Abstract
In this article, we use noise as a metaphor for the overload of information – embodied, technological and online social – that characterizes life with type 1 diabetes. Noise illustrates embodied sensations of fluctuating blood glucose, measurement problems and alarms from digital self-care devices and irrelevant or emotionally disturbing posts on Facebook. Attunement is crucial to the quality of self-care achieved by individuals and comprises: (1) developing skills to receive clear signals from the body, (2) adjusting and individualizing self-care technologies to bodies and daily lives and (3) discerning appropriate distracting and unhelpful self-care information. Ideally, life with type 1 diabetes is harmonious, with clear messages from bodies, technologies and Facebook that enable better self-care.
Introduction
On a spring evening in 2016, the first author attended a peer meeting, held once every 2 months, for adults with type 1 diabetes in a large Danish city. Three men and six women had gathered to discuss issues about their illness; some were participating for the first time, others were already acquainted through Facebook or previous meetings. The meeting had been announced through a Facebook community run entirely by adults with type 1 diabetes. With no predetermined theme, the dialogue among participants drifted between topics related to life with diabetes with which they were currently concerned. In the midst of this calm and relatively quiet environment, Mia, a woman in her 50s, introduced the topic of noise. ‘Noise, noise, noise’, she proclaimed before continuing, ‘It is not only the illness that causes noise, it is also Facebook. I just think that there’s a lot of noise, and it’s infuriating’. Mia had an insulin pump and its alarm had just triggered. Mia’s outburst about noise came moments after John, another meeting participant, had to deal with the alarm on his insulin pump. This is noise in its most literal sense, to which one must be attuned in a life with type 1 diabetes.
In truth, Mia’s alarm was somewhat redundant because she had already recognized the symptoms of low blood glucose and acted drinking half a carton of apple juice to raise her blood glucose. Hence, she could turn off the alarm on her insulin pump without further thought. In contrast, John’s insulin pump had been triggered by the fact that his blood glucose was too high, a situation that could not be immediately resolved with a sugary drink. John’s alarm had already been triggered three times during the meeting, and its insistent noise was clearly making him agitated. He suddenly stood up, trying to extract the pump from the pocket of his tight-fitting jeans as the shrill alarm continued. He shouted at it to ‘shut up’, as he turned it off with exaggerated movements. At that point, it was hard not to concur with Mia’s assertion that life with diabetes is a life filled with noise. As we shall emphasize, however, one inevitably encounters various types of noise when living with type 1 diabetes. Although some noises can be muted with an off switch, others do not provide that luxury. Individuals must learn to attune to these noises.
Throughout this article, we attend to aspects of life with type 1 diabetes that are associated with noise. As we have already seen, people with diabetes experience noise both literally, as in the case of device alarms, and figuratively, as a trope that captures something of the lived experience of the disease. Attunement to life with type 1 diabetes requires a range of knowledge, skills and strategies, which are developed through years of living with the illness (Mattingly et al., 2011; Nielsen and Grøn, 2012; Pols, 2010). It is the practice of attunement that we aim to illuminate here.
In Denmark, with a total population of around 5.7 million people, approximately 30,000 people are diagnosed with type 1 diabetes. Self-care for type 1 diabetes is one of the most complicated of all chronic illnesses, due to the complex, highly specialized practices such as daily insulin infusion by insulin pen or insulin pump, blood glucose measurements, carbohydrate intake calculations and changing vials and needles (Coyle et al., 2013; Funnell and Anderson, 2004). Patients conduct these practices alone, without direct support from healthcare professionals. After diagnosis, patients must gradually learn and assume responsibility for these practices, a process requiring the development of some degree of mastery over the complex interactions between bodies, biologies and technologies occurring in the context and course of daily life (Danholt, 2013; Mol and Law, 2004).
Life with type 1 diabetes can then, to borrow a phrase from Susan Sontag, be reasonably described as an ‘onerous citizenship’ (Sontag, 1978: 3); it places considerable cognitive and emotional demands on people in terms of daily self-management and has potentially serious consequences, both short term and long term, with regard to health outcomes. Sontag may also help us to understand responses to diabetes, for as she suggests in her essay ‘Illness and Metaphor’, the less an illness is understood, the more prone it is to the ‘play’ of metaphor. And while it is true that we do understand much about the way in which to treat type 1 diabetes, often with more and more technologically sophisticated means, there nonetheless remains a good deal of mystique about the illness, both with respect to its causality and, more importantly for present purposes, for the way it is experienced by the individuals who must live with the onerous citizenship it bestows.
