Minutes of the Columbia University Seminar on Appetitive Behavior(#529)

Date: March 5th, 2009

Seminar Title: "Bulimia Nervosa: A Consequence of the Obesity Epidemic"

Speaker's Name and Affiliation:

  • Michael Lowe
  • Drexel University

Presiding Chair: Harry R. Kissileff, Ph.D.

Rapporteur: Kathleen L. Keller, Ph.D.

Attendees and their Affiliation:

Kathleen KellerColumbia/Obesity Research Center
Harry KissileffObesity Research Center
Miriam BocarslyPrinceton
Nicole AvenaRockefeller
Laura BernerNYSPI
Jen BushNYSPI
Eva KovacsUnilever
Jennifer NasserNYORC/Drexel
Carla WolperNYSPI & NYORC
Chris OchnerNYORC
Laurence NolanWagner
Stefani FanaraWagner
Sally Ann LedermanColumbia University Institute of Human Nutrition
Poornima ShankarRutgers
Beverly TepperRutgers
Dara BellaceWeill Cornell
Rachel MarshNYSPI
Marina GershkovichNYSPI
Michael DevlinNYSPI
Barbara SmolekCUMC
JA GrinkerUniversity of Michigan
Susan CarnellORC
Lauren PumaORC
Tatiana UngreddaORC
Allan GeliebterORC

Summary: (Prepared by the speaker)

Body weight has generally been assumed to be relatively unimportant in restrained eaters and in those with BN. However, though most bulimic individuals are in the normal weight range, Russell (1979) initially noted that they previously weighed considerably more. Interestingly, research on restrained eaters, who also are higher than average in weight suppression (highest weight ever ¨C current weight) started at about the same time that BN first emerged. It may be more than coincidental that BN was first identified, and that restrained eating research was initiated, at about the same time that the obesity epidemic was accelerating rapidly (late 1970s-early 1980s).

Though restrained eating is currently viewed as an analogue of the processes that can culminate in eating disorders, it appears more likely that restrained eating and dieting in normal weight women represent efforts to reverse or prevent weight gain than they are attempts to get skinny. Research has shown that restrained eaters (as assessed by the Restraint Scale) have higher BMIs than unrestrained eaters and are also more weight suppressed. Research by Stice and colleagues has shown that restrained eaters eat no less than unrestrained eaters in the natural environment and Lowe and colleagues have shown that chronic dieting predicts future weight gain. Restrained eaters also often show biological characteristics that would make them susceptible to future weight gain (e.g., higher free fatty acid levels; reduced leptin levels).

There is also evidence that individuals with BN are also prone to overweight; they have a heightened risk for childhood obesity and their parents are more likely to be overweight. They also show high levels of weight suppression (mean weight suppression was found to be 9.6 kg in outpatients and 12.0 kg in residential patients). Weight suppression also predicts accelerated weight gain among both nonclinical individuals and in those with BN and predicts poorer outcome in CBT treatment. Finally, recent evidence indicates that roughly half of BN individuals reach their highest-ever weight after the onset of their disorder, indicating that they often experience very large weight gains as their disorder progresses.

As a field, we have 1) over-emphasized the role of undifferentiated "dieting"in BN, 2) largely ignored Russell's initial emphasis on the possible role of weight suppression in the disorder, and 3) under-emphasized the apparent susceptibility of bulimic individuals to weight gain and overweight both before and after the onset of their disorder. In retrospect, it appears that the field "went cognitive"prematurely.

Discussion:

Q. Has anyone ever done a study in a lab where you asked people to eat as much as they would like (not binge)?
A. No, but I think we just identified our next collaboration.

Q. In the 1990s, Heymsfield and colleagues did a study where they used the Minnesota EAT scale and obese people actually scored higher on the "lying"scales and the "wanting to please"scales. Does this relate to what you are talking about?
A. Well, in this study, they used doubly labeled water, and they controlled eating and kept track of body weight. Thus, we know that subjects were not underreporting or lying about how much they ate.

Q. The study referred to in the question before raises a very important issue. How much does social desirability account for the effects you are seeing?
A. Social desirability probably explains in part why restrained eaters and obese individuals under-report their intake the most. However, it doesn't have anything to do with Stice et al.'s (2007) findings that restrained eaters consume no less food in the natural environment than unrestrained eaters.

Q. Did you (or others) define what you meant by the question "are you dieting?"
A. People have started to make those distinctions. We have begun asking about the temporality (eg. "Are you currently dieting?) as well as about subjects' intentions (eg. "are you dieting to lose weight or to avoid gaining weight?").

Q. Are people who say they are dieting typically restrained?
A. Yes, but being on a dieting is definitely not just a case of being more restrained ¨C dieters behave in a fundamentally different way than restrained eaters who are not dieting.

Q. Among those who are dieting and restrained, what % are weight suppressed? Is weight suppression a good predictor of weight gain?
A. We have found weight suppression to be a reliable predictor of subsequent weight gain in both nonclinical women and in bulimic patients.

Q. Where would you put people who diet before they go on a trip? Are they "restrained"?
A. I would call them conscientious Americans; they are compensating before they overeat rather than afterwards, which probably is more effective.

Q. In those data looking at fat oxidation, you'd want to look at the composition of the entire diet (not just response to a meal), right?
A. Yes, but they are comparing restrained to unrestrained (under those meal conditions).

Q. Why don't people focus much on the other subscales of the Three Factor Eating Questionnaire, like "disinhibition"and "hunger"?
A. There are a lot of investigators who study disinhibition, but I suppose people are interested in restraint because it would appear to be critical to preventing and treating obesity. Also of considerable interest are people who are both score high on both the restraint and disinhbition factors of the TFEQ.

Q. In relation to Harry's question, most of our studies show that people who are disinhibited have higher body weights, but we haven't found many relationships with hunger.
A. Yes, that's true.

Q. Even though they are correlated (restraint and disinhibition), some people have one behavior but not the other.
A. Actually, there is little or no correlation between TFEQ restraint and disinhibiton; when correlations are found they are negative.

Q. What % of bulimics were formerly overweight?
A. I'll show those data later.

Q. What is the Y-axis in that graph?
A. Those are the standardized scores. The z-scores.

Q. In your data, are there any genetic or physiological markers to differentiate the restrained from the bulimic who wants to lose a lot of weight?
A. No, but if we could find them that would be very important. There is a lot more depression in bulimics. We still haven't defined all of the behavioral differences. If there is a genetic marker of self-control of food intake, maybe bulimics have it because they have a remarkable ability to control dietary intake when they are in periods of restriction.

Q. Does restricted eating come before the development of bulimia?
A. Yes for about 80% of the cases. When they set their mind to this restriction, they are incredibly successful.

Q. What about current anorexics? How would they end up on this Drive to be Thin Scale? vWe wouldn't be able to assess them with this measure because they are already below 15% of their ideal body weight.

Q. Did you look at the actual weight gain over the course of the study?
A. I will get to that.

Q. Is the drive for thinness in bulimia nervosa internal or external?
A. It's internal.

Q. What % of bulimics are overweight before onset of the disease?
A. About 50%

Q. Is the goal in the future to be able to separate people on the TFEQ from those that are restraining to be healthy from those who are doing it to be thin or control their weight?
A. Yes, we'd like to do this, but we can't as of yet.

Q. My understanding is that the TFEQ was developed by Stunkard and Messick using factor analysis of where a number of different behaviors correlated. Has anybody redone these factor analyses to see if the TFEQ still applies?
A. People have attempted to do this. Rossner (from a Swiss group) has developed a new measure, but the validity has not been tested.