Impact of High-Intensity Interval Exercise With Elastic Bands Versus Continuous Moderate-Intensity Aerobic Exercise on Glycemic Control in People With Type 1 Diabetes

Abstract

Background:

Engaging in physical exercise is recommended to enhance cardiovascular health and manage blood sugar levels in people with type 1 diabetes (T1D).

Hypothesis:

The impact of high-intensity interval exercise with elastic bands (EB-HIIE) versus continuous moderate-intensity aerobic exercise (CONT) on glycemic control is different in men with T1D.

Study Design:

Crossover study design.

Level of Evidence:

Level 3.

Methods:

Participants (39 men with T1D) underwent either an EB-HIIE or a CONT session in randomized order, with a separation of ≥72 hours to avoid carry-over effects. Changes in glucose values during exercise were measured simultaneously from venous blood (YSI) and interstitial fluid (Dexcom G6 glucose sensor). Subsequent 24-hour glucose was monitored using the glucose sensor.

Results:

Blood glucose was lower in CONT vs EB-HIIE (P < .01). Post hoc analysis revealed clinically relevant differences during exercise (–35.1 mg/dl; P = .02), at its end (–49.5 mg/dl; P < .01), and at 10 and 20 minutes after completion (–51.2 mg/dl; P < .01 and −45.9 mg/dl; P < .01, respectively). Time-in-range 24 hours after exercise completion was significantly higher with EB-HIIE than with CONT (66.5% vs 59.3%), although both were significantly better than the previous 24 hours before exercise (50%).

Conclusion:

Results suggest that EB-HIIE is a safe training method for male adults with diabetes, resulting in euglycemia during and immediately after exercise and improving glucose outcomes in the subsequent 24 hours.

Clinical Relevance:

This study provides new evidence and practical information on how to implement safe physical activity in daily life of patients with diabetes. EB-HIIT exhibited lower hypoglycemia risk during exercise and better glycemic control in the subsequent 24 hours. In contrast, practicing CONT exercise is associated with higher risk of hypoglycemia. Healthcare providers should take this information into account when prescribing exercise.

Diabetes mellitus stands as one of the most prevalent and serious metabolic diseases, giving rise to chronic complications that significantly diminish both the quality of life and life expectancy of affected people, while imposing substantial costs on healthcare systems. ¹⁶ Current guidelines for diabetes management underscore the significant benefits of physical activity, as it is associated with a substantial reduction in cardiovascular and overall mortality risks.^6,31 The general recommendation is to incorporate various types of training, including ≥150 minutes of moderate-intensity (or 75 minutes of vigorous intensity) aerobic exercise spread over at least 3 days per week, and 2 to 3 sessions per week of resistance exercise on nonconsecutive days.^6,9

However, the implementation of these recommendations in everyday life encounters at least 2 barriers. The first, common to most people, is the constraint of time for exercising. ³⁴ Adherence to programs involving continuous moderate-intensity aerobic exercise (CONT), ²⁷ remains low, with the minimum recommended duration of 30 minutes of CONT rarely achieved. ⁶ The second barrier, specific to people with type 1 diabetes (T1D) is the exercise-induced glucose variability and the increased risk of hypoglycemia during and after exercise.^3,32

An option to enhance adherence to exercise and potentially overcome the aforementioned barriers is engaging in high-intensity interval exercise (HIIE). ³³ HIIE requires less time commitment to achieve similar or even greater cardio-metabolic effects compared with CONT, even in patients with T1D. ³⁰ Moreover, HIIE has generally demonstrated superiority over CONT in terms of inducing lower glycemic variability. ³⁰ Nevertheless, data on its effects in people with T1D are still limited and obtained primarily using cycle ergometers or treadmills protocols.^{2,10,13,26,29,30} Although these protocols are valuable in research, their applicability in daily life is constrained due to the expensive and inaccessible equipment required.