In our analytical focus on noise in the context of life with type 1 diabetes, we propose a distinction between three types: embodied noise, technological noise and online social noise. For all three, noise evokes the frequently unpleasant experience of managing information of indeterminate relevance from diverse sources, including bodies, self-care technologies and Facebook. Our categorization of noise is thus closely aligned with Serres’ (1982) distinction between biological, informational and social noise. Serres also offers us a way of analysing the resonance of noise and the practice of attunement that is not immediately accessible from the way the notion is mobilized by people with diabetes. Throughout the data we will present here, noise is frequently experienced and portrayed by people with diabetes as that which obfuscates or distorts information and understanding (the signal). Serres’ work problematizes any easy antithesis between noise and signal, for much as he points our attention to the disruptive nature of noise he is at equal pains to highlight the generative outcomes of these disruptions.
Mutual and mutable relations of generative disruption are also captured in Serres’ idea of the parasite. The three types of noise that we identify in our analysis are all, in their respective ways, characterized by the parasitic relationships that they capture. The hosts and parasites that we delineate are mutable, and the system of relations that they constitute are liable to the interruption of noise. Likewise, both host and parasite are equally capable of generating noise. In each case, however, noise is better seen as a catalyst for new understanding rather than an obstacle.
In evoking the idea of noise as interruption in the case of type 1 diabetes, we are also addressing themes previously explored in the pages of this journal (e.g. Shildrick, 2010). It is at the core of the lived experience of type 1 diabetes and its treatment that bodily processes, which are otherwise autonomic and pre-reflective, rise to the surface of consciousness. To use the term coined by Leder (1990), diagnosis with type 1 diabetes can be said to make the pancreas dysappear. This otherwise opaque organ of the body is rendered transparent and immanent and the nature of its treatment demands that it remains so. Skill and mastery of diabetes management practices may serve to alleviate dysappearance, but there is no promise or hope that the pancreas can ever truly return to its previously opaque and transcendent state. The pancreas becomes visible, captured in technologies such as the digitalized insulin pumps that people with diabetes wear close to the body replicating the function of a healthy organ.
The emergence of bodily immanence and transparence is, at the experiential level, often cause for a sense of alienation, a sense that one’s body is no longer recognizable (Sobchack, 2010). Returning to the concept of noise, this sense of dispossession may be evoked in the dictum that in matters of health, people often claim, or are encouraged, to listen to their bodies. This advice clearly addresses the ‘knowing body’, implying a healthy and harmonious body in tune with itself, yet the sounds made by a body with type 1 diabetes are often more discordant. But while these sounds might be described as discordant, it can also be said that few people listen as intently to their bodies as those who live with type 1 diabetes. Without the capacity for endogenous regulation of blood glucose, they are required to listen to their bodies as if, quite literally, their lives depended on it. Yet the signs of dysregulated blood glucose are not always easy to distinguish from signs of everyday life (feeling drowsy or irascible are not necessarily signs of hypoglycaemia – but they might be; sometimes such signs are only experienced sotto voce, first recognized too late before the crescendos of hypo-/hyperglycaemia come crashing into the score and sometimes, in the case of hypoglycaemic unawareness, they don’t appear at all). The chronic, high-maintenance nature of the illness results in a blurred distinction between treatment and life with type 1 diabetes (Guell, 2011; Mol, 2008).
The uncertainty of listening to one’s body when living with type 1 diabetes is precisely why so much of its day-to-day management is technologically mediated. Recent years have born witness to a rapid evolution in diabetes management techniques, including the development of new types of insulin, new modes of insulin delivery and new means of measuring blood glucose. These developments create new opportunities for people with diabetes, but they also present new challenges. For some patients, the resulting uncertainty about their condition and what is required to achieve and maintain effective self-care leads them to seek out peers in online communities. (Kingod et al., 2017; Kingod, 2018).
