Toward Individualized High-Intensity Interval Training in Type 1 Diabetes: A Framework for Safe Implementation ^†

Abstract

High-intensity interval training (HIIT) is presented as a safe, effective, and time-efficient strategy for individuals with type 1 diabetes, offering benefits for glycemic control, cardiovascular function, and physical fitness, with a lower risk of hypoglycemia than other exercise modalities. However, substantial variability exists among protocols, and there is no consensus on optimal dosage. This study reviewed 18 investigations to identify key parameters for safe and effective implementation. Results emphasize the importance of individualized programming, adherence to protocols, frequent glucose monitoring, and professional supervision. A preliminary framework is proposed to guide personalized HIIT programs for people with type 1 diabetes.

1. Introduction

Exercise is one of the key components in the management of type 1 diabetes mellitus (T1DM). In particular, HIIT is a potentially useful exercise modality for individuals with T1DM, as it improves cardiorespiratory fitness, insulin sensitivity, and glycemic control, including 24-h mean glucose levels [1,2,3]. It is commonly associated with a lower risk of acute hypoglycemia during exercise compared with moderate-intensity continuous exercise [4]; however, it may induce transient hyperglycemia or reduce post-exercise hypoglycemia awareness in some patients. HIIT is considered a safe, effective, and time-efficient training option, and it may even facilitate home-based exercise with virtual monitoring, helping to overcome common barriers such as lack of time and fear of hypoglycemia [5].

The safe and effective implementation of HIIT in individuals with T1DM requires strict protocol planning, as multiple complex factors interacting with individual physiology must be taken into account, along with additional variables such as exercise timing, types and doses of insulin used, and glycemic monitoring systems, among other factors [6,7].

The existing literature recognizes HIIT as a key component in the management of diabetes; however, there is currently no clear consensus regarding the optimal type, structure, or dosage of HIIT for individuals with type 1 diabetes mellitus (T1DM). The wide heterogeneity of protocols reported in the literature—in terms of intensity, interval duration, number of repetitions, recovery periods, and weekly frequency—limits comparisons across studies and hinders the development of standardized recommendations. This narrative review aims to analyze and synthesize the available scientific evidence on the effects of HIIT in individuals with type 1 diabetes, identifying key parameters for its safe and effective programming.

2. Materials and Methods

This review aimed to synthesize and critically examine the current literature on the impact of HIIT on individuals with type 1 diabetes, identifying key variables for its safe and individualized application.

The literature search was conducted using the MEDLINE and PubMed Central (PMC) databases, without date restrictions, to identify clinical studies, controlled trials, and reviews evaluating the effects of HIIT in individuals with type 1 diabetes mellitus (T1DM). Combinations of keywords were applied, including high-intensity interval training, HIIT, type 1 diabetes, and cycle ergometer, using Boolean operators such as AND and OR.

All records retrieved from MEDLINE and PMC were imported into a reference management software (Zotero 7.0.32, Corporation for Digital Scholarship, Vienna, VA, USA). Duplicate records were automatically identified and subsequently verified and merged manually prior to screening. Study selection was performed manually through the review of titles, abstracts, and full texts according to predefined inclusion and exclusion criteria. Priority was given to human studies involving participants with a confirmed diagnosis of type 1 diabetes, in which HIIT interventions were applied—preferably using a cycle ergometer—and that reported outcomes related to metabolic, physiological, or neurological variables. Studies involving populations with type 2 diabetes or prediabetes were excluded, as were interventions combining pharmacological or nutritional modifications, studies focused exclusively on diabetes-related technologies such as continuous glucose monitoring or insulin pumps, case studies, and articles without full-text availability.

During the analysis of the selected studies, common and relevant variables for HIIT programming in this population were identified and organized into four main categories: population characteristics, HIIT protocol parameters, reported outcomes and effects, and control and safety criteria (

Table 1. Overview of key variables used to analyze HIIT protocols in individuals with type 1 diabetes.

Population Characteristics	HIIT Protocol Parameters	Reported Outcomes and Effects	Control and Safety Criteria
Number of Participants. Age, Sex. Duration of type 1 diabetes diagnosis. Type of insulin therapy (multiple daily injections vs. insulin pump).	Warm-up: intensity and duration. Protocol format: interval and recovery duration. Exercise intensity (expressed as %VO2max, %HRmax, or relative workload). Weekly frequency and total duration of the intervention.	Physical fitness responses (VO2max, aerobic power, strength). Metabolic, physiological and neurological effects. Adverse events. Other reported outcomes.	Pre- and post-exercise blood glucose monitoring. Hypoglycemia prevention strategies. Training supervision.

