Trial Outcomes & Findings for Comparing Efficacy and Safety of Insulin Degludec/Insulin Aspart and BIAsp 30 in Subjects With Type 2 Diabetes (NCT NCT02762578)

NCT ID: NCT02762578

Last Updated: 2019-04-02

Results Overview

Change from baseline in HbA1c after 26 weeks of treatment. The response and change from baseline in response after 26 weeks are analysed using an analysis of covariance model with treatment, anti-diabetic therapy at screening and sex as fixed factors, age and baseline response as covariate. Missing values imputed using last observed value.

• Recruitment status: COMPLETED
• Study phase: PHASE3
• Target enrollment: 543 participants
• Primary outcome timeframe: At 26 weeks
• Results posted on: 2019-04-02

Participant Flow

A total of 41 sites were approved for recruiting subjects, of which 40 sites screened and randomized the subjects.

Insulin treated subjects on current treatment: basal insulin, premixed insulin or a self-mixed insulin regimen, all administered once daily or twice daily (BID) with or without metformin. The treatment regimen should have remained unchanged for at least 8 weeks prior to randomisation.

Participant milestones

Measure	IDegAsp BID Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Overall Study STARTED	361	182
Overall Study Exposed	360	181
Overall Study COMPLETED	346	172
Overall Study NOT COMPLETED	15	10

Reasons for withdrawal

Measure	IDegAsp BID Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Overall Study Adverse Event	1	6
Overall Study Protocol Violation	4	2
Overall Study Withdrawal by Subject	4	2
Overall Study Unclassified	6	0

Baseline Characteristics

Number of participants analysed=participants with available data for FPG at baseline

Baseline characteristics by cohort

Measure	IDegAsp BID n=361 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=182 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	Total n=543 Participants Total of all reporting groups
Age, Continuous	59.64 years STANDARD_DEVIATION 9.04 • n=361 Participants	58.78 years STANDARD_DEVIATION 9.41 • n=182 Participants	59.35 years STANDARD_DEVIATION 9.16 • n=543 Participants
Sex: Female, Male Female	163 Participants n=361 Participants	83 Participants n=182 Participants	246 Participants n=543 Participants
Sex: Female, Male Male	198 Participants n=361 Participants	99 Participants n=182 Participants	297 Participants n=543 Participants
Ethnicity (NIH/OMB) Hispanic or Latino	0 Participants n=361 Participants	0 Participants n=182 Participants	0 Participants n=543 Participants
Ethnicity (NIH/OMB) Not Hispanic or Latino	361 Participants n=361 Participants	182 Participants n=182 Participants	543 Participants n=543 Participants
Ethnicity (NIH/OMB) Unknown or Not Reported	0 Participants n=361 Participants	0 Participants n=182 Participants	0 Participants n=543 Participants
Race (NIH/OMB) American Indian or Alaska Native	0 Participants n=361 Participants	0 Participants n=182 Participants	0 Participants n=543 Participants
Race (NIH/OMB) Asian	361 Participants n=361 Participants	182 Participants n=182 Participants	543 Participants n=543 Participants
Race (NIH/OMB) Native Hawaiian or Other Pacific Islander	0 Participants n=361 Participants	0 Participants n=182 Participants	0 Participants n=543 Participants
Race (NIH/OMB) Black or African American	0 Participants n=361 Participants	0 Participants n=182 Participants	0 Participants n=543 Participants
Race (NIH/OMB) White	0 Participants n=361 Participants	0 Participants n=182 Participants	0 Participants n=543 Participants
Race (NIH/OMB) More than one race	0 Participants n=361 Participants	0 Participants n=182 Participants	0 Participants n=543 Participants
Race (NIH/OMB) Unknown or Not Reported	0 Participants n=361 Participants	0 Participants n=182 Participants	0 Participants n=543 Participants
Glycosylated haemoglobin (HbA1c)	8.32 percentage of HbA1c STANDARD_DEVIATION 0.76 • n=361 Participants	8.33 percentage of HbA1c STANDARD_DEVIATION 0.77 • n=182 Participants	8.32 percentage of HbA1c STANDARD_DEVIATION 0.76 • n=543 Participants
Fasting plasma glucose (FPG)	9.07 mmol/L STANDARD_DEVIATION 2.21 • n=355 Participants • Number of participants analysed=participants with available data for FPG at baseline	9.07 mmol/L STANDARD_DEVIATION 2.49 • n=180 Participants • Number of participants analysed=participants with available data for FPG at baseline	9.07 mmol/L STANDARD_DEVIATION 2.31 • n=535 Participants • Number of participants analysed=participants with available data for FPG at baseline
Body weight	68.50 kg STANDARD_DEVIATION 11.63 • n=361 Participants	69.41 kg STANDARD_DEVIATION 12.38 • n=182 Participants	68.80 kg STANDARD_DEVIATION 11.88 • n=543 Participants