The fieldwork which this article is based upon was conducted between June 2015 and June 2016. It included a year of observations of interactions within 16 Facebook communities for adults with type 1 diabetes, 12 in-depth interviews and 6 participant observations at offline peer meetings announced through Facebook. In Denmark, the use of Facebook is widespread among people with type 1 diabetes. Recognizing the extensive amount of activity within Facebook communities, individuals were asked during interviews to list the ones they participated in most. Three type 1 diabetes communities on Facebook with different foci (a larger general community on type 1 diabetes, a community on insulin pumps and a community on exercise) were selected for in-depth study on a daily basis, and 13 other communities were observed more occasionally.
Observations were not restricted to interactions in Facebook communities. They also took place during interviews with individuals who navigated between online and offline domains with a smartphone and well-developed digital dexterity. The aim of individual interviews was to illuminate the integration of online and offline sociality. Interviews were structured to understand daily self-care practices with type 1 diabetes and the use of Facebook for peer support and to provide data on individuals’ experiences of using Facebook throughout the day and on how online practices and shared content within communities interfered with daily self-care practices. Interviews were augmented by several informal conversations during and after offline peer meetings, email contact and phone calls with the same individuals to obtain a deeper understanding of daily life with type 1 diabetes (Dalgas, 2016).
Following people with diabetes as they moved seamlessly between online and offline contexts was made somewhat easier by the fact that some participants in the online Facebook groups actively sought to integrate the two.
Embodied Noise
Daily life with type 1 diabetes comprises various noises. While their origins and nature may be diverse, a universal attribute is that they cannot be ignored; these noises demand the attention of the person who is exposed to them. Embodied noise can be exemplified as an experience of an unregulated blood glucose level that requires action on the individual’s part for stabilization.
Embodied noise relates to internal signals caused by a faulty pancreas; if this noise goes unnoticed, it intensifies. High blood glucose over a prolonged period is a proven risk factor for the onset of diabetes complications, whereas repeated instances of hypoglycaemia can reduce an individual’s awareness of symptoms, a condition known as hypoglycaemic unawareness (Lawton et al., 2014). To avoid the consequences and extremes of hyper- and hypoglycaemia, people with diabetes are advised to maintain their blood sugar levels within the ‘normal range’. Although the immediate consequences of blood glucose outside the ‘normal range’ are less obvious than the consequences of the extremes of ketoacidosis and severe hypoglycaemia, the uncertain nature of the threat is one of the factors that make living with diabetes a challenge that requires constant vigilance.
Andy exemplifies this vigilance. Despite being 60 years old, Andy was new to life with type 1 diabetes, having been diagnosed little more than a year before. He monitored his blood glucose with a blood glucometer and took insulin, which he injected with a purpose-designed insulin pen. Andy used his glucometer less frequently than recommended by health professionals, in part because he was concerned that the constant pricking of his fingers would cause permanent nerve damage. He was acutely aware of the danger of infrequent blood glucose measurements and explained his strategy as based on wanting to cultivate a sense for the signs of fluctuating blood glucose. Far from neglecting his diabetes, Andy’s strategy was one that demanded continual attention to his body. His strategy also required him to be able to identify the symptoms as diabetes-related, which was not always immediately obvious. During a conversation in his home, Andy laid out on the table all the equipment he used for regulating his blood glucose: If I am on the wrong side of 8 (blood glucose number in mmol) then I become tired and drowsy. That means that I am aware of my bodily sensation. I measure before I take a run. I take carbohydrates if I am too low, and I always carry an energy bar in my pocket, when I feel that I can use it.
Embodied noise is exemplified when Andy’s blood glucose fluctuated, which he defined as small sensations that would turn into severe alarms without action. Problems arose when it became difficult for him to register the fluctuations. Throughout the day, Andy needed to process data from his body and from numbers, displayed by his self-care technology, because he relied on the technology when his senses failed him. He hoped he could become less reliant on measuring with more practice of attuning to his body. He explained that this was not easy but could be mastered by practicing attention, which was a strategy for him to understand his body and reduce the levels of experienced embodied noise. Andy tried to tune into tiny messages from within that could easily be interrupted by other sensations from his body, like when he was certain that his blood glucose level was low due to the sensation of tiny sweat pearls on his forehead only to realize from his glucometer that the blood glucose was way too high. He had developed a table, inspired by his peers in Facebook communities, in which he noted the hour of the day, his measured blood glucose level, which he compared with his sensed one, carbohydrates and exercise. He used this table as a tool to train attunement to his physical condition. Ingold (1993, 2011) describes attunement as knowledge and skills developed by a practice of bodily fine-tuning and attention. It requires strengthening the senses of watching, listening and feeling (Ingold, 1993). Andy had to sort all data from his body to take timely action on a fluctuating blood glucose level. Returning to the framework of Serres, it seems initially obvious to cast Andy as the host and type 1 diabetes as the parasite, with the latter being the sole source of noise in the relation. Yet this would be to overlook the part that Andy plays in generating embodied noise and on how he seeks to rely on his diabetes to provide him with signals to act upon. Here the question of host and parasite in the relation between the two appears more mutable.
