Improved quality of life and glycemic management after switching from conventional rapid-acting insulin to ultra-rapid lispro in patients with diabetes

Abstract

For optimal postprandial glucose (PPG) management, rapid-acting insulin analogs (RAA) should be administered 15 min before a meal; however, this may not be possible for some individuals. Ultra-rapid lispro (URLi) can be administered 0–2 min before or <20 min after a meal, which may improve patient satisfaction and PPG management. In this pilot study, we evaluated changes in quality of life (QOL) and glycemic management among Japanese outpatients with type 2 diabetes mellitus (T2DM) who switched from RAA to URLi. We enrolled 12 outpatients with T2DM and evaluated QOL using the insulin therapy-related (ITR) QOL questionnaire. The primary endpoint was the change in ITR-QOL scores at 12–15 weeks. Endpoints were evaluated using the one-sample Wilcoxon signed rank test or paired t-tests. URLi was associated with a significant increase in ITR-QOL (+15.1 ± 16.1 points, p < 0.01), perception (+7.2 ± 6.9 points, p < 0.01), and status (+7.9 ± 9.5 points, p < 0.05) scores. At 12–15 weeks, the time in range significantly increased (+8.3 ± 9.2, p < 0.05), time above range significantly decreased (−7.7 ± 10.2, p < 0.05), and time below range showed no significant changes. Thus, switching from RAA to URLi significantly improved ITR-QOL questionnaire scores. In summary, URLi is an effective treatment alternative, providing flexible timing, improved glycemic management, and enhanced patient satisfaction.

Highlights

• •Ultra-rapid lispro (URLi) significantly improved the ITR-QOL questionnaire scores.

• •URLi significantly increased TIR and decreased TAR, without increase in hypoglycemia.

• •URLi offers flexible timing for insulin administration with enhanced satisfaction.

1. Introduction

Ultra-rapid lispro (URLi, Lyumjev®) is a new ultra-rapid-acting insulin commercially released in Japan in 2020. The absorption of lispro (Humalog®) is accelerated by adding treprostinil and citric acid, which enhance local vasodilation and vascular permeability, respectively [1,2]. URLi reaches half-maximal drug concentration within 13 min of administration [3], which is significantly faster than lispro and insulin aspart (NovoRapid®) by 13 and 14 min, respectively (both p < 0.0001) [3]. Additionally, URLi significantly reduces postprandial glucose (PPG) levels compared with lispro and insulin aspart by 21 mg/dL and 29 mg/dL, respectively [3]. Thus, URLi is non-inferior to lispro in terms of HbA1c changes and allows good PPG management in patients with type 1 and type 2 diabetes [[4], [5], [6]].

Poor PPG management increases the risk of diabetes-related complications, including cardiovascular diseases. For optimal PPG management, conventional rapid-acting insulin analogs (RAA) should be administered approximately 15 min before a meal [7,8]; however, this timing may not be feasible for some individuals. An online survey of 382 Japanese adult patients with type 1 or type 2 diabetes using RAA revealed that 38.5 % of participants injected insulin from 2 min to immediately before the start of a meal [9]. Additionally, 8.9 % of individuals injected insulin after the start of meals, with 24.9 % of post-meal injections occurring within the past month. The most common reasons were “ate outside of home,” “forgot to administer before meal,” “had lower glucose levels or was hypoglycemic prior to the meal,” “stayed outside/worked outside,” and “could not find a place to inject.” In contrast, URLi can be administered 0–2 min before a meal or <20 min after the start of a meal. Therefore, URLi may address the burdens and unmet needs associated with the timing requirements of conventional RAA administration.

Accordingly, patients struggling with poor glycemic management using RAA may switch to URLi in order to improve their satisfaction and glycemic management. While some reports have examined glycemic changes, to the best of our knowledge, there are no other studies, either domestically or internationally, that have evaluated the numerical changes in the quality of life (QOL) associated with the use of URLi. Therefore, the aim of this pilot study was to evaluate changes in the QOL of outpatients with type 2 diabetes mellitus (T2DM) who switched from conventional RAA to URLi after URLi was considered more appropriate based on the Insulin Selection Checklist. QOL was assessed using the insulin therapy-related (ITR) QOL measure (ITR-QOL questionnaire).

