Edited by: Kristian Karstoft, Rigshospitalet, Denmark
Reviewed by: Andreas Buch Møller, Aarhus University Hospital, Denmark; George L. King, Joslin Diabetes Center and Harvard Medical School, United States
*Correspondence: Jason P. Pitt,
This article was submitted to Clinical 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.
International charities and health care organizations advocate regular physical activity for health benefit in people with type 1 diabetes. Clinical expert and international diabetes organizations’ position statements support the management of good glycemia during acute physical exercise by adjusting exogenous insulin and/or carbohydrate intake. Yet research has detailed, and patients frequently report, variable blood glucose responses following both the same physical exercise session and insulin to carbohydrate alteration. One important source of this variability is insulin delivery to the circulation. With modern insulin analogs, it is important to understand how different insulins, their delivery methods, and inherent physiological factors, influence the reproducibility of insulin absorption from the injection site into circulation. Furthermore, contrary to the adaptive pancreatic response to exercise in the person without diabetes, the physiological and metabolic shifts with exercise may increase circulating insulin concentrations that may contribute to exercise-related hyperinsulinemia and consequent hypoglycemia. Thus, a furthered understanding of factors underpinning insulin delivery may offer more confidence for healthcare professionals and patients when looking to improve management of glycemia around exercise.
People with type 1 diabetes (T1D) are required to administer exogenous insulin
The pathway of subcutaneously administered exogenous insulin. Insulin is injected/released from formulation in the insulin pen/pump cartridge into the subcutaneous tissue. The insulin oligomers disassociate into monomer units before translocating across the capillary endothelium into blood circulation. Insulin circulates before binding to an insulin receptor to facilitate glucose uptake into the cell (e.g. into the myocyte). Factors at rest, acute exercise, and chronic exercise which affect each stage are listed along the row beside each illustrated stage of the pathway. Insulin diffusion in the subcutaneous layer is adapted with permission from digital images of insulin depot formation 15 to 30 s after bolus injection into porcine subcutaneous tissue; authored by Jockel et al. (
It is of clinical concern when the
This review aims to (i) inform the reader of relevant factors that influence insulin absorption at rest and (ii) critique the research investigating exercise effects on insulin absorption, providing an evidence-based explanation of the underlying mechanisms, where possible.
At concentrations necessary for subcutaneous injection, human insulin self-associates into hexamer crystals which are too large to pass through capillary fenestrae and reach circulation. The significant time required to allow the hexamers in the subcutis to dissociate to monomer units, that are small enough to diffuse into the capillary lumen, results in serum insulin concentrations out of sync with post-absorption blood glucose concentrations. The development of genetic engineering in recombinant DNA technology enabled modifications of insulin’s molecular structure that influence its pharmacokinetic properties. Initial modifications focused on inverting (
Synopsis of the pharmacokinetic properties of exogenous insulin.
Insulin action type | Name | Manufacturer | Chemical modifications | Mechanism | Pharmacokinetic profile | References | ||
---|---|---|---|---|---|---|---|---|
ULTRA-RAPID | Fiasp (Faster-acting insulin aspart) | NovoNordisk | Niacinamide and L-arginine added to solution (insulin structure is that of insulin aspart) | Niacinamide excipient destabilizes hexamer in subcutis and may mediate local vasodilation | 3–5 h | 3–5 min | 45–60 min | Hövelmann et al. ( |