The sensory experience of the signal of a blood glucose fluctuation varied among informants. With smaller fluctuations, it can be sensed and interpreted when it is a signal and an initial warning of a body out of balance, however if it turns into severe embodied noise it becomes difficult to interpret and to act upon in a timely fashion as it would affect physiology and cognition. Some could sense a fluctuation close to normal levels (defined by blood glucose levels of 4–8 mmol); others first sensed fluctuations when several bodily alarming sensations such as inner restlessness or agitation indicated a number that was too low or too high. At times, these fluctuations were visible to outsiders through physiological reactions such as sweating, shaking or talking nonsense, which could only be reversed by consuming carbohydrates to raise blood glucose levels. It was not uncommon for inner alarms to go off several times a week or even many times a day. This had a potentially interrupting effect, as individuals with type 1 diabetes were required to momentarily withdraw from their social lives to undertake self-care practices to either raise or lower levels. Tanja, a 41-year old, who as diagnosed with type 1 diabetes at the age of 21 provided an example of this: At times I am forced to take a break, in order to ‘pull out the plug’. This could be a situation where I have to get some sugar, because I suddenly sense that I am rambling. Even though I know that there are people nearby to assist me if necessary, I always carry some sugar around.
When Andy practiced listening to or sensing his body, it gave him a sense of control over what happened inside him, which is also why he preferred using an insulin pen rather than an insulin pump. An insulin pump is attached to the body and provides continuous small amounts of insulin, replicating the actions of a healthy pancreas. With an insulin pump, Andy feared that he might get lazy and lose the skill of listening to his body. The bodily signs, when processed correctly, guided him in regulating his blood glucose.
Technological Noise
Using an insulin pen requires a continuous attunement to the body cross-referenced with visual comparisons to the number revealed by the glucometer. In contrast, using insulin pump and automated glucose monitors requires continuous attunement to the technology because they must be continuously adapted and adjusted to the body they serve (Winance, 2010). Here noise becomes extended from the body to the technology and therefore the strategies of attunement from using a pen to a pump differ.
Newly diagnosed individuals are often more sensitive to fluctuations in blood glucose, whereas a potential side effect of the illness over time is a decreased capacity to recognize these signs, a risk, in other words, that the body is silenced (DeVries et al., 2004). This was evident within Facebook communities, in which individuals with long-standing type 1 diabetes searched for ways to regain sensitivity towards the signs of hypoglycaemia. When sensitivity to low blood glucose levels deteriorates, known clinically as ‘hypoglycaemia unawareness’, an insulin pump becomes potentially advantageous (Lawton et al., 2014). Lisa had been living with type 1 diabetes for more than 40 years, diagnosed at the age of 8. She had difficulty sensing her blood glucose levels, so the visible number and alarms on her insulin pump acted as a compass, steering her self-care in the right direction. She downloaded the data on her insulin pump and analysed it in conjunction with her healthcare practitioner, which allowed her to adjust the pump to her needs. Her pump became what Koksvik (2016) has defined as a communicator on behalf of the patient. Koksvik describes fluid relationships between human (patients) and non-human (life-saving machines) actors in which the machine communicates for the patient and the patient also communicates through the machine. This is what Akrich and Pasveer (2004) describe in terms of scenario, where competences between artefact and user are shared. Even though Lisa appreciated the machine, she often felt that it was not properly adjusted to her body. In a 2008 study, Mol describes how technologies, daily habits and people’s skills require mutual adjustment and that the human and the technology must be attuned to one another (Mol, 2008). Here again we also see echoes of the essential intersubjectivity of the parasitic relationship.
