GRAPHICAL ABSTRACT

Highlights

• The glycemic benefits of IDegAsp over IGlar may not be broadly applicable.

• Switching to IDegAsp improved glycemic control in individuals overbasalized on IGlar.

• CGM showed IDegAsp improved both prandial and later glycemia after injection.

• HbA1c benefits from IDegAsp were linked to impaired GLP-1 response to oral glucose.

INTRODUCTION

Type 2 diabetes mellitus (T2DM) is characterized by a progressive decline in insulin secretion, and many individuals with T2DM eventually require insulin therapy [1,2]. Generally, a once-daily injection of long-acting insulin analogs is the initial step. If basal insulin therapy alone does not provide adequate glycemic control, patients may be considered adding mealtime (bolus) insulin or switching to a premixed insulin formulation [3,4]. Given that the Basal-Bolus regimen involves multiple injections and intricate titration processes, premixed insulin formulations offer greater convenience for patients. However, there remains a concern regarding potential interactions among the ingredients, which could elevate the risk of hypoglycemia [5].

Insulin degludec/insulin aspart (IDegAsp) is the first combination of a long-acting insulin analog and a rapid-acting one. Because insulin degludec and insulin aspart remain separate entities without chemical interaction in the co-formulation [6], its hypoglycemic profile is improved compared to previous biphasic insulin aspart 30 [7]. The addition of prandial insulin formula is an intensifying therapy, making a once-daily IDegAsp injection theoretically more effective in reducing glycosylated hemoglobin (HbA1c) than basal insulin therapy. Despite the success of a proof-of-concept trial with European participants with T2DM [8], subsequent large-scale trials comparing the efficacy of once-daily IDegAsp and long-acting insulin analogs, such as insulin glargine (IGlar), have shown inconsistent results. A study in Japan observed a significant improvement in HbA1c with IDegAsp compared to IGlar [9], whereas another study found IDegAsp to be similar to IGlar in HbA1c regulation, but associated with more overall hypoglycemia [10].

This inconsistency led us to investigate how to select patients for whom IDegAsp may be a safer and more effective regimen than basal insulin. In a retrospective study involving Koreans with T2DM from three referral hospitals, we found that switching from basal insulin to IDegAsp resulted in a significant reduction in HbA1c after 6 months, especially in those with a lower fasting plasma glucose (FPG) than predicted by regression from the concurrent HbA1c [11,12].

An FPG lower than that predicted by HbA1c may indicate overbasalization, marked prandial hyperglycemia, and/or high glycemic variability including hypoglycemia. However, these could not be assessed in the retrospective analyses. Notably, insulin therapy is often necessary for patients with diminished β-cell function, particularly in elderly populations. Given the higher risk of hypoglycemia in these groups, the addition of bolus insulin must be carefully and precisely controlled.

Therefore, we conducted the current prospective study with the following aims: (1) to confirm glycemic benefit of IDegAsp over basal insulin therapy; (2) to visualize the glycemic profiles using continuous glucose monitoring (CGM); and (3) to investigate the mechanisms mediating its glycemic efficacy. We specifically focused on incretin hormones, given their close relationship with β-cell function and postprandial glycemia [13]. Collectively, these findings may contribute to a more personalized approach to insulin therapy.

METHODS

Trial design

This is a prospective, randomized, open-label, cross-over, treat-to-target clinical study comparing the efficacy and safety of once-daily IDegAsp to IGlar. The randomization table for sequence allocation was generated using random number generation in an Excel program, and the administration sequence was assigned in a 1:1 ratio using a randomization method. Each treatment period lasted 20 weeks without a washout period. Before randomization, all participants underwent baseline blood chemistry, urine analyses, a blinded CGM for 1 week, and an oral glucose tolerance test (OGTT). The blinded CGM was also applied 1 week prior to the completion of each treatment. HbA1c was measured every 10 weeks after randomization (Supplementary Fig. 1).

Participants

The enrollment criteria were as follows: adults aged 18 years or older with T2DM who were regularly prescribed in a referral hospital; on a once-daily basal insulin (glargine U-100, glargine U-300, or degludec) for at least 4 months at a dose of less than 1.0 unit/kg/day; HbA1c levels between 7.0% and 10.0%; an FPG lower than that predicted by concurrent HbA1c at least two times consecutively. The prediction formula (predicted FPG [mg/dL]=12.6×HbA1c [%]+30) was based on our previous study involving 324 Koreans with T2DM on basal insulin therapy [12]. There were no restrictions on the use of non-insulin anti-diabetic drugs for enrollment.

Exclusion criteria included pregnancy; lactation; cardiovascular disease, cancer, or other serious diseases diagnosed within 6 months prior to screening; severe hypertension; a medical condition or medications that may significantly affect blood glucose levels; and anticipated significant lifestyle changes during the study (e.g., highly variable eating habits, night/evening shift work).

Treatment

Participants were administered IDegAsp (100 IU/mL) or IGlar U-100 (100 IU/mL) subcutaneously just prior to the main meal of the day. The initial insulin dose was identical to the previous insulin dose and was then titrated weekly based on the participants’ mean values of pre-breakfast self-measured plasma glucose (SMPG) (target: 91 to 125 mg/dL) according to a pre-defined titration algorithm (Supplementary Table 1). Participants were requested to perform at least three SMPG measurement per week, including two pre-breakfast SMPG. Additional SMPG was recommended when hypoglycemic symptoms occurred. Concomitant other anti-diabetic agents were not changed until the end of study.

