The Relationship between Food Related Compulsivity, Memory, Intellectual Ability, and Body Mass Index in Prader-Willi Syndrome 

Laura Holsen, Jennifer Zarcone*, Mary Caruso, Jamie Young, David Richman, Travis Thompson, Institute for Child Development, University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas City, KS 66160 USA.

Introduction: One of the common characteristics of Prader-Willi Syndrome (PWS) is compulsive food-related behavior. Although the specific cause of compulsive behavior in PWS is unknown, several tasks associated with memory, compulsive behavior, and food seeking may offer clues to the phenotype/genotype relationship in PWS.  In this study, two tasks were used to evaluate several dimensions of these compulsive behaviors.  

Methods: The first task was a delayed match-to-sample (DMTS) task that required participants to recall pictures at 0-, 15-, and 30-second delays. The goal of the task was to determine whether individuals with PWS would perform better if the stimuli used were of food versus nonfood pictures. Seventeen people with PWS, aged 11 to 48, participated.  The second task involved baiting a break room with several low fat food items in sites that varied from “acceptable” (e.g., on a paper plate) to “unacceptable” (e.g., in the trash can). Twenty-nine typically developing adults and children as well as six individuals with mental retardation and 12 individuals with PWS were surveyed to determine the degree of “acceptability” of each site according to each group, similar to the procedure used by Dykens (2000). Finally, nine individuals with PWS and three control participants with mental retardation and compulsive behavior were left alone with leisure activities in the baited break room for a 15-minute session.

Results: On the short-term memory task there was a significant relationship between IQ, compulsivity ratings, and the number of perseverative or compulsive errors made.  Participants with high IQ and low compulsivity made the fewest errors, while those with low IQ and low compulsivity made the most errors.  When looking at the relationship across food and nonfood stimuli, food-related errors were predicted by compulsivity level, while IQ predicted non-food errors. In addition, there was only a moderate degree of correspondence (60%) between PWS and typically developing adults with regard to what they felt were the most and least acceptable food site. When the individuals with PWS were in the baited break room, only three individuals with PWS ate a significant number of food items. These three individuals did not differ from the rest of the group according to IQ, age, or compulsivity score; however, they had significantly lower Body Mass Index (BMI) than the rest of the participants.

Discussion: Performance on a short-term memory task appeared to depend on level of compulsivity when food stimuli were presented and IQ when non-food stimuli were presented.  Highly compulsive people with PWS were capable of performing well on short term memory tasks involving food-related stimuli, which are presumably highly motivating.  On the other hand, on tasks involving less motivationally laden stimuli, IQ played a more important role.  For participants in the food-seeking task, it appears that the degree to which their diets were restricted (resulting in lower BMI scores) may have had an affect on the amount of time spent seeking food and the amount of food covertly consumed.

July 2003