This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR mhealth and uhealth, is properly cited. The complete bibliographic information, a link to the original publication on http://mhealth.jmir.org/, as well as this copyright and license information must be included.
Accurately assessing the diets of children and adolescents can be problematic. Use of technologies, such as mobile apps designed to capture food and beverages consumed at eating occasions with images taken using device-embedded cameras, may address many of the barriers to gathering accurate dietary intake data from adolescents.
The objectives of this study were to assess the willingness of adolescents to take images of food and beverages at their eating occasions using a novel mobile food record (mFR) and to evaluate the usability of the user confirmation component of the mFR app, referred to as the “review process.”
Mixed methods combining quantitative and qualitative protocols were used in this study. Adolescents (11-15-year olds) attending a summer camp were recruited to participate in the study. First, the participants were asked to take images of foods and beverages consumed as meals and snacks for 2 consecutive days using the mFR app running on an iPhone and the number of images taken was noted. This was followed by focus group sessions to evaluate usability, which was analyzed by content and themes. After using the mFR, a think-aloud method was used to evaluate the usability of the mFR method for reviewing system-identified foods (ie, the review process). A usability questionnaire was administered at the end of all activities.
The mFR was accepted by the majority of the 24 boys and 17 girls (n=41) but varied according to gender and eating occasion. Girls were significantly more likely than boys to capture images of their eating occasions (Fisher exact test,
The adolescents’ experiences with and feedback about the mFR highlighted the importance of increased training, reminders, entertainment (eg, games), and training with practice in using the device to capture complete dietary intake as part of their active lifestyles.
Collecting information about dietary intake from children and adolescents is challenging. Developmental stages add to the complexity of deciding whether to gather information from the parent, the child, or both. Burrows and colleagues [
Livingstone et al [
Taking images of eating occasions has been proposed as a method for reducing burden on adolescents participating in dietary studies [
The mFR allows users to take images of their food and beverages at eating occasions, which could address the barriers of hand-written pen and paper or digital entry methods that still require spelling and focused entry. Image analysis methods have the potential to automatically identify foods and beverages in the image and estimate volumes [
The Pew Research Center recently reported that as of 2013, 78% of adolescents aged 12-17 years in the United States have their own mobile phone and nearly half of those (47%) own smart phones [
Diagram of the Technology Assisted Dietary Assessment (TADA) system that starts with capturing an image with the mobile food record (mFR).
Data were collected from adolescents participating in a residential summer camp on the Purdue University campus in 2010. There were 41 adolescents (17 girls and 24 boys) between 11 and 15 years of age who participated in this study. The active camp environment included participation in organized sports, hiking, river rafting, rock-wall climbing, relay races, swimming, bowling, social events, classes, free time, and attending a minor league baseball game. The activities described in this study were an add-on to the original objective of the residential camp, which was to establish dietary calcium requirements for bone health [
This mixed-methods study sequentially used quantitative and qualitative data-collection methods and analyses to better explore adolescents’ willingness to use the mFR [
During the second 3-week session of the camp, a subset of the same children used the iPhones again in groups of 8-11 at a time for breakfast and/or lunch and/or dinner. This data-collection protocol differed from the first session, in that the participants were monitored during the image capture process at each meal. At the end of the day, participants were provided instructions according to the think-aloud method [
Mixed methods study design with quantitative and qualitative protocols.
Beta testing the review component. The user selects an eating occasion to review. The before-eating image is in landscape view as displayed on the left. Foods are denoted with pins and the food label appears on a color coordinated bubble. To correct a misidentified food, the user taps the bubble associated with the food as shown on the right and selects the correct food from a list of foods (not shown).
The camp situation was unique in that the eating occasions were limited to breakfast, lunch, an afternoon snack, dinner, and an evening snack. A 4-day cyclic menu consisting of the aforementioned eating occasions was used. Each camper was provided with adequate food to maintain his/her weight, according to his/her estimated energy requirements, which was monitored on a daily basis. No other foods or beverages were allowed. Over the 2 days of using the mFR, each participant was expected to capture a before- and an after-eating image (image pair) for each eating occasion. It was acceptable for participants to also only capture a before-eating image because they were expected to finish all foods and beverages provided as part of the main study protocol. Thus, a usable image was considered to be either an image pair or just the before-eating image. Usable images were considered indicative of participants’ willingness to comply with using the mFR. The set eating occasions allowed for the identification of the number of usable images and to classify participants as demonstrating high, moderate, or low willingness. One boy was unable to complete the task on both days due to reasons unrelated to this study and was excluded. For the remaining participants, the minimum expected total number of images was 360 (40 participants with 9 eating occasions, including 1 breakfast, 2 lunches, 2 dinners, 2 afternoon snacks, and 2 evening snacks). The boys and girls were classified as having taken 8-9 images (high willingness), 6-7 images (moderate willingness), and 3-5 images (poor willingness). At the end of each day, the research staff downloaded the mFR images captured using the TADA app to an Apple computer. The images were systematically reviewed by a trained analyst to enumerate only those containing images of the known meals or snacks.
