Please note, this is recovered content from the former website of the New York Obesity Research Center website.


“Sip-and-Spit” Model of Taste Mixture Interactions.


Date: March 6th, 2008
Speaker’s Name, Affiliation: Diane Klein, Columbia University
Seminar Title: “Translational Research on Eating Disorders: Development of a “Sip-and-Spit” Model of Modified Sham Feeding

Although it is possible that abnormal meal size in humans is due to altered responsiveness of orosensory excitatory controls of eating, there is no direct evidence for this because food ingested in a test meal stimulates orosensory excitatory and postingestive inhibitory controls. We adapted the modified sham feeding technique (MSF) to measure orosensory excitatory control of intake of a series of sweetened solutions in the absence of postingestive negative feedback of ingested solution.

In the first study, 10 women without a history of eating disorders were randomly presented with cherry Kool Aid solutions sweetened with one of 5 concentrations of sucrose (0-20%) in a closed opaque container fitted with a straw. They were instructed to sip as much as they wanted of the liquid during 1-min trials and to spit the fluid into another opaque container, for 10 solutions per subject.

Nine of the 10 subjects spit out more solution than they swallowed, supporting the validity of this procedure in minimizing postingestive stimulation, and in these subjects, sucrose produced a significant increase in overall intake of sipped solution. In the second study, aspartame was substituted for sucrose in concentrations to produced approximately equivalent sweetness intensity and 10 control women and 11 women with bulimia nervosa (BN) were asked to sip and spit a series of 15 solutions (3 trials of 5 solutions each).

Again, presence of sweetener produced a significant increase in intake both across and within subject groups. Women with BN sipped significantly more of each solution with the exception of the 0.03% aspartame solution (p=0.054). Increased intake appeared to be due to baseline increase as there was no significant differential effect of increasing sweetener concentration on the two groups.

Preliminary data collected from 14 women with anorexia nervosa (AN) who do not engage in binge-eating behavior show decreased intake at each solution that is significant compared with BN subjects and that also appears to be related to difference in baseline intake of the unsweetened solution. Furthermore, with repeated presentation of this solution over trials, intake decreased in patients with AN while it remained stable in control participants and there was a trend towards an increase in BN.

Self-reported liking, wanting, and sweetness of the beverages did not differ among the groups. Results further validate this MSF procedure, its ability to distinguish among eating disorder diagnoses, and hypotheses that women with BN have increased orosensory excitation, while non-binge eating women with AN do not.


Q. What is the recommended % of sugar that the Kool Aid packet recommends adding?
A. My understanding is that they suggest making it with 10% sucrose.

Q. Do you have any impression at how large participants’ sips were?
We collected microstructure data, but we haven’t yet analyzed it.

Q. Why not look at your data in terms of volume?
A. We actually did that first, but the reviewers wanted to see it in grams.

Q. Did you have any preconceived idea of what would happen when you used the MSF procedure with patients with bulimia?
A. We thought they would have a higher intake overall, and higher intake of the sweeter tasting solutions.

Q. Did you test for an interaction of patients * controls for the treatment condition?
A. We did test that and it wasn’t significant. There’s a lot of variability in the data.

Q. What was the scale you used?
A. We used a 10-cm VAS that was anchored with “not at all” and “extremely.”

Q. Where does sucrose fall on this scale (compared to aspartame) when you look at measures of liking and wanting?
A. People give it about the same ratings as aspartame.

Q. Did you correct for menstrual cycle?
A. No, we didn’t.

Comment: The psychophysics do not show any of the characteristics you might expect if the luteal phase of the cylcle were effecting results. I’m not sure that correcting for the phase of the menstrual cycle would change the data any.

Q. What was the structure of the trial again?
A. The order of the solutions was randomized within each set of five trials.

Q. Did you look at the sweetness ratings of the unsweetened solutions?
A. I think we did and my understanding is that it didn’t change.

Comment: This might explain why the anorexics are going down while the patients with BN are increasing. They might be perceiving the odor as sweeter. There are lots of published studies that suggest that when you pair sweet with a kool aid flavor, the flavor becomes sweeter.

Q. What was the wording of the wanting question?
A. How much do you want more of what you just tasted?

Q. How would you characterize the underlying mechanism of why bulimics consume more than controls?
A. That is open to interpretation.

Q. Do you have any idea how much they liked the test drinks compared to drinks in their typical diet?
Many of them said not as much, but some subjects actually really liked them.

Q. I’m intrigued about your controls and why the controls performed the same with sucrose and aspartame? They all had diet drink experience, right? Could there be something different about their sipping styles?
A. We aren’t sure.

Q. Were these all college aged students?
A. The mean age of subjects was about 26 years old.

Comment: You used BMI, but you don’t have other measures of body composition. I would love to see insulin response or resistance. My impression is that bulimics are less insulin sensitive.

Q. What about measured fMRI in patients?
A. It would be very interesting, but we haven’t done that yet.

Q. Does Linda Bartoshuk have data on this?
A. We are currently measuring PROP status in our subjects now to see if there are any differences in taste sensitivity.

Q. Did any of your bulimics report chew and spit behaviors?
A. Yes, some did.

Q. Did you look at the salivary response?
A. We didn’t measure that, but we might predict differences between groups.

Q. If you assess motivation to consume, this type of experiment would produce strong evidence in that regard.
A. We are currently working on something like that where patients work for sweetener packets.

Q. What are the controls for your sweetener packets.
A. We are just having them press a computer key for sweetener packets, there is no alternative reinforcer in this paradigm.

Comment: I wonder if you used a cover story that you were testing a new mouthwash (instead of a drink or food), if participants with eating disorders would behave differently towards the testing procedure?

Q. Have you thought about the therapeutic aspects to this?
A. We haven’t thought of any.

Q. What do you think would happen if you did this with obese and BED patients?
A. We are doing that experiment right now, so that’s one of our future directions.