Lisa was very interested in finding out how to ‘hack’ and reconfigure her device to make self-care easier, including changing the pump’s annoying alarm to a sound from her smartphone. Changing technology to fit individual needs and lifestyles has been stressed in several studies, from self-tracking technologies (Fox, 2017; Lupton, 2016) to negotiating and fitting technologies to bodies and daily lives (Kingod, 2018; Pols, 2012; Winance, 2010). Lisa reflected on how receiving an alarm from her smartphone rather than her insulin pump was less ‘noisy’ because she found the alarm on the pump very disturbing. She had recently considered joining a newly established community on Facebook for practical guidance and information about how to hack her insulin pump and connect it with a smartphone. Membership of this community was growing, and it offered offline meetings with unauthorized tutorials on how to change the technology. However, hacking her insulin pump would only succeed if the accompanying continuous glucose monitor maintained its signal to her insulin pump. Lisa had many examples of how the insulin pump misguided her when the continuous glucose monitor measured her blood glucose level incorrectly, resulting in loud false alarms throughout the day. This technological noise was common among the observed individuals due to difficulty attaching the monitor properly to the body. Lisa explained how these alarms could sound any time of day, sometimes in very inconvenient situations. It is so frustrating when the alarm goes off several times, and I have to get it from underneath my shirt. And it just keeps going off. Stupid pump!!
Even though Lisa had been living most of her life with type 1 diabetes and Andy had been diagnosed just the year before, they both regularly felt lost about how to self-care in specific situations. In general, adults with type 1 diabetes have approximately four annual visits with healthcare professionals, primarily dedicated to numbers-based routinized biomedical practices (measuring blood glucose and blood pressure, weight, vitamin and mineral levels and eyesight). These visits rarely provide knowledge about how to manage type 1 diabetes in various situations that occur in daily life. The need for comprehensive daily self-care practices drove informants to Facebook to find guidance and support about how to silence unwanted illness-associated noise.
Online Social Noise
The informants in this study were members of several Facebook communities and interacted in them to varying degrees. Enrolling in Facebook communities was a strategy to solicit guidance and support, to address doubt and decrease an unwanted focus on the illness. The amount of data from these communities was, however, often overwhelming, causing the opposite effect. When this occurred, Facebook became another device to operate and another source of noise.
When confronted with instances of the embodied and technological noise described here, informants often approached peers on Facebook with questions. However, noise also appeared online. An overload of irrelevant or emotionally disturbing information could occur, especially among people not used to navigating Facebook, who lacked familiarity with the search functions provided by the platform. Selecting particular communities targeting specific aspects of individual self-care was a strategy to reduce the noise caused by information overload.
To use this strategy, individuals first had to find the right communities for their specific self-care needs and regimens; for example, insulin pump users posted most actively in communities targeting their specific brand of insulin pump. The multiplicity of communities reflected various methods of self-caring for type 1 diabetes. Conversely, even though the observed individuals were selective about the communities in which they participated, they still often felt flooded with information. Continual smartphone notifications about community activity increased the focus on illness to the point where informants felt a great deal of time was required to sort out wheat from chaff in terms of their individual information needs.
Although they experienced Facebook as a quick and easy way to receive essential experiential knowledge on how to live with illness, study informants needed to understand the site’s platform, options and communities’ netiquette before Facebook could be truly supportive. Observations of their interactions with the technology revealed that the informants both created and received noise. This was evident when information and questions were posted in the wrong communities or when individuals did not understand platform functions. For example, when questions about insulin pens were posted in the insulin pump communities, members of those communities directed these posts to the larger and more general communities. Irrelevant information sometimes resulted when individuals were unaware of Facebook functions or how to use them within the communities. For example, they often created a post about a question or concern without either using the search function or scrolling through previous content to see whether others had already raised the issue. This type of online navigation generated unnecessary confusion and frustration and caused an unwanted awareness on the illness, here exemplified by the informant Susan: It is the proportion. It is like you get the same question 10,000 times. It is usually something like, ‘I am high in the morning’…then some of the old ones writes back, ‘try to use the search field’. (Susan, living with type 1 diabetes for 24 years) The fact that the same information is posted all the time makes the threads unnecessarily long. There is the risk that concrete and useful information is drowned out. I have often looked for answers that I almost couldn’t find because of all the ‘noise’. I wish that people would just browse through the information…I just wish for shorter and clearer threads. (A woman who created a post that received 57 replies) A thread tends to develop and then so much irrelevant information is posted. You know, just to give an example, of a community on food and type 1 diabetes. I am vegan and there are not a lot of vegans. Then I ask, is this also a vegan group? And then I get 7000 replies (exaggerating). The first replies are typically a ‘yes’ or a ‘no’ and then it develops into ‘I don’t like this’ or ‘I eat this and this’…In this way there is suddenly so much information about something completely different. And it was not what I asked about initially. And then I just sit there and I am tempted to scream out: ‘JUST READ MY QUESTION!’ (Anna 36 years old, and living with type 1 diabetes since she was 14 year old) It is all about being able to sort out, so you won’t drown. If you have to relate to everything, then you get stressed. In some groups, however not these ones, you can only like the content, and then you have to search for your information. But in our groups, we share everything and then you have to be able to filter/sort out the posts.
