Information literacy in food and activity tracking among three communities: parkrunners, people with type 2 diabetes and people with IBS

Background: Tracking food intake and physical activity are increasing and there is evidence of links to improvement in health and well-being as a result. Crucial to the effective and safe Objective : To analyse food and activity tracking from an information literacy perspective. Methods: An online survey was distributed to three communities via parkrun, diabetes.co.uk and the IBS Network. Results: The data showed that there were clear differences in the logging practices that members of the three communities engaged with, and differences in motivations for tracking and extent of sharing of tracked data. Respondents showed a good understanding of the importance of information accuracy, and were confident in their abilities to understand tracked data. There were differences in the extent to which food and activity data was shared, and a lack of understanding of the potential re-use and sharing of data by third parties. Conclusions: Information literacy in this context involves developing awareness of the issues of accurate information recording, and how tracked information can be applied to support specific health goals. Developing awareness of how and when to share data, and of data ownership and privacy are also important aspects of information literacy.


Introduction
Selfcollect information about themselves, which they then review and consider applying to the [1]. While manual recording of personal data has been advocated for many years, the potential of digital devices and apps to monitor and measure the self and share that information with others has huge potential for improving health [2]. Mobile phones are ubiquitous, powerful, connected devices that are highly valued by users, and have the potential to support healthy behaviours through their in-built sensors and downloadable apps [3,4]. Smartphone penetration is high, with 83% of people in the UK owning one [5], and 25% of app users regularly using a health or fitness app [6]. The act of tracking has been shown to be beneficial in terms of increasing desired behaviours in the health arena [7,8], and research has focused on the value of apps and technologies to support health goals in a variety of contexts e.g. menstrual tracking [9]; management of migraine [10]; management of diet and exercise [11,12] and management of chronic disease [13,14].
Lifestyle changes supported by self-management of people with non-communicable diseases are a key factor in their prevention and treatment [15]. It is accepted that to achieve health goals involving weight loss, people must address both diet, in terms of reducing calorific intake, and also increase physical exercise [16]. People who track both diet and physical activity are more likely to lose weight [11,17]. Wearable devices that automatically track physical activity, such as FitBit, are increasingly popular: the market research organisation Mintel estimate that 38% of UK consumers have an interest in wearable technology to monitor health and fitness [18].
However, there are a number of barriers to the effective and safe use of tracking, for example whether tracked data is sufficiently accurate to be valued by health professionals, ease of use of apps and the information and digital literacies required to use them effectively, the threat to personal privacy from re-use of tracked data shared to third parties and developing understanding of the social norms of tracking and of sharing tracked data [3,13,19]. Of particular interest to this paper is the way that levels of information literacy might be one important determinant of effective and safe use of tracking. Information literacy can be T about any information we find and use. It empowers us as citizens to develop informed [20]. Initially, research in information literacy focused on the educational context, but increasingly it has broadened to developing understandings of information literacy across a range of everyday life [21], workplace [22] and health contexts [23]. Information literacy is highly contextual with the set of skills, abilities and practices that are regarded as competencies varying enormously depending on the setting [24]. Previous research into the information literacy aspects of diet and fitness tracking implies skills in a number of inter-related areas [25]: 1) Understanding the importance of quality in data inputs; 2) Ability to interpret tracking information outputs in the context of the limitations of the technology; 3) Awareness of data privacy and ownership; 4) Appropriate management of information sharing.
In order to investigate the role of information literacy in the safe and effective use of tracking in a range of contexts, we selected three contrasting populations to study: participants in parkrun free running events; people with type 2 diabetes; and people with Irritable Bowel Syndrome (IBS). The populations were identified during a previous study as being inclined to want to support their health through tracking [25]; and were selected to capture variations of underlying motivation and need for tracking and variations in tracking behaviours. An investigation across the three groups offers insights into the diversity of tracking practices.
The research questions for the study were: 1. What do people in the three communities track and why?
2. What barriers to effective and safe use do they encounter, particularly in relation to information literacy?
This study is the first to investigate self-tracking for health and wellbeing in these three specific communities, and offers a novel comparative perspective on the attitudes and behaviours of people with regard to supporting health. Framing tracking behaviours within an information literacy perspective s of competence in using information to meet their goals, and this contributes to an increased understanding of the way people engage with information in the health arena.