Focus groups were convened to explore the issues influencing participants’ willingness to capture images. The same script was used for all of the focus group sessions. All the girls attended 1 of the 2 girl-only focus group sessions. Three boys were unable to attend 1 of the 2 boy-only focus group sessions due to reasons unrelated to this study. Each focus group was led by 2 experienced moderators. Examples of questions relevant to this paper were as follows: “If we were to use these phones with other groups of teenagers, what could we do to help them remember that the main reason they have the phone is to take images of their meals and snacks?” Follow-up questions to this topic were “We have heard a lot of ideas about getting people to use the app, I just want to follow-up and see if anyone has more to add, what would motivate you to use the app each time you eat?” Another topic was “What did you like most about using the phones?” Copies of the entire script can be requested from the corresponding author of this paper. Two support staff hand-recorded each session and then pooled their notes. All staff present at a session reviewed the notes to achieve final consensus on the nature of the participants’ responses and any unscripted probes from the moderators. Two staff, other than the recorders, independently reviewed and analyzed the notes, categorizing them according to content and common themes. After consensus, these were finalized.
Usability testing of a beta version of the user review process component of the app was used to identify potential improvements. Before the usability testing of the image review component, the children took images of at least 2 meals using the mFR. Thus, during the usability testing, all children were reviewing images of their own meals and were instructed to verbalize everything they were thinking as they decided which button to press first, and in every subsequent step, using the think-aloud method [
A usability questionnaire was used to further inform behaviors related to using the mFR. The questionnaire was mainly composed of forced choice questions with 5 responses of “strongly disagree” to “strongly agree” and several open-ended questions [
Willingness to take images was evaluated objectively by enumerating images captured of the known eating occasions. Each image was classified as being taken of a particular meal or snack. Information about each image was entered into a database, including whether the image was the before- or after-eating image and useful metadata (ie, date and time stamps). Data from the image evaluation and the questionnaire were entered into a Microsoft Access database (Microsoft, Redmond, WA, USA) and then imported into an SPSS version 17 database (SPSS Inc, Chicago, IL, USA) for data analysis. Statistical examination to compare girls and boys included frequencies, two-by-two tables, Chi-square test, Fisher exact test, and independent samples
The classification of participants by ethnicity is shown in
Classification of study participants by ethnicity.
Ethnic group | Boys |
Girls |
Totala
|
Hispanic | 22 (91.7) | 17 (100) | 39 (95.1) |
Non-Hispanic | 2 (8.3) | 0 (0) | 2 (4.9) |
aThere were a total of 24 boys (59%) and 17 girls (41%) in this study (N=41).
Characteristics of adolescents using the mobile phone food record (mFR) while attending a summer camp.
Characteristics | Boys | Girls | ||
|
Mean | SD | Mean | SD |
Age, years | 13.9 | 0.9 | 13.5 | 1.0 |
Weight, kg | 73.0 | 25 | 67.6 | 20 |
Height, cm | 167.1 | 9.1 | 158.6 | 5.1 |
Body mass index | 25.7 | 7.2 | 26.9 | 8.0 |
In the focus groups, participants were asked what they liked best about using the mobile phones; the theme mentioned most often were games and other forms of entertainment installed on the device (
What did you like most about using the phones?
Games
Internet
It had a case
Calling
Tracking device that provides information if people steal it.
What would motivate you to use the app each time you eat?
Knowing that others are using/doing the same thing as you.
Reward for a week straight with adding fun apps if we take all our pictures. Even a reward after 2 straight days of using the app.
Different funny pictures pop-up every time we take a picture. Maybe even saying “good job” and laughing.
Provide entertainment with pop-up pictures.
What did you like most about using the phones?
Games
YouTube
Music
Good for passing time
Apps
Did not get in trouble for using it
Keeping up with the World Cup
Internet
What would motivate you to use the app each time you eat?
Money
Extra money
A picture that gives an “awesome” after taking the image or “congratulations” with fireworks in the background.
A contest—whoever remembers to take the most photos of their meals wins something (eg, money).
If the phone had a bleep or vibrate around meal time.
Something pops up on the screen around meal time to remind you
Receiving a text message from TADA.
Alarms
The usability findings of the review process were useful (
The usability findings from the questionnaire pertaining to the app as a whole are shown in
Responses to statements in usability questionnaire among adolescents after using the Technology Assisted Dietary Assessment iPhone (n=40).
Statement, as presented | Agreed by boysa
|
Agreed by girlsb
|
|
|
|
|
|
|
Use | 21 (91) | 11 (65) |
|
Carry around | 20 (87) | 11 (65) |
|
Include in the picture of my meals | 20 (87) | 12 (71) |
Understanding the purpose of the TADA iPhone app would have been helpful | 20 (87) | 17 (100) | |
Carrying 2 telephones is easy (n=31c) | 16 (84) | 12 (100) |
aBoys (n) = 23
bGirls (n) = 17
cRefers to carrying TADA iPhone and the personal mobile phone. The numbers of boys and girls who were carrying both were 19 and 12, respectively.