2. Methods

2.1. Participants

We enrolled 12 outpatients with T2DM who visited Minami Osaka Hospital in Japan. The inclusion criteria were as follows: (1) previously used conventional RAA for ≥6 months before study commencement, (2) URLi was considered more appropriate based on the Insulin Selection Checklist, and (3) gave consent to participate in this study. The exclusion criteria were as follows: (1) a history of allergy to lispro; (2) inability to administer insulin within 2 min before a meal or 20 min after starting a meal; (3) patients with T1DM; (4) severe infection or serious injury; (5) surgery planned/undertaken during the study period; (6) a history of severe ketosis, diabetic coma, or pre-coma within 6 months before study commencement; (7) pregnancy, breastfeeding, or planned pregnancy; (8) severe liver dysfunction (Child-Pugh score: ≥10 and ≤ 15); and (9) patients considered unsuitable for participation by the investigator.

2.2. Study design

This single-center, single-arm, open-label, prospective pilot study was approved by the Institutional Review Board of Minami Osaka Hospital (No. 2022-4) on July 04, 2022. The study was registered in the University Hospital Medical Information Network Clinical Trial Registry (UMIN 000048528). In accordance with the 1964 Declaration of Helsinki, the purpose of this study was explained to the patients and written informed consent was obtained.

The study flowchart is illustrated in Fig. 2. At the start of the study, participants were asked to wear a proCGM device (Freestyle Libre Pro; Abbott Diabetes Care, Alameda, CA, USA) for 14 days. All participants received instructions on URLi injection techniques and timing (0–2 min before a meal or within 20 min after the start of a meal) from a physician during the consultation and from a nurse afterwards. Patients were advised to maintain their usual diet. Seven days after wearing the proCGM, participants switched from conventional RAA to URLi with the same insulin units. Insulin units could then be adjusted during regular checkups. Oral medications were continued without any change in regimen. After 12–15 weeks, participants completed the ITR-QOL questionnaire and wore the proCGM again.

Fig. 2.

Flowchart showing the design of this pilot study ITR-QOL: insulin therapy-related quality-of-life measure; proCGM: professional continuous glucose monitoring.

The Insulin Selection Checklist, developed at our hospital to compare the suitability of conventional RAA and URLi, consists of six questions. As URLi is characterized by a rapid onset of action, the following questions were included to understand how much time a patient spends on meals: “Q1. Do you have sufficient time to eat?” and “Q2. How much time do you spend on meals?” Additionally, as some patients may inject insulin after meals, we also included “Q3. Is the amount of food that you eat inconsistent?” To identify the most convenient time for insulin administration, we included “Q4. What is the best time to inject insulin at home?” and “Q5. What is the best time to inject insulin while eating out?” Finally, related to Q5, we included “Q6. Do you eat out often?” to determine the frequency of eating out (Supplementary Fig. 2). Participants were deemed suitable for URLi if the following responses were given: Q1, Difficult to find sufficient time because of work or other reasons; Q2, <20 min; Q3, Yes; Q4, Just before a meal (0–2 min before) or within 20 min after starting a meal; Q5, Just before a meal (0–2 min before) or within 20 min after starting a meal; and Q6, Often. As shown in Supplementary Fig. 2, the decision on whether conventional RAA or URLi was more suitable was based on the number of filled checkboxes.

The ITR-QOL questionnaire was developed by Ishii et al. in 2001 [10]. This questionnaire consists of 23 questions categorized into two domains. The perception domain includes 11 items that assess the extent of perceived psychological burden associated with insulin injections. The status domain includes 12 items that assess the extent to which insulin injections affect patient daily activities (Supplementary Table 1). For Q1–22, the sum of the scores for each item were added, whereas the reverse score was applied for Q23, with their sum being used as the total ITR-QOL questionnaire score. The total score ranges from 23 to 115 points, with higher scores indicating better QOL. The questionnaire items assess factors that affect patient QOL, including the timing of insulin injections, blood glucose management status, and frequency of severe hypoglycemia. The questionnaires were completed during the hospital waiting period.