).

3. Results

The literature search identified a total of 353 records, including 287 records from PubMed Central (PMC) and 66 records from MEDLINE. After removal of 32 duplicate records using reference management software, 321 records remained for screening. Following title and abstract screening based on the predefined inclusion and exclusion criteria, 303 records were excluded. Full-text assessment of the remaining articles resulted in the inclusion of 18 studies in the final analysis. The study selection process is summarized in the PRISMA flow diagram (Figure 1).

An exhaustive evaluation of the 18 selected studies revealed substantial heterogeneity in participant characteristics. Overall, the included studies analyzed 238 individuals with type 1 diabetes (126 males and 112 females), with ages ranging from 10 to 57 years. For the purposes of this review, in studies evaluating multiple exercise modalities, only participants allocated to HIIT interventions were considered for analysis, and only the effects attributable to this exercise modality were extracted; participants assigned to control conditions or alternative exercise protocols were not included in the aggregated analysis. A summary of the protocols is presented in

Table 2. Main characteristics of the HIIT protocols applied in individuals with type 1 diabetes across the analyzed studies.

Study	Total Time (Weeks)	Weekly Freq.	Session Time (min)	Warm-Up (min/Intensity)	HIIT Bouts	Effort Phase (sec/Intensity)	Recovery Phase (sec/Intensity)	Cooldown (min)
Harmer et al. (2008) [8]	7	3	25–35	NR	4-6-8-10	30 All-out	240/ passive	NR
Tonoli et al. (2015) [9]	Acute (1 day)	1	22	2/100 W	10	60–90% VO2max	60/50 W	NR
Rooijackers et al. (2017) [10]	Acute (1 day)	1	15	4/50 W	3	30-All-out (Borg > 15)	240/50 W	NR
Wiegers et al. (2017) [11]	Acute (1 day)	1	15	4/50 W	3	30 All-out	240/50 W	NR
Farinha et al. (2017) [12]	10	3	25	3/50 W	10	60–90% HRmax	60/50 W	2/ 50 W
Riddell et al. (2019) [5]	Acute (4 sessions)	1	25	0.5/50% VO2peak	10	100–130% PPO	30/50% VO2peak	0.5
Scott et al. (2019) [1]	6	3	15–23	3/low-intensity	6-8-10 *	60–100% VO2peak	60/50 W	NR
Boff et al. (2019) [13]	8	3	40	5/60% HRmax	6	60–85% HRmax	240/50% HRmax	5
Farinha et al. (2019) [14]	10	3	25	3/50 W	10	60–≥90% HRmax	60/50 W	NR
Scott et al. (2019) [15]	12	3	12–20	2/Jogging	6-8-10 *	60–85% HRmax	60/ Passive	NR
Lee et al. (2020) [16]	12	3	33	5/60% HRpeak	4	240–85–95% HRpeak	180/50–70% HRpeak	3
Lee et al. (2020) [7]	Acute (4 sessions)	Variable	33	5/60% HRpeak	4	240– High intensity	180	3
Minnebeck et al. (2020) [17]	4	2	16–20	5/20 W	4 to 6	60-All-out (≥95% HRmax)	60/ Passive	3
Alarcón-Gómez et al. (2021) [2]	6	3	28–40	5/50 W	12-16-20 *	30–85% PPO	60/ 40% PPO	5/ 50 W
Alarcón-Gómez et al. (2021) [18]	6	3	28–40	5/50 W	12-16-20 *	30–85% PPO	60/ 40% PPO	5/ 50 W
Mascarenhas et al. (2022) [19]	Acute (1 day)	1	30	NR	5	10– Max intensity	300/60% VO2max	NR
Farrell et al. (2024) [20]	4	3	20	5/gentle cycling	4	30–≥90% HRpeak	120/ Active	5
Scoubeau et al. (2024) [21]	12	3	35	3	8	120–90% HRmax	120/ Active (VT1)	NR

NR: not reported, min: minutes, sec: seconds, VO₂max and VO₂peak: maximal oxygen uptake, W: watts, HRmax: maximal heart rate, HRpeak: peak heart rate, PPO: peak power output, VT1: first ventilatory threshold. (*) Progressive volume protocols where the number of bouts increased throughout the weeks.

.

