Authors T1D subjects included Exercise description Exercise time (vs. meal & insulin) Glucose evolution during exercise Glucose evolution during recovery
Maximal graded exercise
Continuous moderate intensity exercise
[77] N=10; 10-19 yo Age: 15.7 ± 3 years HbA1c: < 10% With insulin pump Continuous moderate intensity exercise, 40 to 45 min at 60% VO2max on ergocycle 2 exercise conditions, in randomized order, staggered by 7days: with basal insulin rate reduced by 50% with the pump suspended basal. Back to normal basal rate 5min after exercise No physical activity in the previous 24h 2h after standardized breakfast and usual insulin bolus, not reduced (in order to mimic conditions of an unplaned exercise) Glucide intake (20g) before exercise and 15 min after exercise. Glycemia before exercise between 1 and 3 g/L, with no ketosis ↓ glycemia in average of 29% (2 hypoglycemia i.e. <70mg/dL) and 35.5% (2 hypoglycemia) for exercise with 50% basal rate, and exercise with suspended basal rate (no significant difference between both conditions) (and no significant difference in cortisol, GH, NE) Insulin concentration ↑ at the beginning of exercise (until 20min) then ↓ back to initial values at the end of exercise, and continues to ↓ until 45min recovery (no significant difference between both conditions) Subjects with hypoglycemia during exercise started it with <glycemia level and > insulin concentration  vs. other subjects Glycemia goes back to initial values after 45min recovery Data from continuous glucose sensor Every subject had 1 to 3 delayed hypoglycemia 1 after each exercise (after 2.5h to 12h), not always symptomatic. 9 and 6 delayed hypoglycemia after 50% basal rate and suspended basal rate respectively (no difference)
[114] N=50; 11-17yo Exercise: 4 sessions of 15 min exercise on treadmill at 140 bpm, with 5min recovery in between (140bpm ≈ 55%VO2max) Comparison between a sedentary and exercising day staggered by 1 to 4 weeks in randomized order (with meal and same insulin dose) Exercise at 4pm-5pm Same insulin treatment as usual Glycemia before exercise between 1 and 2g/L If glycemia< 0.60 g/L during exercise or during the night àglucide intake 22%  of patients had a hypoglycemic episode Averaged glycemia from 10pm to 6am < after the afternoon with exercise vs. sedentary Nocturnal hypoglycemia occurrence frequency > the next night after exercise vs. sedentary No nocturnal hypoglycemia the sedentary day when glycemia before bedtime >1.30g/L No significant difference related to counterregulatory hormones (GH, E, NE, Cortisol, Glucagon) during the night after exercise vs. sedentary
[115]
[75] N=49; 8-17 yo; with insulin pump 2 days of exercise including one with suspended insulin basal rate during exercise and the 45min recovery, staggered by 6 to 36days in randomized order (with the same meal) Glycemia before exercise between 1.2 and 2g/L Hypoglycemia (<0.7g/L) during exercise are less frequent when basal insulin rate is suspended (16% vs. 43%) ↓ glycemia during exercise < when basal rate suspended (-28% vs -41%) After 45min of recovery: Hyperglycemia (↑20% compared to value at the end of exercise, or > 2g/L) more frequent when insulin basal rate is suspended (27% vs. 4%) No abnormal ketonemia
[116] N=10 adolescent; 15.2 yo in average; 6.9% HbA1c in average   Exercise at around 3pm Basal insulin rate maintained during exercise Averaged glycemia before exercise: 1.62g/L Glycemia ↓ by 52% in average during exercise  
[76] N=16; 10-17yo; HbA1c<10%; average =7.5%; with insulin pump 3 conditions (randomized order): 2.5mg oral Terbutaline at bedtime; 20% basal insulin rate reduction ; no modification of usual treatment (=control) Staggered by at least 72h Standardized lunch and  normal insulin bolus Exercise at 4pm (glycemia before exercise between 1.2 and 2g/L) Suspended basal insulin rate during exercise then decreased by 50% during the next 45min after exercise Glucide snack if glycemia< 0.7g/L during the 5 min recovery periods and during the following recovery   Averaged glycemia at midnight: 1.88g/L after terbutaline; 1.72g/L after basal insulin reduction < vs. 1.27g/L control Terbutaline: No more nocturnal hypoglycemia but more frequent hyperglycemia >2.5g/L vs. control Basal insulin rate decrease : less glycemia< 0 8 and 0.7 g/L but more >2.5g/L vs. control
[117] N=9; 16 yo in average; average HbA1c 7.8% 1 exercise day (45min at 95% of second lactic threshold = 54.9 ± 2.8% VO2peak on ergocycle) and 1 sedentary day, staggered by 4 weeks, in a randomized order Exerciseat 4pm Morning: usual breakfast and usual rapid acting insulin injection, but no intermediate/long acting. From 11am: iv insulin infusion at an usual basal rate (=50% of total insulin dose /day) At noon: standardized meal + usual insulin bolus Comparable insulin concentration between both days Comparable previous 24h diet 48h before: no hypoglycemia; 24h before: no exercise ♀: in the middle of  menstrual  cycle follicular phase iv infused glucose in order to maintain euglycemia (during 18h): glucose infusion rate in order to maintain glycemia> during exercise Glucose infusion rate in order to maintain glycemia> at the beginning of recovery after exercise (90min recovery) and again 7-11h after exercise (=early or delayed hypoglycemia risk) No difference in counterregulatory hormones level between exercise and sedentary days except during early recovery: peaks E, NE, GH, Cortisol > (à compromised nocturnal hypoglycemia counterregulatory response after exercise)
High intensity intermittent exercise
[118] N=12; 12.2-15.8 yo; Tanner stages 2-4 Tanner; HbA1c 6.5-10.5% 10 bouts of 2min at 80% VO2peak with 1 minute recovery between each (on ergocycle) In the norning, after  light breakfast Glycemiamaintained  between 0.9 and 1.10 g/L during 90min before exercise by infusing insulin (clamp) then insulin rate infusion maintained at the same level during exercise. Glucose iv infusion in order to maintain euglycemia during exercise: average infusion rate = 1.5 mg/kg/min for 90 min before exercise, 1.7 mg/kg/min at the end of exercise, and 1.7 mg/kg/min after 30min recovery  
Moderate intensity continuous vs. High intensity intermittent exercise
[86] N=12; 14-19 yo, Tanner stages 4-5; (♀ and ♂) HbA1c 6.6-9.6% Actives 2 adolescents excluded from the analysis because of a nocturnal hypoglycemia the night before continuous exercise Maximal graded test on ergocycle 2h in average after  breakfast Insulin dose according to subject’s habits 3 exercises performed on 3 consecutive days, always in the same order: graded test, continuous, and intermittent ↓ glycemia (-1,5 mM in average) (final value = 11.2 mM in average) No hypoglycemic episode ↑↑E (*10) and NE (*8)  
Continuous 60min at 40% of VO2max on ergocyle ↓↓ glycemia (-4,1 mM in average) (final value = 11.2 mM in average) No hypoglycemic episode ↑E (*5) et NE (*3)  
Intermittent exercise on ergocycle: 5min warm-up, and 6 times 3min at 70% of VO2max interspersed with 1.5min active recovery (light resistance cycling), and 5 min resting time ↓ glycemia (-2.7 mM in average) (final value = 11.4 mM in average) No hypoglycemic episode ↑↑E (*8) and NE (*7) ↑ more important for cortisol vs. Other exercises  
Everyday life exercises
[119] n=30; 12-18 yo; HbA1c < 12% Parallel between accelerometer data and continuous intersticial glucose concentration measurement (subcutaneous sensor)   Hyperglycemia rebound after exercise induced glycemia decrease. Hypothesis: This rebound could be linked to : SNS effect during physical activity, a strategy before physical activity in order to prevent exercise induced hypoglycemia, or an excessive treatment for low glycemia.  
↓: Significant Decrease ; ↑: Significant Increase ; Δ: Significant Variation; ♀: Girl; ♂: Boy ; E: Epinephrine; NE: Norepinephrine GH: Growth Hormone ; VO2: Oxygen Comsumption; HbA1c: A1 Glycated Hemoglobin ; NS: Not Significant In blue: studies about the effect of insulin dose variation on exercise glycemia; In green: results about the effect of glycemia variation on performance; in orange: associated hormones response
Table 3: Evolution of blood and plasma glucose concentration in T1D according to different exercise type.