2.3. Study endpoints

The primary endpoints were changes in the ITR-QOL questionnaire scores at 12–15 weeks. The secondary endpoints were as follows: changes in HbA1c levels; 24-h mean glucose level; TIR (time spent in the target glucose range [70–180 mg/dL]); TAR (time spent above the target glucose range [>180 mg/dL]); TBR (time spent below the glucose level [<70 mg/dL]); percentage of time with nocturnal hypoglycemia (time spent in the glucose level <70 mg/dL during 00:00–00:59); 24-h SD of glucose levels, 24-h CV of glucose levels [11]; 24-h M-value (target glucose level 100 mg/dL) [12]; MAGE [12]; MODD for a 24-h period (average of the difference in proCGM data for days 1–2 and days 2–3 over 3 consecutive days) [12]; 24-h AUC [13]; and body weight at 12–15 weeks. The analysis was performed using 3 days of proCGM data (3–5 days after wearing the proCGM).

2.4. Statistical analyses

All statistical analyses were performed using EZR software version 1.54 (Saitama Medical Center, Jichi Medical University, Saitama, Japan). The Shapiro–Wilk test was used to determine the normality of data distribution. The primary and secondary endpoints were evaluated using the Wilcoxon signed rank test (one-sample Wilcoxon test) or paired t-tests. Data are presented as the mean ± SD. Statistical significance was set to p < 0.05. This was an exploratory pilot study and, therefore, a formal power analysis was not conducted.

3. Results

3.1. Patient characteristics

The patient characteristics are shown in Table 1. Among the 12 patients, 1 patient dropped out because of hospitalization for a urinary tract infection, leaving 11 patients (5 men and 6 women) in the final analysis (Supplementary Fig. 1). Regarding adverse events, the professional continuous glucose monitoring (proCGM) data revealed no incidence of hypoglycemia (<70 mg/dL) or severe hypoglycemia (glucose levels <54 mg/dL). Additionally, no allergic reactions to URLi were observed during the study period.

Table 1.

Patient characteristics (n = 11).

	Mean ± SD
Age (years)	68.3 ± 10.2
Sex (male/female)	male 5, female 6
HbA1c (%)	7.8 ± 0.9
Height (cm)	1.6 ± 0.1
Body weight (kg)	69.0 ± 12.8
BMI (kg/m2)	27.7 ± 4.9
Duration of diabetes (years)	16.2 ± 6.7
Duration of insulin therapy (years)	12.5 ± 7.4
Frequency of injection per day (times)	Number (%)
2	3 (27)
≧3	8 (73)
Antihyperglycemic drugs
Metformin	5 (45)
DPP-4 inhibitor	3 (27)
GLP-1 receptor agonist	3 (27)
SGLT-2 inhibitor	6 (55)
Sulfonylurea	0 (0)
Glinide	0 (0)
α-Glucosidase inhibitor	0 (0)
Thiazolidinediones	0 (0)

Data are presented as mean ± standard deviation. HbA1c: glycated hemoglobin; BMI: body mass index; DPP-4: dipeptidyl peptidase-4; GLP-1: glucagon-like peptide-1; SGLT-2: Sodium-glucose transporter 2.

3.2. ITR-QOL questionnaire

At the baseline (0 weeks), the average total ITR-QOL questionnaire score of the 11 patients was 94.7 ± 15.4 points. From Q1 to Q22, the items with an average score <4 included “Q3. The length of time spent away from the home is restricted because of insulin injections” (3.8 ± 1.1 points), “Q9. I am aware of the symptoms of low blood glucose levels” (3.8 ± 1.3 points), “Q15. I am always worried about and feel burdened by the timing of insulin injections” (3.9 ± 1.1 points), “Q18. I am worried that my blood glucose levels may become too low after insulin injection” (3.7 ± 1.3 points), and “Q21. It is difficult to inject insulin while eating out” (3.8 ± 1.3 points). Table 2 shows the findings of the primary endpoints at 12–15 weeks. At 12–15 weeks, participants taking URLi reported a significant increase in their total ITR-QOL questionnaire score (+15.1 ± 16.1 points, p < 0.01), with 13 of 23 items showing statistically significant improvements. The perception domain score increased significantly (Q13– Q23) (+7.2 ± 6.9 points, p < 0.01). Specific items that showed a significant score increase included Q13, Q15, Q16, Q18, Q19, and Q21. The status domain score (Q1–Q12) also showed a significant increase (+7.9 ± 9.5 points, p < 0.05). Specific items that showed a significant increase included Q3, Q4, Q5, Q6, Q7, Q8, and Q12. Taken together, patients’ QOL was significantly improved by switching from conventional RAA to URLi. Notably, HbA1c values showed no significant changes (−0.1 ± 0.4 %, p = 0.483).

