Edited by: Ondrej Šeda, Charles University, Czechia
Reviewed by: George A. Brooks, University of California, Berkeley, United States; Heikki Olavi Tikkanen, University of Eastern Finland, Finland; Jan Brož, Charles University, Czechia
This article was submitted to Diabetes, a section of the journal Frontiers in Endocrinology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Studies have shown that individuals with diabetes type 1 have the same aerobic capacity as healthy individuals, provided that blood glucose (BG) is maintained in the euglycemic range (
In healthy subjects, new guidelines are available (
For individuals with diabetes type 1 there are currently no recommendations concerning insulin adjustments and/or the CHO-intake during prolonged endurance sports when physical performance is to be considered. In the absence of such recommendations, the individuals must adopt a trial-and-error approach based on their past experience of BG responses to similar activities (
In healthy individuals an increased intake of carbohydrates (CHO-loading) during the days before prolonged PE (>1.5 h) was shown to improve the physical performance (
A common strategy prior to the prolonged PE in individuals with diabetes type 1 is to reduce the basal insulin doses by about 30–80% to reduce the risk of hypoglycemia. However, during competition, problems with hyperglycemia is often seen and is due to stress-induced release of glucose elevating stress hormones such as adrenaline and noradrenaline (
Furthermore, during prolonged PE, the carbohydrate intake has primarily been governed by the current blood glucose value or trend in previous studies (
With technological developments in recent years it is not only possible to accurately assess glucose control but also to take preventive actions. Real-Time Continuous Glucose Monitoring (rtCGM) has been shown to improve glycated hemoglobin (A1c) as well as reducing the risk of hypoglycemia (
The aim of this study was to evaluate CHO-loading prior to and intermittent high carbohydrate intake during prolonged PE in individuals with type 1 diabetes and its impact on glycemic control when applied along with rtCGM.
Physically active individuals with type 1 diabetes from different regions of Sweden reported their interest to participate in the study and 10 individuals were selected.
Inclusion criteria: age 18–50 years, type 1 diabetes with a diabetes duration >1 year, exercising regularly ≥3 workout/week, previous experience from long-distance cross-country skiing and willingness to follow the study according to the protocol.
Exclusion criteria: A1c 8.6% NGSP (70 mmol/mol IFCC) proliferative retinopathy, known nephropathy or cardiac failure, and presently following a low-carbohydrate diet.
This is a descriptive, single arm, non-randomized interventional study. The intervention consisted of CHO-loading prior to and intermittent high CHO-intake during PE and a proactive use of rtCGM to achieve and maintain glucose control. A 3-day long sports camp was performed 1.5 months before a 90 km cross-country skiing race (Vasaloppet). The objective of this sport camp was to educate and prepare the participants on CHO-loading and the extended use of CHO during PE. All participants used rtCGM during the sports camp, the time between the sports camp and the Vasaloppet as well as during the race.
All participants conducted a 2-day CHO-loading twice, first during the time interval between the sports camp and the Vasaloppet and later during the 2 days prior to the race. The first occasion was used as an exercise in adjusting the insulin doses. The CHO-loading procedure consisted of the usual diet extended by 2 g of CHO/kg/day for 2 days in the form of a sports drink mixed in 1 liter of water which was ingested, 0.4 dl/h every half hour, intermittently for 12 h (08:00 a.m.−08:00 p.m.). The extended CHO-intake was simultaneously balanced with an increased amount of basal insulin where the individual carbohydrate to insulin ratio was used to find the appropriate dose to add as a basal dose during the subsequent 12 h. Furthermore, before bedtime after the start of CHO-loading the basal insulin was increased by ~20% day 1 and 30% during day 2 lasting throughout the night until breakfast on both nights. The participants on multiple daily injection (MDI) therapy administered the long-acting insulin twice daily with a 50–50% distribution morning and evening. During the CHO-loading, the long-acting insulin taken in the evening was increased by 20% prior to night 1 and by 30% prior to night 2. Extra bolus doses were used during daytime corresponding to the added carbohydrate amount.