Despite their different objectives, acute (single-session) and longitudinal HIIT protocols in individuals with type 1 diabetes share several key characteristics. Both typically require the attainment of very high exercise intensities (≥85–90% of HRmax or HRpeak, or VO₂max), with some protocols incorporating maximal or supramaximal efforts (e.g., 100–130% of peak power output). Most interventions include a standardized warm-up phase (approximately 2–5 min) performed at low-to-moderate intensity. Exercise intensity is prescribed using comparable physiological metrics, such as heart rate, VO₂max, or peak power output, reflecting a consistent methodological framework across study designs.

Regarding session structure, effort phases are generally short (approximately 30–60 s), and the average duration of a HIIT session is around 25 min, with reported protocols ranging from 12 to 40 min. Across studies, HIIT programs are primarily designed with participant safety as a central consideration, incorporating appropriate work-to-rest ratios and appropriate supervision to minimize adverse events, including severe hypoglycemia.

Table 3. Summary of the main metabolic, physiological, and neurological effects of HIIT in individuals with type 1 diabetes.

Study	n	Age (Years)	Outcomes			Other Effects
			Physical Fitness	Metabolics	Physiological/Neurological
Harmer et al. (2008) [8]	8	21–29	NR	↑ Exercise-induced plasma glucose (acute exercise) ↔ HbA1c ↔ Resting plasma glucose	↓ Exercise-induced hyperglycemia (post-exercise), ↓ Exercise-induced lactate accumulation (post-exercise)	↓ VCO2peak ↓ VEpeak (post-training), ↔ Acid–base status within physiological range No hypoglycemia reported during exercise
Tonoli et al. (2015) [9]	10	18–44	NR	↓ post-blood glucose levels	↑ BDNF ↑ IGF-1 postexercise; improves executive function	Serum-free insulin, and blood glucose serum was correlated with serum BDNF and negatively correlated to serum IGF-1
Rooijackers et al. (2017) [10]	30 (10 NAH, 10 IAH, 10 healthy)	19–37	NR	↑ blood glucose (normal range), ↑ Blood Lactate ↓ pH	↓ Awareness of hypoglycemia ↑ adrenaline ↑ noradrenaline ↑ GH ↑ Cortisol (remains ↑ IAH)	HIIT mitigates cognitive impairment during hypoglycemia in NAH
Wiegers et al. (2017) [11]	18 (6 IAH, 6 NAH, 6 healthy)	19–30	NR	↑ Brain Lactate, most pronounced increase in IAH, ↓↓ Brain lactate in hypoglycemic phase (IAH); identical lactate decay rate (all groups)	↓ Adrenaline response to hypoglycemia (IAH < NAH/Healthy)	↑ cerebral lactate transport/oxidation in IAH.
Farinha et al. (2017) [12]	9	18–40	↑ Strength and cardiopulmonary fitness, ↑ VO2peak, ↑ FFM	↓ HbA1c and ↓ Fasting glucose. ↑ Antioxidants (TAC, CAT, SOD) ↓ sRAGE (↓ inflammation ↓ vascular damage)	↑ iHSP70 protein (cell protection biomarker)	ST+HIIT ↓ insulin dose ↑ adherence to exercise
Riddell et al. (2019) [5]	16	24–44	NR	↑ Plasma glucose (post-exercise hyperglycemia), ↑ Lactate, ↓ Ketone bodies, ↓ Free fatty acids, ↑ Serum insulin	↑ HR (≈HRmax) ↑ Oxygen uptake (≈VO2peak). ↑ Catecholamines, ↑ GH, ↑ RPE (near-maximal)	Similar inter-subject glycemic, hormonal, and lactate responses No hypoglycemia risk (fasted HIIT)
Scott et al. (2019) [1]	7	26–32	↑ VO2peak, ↑ Wmax	↔ glucose during HIIT (fed state); ↑ Nocturnal hyperglycemia.	↓ aPWV; ↔ CHO ↔ TG	HIIT prevents acute intra-session glucose drops
Boff et al. (2019) [13]	9	19–33	↑ VO2peak	↔ Glycemic control (HbA1c/Glucose)	↑ FMD, superior to MCT	Improvement in FMD and VO2peak are positively correlated
Farinha et al. (2019) [14]	9	18–40	NR	↓ blood glucose	↓ HbA1c ↓ RoCE	-
Scott et al. (2019) [15]	11	27–33	↑ VO2peak, ↑ Perceived physical fitness	↔ Glycemic stability (during and 1 h post-exercise), ↔ BMI	No Severe hypoglycemia ↔ Nocturnal glucose levels	↓ Short-acting insulin dose ↑ Adherence and compliance ↓ Perceived barriers (time constraints, fear of hypoglycemia)
Lee et al. (2020) [16]	24	30–50	↑ Maximal treadmill exercise test time, ↑ Leg strength	↓ Mean glucose, ↔ Body weight, ↔ Lipid profile	↓ HbA1c, ↑ Leptin	↔ Total daily insulin dose, ↔ Adiponectin, ↔ Hypoglycemia incidence, ↔ Glycemic safety (24 h) (with appropriate self-management)
Lee et al. (2020) [7]	12	44 ± 10	↑ Strength (Leg) ↑ VO2peak and time on Bruce protocol	↓ Mean 24h glucose, ↑ total body lean mass	↓ HbA1c in subgroup with ≥50% adherence ↑ Leptin	↑ Adherence and compliance severe hypoglycemic (1 event) Targeted at a high-risk population
Minnebeck et al. (2020) [17]	22 (11 Overweight, 11 Normal Weight)	26–57	↑ VO2peak ↑ maximal exercise capacity ↑ physical fitness	↓ Daily basal and bolus insulin requirements post-HIIT	HbA1c: downward trend in overweight subgroup; ↓ LDL ↓ UA	↑ HRQoL (physical role limitations)
Alarcón-Gómez et al. (2021) [2]	11	33–43	↑ VO2max; ↑ FFM, 3.4%), ↓ FM (6.4%).	↓ Fasting Glucose (7.8%). Low rate of mild hypoglycemia (1.5%).	Improves HRV ↑ HRQoL (physical and social domains) and sleep quality	No severe hypoglycemia; HIIT is safe and efficient for ↓ CV risk. ↑Adherence
Alarcón-Gómez et al. (2021) [18]	11	33–43	NR	Low incidence of mild hypoglycemia No severe hyperglycemias. No increase in nocturnal hyperglycemia.	↑ HRQoL (physical and social domains), ↑ SQ	↑ exercise enjoyment, ↑ motivation (intrinsic and identified regulation) HIIT is safe for improving well-being
Mascarenhas et al. (2022) [19]	31	10–15	NR	↓ RoCE (Lower reduction in blood glucose during exercise). Better recovery (RoCR) in the 30 min post-exercise	Reduces the risk of hypoglycemia	NR
Farrell et al. (2024) [20]	18	20–54	NR	↓ hypoglycemia ↑ plasma glucagon during hypoglycemia	improvements CCR ↑ response to hypoglycemia Improved glucagon and norepinephrine response	Maintained hypoglycemia symptom awareness (total/autonomic)/HIIT is safe in IAH
Scoubeau et al. (2024) [21]	10	30–56	↑ VO2peak, ↑ VT1, ↑ Maximal O2pulse	↔ HbA1c ↔ Lipid profile	NR	NR

