All examinations were carried out with written informed consent from study participants. The NutriDiary evaluation study was approved by the Ethics Committee of the University of Bonn (project identification 445/23). All data collected in this study has been pseudonymized to protect participant privacy. No reimbursement was provided to participants for their involvement in this study.

To our knowledge, an app-based WDR system with barcode scanning function for scientific studies in Germany does not exist so far. The popularity and widespread use of smartphones make NutriDiary a promising alternative to the traditional pen-and-paper approach. Different ways for entering food items (text search, barcode scanning, and free text entry) and the NutriScan function enable users to record the products they consume in a very detailed manner. The digital data output has the potential to reduce the burden for researchers also. In this study, we evaluated NutriDiary in a convenience sample of experts and laypersons and found good usability and acceptability.

The median SUS score of 75 indicated good usability of NutriDiary. The SUS score has been developed as a means to measure the overall perceived usability of a system [26,27]. However, technology-based dietary assessment instruments for scientific purposes are very specific tools that are not designed for everyday use but to generate scientifically useful data. The collection of these data is often challenging for the participants. The first statement of the SUS questionnaire is “I think that I would like to use this system frequently.” This statement is unlikely to be agreed upon by many people in the case of a WDR due to the burden on participants by the method itself. Nevertheless, the advantage of the SUS score is that it offers the possibility to compare systems used in the same context [27]. The median SUS score of 75 for NutriDiary (mean SUS score: 74, SD 15, for comparison) is comparable with SUS scores for other similar technology-based food records such as “METADIETA-web” (mean SUS score: 68, SD 15, n=26) [31], the “Eat and Track app” (mean SUS score: 69, n=15) [32] or the “Traqq app” (mean SUS score: 79, SD 15, n=22) [33].

Age has been discussed as a major limiting factor for the use of technology-based systems [29] or innovative dietary assessment instruments [31,34,35]. Consistently, the results of the backward selection procedure showed that age was identified as a potential predictor for a lower SUS score in the examined sample. Further, older participants took slightly longer to complete a WDR with NutriDiary than younger ones. Feasibility testing of the digital food record METADIETA-web by Vitale et al [31] also showed that the preference for using the digital tool instead of the traditional pen-and-paper method decreased with increasing age [31]. Older generations did not grow up with smartphones and computers and are therefore likely to be less intuitive with apps and web-based tools than younger generations. However, longer completion times could also be explained by higher accuracy in entering items and greater patience [36], which would be in line with the observation that older participants used fewer estimated household measurements instead of weighing than younger ones in our study (0%‐23%, 4% for age group 3 vs 0%‐31%, 7% and 0%‐26%, 8% for age group 2 and 1, respectively). Nevertheless, an understandable introduction to the use of the instrument might be particularly important in the group of older adults. For practical reasons, only an introductory video was sent out in this study. Our experience in training users to use the NutriDiary app shows that older people tend to ask more questions. This can only be addressed in a face-to-face conversation where it can be ensured that all relevant information is conveyed. Therefore, we recommend a personal introduction when using NutriDiary in epidemiological studies, if feasible.

The median SUS score was higher in women than in men and higher in the group of experts than in the group of laypersons. This may partly be explained by the fact that the proportion of older people was higher in men than in women and higher in the group of laypersons than in the group of experts. The group of laypersons also included students of nutritional science. Although this group certainly cannot be described as experts, the students may have some background knowledge that could have influenced the outcome. Whether this makes them more critical or less critical in their judgment remains questionable.

WDRs have the potential to provide the most accurate description of the types and amounts of the foods consumed over a specified period of time, but they are also considered to be one of the most burdensome and elaborate dietary assessment methods. When conducting a WDR, participants have to weigh and write down everything they eat and drink and always carry their kitchen scales, the record sheet, and a pen with them. This process is exhausting and time-consuming and requires a high level of cooperation from the participants [37]. The median completion time for conducting a WDR with NutriDiary was 35 minutes (1.6 min/item, Table 3) and can be rated as acceptable. When looking at the median relative completion time of the sample meal (here, the amount of the presented foods was predefined), it can be assumed that the single entering process (without weighing) takes around 0.8 minutes per item. This result suggests that weighing accounts for about half the completion time. Keeping this in mind, the single entering process is roughly comparable with the reported average completion time of other text-based DR apps, for example, with “My Meal Mate.” Here, participants needed on average 22 minutes per day to complete a record with estimated (not weighed) portion sizes [38].