Some previous studies have proposed HIIE using elastic bands (EB-HIIE) as a more accessible and cost-effective training alternative suitable for both research and home use. ²¹ In addition, EB-HIIE incorporates resistance training, combining cardiovascular and strength training, potentially offering an advantage over modalities based solely on cyclical movement (eg, running or cycling). Finally, EB-HIIE has not yet been compared in terms of safety, particularly in preventing hypoglycemia or ensuring short-term glucose control compared with CONT. Therefore, the objective of this study is to compare the efficacy of glucose control and safety between EB-HIIE and CONT in people with T1D.

Methods

Study Design

Patients with T1D underwent a graded exercise test (GXT) to determine their CONT power. Subsequently, participants engaged in either EB-HIIE or a CONT session, randomized using a number system and spaced ≥72 hours apart to avoid interactions. All exercise sessions were conducted around 6:00 p.m. in the Clinical Research Laboratory of the Physiotherapy Department at the University of Valencia, Spain, supervised by a physiotherapist with expertise in this exercise methodology.

To compare the effects of EB-HIIE versus CONT, we employed a repeated measures design focusing on glucose values. The experimental protocol received approval from the Ethics Committee of the University of Valencia (Ethics Committee, no. 1587001) and it was registered retrospectively at https://clinicaltrials.gov/ (NCT06080542).

Patients

Inclusion criteria were (1) age between 18 and 40 years old; (2) T1D with a disease duration >2 years; (3) baseline hemoglobin A1c <8.5%; (4) only patients on multiple daily injections; and (5) engage in weekly physical activity of ≥90 minutes, without participating in any sport as an amateur or professional.

Eligible patients were excluded due to clinical conditions or the use of medications (other than insulin) known to affect glycemic control. Patients were provided with information regarding the potential benefits and risks of the project, and all participants signed a dedicated informed consent form. On the initial day, age, weight, height, and information on regular exercise activity for each patient, assessed using the International Physical Activity Questionnaire Short Form (IPAQ), were recorded.

Procedures

The GXT was conducted on a cycle ergometer to determine the individual working power for the subsequent CONT session. The test began with a 3-minute period of rest (0 W), followed by a 3-minute warm-up cycling at a workload of 60 W. Workload increased by 40 W every 3 minutes until volitional exhaustion. ¹⁹ Active recovery at 60 W for 3 minutes on the cycle ergometer was followed by a 3-minute passive recovery (0 W). LTP1 and the power at that level were determined to prescribe CONT intensity using the 0.5 mM method, ie, the point that precedes the rise in blood lactate concentration of >0.5 mM above baseline.^7,19 Capillary blood samples were collected from the earlobe at rest, in the last 10 seconds of each GXT step, and at the end of the test to measure lactate concentration (Lactate Pro 2, Arkray). Heartrate was measured continuously via chest belt telemetry using a Polar H10 heart rate monitor during GXT and both CONT and EB-HIIE sessions. An hour after the GXT, patients had an acquaintance session with strength exercises using elastic bands, performing 2 sets of 12 repetitions of the exercises.

The CONT session began with a 3-minute resting period (sitting quietly on the cycle ergometer at 0 W), followed by a 3-minute warm-up at 60 W. The intensity increased stepwise by 20 W/min, reaching the power at LTP1, which was maintained for 30 minutes. Active and passive recovery periods were the same as in the GXT. ²⁵

The EB-HIIE session was a modification of previous protocols,^23,28 using TheraBand CLX (Hygenic Corporation) instead of bodyweight exercises. ¹⁸ Four upper limb exercises (bench press, seated dumbbell, biceps curl, and seated row) and 4 lower limb exercises (squats, stiff-legged deadlifts, hamstring curl exercise, and quadriceps curl exercise) were chosen. Resistance band strength was selected based on a pilot study ¹ : golden band, which provides a resistance of 78 N at 100% of its elongation, for quadriceps extension and femoral curl and bench press; silver band, which provides a resistance of 60 N at 100% of its elongation, for seated dumbbell, bicep curl, and seated row; and combined gold and silver bands for the rest. Exercises were programmed to avoid 2 consecutive exercises targeting the same area. ²¹ The warm-up included 3 minutes at 60 W on the cycle ergometer and 15 no-load shoulder flexo-extensions. The session comprised 2 series of exercise intervals (20 seconds on, 10 seconds off) with a rest of 3 minutes between series. Participants chose the band width for maximum effort, aiming for correct form. Patients were encouraged to perform as many repetitions as possible while maintaining proper form, so the number of repetitions per set depended on each exercise and participant.