NR: Not reported, ↑: increase, ↓: decrease, ↔: no significant change, ≈: approximately equal. HbA1c: Glycated hemoglobin, RoCE: rate of change in glycemia during exercise, RoCR: rate of change in the initial recovery period, BDNF: brain-derived neurotrophic factor, IGF-1: insulin-like growth factor 1, FMD: endothelial function, GH: growth hormone, CCR: counterregulatory responses, TG: triglycerides, FFM: fat free mass, FM: fat mass, LDL: low-density lipoprotein, CHO: cholesterol, UA: uric acid, iHSP70: intracellular Heat Shock Protein 70, TAC: total antioxidant capacity, CAT: catalase, SOD: superoxide dismutase, sRAGE: soluble receptors for advanced glycation end products, PH: hydrogen potential, VCO₂peak: peak carbon dioxide production, VEpeak: peak ventilation, Wmax: maximal aerobic power output, VO₂max and VO₂peak: maximal oxygen uptake, W: watts, HRmax: maximal heart rate, VT1: first ventilatory threshold, RPE: Rating of perceived exertion, ST: strength training, aPWV: aortic pulse wave velocity, NAH: normal awareness of hypoglycemia, IAH: impaired awareness of hypoglycemia, HRQoL: health-related quality of life, SQ: sleep quality.

and Figure 2 below provide an overview of the main metabolic, physiological, neurological, and other effects reported after HIIT in people with type 1 diabetes in the included studies.

4. Discussion

The effects observed following the application of HIIT protocols are heterogeneous; however, they can be systematically examined across metabolic, physiological, neurological, and physical fitness domains. Regarding glycemic responses, HIIT has been consistently associated with a smaller decline in blood glucose levels during exercise compared with continuous moderate-intensity exercise. In the analyzed studies, the most commonly used HIIT protocols consist of short intervals of 30–60 s performed at high intensities (≥85–90% HRmax). Under this scheme, no significant hypoglycemic or hyperglycemic events are reported during the session, and greater glycemic stability is described. Moreover, the findings consistently indicate that this exercise modality substantially reduces the risk of exercise-induced hypoglycemia, both during and after physical activity, thereby improving treatment adherence and enhancing its therapeutic feasibility [5,9,10,11,15]. These effects are explained, at least in part, by an increased release of counter-regulatory hormones, particularly catecholamines, which stimulate hepatic glucose production [2].

A recent meta-analysis demonstrated that HIIT interventions are associated with significant improvements in 24-h mean glucose levels. Regarding HbA1c, the available evidence remains heterogeneous; while some studies report no statistically significant changes at the group level, others describe clinically meaningful reductions, with decreases of up to 0.64% observed in participants with high adherence (>50% of prescribed HIIT sessions). Notably, none of the analyzed studies reported increases in HbA1c; on the contrary, most demonstrated reductions, and only two found no significant changes. Fasting plasma glucose has also been shown to exhibit modest reductions following structured HIIT programs, and selected protocols—particularly those incorporating resistance training—have reported decreases in total daily insulin requirements [17]. Beyond glycemic outcomes, HIIT elicits meaningful cardiometabolic adaptations in individuals with type 1 diabetes, including improvements in cardiorespiratory fitness, vascular function, body composition (especially in overweight individuals), lipid-related risk factors, and heart rate variability, collectively contributing to a more favorable metabolic and cardiovascular risk profile [21].

Regarding the programming and the safe and effective implementation of HIIT in individuals with T1DM, an interdisciplinary approach is required to appropriately manage insulin dosing, nutritional strategies, and training loads, together with frequent glucose monitoring—preferably using continuous glucose monitoring systems, with capillary confirmation when necessary [1]. Furthermore, it is essential to consider current consensus recommendations, which advise a pre-exercise blood glucose range of 126–180 mg/dL (7.0–10.0 mmol/L) or 90–270 mg/dL (5.0–15.0 mmol/L) in the pediatric population. Exercise should also be postponed in the presence of blood glucose ≥ 270 mg/dL accompanied by elevated ketones (≥1.5 mmol/L in blood or 4.0 mmol/L in urine), a severe hypoglycemic episode within the previous 24 h, or acute infection or injury, given the associated risk of metabolic instability [22,23,24,25]. Given the lower treatment adherence and the higher prevalence of physical inactivity observed in this population, the available evidence supports the use of pragmatic HIIT designs characterized by shorter exercise sessions, reduced total training volume, and submaximal rather than maximal intensities, in order to enhance feasibility and long-term adherence without compromising safety. This perspective is consistent with contemporary frameworks emphasizing the need to optimize HIIT prescriptions according to training status and population-specific characteristics, particularly in health-oriented contexts where sustainability and individual responsiveness are prioritized over maximal performance outcomes [26]. Furthermore, the marked heterogeneity in HIIT protocols and inter-individual responses underscores the need for individualized programming based on patient-specific characteristics, including age, sex, fitness level, duration of diabetes, and hypoglycemia awareness [5].

5. Conclusions

Overall, the evidence synthesized in this review indicates that high-intensity interval training represents a safe, time-efficient, and feasible exercise strategy for individuals with type 1 diabetes. Across acute and longitudinal interventions, HIIT is consistently associated with a more stable glycemic response during and after exercise, without increasing the incidence of severe hypo- or hyperglycemic events, supporting its clinical relevance in diabetes management.

Beyond metabolic control, HIIT elicits meaningful adaptations at the physiological and neuroregulatory levels, including improvements in physical fitness, cardiovascular function, and autonomic balance. These adaptations are accompanied by favorable effects on perceived well-being, enjoyment, and motivation toward physical activity, which collectively contribute to greater exercise adherence and improved quality of life. Taken together, these findings support HIIT as a comprehensive non-pharmacological intervention capable of addressing both the physiological and behavioral barriers to physical activity commonly observed in this population.