Table 2.

Results of the ITR-QOL questionnaire.

	0 weeks	12–15 weeks	Amount of change	One-sample Wilcoxson test, p-value
Q1	4.5 ± 0.9	4.8 ± 0.4	0.4 ± 1.1	0.177
Q2	4.2 ± 0.8	4.7 ± 0.6	0.5 ± 1.1	0.081
Q3	3.8 ± 1.1	5.0 ± 0.0	1.2 ± 1.1	<0.01∗
Q4	4.0 ± 1.3	4.8 ± 0.6	0.8 ± 1.3	<0.05∗
Q5	4.3 ± 0.7	4.8 ± 0.4	0.5 ± 0.8	<0.05∗
Q6	4.0 ± 1.3	4.9 ± 0.3	0.9 ± 1.4	<0.05∗
Q7	4.1 ± 1.2	4.8 ± 0.4	0.7 ± 1.3	<0.05∗
Q8	4.0 ± 1.0	4.9 ± 0.3	0.9 ± 1.0	<0.05∗
Q9	3.8 ± 1.3	4.5 ± 0.8	0.7 ± 1.4	0.053
Q10	4.5 ± 0.9	4.8 ± 0.4	0.3 ± 1.0	0.213
Q11	4.5 ± 0.9	4.9 ± 0.3	0.5 ± 0.9	0.055
Q12	4.3 ± 1.0	4.7 ± 0.6	0.5 ± 0.7	<0.05∗
Q13	4.1 ± 1.0	4.9 ± 0.3	0.8 ± 0.9	<0.05∗
Q14	4.3 ± 1.1	4.8 ± 0.4	0.5 ± 1.2	0.055
Q15	3.9 ± 1.1	4.8 ± 0.4	0.9 ± 1.0	<0.05∗
Q16	4.1 ± 0.9	5.0 ± 0.0	0.9 ± 0.9	<0.01∗
Q17	4.8 ± 0.4	5.0 ± 0.0	0.2 ± 0.4	0.090
Q18	3.7 ± 1.3	4.8 ± 0.4	1.1 ± 1.2	<0.05∗
Q19	4.3 ± 1.0	4.9 ± 0.3	0.6 ± 1.1	<0.05∗
Q20	4.5 ± 0.9	5.0 ± 0.0	0.5 ± 0.9	0.055
Q21	3.8 ± 1.3	4.9 ± 0.3	1.1 ± 1.3	<0.05∗
Q22	4.8 ± 0.4	4.9 ± 0.3	0.1 ± 0.5	0.297
Q23	2.5 ± 0.5	2.1 ± 0.9	−0.5 ± 1.1	0.087
Total sum	94.7 ± 15.4	109.8 ± 3.6	15.1 ± 16.1	<0.01∗
Status domain sum	49.9 ± 9.1	57.8 ± 2.6	7.9 ± 9.5	<0.05∗
Perception domain sum	44.8 ± 6.8	52.0 ± 1.5	7.2 ± 6.9	<0.01∗

Data are presented as mean ± standard deviation. ∗Indicates statistical significance. Wilcoxon signed rank test (versus 0 weeks), ∗p < 0.05. “Amount of change” is the change between 0 and 12–15 weeks. Status domain includes Q1–12; perception domain includes Q13–22. For Q1–22, each score selected by the patient was added, whereas for Q23, the reverse score was used as the score; the sum of these two scores was used as the total ITR-QOL questionnaire score. ITR-QOL: insulin therapy-related quality-of-life measure; sum: summation.