Whether the NutriDiary app significantly reduces the completion time compared to the traditional pen-and-paper method is questionable, because the digital version of a WDR does not change the fact that weighing is required for this dietary assessment method. However, NutriDiary offers some advantages that can make its use more attractive than the traditional method. First, most people always carry their smartphone with them and recording “on the go” is more practical than using pen-and-paper. Presumably, this makes it less likely that participants forget their records and need to add the meal at a later time. Second, features such as the recipe book and the barcode scanner can make recording easier, as frequently consumed food combinations and recipes can be stored and retrieved later, and consumed products can be added more easily. Furthermore, integrated standard household measurements make it easier to estimate the quantity if weighing is not possible, for example, when eating out of home. The photo function allows participants to add a photo of the meal, which can help the postprocessing. To figure out whether NutriDiary is more attractive to study participants than the traditional pen-and-paper method, we added 2 more questions to the SUS questionnaire (, Multimedia Appendix 2). The result clearly showed that the vast majority would prefer using the NutriDiary app instead of a paper-based food record. To avoid errors caused by application problems using technology-based instruments, an understandable introduction and information structure are necessary to prevent frustration and enable participants to help themselves quickly if they have difficulties. For this purpose, we designed a website where all information and support materials are centralized in 1 location [24].

When using traditional DRs in epidemiological studies, the postprocessing of the DRs is very time-consuming. The data needs to be digitized and manually coded by the study staff. In this study, 68% (IQR 55%‐82%) of the food entered in NutriDiary was already coded. Furthermore, the data was already available in digital form, suggesting less costs and time in data postprocessing, compared to the pen-and-paper method. NutriDiary was designed as an app for keeping WDRs, meaning that participants are asked to weigh all the food and drinks they consume. However, as experience showed that this is not always possible, the app also offers the choice of standard household measurements (eg, teaspoon, glass, or portion) to estimate the quantity in situations where weighing is not possible. Considering the burden of weighing all foods, this selection option might tempt participants to estimate rather than weigh. However, the proportion of estimated quantities was rather low and can be considered manageable in this study.

Within the development process of NutriDiary, building an appropriate food database structure and the collection, standardization, and integration of barcode and food packaging information to enable food entry via barcode scanning was one of the most challenging tasks. The underlying database determines the users’ success in searching for food items and is essential for the functionality and accuracy of technology based, self-administered dietary assessment tools [3,39]. Therefore, we aimed to make the database as complete as possible. Due to a high number of available branded foods and the frequently changing food market, this is a major challenge [12,40-43]. According to the Food Federation Germany [44], there are more than 170,000 food products available on the German food market. Every year, about 40,000 new products are launched and just as many disappear [44]. It also happens that manufacturers change food recipes, which can also change the ingredient lists and nutritional values. These circumstances mean that the database needs to be constantly updated. To make this possible, the NutriScan function described above was developed and integrated. If the app is used regularly, the database is regularly updated with new products recorded by users.

The NutriDiary evaluation study had some strengths and limitations. As NutriDiary was developed for scientific use, practicing nutritionists were also recruited as experts. Although experts are not intended users of the app, their evaluation allows for a professional view of the usability of NutriDiary. Furthermore, experts decide on the use of the assessment instrument in scientific studies, which is why their opinion on usability is of particular importance in this context. To assess user-friendliness in older age groups as well, we specifically recruited older participants. However, this study population was not representative of the general German population. Most participants were female, young, and highly educated. This limited the generalizability of our results. Nevertheless, it is important to acknowledge that participants in epidemiological studies frequently possess a higher educational status compared to the general population. In addition, participants were aware of the purpose of this study, which was to evaluate the usability and acceptability of NutriDiary, making it difficult to assess the quality and transferability of the data on the completion time and proportion of items already coded. This study aimed to assess the usability and acceptability of NutriDiary and does not contain any data on the validity of the NutriDiary app. However, providing information on the app’s validity and psychometric properties is a critical prerequisite for its effective application. Therefore, a validation study has already been initiated and will give insights into the validity and quality of the assessed nutritional data.

NutriDiary is the first smartphone based WDR app with integrated barcode scanning function for scientific purposes in Germany. The evaluation by experts and laypersons indicated an acceptable completion time, good usability and acceptability on the users’ side, whereby younger experts and laypersons tended to rate the app better than older ones. Future research will give insights into the validity and feasibility of NutriDiary in different study populations.