This SoliRam sub-analysis included all 420 eligible participants involved in the overall study. Of the 420 eligible participants, 174 participants were using SGLT-2i in addition to iGlarLixi and 246 participants were not using SGLT-2i, henceforth, referred to as SGLT-2i user and non-user, respectively.

Baseline Demographics and Participant Characteristics

More than 50% of participants were males in both groups (Table 1). The mean age of participants in SGLT-2i-user group (55 years) was lower than that in SGLT-2i-non-user group (58 years) and more participants in SGLT-2i-non-user group were aged ≥ 65 years (Table 1). The mean weight, duration of diabetes, HbA1c, and FPG were similar in both groups. Any complications and/or comorbidities were higher in SGLT-2i-non-user group, whereas the use of OADs was higher in SGLT-2i-user group (Table 1).

Table 1 Baseline demographics and clinical characteristicsFasting During Ramadan

Around 97% of the participants in both groups were able to fast for ≥ 25 days. Most participants in SGLT-2i-user (93.6%) and SGLT-2i-non-user (66.7%) groups considered their personal decision as motivation for fasting. Other reasons for fasting included improving health (SGLT-2i user, 2.3% and SGLT-2i-non-user, 1.3%), coping with family and community (SGLT-2i user, 0.6% and SGLT-2i-non-user, 2.5%), enhancing self-discipline (SGLT-2i user, 1.7% and SGLT-2i-non-user, 3.4%) and other (SGLT-2i user, 1.7% and SGLT-2i-non-user, 26.2%). Meanwhile, 16 (9.3%) and 15 (6.3%) participants in SGLT-2i-user and SGLT-2i-non-user groups, respectively, broke their fast. The main reasons for breaking the fast were traveling (SGLT-2i user, n = 4 and SGLT-2i-non-user, n = 6), pre-existing conditions (SGLT-2i user, n = 1 and SGLT-2i-non-user, n = 6), AEs (SGLT-2i user, n = 4 and SGLT-2i-non-user, n = 0), hypoglycemia (SGLT-2i user, n = 3 and SGLT-2i-non-user, n = 1) or menses (SGLT-2i user, n = 4 and SGLT-2i-non-user, n = 2). More than 60% of the participants received diabetes education specific to Ramadan either by educational programs or face-to-face sessions or website/self-education (Supplementary Table S2).

Changes in Anti-hyperglycemic TherapyiGlarLixi Treatment

At the study initiation visit, 67.2% and 73.6% participants in SGLT-2i-user and SGLT-2i-non-user groups, respectively, used an iGlarLixi 10–40 pen (insulin glargine 100 units/ml + lixisenatide 50 μg/ml) and 32.8% and 26.4% participants in SGLT-2i-user and SGLT-2i-non-user groups, respectively, used an iGlarLixi 30–60 pen (insulin glargine 100 units/ml + lixisenatide 33 μg/ml).

During pre-Ramadan period, 45.1%, 32.9%, and 22.0% of participants in SGLT-2i-user group and 39.2%, 38.8%, and 22.0% of participants in SGLT-2i-non-user group administered iGlarLixi at breakfast, lunch, and dinner, respectively. During Ramadan, as aligned with the recommendations, most participants in SGLT-2i-user (93.0%) and SGLT-2i-non-user (85.7%) groups injected iGlarLixi in the evening at Iftar, remaining 7% participants in SGLT-2i-user group and 13.9% in SGLT-2i-non-user group injected iGlarLixi in the morning (prior to Suhur).

In SGLT-2i-user group, the mean ± standard deviation (SD) iGlarLixi daily dose was 27.8 ± 11.6, 26.7 ± 10.8 and 27.8 ± 11.9 dose steps during pre-Ramadan, Ramadan, and post-Ramadan periods, respectively; iGlarLixi daily dose increased by 0.1 ± 3.4 dose steps from pre- to post-Ramadan. In SGLT-2i-non-user group, the mean ± SD iGlarLixi daily dose was 24.3 ± 10.3, 23.5 ± 9.0, and 24.4 ± 9.9 dose steps during pre-Ramadan, Ramadan and post-Ramadan periods, respectively; iGlarLixi daily dose increased by 0.4 ± 3.3 dose steps from pre- to post-Ramadan.

The iGlarLixi interruption was reported for 3 days in SGLT-2i-user and 5 days in SGLT-2i-non-user groups; reasons for iGlarLixi interruption were hypoglycemia (SGLT-2i user, n = 1 and SGLT-2i-non-user, n = 0), pre-existing conditions (SGLT-2i user, n = 0 and SGLT-2i-non-user, n = 2), AEs (SGLT-2i user, n = 1 and SGLT-2i-non-user, n = 0) and other (SGLT-2i user, n = 1 and SGLT-2i-non-user, n = 3).

