Low-dose glucagon has been proposed as a therapeutic option for prevention and treatment of mild hypoglycemia in individuals with type 1 diabetes (T1D) 1, 2. It provides a non-caloric treatment alternative, particularly when oral glucose intake is less desirable, including immediately after meals and during nighttime. However, its efficacy may be influenced by factors such as high circulating insulin levels and depleted hepatic glycogen stores 3-5. In this study, we aimed to investigate whether dietary macronutrient composition affects the glycemic response to low-dose glucagon following fasted exercise in adults with insulin pump-treated T1D. In a randomised crossover design, participants followed three distinct diets for 7 days each. On day eight of each intervention period, they attended an in-clinic visit, where they undertook exercise in an overnight fasted state (~10–12 h) followed by the administration of low-dose glucagon. Between intervention periods there was a washout of 5–35 days. Participants were recruited from Steno Diabetes Center Copenhagen, Denmark. After providing informed consent, they were screened for eligibility based on the following inclusion criteria: age ≥ 18 years, diagnosed with T1D for ≥ 5 years, use of non-automated insulin pump for ≥ 1 year, use of sensor-based glucose monitoring for ≥ 3 months, HbA1c ≤ 69 mmol/mol, self-reported hypoglycemia awareness and exercising at least 30 min at moderate or vigorous intensity twice per week. The main exclusion criteria were: use of drugs affecting glucose metabolism (other than insulin) during the study or within 30 days prior to its start, severe cardiopulmonary disease, pregnancy, or breastfeeding. Detailed methods and results of the dietary and exercise study components of the study have been published previously 6, 7. In brief, the three diets were isocaloric but differed in macronutrient composition. One had a high carbohydrate (HC) content (minimum 250 g/day) and two had a low content (maximum 100 g/day). The HC diet provided 48% of energy from carbohydrates, 33% from fat and 19% from protein. The two low carbohydrate diets included a low carbohydrate, high fat (LCHF) diet with 19% of energy from carbohydrates, 62% from fat and 19% from protein and a low carbohydrate, high protein (LCHP) diet, which provided 19% of energy from carbohydrates, 57% from fat and 24% from protein. All diets were tailored by a registered dietitian to meet participant preferences and estimated energy needs. After an overnight fast, exercise was performed in the morning on a stationary bicycle for 45 min at 60% of the individual's highest rate of oxygen consumption (V̇O2peak), which had been assessed prior to the first intervention period using a cardiopulmonary exercise test to volitional exhaustion. The bicycling was followed by a 90-min rest. Then participants received a subcutaneous injection of 150 μg of glucagon (GlucaGen, Novo Nordisk, Denmark) and were observed for an additional 120 min of rest. The target plasma glucose level at fasted exercise onset was 7–10 mmol/L. To achieve this, minor insulin or carbohydrate corrections were permitted. Basal insulin rates were identical across intervention periods and independent of diet type. Rates were temporarily reduced by 50% from 90 min before to immediately after exercise and then returned to 100%. Automated suspend functions were deactivated during the entire in-clinic stay. Blood samples were drawn frequently from an intravenous catheter placed in an antecubital vein for analysis of plasma glucose using YSI (YSI Inc. Ohio, USA), serum insulin using Mercodia Iso-Insulin Elisa (Mercodia AB, Uppsala, Sweden) and plasma glucagon via a sensitive C-terminal specific glucagon radioimmunoassay 8. To account for potential differences in baseline plasma glucose, both absolute and incremental values were calculated following glucagon administration. Incremental peak was defined as the maximum increase relative to the baseline at glucagon injection and incremental area under the curve (AUC) as the area between the glucose curve and this baseline value over the observation period. Data reported here were collected as predefined, exploratory endpoints of the randomised cross-over study; however, no power calculation was conducted for this specific substudy as the sample size was determined based on the primary outcomes published previously 7. All analyses followed the intention-to-treat principle, analysing participants in their originally assigned groups, regardless of protocol deviations. A linear mixed-effects model was used to compare continuous data. For non-parametric data, Friedman's test was applied. All statistical analyses were conducted using SAS 9.4 (SAS Institute, NC, USA). The study was approved by the Regional Committee on Health Research Ethics (H-21042230) and the Danish Data Protection Agency (P-2021-826), carried out in accordance with the principles of the Declaration of Helsinki and registered