Information literacy and health
There is increasing interest in the contextual nature of information literacy which, in the health field, is often [26]. The interest in the relationship between information and health is driven by the increasing demand for such information among the population, and the changing nature of the relationship between people and healthcare providers [27]. Self-tracking could also be understood as a response to a growing perception of individual responsibility for health [28]. Phenomenographic studies have demonstrated substantial variation in conceptions of health information literacy; which can mean striving for or reaffirming wellness, knowing or protecting oneself; screening, storing, or creating knowledge; using information to choose a treatment path; paying attention to the body; or participating in learning communities [26,27]. This variation underlines the complexity of both the concept of health information literacy, and in the multiple and distinctive ways in which people engage with and use health information in their lives. Health literacy has been defined specifically within an electronic context: model which focuses on abilities to interact with technology, other users and apply the information for improved health [29].
Understanding how people engage with and use health information, and develop their information literacy is of interest to public health bodies as they attempt to design health ehaviours [30]. The view that health information literacy is an example of a contextual application of information literacy is adopted in this paper [24].

Food and activity tracking
Research has shown that use of apps can motivate people to adopt healthy behaviours, including a healthy diet, increased physical activity and weight loss [11,31]. Selfmanagement of diet is seen to be a critical issue in some chronic disease management [32], and it has been found that mobile apps for dietary assessment are as valid and reliable as more traditional methods of food diarying [33]. The MyFitnessPal app, popular with both health professionals and the general public, has been found to promote positive changes to lifestyles of people suffering from diabetes [16]. Tracking can give people a sense that they are taking control of aspects of their life, that they are developing enhanced self-knowledge and self-management, and that they have improved understanding of their own bodies [1,34]. ' al tracking: d to simply record bodily information; diagnostic tracking to link different aspects of behaviour; collecting rewards as a way to register achievement and fetishized tracking characterised by an interest in gadgets and technology [35]. People gain enjoyment from setting and achieving personal goals from tracked data [1]. Use of multiple devices is common among those who actively engage in tracking behaviour [36].
However, there are a number of potential issues associated with tracking practices. Tracking can radically alter eating practices, and can de-pleasure food [1], and tracking can remind people of the negative aspects of a chronic disease [13]. There are concerns that users may fetishize data and develop unhealthy obsessions [1,12,13]. Apps tend, on the whole, not to be based on any behaviour change theory [37,38]. People can find the apps very timeconsuming to use, leading to a culture of temporary use, particularly if apps do not meet expectations [3,32,39]. In addition to these issues, there are a number revolving specifically around the information literacies required to make effective and safe use of tracking.
Accuracy of data input in tracking is important, but people recognise that their own recording practices may not be sufficiently diligent [39]; and people who use apps should have concerns around their ability to enter information accurately and avoid issues of selfdeception [3]. Understanding quality in data input is one key aspect of information literacy in tracking.
Equally, tracking devices are not necessarily scientifically reliable. They remain unregulated and there has been considerable speculation about their accuracy [40], and the extent to which valuable bodily data cannot be recorded with apps and devices [12]. So an information literate individual would be aware of these issues and either find ways to take them into account or not use them at all. Yang et al. [41] investigated how people themselves attempted to test accuracy, though folk approaches to doing this were often flawed. Furthermore, the outputs of apps are not necessarily understandable by those who use them: tracking demands the ability to interpret information outputs [19].
The extent to which people are aware of issues to do with the privacy of their personal data held in mobile apps or shared online is also an aspect of information literacy. Although market research in the UK has shown a majority of app users express concern about privacy and the extent to which apps share information about them, they are not always wary of using a social media account to access app functions, indicating a lack of awareness about potential re-use of data [6]. This parallels what has been dubbed the privacy paradox in social media: that people are concerned about privacy but do risky things anyway [42]. This could be because they are not sure how to protect themselves, because they are not fully aware of the risk, or because of cynicism about having any privacy in a connected world.
Further, research has found that many apps lack a data privacy statement, and often share data with third party organisations [43]. There have been several high profile data breaches of consumer data including, in 2018, 150 million users of the popular MyFitnessPal app [44].
A US study found that users were confident that apps kept their personal data secure [38], but other studies have found that users do have concerns about the privacy of their health data, particularly if data was sold to third parties [3,39]. A further area of concern is long term access to data, whether because of the disappearance of platforms or the difficulty of exporting data when moving between devices. Thus, issues around data privacy and ownership constitute another area of information literacy relevant to tracking.
There are also aspects of information literacy bound up with appropriate data sharing.
Research has shown that people are much more comfortable sharing activity and exercise data than they are sharing food and diet data [3,11,25]. Some studies have found positive perceptions of sharing exercise data e.g. people can enjoy a competitive relationship with friends and family relating to physical activity [11]. Digital health communities, where people share tracked data, have been identified as a motivating factor in increasing exercise [45], and it is possible to gain intimacy and social support through sharing, to benefit from crowd-sourced expertise, and to learn from others who have the same chronic condition [1,34]. However, other studies have shown that there are sensitivities to do with sharing tracked health data online, with some people considering sharing some health data (including diet information) completely unacceptable [3,25]. Unwanted automatic sharing of data with friends is a reason why people discontinue app use [38].
There seems to be a problematic relationship between people, their tracked health data and health professionals, and relatively few people report sharing data with a healthcare provider [36]. Yet mobile apps that record diet have been identified as potentially useful, particularly for dietetic professionals [16], and people can see value in being able to provide accurate data to health professionals [10,38]. In particular, people with IBS are often advised to keep a diary of their symptoms and diet in order to share with a doctor [46,47].
However, healthcare providers often regard self-tracked data as unreliable, partly due to lack of diligence on the part of the patient, and their supposed unwillingness to admit to negative data [13]. There is also a perception among healthcare professionals that using apps in the context of managing a specific health problem could cause people to undertake inappropriate or dangerous behaviours [19], or promote obsessive or compulsive behaviours [13]. Patients feel that healthcare professionals are dismissive of their ability to collect accurate data or to know their own bodies [34].
In summary, the literature review identified that the adoption and use of tracking behaviours and technologies requires people to develop information literacy, both to understand the collection and interpretation of their own data, but also the social constraints around the sharing of that data. Understanding potential issues around privacy and security of data is also an aspect of information literacy in this context.