Non-insulin Anti-hyperglycemic Treatments

All participants in SGLT-2i-user (n = 174) and 75.6% (n = 186) of participants in SGLT-2i-non-user groups were taking OADs in addition to iGlarLixi. In SGLT-2i-user group, 80 (46.0%) and 34 (19.5%) participants were on 2 and ≥ 3 OADs, respectively, whereas in the SGLT-2i-non-user group, 63 (25.6%) and 2 (0.8%) participants were taking 2 and ≥ 3 OADs, respectively (Table 2). The most commonly used OADs in the SGLT-2i-user group were sulphonylureas (31.6%), biguanides (86.2%) and SGLT-2i (100%). In the SGLT-2i-non-user group, the most frequently used OADs were biguanides (34.6%) and sulphonylureas (54.5%). There was no substantial change in the proportion of participants taking OADs from pre-Ramadan to Ramadan in both groups (Supplementary Table S3).

Table 2 Number of concomitant OADs taken during the overall study period, pre-Ramadan, Ramadan, and post-Ramadan periodHypoglycemiaPrimary Endpoint

The incidence of severe and/or symptomatic documented (< 70 mg/dl [< 3.9 mmol/l]) hypoglycemia events was low throughout the study for both groups. During the last month pre-Ramadan, Ramadan and first month post-Ramadan, 0.6%, 4.2%, and 0.6% participants in SGLT-2i-user and 1.3%, 0.9%, and 0% participants in SGLT-2i-non-user groups, respectively, experienced severe and/or symptomatic documented (< 70 mg/dl [< 3.9 mmol/l]) hypoglycemia (Fig. 1A).

Fig. 1

Proportion of participants with severe and/or symptomatic documented hypoglycemia: A < 70 mg/dl (< 3.9 mmol/l) and B < 54 mg/dl (< 3.0 mmol/l). *Last month of pre-Ramadan. †First month of post-Ramadan. The analysis included eligible population, defined as all participants included in the study who met the inclusion criteria and did not meet exclusion criteria. CI confidence interval, n number of participants reporting hypoglycemia event, N total number of participants, NA not applicable, SGLT-2i sodium-glucose co-transporter-2 inhibitor

Secondary Hypoglycemia Endpoints

The incidence of severe and/or symptomatic documented (< 54 mg/dl [< 3.0 mmol/l]) hypoglycemia events was also extremely low throughout the study for both groups, including Ramadan (Fig. 1B). Events of severe and/or symptomatic documented hypoglycemia (< 70 mg/dl [< 3.9 mmol/l] or < 54 mg/dl [< 3.0 mmol/l]) during Ramadan were reported more frequently during fasting hours (between Suhur and Iftar) than during non-fasting hours (Supplementary Fig. 1A and 1B). Only one severe hypoglycemia event was reported during the whole study period, which occurred in SGLT-2i-user group during pre-Ramadan period.

Change in Glycemic Endpoints (HbA1c, FPG, and SMPG)

A mean ± SD reduction in HbA1c was observed from pre- to post-Ramadan in both SGLT-2i-user (0.69 ± 1.20%) and SGLT-2i-non-user groups (0.78 ± 0.92%) (Fig. 2A). The proportion of participants with HbA1c < 7% increased from pre- to post-Ramadan in both groups (SGLT-2i user, 16.3% to 39.3% and SGLT-2i-non-user, 6.7% to 35.1%). A mean ± SD reduction of 16.0 ± 40.4 mg/dl and 19.8 ± 26.6 mg/dl was observed in FPG from pre- to post-Ramadan in SGLT-2i-user and SGLT-2i-non-user groups, respectively (Fig. 2B). Fasting SMPG also reduced from pre- to post-Ramadan in both groups; 5.4 ± 26.9 mg/dl in SGLT-2i-user and 13.5 ± 25.0 mg/dl in SGLT-2i-non-user groups (Fig. 2C).

Fig. 2

Change in A HbA1c (%), B FPG (mg/dl), C SMPG (mg/dl), and D weight (kg) from the pre-Ramadan to the post-Ramadan period. *SGLT-2i user, n = 171 and SGLT-2i-non-user, n = 245. †SGLT-2i user, n = 141 and SGLT-2i-non-user, n = 215. The analysis included the evaluable population, defined as all participants who had glycemic endpoint for both pre-Ramadan and post-Ramadan periods (HbA1c, FPG, and fasting SMPG). CI confidence interval, FPG fasting plasma glucose, HbA1c glycated hemoglobin, SMPG self-monitoring of plasma glucose, SD standard deviation, SGLT-2i sodium-glucose co-transporter-2 inhibitor

Change in Weight

A mean ± SD reduction of 0.52 ± 2.50 kg and 1.31 ± 2.04 kg was observed in weight from pre- to post-Ramadan period in SGLT-2i user and SGLT-2i-non-user groups, respectively (Fig. 2D).

Adverse Events

A total of 14 (7.9%) participants in SGLT-2i-user and six (2.4%) participants in SGLT-2i-non-user groups experienced at least one AE throughout the study period. Of these, five (2.9%) participants in SGLT-2i-user and one (0.4%) participant in SGLT-2i-non-user groups experienced at least one AE during Ramadan (Table 3). There were no AEs leading to permanent treatment discontinuation in SGLT-2i-user group, whereas one (0.4%) participant in the SGLT-2i-non-user group had such an event during the whole study period. No serious AEs, treatment-related AEs, and AEs leading to death were reported in either of the treatment groups throughout the study period.

Table 3 Incidence of AEs