PRIMARY outcome

Timeframe: At 26 weeks

Population: Full analysis set. Number of subject analysed=subjects with available data for HbA1c

Change from baseline in HbA1c after 26 weeks of treatment. The response and change from baseline in response after 26 weeks are analysed using an analysis of covariance model with treatment, anti-diabetic therapy at screening and sex as fixed factors, age and baseline response as covariate. Missing values imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=359 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=182 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Change From Baseline in HbA1c (%) (Glycosylated Haemoglobin)	-1.48 percentage of HbA1c Standard Error 0.05	-1.40 percentage of HbA1c Standard Error 0.06

SECONDARY outcome

Timeframe: At 26 weeks

Population: Full analysis set. Number of subject analysed=subjects with available data for FPG.

Change from baseline in FPG at week 26. The response and change from baseline in response after 26 weeks are analysed using an analysis of covariance model with treatment, anti-diabetic therapy at screening and sex as fixed factors, age and baseline response as covariate. Missing values imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=354 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=180 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Change From Baseline in FPG (Fasting Plasma Glucose)	-2.99 mmol/L Standard Error 0.09	-1.57 mmol/L Standard Error 0.13

SECONDARY outcome

Timeframe: Weeks 0-26

Population: Full analysis set. Number of subjects analysed=subjects with data available for nocturnal confirmed hypoglycaemic episodes.

The number of events was analysed using a negative binomial model with a log-link function and the logarithm of the exposure time (100 years) for which a hypoglycaemic episode is considered treatment emergent as offset. The model included treatment, anti-diabetic therapy at screening and sex as fixed factors, and age as covariate. Confirmed hypoglycaemia is defined as either severe episodes or episodes with plasma glucose < 3.1 mmol/L (56 mg/dL) with or without symptoms. The nocturnal period was defined as the period between 00:01 and 05:59 a.m. (both inclusive). A treatment emergent hypoglycaemic episode was defined as an episode that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment.

Outcome measures

Measure	IDegAsp BID n=359 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=181 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Number of Treatment Emergent Nocturnal Confirmed Hypoglycaemic Episodes	31.18 events per 100 patient years of exposure	58.55 events per 100 patient years of exposure

SECONDARY outcome

Timeframe: Weeks 0-26

Population: Full analysis set. Number of subjects analysed=subjects with data available for confirmed hypoglycaemic episodes.

The number of events was analysed using a negative binomial model with a log-link function and the logarithm of the exposure time (100 years) for which a hypoglycaemic episode is considered treatment emergent as offset. The model included treatment, anti-diabetic therapy at screening and sex as fixed factors, and age as covariate. Confirmed hypoglycaemia was defined as either severe episodes or episodes with plasma glucose < 3.1 mmol/L (56 mg/dL) with or without symptoms. A treatment emergent hypoglycaemic episode was defined as an episode that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment.

Outcome measures

Measure	IDegAsp BID n=359 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=181 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Number of Treatment Emergent Confirmed Hypoglycaemic Episodes	223.56 events per 100 patient years of exposure	394.17 events per 100 patient years of exposure

SECONDARY outcome

Timeframe: Week 0, Week 26

Population: Full analysis set. Number of subject analysed=subjects with data available for body weight.

Change from baseline in body week at week 26. The response and change from baseline in response after 26 weeks are analysed using an ANCOVA model with treatment, anti-diabetic therapy at screening and sex as fixed factors, age and baseline response as covariate. Missing values imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=359 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=182 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Change From Baseline in Body Weight	2.82 kg Standard Error 0.14	2.21 kg Standard Error 0.19

SECONDARY outcome

Timeframe: At 26 weeks

Population: Full analysis set. Number of subjects analysed=subjects who have been exposed for at least 12 treatment weeks.