Study populations
The selection of these three specific communities is based on findings from a previous qualitative study [25], which highlighted how users with IBS and type 2 diabetes could benefit from food and activity tracking. In addition, the study identified a difference in tracking behaviours between diet and fitness tracking. Therefore, the present study aims to explore in more detail how very diverse groups of users make use of food and activity tracking functions.
Founded in the UK in 2004, parkrun is a not-for-profit organisation that organises weekly timed 5 Kilometre runs in public spaces [51,52]. Events are free to enter and organised by volunteers, and p s ethos emphasises inclusivity. Most participants were not regular runners before registering for parkrun. Evidence suggests that running has positive impacts on physical health and well-being, so mobilising an inclusive community around running has significant potential public health benefits [53].
Type 2 diabetes is a lifelong condition which occurs when the human body cannot use insulin effectively and blood glucose (sugar) levels rise to higher than normal values [54].
Even though type 2 diabetes is mostly diagnosed in adults, it can develop from a young age and can be controlled if treated properly in its early stages by adopting healthy habits and lifestyle such as exercising regularly, maintaining a normal weight and following a lowcarbohydrate diet [55]. Type 2 diabetes, if not managed correctly, can lead to additional health complications, such as heart disease, stroke, blindness, kidney failure and foot or leg amputations [56].
IBS has been functional bowel disorder characterized by symptoms of abdom [46] It is not understood as a single disease, but instead as a range of physiological factors that contribute to commonly experienced symptoms [46]. The cause is unknown, but it is strongly linked to diet and stress, and diet changes are recommended as a way to control symptoms which can vary enormously from person to person [47,57]. One commentator has estimated that around 11% of the global population has IBS [58]. Those self-identifying as suffering with IBS are predominately women [59].

Data analysis
All numerical data were entered in IBM SPSS version 24 and analysed using descriptive statistics. The results of the 5-point Likert scale statements were aggregated to produce overall . In, addition, Independent Samples ttests were performed to identify potential differences in attitudes between men and women and depending on the level of education of the respondents.
The qualitative responses were manually coded independently by two members of the research team using thematic analysis [60], and the central themes surfaced for discussion alongside the quantitative data.

Results
The demographic data (questionnaire section A) of the three participant groups are reported in table 1 below. A summary of the reported app usage, section B of the questionnaire, is reported in figure 1, below, which presents the apps used, reasons for tracking and who data is shared with. below. Table 2 reports on frequency of tracking behaviours in the three respondent groups, and figure 2 presents opinions and view of respondents of their own tracking behaviours. Diabetes respondents displayed a technical knowledge of their condition and the factors that they could log in order to manage it:

It is the main cause [that] my HCA1b is now in the 34 area which is normal non-diabetic level, arb intake around 280 grams a day. (Diabetes)
Logging provided an element of control over the condition: The process of logging helps me stay focussed. (Diabetes)

IBS
Although IBS is a condition that often involves sensitivity towards certain foods, surprisingly few (8) IBS respondents (17.8% use every day) were current users of food logging apps.
However, over a third of respondents (35.6%, n=16) had used one in the past, indicating that logging food could be valuable, but possibly only over the short term to identify trigger foods:

Great to start but cumbersome, especially if you have to log each ingredient every time. I tend to get bored and apps stop getting used. (IBS)
IBS respondents were concerned about accuracy in data entry in common with the other groups, and the qualitative comments revealed a particular focus on perceived inaccuracies in food logging apps could make the practice pointless: Often a discourse of addiction or obsession was used in relation to tracking, but more often with food logging, and it was perceived as more dangerous with food than activity. Thus one person commented satirically on their obsession with recording running data: The demanding requirement to gather accurate data throughout the day could be seen as creating this obsessive element. Thus part of the information literacy of food tracking could be the management of risk around becoming obsessed with collecting data in a counterproductive way. Food tracking seemed often to be undertaken for short periods, probably for this reason. In contrast, comments on activity/running tracking often emphasised long term practice, and enjoyment, because it was motivating, because of the online community element and because it was easy to do:

Run logging is fun and easy. (parkrun)
Where they did persevere with food logging, a number of solutions to data quality issues had been developed by participants: 1. becoming a data creator, e.g. entering information from recipes into the app, 2. being particular about weighing food, 3. modifying interpretation of the results to take account of perceived inaccuracies. Participants were asked about the extent to which they were concerned about how service providers used their logged data. Parkrunners were most likely to be unconcerned (44.8%, n=64 disagree). Diabetes respondents were more worried about re-use of their data, with 24.3% n=34 agreeing strongly (40% n=56 agree overall). The most common response from IBS were evenly distributed /agree strongly and neutral with (37.8% n=17) choosing these options. For each group therefore less than half of the respondents had concerns about potential re-use of their data.
Respondents were also relatively unconcerned about threats to the long term access to their data. 62  . This probably reflects that rather than activity data, they were collecting data about a medical condition or weight and diet, which was seen as more private. Sharing different types of data reflects an awareness of social norms surrounding tracked data.
A few strategies were mentioned as part of maintaining privacy, such as manual data tracking:

I strongly disagree with 'cloud' based apps where I can't restrict data
sharing. That's both for privacy, and also risk of losing access. (Diabetes) Similar sorts of sensitivities were reflected in who data was shared with. Partners were the most popular people to share data with across diabetes (41.4%, n=58) and IBS (33.3%, n=14) respondents, but friends were the most popular for parkrunners (55.9%, n=80). Diabetes respondents were the most likely to share data with a health practitioner, but the numbers were still quite low (26.4%, n=37). Less than 10% of the other two groups shared their data with an expert such as a trainer or doctor.
In summary, the results present the different flavours of tracking practice among the three communities studied, and demonstrate that there are significant differences in motivations for tracking and the uses to which the data are put. People who engage in self-tracking show evidence of information literacy through ensuring data quality; understanding the information produced by tracking technologies and how this relates to their particular situation or medical condition; developing awareness of when and how to share their data; and developing understanding how who has access to their data and the potential for sharing and re-use without their explicit consent.

