nascent stages — you may not be able to shift the blame just yet.