Limitations with traditional dietary assessment methods are well documented. As a result, novel methods of dietary intake assessment and portion size estimation are constantly being developed(1). Food photographs have been shown to help individuals estimate food portion sizes more accurately(2). Therefore the aim of this study is to assess the accuracy of portion size estimation when facilitated by a digital food atlas intended for use in a smartphone app. 
Food portion photos were taken according to a standard procedure developed from published literature(3) and prepared for presentation on a smartphone. Each food had four portion images accompanied by the portion measure (g) and description. Ten foods from two food groups were selected for testing. Participants were recruited opportunistically at a stand in University College Dublin (UCD) in association with UCD healthy eating week. General demographic information was gathered via an online questionnaire.  Participants were presented with a pre-weighed serving of each food and were asked to select the corresponding portion image from a smartphone display. The percentage correctly identifying the portion sizes was assessed across the total population using SPSS© and differences across genders were analysed using a Chi-square test. Significant differences between genders was compared using a chi-squared test; p<0.05
In total 245 participants participated in the survey, 151 females and 94 males. The majority (48%) were aged between 18-20 years, 31% were 21-23 years and the remaining 21% >24 years. In the total population, across all foods, 70% of participants selected the correct portion image when guided by a digital food atlas. The food that was guessed correctly by most participants was cottage cheese (92%) and the food that was guessed correctly by the fewest participants was yoghurt (46%). Significantly more males (81%) than females (70%) guessed the correct portion size for potato chips whereas significantly more females (74%) than males (60%) guessed the correct portion size for pasta (p<0.05). Where incorrect portion sizes were selected, the tendency was toward over-estimation which is in line with other studies(4).
In conclusion, this study suggests that the use of a digital food atlas within a smartphone app may help with the estimation of portion size.
1.Gemming L, Utter J, Ni Mhurchu C, et al. (2015) J Acad Nutr Diet 115, 64-77
2.Naska A, Valanou E, Peppa E, et al. (2016) Public Health Nutr19, 2369-2376
3. Bouchoucha M, Akrout M, Bellali H, et al. (2016) Libyan J Med 11, 32676
4. Nelson, M., Atkinson, M., & Darbyshire, S. (1994) Brit J Nutr, 72(5), 649-663

05. Assessing the usefulness of a digital food atlas in estimation of portion size when presented in a smartphone app.