Abstract
Background: Research has shown a reduction in eating rate may be beneficial in the reduction of energy intake, obesity and diabetes risk. There is currently limited research as to what constitutes a fast eater and therefore how these individuals could be identified. Therefore the purpose of this study was to assess the effectiveness of self- reported eating rate at predicting objectively measured eating rate.
Methods: Participants were healthy young adults (n=78), mostly female (74%), at the University of Otago. A self-reported eating rate question was embedded in an eating habits questionnaire and administered at baseline and at two months. The question used was “On a scale of 1-5, how fast do you believe you eat?” the options provided were: 1- very slow, 2-relatively slow, 3-medium, 4-relatively fast, and 5-very fast. Each participant consumed a hot lunch meal comprising beef mince in a Bolognese sauce, a starchy carbohydrate (rice or pasta), and non-starchy vegetables. All meals weighed 550g and were consumed on three occasions. Using a digital clock, the candidate, while unobserved, recorded the start and finish times of the meals for each participant. The recorded times were used to calculate meal duration in minutes and eating rate in grams of food per minute. Self-reported eating rate categories and objective eating rates were compared using a mixed effects regression model, percentage agreement, and kappa coefficients.
Results: The number of people who self-identified as very slow, relatively slow, medium, relatively fast and very fast were 1, 9, 31, 35, and 2, respectively. Due to the small numbers in the very slow and very fast categories, very slow and relatively slow were combined to form the slow category (n=10), and very fast and relatively fast were combined to form the fast category (n=37). On average, self-reported fast, medium, and slow eaters ate at a rate of 48.95±13.65g/min, 41.77±12.19g/min, and 35.28±10.46g/min, respectively. There was a mean difference of 14 (5.27-22.07, 95%CI) g/min between self-reported fast eaters and self-reported slow eaters (p=0.004). There was no significant difference in eating rate between the medium and fast self- reported categories and between the medium and slow self-reported categories. Each self-reported eating rate category had a wide range of eating rates with considerable overlap, and thus a low sensitivity of 56.8% was found when identifying fast eaters. Individual analyses revealed ‘fair’ agreement between self-reported eating rate and objective eating rate (κ=0.219), with exact and adjunct agreements of 47.4% and 48.7%.
Conclusions: Asking someone about their perceived speed of eating is highly subjective and unreliable. At a group-level, self-reported eating rate was sufficient to detect group mean differences in eating rate between self identified fast and slow eaters, however self-reported eating rate had poor sensitivity for classification of individuals. Hence, self-reported eating rate is unlikely to provide reliable information on an individual’s actual speed of eating. However, even a timed rate of eating is quantitatively vague given there are currently no population reference ranges of slow, medium and fast eating speeds by which to compare.