Outcomes

The primary endpoint was the change in HbA1c after 20 weeks of treatment from baseline. Key secondary endpoints included differences in hypoglycemic events and CGM parameters between the treatments. Additional outcomes included differences in body weight, insulin dose, and the occurrence of adverse events. Overall hypoglycemia was evaluated using self-reported hypoglycemic events. Confirmed hypoglycemia was defined as a glucose level of less than 70 mg/dL measured by SMPG.

Laboratory measurements and incretin response

A 75-g OGTT was performed after an overnight fasting. Venous blood was drawn every 30 minutes for 120 minutes for the measurement of plasma glucose using the hexokinase method (Roche, Basel, Switzerland). Concentrations of C-peptide, glucagon, glucose-dependent insulinotropic polypeptide (GIP), and glucagon-like peptide-1 (GLP-1) were measured at 0, 30, and 60 minutes, using enzyme-linked immunosorbent assay kits: Mercodia AB, Sylveniusgatan, Sweden (glucagon, sensitivity 1.5 pM, inter-assay coefficient of variation [CV] <15%), Beckman coulter, CA, USA (C-peptide, imprecision ≤10% at concentrations ≥0.1 ng/mL), and Millipore, Burlington, MA, USA (total GIP [sensitivity 4.2 pg/mL, inter-assay CV <10%], and total GLP-1 [sensitivity 1.5 pM, inter-assay CV <12%]).

The incretin response was assessed by calculating the ratio of 30-minute measurements to baseline measurements during an OGTT. Participants with a ratio below the median were classified as having an impaired incretin response.

Continuous glucose monitoring

CGM was performed for 1 week before randomization and at the end of each treatment using the Medtronic iPro2 CGM system (MMT-7745) and Enlite sensor (Medtronic, Minneapolis, MN, USA), along with a food diary. Various glycemic parameters including hypoglycemic events (glucose measures less than 70 mg/dL for 15 consecutive minutes or longer) were analyzed [14]. Twenty-four-hour glucose profiles were depicted by the medians of each person.

Statistical analysis

Assuming a decrease in HbA1c of 0.3% [15] and a standard deviation (SD) of 0.75%, the sample size was calculated to detect a difference in HbA1c changes with a significance level of 0.05 and power of 0.8. The required number of participants was calculated as 52, and considering a dropout rate of 6%, a total of 55 were planned to be recruited.

Values for continuous variables are presented as mean±SD and median (interquartile range). Categorical variables are presented as numbers (%). The outcomes were analyzed by intention-to-treat analyses. Changes in HbA1c after 20 weeks of administration of IDegAsp or IGlar were compared using a linear mixed model with sequence, period, and treatment as fixed effects, and subjects as random effects. Hypoglycemic episodes were analyzed using a negative binomial log-linear mixed model with subjects as a random effect; sequence, period, and treatment as fixed effects; and the logarithm of the treatment period as offset. CGM parameters at the end of each treatment were compared by paired t-test or Wilcoxon signed rank test. Differences in the baseline characteristics were analyzed using Student’s t-test and Mann-Whitney test for continuous variables and chi-squared test for categorical variables. Dynamics parameters during the OGTT were analyzed using a two-way repeated measures analysis of variance (ANOVA). Relationships between variables were evaluated using Spearman rank-order correlation analysis.

P values <0.05 were considered statistically significant. Statistical analyses were performed using SPSS for Windows version 27.0 (IBM Corp., Armonk, NY, USA), Prism version 9.5.1 (GraphPad Software Inc., San Diego, CA, USA), and SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).

Ethical statement

This study was conducted in accordance with the Declaration of Helsinki. The study protocol was approved by the Institutional Review Board of Seoul National University Hospital (IRB No. 1911-023-1076), and registered at cris.nih.go.kr (KCT0004597). Participants voluntarily agreed to participate in the study after being informed of and understanding the full explanation of the study. Written informed consents was obtained from all participants.

RESULTS

Participants characteristics

Among 469 screened patients, 204 were identified as eligible, and 149 declined to participate. A total of 55 participants were enrolled, with 28 randomized to the IDegAsp-IGlar sequence, and 27 to the IGlar-IDegAsp sequence. While four participants dropped out during IGlar treatment, only one participant dropped out during IDegAsp treatment. Overall, 50 participants completed the study (Supplementary Fig. 2).

Baseline characteristics of the participants are summarized in Table 1. In brief, the mean age was 65±10 years, 22 participants (40%) were male, and the mean body mass index (BMI) was 25.0±3.4 kg/m². About half of the participants had retinopathy, and a third had cardiovascular diseases. Most participants were taking IGlar at baseline (median dose 22 unit/day), with mean HbA1c of 8.3%±0.6% and FPG of 102.6±19.5 mg/dL. There were no significant differences in the clinical features between participants in the IDegAsp-IGlar sequence and those in the IGlar-IDegAsp sequence, including HbA1c and FPG.