3.3. proCGM data and body weight

The results of the secondary endpoints are presented in Table 3. At 12–15 weeks, the time in range (TIR) increased significantly (+8.3 ± 9.2 %, p < 0.05), and the time above range (TAR) decreased significantly (−7.7 ± 10.2 %, p < 0.05) (Fig. 1). Moreover, the 24-h standard deviation (SD; −9.3 ± 12.6 mg/dL, p < 0.001), 24-h coefficient of variation (CV; −3.7 ± 7.0 %, p < 0.01), 24-h M-value (−8.3 ± 8.4 mg/dL, p < 0.05), mean amplitude of glycemic excursion (MAGE; −31.5 ± 19.2 mg/dL, p < 0.001), mean of daily difference (MODD; −6.7 ± 9.2 mg/dL, p < 0.05), and 24-h area under the curve (AUC; −287.6 ± 624.0 mg/dL h, p < 0.05) all showed significant decreases, but the time below range (TBR) and percentage of time with nocturnal hypoglycemia showed no significant change. Thus, switching from RAA to URLi improved PPG management without increasing the risk of hypoglycemia, leading to an increase in the TIR. Notably, body weight showed no significant change.

Table 3.

Results of secondary endpoints at 12–15 weeks.

	0 weeks	12–15 weeks	Amount of change	p-value
HbA1c (%)	7.8 ± 0.9	7.7 ± 1.0	−0.1 ± 0.4	0.483
Body weight (kg)	69.0 ± 12.8	69.1 ± 12.8	0.1 ± 0.4	0.496
Mean glucose level (mg/dL)	184.1 ± 52.8	172.1 ± 47.4	−12.0 ± 19.9	0.086
TAR (%)	46.1 ± 29.0	38.4 ± 28.8	−7.7 ± 10.2	<0.05∗
TIR (%)	51.3 ± 26.8	59.6 ± 27.6	8.3 ± 9.2	<0.05∗
TBR (%)	0.0 [0.0–0.5]	0.0 [0.0–1.7]	0.0 [0.0–0.5]	1.000
Nocturnal TBR (00:00–00:59 h)	0.0 [0.0–0.0]	0.0 [0.0–0.7]	0.0 [0.0–0.7]	0.675
24-h SD of glycemic variability (mg/dL)	52.5 ± 15.5	43.2 ± 15.5	−9.3 ± 12.6	<0.001∗
24-h CV of glycemic variability (%)	29.5 ± 8.0	25.8 ± 8.2	−3.7 ± 7.0	<0.01∗
24-h M-value (target glucose level: 100 mg/dL)	34.5 ± 31.9	26.2 ± 27.4	−8.3 ± 8.4	<0.05∗
MAGE (mg/dL)	114.1 ± 26.8	82.6 ± 26.2	−31.5 ± 19.2	<0.001∗
MODD in glucose level (mg/dL)	48.5 ± 17.5	41.9 ± 16.2	−6.7 ± 9.2	<0.05∗
24-h AUC (mg/dL h)	4404.7 ± 1306.8	4117.1 ± 1204.0	−287.6 ± 624.0	<0.05∗

Data are presented as mean ± standard deviation or median (interquartile range). ∗Indicates statistical significance. Paired t-test (versus 0 weeks), ∗p < 0.05, Wilcoxon signed rank test (versus 0 weeks), ∗p < 0.05. “Amount of change” is the change between 0 and 12–15 weeks. TIR: percentage of time in the glucose range of 70–180 mg/dL; TAR: percentage of time in the glucose range >180 mg/dL; TBR: percentage of time in the glucose range <70 mg/dL; HbA1c: glycated hemoglobin; BMI: body mass index; SD: standard deviation; CV: coefficient of variation; MAGE: mean amplitude of glycemic excursion; MODD: mean of daily difference; AUC: area under the curve.

Fig. 1.