Lyumjev (Ultra-rapid lispro) | Eli Lilly and Company | Treprostinil and citrate added to solution (insulin structure is that of lispro) | Citrate increases local vascular permeability and treprostinil increases local vasodilation | 5 h | 2 min | 45–60 min | Leohr et al. ( |
|
RAPID | Humalog (Lispro) | Eli Lilly and Company | Inverted |
Distortion at the dimer interface destabilizes hexamer | 3–5 h | 5–20 min | 45–60 min | Howey et al. ( |
Apidra (Glulisine) | Sanofi | Lower isoelectric point improves solubility at physiological pH | 3–5 h | 10 min | 45–60 min | Danne et al. ( |
||
Novorapid (Aspart) | NovoNordisk | Removing |
3–5 h | 10 min | 45–60 min | Plank et al. ( |
||
SHORT | Actrapid | NovoNordisk | Regular human insulin | Hexamer formation in storage delays appearance in circulation | 8 h | 30 min | 1–2.5 h | Mortensen et al. ( |
INTERMEDIATE | Novo NPH | NovoNordisk | Protamine added to insulin solution | Formation of crystals in solution | 10–14 h | 1.5 h | 4 h | Lepore et al. ( |
LONG | Levemir (Detemir) | NovoNordisk | C14 fatty acid is bound to |
Human serum albumin binding and dodecamer formation | 20–24 h | 2.5 h | None | Porcellati et al. ( |
Lantus (Glargine) | Sanofi | Isoelectric point ~7 leads to precipitation in subcutis | 20–24 h | 1.5 h | None | Lepore et al. ( |
||
ULTRA-LONG | Degludec (Tresiba) | NovoNordisk | C16 fatty diacid γ- |
Formation of multi-hexamer units | 24–42 h | 30–90 min | None | Haahr & Heise ( |
Glargine U300 (Toujeo) | Sanofi | Larger precipitate than U100 glargine delays dissolution | >30 h | 30–90 min | None | Becker et al. ( |
Arg, Arginine; Asp, aspartic acid; Glu, glutamic acid; Gly, glycine; Lys, lysine; NPH, neutral protamine Hagedorn; Pro, proline; Thr, threonine.
The subcutaneous tissue consists primarily of adipocytes and an extracellular matrix made up of connective tissue and interstitial fluid, which present barriers of different resistances to insulin in its pathway to the vascular system (
Several studies demonstrate the importance of subcutaneous injection in tempering, and reducing the variation of, the pharmacokinetic/pharmacodynamic profiles of injected insulin compared to intramuscular injection at rest (
Combined with data collected on the thickness of the subcutaneous tissue in patients (
The effects of intramuscular insulin are more rapid and variable than subcutaneous insulin injections at rest and during exercise.
An inverse relationship exists between subcutaneous thickness and the rate of insulin absorption. In healthy participants, weak-moderate negative correlations exist between subcutaneous fat layer thickness and serum insulin appearance rate (
Greater subcutaneous adipose tissue layer thickness is associated with a tempered absorption profile of injected insulin.
The effect of temperature on exogenous insulin absorption has been investigated in those with and without T1D. One study investigated insulin absorption in individuals with T1D who injected insulin actrapid 60 min before two 25-min bouts of sitting in a sauna at 85°C (
Increased ambient temperature or local warming of the injection site accelerates insulin absorption.
Muscular exercise induces rapid changes in the physiological systems of the person with type 1 diabetes to supply working muscles with oxygen and nutrients. The exercise pressor reflex (i.e. a peripheral neural reflex arising from skeletal muscle contraction) prompts cardiovascular changes, namely: an increase in cardiac output, blood pressure, and a shunting of blood away from the viscera towards the working muscles, aided by increased concentrations of adrenaline, noradrenaline, and cortisol (
For the individual with T1D, the physiological changes induced by exercise pose a problem for maintaining glucose control. The synergistic effect of relative hyperinsulinemia (from the previous exogenous injection) and exercise-induced insulin-independent pathways cause the uptake of glucose into myocytes to exceed hepatic glucose release and a decline in blood glucose during continuous steady-state exercise (
Randomized controlled trials investigating the effect of exercise compared to rest on insulin absorption in people with type 1 diabetes or healthy individuals.
Authors and date (arrow indicating exercise-induced change in insulin absorption) | Investigated insulin (units injected) | Site of injection | Insulin absorption measurement | Exercise methodology | Insulin absorption outcome |
---|---|---|---|---|---|
Ferrannini et al. ( |
Actrapid (8 U) | Thigh and abdomen | 125I-labeled actrapid (radioactivity count) | Healthy participants (n = 8; undefined M/F) performed 20 min of moderate-intensity continuous exercise (ending in 170 bpm HR) on cycle ergometer | Increased RIA during exercise in leg injection (exercise 1.12 ± 0.12 vs Rest 0.68 ± 0.15%.min−1; p < 0.05). |
Kemmer et al. ( |
Actrapid (20 U) | Leg and arm (undefined) | 125I-labeled actrapid (radioactivity count) | Participants with T1D (n=9; M 8/F 1) performed 10 min bouts separated by 5 min rest, for 30 min total exercising, continuous low-to-moderate intensity exercise (125 ± 5 bpm) on cycle ergometer | Increased RIA after exercise in leg injection compared to same time period at rest (undefined, statistically significant); however, no change during exercise. No change in RIA at any timepoint in arm injection compared to rest |
Kolendorf et al. (1979) ( |
Actrapid (8 U) | Thigh | 131I-labeled actrapid insulin (radioactivity count) | Participants with T1D (n = 5; undefined M/F) performed four 10-min periods, with 400-sec intervals, of moderate-intensity continuous exercise (120 ± 10 bpm) on cycle ergometer | Increased RIA during exercise compared to rest (Exercise 0.71 ± 0.18 vs Rest 0.41 ± 0.15%.min−1; |
McAuley et al. ( |
Aspart (pump) (TDD 0.55 ± 0.10 U.kg−1.day−1) | Abdomen | Venous blood sampling (radioimmunoassay) | Participants with type 1 diabetes (n = 14; M 7/F 7) performed 30 min, including a 5 min warm up, of moderate-intensity continuous exercise (65–70% age-predicted maximal heart rate on a cycle ergometer) | Significant increase of mean free insulin concentration during exercise by 6 ± 2 pmol.L−1 compared to rest ( |
Ronnemaa & Koivisto ( |
Actrapid (5 ± 1 U) | Thigh | Venous blood sampling (radioimmunoassay) | Participants with type 1 diabetes (C-peptide negative) (n = 9; M 9/F 0) performed three 15-min periods, with 5-min rest intervals, of moderate-intensity continuous exercise (3-min warm-up, then 12-min at 60% VO2max) on cycle ergometer, in either cold (10°C) or warm (30°C) ambient temperatures | Significant difference in plasma free insulin (average difference over whole exercise bout) between exercise and rest in 10°C, 2.7 mU.L−1 ( |
Thow et al. ( |
NPH (0.25 U.kg−1) | Thigh | Venous blood sampling (radioimmunoassay) | Healthy participants (n=7; M 7/F 0) performed 60 min low-to-moderate-intensity continuous exercise (5 km.h−1 at 5° gradient) on treadmill | Increased serum insulin concentration from pre-exercise rest to average peak in exercise (13.7 ± 1.2 vs 27.3 ± 3.2 mU.L−1; NSR) |
Susstrunk et al. ( |
Actrapid (0.12 U.kg−1) | Abdomen or Thigh | Venous blood sampling (radioimmunoassay) | Healthy volunteers (n = 4; undefined M/F) performed three 15-min bouts exercise, separated by 5-min rest periods, of continuous exercise at low-to-moderate-intensity (50% maximum power capacity) on a cycle ergometer | Rate of insulin absorption was higher upon injecting into the abdomen (0.039 U.min−1) than into the thigh (0.027 U.min−1; |
Peter et al. ( |
Glargine (27.2 ± 9.1 U) | Thigh | 125I-labeled Glargine | Participants with type 1 diabetes (n = 13; M 12/F 1) performed 30 min of moderate-intensity continuous exercise (65% VO2max) on cycle ergometer | No significant change in RIA between exercise and rest trial days (NDR; |
Turner et al. ( |
Glargine (27.5 ± 3.1 U) | NDR | Venous blood samples (immunometric assay) | Participants with type 1 diabetes (n = 8; M 7/F 1) performed either control (rest), 1, 2, or 3 sets of moderate to high intensity (67 ± 3% 1RM) resistance exercise | No significant change in plasma insulin concentrations between or within trials (during exercise = NDR, post exercise |
Turner et al. ( |
Glargine (27.5 ± 3.1 U) | NDR | Venous blood samples (immunometric assay) | Participants with type 1 diabetes (n = 8; M 7/F 1) performed either control (rest), 1, 2, or 3 sets of moderate-to-high intensity (60–70% 1RM) resistance exercise | No significant change in plasma insulin concentrations between any exercise trials and control, at any timepoints after exercise (during exercise = NDR) |
bpm, beats per minute; F, females; HR, heart rate; M, males; NDR, no data reporting; NPH, neutral protamine Hagedorn insulin; NSR, no statistical reporting; RIA, rate of insulin absorption; RM, repetition maximum; TDD, total daily dose; U, units (of insulin); VO2max, peak rate of oxygen uptake.
Increasing exercise workload to a high intensity, such as heavy weightlifting in resistance exercise, is associated with large increases in adrenaline and noradrenaline levels, in addition to high rates of H+ generation and efflux from muscle cells. In people with T1D, elevations in catecholamine concentrations stimulate hepatic glucose release to a degree which exceeds muscular glucose uptake, contributing to an increase in blood glucose that contrasts the decline typically observed at lower intensities. Turner and colleagues conducted two studies from which point-concentrations of plasma insulin can be compared between resistance exercise protocols and a rested control session (
Physical exercise increases the rate of insulin absorption in intermediate-, short-, and rapid-acting insulins but not in older long-acting insulins. There remains a dearth in the literature studying this effect in modern insulins and with exercise modalities other than sub-maximal endurance activities.
The protracted mechanism of action of intermediate- and long-acting insulins is primarily dependent on the slowed movement and delayed dissociation of insulin oligomers into monomeric form to cross the endothelial layer into systemic circulation. Exercise has limited influence on the rate of insulin dissociation, and consequent availability for absorption, as its initial location is confined to the subcutaneous interstitium (
The decision to inject into a specific injection site around exercise may be hampered by logistical reasons (e.g. a rugby player removing their pump prior to a match, or an endurance cyclist unable to inject into the thigh during a ride) and also by a lack of knowledge as to any potential effects that are consequent of choosing one location over another. Few studies have compared the use of different injection sites during exercise. One study demonstrated the rate of absorption increased when injecting 125I-labeled actrapid into the exercising limb (thigh) compared to a non-exercising limb (arm) in people with T1D performing bouts of moderate-intensity bicycle exercise (
The distribution of blood flow to the periphery for the purposes of thermoregulatory heat dissipation has been suggested as the underlying mechanism that explains the influence of temperature (
Diffusion of insulin into the circulation is dependent on the concentration gradient (i.e. a smaller concentration in the blood than the depot); hence, greater blood flow that transports insulin away from the vasculature, local to the depot, may indirectly promote the diffusion of insulin monomers away from the injection site by enabling a higher concentration gradient (
The exercise-induced increased rate of insulin absorption is likely due to a combination of factors relevant to the changes at injection site during exercise. The dissociation of insulin oligomers into biologically active units remains the initial rate-limiting step.
Insulin absorption rate into circulation is influenced by different factors both at rest and during exercise. Compared to the same individual at rest, the exercise-induced increased appearance of insulin in the blood leads to a greater reduction in blood glucose. This phenomenon is often over-looked by individuals performing spontaneous bouts of activity or planning insulin adjustments around structured exercise. There is some evidence to suggest that the choice of location and depth of injection causes additional variability to absorption rates, whereby injections that are deeper and local to the working muscles are susceptible to even higher rates of insulin absorption. Overall, the cause of the increase in absorption during exercise is likely due to a myriad of factors including capillary recruitment, massage-effect, blood flow, temperature, and flushing effect; however, further studies are required to clarify their relative importance. Furthermore, the studies that have investigated the effects of exercise on absorption are now dated, using insulin types that are becoming increasingly less common among the T1D population. Studies using modern ultra-rapid and ultra-long acting insulins are required to determine whether the exercise-induced increase in the rate of absorption is still applicable. Patients and healthcare providers should be aware that the insulin pharmacokinetics around exercise may differ to resting profiles, enabling proactive avoidance of low blood glucose concentrations.
JP—Main investigator. Led literature search, draft composure, and draft review. OM—Aided literature search, draft composure, and draft review. TH-J—Aided draft composure, provided expert knowledge, and draft review. BW—Aided literature search, draft composure, and draft review. RB—Aided literature search, draft composure, and draft review and provided expert knowledge. All authors contributed to the article and approved the submitted version.
TH-J is an employee of Novo Nordisk A/S.
The remaining 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.