Abstract
The paper of Sazonov et al (2008 Physiol. Meas. 29 525–41) addresses the topic of on-body sensor-based measurement and analysis of food intake and eating behaviour. The authors rightly pinpoint a lack of solutions to estimate eating behaviour and energy intake in contrast to the active development of energy expenditure prediction tools. Unfortunately, Sazonov and colleagues have missed reviewing a considerable amount of published research in the field of ubiquitous and wearable computing. Moreover, it should be noted that objective measurement techniques exist for laboratory studies of chewing and swallowing that could have served for the validation of their work. This letter summarizes relevant related works and identifies refinements of the study methodology suggested by Sazonov et al. Food intake behaviour is very variable and hard to capture. Nevertheless, the approaches towards automatic dietary monitoring (ADM) cited in this letter confirm the broad potential for sensing and pattern recognition techniques. ADM could eventually supplement or replace intake diaries.