Changes in proCGM parameters between 0 and 12–15 weeks TAR: time above target glucose range; TIR: time in target glucose range; TBR: time below target glucose range; proCGM: professional continuous glucose monitoring.

4. Discussion

The total ITR-QOL questionnaire score, perception domain score, and status domain score at 12–15 weeks were significantly improved after switching to URLi. Among items with an average score ≤4 at baseline (Q3, Q9, Q15, Q18, and Q21), Q3, Q15, Q18, and Q21 showed significant improvement at 12–15 weeks. To our knowledge, the original publication that introduced the ITR-QOL questionnaire does not define a minimally important difference for either the total score or individual item scores. However, in subsequent studies using the same measure and similar statistical analyses, even a change of approximately 0.5 points has been interpreted as clinically meaningful [14]. The main reason for administering insulin during or after meals in patients with T2DM is concerns regarding hypoglycemia before meals [15,16]. Thus, the timing of URLi injection may reduce the fear of pre-meal hypoglycemia. Additionally, the timing of URLi may explain the significant improvements in scores related to the timing of insulin injections. The significant improvement in Q3, Q18, and Q21 could be attributed to reduced anxiety about experiencing hypoglycemia while waiting for a meal that may not arrive soon. Given the improvements in Q15 and Q18, as well as the increase in TIR without an increase in TBR, we suggest that switching from conventional RAA to URLi and careful optimization of insulin units can alleviate concerns about hypoglycemic symptoms.

For comparison, we have added supporting data in Supplementary Table 2 consisting of ITR-QOL questionnaire scores from patients who were deemed more suitable for continued RAA therapy rather than switching to URLi, based on the Insulin Selection Checklist. For patients who remained on RAA therapy, we assumed that their QOL would remain theoretically unchanged and, accordingly, did not collect follow-up data at 12–15 weeks. At baseline (0 weeks), the total ITR-QOL questionnaire score in the patients suitable with RAA was 100.8 ± 9.7 points, and items with an average score ≤4 were “Q9. I am aware of the symptoms of low blood glucose " (3.4 ± 0.9 points) and “Q21. It is difficult to inject insulin when I eat out” (3.7 ± 1.6 points).

HbA1c levels were not significantly altered (−0.1 ± 0.4 %, p = 0.483) at 12–15 weeks. This may be attributed to the relatively short observation period (12–15 weeks). Furthermore, as seven participants were aged ≥65 years, with four having already reached their target HbA1c levels, many of the participants did not require any strict changes in insulin units during the study period, which may have contributed to this result.

Switching to URLi did not increase the risk of hypoglycemia but effectively suppressed PPG, which significantly increased the TIR and decreased the TAR. The PRONTO-Time in Range study [17] reported a significant increase in both daytime and 24-h TIR for URLi injection compared to RAA injection. Similarly, our findings revealed a significant improvement in the TIR for URLi. Furthermore, switching to URLi effectively managed PPG levels, resulting in a significant decrease in the TAR, 24-h M-value, MAGE, and 24-h AUC. PPG levels have a significant positive correlation with the progression of diabetic microangiopathy [18], incidence of major cardiovascular events, and risk of all-cause mortality and cardiovascular death [19,20]. These risks can be reduced by improved PPG management through the use of URLi.

One of the advantages of URLi is its rapid onset and offset of insulin activity. Compared to lispro, URLi has a faster onset of approximately 5 min, approximately eight times greater exposure within the first 15 min, and a shorter exposure time by 68 min [21]. Given this rapid onset and short exposure time, URLi reduces the risk of late postprandial hypoglycemic episodes occurring ≥3 h after a meal [5,22]. However, URLi may increase the incidence of hypoglycemia within 2 h of a meal [23]. Therefore, when switching from conventional RAA to URLi, although the dosage of insulin units can typically remain the same, careful dose adjustment and follow-up should be performed for each individual. In our study, we observed a non-significant increase in the total amount of insulin units (basal insulin +1.2 ± 2.7 units [p = 0.19], bolus insulin +0.7 ± 3.9 [p = 0.57]). Previous reports on changes in insulin units are conflicting, with one indicating a significant increase in basal, bolus, and total insulin doses at week 12 [17] and another suggesting no change in the average daily bolus insulin dose [24].

