Race Across America: First Athlete With Type 1 Diabetes to Finish Solo With Diabetes Technology Support

T.H., a 53-year-old Austrian amateur cyclist working as a nurse in a retirement home, qualified for the 2018 RAAM after regular endurance bike training and competitions, as well as previous participation in marathon runs and triathlons. Diagnosed with type 1 diabetes in 2003, he had no evidence of diabetes complications and had intact hypoglycemia awareness and no history of severe hypoglycemia.

Having started with basal-bolus insulin treatment after diagnosis, he soon began using a sensor-augmented insulin pump. In 2014, he first experienced high glucose excursions, with an A1C of 8.3% (67.2 mmol/mol), as a result of intensive athletic training. He then changed from intermittently high carbohydrate intake to a ketogenic diet until the RAAM. From October 2015 on, he used a sensor-augmented pump (MiniMed 640G, Medtronic, Northridge, CA). His A1C values ranged from 5.5 to 7.3% (36.6 to 56.3 mmol/mol).

In summer 2017, he qualified for the 2018 RAAM through the Glocknerman Race, a challenging Austrian bicycle race encompassing a distance of 1,000 km and 16,000 vertical meters to overcome. In October 2017, we equipped him with an implanted continuous glucose monitoring (CGM) sensor (Eversense XL, Roche Diabetes Care, Vienna, Austria) to provide additional safety during the race without a danger of sensor loss or dysfunction as a result of sweating, extensive motion, or temperature changes. With a sensor-matching digital app (Eversense Now, Senseonics, Germantown, MD), his glucose values could be tracked continuously by the crew in an escort vehicle. Immediately before his trip to the United States, he was in excellent physical shape, with an A1C of 7.0% (53.0 mmol/mol), an average pump basal rate of 0.5 units/hour, and an insulin-to-carbohydrate ratio of 1:24 (IU:g).

T.H. started his race on 12 June 2018 and reached the finish line after 11 days, 12 hours, and 5 minutes. His race statistics provide a glimpse into the rigors of the race; he cycled 3,069 miles and climbed 179,465 vertical feet, in temperatures of 7–46°C (44.6–114.8°F), across four time zones in 12 U.S. states, with only 3.5 hours of sleep per day. He finished the race in sixth place overall and first place in his age-group (50–59 years). With a mean intake of 5,568 kcal (23.296 kJ) mainly from ketogenic sources and up to 15 L water per day, he managed to keep his glucose levels in an acceptable range for physical and mental ability while having enough strength to cope with the daily challenges of the race. Given his ketogenic diet, his meal composition was almost entirely protein and fat (Table 1).

T.H.’s physical and mental stress was enormous at the start of the race, followed by a phase of uncertainty. His glucose values as detected by the sensor system reflected these difficulties, showing deterioration during the first 4 days, with adaptation to values mostly within his recommended target thereafter (Figure 1A). His mean glucose variability over the whole race period is depicted in Figure 1B. His time in range (individually defined for the race as 80–170 mg/dL [4.4–9.4 mmol/L]), time above range, and time below range were 51.9, 32.6, and 8.3% of values measured, respectively, with an overall sensor glucose mean of 148 mg/dL (8.2 mmol/L). His CGM system was calibrated three times daily throughout the race using a glucose meter (AccuChek Guide, Roche Diabetes Care, Vienna, Austria). T.H. had only a few mild to moderate hypoglycemic reactions (grade 1–2) and a couple of glucose values >250 mg/dL. Laboratory parameters (electrolytes, acid-base balance, red cell parameters, and blood gas analyses) were stable throughout the race and depicted optimal training and race performance (Supplementary Figure S1). Figure 2 shows T.H. wearing his implanted glucose sensor.

Some studies of endurance sports have shown better performance and results with high-carbohydrate diets compared with a ketogenic diet (3,4). In particular, reduced peak performance and time to exhaustion is reported to be overcome by higher glucose availability and glycogen storage capacity (5–8). On the other hand, energy sources in ultra-endurance competitive sports rely up to 90% on β-oxidation, as athletes perform ∼10% below their individual first ventilatory threshold and metabolic threshold (9,10). T.H. was convinced by his positive experience with a ketogenic diet over time. Flat glucose profiles with reduced glucose variability compared with former high glucose excursions and more hypoglycemic reactions with a diet high in carbohydrate content reinforced his dedication to a diet high in protein and fat content. Despite the possibly negative impacts of an unbalanced diet on performance, such diets are not uncommon, particularly for nonelite multisport endurance athletes (11,12).

A sample race day for T.H., including food intake of ∼6,500 kcal (27,196 kJ), glucose values, and accordingly adapted insulin doses, is provided in Supplementary Figure S2. His daily fluid intake was, according to recommendations, 400–800 mL of fluid/hour, adapted to thirst and ambient temperature (13). His sodium and nitrate intake was provided by salted soup, mineral drinks, fruit shakes, and yogurt, the latter providing a source of probiotics, which are considered advantageous in endurance sports (14) and cited as preventing digestive problems such as meteorism, cramps, and diarrhea. Furthermore, T.H. took up to four times daily 10,000 IU of pancreatic enzymes, together with his high-fat creamy meals, to support fat resorption. In addition, caffeine-enriched fluids and caffeine tablets were intermittently used to improve cycling endurance (15). T.H. did not use dietary supplements other than potassium and magnesium, which are often used by endurance athletes.

Adaptation of insulin doses during the race were suggested by the support team based on acquired sensor glucose readings. A backup CGM system (FreeStyle Libre, Abbott, Vienna, Austria) was used in addition to the implanted system to overcome possible dysfunction of the implanted sensor. Data from both CGM systems were tracked by the escort team using Cloud-based apps. As a result of problems with sweat and rapid arm motion, the FreeStyle Libre sensor was clearly inferior to the implanted sensor and was used only during the first half of the race. To provide standardized support for insulin adaptation, a predefined dose correction algorithm was developed, following scientifically based recommendations and individual experience (Table 2) (16,17). In addition to the glucose and nutrition adaptation plan, real-time CGM using an implanted sensor provided an optimal data feed for the athlete to finish his race without major problems caused by energy shortcomings or glucose excursions, and especially without moderate or major hypoglycemic reactions. Individualized diabetes care, including insulin adjustment and the combined support of a diabetologist, a sport medicine specialist, and the escort team, allowed for a successful race. A comparable team effort for endurance sport activities based on physiological background and experience has been reported and recommended in other reviews (18–20).

This article documents the case of the first individual with type 1 diabetes to finish the RAAM as a solo athlete. It is also the first reported use of an implanted glucose sensor in an extreme endurance sport activity (21–23). The use of diabetes technology, including the combination of an insulin pump and a CGM system, has the potential to increase self-assurance and safety for individuals participating in extreme sports (24–27) and level the playing field for people living with type 1 diabetes in sports dominated by athletes without diabetes.

The authors thank Prof. Dr. Tom Elliott, BC Diabetes, Vancouver, Canada, for his editing assistance. T.H. provided written consent to publish the data on his RAAM and his photograph in this article.

R.W. is an advisory board member and lecturer for Roche Diabetes Care Austria and Abbott Diabetes Care Austria. No other potential conflicts of interest relevant to this article were reported.

R.W. provided diabetes care for T.H., and H.O. and S.F., with their sports medicine expertise, provided training and nutritional advice for the race. S.F. accompanied the athlete on the race, taking responsibility for medical support and care. All of the authors collected data, and R.W. wrote the manuscript, with contributions from H.O. and S.F. R.W. is the guarantor of this work and, as such, had full access to all the data reported and takes responsibility for the integrity and accuracy of the case study presented.

This article contains supplementary material online at https://doi.org/10.2337/figshare.17139353.