Discussion
Tracking is used in different ways by different groups, but in all contexts it is an information intense activity, based on gathering, interpreting and managing data mediated by various devices and apps. The question of how information literate trackers are -how good is their critical understanding of the information they are using -thus becomes central to effective and safe tracking. This is one of the first papers to bring this perspective on tracking explicitly to the fore, and complements research that has examined self-tracking from a Human Computer Interaction perspective [12,62,63], a health-behaviour change perspective [3] and a sociological perspective [13].
Respondents showed an understanding of the importance of their own accurate data entry, but also a sceptical awareness of its limits, especially in the context of food logging. In some cases, it was this critical understanding that led to non-use; in others, people found approaches to ensuring data quality or only used it intermittently. This is consistent with previous studies that show that simplifying diet and nutrition apps to make data entry less time consuming and more automatic was a key improvement desired by users [38]. It would make food logging much easier and also remove one aspect that created a fear of obsession, which is a common issue identified in self-tracking research [1,13].
While data accuracy is an important aspect of successful tracking, previous research into self-tracking has highlighted a tension between trusting data, or trusting bodily sensations [12], with speculation regarding the relationship and potential value of each. In Information Literacy research, the role of corporeal information as a valuable source of information alongside social and epistemic or formal sources of information is widely understood [22,25,64]. Conceptions of health information literacy indicate that assessing and evaluating information is a key activity, and that paying attention to the body, and developing selfawareness, support the interpretation of other health information [26]. Diabetes respondents actively used information, often manually recorded, to manage their condition, which could be seen as an example of diagnostic tracking [35]. This extends conceptions of self-tracking broader understanding of their relationship with information [63]. Previous research into the information behaviour of people with type 2 diabetes found that connecting together information gathered from different objective and subjective bodily observations was an aspect of effectively managing the condition [34]. Integrating bodily information with app related information has also been shown to be an important aspect of elite runner personal informatics practice [12]. Becoming information literate with regard to selftracking, therefore, involves developing understanding of how to integrate app data with corporeal information in order to achieve specific health goals.
Although the app MyFitnessPal was popular with participants in this study, perceived , were also barriers to food logging. This is exacerbated by the acknowledged US bias of the food and measurements in [16]. Many people therefore log food for only as long as it takes either to learn better food habits, or to learn which foods trigger aspects of their condition.
Discontinuing use can also occur due to the burdensome nature of tracking [38].
Information literacy therefore revolves around learning at what point the information needs have been met, and when to modify or discard the logging practice.
As regards interpreting information from tracking, respondents were confident in their own information literacy in interpreting the data output by the apps, and often used multiple devices in rather sophisticated tracking practices. They also made some critical comments on the questionable assumptions or expectations built into app design. This is consistent with previous research that has also found that users can be very capable of taking critical stance towards apps [65]. An area of rather more concern, consistent with some previous studies [40] is that many respondents were not worried about the use of the data by the platform, or about long term access to their data, particularly in the current climate where app data is widely shared with 3 rd parties without the express consent of the user [1,43].
Fortunately, some studies have found that data reuse is a serious issue for participants [3], Opinions may also be shifting because of recent cases in the news.
Data sharing with friends and even online communities was found to be central to activity tracking for many participants, and this is consistent with previous research that has found sharing to be a fun aspect of tracking [12]. However, participants in this study were more reluctant to share data about health conditions, diet and weight, a point also identified by previous authors [3]. Information literacy research has found that sharing information about a chronic disease is a way for people to draw friends and family into their landscape and create a narrative about a disease [24], but this does not seem to be the case for participants in this study. Consistent with previous research [36,38], despite the potential benefits, data was not often being shared with a trainer or doctor, but this might reflect lack of interest by practitioners rather than willingness to share data, since they were already sharing with others, such as partners.
In summary, from an information literacy perspective, users seemed to be information literate in many aspects of logging practice. The relatively low use of food apps seems to reflect a critical information literate perspective on the effort required to use them, their accuracy and their potentially obsessive effects.

Limitations
This is a small-scale exploratory study, which only begins to identify the information literacy aspects of tracking behaviours for the three participant groups. All three groups of respondents reported a high level of prior education which may not be representative of the populations as a whole. This may reflect a higher use of logging by higher socio economic groups [66]. It is not surprising that more educated users seemed to have more confidence in their information literacy capabilities. In several respects we do not know how well the respondents represent the wider target population, partly because we do not have data about the demographics of the wider population, and respondents were self-selecting. The skew towards women among parkrunners, for example may reflect greater willingness to participate in surveys, rather than the actual proportions in the population [53]. The response rate from IBS sufferers was lower than for the other groups, and so the results for this group should be treated with additional caution.
By definition, respondents to the questionnaire were current or past users of tracking apps.
Many had not used or lapsed from use of particular forms of tracking; but the sample did not include those who had never tracked at all. This places a limitation on the data as a means of understanding of barriers to tracking in all contexts. However, studies of non-users are inherently difficult.
The survey was based on asking participants to self-evaluate some of their information literacy skills, e.g. their ability to understand charts produced by apps. This may differ from actual competence. Over confidence in information literacy is a known phenomenon [67].
However, levels of information literacy were implicit in many of the qualitative comments, which reflected complex, personalised practices of use.

Conclusions and implications
An information literacy perspective is of value because tracking is an information intensive activity, involving the user in entering data, interpreting the information outputs of the device, and managing access to the data. Effective and safe use of tracking depends on information literacy. The study showed that in three very different domains devices were used quite differently, and levels of information literacy were also variable. In terms of understanding data entry quality, interpreting information and appropriate sharing respondents seemed to demonstrate good information literacy. A greater area of concern could be around the awareness of risks around platform use of data and continuity of access. This implies the need for much better public awareness around data ownership, and simplified privacy statements would assist in this. Organisations such as parkrun, Diabetes UK and the IBS network could consider this issue when providing advice and support around using mobile apps to their communities. The GDPR is a move in a favourable direction in increasing protection of . Simple tools to extract data and maintain access to personal tracking data in the long term are also needed. Additionally, there seems to be a gap in the market regarding mobile apps to support both the management of type 2 diabetes and IBS, given the reported manual tracking of one community, and the pattern of app use and non-use of the other.