Characteristics of smartphone-based dietary assessment tools: A systematic review

Abstract

Smartphones have become popular in assessing eating behaviour in real-life and real-time. This systematic review provides a comprehensive overview of smartphone-based dietary assessment tools, focusing on how dietary data is assessed and its completeness ensured. Seven databases from behavioural, social and computer science were searched in March 2020. All observational, experimental or intervention studies and study protocols using a smartphone-based assessment tool for dietary intake were included if they reported data collected by adults and were published in English. Out of 21,722 records initially screened, 117 publications using 129 tools were included. Five core assessment features were identified: photo-based assessment (48.8% of tools), assessed serving/ portion sizes (48.8%), free-text descriptions of food intake (42.6%), food databases (30.2%), classification systems (27.9%). On average, a tool used two features. The majority of studies did not implement any features to improve completeness of the records. This review provides a comprehensive overview and classification scheme of smartphone-based dietary assessment tools to help researchers identify suitable assessment tools for their studies. Future research needs to address the potential impact of specific dietary assessment methods on data quality and participants’ willingness to record to ultimately improve the quality of smartphone-based dietary assessment for health research.

A systematic review and meta-analysis of studies of reactivity to digital in-the-moment measurement of health behaviour

“…Since truly objective digital assessment of consumption behaviours in daily life including smoking, alcohol consumption, and food intake is still in its early stages (21,48), it was decided prior to screening titles and abstracts that Ecological Momentary Assessment (EMA) (18) of consumption behaviours would be considered as long as participants were asked to record the occasion immediately before, during or after consumption, preferably using objective markers such as a photo (21) which minimises recall bias (18).…”

“…Consumption behaviours such as medication adherence, alcohol consumption, smoking, and eating behaviour were rarely studied, which may be the case because they are more difficult to assess digitally than physical activity, for which passive tracking is well established. Most notably, although digital assessment of dietary intake including objective indicators such as photos is common in behavioural research (21), measurement reactivity has not yet been studied in this domain: Apart from one study that investigated the consumption of both alcoholic and non-alcoholic drinks during nights out, no studies on dietary intake were identified. Future research needs to address this gap to provide important information on the validity of digital assessment of dietary intake as well as other consumption behaviours.…”

“…For instance, health-related behaviours such as physical activity and sedentary behaviour are increasingly tracked using wearable devices (20), while consumption behaviours such as eating are increasingly studied using smartphones (21) and body-worn sensors (22). These methods provide detailed insights into health behaviours in daily life and allow individuals to measure behaviour when it occurs, thereby reducing recall bias and increasing the validity of the collected data (18).…”

Sustainable healthy diets from the lens of behavioural science

“…The utilisation of more objective data collection, in parallel with self-reported behavioural outcomes, such as food photos, could also help improving the evaluation of these interventions. 75…”

Image Based Food Energy Estimation With Depth Domain Adaptation

“…The ubiquity of smartphones and their integration into every facet of life has made it possible to monitor one's health status such as vital signs, exercise and sleep statistics, etc. What and how much a person eats and drinks in a day is another key contributor to one's overall health, yet very difficult to assess [1]. Many existing mobile applications rely on manual user input to capture information about the foods they consumed [2].…”