Plasma glucose was measured at 10-minute intervals, starting 20 minutes before each session and ending 20 minutes after exercise. Continuous subcutaneous glucose measurements were obtained during and 24 hours after exercise using the Dexcom G6 continuous glucose monitoring (CGM) system. The Dexcom G6 sensor was inserted 72 hours before commencement of the initial exercise session and was kept in place for a minimum of 48 hours postexercise. A catheter was inserted into the forearm for venous blood sampling during the observation period. Glucose and lactate were determined using the YSI 2500.

If glucose was ≤60 mg/dl, GXT or exercise session initiation or continuation was halted to prevent hypoglycemia. ¹¹ In cases of mild hypoglycemia (60-69 mg/dl), patients received 200 ml of orange juice (containing 10.4 g carbohydrate/100 ml), and blood glucose was rechecked after 10 minutes.

During the observation period, patients maintained their usual insulin doses and meals. They were instructed to avoid fast-acting insulin dosage 3 hours before each session to minimize the impact of circulating insulin during exercise.

Statistical Analysis

Plasma glucose (YSI) measurements underwent a 2-way repeated samples analysis of variance with exercise type (CONT vs EB_HIIE) and study period (before, during, and after) as within-subjects factors. Two measurements were taken before exercise (−20 minutes and −10 minutes), 1 at the beginning, 1 during exercise (in CONT it was at 20 minutes and in EB-HIIE it was between the 2 series), at the end of exercise, and after the completion of the session (10 minutes and 20 minutes after end of exercise), of both types of exercise (CONT or EB-HIIE). Post hoc comparisons were carried out to obtain specific differences on significant interaction terms. The level for significance was set at P < .05 (with Bonferroni adjustments for multiple comparisons to maintain an overall type I error rate at 5%). Effect size was evaluated with η² (partial eta-squared), where 0.01 < η² < 0.06 constitutes a small effect, a medium effect being 0.06 < η² < 0.14, and a large effect being η² > 0.14. ⁵

To compare the effects of each type of exercise on lactate, paired t tests were used to examine at baseline the differences in lactate between exercises. In addition, paired t tests were used in each exercise to analyze the differences between the lactate at the start and the end both in CONT and EB-HIIE. Finally, paired t tests were used to detect between exercises differences in lactate, using pre- to postdifferential value as dependent variable. Cohen’s d was also calculated to evaluate the effect size (d < 0.2, trivial; 0.2-0.5, small; 0.5-0.8, medium; and >0.8, large). ⁵

For effects on glucose concentrations in the 24-hour period after exercise, mean glucose, glycemic variability (measured using the coefficient of variation [CV]) and the time spent at each glucose range (euglycemia [70-180 mg/dl] or time in range [TIR], level 1 or mild hypoglycemia [54-70 mg/dl], level 2 or severe hypoglycemia [<54 mg/dl], level 1 hyperglycemia [180-250 mg/dl] and level 2 hyperglycemia [>250 mg/dl]) were evaluated with Dexcom G6 for 6 h, 12 h, and 24 h after both exercise sessions. Glucose before exercise was monitored using samples from a Dexcom G6 corresponding to a window of 24 hours before exercise that corresponded exactly to the clock times of the patient in the 24 hours after exercise.

CGM data were formatted and preprocessed with MATLAB R2022a (MathWorks Inc). CGM reported P values are 2-sided and were not adjusted for multiple testing. All glycemic metrics were calculated for each subject, observation window, and exercise session. One-sample Kolmogorov-Smirnov tests were applied to the data to demonstrate a normal distribution. If the null hypothesis of normality of the data could not be rejected, a 2-sample Student’s test was used to compute P values. If the normality test failed, a Wilcoxon Signed Rank test was used to determine the statistical significance of the metrics.

