Comparison of self-reported speed of eating with an objective measure of eating rate

Abstract

Slow eating may be beneficial in reducing energy intake although there is limited research quantifying eating rate. Perceived speed of eating was self-reported by 78 adults using a standard question “On a scale of 1–5 (very slow–very fast), how fast do you believe you eat?” Timing the completion of meals on three occasions was used to assess objective eating rate. The mean (SD) speeds of eating by self-reported categories were 49 (13.7), 42 (12.2), and 35 (10.5) g/min for fast, medium, and slow eaters, respectively. Within each self-reported category, the range of timed speed of eating resulted in considerable overlap between self-identified ‘fast’, ‘medium’ and ‘slow’ eaters. There was 47.4% agreement (fair) between self-reported speed of eating and the objective measure of eating rate (κ = 0.219). Self-reported speed of eating was sufficient at a group level to detect a significant difference (10.9 g/min (95% CI: 2.7, 19.2 g/min, p = 0.009)) between fast and slow; and fast and medium eaters (6.0 g/min (0.5, 11.6 g/min p = 0.033)). The mean difference (95% CI) between slow and medium eaters was 4.9 (-3.4, 12.2) g/min (p = 0.250). At an individual level, self-report had poor sensitivity. Compared to objectively measured speed of eating, self-reported speed of eating was found to be an unreliable means of assessing an individual’s eating rate. There are no standard protocols for assessing speed of eating or eating rate. Establishing such protocols would enable the development of population reference ranges across various demographic groups that may be applicable for public health messages and in clinical management.