Can a web-based food record accurately assess energy intake in overweight and obese women? A pilot study

Background: Innovative dietary intake measurement tools, such as web-based food records, are becoming increasingly available for self-monitoring. However, the accuracy of this method has not been well studied. This pilot study aimed to evaluate the accuracy of energy intake (EI) estimated by a web-based food record, by comparison with total energy expenditure (TEE) measured by doubly-labelled water (DLW) in overweight and obese women. Methods: Total energy expenditure (TEE) was assessed in weight stable (-1 kg) women (n = 9), with a mean (SD) age of 34.5 (11.3) years and body mass index of 29.2 (1.4) kg m-2 over 10 days using the DLW technique. All food and beverages were self-reported for 9-days using a web-based food record and mean daily EI calculated. Food record accuracy was assessed by calculating the absolute (EI - TEE) and percentage (EI/TEE - 100) differences between EI and TEE. Women were identified as under-reporters of EI based on the 95 confidence limits of the expected EI : TEE of 1. Results: The mean (SD) self-reported EI was 8351 (1225) kJ day-1 [1996 (293) kcal day-1] and TEE was 10 648 (1774) kJ day-1 [2545 (424) kcal day-1]. The mean (SD) absolute difference in self-reported EI and TEE was -2301 (1535) kJ day-1 [-550 (367) kcal day-1], representing a mean reporting accuracy of 79.6 (14.1 ), with four participants under-reporting EI. Conclusions: This pilot study highlights the opportunity for the use of the Internet as a novel medium for recording and assessing dietary intake. Although further research is needed in more diverse population groups, the accuracy of web-based food records for assessing EI appears to be consistent with other published dietary intake methods.