Another area identified with online social noise related to posts with an emotionally disturbing effect. These posts primarily included information about late complications such as nerve damage, retinopathy, kidney disease or foot problems as side effects of high blood glucose over time (Steffens and Anderson, 2013). This finding is consistent with a study among women with breast cancer who withdrew from an online community to avoid painful and anxiety-provoking information (Sandaunet, 2008). Studies have shown a tendency for participants to manage noise by seeking affirming information online, fostering hope instead of fear (Kaufman and Whitehead, 2016; Radin, 2006; van Uden-Kraan et al., 2008). The Facebook communities observed in this study often identified desirable content in introductory comments about community netiquette on the welcome page. Some communities emphasized strong emotional support and free sharing of difficult stories about illness; others aimed to empower individuals through a decreased focus on the negative aspects of the illness and an increased focus on empowering messages, borrowing from positive psychology. Lisa explained how she established a more empowering community in response to a general community in which she was exposed to stories of late complications, often combined with uploaded photos of foot wounds. This made it more difficult for her to cope with her illness: I think that most people with type 1 diabetes want it to take up as little space as possible. It is like now when we talk, I don’t want to only sit in a diabetes energy. Because it is very difficult to talk about all the time, I can tell you. I would break down. It becomes too much diabetes…I like to have access to information, but not to the posts that are disempowering. (Lisa)
Attuning to Noise and Life with Type 1 Diabetes
The metaphor of noise is pervasive and certainly captures something of the mystical or dangerous aspects of diabetes as an illness and how people struggle with self-care in the 21st century (cf. Sontag, 1978). The metaphor is, however, firmly rooted in the literal experience of diabetes, not least because the technology designed to support and enhance self-management of the condition nonetheless manages to recapitulate its intransigent unpredictability. But as Serres suggests, the notion that ‘sickness is noise’ is ultimately not a metaphor: Sickness of whatever variety, intercepts a function; it is a noise that mixes up messages in the circuits of the organism, parasiting their ordinary circulation. (Serres, 1982: 197).
Whenever there is a channel of communication (be it biological, informational or social), there is room for noise (Serres, 1982: 79). Yet while noise may be said to have generally negative connotations, it is not something which can be ignored and in certain respects people with diabetes are compelled towards the noises that they encounter. Indeed, when informants could process data from bodies, technologies and Facebook to receive a clear signal, noise changed into useful information. Although type 1 diabetes causes unwanted noise, being able to understand noise as a message can also contribute to self-management when individuals can ‘tune in’ to what is taking place inside the body and then apply correct practical knowledge to deal with it. The process of attunement was adapted to bodies, technologies and Facebook, which required continuous development of skills and effective practices. Even though processes and practices of attunement were seen to be individualized and situated, they had generalizable elements that could be shared among peers on Facebook.
We can use the concept of noise as dysappearance in daily life with type 1 diabetes to critically think and rethink how the technological advancements of the 21st century are reshaping self-care and daily life with type 1 diabetes. Allowing for the fact that some noise is an inherent aspect of communication and that communication is a prerequisite for relations, our data point to developments in the way people live with type 1 diabetes in which there is an expansion of relational channels for noise to filter through. The challenge for people with type 1 diabetes and those that support them, be it professionally or personally, is to embrace the opportunities allowed by such new relations while at the same time evolving and refining their ability to discern what is and what is not noise.
Footnotes
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Danish Innovation Fund and Diabetes Management and Steno Diabetes Center Copenhagen. As a member of the ‘The Vitality of Disease’ project (VITAL), the preparation of this article was supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant Agreement No. ERC-2014-STG-639275).