To avoid the stress-induced hyperglycemia prior to the start of the Vasaloppet, the participants practiced to deliberately increase the mealtime insulin dose prior to PE during preparations prior to the race aiming to reach a glucose target of 90–144 mg/dl (5–8 mmol/l) at the start of PE. The breakfast was consumed at least 2 h before the start of PE and the participants were given a breakfast they were used to consume in relation to PE. The participants were informed to consume extra CHO (about 20–25 g) in the form of a sport drink, or to use a 30-min pump suspension, if the glucose level tended to decrease toward hypoglycemia before start, thus using the rtCGM proactively.
During all exercise sessions, the participants consumed a glucose–fructose containing liquid, three times per hour. The use of this mixed glucose-fructose sports drink enabled the extended use of CHO, corresponding to 1.00 ± 0.15 g CHO/kg body weight. The participants with type 1 diabetes were instructed to balance this CHO-intake with an appropriate insulin dose aiming at glucose values in range 72–180 mg/dl (4–10 mmol/l).
The Dexcom G5 Platinum system (Dexcom, San Diego, CA) was used during the study period. The sensors were inserted and calibrated according to company recommendation. The rtCGM devices and insulin pumps were downloaded via Diasend (Glooko Inc., Mountain View, CA) downloading system. HemoCue Glucose 201 RT (HemoCue, Ängelholm, Sweden), glucose measuring range 0–180 mg/dl (0–24.6 mmol/l), coefficient of variation (CV) 1.8%, was used for calibration of the rtCGM. All participants utilized the information they received via rtCGM but received no further instructions on glucose management during the race.
Each participant was informed to take proactive actions against hyperglycemia, >180 mg/dl (>10 mmol/l) as well as against hypoglycemia, <72 mg/dl (<4 mmol/l).
Hyperglycemia: Glucose value >180 mg/dl (>10 mmol/l) and stable alternatively increasing trend—Action: Bolus correction aiming at 108 mg/dl (6 mmol/l) taking into account a doubled Insulin Sensitivity Factor (ISF) during exercise.
Hypoglycemia: Glucose value 86.4–102.6 mg/dl (4.8–5.7 mmol/l) + arrow trend obliquely downwards, 102.6–117 mg/dl (5.7–6.5 mmol/l) + single arrow downwards and 117 mg/dl (>6.5 mmol/l) + double arrow downwards—Action: Immediate response by extra CHO-intake (about 20–25 g) or to use a 30-min insulin pump suspension.
The statistical package for the social sciences (SPSS) version 17.0 (SPSS Inc., Chicago, IL) was used for statistical analysis. Glucose values are presented as mean ± standard deviation unless otherwise indicated.
Ten individuals, mean age 36.3 ± 4.9 years (range 31–43), mean diabetes duration 15.3 ± 10.9 years (range 1.6–30) and the mean A1c was 7.2% (55 mmol/mol) were included. The baseline characteristics of the study population is shown in
Characteristics of study population.
Number ( |
10 |
Gender (female/male), number | 2/8 |
Age (years), mean ± SD (range) | 36.5 ± 4.9 (27–43) |
BMI (kg/m2), mean ± SD (range) | 24.7 ± 2.6 (20.1–27.8) |
Weight (lb), mean ± SD (range) | 180.6 ± 26 (143.3–233.7) |
Weight (kg), mean ± SD (range) | 81.9 ± 11.8 (65–106) |
Diabetes duration (years), mean ± SD (range) | 15.3 ± 10.9 (1.6–30) |
Treatment regimen (CSII/MDI), number | 7/3 |
Total daily dose of insulin (IU/kg), mean (range) | 0.47 (0.17–0.81) |
A1C (NGSP, %), mean (range) | 7.2 (6.3–8.3) |
A1C (IFCC, mmol/mol), mean (range) | 55 (45–67) |
During the CHO-loading prior to the Vasaloppet, one participant had problems with the rtCGM equipment during day 1 resulting in missing data. Mean glucose during the 2-day CHO-loading were: for both days; 129.6 ± 43.2 mg/dl (7.2 ± 2.4 mmol/l), day 1; 140.4 ± 45.0 mg/dl (7.8 ± 2.5 mmol/l) and day 2; 120.6 ± 41.4 mg/dl (6.7 ± 2.3 mmol/l) (
Mean glucose levels measured by rtCGM in individuals with type 1 diabetes during a 2-day carbohydrate loading prior to Vasaloppet.
A | 153 (8.5) | 59 (3.3) | 40 (2.2) | 277 (15.4) | 140 (7.8) | 56 (3.1) | 40 (2.2) | 259 (14.4) |
B | 126 (7.0) | 49 (2.7) | 40 (2.2) | 275 (15.3) | 101 (5.6) | 40 (2.2) | 40 (2.2) | 220 (12.2) |
C | 113 (6.3) | 22 (1.2) | 47 (3.6) | 173 (9.6) | 106 (5.9) | 23 (1.3) | 52 (2.9) | 185 (10.3) |
D | 155 (8.6) | 31 (1.7) | 88 (4.9) | 232 (12.9) | 142 (7.9) | 43 (2.4) | 70 (3.9) | 245 (13.6) |
E | 95 (5.3) | 31 (1.7) | 40 (2.2) | 194 (10.8) | 94 (5.2) | 34 (1.9) | 41 (2.3) | 230 (12.8) |
F | 162 (9.0) | 49 (2.7) | 56 (3.1) | 286 (15.9) | 149 (8.3) | 34 (1.9) | 68 (3.8) | 212 (11.8) |
G | 180 (10.0) | 67 (3.7) | 49 (2.7) | 297 (16.5) | 119 (6.6) | 45 (2.5) | 54 (3.0) | 268 (14.9) |
H | NA | NA | NA | NA | 146 (8.1) | 61 (3.4) | 59 (3.3) | 238 (13.2) |
I | 155 (8.6) | 36 (2.0) | 45 (2.5) | 236 (13.1) | 119 (6.6) | 41 (2.3) | 47 (2.6) | 218 (12.1) |
J | 131 (7.3) | 67 (3.7) | 40 (2.2) | 326 (18.1) | 117 (6.5) | 52 (2.9) | 43 (2.4) | 265 (14.7) |
All | 140 (7.8) | 45 (2.5) | 50 (2.8) | 256 (14.2) | 121 (6.7) | 41 (2.3) | 47 (2.6) | 234 (13.0) |
Glucose values in individuals with type 1 diabetes during 2 days of carbohydrate loading and during a 90 km cross-country skiing race (Vasaloppet).
Time in range (% of total time): 72–180 mg/dl (4–10 mmol/l) | 70.7 | 78.3 | 74.7 | 100 | 94.3 |
Time spent in hypoglycaemia (% of total time): (<72 mg/dl) (<4 mmol/l) | 9.9 | 10.8 | 10.4 | 0 | 0.6 |
Time spent in hyperglycemia (% of total time): (>180 mg/dl) (>10 mmol/l) | 19.4 | 10.9 | 14.9 | 0 | 5.2 |
The mean insulin bolus dose before the race was increased by 55.5%, from the calculated dose of 5.4 ± 3.0 IU to 8.4 ± 4.0 IU. Mean sensor glucose levels at the start (07:00 a.m.) of the race was 126.0 ± 25 mg/dl (7.0 ± 1.4 mmol/l), with a range of 81–157 mg/dl (4.5–8.7 mmol/l) (
Mean glucose levels, carbohydrate intake, and adjustment of basal insulin in individuals with type 1 diabetes during a 90 km cross-country skiing race (Vasaloppet) and the duration of the physical activity (A–J).
A | MDI | 94 (5.2) | 86 (4.8) | 8 (0.5) | 74–99 |
90 | 0 | 05:33 |
B | MDI | 140 |
138 |
9 |
119–158 |
75 | 0 | 05:21 |
C | CSII | 104 |
129 |
33 |
86–194 |
75 | 0 | 07:00 |
D | MDI | 157 |
110 |
22 |
77–196 |
75 | 0 | 08:23 |
E | CSII | 153 |
138 |
18 |
99–162 |
84 | 0 | 05:54 |
F | CSII | 119 |
150 |
26 |
104–184 |
84 | 0 | 06:45 |
G | CSII | 81 |
141 |
31 |
81–187 |
100 | 0 | 06:24 |
H | CSII | 142 |
132 |
39 |
68–203 |
75 | 0 | 07:20 |
I | CSII | 128 |
128 |
21 |
85–158 |
75 | +27 | 06:02 |
J | CSII | 139 |
126 |
38 |
83–220 |
75 | Start: 0 |
06:07 |
All | 126 |
128 |
25 |
68-220 |
80.8 | 06:28 |
Mean sensor glucose during the race was 127.8 ± 25.2 mg/dl (7.1 ± 1.4 mmol/l) (
Five of the participants experienced hyperglycemia during the race with a max sensor glucose value 220 mg/dl (12.2 mmol/l) and one subject had a hypoglycemia with a nadir sensor glucose value of 68 mg/dl (3.8 mmol/l). Each participant's glucose graph is shown in
Continuous glucose monitoring graphs of 10 individuals with type 1 diabetes during the 90 km cross-country skiing race (Vasaloppet).
The time needed to complete the race, CHO-intake and basal insulin dose adjustments are illustrated in
In this study we have shown that it was possible to achieve and maintain good glycemic control, even during extraordinary challenges such as 2 days of CHO-loading followed by high intermittent CHO-intake during a 90 km long cross-country skiing race. A proactive use of rtCGM enabled individual insulin dose adjustments under these conditions.
To our knowledge few studies have been published regarding models for carbohydrate loading in type 1 diabetes individuals and the same also applies for the use of high carbohydrate intake during prolonged physical exercise. At the same time, we want to emphasize that attempts have been made exploring this area, for example via projects as “Team Novo Nordisk Pro Cycling,” but published data are limited or missing.
In the study by McKewen et al. (
Prior to competitions, stress is common, causing increased release of adrenaline and noradrenaline (
Both hypoglycemia and hyperglycemia are important to avoid during PE when performance also is regarded as an important factor. Hyperglycemia increases the release of free fatty acids (FFA) (
Despite the prolonged duration of the PE in the current study, the participants spent 94.3% of the TIR. Our results differ in comparison to a previous study in individuals with type 1 diabetes who performed a 75 km cross-country skiing race (>7 h) at two consecutive years (1986 and 1987), where a CHO-intake corresponding to 40 g of CHO/h appeared to prevent hypoglycemia in the majority of the participants (8 of 9) in both competitions (
A possible mechanism behind the good results in our study could be the combination of carbohydrate loading and subsequent high carbohydrate intake during prolonged physical exercise. This procedure ensured a relatively high glycogen content in the liver and muscles prior to exercise. Exercise performance/capacity and a stable plasma glucose level is often limited by endogenous carbohydrate availability during prolonged physical activity (
Gastrointestinal discomfort is very common symptom during exercise, especially in prolonged endurance races (
This study was performed as an exploratory study which includes the benefits of this being a real world situation. As opposed to this, there are of course also limitations as a control group is missing and where the environment made it difficult to carry out parallel sampling to evaluate the mechanisms behind the good glucose control achieved in the study.
Furthermore, the participants was not randomly selected and could thus limit the generalisability of the study. Vasaloppet is a 90 km long cross-country skiing race which is very demanding in terms of individual physical performance and the participants had to have relative high level of fitness to be able to complete the race. Therefore, the results could be seen as a description of real-world data in this specific group.
We conclude, that high intermittent CHO-intake during prolonged PE was associated with good glucose control in individuals with type 1 diabetes. However, the proportion of time spent in hypoglycemia during the 2-days of CHO-loading was 10.4% and a lower insulin dose might have been required to reduce time spent in hypoglycemia. rtCGM could be beneficial when used proactively to maintain sensor glucose values within target range before and during PE. These strategies and the mechanisms that create the conditions for good glucose control during prolonged physical exercise needs to be further evaluated in randomized controlled studies.
All procedures performed in this study involving human participants were in accordance with the ethical standards of the national research committee and with the 1964 Helsinki declaration and its later amendments. Signed informed consent was collected from all participants prior to study start. This study was approved by the Regional Ethical Review Board in Uppsala, Sweden (DNR: 2012/159).
SM and PA conceived and designed research, conducted the experiments, and analyzed data. JJ participated in the planning of the study. SM, PA, and JJ did all participate during the preparatory sports camp. SM wrote the manuscript. PA and JJ reviewed the manuscript. All authors read and approved the manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We would like to thank all participants in the study providing us with new knowledge.
The Supplementary Material for this article can be found online at:
Glycated Hemoglobin
Blood Glucose
Body Mass Index
Carbohydrates
Continuous Subcutaneous Insulin Infusion
International Federation of Clinical Chemistry
Multiple Daily Injections
National Glycohemoglobin Standardization Program
Physical Exercise
Real-Time Continuous Glucose Monitoring
Time in range.