at clinicaltrials.gov. Twelve participants (eight men and four women) with a median age of 50 (range 22–70) years and diabetes duration of 25 (11–52) years were included. Median HbA1c was 57 (37–67) mmol/mol and BMI was 27.3 (21.3–35.9) kg/m2. During the previous 2 weeks median carbohydrate intake was 166 (62–263) g/day and total insulin dose was 35.7 (18.4–127.0) U/day, corresponding to 0.4 (0.3–1.0 U/kg/day). Nine used insulin aspart (Novorapid, Novo Nordisk, Denmark) and three used faster-acting insulin aspart (Fiasp, Novo Nordisk, Denmark). During the three intervention periods, mean (IQR) deviations from the planned carbohydrate intake were: HC −4.8 (−25.3; 9.9) g/day, LCHF 10.7 (−3.1; 28.4) g/day and LCHP 5.7 (−12.2; 2.5) g/day. Mean plasma glucose concentrations at the time of low-dose glucagon administration and subsequent peak glucose values were significantly different between the three diets (p = 0.003 and p = 0.021, respectively) (Table 1 and Figure 1, panel A). However, there were no significant differences in neither incremental peak plasma glucose values (p = 0.348) nor in the incremental area under the plasma glucose curve (p = 0.550) in the 120-min observation period after low-dose glucagon administration (Figure 1, panel B). There were no hypoglycemia events. Plasma glucagon concentrations increased significantly after the administration of low-dose glucagon. There was no difference in baseline values, peak values, or the area under the curve between the three diets (p = 0.106, p = 0.656, p = 0.823, respectively) (Figure 1, panel C). Serum insulin concentrations were similar between diets at the time of low-dose glucagon administration (p = 0.295) and remained stable throughout the observation period (Figure 1, panel D). Diet composition in the preceding 7 days did not affect the glycemic response to 150 μg of glucagon administered 90 min after fasted morning exercise. Although glucagon is currently approved only for the treatment of severe hypoglycemia, its feasibility as part of dual-hormone therapy for T1D has been demonstrated in both open- and closed-loop systems using multiple daily injections and insulin pumps 9-13. We have previously shown that the efficacy of low-dose glucagon in treating insulin-induced hypoglycemia was diminished after 7 days of low carbohydrate intake ( 250 g/day) 3. In this study we explored whether this difference would be modulated if study participants undertook fasted exercise prior to glucagon administration. Contrary to our previous findings, in the present study we did not observe a difference in the glycemic response to low-dose glucagon administration between diet types. Potential explanations remain speculative and include the higher daily carbohydrate intake (~100 vs. ~50 g/day), the glycogen-depleting effect of fasted exercise, which likely dominated and masked any diet-dependent differences and the relatively high glucose levels at the time of glucagon administration, which differed between diets and may have acted as a confounding attenuating observable differences. However, this exploratory substudy included only 12 participants and was not formally powered; therefore, the null finding should be interpreted with caution. Finally, the 150 μg dose was selected based on prior experience to elicit a measurable yet clinically relevant glucose response, although it cannot be excluded that different glucagon doses might have revealed diet-dependent differences. A limitation of the study is the lack of direct assessment of hepatic glycogen stores. Given that the study hypothesis relates to glycogen availability, this limits mechanistic interpretation and conclusions must therefore remain inferential. Nonetheless, this study contributes to our understanding of the interaction between diet, fasted exercise and the main glucoregulatory drugs in type 1 diabetes, insulin and glucagon—an interaction that should be further investigated in parallel with the development of new dual-hormone drug delivery systems. The authors have nothing to report. S.S has received speaker's fees from Novo Nordisk and Nordic Infucare, has received consulting fees from Novo Nordisk, Hedia and Abbott and has been employed with Novo Nordisk. K.B.K. and O.M.M. have no conflicts of interest. R.M.B. has received fees for participating in educational activities by Novo Nordisk, Abbott, Beneo and Sanofi. J.J.H. author declares no conflicts of interest. K.N. has received fees to her institution for participating in advisory boards from Abbott, Medtronic, Tandem, Novo Nordisk and Convatec and for lecturing from Sanofi, Novo Nordisk, Roche, Medtronic, Abbott and Dexcom. Her institution received funding for studies she performed from Zealand Pharma, RSP Systems, Novo Nordisk, Medtronic and Dexcom. A.G.R. author declares no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request. The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/dom.70777.
SCHMIDT et al. (Fri,) studied this question.