HbA1c changes

After 20 weeks of treatment, HbA1c decreased by 0.24%±0.72% points with IGlar and by 0.46%±0.83% points with IDegAsp (Fig. 1A). The mean estimated treatment difference in HbA1c changes was –0.21% points (95% confidence interval, –0.39 to –0.02; P=0.031). HbA1c levels tracked over time showed that the treatment effect of IGlar appeared to rebound after 10 weeks to 8.02%±0.74% at 20 weeks, although this increase was not statistically significant. In contrast, the HbA1c levels for IDegAsp remained stable at 7.82%±0.76%, leading to a significant difference from IGlar at 20 weeks (Fig. 1B). Since HbA1c levels below 8% are primarily influenced by postprandial glucose excursions [16], a subgroup analysis was performed on participants with a baseline HbA1c below 8% (n=18). It demonstrated a similar benefit of IDegAsp over IGlar in reducing HbA1c, with a mean estimated treatment difference of –0.20% points; however, statistical significance was not achieved due to the small sample size.

Hypoglycemia and other results

Throughout the entire treatment periods, similar proportions of participants experienced at least one episode of overall hypoglycemia and confirmed hypoglycemia (approximately 47% and 33%, respectively). The estimated annual incidence of overall hypoglycemia was 3.2 episodes per person with IDegAsp and 4.0 with IGlar (P=0.625). That of confirmed hypoglycemia was 2.3 with IDegAsp and 3.1 with IGlar (P=0.924) (Table 2). The frequencies of hypoglycemia-regardless of the level and time-detected by 1-week blinded CGM were also comparable between the treatments (Supplementary Table 2). The comparable distributions of hypoglycemia frequencies are depicted in Supplementary Fig. 3. Furthermore, subgroup analysis of participants with a baseline HbA1c below 8% (n=18) showed similar trends: the estimated annual incidence of overall hypoglycemia was 2.3 episodes per person with IDegAsp compared to 3.3 with IGlar.

At the end of each treatment period, daily insulin doses significantly increased and were comparable between the treatments (P=0.881): 25.9±13.2 units (0.40±0.17 units/kg) for IDegAsp and 25.5±13.4 units (0.39±0.18 units/kg) for IGlar. Body weight didn’t increase significantly with either treatment: the mean weight gain was 0.13±1.92 kg with IDegAsp and –0.03±2.04 kg with IGlar (P=0.681). No other drug-related adverse events were reported.

CGM changes

Twenty-four-hour glucose profiles from the blinded CGM conducted before randomization and at the end of each treatment period are presented in Fig. 1C. Insulin was administered with breakfast in 75% of participants, with lunch in 8%, and with dinner in 17%. When the CGM measurements at the end of each treatment period were compared, favorable prandial excursions were noted with IDegAsp compared to IGlar, with statistical significance in median glucose levels between the two treatments at several time points (Table 3). Conversely, glucose levels at 5:00 AM were significantly higher on IDegAsp than on IGlar. There were no significant differences in other parameters, including time in range (70 to 180 mg/dL), time above range, time below range, and SD.

Subsequently, participants were sorted according to their HbA1c responses to IDegAsp. Participants divided by the median of HbA1c differences between the two treatments demonstrated comparable baseline characteristics, including HbA1c and FPG (Supplementary Table 3). The CGM profiles at the end of treatments for higher responders to IDegAsp indicated that the glycemic benefits of IDegAsp were observed from pre-lunch to bedtime, as much as at post-breakfast when rapid-acting insulin was most frequently administered (Fig. 1D, ‘Higher responders to IDegAsp’ in Supplementary Table 4). Given the low rates of insulin injection with lunch and dinner, these findings may result not only from the direct effects of insulin injections. Lower responders to IDegAsp exhibited comparable prandial excursions between the treatments, but statistically higher glucose levels at 5:00 AM on IDegAsp than on IGlar (P<0.05) (‘Lower responders to IDegAsp’ in Supplementary Table 4), potentially due to a smaller amount of the basal insulin component.

When the CGM profiles were depicted according to the insulin regimens, discernable differences in IDegAsp efficacy between the two responder groups were observed mainly in the afternoon and evening (Fig. 1E), but there was no statistical significance (P=0.073 at 6:00 PM). The CGM profiles on IGlar were comparable between the responders to IDegAsp (data not shown).

Clinical characteristics associated with HbA1c benefit of IDegAsp compared to IGlar

Some parameters were explored to identify characteristics associated with these differences in the response to IDegAsp. First, the ratio of FPG to estimated average glucose (eAG) [17] was examined based on the study’s hypothesis that a low FPG considering HbA1c would predict the benefit. However, this ratio was not related to the difference in HbA1c changes between the treatments (Fig. 2A).

Next, incretin response during an OGTT was examined. When comparing the 30-minute incretin ratios relative to baseline between high and low responders to IDegAsp, the GLP-1 ratio was significantly lower in high responders, while the GIP ratio was comparable between the groups (Fig. 2B, Supplementary Table 3). The GLP-1 ratio was significantly correlated with the difference in HbA1c changes between treatments (r=0.301, P<0.05) (Fig. 2C), whereas the GIP ratio was not (Fig. 2D). Participants were then divided into groups based on the median GLP-1 ratio of 1.2, which was also identified as the optimal cutoff for assessing the benefit of IDegAsp through receiver operating characteristic analysis (sensitivity: 0.739, specificity: 0.654). The difference in HbA1c changes between the two treatments was significantly greater in those with a lower GLP-1 ratio compared to those with a higher GLP-1 ratio (–0.525% vs. –0.011%, P<0.01) (Fig. 2E).

Hormonal profiles according to GLP-1 response

Subsequently, clinical and hormonal characteristics were analyzed based on the GLP-1 ratio of 1.2 (Supplementary Table 5). The lower GLP-1 responder group included slightly more males and demonstrated slightly higher HbA1c and fasting glucose levels during the OGTT compared to the other group (P=0.05–0.10). Between the GLP-1 responder groups, there were no significant differences in age, BMI, the FPG-to-eAG ratio, or the discrepancy between measured and predicted FPG.

During the blinded CGM before randomization, mean glucose and time above range were significantly higher, while time in range was significantly lower in the lower GLP-1 responders. The CV was comparable between GLP-1 responders (Supplementary Fig. 4).

During the OGTT, the lower GLP-1 responders maintained a lower ratio at 60 minutes, while the baseline GLP-1 levels were comparable (Fig. 3A). We found that the GIP profile was contrary to the GLP-1 profile: the lower GLP-1 responders demonstrated higher GIP responses, but the GIP ratios to baseline were comparable between the GLP-1 responders (Fig. 3B). The C-peptide profile was not statistically different by the GLP-1 ratio (Fig. 3C). However, a lower GLP-1 response resulted in a lower C-peptide ratio (P<0.005) (Supplementary Fig. 5) and a lower incremental area under the curve (AUC) for C-peptide adjusted for the glucose incremental AUC (P=0.084) (Supplementary Table 5). The difference in glucagon suppression in response to glucose loading (available for only 27 participants) was also significant between the GLP-1 responders (Fig. 3D). While glucose excursions were comparable between the groups (Fig. 3E), the increase at 120 minutes relative to baseline was significantly lower in the low GLP-1 responders (2.78 [2.05 to 3.29] vs. 3.35 [2.73 to 4.14], P<0.05).

DISCUSSION

In this prospective cross-over study on participants with signs of overbasalization, we observed that once-daily IDegAsp yielded a significant reduction in HbA1c compared to IGlar, with no significant differences in the occurrence of hypoglycemia. The CGM profiles suggested that the HbA1c efficacy of IDegAsp over IGlar may result not only from improved prandial glycemia at the meals with insulin injection, but also from improved premeal glycemia following the relived prandial excursions. It was not the FPG-to-eAG ratio but the GLP-1 response that significantly correlated with the HbA1c benefits of IDegAsp compared to IGlar. Even though real-world data indicate that switching to IDegAsp in T2DM is associated with a significant reduction in HbA1c and a lower incidence of hypoglycemia compared to previous medications [18,19], randomized controlled trials have shown inconsistent efficacy of once-daily IDegAsp compared to long-acting insulin analogs [9,10, 20,21]. This suggests the need to identify a subset of patients who would derive greater benefits from the switch.

β-Cell insufficiency might be a predictive factor for the benefit of IDegAsp, because a large randomized controlled trial demonstrating the benefit was conducted with Japanese participants [9], and Asian individuals are known to have lower insulin secretion compared to white individuals [22]. However, in our study conducted among Asian participants, differences in insulin secretion, as evaluated during an OGTT, were not correlated with the glycemic efficacy of IDegAsp compared to IGlar.

Instead, we identified a novel indicator of the benefit: a low GLP-1 response to oral glucose loading. GLP-1 and GIP are incretin hormones secreted from the small intestine that increase after nutrient ingestion, stimulating insulin secretion from pancreatic β-cells. Since the endocrine pancreas remains responsive to GLP-1 in T2DM [23], and a blunted GLP-1 response is associated with decreased glucose-stimulated insulin secretion, we can expect that bolus insulin administration with IDegAsp ultimately resulted in a better postprandial glucose-lowering effect in this group. Unlike GLP-1, the insulinotropic effect of GIP is significantly reduced in T2DM [24]. In our study, changes in GIP were not associated with changes in HbA1c or glucose-stimulated insulin secretion, suggesting that the GIP response likely did not contribute to the benefit of IDegAsp compared to IGlar.

In addition to its incretin effect, GLP-1 also influences gastric emptying by slowing it down, which helps attenuate postprandial glycemic excursions [25]. On the other hand, GIP has no effects on gastric emptying. It is recognized that gastric emptying is abnormally slow in 30% to 50% of individuals with long-standing, complicated diabetes, such as the participants in the current study [26]. However, those with a lower GLP-1 ratio may have faster gastric emptying, as suggested by CGM findings showing a steep excursion of prandial glycemia (Supplementary Fig. 4). Additionally, the earlier glycemic drop at 120 minutes during the OGTT in the lower GLP-1 responders may also indicate facilitated gastric emptying (Fig. 3E). Notably, in insulin-treated patients, disordered gastric emptying can cause a mismatch between the timing of insulin action and the availability of absorbed carbohydrate [26]. In this context, impaired GLP-1 release among patients on basal insulin therapy is likely to lead to postprandial glycemic patterns that are more closely aligned with the pharmacokinetic profile of IDegAsp, potentially enhancing its effectiveness.

Because gastric emptying regulates the entry of nutrients into the small intestine, it modulates the secretion of both incretins [26]. Regardless of the presence of diabetes, rapid duodenal glucose infusion has been shown to induce an increase in plasma incretin levels [27]. The higher GIP response observed among the lower GLP-1 responders in this study further suggests accelerated gastric emptying in this group.

If a blunted GLP-1 response is a determinant of the glycemic benefit of IDegAsp, previous findings that IDegAsp is more effective at improving hyperglycemia when administered with dinner rather than breakfast [21,28] can be explained by this mechanism. Since the GLP-1 response is more pronounced in the morning compared to the afternoon [29,30], administering IDegAsp at bedtime can be more effective. In our current study, the benefits of prandial glucose excursions by IDegAsp were not distinctive after dinner; however, this could be attributed to the low injection rates at dinner (17%). When we conducted a subgroup analysis with those who injected at dinner, glucose excursions at 9:00 PM were significantly lower with IDegAsp compared to IGlar (167.6±35.1 mg/dL vs. 212.0±57.4 mg/dL, P=0.043).

Plasma GLP-1 levels have been suggested to be influenced by various factors such as plasma dipeptidyl peptidase-4 (DPP-4) levels, race, age, gender, BMI, metformin use, and diabetic status [31-36]. Genetic factors may also play a role [37,38]. Participants in our study were taking metformin, with/without DPP-4 inhibitors, which could have affected GLP-1 levels. However, the crossover design and protocol to keep anti-diabetic medications consistent throughout the study period—except for insulin dosing—likely minimized their potential confounding effects.

Considering that a low GLP-1 response may be a key determinant to intensify basal insulin therapy, a combination with a GLP-1 receptor agonist (GLP-1RA), rather than rapid-acting insulin, could be a more effective option. In fact, most treatment guidelines prioritize GLP-1RAs over insulin [3,4]. Particularly, GLP-1RAs effectively slow gastric emptying, with this effect being more pronounced in short-acting GLP-1RAs compared to long-acting ones [26], further attenuating postprandial hyperglycemia. Moreover, the weight-reducing effects of GLP-1RAs may improve insulin sensitivity, providing additional metabolic benefits. These effects are difficult to achieve with rapid-acting insulins.

However, GLP-1RAs suppress glucagon secretion, which could increase the risk of hypoglycemia in patients with β-cell deficiency: especially those with a lower GLP-1 response, who demonstrated a higher tendency for time below range (Supplementary Table 5). This concern has been highlighted in a trial involving participants with type 1 diabetes mellitus [39]. Therefore, head-to-head comparisons in these specific groups are necessary, particularly for single-injection therapies using fixed-ratio regimens, since randomized controlled trials in this context remain limited.

In this study, there was no significant difference in the hypoglycemia between IGlar and IDegAsp, including nocturnal hypoglycemia evaluated by the CGM. Historically, IDegAsp has been reported to result in less frequent nocturnal hypoglycemia compared to IGlar, even in participants experiencing more overall hypoglycemia [10,20]. This favorable effect on the nocturnal hypoglycemia with IDegAsp has been attributed to its stable and ultra-long action component [21,40]. Additionally, relief from overbasalization may play a role in the benefit of IDegAsp on hypoglycemia, as reflected by the CGM analyses demonstrating higher glucose levels at dawn with IDegAsp compared to IGlar.

The strength of this study lies in its crossover design, wherein participants underwent both treatments, minimizing variability based on individual characteristics. In addition, CGM also allowed us to investigate diurnal glycemic fluctuations, including pre- and post-meal periods, and facilitated a more comprehensive analysis of hypoglycemia. Furthermore, the evaluation of incretin secretion and β-cell function contributed to predict treatment efficacy.

A limitation of our study is that there was no washout period for drug crossovers. Considering potential carryover effects, this may weaken the robustness of the results. However, stopping insulin for a washout period could be harmful for participants already undergoing insulin therapy. Given that the primary outcome was HbA1c which reflects accumulated glycemic control over several months, and that the 20-week treatment period was sufficiently long, we considered the impact of omitting the washout period to be acceptable.

Next, insulin levels and active GLP-1 were not assessed during the OGTT, and stimulated hormones were measured only at the 30- and 60-minute points, rendering the study incomplete. Furthermore, while we have analyzed a simple ratio of total GLP-1 levels which may be feasible in clinical practice, a more comprehensive approach incorporating active GLP-1 and dynamic modeling would enhance the robustness of GLP-1 response assessment.

Additionally, several potential confounders should be considered: the open-label design, the relatively small sample size, and the choice of IGlar as the comparator. Insulin degludec would have been a more suitable comparator for mechanistic comparisons with IDegAsp, as differences in outcomes might be partially due to variations in the basal insulin properties rather than the inclusion of the prandial component in IDegAsp. However, IGlar was chosen in this study to build upon previous research and to align with real-world practice, as it was the representative basal insulin in routine clinical use at the time [9,11,12,20]. Finally, all participants were Korean patients from a tertiary hospital, which may limit generalizability.

In conclusion, our study demonstrated that once-daily IDegAsp provides a greater reduction in HbA1c levels compared to IGlar, with this benefit significantly correlated with a blunted GLP-1 response. Given further investigations in larger, diverse clinical populations, assessing the GLP-1 response to glucose may help identify patients who could benefit from transitioning to IDegAsp, contributing to a more personalized medicine.

SUPPLEMENTARY MATERIALS

NOTES

CONFLICTS OF INTEREST

Hye Seung Jung received a research fund from Novo Nordisk. Andrea Mari is a consultant for Novo Nordisk and Lilly.

Soo Heon Kwak has been associate editor of the Diabetes & Metabolism Journal since 2022. Kyong Soo Park has been honorary editor of the Diabetes & Metabolism Journal since 2020. Hak Chul Jang has been International Editorial Board Member of the Diabetes & Metabolism Journal since 2023. They were not involved in the review process of this article. Otherwise, there was no conflict of interest.

AUTHOR CONTRIBUTIONS

Conception or design: A.M., H.C.J., H.S.J.

Acquisition, analysis, or interpretation of data: all authors.

Drafting the work or revising: all authors.

Final approval of the manuscript: all authors.

FUNDING

This study was supported by grants from the NRF of the Korean Ministry of Science and ICT (2022R1A2C2004570), and funded by Novo Nordisk. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

ACKNOWLEDGMENTS

During the preparation of this work the author(s) used Chat GPT in order to improve language and readability. After using this tool, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.

All data supporting the findings of this study are available within the paper and its Supplementary Information. Additional data sets generated during the current study are available from the corresponding author on reasonable request.

Fig. 1.

Changes in glycemic control by treatments. (A) Changes in glycosylated hemoglobin (HbA1c) from baseline at 20 weeks. (B) Changes in HbA1c over time. (C, D, E) Twenty-four-hour glucose profiles from blinded continuous glucose monitoring (CGM) during the last week of treatments: (C) all participants, including the baseline profile (dotted line); (D) profiles according to response to insulin degludec/insulin aspart (IDegAsp); (E) profiles according to insulin regimens. Data are presented as (A, B) mean±standard error, (C, D, E) median (interquartile range), and median only for baseline measures in (C). IGlar, insulin glargine; NS, no significant. ^aP<0.05 using a linear mixed model, ^bP<0.05 by paired t-test between the treatments.

Fig. 2.

Clinical characteristics associated with glycosylated hemoglobin (HbA1c) benefit of insulin degludec/insulin aspart (IDegAsp) compared to insulin glargine (IGlar). (A) Correlation between the difference in HbA1c change (IDegAsp–IGlar) and the fasting plasma glucose (FPG)-to-estimated average glucose (eAG) ratio. (B) Before randomization, participants underwent an overnight fast, followed by a 75-g oral glucose tolerance test. Plasma hormone levels were measured, and incretin responses were assessed by calculating the ratios of 30-minute levels to baseline. These ratios were compared between the IDegAsp responder groups. (C, D) Correlation between the difference in HbA1c change (IDegAsp–IGlar) and the glucagon-like peptide-1 (GLP-1) response (C) and glucose-dependent insulinotropic polypeptide (GIP) response (D). (E) Difference in HbA1c change (IDegAsp–IGlar) between participants with a GLP-1 ratio of 1.2. Data are presented as (B) median (interquartile range) and (E) mean±standard error. Spearman rank-order correlation analysis for (A), (C), and (D). NS, no significant. ^aP<0.05, ^bP<0.01 by Mann Whitney test or Student’s t-test.

Fig. 3.

Responses to a 75-g oral glucose tolerance test according to glucagon-like peptide-1 (GLP-1) ratio. Results are presented based on the GLP-1 ratio at 30 minutes compared to baseline measure. (A) Total GLP-1 levels. (B) Total glucose-dependent insulinotropic polypeptide (GIP) levels. (C) C-peptide levels. (D) Glucagon levels (n=27). (E) Glucose levels. Data are presented as mean±standard error. The P value for each graph indicates the differences between the GLP-1 ratio groups, analyzed using a two-way repeated measures analysis of variance (ANOVA). C-peptide and glucagon levels were analyzed using log-transformed data. ^aP<0.05, ^bP<0.01 by Student’s t-test.

REFERENCES

• 1. Fonseca VA. Defining and characterizing the progression of type 2 diabetes. Diabetes Care 2009;32 Suppl 2:S151-6.ArticlePubMedPMCPDF
• 2. Rudenski AS, Hadden DR, Atkinson AB, Kennedy L, Matthews DR, Merrett JD, et al. Natural history of pancreatic islet B-cell function in type 2 diabetes mellitus studied over six years by homeostasis model assessment. Diabet Med 1988;5:36-41.ArticlePubMed
• 3. American Diabetes Association Professional Practice Committee. 9. Pharmacologic approaches to glycemic treatment: standards of care in diabetes-2024. Diabetes Care 2024;47(Suppl 1):S158-78.PubMedPMC
• 4. Choi JH, Lee KA, Moon JH, Chon S, Kim DJ, Kim HJ, et al. 2023 Clinical practice guidelines for diabetes mellitus of the Korean Diabetes Association. Diabetes Metab J 2023;47:575-94.ArticlePubMedPMCPDF
• 5. Anyanwagu U, Mamza J, Gordon J, Donnelly R, Idris I. Premixed vs basal-bolus insulin regimen in type 2 diabetes: comparison of clinical outcomes from randomized controlled trials and real-world data. Diabet Med 2017;34:1728-36.ArticlePubMedPDF
• 6. Havelund S, Ribel U, Hubalek F, Hoeg-Jensen T, Wahlund PO, Jonassen I. Investigation of the physico-chemical properties that enable co-formulation of basal insulin degludec with fast-acting insulin aspart. Pharm Res 2015;32:2250-8.ArticlePubMedPMCPDF
• 7. Fulcher GR, Christiansen JS, Bantwal G, Polaszewska-Muszynska M, Mersebach H, Andersen TH, et al. Comparison of insulin degludec/insulin aspart and biphasic insulin aspart 30 in uncontrolled, insulin-treated type 2 diabetes: a phase 3a, randomized, treat-to-target trial. Diabetes Care 2014;37:2084-90.ArticlePubMedPDF
• 8. Heise T, Tack CJ, Cuddihy R, Davidson J, Gouet D, Liebl A, et al. A new-generation ultra-long-acting basal insulin with a bolus boost compared with insulin glargine in insulin-naive people with type 2 diabetes: a randomized, controlled trial. Diabetes Care 2011;34:669-74.PubMedPMC
• 9. Onishi Y, Ono Y, Rabol R, Endahl L, Nakamura S. Superior glycaemic control with once-daily insulin degludec/insulin aspart versus insulin glargine in Japanese adults with type 2 diabetes inadequately controlled with oral drugs: a randomized, controlled phase 3 trial. Diabetes Obes Metab 2013;15:826-32.ArticlePubMedPDF
• 10. Kumar A, Franek E, Wise J, Niemeyer M, Mersebach H, Simo R. Efficacy and safety of once-daily insulin degludec/insulin aspart versus insulin glargine (U100) for 52 weeks in insulin-naïve patients with type 2 diabetes: a randomized controlled trial. PLoS One 2016;11:e0163350.ArticlePubMedPMC
• 11. Jang HN, Yang YS, Lee SO, Oh TJ, Koo BK, Jung HS. Favorable glycemic control with once-daily insulin degludec/insulin aspart after changing from basal insulin in adults with type 2 diabetes. Endocrinol Metab (Seoul) 2019;34:382-9.PubMedPMC
• 12. Jang HN, Yang YS, Oh TJ, Koo BK, Lee SO, Park KS, et al. Low fasting glucose-to-estimated average glucose ratio was associated with superior response to insulin degludec/aspart compared with basal insulin in patients with type 2 diabetes. J Diabetes Investig 2022;13:85-93.ArticlePubMedPMCPDF
• 13. Nauck MA, Muller TD. Incretin hormones and type 2 diabetes. Diabetologia 2023;66:1780-95.ArticlePubMedPMCPDF
• 14. Battelino T, Alexander CM, Amiel SA, Arreaza-Rubin G, Beck RW, Bergenstal RM, et al. Continuous glucose monitoring and metrics for clinical trials: an international consensus statement. Lancet Diabetes Endocrinol 2023;11:42-57.PubMed
• 15. European Medicines Agency. Guideline on clinical investigation of medicinal products in the treatment or prevention of diabetes mellitus (2023). Available from: https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-clinical-investigation-medicinal-products-treatment-or-prevention-diabetes-mellitus-revision-2_en.pdf (cited 2025 Jul 7).
• 16. Riddle MC. Basal glucose can be controlled, but the prandial problem persists: it’s the next target! Diabetes Care 2017;40:291-300.ArticlePubMedPDF
• 17. Nathan DM, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine RJ, et al. Translating the A1C assay into estimated average glucose values. Diabetes Care 2008;31:1473-8.ArticlePubMedPMCPDF
• 18. Fulcher GR, Akhtar S, Al-Jaser SJ, Medina J, Mohamed M, Nicodemus NA Jr, et al. Initiating or switching to insulin degludec/insulin aspart in adults with type 2 diabetes: a real-world, prospective, non-interventional study across six countries. Adv Ther 2022;39:3735-48.ArticlePubMedPMCPDF
• 19. Kesavadev J, Gowda A, Kumar H, Yalamanchi SR, Lodha S, Singh KP, et al. Safety of insulin degludec/insulin aspart in patients with diabetes mellitus over a period of 1 year during routine clinical care in India: SMART (Study of Management of Diabetes with Ryzodeg^TM Treatment). Med Sci (Basel) 2021;10:1.PubMedPMC
• 20. Kumar S, Jang HC, Demirag NG, Skjoth TV, Endahl L, Bode B. Efficacy and safety of once-daily insulin degludec/insulin aspart compared with once-daily insulin glargine in participants with type 2 diabetes: a randomized, treat-to-target study. Diabet Med 2017;34:180-8.ArticlePubMedPMCPDF
• 21. Liebl A, Davidson J, Mersebach H, Dykiel P, Tack CJ, Heise T. A novel insulin combination of insulin degludec and insulin aspart achieves a more stable overnight glucose profile than insulin glargine: results from continuous glucose monitoring in a proof-of-concept trial. J Diabetes Sci Technol 2013;7:1328-36.ArticlePubMedPMCPDF
• 22. Ma RC, Chan JC. Type 2 diabetes in East Asians: similarities and differences with populations in Europe and the United States. Ann N Y Acad Sci 2013;1281:64-91.ArticlePubMedPMCPDF
• 23. Nauck MA, Meier JJ. The incretin effect in healthy individuals and those with type 2 diabetes: physiology, pathophysiology, and response to therapeutic interventions. Lancet Diabetes Endocrinol 2016;4:525-36.ArticlePubMed
• 24. Vilsboll T, Krarup T, Madsbad S, Holst JJ. Defective amplification of the late phase insulin response to glucose by GIP in obese type II diabetic patients. Diabetologia 2002;45:1111-9.ArticlePubMedPDF
• 25. Umapathysivam MM, Lee MY, Jones KL, Annink CE, Cousins CE, Trahair LG, et al. Comparative effects of prolonged and intermittent stimulation of the glucagon-like peptide 1 receptor on gastric emptying and glycemia. Diabetes 2014;63:785-90.ArticlePubMedPMCPDF
• 26. Jalleh RJ, Jones KL, Rayner CK, Marathe CS, Wu T, Horowitz M. Normal and disordered gastric emptying in diabetes: recent insights into (patho)physiology, management and impact on glycaemic control. Diabetologia 2022;65:1981-93.ArticlePubMedPMCPDF
• 27. Marathe CS, Rayner CK, Bound M, Checklin H, Standfield S, Wishart J, et al. Small intestinal glucose exposure determines the magnitude of the incretin effect in health and type 2 diabetes. Diabetes 2014;63:2668-75.ArticlePubMedPDF
• 28. Cho KY, Nakamura A, Oba-Yamamoto C, Tsuchida K, Yanagiya S, Manda N, et al. Switching to once-daily insulin degludec/insulin aspart from basal insulin improves postprandial glycemia in patients with type 2 diabetes mellitus: randomized controlled trial. Diabetes Metab J 2020;44:532-41.ArticlePubMedPMCPDF
• 29. Zilstorff DB, Richter MM, Hannibal J, Jorgensen HL, Sennels HP, Wewer Albrechtsen NJ. Secretion of glucagon, GLP-1 and GIP may be affected by circadian rhythm in healthy males. BMC Endocr Disord 2024;24:38.ArticlePubMedPMCPDF
• 30. Lindgren O, Mari A, Deacon CF, Carr RD, Winzell MS, Vikman J, et al. Differential islet and incretin hormone responses in morning versus afternoon after standardized meal in healthy men. J Clin Endocrinol Metab 2009;94:2887-92.ArticlePubMed
• 31. Toft-Nielsen MB, Damholt MB, Madsbad S, Hilsted LM, Hughes TE, Michelsen BK, et al. Determinants of the impaired secretion of glucagon-like peptide-1 in type 2 diabetic patients. J Clin Endocrinol Metab 2001;86:3717-23.ArticlePubMed
• 32. Stenlid R, Manell H, Halldin M, Kullberg J, Ahlstrom H, Manukyan L, et al. High DPP-4 concentrations in adolescents are associated with low intact GLP-1. J Clin Endocrinol Metab 2018;103:2958-66.ArticlePubMedPDF
• 33. Sleddering MA, Bakker LE, Janssen LG, Meinders AE, Jazet IM. Higher insulin and glucagon-like peptide-1 (GLP-1) levels in healthy, young South Asians as compared to Caucasians during an oral glucose tolerance test. Metabolism 2014;63:226-32.ArticlePubMed
• 34. Huang Y, Lou X, Jiang C, Ji X, Tao X, Sun J, et al. Gut microbiota is correlated with gastrointestinal adverse events of metformin in patients with type 2 diabetes. Front Endocrinol (Lausanne) 2022;13:1044030.ArticlePubMedPMC
• 35. Vollmer K, Holst JJ, Baller B, Ellrichmann M, Nauck MA, Schmidt WE, et al. Predictors of incretin concentrations in subjects with normal, impaired, and diabetic glucose tolerance. Diabetes 2008;57:678-87.ArticlePubMedPDF
• 36. Færch K, Torekov SS, Vistisen D, Johansen NB, Witte DR, Jonsson A, et al. GLP-1 response to oral glucose is reduced in prediabetes, screen-detected type 2 diabetes, and obesity and influenced by sex: the ADDITION-PRO study. Diabetes 2015;64:2513-25.ArticlePubMedPDF
• 37. Gjesing AP, Ekstrøm CT, Eiberg H, Urhammer SA, Holst JJ, Pedersen O, et al. Fasting and oral glucose-stimulated levels of glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (GLP-1) are highly familial traits. Diabetologia 2012;55:1338-45.ArticlePubMedPDF
• 38. Almgren P, Lindqvist A, Krus U, Hakaste L, Ottosson-Laakso E, Asplund O, et al. Genetic determinants of circulating GIP and GLP-1 concentrations. JCI Insight 2017;2:e93306.ArticlePubMedPMC
• 39. Mathieu C, Zinman B, Hemmingsson JU, Woo V, Colman P, Christiansen E, et al. Efficacy and safety of liraglutide added to insulin treatment in type 1 diabetes: the ADJUNCT ONE treat-to-target randomized trial. Diabetes Care 2016;39:1702-10.ArticlePubMedPMCPDF
• 40. Heise T, Nosek L, Bottcher SG, Hastrup H, Haahr H. Ultralong-acting insulin degludec has a flat and stable glucose-lowering effect in type 2 diabetes. Diabetes Obes Metab 2012;14:944-50.ArticlePubMedPDF

Figure & Data

References

Citations

Citations to this article as recorded by  

About this article

Jang HN, Hong ES, Yang YS, Lee SO, Jang Mj, Mari A, Kwak SH, Park KS, Jang HC, Jung HS. Glycemic Benefit of Insulin Degludec/Insulin Aspart Compared to Basal Insulin in Type 2 Diabetes Mellitus Associated with Impaired Glucagon-Like Peptide-1 Response: A Randomized Crossover Trial. Diabetes Metab J. 2026;50(3):552-564.

Received: Nov 21, 2024; Accepted: Apr 28, 2025

DOI: https://doi.org/10.4093/dmj.2024.0741.

Download citation ↓