The timing of URLi enables more flexible treatment and addresses the unmet needs of patients related to the strict requirements for pre-meal insulin injections. This was reflected in the significant improvement in ITR-QOL questionnaire scores. A choice of several protocols for insulin administration is beneficial; however, caution is required when URLi is administered after meals. Although post-meal URLi administration was not inferior to pre-meal conventional RAA administration in terms of HbA1c levels, it is less effective at lowering blood glucose levels than pre-meal URLi administration [5,23].

URLi may not adequately suppress PPG levels in individuals who take a long time to eat, those who eat fatty meals, and those with slow gastric emptying. This is because the rapid onset and short duration of URLi action is unsuitable for glucose fluctuations. Therefore, caution should be exercised when prescribing URLi to these patients. Screening using our Insulin Selection Checklist can help exclude such individuals to some extent.

This pilot study has several limitations. First, the sample size was small (11 patients) and the study period was short (12–15 weeks). Therefore, a large-scale, long-term study is warranted to investigate the durability of these changes. Second, there was no control group consisting of patients who continued using conventional RAA. Including a parallel comparison group using RAA would have strengthened the validity of the study. We have added supporting data in Supplementary Table 2, which includes ITR-QOL questionnaire scores from patients who were deemed more suitable for continued RAA therapy rather than switching to URLi, based on the Insulin Selection Checklist. However, follow-up data at 12–15 weeks were not collected because we assumed their QOL would remain theoretically unchanged. Third, although the ITR-QOL questionnaire has not been specifically validated for use with URLi, it was originally developed and validated to assess treatment-related QOL in Japanese patients with diabetes receiving insulin therapy. The questionnaire is designed to be applicable across different insulin formulations, as it focuses on general aspects of insulin treatment, such as injection burden and impact on daily life. Therefore, its use in the context of URLi appears appropriate and may provide meaningful insights. Fourth, the changes in HbA1c did not reach statistical significance. This may be attributed to the relatively short observation period; moreover, seven participants were aged ≥65 years, and four had already reached their target HbA1c levels, meaning many participants did not require strict changes in insulin doses during the study period. Fifth, reliance on self-reported QOL measures introduces potential biases, particularly expectation bias. Sixth, since the study participants consisted solely of Japanese patients from a single center, the external validity and generalizability of the findings to other populations are limited. Seventh, no formal sample size calculation or power analysis was performed due to the exploratory nature of the study. One participant dropped out for reasons unrelated to the intervention, resulting in a final sample size of 11. Given the small sample size, even a single withdrawal could have impacted the outcomes to some extent. Eighth, some individuals taking conventional RAA refused to switch to URLi despite having their suitability confirmed by the Insulin Selection Checklist; this was because these patients had a long history of insulin use and were accustomed to injecting RAA 2–15 min before a meal. Ninth, no adjustments were made regarding the units of insulin degludec/insulin aspart or insulin degludec/liraglutide. Adjustment of these units may result in different changes in the units of URLi due to fluctuations in the amount of the GLP-1 component. Tenth, this study included patients at different stages with respect to blood glucose management, insulin administration, and frequency of insulin administration. Nevertheless, this study included real-world data that reflects actual clinical practice.

5. Conclusion

Switching from conventional RAA to URLi significantly improved ITR-QOL questionnaire scores, which indicated an increase in the patient QOL. Further, switching to URLi significantly increased TIR and decreased TAR with no increase in the risk of hypoglycemia, thereby improving glycemic management. In summary, URLi is an effective treatment option that offers flexible timing for insulin administration, improved glycemic management, and enhanced patient satisfaction.

CRediT authorship contribution statement

Yuriko Hajika: Writing – original draft, Visualization, Formal analysis, Data curation. Yuji Kawaguchi: Writing – review & editing.

Data availability

Data supporting the findings can be made available upon reasonable request, excluding confidential information such as patient data.

Funding

This research received no funding.

Competing interests statement

The authors declare no competing interests.

Appendix A. Supplementary data

The following is/are the supplementary data to this article: