Participants:
Men with T1D were recruited directly from the Waikato Regional Diabetes Service (WRDS, Hamilton, New Zealand) between September and November 2019. To be eligible, all participants needed to have had T1D for at least two years, be aged 18 - 60 years, have a recent HbA1c (within the last 3 months) of 45 - 90 mmol/mol (6.3 – 13.9%); be able to walk/run unaided for 45 minutes at a speed of 7.24 km.h-1 and be fully competent with diabetes self-management. Participants were excluded if they were taking testosterone therapy, had any comorbidities that would impact on their ability to complete the study (e.g. cardiovascular disease, cerebrovascular disease, severe diabetic retinopathy, nephropathy, neuropathy); were taking medications that altered heart rate (e.g. beta-blockers); or were currently using closed-loop insulin pump therapy. All potential participants were screened at the WRDS Clinic and upon recruitment were required to attend the clinic again 38-48 hours prior to the first test session to have a Flash Glucose monitoring (FGM) device (Abbott Freestyle Libre, Chicago, IL) inserted to allow for continuous glucose monitoring [12]. All participants were able to wear the Freestyle Libre device for 14 days from the date of insertion. The study was successfully registered under the Australia New Zealand Clinical Trial Registry (ACTRN12619001397101p) and was approved by the New Zealand Health and Disability Ethics Committee (ref 19/NTB/175).
Study Design:
This study was a randomised, counterbalanced, crossover design, in which men with T1D evaluated two different interventions for managing glycaemia whilst undertaking a pre-defined exercise protocol. The two sessions were carried out a week apart at the University of Waikato gymnasium in a controlled, indoor environment with meals (breakfast and lunch) provided. The sessions were conducted by a research team consisting of a diabetologist/endocrinologist, a clinical research specialist, a clinical nurse, a sport physiologist and a health research student. The night before each session, participants were required to consume a meal that contained at least 1 g of carbohydrate/kg of body weight [6], and they were asked to eat the same meal before both intervention sessions. Consumption of alcohol, caffeine-based beverages and performing strenuous exercise were prohibited during the 12 hours before each aerobic exercise session. Participants then fasted following dinner on the previous day until 9am the following morning.
The two interventions assessed were i) reducing their insulin (INS) or ii) consuming carbohydrates without the usual insulin bolus (CARBS)) along with the breakfast meal. The same breakfast meal was provided at 9 am in both groups, and consisted of bananas, apple juice and 100 % Isolate protein powder (Musclepharm COMBAT, New Zealand), with quantities individualised to contain 0.66 g of carbohydrate/kg of body weight and 3.375 kcal/kg of body weight, meeting both the requirements of exercise carbohydrates guidelines [6] and calculated energy expenditure using the Browning Walking Metabolic Prediction equation [13]. Dietitian advice was sought to ensure all meals, including the meal the night before the exercise, met the required recommendations. The accurate intake of breakfast was observed by the investigators and all participants confirmed that they had consumed the recommended meal the night prior.
Participants were either on multiple daily injections of insulin (MDI) or continuous subcutaneous infusion of insulin (CSII; insulin pump therapy). All patients on MDI were on a long-acting basal insulin (insulin glargine) and rapid-acting bolus insulin (either insulin aspart or insulin lispro). All patients on insulin pump therapy used a continuous infusion of insulin aspart. During the INS intervention, participants on MDI did not alter their basal insulin, but halved their bolus insulin of their normal dose immediately before consuming the meal as a single bolus as per the consensus guidelines [7]. Those participants in the INS intervention on CSII started a 50% temporary basal rate reduction 90 minutes before aerobic exercise and halved their normal bolus dose for their meal. The 50% temporary basal rate reduction stopped 60 minutes after completion of the aerobic exercise [7]. The participants in the CARBS intervention consumed their breakfast meal without the administration of any bolus insulin and did not alter their basal insulin [6]. Throughout the test sessions the participants were not allowed to further adjust their insulin doses or carbohydrates intake. Any adjustments (required or accidental) would trigger the cessation of the session for this participant.
All participants began the aerobic exercise at 10.00 am and were required to walk on a motorised treadmill (Life Fitness, Chicago, IL; USA) at a fixed speed of 7.24 km.h-1 (4.5 miles per hour) at a 1% gradient for 45 minutes. According to the ACSM classification, this would be classified as “vigorous” intensity exercise. Ten minutes following completion of the treadmill walk, participants then performed the 6MWT [14] by walking as quickly as possible for six minutes around the perimeter of an inside arena. This test is considered as a validated measure of predicted VO2 max, and this was calculated using the equation as follows [15]. VO2 max (mL⋅kg−1⋅min−1) = 70.161 + (0.023 × six-minute walk distance covered [m]) – (0.276 × weight [kg]) – (6.79 × sex, where male = 0, female = 1) − (0.193 × resting heart rate [BPM]) – (0.191 × age [years]). For the 6MWT, distances were marked off every 3m, and the total distance walked in the six minutes was recorded. Approximately sixty minutes after completion of the exercise activities, participants had lunch (One Square Meal, OSM: Queenstown, New Zealand), individualised to consist of 1 g of carbohydrate/kg of body weight and administered their normal dose of bolus insulin [7]. Water was allowed ad libitum throughout each session and no additional adjustments (including self-adjustment) of insulin or carbohydrate were allowed until three hours after lunch.
Capillary blood glucose was initially measured before breakfast. Capillary glucose and lactate were measured immediately at the start of the 45-minute aerobic exercise, every 15 minutes during the aerobic exercise, at the end of the 45-minute aerobic exercise, within 1-3 minutes after completion of the 6MWT [16] and before lunch. Blood ketone level measurements were taken at the beginning and the end of the 45-minute aerobic exercise and within 1-3 minutes after completion of the 6MWT. Blood lactate was measured using a lactate analyser (Lactate Pro 2, Arkray, Japan). Glucose levels were monitored directly using the FGM device, and also via capillary sampling (Abbott Optium glucose strips) to minimise any delay between the interstitial and blood glucose levels during physical activity [12]. Self-perceived rate of exertion (RPE) was also assessed every 15 minutes during the aerobic exercise, and immediately following the 6MWT using the Borg’s RPE scale [17]. After completion of both sessions’, participants were asked which intervention, if either, they preferred for future exercise.
Statistical Analysis:
Descriptive statistics and data are expressed as mean ± standard deviation (SD). Paired T-tests (two-tailed) were performed on blood glucose levels, blood ketone, blood lactate, RPE, distance covered and the predicted VO2 max for the two interventions. Repeated measures ANOVA was performed to compare time in the normal blood glucose range (primary outcome), time in hyperglycaemia and time in hypoglycaemia for INS and CARBS. Thresholds for normal range or normoglycaemia were 3.9 – 10 mmol/L; mild hyperglycaemia between 10.1 – 13.9 mmol/L; severe hyperglycaemia ≥ 14 mmol/L; mild hypoglycaemia < 3.9 mmol/L and significant hypoglycaemia (< 3 mmol/L).The time in each range was calculated from the FGM but importantly there were no significant differences or lag between capillary and interstitial glucose levels in any participants. Statistical significance was accepted at a level of P < 0.05.