This study assessed adolescents’ willingness to take images of eating occasions using a mobile phone and observed their interactions with completing the beta version of the review process component of the app. A notable strength of this study was an objective marker for eating occasions. Just under half of the participants were moderately willing to capture images of their eating occasions indicating that some type of reminder system would be important for capturing total diet record. Girls appeared to be more willing to take images than boys, suggesting that more effort is needed to make this type of activity more compelling to boys. During the focus group sessions, the participants volunteered the importance of incentives and reminders to assist them in remembering to take images of every eating occasion. Whereas only basic instruction was provided about how to take the images, more extensive training may clarify the importance of taking images and might enhance willingness and cooperation [
Adolescents especially liked the game apps loaded on the iPhone. All of the individuals objectively classified as demonstrating high willingness also wrote “games” on the questionnaire as the item they liked most about the telephone. This would suggest that games may enhance the level of interaction with the telephone, which then transfers to remembering to take images of foods and beverages at eating occasions. Providing a mobile phone with preloaded games and apps or offering a selection of games or apps may provide additional motivation for cooperation. A study by Baranowski and colleagues [
During the focus group sessions, both boys and girls expressed a desire to know why taking images of what they eat would be useful or important. Based on the questionnaire responses, nearly all participants agreed that understanding the purpose of the mFR would be helpful. Thus, full transparency would be important in gaining commitment and cooperation from adolescents. In this study, the amount of information initially shared with the participants was limited to determine salient issues. Thus, even with modern technology, the need for full disclosure remains important to adolescents.
Adherence was most influenced by the type of eating occasion (eg, snacks being the worst). Compliance with the dinner meals was substantially lower than the breakfast and lunches. The lower compliance with the dinner meals can partially be explained by the 1-dinner meal that was a picnic on the grass during a minor league baseball game. This particular meal was the most noncompliant meal and reinforces the importance of incorporating reminders to engage an individual when such occasions occur. The minimal training provided in this study was likely inadequate and highlights the need to develop appealing training to aid in the aforementioned situations [
Although carrying and using the fiducial marker and carrying 2 telephones intuitively seem burdensome, the majority of the participants agreed that dealing with the fiducial maker and 2 mobile phones was easy. Boys appeared to be more positive than girls with regard to using and carrying the fiducial marker. By contrast, the girls were more likely to state that the protective covers on the telephones were attractive, and if they also included space for the fiducial marker this may be additionally useful. Such a feature could promote a more positive preference among girls regarding the use of a fiducial marker. However, to improve the number of images including a fiducial marker, participants suggested incorporating an automated system into the app to alert the user when the fiducial marker was not visible. In response to this, the opening screen of mFR TADA app has been modified to include a cartoon of the fiducial marker as a reminder [
A limitation of this study was the hand recording of the think-aloud method, as well as questions and discussions during the focus group sessions being interpreted through the understanding of the recorders and not amenable to independent reanalysis by other coders. Another limitation was the time of year. The summer is an active period for adolescents and this could influence their ability to remember to take images of their eating occasions. During the school year, there may be other barriers, such as restriction of mobile devices in the school environment. By contrast, daily activities during the school year may be more routine than in an active camp environment. Nonetheless, an exploration of challenges with the community dwelling situation and the school environment is in order. Some additional next steps would include comparing results with biomarkers and determining the lowest age at which children are able to capture images either by themselves or with the aid of parents or other adults. Surrogate parent and care-provider reporting was found to be promising with few limitations [
The mFR was accepted by the majority of adolescents in the study, but varied according to gender and eating occasion. The adolescents’ experience of using this novel method of technology to assess diet and their feedback highlighted the importance of reminders, games, and more training and practice in using the TADA app. Addressing these recommendations may assist adolescents with measuring their dietary intake as part of their active lifestyles.
mobile food record
Technology Assisted Dietary Assessment
We wish to thank Professor Connie Weaver and Dr Berdine Martin for facilitating access to “Camp Calcium.” We also thank the study participants. This study was supported by grants from the National Cancer Institute (Grant No 1U01CA130784-01) and the National Institute of Diabetes, Digestive, and Kidney Disorders (Grant No 1R01-DK073711-01A1).
The study was conceived and designed by CJB, DAK, TES, and EJD. Data were acquired by CJB, DAK, TES, SP, MB, and ZA. Data were analyzed and interpreted by CJB, DAK, TES, TA, MB, and ZA. The paper was drafted by CJB, TES, SP, and AJH. The paper was revised for intellectual content by CJB, DAK, TES, TA, and AJH. All authors approved the final content of the paper.
None declared.