Percentage of time with glucose within glycemic range (4.4-7.8 mmol/L) was assessed by the total time within glycemic range divided by the length of the assessment interval.

Continuous glucose monitoring (CGM) population: All participants who received at least 1 dose, had evaluable post-baseline CGM data, irrespective of compliance. Number of participants analyzed = participants with baseline, Weeks 7-8 (Period A) and/or Weeks 15-16 (Period B) CGM assessment. Missing data imputed using last observation carried forward.

Percentage of time with glucose above the upper limit of glycemic range (>7.8 mmol/L) was assessed by the total time above the upper limit of glycemic range divided by the length of the assessment interval.

CGM population. Number of participants analyzed = participants with baseline, Weeks 7, 8 (Period A), and/or Weeks 15, 16 (Period B) CGM assessment. Missing data imputed using last observation carried forward (LOCF).

Percentage of time with glucose below the lower limit of glycemic range (<4.4 mmol/L) was assessed by the total time below the lower limit of glycemic range divided by the length of the assessment interval.

CGM population. Number of participants analyzed = participants with baseline, Weeks 7, 8 (Period A), and/or Weeks 15, 16 (Period B) CGM assessment. Missing data imputed using LOCF.

The diurnal glucose exposure is measured as the average diurnal glucose concentration, diurnal glucose variability is measured by interquartile range (IQR), that is, average distance between the 25th and the 75th point-wise percentiles and diurnal glucose stability is assessed in terms of the mean absolute rate of change (mmol/l), that is, the area under the absolute rate of change of the median curve (based on the median point values between two adjacent hourly basket intervals), divided by the length of the assessment interval.

CGM population. Number of participants analyzed = participants with baseline, Weeks 7, 8 (Period A), and/or Weeks 15, 16 (Period B) CGM assessment. Missing data imputed using LOCF.

Percentage of time with glucose within glycemic range (4.4-7.8 mmol/L) was assessed by the total time within glycemic range divided by the length of the assessment interval.

CGM population. Number of participants analyzed = participants with baseline, Weeks 7, 8 (Period A), and/or Weeks 15, 16 (Period B) CGM assessment, and n = participants with assessment at specified time-point.

Modified Intent-to-Treat population: randomized participants who received at least 1 dose; had baseline, at least 1 post-baseline efficacy assessment; irrespective of compliance. Number of participants analyzed = participants with baseline, Week 8 and/or 16 HbA1c assessment, n = participants with HbA1c assessment at specified time. LOCF applied.

Modified Intent-to-Treat population. Number of participants analyzed = participants with baseline, Week 8 and/or 16 FPG assessment, n = participants with FPG assessment at specified time. Missing data imputed using LOCF.

Change in average of 7-point SMPG. 7-point SMPG was assessed starting with a measurement at before breakfast and 2 hours after breakfast; before and 2 hours after lunch; before and 2 hours after dinner; at bedtime.

Modified Intent-to-Treat population. Number of participants analyzed = participants with baseline, Week 8 and/or 16 7-point SMPG assessment, n = participants with 7-point SMPG assessment at specified time. Missing data imputed using LOCF.

Modified Intent-to-Treat population. Here n = participants with basal insulin dose assessment at specified time-point. Missing data imputed using LOCF.

Hypoglycemia events were Severe hypoglycemia (an event that required assistance of another person to actively administer carbohydrate, glucagon, or other resuscitative actions); Documented symptomatic hypoglycemia (typical symptoms of hypoglycemia with plasma glucose level of <=3.9 mmol/L [70 mg/dL]); Asymptomatic hypoglycemia (no typical symptoms of hypoglycemia but plasma glucose level <=3.9 mmol/L); Probable symptomatic hypoglycemia (an event during which symptoms of hypoglycemia were not accompanied by a plasma glucose determination, but was presumably caused by a plasma glucose level <=3.9 mmol/L, symptoms treated with oral carbohydrate without a test of plasma glucose); Relative hypoglycemia (an event during which the person with diabetes reported any of the typical symptoms of hypoglycemia, and interpreted the symptoms as indicative of hypoglycemia, but plasma glucose level >3.9 mmol/L); Severe and/or confirmed a hypoglycemia (plasma glucose <=3.9 mmol/L).

Safety population: all randomized participants who were exposed to at least one dose, regardless of amount of treatment administered. In the event of participants having received treatments different from those assigned according to the randomization schedule, safety analyses were conducted according to treatment received.