Sample size calculation was based on the difference in mean glucose, which was assumed to be similar to that of a recent pilot study. ²¹ Power analysis showed that ≥32 patients would be required to achieve 80% power at significance level of 0.05 to detect a change of 20 mg/dl in mean glucose. To account for patient attrition, the sample size was adjusted to be 40 overall, with 20 participants per branch.

Results

Demographic parameters for all patients are listed in Table 1. A CONSORT 2010 flowchart is reported in Figure 1.

Table 1.

Baseline patient characteristics

n	39
Age, y	29.5 (6.2)
Diabetes duration, y	15.2 (6.2)
Weight, kg	78.76 (13.79)
BMI, kg/m2	25.01 (3.61)
HbA1c, %	7.3 (0.9)
Insulin total daily dose, U/day	51.4 (16.1)
U/kg/day	0.65 (0.19)
Physical activity, METS	4013.5 (3358.0)

BMI, body mass index; Hb1Ac, hemoglobin 1Ac; METS, metabolic equivalents.

Data presented as mean (SD).

Figure 1.

CONSORT flow diagram.

Figure 2 and Table 2 present the YSI glucose measurements and the glycemic outcomes during and immediately after exercise for each type of exercise, corresponding to the trial window. Regarding comparisons between plasma glucose measurements (Figure 2), significant differences existed both between moments (P < .01; η2 = 0.29) and between types of exercise (P = .04; η ² = 0.10, respectively) In addition, there was a significant 2-way interaction between the measurement moment and the type of exercise (P < .01; η ² = 0.42), showing differences in the post hoc analysis in particular during exercise (–35.1 mg/dl; P = .02), at the end of the exercise (–49.5 mg/dl; P < .01), and at 10 and 20 minutes after finishing the exercise (–51.2 mg/dl; P < .01 and 45.9 mg/dl; P < .01, respectively). In addition, whereas there were no differences in plasma glucose measurements in EB-HIIE, CONT showed differences in pairwise comparisons between multiple pairs of moments. Importantly, CV, time in level 1, and level 2 hypoglycemia were significantly lower in EB-HIIE (Table 2), due mostly to differences at time “end of exercise,” +10, and +20 minutes. In CONT, 9 and 1 patients went into level 2 and level 1 hypoglycemia, respectively, in contrast with 0 and 1 patients for EB-HIIE.

Figure 2.

Glucose measurements during and immediately after exercise by CONT and EB-HIIT sessions. *Measured at 4 minutes of EB-HIIT (ie, at the end of the first series) and at 20 minutes of CONT. ^†Significant differences (P < .05) between CONT and EB-HIIT. CONT, continuous moderate intensity aerobic exercise; EB-HIIT, elastic band high intensity interval training.

Table 2.

Glycemic outcomes of the clinical trial period measured by YSI

	Mean	CV	<54, %	54-70, %	70-180, %	180-250, %	>180, %	>250, %
EB-HIIE	151.3	9.4	0	0.7	68.6	25.7	30.4	4.6
CONT	123.6	21.1	5	11.9	62.8	15.9	20	4.1
P values	.04	<.01	<.01	<.01	.51	.15	.23	.88

CONT, continuous moderate intensity aerobic exercise; CV, coefficient of variation; EB-HIIE, elastic band high intensity interval exercise.

As appreciated in Figure 3, at the start of both exercise sessions, patient lactate values were similar, at 1.40 ± 0.15 mmol/l for CONT and 1.24 ± 0.10 mmol/l for EB-HIIE. Both exercises presented significant differences at the end, with an increase of 2.51 ± 0.34 mmol/l (P < .01) for CONT and 8.71±0.41 mmol/l (P < .01 for EB-HIIE). These lactate increases were greater for EB-HIIE compared with CONT, with a 6.20±2.57 mmol/l difference between exercises (P < .01).

Figure 3.

Lactate measurements at start and end of exercise in CONT and EB-HIIT sessions. *Significant differences (P < .05) between CONT and EB-HIIT. CONT, continuous moderate intensity aerobic exercise; EB-HIIT, elastic band high intensity interval training.

CGM metrics are reported in Table 3. Statistical significance is appreciated for the 24-hour window: TIR was statistically better for the EB-HIIE than CONT (68.2% vs 59.2%), without incurring in more time in mild or severe hypoglycemia for the patients. CV was also significantly reduced for the post-EB-HIIE observed periods (4.2% absolute difference, 31.5% post-EB-HIIE vs 35.7% CONT).

Table 3.

CGM metrics for EB-HIIE and CONT exercise by window of study

	EB-HIIE	CONT	P values
	6 hours
Mean, mg/dl	168.4 ± 40.9	172.2 ± 45.2	.71
CV, %	24.7 ± 10.8	26.9 ± 11.1	.34
<54, %	0 [0, 0]	0 [0, 0]	.75
54-70, %	0 [0, 0]	0 [0, 0]	.56
70-180, %	61.1 [36.4, 88]	61.1 [35.1, 87.5]	.80
180-250, %	31.9 [9.23, 43.8]	16.4 [8.3, 32.9]	.22
>180, %	37.5 [9.23, 63.5]	34.7 [11.1, 64.9]	.94
>250, %	0 [0, 14]	4.2 [0, 33]	.14
	12 hours
Mean, mg/dl	164.9 ± 39.7	169.9 ± 42	.524
CV, %	28.3 ± 11.3	30.2 ± 12	.354
<54, %	0 [0, 0]	0 [0, 0]	.430
54-70, %	0 [0, 2.08]	0 [0, 2.43]	.643
70-180, %	61.4 [41.2, 78.4]	52.8 [42.2, 80.6]	.704
180-250, %	29.3 [9.73, 47.2]	16 [6.11, 49.1]	.456
>180, %	36.5 [16.7, 58.1]	46.2 [11.5, 57.7]	.848
>250, %	0.35 [0, 12.2]	4.53 [0, 25.3]	.266
	24 hours
Mean, mg/dl	153.75 ± 28.4	163.04 ± 32.2	.113
CV, %	31.53 ± 9.9	35.77 ± 10	.026
<54, %	0 [0, 0.17]	0 [0, 0.17]	.620
54-70, %	0.88 [0, 2.68]	1.22 [0, 4.85]	.285
70-180, %	68.2 [52.3, 79]	59.2 [45.7, 72.4]	.048
180-250, %	21.4 [11.8, 35.5]	22.9 [13.6, 33]	.457
>180, %	28.5 [15.2, 43.8]	36.9 [21.3, 47.8]	.098
>250, %	1.22 [0, 12.3]	6.54 [0, 20.1]	.09

CGM, continuous glucose monitoring; CONT, continuous moderate intensity aerobic exercise; CV, coefficient of variation; EB-HIIE, elastic band high intensity interval exercise.

Mean and CV are shown as mean ± SD. All other metrics shown as median [interquartile range].

The 24 hours in the pre- and postexercise periods are listed in Table 4 and illustrated in Figure 4. Both types of exercise showed a significant improvement in the TIR for the period postexercise in contrast to pre-exercise, and this effect was observed more clearly in the EB-HIIE exercise, as also listed in Table 3. Time <70 mg/dl was not significantly different for CONT or for EB-HIIE, and no statistically meaningful differences were appreciated for CV either. Mean average glucose value was reduced significantly by EB-HIIE, going down to 153.75 mg/dl from 169.77 mg/dl, suggesting a more profound impact of the EB-HIIE exercise on overall glucose control. Note that the 24 hours pre-exercise columns for both exercises are very similar, showing analogous baseline points for the metrics of the study (Table 2). The first hours after CONT showed a trend of recovery from hypoglycemia that was not observed in the 24 hours before exercise or in the EB-HIIE. After that recovery period, the 24 hours after CONT showed very little difference in the average values of glucose compared with the 24hpre data, which is similar to the results in Table 3. In the EB-HIIE, the mean average glucose value was lower in the 24-hour period after exercise than in the 24 hours pre-exercise.

Table 4.

Glycemic metrics for the 24-hour pre- and postexercise periods

	EB-HIIE			CONT
	24 hours pre-exercise	24 hours postexercise	P value	24 hours pre-exercise	24 hours postexercise	P value
Mean, mg/dl	169.8 ± 29.3	153.7 ± 28.4	<.01	168.44 ± 36.9	163 ± 32.2	.47
CV, %	31.7 ± 8.3	31.5 ± 9.95	.94	33.08 ± 9.96	35.8 ± 10	.07
G < 54, %	0[0, 0]	0 [0, 0.17]	.17	0 [0, 0]	0 [0, 0.17]	.66
54 < G < 70, %	0.69[0, 2.69]	0.88 [0, 2.68]	.76	1.39 [0, 3.13]	1.22 [0, 4.85]	.72
70 < G < 180, %	54.2[36.2, 60.3]	68.2 [52.3, 79]	<.01	50 [37, 65.5]	59.2 [45.7, 72.4]	.02
180 < G<250, %	25.7[16.2, 37.7]	21.4 [11.8, 35.5]	.13	19.4 [14, 27.3]	22.9 [13.6, 33]	.19
G > 180, %	36.1[20.1, 50.9]	28.5 [15.2, 43.8]	.10	31.3 [20.7, 41.5]	36.9 [21.3, 47.8]	.30
G > 250, %	6.94[2.08, 14.2]	1.22 [0, 12.3]	.35	5.56 [0.17, 18.5]	6.54 [0, 20.1]	.87

CONT, continuous moderate intensity aerobic exercise; CV, coefficient of variation; EB-HIIE, elastic band high intensity interval exercise; G, glycemic metric (expressed as percentage).

Mean and CV are shown as mean ± SD. All other metrics shown as median [interquartile range].

Figure 4.

Graphical comparison of the 24-hour pre- and 24-hour postexercise periods for CONT (top) and EB-HIIT (bottom). CONT, continuous moderate intensity aerobic exercise; EB-HIIT, elastic band high intensity interval training.

Discussion

The present study was designed to comprehensively assess the impact of EB-HIIE and CONT on glucose homeostasis during, immediately after, and in the subsequent 24 hours of exercise in active male adults with well-controlled T1D. Three important findings emerged: EB-HIIE is associated with minor glucose changes during, or immediately after, exercise compared with CONT; both exercises improved TIR in the 24 hours after exercise compared with the previous 24 hours; and EB-HIIE was associated with higher TIR values in the 24 hours postexercise compared with CONT.

Using elastic bands, HIIE demonstrated stable blood glucose values during exercise and immediately after its completion, with a TIR of 68.6% and minimal time spent in level 1 hypoglycemia (<1%). In contrast, CONT exhibited a more pronounced glucose-lowering effect during exercise, persisting up to 20 minutes later, with a TIR of 62.8% and a level 1 hypoglycemia of 16.9%. This reduction in glucose levels during a 30-minute CONT session aligns with findings from previous studies with the same exercise duration.^15,30 Moreover, these studies also reported that CONT had a greater hypoglycemic effect than HIIE,^15,30 which has been attributed to potential metabolic and hormonal responses counteracting declining blood glucose levels. ¹⁴

In terms of lactate levels, we observed a greater increase with HIIE compared with CONT (8.71 mmol/l vs 2.51 mmol/l). This may suggest that HIIE could have a hypoglycemia-protective effect by reducing insulin-mediated peripheral glucose uptake by skeletal muscles, ³⁵ inhibiting glycolysis, ⁴ and supporting hepatic glucose production via gluconeogenesis. ²⁴ Alternatively, the increase in catecholamine levels during HIIE might enhance hepatic glucose production and reduce insulin-mediated glucose uptake.^8,20

Regarding the long-term effects of both exercises on glucose homeostasis, we observed an improvement in TIR up to 24 hours after exercise compared with the 24 hours before exercise, both with CONT (48.9% before, 59.3% after) and with EB-HIIE (50.0% before, 66.5% after). We consider this improvement to be clinically relevant, as each 5% increase in TIR is known to have clinical benefits in adults with T1D, as reported in the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) consensus. ¹⁷ In terms of glucose control, EB-HIIE demonstrated superiority over CONT in TIR improvement, with an increase of 16.5% for EB-HIIE versus 10.4% for CONT. In addition, EB-HIIE exhibited a reduction in mean 24-hour interstitial glucose from 169.8 to 153.8 mg/dl. This decrease, along with the improvement in TIR, is noteworthy, considering that the 2022 consensus sets a target mean glucose concentration of 154 mg/dl as an acceptable clinical goal for T1D patients. ¹⁷ These findings of the superiority of EB-HIIE compared with CONT in 24-hour metrics contrast with those in previous studies, where improvements were observed only in mean glucose values but not in TIR, ²⁶ or no significant differences were found at all. ²⁹ However, the differences in the methodology of each study make comparisons challenging, especially due to variations in HIIE modalities, ²² or other aspects of the study design, such as the recruitment of sedentary subjects,^26,29 or not using the same time of day for exercising for all patients. ²⁹

The main findings of this study indicate that as an exercise modality, HIIE, specifically EB-HIIE used in this study, is more effective in improving glycemic control and safer compared with CONT in people with T1D. The use of EB-HIIE is highlighted for its advantages, being easily reproducible in any environment and cost-effective, as the elastic bands used are inexpensive and user-friendly exercise tools. The clinical implications of these findings suggest that EB-HIIE, due to its affordability and accessibility, may be a practical and feasible option for implementation in tele-rehabilitation programs. The study emphasizes the importance of further research to explore the application of EB-HIIE in therapeutic programs, assess its long-term glycemic effects, and compare its adherence with other HIIE modalities or types of exercise, including CONT or resistance exercise.

The main strengths of our study were the sample size and the design. First, the sample size of 39 adult men with T1D is a significant improvement compared with previous studies in this field. Previous studies that have quantified acute glucose changes with exercise in T1D, ¹² or the specific review of HIIE in subjects with T1D, ²² included only 6 to 17 subjects. In addition, the use of a crossover design allowed each subject to serve as their own control in different experimental situations, contributing to more robust paired samples analysis and reducing uncertainty related to interpatient variability. The incorporation of a pilot test for each person to tailor the GXT to individual physical conditions is another strength. The employment of the YSI for blood glucose measurement, considered the gold standard, and the measurement of blood lactate concentrations to assess workout intensity during exercise further enhance the study’s reliability.

However, the study has limitations. First, only 1 repetition of each exercise was performed for each patient, limiting the ability to estimate within-exercise variability. Second, the study population comprised exclusively exercise-active men with T1D using multiple daily injections, restricting the generalizability of the results to other populations, such as women, more sedentary people, or those with diabetes using pump therapy. To address these limitations, future studies should aim for a broader and more representative sample of people with T1D.

In conclusion, this study shows that EB-HIIE is a safe and effective training method for male adults with T1D, exhibiting positive effects on glucose homeostasis. Specifically, EB-HIIE improves TIR in the 24 hours after its completion, while maintaining euglycemia during and immediately after the exercise session. In contrast, CONT, although also showing improvements in TIR 24 hours after exercise, is associated with more pronounced immediate glucose-lowering effects coupled with a longer time in level 1 hypoglycemia, necessitating increased surveillance during implementation. Furthermore, EB-HIIE demonstrates a higher improvement in TIR compared with CONT. In summary, the results indicate that EB-HIIE is a more convenient and affordable exercise modality for people with T1D, offering advantages over the traditional CONT, especially when aiming to incorporate exercise into daily life conditions.