Diabetologia (2020) 63:1475–1490 https://doi.org/10.1007/s00125-020-05183-8 REVIEW The competitive athlete with type 1 diabetes Michael C. Riddell 1,2 & Sam N. Scott 3,4 & Paul A. Fournier 5 & Sheri R. Colberg 6 & Ian W. Gallen 7 & Othmar Moser 8 & Christoph Stettler 3 & Jane E. Yardley 9,10,11 & Dessi P. Zaharieva 12 & Peter Adolfsson 13,14 & Richard M. Bracken 15 Received: 17 February 2020 / Accepted: 17 April 2020 / Published online: 12 June 2020 # The Author(s) 2020 Abstract Regular exercise is important for health, fitness and longevity in people living with type 1 diabetes, and many individuals seek to train and compete while living with the condition. Muscle, liver and glycogen metabolism can be normal in athletes with diabetes with good overall glucose management, and exercise performance can be facilitated by modifications to insulin dose and nutrition. However, maintaining normal glucose levels during training, travel and competition can be a major challenge for athletes living with type 1 diabetes. Some athletes have low-to-moderate levels of carbohydrate intake during training and rest days but tend to benefit, from both a glucose and performance perspective, from high rates of carbohydrate feeding during long- distance events. This review highlights the unique metabolic responses to various types of exercise in athletes living with type 1 diabetes. Keywords Athlete . Carbohydrate . Competition . Continuous glucose monitoring . Exercise . Glucose . Insulin . Nutrition . Review . Sport . Type 1 diabetes Abbreviations CSII Continuous subcutaneous insulin infusion AGP Ambulatory glucose profile isCSM Intermittently scanned continuous CGM Continuous glucose monitoring glucose monitoring Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00125-020-05183-8) contains a slideset of the figures for download, which is available to authorised users. * Michael C. Riddell 7 Royal Berkshire NHS Foundation Trust Centre for Diabetes and mriddell@yorku.ca Endocrinology, Royal Berkshire Hospital, Reading, UK * Richard M. Bracken 8 Cardiovascular Diabetology Research Group, Division of r.m.bracken@swansea.ac.uk Endocrinology and Diabetology, Department of Internal Medicine, Medical University of Graz, Graz, Austria 1 School of Kinesiology and Health Science, Faculty of Health, 9 Augustana Faculty, University of Alberta, Edmonton, AB, Canada Muscle Health Research Centre and Physical Activity & Chronic 10 Disease Unit, York University, 4700 Keele Street, Toronto, ON M3J Alberta Diabetes Institute, Edmonton, AB, Canada 1P3, Canada 11 Women’s and Children’s Health Research Institute, Edmonton, AB, 2 LMC Diabetes & Endocrinology, Toronto, ON, Canada Canada 12 3 Department of Pediatrics, Stanford University School of Medicine, Department of Diabetes, Endocrinology, Nutritional Medicine and Stanford, CA, USA Metabolism, Bern University Hospital, University of Bern, 13 Bern, Switzerland Department of Pediatrics, The Hospital of Halland, 4 Kungsbacka, Sweden Team Novo Nordisk Professional Cycling Team, Atlanta, GA, USA 14 Institute of Clinical Sciences, Sahlgrenska Academy, University of 5 School of Human Sciences, Division Sport Science, Exercise and Gothenburg, Gothenburg, Sweden Health, University of Western Australia, Crawley, WA, Australia 15 Applied Sport, Technology, Exercise and Medicine Research Centre 6 Human Movement Sciences Department, Old Dominion University, (A-STEM), Swansea University, A111 Engineering East, Fabian Norfolk, VA, USA Way, Crymlyn Burrows, Swansea SA1 8EN, UK 1476 Diabetologia (2020) 63:1475–1490 MDI Multiple daily injections may exist in some individuals with the disease in whom blood rtCGM Real-time continuous glucose monitoring glucose levels are not tightly managed with insulin therapy [5, TIR Time in range 6]. Insulin deprivation and/or sustained hyperglycaemia can impair mitochondrial function, promote mitophagy, lower ATP provision and increase reactive oxygen species production Introduction in muscle, heart, kidney and brain [7]. As we approach the 100-year mark of the discovery of insulin, Carbohydrates During high-intensity exercise, carbohydrate is a people with type 1 diabetes may achieve a near normal life primary fuel source. Glucose stores within liver and skeletal expectancy with an overall high quality of life, but this requires muscle, in the form of glycogen, depend on the size and training tight maintenance of on-target blood glucose levels and good status of the individual and are the body’s primary carbohydrate cardiovascular health [1]. Both of these aspects of diabetes stores. In the average adult male weighing 70 kg, up to 160 g of management are still very challenging for individuals with type glucose can be stored in the liver, while up to 700 g of glucose 1 diabetes, even with access to specialised diabetes care [2]. can be stored in the muscle [8]. A ‘normal’ blood glucose Being regularly active with the disease improves cardiometa- concentration of ~5–7 mmol/l amounts to only ~4–6 g of total bolic health [3] and is associated with increased longevity [4]. blood glucose, depending on the person’s size. In individuals Leading up to the next Summer Olympic Games, numer- without diabetes, intense exercise causes a transient rise in ous athletes with type 1 diabetes will train and compete at the glucose by ~2 mmol/l [9], while prolonged moderate-intensity elite level, with some aspiring to pursue their podium dreams. exercise induces a small and transient drop in glucose by ~2 The day-to-day management of the condition remains oner- mmol/l [10], albeit responses are highly variable. Glucose ous, however, given the monotonous tasks of monitoring production, predominantly by the liver via glycogenolysis and glucose, carbohydrate/macronutrient counting, insulin dosing, gluconeogenesis, as well as oral carbohydrates, help support and managing stress/sick days, particularly while training and normal blood glucose levels [11] (Fig. 2). Individuals with type preparing for competition (Fig. 1). Ongoing research is 1 diabetes can have normal levels of muscle and liver glycogen increasingly focusing on the unique physiology of such content if they are adequately fed, take insulin and have good high-level athletes with type 1 diabetes, while also investigat- glycaemic control (HbA1c <58 mmol/mol [7.5%]) [12, 13]. ing how new insulin analogues and other therapeutic agents/ Hepatic glycogen levels are lowered by poor glycaemic control technologies might improve their glycaemic management. in individuals with type 1 diabetes [14], with only a partial resto- This review highlights the challenges of high-level training ration with short-term improvements in glycaemic control [15]. and competition in athletes with type 1 diabetes and identifies The flux of glucose from liver to muscle during exercise is some of the knowledge gaps that limit our capacity to provide impacted by insulin treatment, which can result in either hypo- evidence-based strategies to optimise their performance. or hyperglycaemia [16]. High insulin levels limit hepatic glucose mobilisation and increase muscle glucose disposal, thereby causing hypoglycaemia. Inadequate insulin levels Energy metabolism cause hyperglycaemia, as glucose production exceeds utilisation [16]. Physical activity at all levels requires the mobilisation of vari- ous fuel sources. To help better understand the unique Lipids Adipose tissue and skeletal muscle lipid stores are plen- responses to exercise in type 1 diabetes, we briefly describe tiful, even in lean individuals. Lipids are used heavily during the main energy systems used during various forms of exer- prolonged exercise, particularly as the activity duration cise in the following sections. Possible alterations in energy increases (Fig. 2). Peak absolute lipid oxidation rates occur metabolism caused by the disease are highlighted. at ~55–60% of maximal aerobic rate in trained individuals [17]. Intensive insulin therapy in type 1 diabetes often ATP and phosphocreatine During skeletal muscle contraction, increases body fat stores and body weight [18], although this energy is provided from ATP, which is immediately effect can be attenuated with dietary restriction [19] and/or resynthesised from phosphocreatine. The limited phosphocrea- endurance training [20]. Lipolytic potential may be elevated tine stores require that ATP resynthesis occurs by catabolising in type 1 diabetes, perhaps because of increased β- other fuel sources (lipid and carbohydrates) for exercise events adrenoceptors on fat cells [21]. However, a high insulin level lasting more than a few seconds (Fig. 2). With insulin therapy, during exercise suppresses lipolysis/fat oxidation, as and in the absence of nephropathy, ATP and phosphocreatine compared with basal insulin concentrations [22] (see below). levels at rest and post exercise appear normal in individuals with type 1 diabetes [5]. However, a slower phosphocreatine Protein Although protein is a major component of lean tissue, recovery time and impaired mitochondrial function/capacity it does not normally contribute significantly to energy Diabetologia (2020) 63:1475–1490 1477 a Hydration Resistance Aerobic/endurance workout workout Computer work, Social Wake up Breakfast Snack Lunch Rest Hypoglycaemia Snack Dinner social media, etc Snack time Bedtime treatment 10 Glucose (mmol/l) 4 Time 06:00 12:00 18:00 23:00 Glucose monitoring Hypo- and hyperglycaemia mitigation strategies (dextrose tabs, correction boluses, temporary basal rates if on CSII, etc.) Prandial or bolus correction insulin Basal insulin Diabetes supplies on hand (glucose meter, lancing device, lancets, CGM, insulin [needles or pump supplies], carbohydrate snack, glucagon, medical identification) b Stress management Hydration Post-event The race/event activities Travel to Pre-race Medical (debriefing, Social Wake up Breakfast Pack race location snack checkup interviews, etc.) Dinner time Snack Bedtime Drink 10 Glucose (mmol/l) 4 Time 06:00 12:00 18:00 23:00 Glucose monitoring Hypo- and hyperglycaemia mitigation strategies (dextrose tabs, correction boluses, temporary Prandial insulin basal rates if on CSII, etc.) Basal insulin Diabetes supplies (syringes, insulin, glucagon, rtCGM/isCGM supplies, glucose meter, ketone meter, pump etc, spare supplies, carbohydrate snacks, etc.) Fig. 1 Example training day (a) and competition day (b) for a competi- with regard to the timing of training and meals, this flexibility is lost on tive athlete with type 1 diabetes. A number of variables need to be consid- competition days due to strict competition schedules. Note that this is an ered and controlled by an athlete with type 1 diabetes, including glucose example and will differ depending on numerous factors such as the event monitoring, basal and bolus insulin-dose modifications, snacks and that the athlete competes in. This figure is available as part of a meals, hypo- and hyperglycaemia mitigation, hydration and stress downloadable slideset management. Although some flexibility may be allowed on training days metabolism. However, some protein-derived amino acids, protein-derived and free amino acids into glucose during exer- such as leucine or alanine, can contribute minimally to skeletal cise is upregulated in type 1 diabetes if insulin is withheld muscle energy needs, especially when carbohydrate availabil- [24]. Insulin deficiency for as little as 8 h in type 1 diabetes, ity is restricted (i.e. by low-carbohydrate diets, periods of perhaps in combination with other factors (hyperglycaemia, insulin deficiency) [23]. The gluconeogenic conversion of elevated cortisol, inflammation, etc.), rapidly promotes 1478 Diabetologia (2020) 63:1475–1490 Fig. 2 Energy substrates for exercise. The source of energy Anaerobic Muscle triacylglycerol substrates during exercise varies glycolysis depending on exercise duration. During skeletal muscle contraction, in the first few Plasma NEFA seconds of exercise, energy is provided from ATP, which is immediately resynthesised from Blood glucose from phosphocreatine (PC). For hepatic and oral sources Aerobic exercise of longer duration, ATP metabolism resynthesis occurs by catabolising other fuel sources (lipids and Muscle glycogen carbohydrates). Figure based on ATP-PC previously published data [102, 10 s 30 s 5 min 1h 2h 3h 4h 5h 103]. This figure is available as Exercise part of a downloadable slideset time protein catabolism, likely via activation of muscle-specific Selecting an insulin delivery method transcription factors [25]. The primary goal of exercise management in athletes with type 1 diabetes should be to limit dysglycaemia, with a secondary goal of attempting to replace insulin to healthy physiological Insulin regulation and dysregulation insulin levels. Complete restoration of insulin to physiological during exercise levels is impossible since insulin is administered subcutaneous- ly rather than released into the portal circulation. While some Insulin mediates glucose disposal into skeletal muscle and athletes with type 1 diabetes perform well using multiple daily adipose tissue via increased glucose transporter type 4 translo- injections (MDI) of insulin [31], others prefer the flexibility cation. In liver, insulin signalling supresses glucose production afforded by continuous subcutaneous insulin infusion (CSII) and activates glycogen synthesis via activation of various [32]. The latter allows for temporary basal rate reductions in enzymes, including glucokinase and glycogen synthase [26]. anticipation of and/or recovery from prolonged aerobic exer- During endurance exercise in individuals without diabetes, cise, temporary basal rate increases for very intensive aerobic/ insulin secretion decreases via increased sympathoadrenal anaerobic work, and for basal rate reductions overnight, if drive, with the magnitude of decline closely linked to activity nocturnal hypoglycaemia is an issue. Hybrid closed-loop tech- intensity and duration [27]. This drop in insulin secretion facil- nology may support glycaemic management in athletes better itates lipid and glucose mobilisation from stores outside of the than traditional pump therapy as insulin delivery is informed by muscle, while minimising the risk for hypoglycaemia as current glucose levels, glucose predictions, previous insulin contraction-mediated glucose disposal increases [16]. With delivery and other features of proprietary algorithms that brief intensive exercise bouts, insulin secretion increases during improve overall ‘time in range’ (TIR; the percentage of time early recovery to offset rising glucose concentrations [9]. that an individual’s blood glucose is within the target level) In individuals with type 1 diabetes, circulating insulin levels [33]. Currently approved hybrid closed-loop devices are suit- depend on the amount and location of insulin administration. able for prolonged aerobic exercise if a temporary (higher) Because insulin levels cannot immediately be lowered at exer- glucose target is set well before the start of exercise (i.e. 45– cise onset, individuals with type 1 diabetes are often 90 min before the exercise start time). hyperinsulinaemic during their activity (Fig. 3). The relative In spite of these benefits, many individuals report that CSII hyperinsulinaemia during prolonged moderate-intensity exer- interferes with their sporting activities or that they would rather cise supresses lipolysis/fat oxidation [22] while increasing not be attached to a medical device [34]. Maintaining insulin whole-body glucose utilisation and hypoglycaemia risk [16]. infusion sets and glucose monitoring devices during exercise Exercise increases absorption rates of some [28], but not all and sport is challenging when there is increased perspiration [29] forms of insulin, which can exacerbate the risk for and the potential for sport contact and/or friction. For athletes hypoglycaemia. With intensive exercise, hyperglycaemia post who prefer pump removal during exercise, a hybrid approach exercise is aggravated by the inability to automatically increase that combines basal insulin delivery split between an ultra-long- insulin delivery into the portal circulation [9]. Omitting insulin acting insulin and 50% reduced basal insulin delivery by CSII altogether, well in advance of exercise, promotes excessive [35] can be used. The addition of continuous glucose monitor- hyperglycaemia and ketone production [30]. ing (CGM) is beneficial, as athletes (particularly those with Diabetologia (2020) 63:1475–1490 1479 1000 900 Rest Post exercise 800 700 Insulin (pmol/l) 600 500 400 300 200 100 0 Non-DM HGI-CHO LGI-CHO Full bolus 75% bolus 50% bolus 50% bolus Full bolus 75% bolus 50% bolus 25% bolus Full basal/50% bolus 80% basal/50% bolus Low insulin High insulin Pump suspend 50% basal rate 20% basal rate Exercise 60 min 45 min 45 min 45 min 45 min 45 min 45 min 45 min 45 min 45 min 45 min 45 min 45 min 45 min 45 min 60 min 60 min 60 min mode run run run run run run run run run run run run run cycle cycle run run run Reference [104] [105] [106] [107] [108] [41] [22] [39] Fig. 3 Circulating insulin levels in physically active individuals with type insulin-dose reductions, along with bolus dose reductions for MDI (full 1 diabetes. The values shown represent the insulin concentration as basal/50% bolus; 80% basal/50% bolus); participants with type 1 diabetes measured before (rest) and soon after the end of exercise in a variety of who had low (15 mU m−2 min−1) or high (50 mU m−2 min−1) intravenous previously published studies, which included various cohorts/conditions: insulin infusions; and type 1 diabetic participants who underwent basal non-diabetic control participants (Non-DM); participants with type 1 insulin rate reductions for CSII (pump suspend; 50% basal rate; 20% diabetes who underwent high- (HGI-CHO) or low- (LGI-CHO) carbohy- basal rate). The mode of exercise and duration of activity is shown on drate feeding interventions; participants with type 1 diabetes who the x-axis. Data are from select studies [22, 39, 41, 104–108] and were underwent bolus insulin-dose reductions (full bolus; 75% bolus; 50% analysed by R. M. Bracken. This figure is available as part of a bolus; 25% bolus); participants with type 1 diabetes who underwent basal downloadable slideset hypoglycaemia unawareness [36]) can gather glucose data, event [42]. For other long-acting basal insulins (e.g. insulin respond to glucose trend arrows and alerts/alarms, and optimise glargine, insulin detemir), the total basal insulin dose can be therapy [37]. Real-time CGM (rtCGM) offers the advantage of divided into a morning and evening dose to allow for more alerts and alarms when glucose drifts away from target; howev- flexible adjustments. As an alternative (or complement) to basal er, exercise itself has an impact on sensor accuracy [38]. insulin-dose reduction, simple carbohydrate consumption (up to 70–90 g/h) during prolonged aerobic activities can help prevent hypoglycaemia and support performance [43]. Strategies to address relative In addition to the inability to lower insulin secretion into hyperinsulinaemia during prolonged exercise the portal circulation at exercise onset, glucagon fails to rise normally during prolonged exercise in type 1 diabetes, predis- Relative hyperinsulinaemia during prolonged aerobic exercise posing athletes to developing hypoglycaemia during some can be offset by basal and/or prandial insulin-dose reductions activities [44]. Administering a mini dose of glucagon [45] and/or by increased carbohydrate feeding. For those using or glucagon in a dual-hormone closed-loop pump [46] helps CSII, basal insulin delivery can be reduced by 50–80% to eliminate hypoglycaemia; however, this has never been 90 min before exercise [39]. Suspending insulin delivery at tested in a setting of competition. exercise onset is safe, albeit less effective in mitigating the drop in blood glucose level [40]. Basal insulin delivery can be resumed immediately post exercise, allowing circulating Strategies to address relative insulin levels to rise before the recovery meal. hypoinsulinaemia post exercise For individuals using MDI, the basal insulin dose can be reduced by 20–50% before exercise to mitigate hypoglycaemia Managing competition-related hyperglycaemia, particularly at risk [41]. Even insulin degludec can be reduced by ~25%, but the start of an event, can be challenging [47]. Psychological strat- this reduction should be initiated 3 days before the exercise egies, such as cognitive restructuring and overlearning of skills, 1480 Diabetologia (2020) 63:1475–1490 may help offset the stress effects [48]. Some athletes will tolerate, and duration will largely dictate the strategies employed for or even plan for, a slightly elevated blood glucose level when active days [16]. In general, prolonged predominantly aerobic starting an event; others may choose to use a temporary basal rate exercise promotes a drop in blood glucose concentration, while increase (if using CSII), a partial bolus insulin correction or a more intensive aerobic and anaerobic events promote a glucose prolonged aerobic warm-up to correct hyperglycaemia. Giving rise [16]. The rise in blood glucose during intensive exercise in a standard (i.e. full) insulin bolus correction before a prolonged the fasted state is reproducible and tends to be associated with a aerobic exercise event is not recommended unless ketones are rise in lactate [53]. For endurance events, such as marathons elevated, since doing so increases hypoglycaemia risk [11]. and road cycling, athletes often have elevated glucose levels Many athletes have difficulty managing immediate post- prior to the event, sometimes because of psycho-physiological event hyperglycaemia [42]. When fasted, a bolus insulin stress responses [54] or as a purposeful coping strategy to limit correction can be given after intense aerobic exercise [49] or the likelihood of developing hypoglycaemia during the event after resistance exercise [50]. However, with most prolonged [55]. Typically, carbohydrate consumption is needed to main- exercise events, late-onset hypoglycaemia remains common tain performance and prevent hypoglycaemia in endurance for athletes [51] and, thus, basal insulin-dose reduction and/or events lasting ≥60 min [16]. More aerobically fit individuals bedtime snack strategies are recommended [16]. may have higher hypoglycaemic risk during exercise than those who are less fit [56], potentially due to higher absolute power outputs and greater rates of carbohydrate oxidation. Planning for glucose management Conversely, having insulin at near basal levels or lower typical- with dynamic training protocols ly causes a rise in glucose during burst events, like pole vault- ing, power lifting, sprinting or wrestling [57]. Athletes partake in varied training regimens, often differing daily or seasonally with regard to training mode, intensity and duration. Professional athletes frequently use ‘polarised’ train- Optimising performance with nutrition ing strategies, starting early in the season with low-intensity, high-volume work, followed by high-intensity, lower volume Several evidence-informed nutritional strategies exist to work later in the season. Before competition, training volume support athletes in various settings [58]. However, for athletes generally tapers. Such training varieties may make glucose with type 1 diabetes, it is unclear if special or additional management challenging for athletes with type 1 diabetes. considerations are required to optimise performance. Like However, by individualising standard recommendations, athletes without diabetes, those with diabetes follow the full athletes can personalise strategies through trial and error to spectrum of carbohydrate intake strategies, depending on their temper glycaemic excursions [47]. activity and training regimens. Even when athletes with type 1 diabetes have well-honed strategies, it is often useful to work with endocrinologists and Carbohydrate intake While some athletes use carbohydrate other healthcare providers [47]. The clinical team should first counting to determine meal- and snack-based insulin-dose review the ambulatory glucose profile (AGP) if rt-CGM or inter- adjustments, this procedure often lacks precision, particularly mittently scanned CGM (isCGM) is used, along with the athlete’s with high-carbohydrate feeding [59]. Moreover, carbohy- current strategies for glucose management around training and drates with differing glycaemic indices and mixed meals make competition. Clinicians should review glucose monitoring down- this practice difficult. If exercise occurs soon after a meal, loads to ensure adequate basal insulin dosing and correct bolus glucose disposal from the meal may be stimulated by both insulin usage to cover meals and hyperglycaemic excursions insulin-dependent and insulin-independent signalling [60]. [52]. The clinical team should offer reasonable initial strategies Although high-glycaemic index meals/snacks generally for athletes who are newly diagnosed with type 1 diabetes, such increase insulin resistance in people without diabetes [61], as the use of self-monitoring of blood glucose, nutrition counsel- carbohydrate loading pre exercise and/or carbohydrate feed- ling, newer insulin analogues and CSII with rtCGM or isCGM, ing during competition with simple sugars is feasible [43] and as appropriate. Various features, such as cost, comfort and accu- likely to be important for performance and glycaemic manage- racy, are considerations for product choice. ment during competition and training. According to self-report, some athletes with type 1 diabetes adopt low or moderate carbohydrate diets to improve Strategies to manage different modalities glycaemic management (Fig. 4a). It is currently unclear if this and durations of exercise dietary approach has an impact on performance. Good long- term glucose management improves performance in athletes Assuming that glycaemic management has been optimised for with type 1 diabetes: those with lower HbA 1c levels non-exercise days, the exercise type (aerobic, anaerobic, mixed) (~48 mmol/mol [6.5%]) have superior cardiorespiratory Diabetologia (2020) 63:1475–1490 1481 a 13% 17% 39% 31% High (>200 g/day) Moderate (100–200 g/day) Low (40–99 g/day) Very low (<40 g/day) b 1.8 1.6 Carbohydrate ingestion rate (g [kg BM]−1 h−1) 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0 01:46 01:48 02:01 02:52 03:35 04:12 04:28 05:30 05:47 05:53 02:54 03:14 04:08 06:36 04:58 05:53 06:00 06:28 07:30 08:12/day 13:30 13:30 15:42 15:42 n value n=14 n=14 n=14 n=14 n=15 n=15 n=15 n=15 n=15 n=15 n=15 n=15 n=15 n=15 n=15 n=15 n=15 n=15 n=15 n=15 n=6 n=6 n=6 n=6 n=6 n=6 n=6 n=6 n=6 n=6 n=6 n=6 n=6 n=6 n=10 n=10 n=9 n=10 n=9 n=10 n=1 n=1 n=1 n=1 n=1 n=1 n=1 n=1 Distance 21.1 21.1 56 90 110 134 144 164 182 183 141 143 128 195 219 213 207 90 75 42.2 82 82 82 82 (km) ×4 Exercise Run Run Cycle Cycle Cycle Cycle Cycle Cycle Cycle Cycle Cycle Cycle Cycle Cycle Cycle Cycle Cycle Ski Ski 4 day Run Run Run Run mode walk Reference [72] Team Novo Nordisk (Training) [73] Team Novo Nordisk (Tour of California) [31] [43] [74] [75] [76] Fig. 4 Carbohydrate intake patterns in athletic individuals with type 1 participants), with a meal plan of carbohydrate intake of <40 g/day (S. R. diabetes. (a) Daily macronutrient intake from carbohydrates in 252 active Colberg, unpublished data). (b) Carbohydrate intake rates during endur- adults with type 1 diabetes and of varying athletic level who were ance training and competition events: Data are from select studies and surveyed about their carbohydrate intake patterns. Approximately 40% were analysed by R. M. Bracken [31, 43, 72–76]. The x-axis provides self-reported carbohydrate consumption within the current acceptable information on exercise duration (h:min) and also profiles the exercise macronutrient distribution range of 45–65% of energy intake (>200 g/ distance and mode used in each of the published studies. Mean carbohy- day) [109], whereas ~30% consciously moderated their carbohydrate drate intake rate across the studies analysed was 0.70 ± 0.26 g [kg body intake (100–200 g/day), typically by avoiding starchy or sugary foods. mass]−1 h−1 (50 ± 18 g/h). BM, body mass. This figure is available as part The remainder described following either a low-carbohydrate diet (17% of a downloadable slideset of participants; ~40–99 g/day) or a very-low-carbohydrate diet (13% of fitness and pulmonary function than those with higher HbA1c moderate carbohydrate intake (50% of total energy) was supe- levels (~62 mmol/mol [7.8%]) [62]. However, it is unclear if rior to high carbohydrate intake (59% of total energy) for achieving this via restricted carbohydrate feeding, rather than glycogen replenishment, glycaemic management and perfor- by administering more insulin or by some other means, may mance [64]. If carbohydrate intake is limited after prolonged compromise endurance performance and/or increase exercise, restoring muscle glycogen levels is likely to take hypoglycaemia or ketoacidosis risk [63]. longer [65], which may increase nocturnal hypoglycaemia Muscle glycogen replenishment following exercise risk [63]. Co-ingesting protein with moderate amounts of requires effective blood glucose management and balancing carbohydrate (e.g. 0.8 g kg−1 h−1) post exercise may provide of carbohydrate intake with insulin dosing. In one study, a feasible option for normal muscle glycogen repletion, while 1482 Diabetologia (2020) 63:1475–1490 still balancing blood glucose levels [66]. However, high trend arrows and rate of change data [70]. Glucose data should dietary protein intake does not appear to increase muscle be analysed together with a connected smart pen that can glycogen repletion rates further in those already consuming automatically log insulin administration [80], or with pump enough carbohydrate [67]. data [81], to better manage complex situations that may arise During exercise, carbohydrate requirements depend on the due to exercise. With multi-day training, monitoring the AGP use of insulin or other medications, exercise timing, activity can help athletes and clinicians to define achievable blood undertaken and starting blood glucose levels. Because of a glucose (and, consequently, performance) goals [82]. Due to tendency for lower insulin levels and/or elevation in morning the unique challenges of glycaemic management during cortisol and growth hormone levels, training before breakfast competition, athletes with type 1 diabetes should engage in may require little-to-no carbohydrate ingestion during activi- several training sessions that closely mimic competition-day ty, as compared with afternoon exercise [68]. An elevated pre- conditions to optimise management strategies. exercise blood glucose level in the morning or afternoon The glycaemic targets for health and performance of reduces carbohydrate intake needs. Participation in resistance athletes with type 1 diabetes should be individualised. exercise [69] and high-intensity interval-type training [53] However, we propose that for any training period, athletes may not require carbohydrate intake since glucose levels tend should aim for >70% TIR (3.9–10.0 mmol/l), with <4% below not to drop. 3.9 mmol/l and <1% below 3.0 mmol/l, identical to the recom- Carbohydrate intake and/or insulin reduction is typically mendations for the type 1 diabetes adult population [83] (Fig. required for activities lasting >30 min in a non-fasting state, 5). Since hypoglycaemia during exercise can severely impact to prevent hypoglycaemia. For low-to-moderate-intensity performance and, potentially, heart rate variability [84], aerobic activities lasting 30–60 min that are undertaken when athletes should aim for <1% time below target and >75% circulating insulin is at basal levels, the intake of small TIR during competition. Reducing glycaemic variability, as amounts (8–20 g) of carbohydrate may suffice to limit measured by a coefficient of variation of ≤36% for CGM hypoglycaemia, but are not likely to affect performance [70]. values, is also recommended since values above this threshold With higher circulating insulin exposure due to bolus insulin appear to correlate with increased hypoglycaemia risk [85]. administration, 30–60 g/h carbohydrate may be needed when While we acknowledge that these targets are ambitious, they the exercise duration lasts >30 min [71]. Carbohydrate intake may be achievable with newer technologies and dedication. rates of 0.4 g to 1.3 g carbohydrate per kg body mass per h have been reported for athletes with type 1 diabetes exercising in performance settings lasting ≥60 min (Fig. 4b). These stud- Additional considerations ies found that carbohydrate intake within this range prevented hypoglycaemia and enhanced endurance performance in Many competing athletes deal with additional factors that can prolonged exercise [31, 43, 72–76]. affect performance. Poor recognition of hypoglycaemia, trav- elling, optimisation of body weight and/or menstrual cycle Hydration and electrolyte balance Adequate hydration during variations in insulin sensitivity are a few factors that may have training and competition is required to maintain blood volume an impact on glucose control and performance. and for thermoregulation [77]. Athletes with type 1 diabetes may experience mild to moderate dehydration during exercise Hypoglycaemia unawareness Individuals with type 1 diabetes if their blood glucose is elevated, which can be exacerbated by often develop impaired awareness of hypoglycaemia, which the fact that hyperglycaemia increases urinary water loss. increases the risk for a severe hypoglycaemic event by Fluid intake during training tends to be higher in type 1 diabe- approximately sixfold [86]. Active individuals may be at an tes, as compared with control individuals, perhaps because of elevated risk for developing impaired hypoglycaemia aware- elevated thirst caused by hyperglycaemia [78]. In general, ness and counterregulatory failure during exercise: routine plain water or a carbohydrate–electrolyte beverage, depending exercise blunts counterregulation during a hypoglycaemic on glucose level, should be consumed at a rate of ~1 l/h [79]. event [87], which may be a form of habituation. Altering the training exposure to a novel stimulus, such as high-intensity interval training, may help dishabituation and may improve Recommendations for rtCGM/isCGM use both hypoglycaemia symptom recognition and counterregulation [88]. rtCGM and isCGM may allow athletes with type 1 diabetes to better manage their glucose levels during training, competi- Weight management Sports like gymnastics and cycling tion and recovery. When used during prolonged exercise, the require low body weight (and/or fat mass) for performance, initiation of carbohydrate feeding can be based on glucose while others benefit from maximised body mass (e.g. Olympic concentrations (e.g. sensor glucose <8.0 mmol/l), glucose deadlifts). Combat athletes aiming to compete in the lightest Diabetologia (2020) 63:1475–1490 1483 a Training b Compeon Target Target Glucose range Glucose range (% of me) (% of me) >13.9 mmol/l <5% >13.9 mmol/l <5% >10 mmol/l <25%a >10 mmol/l <25%a 3.9–10 mmol/l >70% 3.9–10 mmol/l >75% <3.9 mmol/l <4%b <3.0 mmol/l <1% <3.9 mmol/l <1%b Fig. 5 Proposed CGM-based targets for athletes with type 1 diabetes competition in individuals aged <25 years, if the HbA1c goal is during training (a) and competition (b). Targets for training days are <58 mmol/mol [7.5%], then the TIR target should be set to ~60% but a based on the international consensus [83], while the targets for competi- goal of <4% time below target range (<3.9 mmol/l glucose) should be tion are based on the opinion of the authors. Individual targets should be maintained. aIncludes percentage of values >13.9 mmol/l. bIncludes set by the individual’s healthcare provider with consideration of several percentage of values <3.0 mmol/l. This figure is available as part of a variables, including age, duration of diabetes, diabetes-related complica- downloadable slideset tions and level of hypoglycaemia awareness. In both training and weight category possible often must lose weight for pre-event basal insulin delivery rate [94]. Since the luteal phase is also weigh-ins [89]. These athletes typically combine chronic and associated with high oestrogen levels and rising progesterone acute strategies to achieve target weights, including energy levels, hyperglycaemia is more prevalent [95] and an restriction and dehydration [90]. Such high-risk practices increased reliance on lipids as a fuel source during training may increase the likelihood of severe dehydration and, possi- and recovery may occur [96]. Moreover, the luteal phase is bly, even death [91]. Safe and effective weight management associated with less muscle glycogen mobilisation during strategies are possible in athletes with type 1 diabetes. Since endurance exercise, at least in those without diabetes [97], insulin is an anabolic/anti-catabolic hormone [18], gradual implying that less carbohydrate intake may be required for reductions in both energy intake and insulin daily dose are post-exercise glycogen replenishment. effective for gradually lowering fat mass without compromis- ing muscle mass and safety. It should be noted that acute Travel Regular travel, a key part of being a modern-day episodes of hypoglycaemia are associated with food cravings, athlete, can present a significant challenge to athletes with which can cause disinhibited eating behaviours [92]. Training type 1 diabetes. Individuals need to be well prepared for their in settings of low circulating insulin levels should maximise journey by ensuring they have enough accessible supplies energy provision and training adaptations without requiring (Fig. 6) [98]. Difficulties may arise from practical decisions excessive snacking, if weight loss is desired [63]. about packing insulin properly and bringing spare diabetes- related supplies (e.g. meters, sensors, pumps, needles, gluca- The female athlete Female athletes with type 1 diabetes may gon, snacks, etc.) in carry-on luggage. Choosing appropriate have unique glycaemic responses to training and competition travel insurance, dealing with airport security procedures, depending on the stage of the menstrual cycle that they are delayed flights and choosing appropriate on-board meals are currently in, and may have a reduced risk for hypoglycaemia also important considerations. When flying long distances and as compared with male athletes [93]. Female athletes should crossing multiple time zones, individuals must develop strat- be aware that insulin and carbohydrate needs before and after egies to adapt to new time zones, limit the effects of jet lag/ exercise/training may differ throughout their menstrual cycle. travel on insulin needs and be hypervigilant to manage blood In general, higher blood glucose levels are found during the glucose levels [99]. Athletes should prepare for the possibility luteal phase, which is often not fully abolished by increasing of losing diabetes-related supplies, consuming unfamiliar 1484 Diabetologia (2020) 63:1475–1490 Risk of Choosing an lost appropriate baggage on-board meal Effects of flying on insulin pump Pack spare insulin Change in climate on blood glucose Getting glucose monitors and insulin pumps Snacks to prevent and through treat hypoglycaemia Travel security F C 120 50 100 80 40 30 Storing Consequences 60 40 20 10 insulin at of delayed/late 20 0 0 -10 correct flights on -20 -40 -20 -30 temperature blood glucose -50 -40 Jet lag Reduce sedentary time/increase physical activity whilst travelling Fig. 6 Additional travel considerations for athletes with type 1 diabetes. hyperglycaemia [99]. Increased vigilance around glucose monitoring A summary of practical considerations that an athlete with type 1 diabetes and insulin-dose alterations, as well as access to healthy dietary options, should take into account when travelling for athletic competition. Long- diabetes supplies and, at least, light physical activity (not shown), should distance travel typically increases sedentary time (not shown), can alter be considered. This figure is available as part of a downloadable slideset food choices and tends to be associated with risk of hypo- and foods, and managing changes in climate and other environ- Summary mental conditions. If significant time zone changes will occur, those using MDI may need to alter their basal insulin strategy, Despite the challenges, athletes with type 1 diabetes such as by splitting the basal dose into two doses spaced ~12 h continue to excel at all levels of competition, with some apart before departure [100], or use insulin degludec, which even achieving gold medals at the Olympic Games. has a long half-life (>25 h) and is more flexible with respect to Several strategies can be implemented to help manage dose timing than insulin glargine (~12 h half-life) [101]. athletes with type 1 diabetes (see Text box). Recent Diabetologia (2020) 63:1475–1490 1485 Recommendations for glycaemic management in competitive athletes with type 1 diabetes Considerations Strategies Relative hyperinsulinaemia during If using CSII: prolonged exercise 50–80% basal rate reduction set 90 min pre exercise or Pump suspension at exercise onset (less effective at mitigating hypoglycaemia) If using MDI: 20–50% basal rate reduction before exercise event and/or Carbohydrate consumption (up to 70–90 g/h) during exercise All athletes: Mini-dose glucagon pre exercise (not currently approved) Relative hypoinsulinaemia post If using CSII: intensive exercise Temporary insulin basal rate increase If using MDI or CSII: Hydrate More prolonged aerobic warm-up Restructuring and overlearning of skills to offset stress Glucose management with If using rtCGM or isCGM: dynamic training Sugar-free hydration if glucose is elevated (>10.0 mmol/l) Initiate carbohydrate feeding if glucose level drops below 8.0 mmol/l based on self-monitored blood glucose or CGM Review AGP with healthcare team Review current strategies for management around competition and training During training, aim for >70% TIR, <4% below 3.9 mmol/l and <1% below 3.0 mmol/l During competition, aim for >75% TIR and <1% below 3.9 mmol/l Glucose management with If partaking in aerobic exercise: different modalities and durations Generally, causes drop in blood glucose levels of exercise If partaking in more intensive aerobic/anaerobic exercise: Generally, may cause rise in blood glucose in the fasted state and a rise in lactate Reductions in insulin not recommended Post-exercise hyperglycaemia can be managed with a conservative insulin bolus (50% of usual correction dose) Nutrition and carbohydrate intake Some athletes with type 1 diabetes claim to moderate carbohydrate intake to help with glucose management Training before breakfast is likely to require little-to-no carbohydrate ingestion during activity Carbohydrate loading pre exercise and/or feeding during competition is feasible as long as insulin is matched to the amount of carbohydrate If carbohydrate intake is limited following prolonged exercise, restoring muscle glycogen may take longer If partaking in activities for 30–60 min: Under basal circulating insulin levels, 8–20 g carbohydrate will suffice With higher circulating insulin levels, 30–60 g/h carbohydrate may be needed If partaking in activities for >60 min: Likely to require carbohydrate intake and/or insulin reduction 0.4–1.3 g carbohydrate per kg of body mass per h recommended to prevent hypoglycaemia and enhance performance Hydration Plain water or carbohydrate–electrolyte beverage should be consumed at a rate of ~1 l/h based on performance goals and blood glucose level rtCGM or isCGM Carbohydrate consumption should be initiated when interstitial glucose drops below ~8.0 mmol/l during endurance events During training, aim for >70% TIR, <4% below 3.9 mmol/l, and <1% below 3.0 mmol/l During competition, aim for >75% TIR 1486 Diabetologia (2020) 63:1475–1490 advances in glucose monitoring technologies allow for References minute by minute manipulations in insulin administra- tion and nutrient intake to achieve near optimal 1. Stadler M, Peric S, Strohner-Kaestenbauer H et al (2014) Mortality and incidence of renal replacement therapy in people glycaemic control. In general, athletes with type 1 with type 1 diabetes mellitus–a three decade long prospective diabetes perform training and competition with elevated observational study in the Lainz T1DM cohort. J Clin circulating insulin levels and blunted glucagon responses Endocrinol Metab 99(12):4523–4530. https://doi.org/10.1210/jc. that typically require a high rate of carbohydrate 2014-2701 consumption in race events. However, some athletes 2. Foster NC, Beck RW, Miller KM et al (2019) State of type 1 diabetes management and outcomes from the T1D Exchange in follow a low-to-moderate-carbohydrate diet on non-race 2016–2018. Diabetes Technol Ther 21(2):66–72. https://doi.org/ days, which appears to improve overall glycaemic 10.1089/dia.2018.0384 control and preserve muscle glycogen storage capacity. 3. Wu N, Bredin SSD, Guan Y et al (2019) Cardiovascular health Future research is needed to better define the optimal benefits of exercise training in persons living with type 1 diabetes: a systematic review and meta-analysis. J Clin Med 8(2):253. macronutrient diet for training and competition in these https://doi.org/10.3390/jcm8020253 exceptional athletes. Maintaining a high TIR should 4. Moy CS, Songer TJ, LaPorte RE et al (1993) Insulin-dependent allow for maximal performance and safety during diabetes mellitus, physical activity, and death. Am J Epidemiol periods of training, travel and competition. 137(1):74–81. https://doi.org/10.1093/oxfordjournals.aje. a116604 Acknowledgements The authors would like to thank F. Fontana, Head of 5. Sedivy P, Dezortova M, Drobny M, Vlasakova Z, Herynek V, Performance at Team Novo Nordisk (Verona, Italy), for his input and Hajek M (2018) Differences in muscle metabolism in patients edits to Fig. 1. with type I diabetes – influence of gender and nephropathy studied by 31P MR spectroscopy. Physiol Res 67(3):433–441. https://doi. Authors’ relationships and activities MCR has received lecture fees org/10.33549/physiolres.933745 from Medtronic Diabetes, Novo Nordisk and Insulet, consulting/ 6. Monaco CMF, Hughes MC, Ramos SV et al (2018) Altered mito- advisory board fees from Xeris Pharmaceticles and Zucara thera- chondrial bioenergetics and ultrastructure in the skeletal muscle of peutics, research grants from Sanofi and Novo Nordisk and mate- young adults with type 1 diabetes. Diabetologia 61(6):1411–1423. rial funding from Dexcom. SNS is employed by Team Novo https://doi.org/10.1007/s00125-018-4602-6 Nordisk Professional Cycling Team. OM has received lecture fees 7. Ruegsegger GN, Creo AL, Cortes TM, Dasari S, Nair KS (2018) from Medtronic, travel grants from Novo Nordisk, Medtronic and Altered mitochondrial function in insulin-deficient and insulin- Sanofi-Aventis, research grants from Sêr Cymru II COFUND resistant states. J Clin Invest 128(9):3671–3681. https://doi.org/ fellowship/European Union and Novo Nordisk and material 10.1172/JCI120843 funding from Abbott Diabetes Care. JEY has received lecture fees 8. Murray B, Rosenbloom C (2018) Fundamentals of glycogen from Dexcom and in-kind research support from Dexcom, metabolism for coaches and athletes. Nutr Rev 76(4):243–259. Lifescan Canada and Abbott Nutrition. DPZ has received lecture https://doi.org/10.1093/nutrit/nuy001 fees from Medtronic Diabetes, Ascensia Diabetes and Insulet 9. Marliss EB, Vranic M (2002) Intense exercise has unique effects Corporation. PA has received unrestricted research grants from on both insulin release and its roles in glucoregulation: implica- Novo Nordisk and has been part of advisory boards for Abbot, tions for diabetes. Diabetes 51(Suppl 1):S271–S283. https://doi. Eli Lilly, Novo Nordisk and Roche during the last 2 years. RMB org/10.2337/diabetes.51.2007.s271 reports having received honoraria, travel and educational grant 10. DuBose SN, Li Z, Sherr JL, Beck RW, Tamborlane WV, Shah support from Beneo, Boehringer-Ingelheim, Eli Lily and VN (2020) Effect of exercise and meals on continuous glucose Company, Novo Nordisk and Sanofi-Aventis. PAF, SRC, IWG monitor data in healthy individuals without diabetes. J Diabetes and CS declare that there are no relationships or activities that Sci Technol. https://doi.org/10.1177/1932296820905904 might bias, or be perceived to bias, their work. 11. Rothman DL, Magnusson I, Katz LD, Shulman RG, Shulman GI (1991) Quantitation of hepatic glycogenolysis and gluconeogene- Contribution statement All authors were responsible for drafting the sis in fasting humans with 13C NMR. Science 254(5031):573– article and revising it critically for important intellectual content. All 576. https://doi.org/10.1126/science.1948033 authors approved the version to be published. 12. Buehler T, Bally L, Dokumaci AS, Stettler C, Boesch C (2016) Methodological and physiological test-retest reliability of 13C- Open Access This article is licensed under a Creative Commons MRS glycogen measurements in liver and in skeletal muscle of Attribution 4.0 International License, which permits use, sharing, adap- patients with type 1 diabetes and matched healthy controls. NMR tation, distribution and reproduction in any medium or format, as long as Biomed 29(6):796–805. https://doi.org/10.1002/nbm.3531 you give appropriate credit to the original author(s) and the source, 13. Bischof MG, Bernroider E, Krssak M et al (2002) Hepatic glyco- provide a link to the Creative Commons licence, and indicate if changes gen metabolism in type 1 diabetes after long-term near were made. The images or other third party material in this article are normoglycemia. Diabetes 51(1):49–54. https://doi.org/10.2337/ included in the article's Creative Commons licence, unless indicated diabetes.51.1.49 otherwise in a credit line to the material. If material is not included in 14. Hwang JH, Perseghin G, Rothman DL et al (1995) Impaired net the article's Creative Commons licence and your intended use is not hepatic glycogen synthesis in insulin-dependent diabetic subjects permitted by statutory regulation or exceeds the permitted use, you will during mixed meal ingestion. A 13C nuclear magnetic resonance need to obtain permission directly from the copyright holder. To view a spectroscopy study. J Clin Invest 95(2):783–787. https://doi.org/ copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 10.1172/JCI117727 Diabetologia (2020) 63:1475–1490 1487 15. Bischof MG, Krssak M, Krebs M et al (2001) Effects of short-term 32. Yardley JE, Iscoe KE, Sigal RJ, Kenny GP, Perkins BA, Riddell improvement of insulin treatment and glycemia on hepatic glyco- MC (2013) Insulin pump therapy is associated with less post- gen metabolism in type 1 diabetes. Diabetes 50(2):392–398 exercise hyperglycemia than multiple daily injections: an obser- 16. Riddell MC, Gallen IW, Smart CE et al (2017) Exercise manage- vational study of physically active type 1 diabetes patients. ment in type 1 diabetes: a consensus statement. Lancet Diabetes Diabetes Technol Ther 15(1):84–88. https://doi.org/10.1089/dia. Endocrinol 5(5):377–390. https://doi.org/10.1016/S2213- 2012.0168 8587(17)30014-1 33. Tagougui S, Taleb N, Molvau J, Nguyen É, Raffray M, Rabasa- 17. Achten J, Jeukendrup AE (2004) Optimizing fat oxidation through Lhoret R (2019) Artificial pancreas systems and physical activity exercise and diet. Nutrition 20(7–8):716–727. https://doi.org/10. in patients with type 1 diabetes: challenges, adopted approaches, 1016/j.nut.2004.04.005 and future perspectives. J Diabetes Sci Technol 13(6):1077–1090. 18. Russell-Jones D, Khan R (2007) Insulin-associated weight gain in https://doi.org/10.1177/1932296819869310 diabetes – causes, effects and coping strategies. Diabetes Obes 34. Seereiner S, Neeser K, Weber C et al (2010) Attitudes towards Metab 9(6):799–812. https://doi.org/10.1111/j.1463-1326.2006. insulin pump therapy among adolescents and young people. 00686.x Diabetes Technol Ther 12(1):89–94. https://doi.org/10.1089/dia. 19. Ahola AJ, Forsblom C, Harjutsalo V, Groop P-H (2019) Dietary 2009.0080 carbohydrate intake and cardio-metabolic risk factors in type 1 35. Aronson R, Li A, Brown RE, McGaugh S, Riddell MC (2020) diabetes. Diabetes Res Clin Pract 155:107818. https://doi.org/10. Flexible insulin therapy with a hybrid regimen of insulin degludec 1016/j.diabres.2019.107818 and continuous subcutaneous insulin infusion with pump suspen- 20. Mosher PE, Nash MS, Perry AC, LaPerriere AR, Goldberg RB sion before exercise in physically active adults with type 1 diabe- (1998) Aerobic circuit exercise training: effect on adolescents with tes (FIT Unthethered): a single-centre, open-label, proof-of- well-controlled insulin-dependent diabetes mellitus. Arch Phys concept, randomised crossover trial. Lancet Diabetes Endocrinol Med Rehabil 79(6):652–657 8(6):511–523. https://doi.org/10.1016/S2213-8587(20)30114-5 21. Wahrenberg H, Lönnqvist F, Engfeldt P, Arner P (1989) 36. Lucidi P, Porcellati F, Bolli GB, Fanelli CG (2018) Prevention and Abnormal action of catecholamines on lipolysis in adipocytes of management of severe hypoglycemia and hypoglycemia unaware- type I diabetic patients treated with insulin. Diabetes 38(4):524– ness: incorporating sensor technology. Curr Diab Rep 18(10):83 533. https://doi.org/10.2337/diab.38.4.524 37. Moser O, Yardley JE, Bracken RM (2018) Interstitial glucose and 22. Chokkalingam K, Tsintzas K, Norton L, Jewell K, Macdonald IA, physical exercise in type 1 diabetes: integrative physiology, tech- Mansell PI (2007) Exercise under hyperinsulinaemic conditions nology, and the gap in-between. Nutrients 10(1):93. https://doi. increases whole-body glucose disposal without affecting muscle org/10.3390/nu10010093 glycogen utilisation in type 1 diabetes. Diabetologia 50(2):414– 38. Moser O, Eckstein ML, McCarthy O et al (2019) Performance of 421. https://doi.org/10.1007/s00125-006-0520-0 the Freestyle Libre flash glucose monitoring (flash GM) system in 23. Rennie MJ, Bohé J, Smith K, Wackerhage H, Greenhaff P (2006) individuals with type 1 diabetes: a secondary outcome analysis of Branched-chain amino acids as fuels and anabolic signals in a randomized crossover trial. Diabetes Obes Metab 21(11):2505– human muscle. J Nutr 136(1):264S–268S. https://doi.org/10. 2512. https://doi.org/10.1111/dom.13835 1093/jn/136.1.264S 39. Zaharieva DP, McGaugh S, Pooni R, Vienneau T, Ly T, Riddell 24. Wahren J, Hagenfeldt L, Felig P (1975) Splanchnic and leg MC (2019) Improved open-loop glucose control with basal insulin exchange of glucose, amino acids, and free fatty acids during reduction 90 minutes before aerobic exercise in patients with type exercise in diabetes mellitus. J Clin Invest 55(6):1303–1314. 1 diabetes on continuous subcutaneous insulin infusion. Diabetes https://doi.org/10.1172/JCI108050 Care 42(5):824–831. https://doi.org/10.2337/dc18-2204 25. O’Neill BT, Bhardwaj G, Penniman CM et al (2019) FoxO tran- 40. Zaharieva DP, Cinar A, Yavelberg L, Jamnik V, Riddell MC scription factors are critical regulators of diabetes-related muscle (2019) No disadvantage to insulin pump off vs pump on during atrophy. Diabetes 68(3):556–570. https://doi.org/10.2337/db18- intermittent high-intensity exercise in adults with type 1 diabetes. 0416 Can J Diabetes 44(2):162–168. https://doi.org/10.1016/j.jcjd. 26. Petersen MC, Shulman GI (2018) Mechanisms of insulin action 2019.05.015 and insulin resistance. Physiol Rev 98(4):2133–2223. https://doi. 41. Campbell MD, Walker M, Bracken RM et al (2015) Insulin ther- org/10.1152/physrev.00063.2017 apy and dietary adjustments to normalize glycemia and prevent 27. Aarnio P, Lauritsen T, Dela F (2001) Insulin secretion and glucose nocturnal hypoglycemia after evening exercise in type 1 diabetes: kinetics during exercise with and without pharmacological α1- a randomized controlled trial. BMJ Open Diabetes Res Care 3(1): and α2-receptor blockade. Diabetes 50(8):1834–1843. https:// e000085. https://doi.org/10.1136/bmjdrc-2015-000085 doi.org/10.2337/diabetes.50.8.1834 42. Moser O, Eckstein ML, Mueller A et al (2019) Reduction in insu- 28. Koivisto VA, Felig P (1978) Effects of leg exercise on insulin lin degludec dosing for multiple exercise sessions improves time absorption in diabetic patients. N Engl J Med 298(2):79–83. spent in euglycaemia in people with type 1 diabetes: a randomized https://doi.org/10.1056/NEJM197801122980205 crossover trial. Diabetes Obes Metab 21(2):349–356. https://doi. 29. Peter R, Luzio SD, Dunseath G et al (2005) Effects of exercise on org/10.1111/dom.13534 the absorption of insulin glargine in patients with type 1 diabetes. 43. Mattsson S, Jendle J, Adolfsson P (2019) Carbohydrate loading Diabetes Care 28(3):560–565 followed by high carbohydrate intake during prolonged physical 30. Berger M, Berchtold P, Cuppers HJ et al (1977) Metabolic and exercise and its impact on glucose control in individuals with hormonal effects of muscular exercise in juvenile type diabetics. diabetes type 1-an exploratory study. Front Endocrinol 10:571. Diabetologia 13(4):355–365 https://doi.org/10.3389/fendo.2019.00571 31. Scott SN, Christiansen MP, Fontana FY et al (2020) Evaluation of 44. Mallad A, Hinshaw L, Schiavon M et al (2015) Exercise effects on factors related to glycemic management in professional cyclists postprandial glucose metabolism in type 1 diabetes: a triple-tracer with type 1 diabetes over a 7-day stage race. Diabetes Care approach. Am J Physiol Endocrinol Metab 308(12):E1106– 43(5):1142–1145. https://doi.org/10.2337/dc19-2302 E1115. https://doi.org/10.1152/ajpendo.00014.2015 1488 Diabetologia (2020) 63:1475–1490 45. Rickels MR, DuBose SN, Toschi E et al (2018) Mini-dose gluca- 61. Brand-Miller J, Buyken AE (2012) The glycemic index issue. gon as a novel approach to prevent exercise-induced hypoglyce- Curr Opin Lipidol 23(1):62–67. https://doi.org/10.1097/MOL. mia in type 1 diabetes. Diabetes Care 41(9):1909–1916. https:// 0b013e32834ec705 doi.org/10.2337/dc18-0051 62. Baldi JC, Cassuto NA, Foxx-Lupo WT, Wheatley CM, Snyder EM 46. Castle JR, El Youssef J, Wilson LM et al (2018) Randomized (2010) Glycemic status affects cardiopulmonary exercise response in outpatient trial of single- and dual-hormone closed-loop systems athletes with type I diabetes. Med Sci Sports Exerc 42(8):1454–1459. that adapt to exercise using wearable sensors. Diabetes Care 41(7): https://doi.org/10.1249/MSS.0b013e3181d1fdb3 1471–1477. https://doi.org/10.2337/dc18-0228 63. Scott SN, Anderson L, Morton JP, Wagenmakers AJM, Riddell 47. Dizon S, Malcolm J, Rowan M, Keely EJ (2019) Patient perspec- MC (2019) Carbohydrate restriction in type 1 diabetes: a realistic tives on managing type 1 diabetes during high-performance exer- therapy for improved glycaemic control and athletic performance? cise: what resources do they want? Diabetes Spectr 32(1):36–45. Nutrients 11(5):1022. https://doi.org/10.3390/nu11051022 https://doi.org/10.2337/ds18-0016 64. McKewen MW, Rehrer NJ, Cox C, Mann J (1999) Glycaemic 48. Hanton S, Wadey R, Mellalieu SD (2008) Advanced psycholog- control, muscle glycogen and exercise performance in IDDM ical strategies and anxiety responses in sport. Sport Psychol 22(4): athletes on diets of varying carbohydrate content. Int J Sports 472–490. https://doi.org/10.1123/tsp.22.4.472 Med 20(6):349–353. https://doi.org/10.1055/s-2007-971143 49. Aronson R, Brown RE, Li A, Riddell MC (2019) Optimal insulin 65. Jensen J, Rustad PI, Kolnes AJ, Lai Y-C (2011) The role of skel- correction factor in post-high-intensity exercise hyperglycemia in etal muscle glycogen breakdown for regulation of insulin sensitiv- adults with type 1 diabetes: the FIT study. Diabetes Care 42(1): ity by exercise. Front Physiol 2:112. https://doi.org/10.3389/ 10–16. https://doi.org/10.2337/dc18-1475 fphys.2011.00112 50. Turner D, Luzio S, Gray BJ et al (2016) Algorithm that delivers an 66. Alghannam AF, Gonzalez JT, Betts JA (2018) Restoration of individualized rapid-acting insulin dose after morning resistance muscle glycogen and functional capacity: role of post-exercise exercise counters post-exercise hyperglycaemia in people with carbohydrate and protein co-ingestion. Nutrients 10(2):253. type 1 diabetes. Diabet Med 33(4):506–510. https://doi.org/10. https://doi.org/10.3390/nu10020253 1111/dme.12870 67. Jentjens RL, van Loon LJ, Mann CH, Wagenmakers AJ, 51. Yardley JE, Zaharieva DP, Jarvis C, Riddell MC (2015) The “ups” Jeukendrup AE (2001) Addition of protein and amino acids to and “downs” of a bike race in people with type 1 diabetes: dramat- carbohydrates does not enhance postexercise muscle glycogen ic differences in strategies and blood glucose responses in the synthesis. J Appl Physiol 91(2):839–846. https://doi.org/10. Paris-to-Ancaster Spring Classic. Can J Diabetes 39(2):105–110. 1152/jappl.2001.91.2.839 https://doi.org/10.1016/j.jcjd.2014.09.003 68. Gomez AM, Gomez C, Aschner P et al (2015) Effects of 52. Gallen IW, Hume C, Lumb A (2011) Fuelling the athlete with type performing morning versus afternoon exercise on glycemic 1 diabetes. Diabetes Obes Metab 13(2):130–136. https://doi.org/ control and hypoglycemia frequency in type 1 diabetes patients 10.1111/j.1463-1326.2010.01319.x on sensor-augmented insulin pump therapy. J Diabetes Sci 53. Riddell MC, Pooni R, Yavelberg L et al (2019) Reproducibility in Technol 9(3):619–624. https://doi.org/10.1177/ the cardiometabolic responses to high-intensity interval exercise in 1932296814566233 adults with type 1 diabetes. Diabetes Res Clin Pract 148:137–143. 69. Turner D, Luzio S, Gray BJ et al (2015) Impact of single and multiple https://doi.org/10.1016/j.diabres.2019.01.003 sets of resistance exercise in type 1 diabetes. Scand J Med Sci Sports 54. De Pero R, Cibelli G, Cortis C, Sbriccoli P, Capranica L, 25(1):e99–e109. https://doi.org/10.1111/sms.12202 Piacentini MF (2016) Stress related changes during TeamGym 70. Riddell MC, Milliken J (2011) Preventing exercise-induced hypo- competition. J Sports Med Phys Fitness 56(5):639–647 glycemia in type 1 diabetes using real-time continuous glucose 55. Pinsker JE, Kraus A, Gianferante D et al (2016) Techniques for exer- monitoring and a new carbohydrate intake algorithm: an observa- cise preparation and management in adults with type 1 diabetes. Can J tional field study. Diabetes Technol Ther 13(8):819–825. https:// Diabetes 40(6):503–508. https://doi.org/10.1016/j.jcjd.2016.04.010 doi.org/10.1089/dia.2011.0052 56. Al Khalifah RA, Suppère C, Haidar A, Rabasa-Lhoret R, Ladouceur M, Legault L (2016) Association of aerobic fitness level with 71. Francescato MP, Stel G, Stenner E, Geat M (2015) Prolonged exercise-induced hypoglycaemia in type 1 diabetes. Diabet Med exercise in type 1 diabetes: performance of a customizable algo- 33(12):1686–1690. https://doi.org/10.1111/dme.13070 rithm to estimate the carbohydrate supplements to minimize glycemic imbalances. PLoS One 10(4):e0125220. https://doi.org/ 57. Fahey AJ, Paramalingam N, Davey RJ, Davis EA, Jones TW, 10.1371/journal.pone.0125220 Fournier PA (2012) The effect of a short sprint on postexercise whole-body glucose production and utilization rates in individuals 72. Murillo S, Brugnara L, Novials A (2010) One year follow-up in a with type 1 diabetes mellitus. J Clin Endocrinol Metab 97(11): group of half-marathon runners with type-1 diabetes treated with 4193–4200. https://doi.org/10.1210/jc.2012-1604 insulin analogues. J Sports Med Phys Fitness 50(4):506–510 58. Burke LM, Castell LM, Casa DJ et al (2019) International 73. McCarthy O, Eckstein ML, Scott SN et al (2020) Glycemic responses Association of Athletics Federations consensus statement 2019: to strenuous training in male professional cyclists with type 1 diabetes: nutrition for athletics. Int J Sport Nutr Exerc Metab 29(2):73–84. a prospective observational study. BMJ Open Diabetes Res Care 8(1): https://doi.org/10.1123/ijsnem.2019-0065 e001245. https://doi.org/10.1136/bmjdrc-2020-001245 59. Brazeau AS, Mircescu H, Desjardins K et al (2013) Carbohydrate 74. Sane T, Helve E, Pelkonen R, Koivisto VA (1988) The adjustment counting accuracy and blood glucose variability in adults with of diet and insulin dose during long-term endurance exercise in type 1 diabetes. Diabetes Res Clin Pract 99(1):19–23. https://doi. type 1 (insulin-dependent) diabetic men. Diabetologia 31(1):35– org/10.1016/j.diabres.2012.10.024 40. https://doi.org/10.1007/bf00279130 60. Sylow L, Kleinert M, Richter EA, Jensen TE (2017) Exercise- 75. van Dijk J-W, Eijsvogels TM, Nyakayiru J et al (2016) Glycemic stimulated glucose uptake - regulation and implications for control during consecutive days with prolonged walking exercise glycaemic control. Nat Rev Endocrinol 13(3):133–148. https:// in individuals with type 1 diabetes mellitus. Diabetes Res Clin doi.org/10.1038/nrendo.2016.162 Pract 117:74–81. https://doi.org/10.1016/j.diabres.2016.04.053 Diabetologia (2020) 63:1475–1490 1489 76. Gawrecki A, Zozulinska-Ziolkiewicz D, Matejko B, Hohendorff systematic review. Int J Sport Nutr Exerc Metab 29(4):441–452. J, Malecki MT, Klupa T (2018) Safe completion of a trail running https://doi.org/10.1123/ijsnem.2018-0165 ultramarathon by four men with type 1 diabetes. Diabetes Technol 92. Strachan MWJ, Ewing FME, Frier BM, Harper A, Deary IJ (2004) Ther 20(2):147–152. https://doi.org/10.1089/dia.2017.0296 Food cravings during acute hypoglycaemia in adults with type 1 77. APA (2016) Nutrition and athletic performance. Med Sci Sports diabetes. Physiol Behav 80(5):675–682. https://doi.org/10.1016/j. Exerc 48(3):543–568. https://doi.org/10.1249/MSS. physbeh.2003.12.003 0000000000000852 93. Brockman NK, Sigal RJ, Kenny GP, Riddell MC, Perkins BA, 78. Buoite Stella A, Yardley J, Francescato MP, Morrison SA (2018) Yardley JE (2019) Sex-related differences in blood glucose Fluid intake habits in type 1 diabetes individuals during typical responses to resistance exercise in adults with type 1 diabetes: a training bouts. Ann Nutr Metab 73(1):10–18. https://doi.org/10. secondary data analysis. Can J Diabetes 44(3):267–273. https:// 1159/000489823 doi.org/10.1016/j.jcjd.2019.08.006 79. Yardley JE, Colberg SR (2017) Update on management of type 1 94. Lunt H, Brown LJ (1996) Self-reported changes in capillary diabetes and type 2 diabetes in athletes. Curr Sports Med Rep glucose and insulin requirements during the menstrual cycle. 16(1):38–44. https://doi.org/10.1249/JSR.0000000000000327 Diabet Med 13(6):525–530. https://doi.org/10.1002/(SICI)1096- 80. Adolfsson P, Hartvig NV, Kaas A, Møller JB, Hellman J (2020) 9136(199606)13:6<525::AID-DIA123>3.0.CO;2-D Increased time in range and improved insulin adherence after 95. Brown SA, Jiang B, McElwee-Malloy M, Wakeman C, Breton introduction of a smart connected insulin pen. Diabetes Technol MD (2015) Fluctuations of hyperglycemia and insulin sensitivity Ther. https://doi.org/10.1089/dia.2019.0411 are linked to menstrual cycle phases in women with T1D. J 81. Tagougui S, Taleb N, Rabasa-Lhoret R (2019) The benefits and Diabetes Sci Technol 9(6):1192–1199. https://doi.org/10.1177/ limits of technological advances in glucose management around 1932296815608400 physical activity in patients type 1 diabetes. Front Endocrinol 9: 96. Riddell MC, Partington SL, Stupka N, Armstrong D, Rennie C, 818. https://doi.org/10.3389/fendo.2018.00818 Tarnopolsky MA (2003) Substrate utilization during exercise 82. Kruger DF, Edelman SV, Hinnen DA, Parkin CG (2019) performed with and without glucose ingestion in female and male Reference guide for integrating continuous glucose monitoring endurance trained athletes. Int J Sport Nutr Exerc Metab 13(4): into clinical practice. Diabetes Educ 45(1_suppl):3S–20S. 407–421. https://doi.org/10.1123/ijsnem.13.4.407 https://doi.org/10.1177/0145721718818066 97. Devries MC, Hamadeh MJ, Phillips SM, Tarnopolsky MA (2006) 83. Battelino T, Danne T, Bergenstal RM et al (2019) Clinical targets for Menstrual cycle phase and sex influence muscle glycogen utiliza- continuous glucose monitoring data interpretation: recommendations tion and glucose turnover during moderate-intensity endurance from the International Consensus on Time in Range. Diabetes Care exercise. Am J Physiol Regul Integr Comp Physiol 291(4): 42(8):1593–1603. https://doi.org/10.2337/dci19-0028 R1120–R1128. https://doi.org/10.1152/ajpregu.00700.2005 84. Koivikko ML, Tulppo MP, Kiviniemi AM et al (2012) Autonomic 98. Burnett JCD (2006) Long- and short-haul travel by air: issues for cardiac regulation during spontaneous nocturnal hypoglycemia in people with diabetes on insulin. J Travel Med 13(5):255–260. patients with type 1 diabetes. Diabetes Care 35(7):1585–1590. https://doi.org/10.1111/j.1708-8305.2006.00057.x https://doi.org/10.2337/dc11-2120 99. Pinsker JE, Schoenberg BE, Garey C, Runion A, Larez A, Kerr D 85. Monnier L, Colette C, Wojtusciszyn A et al (2017) Toward defin- (2017) Perspectives on long-distance air travel with type 1 diabe- ing the threshold between low and high glucose variability in tes. Diabetes Technol Ther 19(12):744–748. https://doi.org/10. diabetes. Diabetes Care 40(7):832–838. https://doi.org/10.2337/ 1089/dia.2017.0259 dc16-1769 100. Garg SK, Gottlieb PA, Hisatomi ME et al (2004) Improved glyce- 86. Gold AE, MacLeod KM, Frier BM (1994) Frequency of severe mic control without an increase in severe hypoglycemic episodes hypoglycemia in patients with type I diabetes with impaired in intensively treated patients with type 1 diabetes receiving morn- awareness of hypoglycemia. Diabetes Care 17(7):697–703. ing, evening, or split dose insulin glargine. Diabetes Res Clin Pract https://doi.org/10.2337/diacare.17.7.697 66(1):49–56. https://doi.org/10.1016/j.diabres.2004.02.008 87. Sandoval DA, Guy DLA, Richardson MA, Ertl AC, Davis SN 101. Haahr H, Heise T (2014) A review of the pharmacological properties (2004) Effects of low and moderate antecedent exercise on of insulin degludec and their clinical relevance. Clin Pharmacokinet counterregulatory responses to subsequent hypoglycemia in type 53(9):787–800. https://doi.org/10.1007/s40262-014-0165-y 1 diabetes. Diabetes 53(7):1798–1806 102. Coyle EF, Coggan AR, Hemmert MK, Ivy JL (1986) Muscle 88. Farrell CM, McNeilly AD, Fournier P et al (2020) A randomised glycogen utilization during prolonged strenuous exercise when controlled study of high intensity exercise as a dishabituating stim- fed carbohydrate. J Appl Physiol 61(1):165–172. https://doi.org/ ulus to improve hypoglycaemia awareness in people with type 1 10.1152/jappl.1986.61.1.165 diabetes: a proof-of-concept study. Diabetologia 63(4):853–863. 103. Rowlands DS, Hopkins WG (2002) Effects of high-fat and high- https://doi.org/10.1007/s00125-019-05076-5 carbohydrate diets on metabolism and performance in cycling. 89. Reale R, Slater G, Burke LM (2017) Acute-weight-loss strategies Metab Clin Exp 51(6):678–690. https://doi.org/10.1053/meta.2002. for combat sports and applications to Olympic success. Int J Sports 32723 Physiol Perform 12(2):142–151. https://doi.org/10.1123/ijspp. 104. Bracken RM, Gray BJ, Turner D (2014) Comparison of the meta- 2016-0211 bolic responses to ingestion of hydrothermally processed high- 90. Reale R, Slater G, Burke LM (2017) Individualised dietary strat- amylopectin content maize, uncooked maize starch or dextrose egies for Olympic combat sports: acute weight loss, recovery and in healthy individuals. Br J Nutr 111(7):1231–1238. https://doi. competition nutrition. Eur J Sport Sci 17(6):727–740. https://doi. org/10.1017/S0007114513003619 org/10.1080/17461391.2017.1297489 105. Campbell MD, Walker M, Trenell MI et al (2014) A low-glycemic 91. Matthews JJ, Stanhope EN, Godwin MS, Holmes MEJ, Artioli index meal and bedtime snack prevents postprandial hyperglyce- GG (2019) The magnitude of rapid weight loss and rapid weight mia and associated rises in inflammatory markers, providing gain in combat sport athletes preparing for competition: a protection from early but not late nocturnal hypoglycemia 1490 Diabetologia (2020) 63:1475–1490 following evening exercise in type 1 diabetes. Diabetes Care 108. West DJ, Morton RD, Bain SC, Stephens JW, Bracken RM (2010) 37(7):1845–1853. https://doi.org/10.2337/dc14-0186 Blood glucose responses to reductions in pre-exercise rapid-acting 106. Campbell MD, Walker M, Trenell MI et al (2013) Large pre- and insulin for 24 h after running in individuals with type 1 diabetes. J postexercise rapid-acting insulin reductions preserve glycemia and Sp orts Sci 28 (7):78 1– 788 . https ://doi.org/10.1080/ prevent early- but not late-onset hypoglycemia in patients with type 02640411003734093 1 diabetes. Diabetes Care 36(8):2217–2224. https://doi.org/10.2337/ 109. Manore MM (2005) Exercise and the Institute of Medicine recom- dc12-2467 mendations for nutrition. Curr Sports Med Rep 4(4):193–198. 107. Campbell MD, Walker M, Trenell MI et al (2014) Metabolic https://doi.org/10.1097/01.csmr.0000306206.72186.00 implications when employing heavy pre- and post-exercise rapid-acting insulin reductions to prevent hypoglycaemia in type Publisher’s note Springer Nature remains neutral with regard to jurisdic- 1 diabetes patients: a randomised clinical trial. PLoS One 9(5): tional claims in published maps and institutional affiliations. e97143. https://doi.org/10.1371/journal.pone.0097143