A responder for HbA1c without confirmed hypoglycaemia was defined as a subject who meets the HbA1c target (<7.0%) at end of trial without treatment emergent confirmed hypoglycaemia during the last 12 weeks of treatment or within 7 days after the last randomised treatment. Confirmed hypoglycaemia was defined as either severe episodes or episodes with plasma glucose < 3.1 mmol/L (56 mg/dL) with or without symptoms. Severe hypoglycaemia: An episode requiring assistance of another person to actively administer carbohydrate, glucagon, or other resuscitative actions. Missing data is imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=348 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=173 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Responder Without Confirmed Hypoglycaemic Episodes HbA1c Below 7.0% Yes	153 Participants	48 Participants
Responder Without Confirmed Hypoglycaemic Episodes HbA1c Below 7.0% No	195 Participants	125 Participants

SECONDARY outcome

Timeframe: Weeks 0-26

Population: Safety analysis set included all subjects receiving at least one dose of IDegAsp or BIAsp.

A treatment emergent adverse event was defined as an episode that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Incidence of AEs was reported in terms of rate, defined as number of adverse events divided by patient years of exposure (PYE) multiplied by 100 (1 PYE=365.25 days).

Outcome measures

Measure	IDegAsp BID n=360 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=181 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Incidence of Treatment Emergent Adverse Events (TEAEs)	321.74 events per 100 PYE	348.82 events per 100 PYE

SECONDARY outcome

Timeframe: Week 26

Population: Full analysis set.

Number of subjects with HbA1c <7% after 26 weeks of treatment. Missing HbA1c data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=361 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=182 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Responder for HbA1c (HbA1c <7%) After 26 Weeks of Treatment Yes	201 Participants	88 Participants
Responder for HbA1c (HbA1c <7%) After 26 Weeks of Treatment No	160 Participants	94 Participants

SECONDARY outcome

Timeframe: Week 26

Population: Full analysis set.

Number of subjects with HbA1c <=6.5% after 26 weeks of treatment. Missing HbA1c data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=361 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=182 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Responder for HbA1c (HbA1c <=6.5%) After 26 Weeks of Treatment Yes	106 Participants	54 Participants
Responder for HbA1c (HbA1c <=6.5%) After 26 Weeks of Treatment No	255 Participants	128 Participants

SECONDARY outcome

Timeframe: At week 26

Population: Full analysis set. Number of subjects analysed=subjects who have been exposed for at least 12 treatment weeks.

A responder for HbA1c without severe hypoglycaemia was defined as a subject who meets the HbA1c target (<7%) at end of trial without severe treatment emergent hypoglycaemia during the last 12 weeks of treatment or within 7 days after the last randomised treatment. Severe hypoglycaemia: An episode requiring assistance of another person to actively administer carbohydrate, glucagon, or other resuscitative actions. Missing HbA1c data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=348 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=173 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Responder for HbA1c (HbA1c <7%) Without Severe Hypoglycaemic Episodes Yes	197 Participants	84 Participants
Responder for HbA1c (HbA1c <7%) Without Severe Hypoglycaemic Episodes No	151 Participants	89 Participants

SECONDARY outcome

Timeframe: At 26 weeks

Population: Full analysis set. Number of subjects analysed=subjects who have been exposed for at least 12 treatment weeks.

A responder for HbA1c without confirmed hypoglycaemia was defined as a subject who meets the HbA1c target (≤6.5%) at end of trial without treatment emergent confirmed hypoglycaemia during the last 12 weeks of treatment or within 7 days after the last randomised treatment. Confirmed hypoglycaemia was defined as either severe episodes or episodes with plasma glucose < 3.1 mmol/L (56 mg/dL) with or without symptoms. Severe hypoglycaemia: An episode requiring assistance of another person to actively administer carbohydrate, glucagon, or other resuscitative actions. Missing data is imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=348 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=173 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Responder for HbA1c (HbA1c ≤6.5%) Without Confirmed Hypoglycaemic Episodes Yes	82 Participants	26 Participants
Responder for HbA1c (HbA1c ≤6.5%) Without Confirmed Hypoglycaemic Episodes No	266 Participants	147 Participants

SECONDARY outcome

Timeframe: At week 26

Population: Full analysis set. Number of subjects analysed=subjects who have been exposed for at least 12 treatment weeks.

A responder for HbA1c without severe hypoglycaemia was defined as a subject who meets the HbA1c target (≤6.5%) at end of trial without severe treatment emergent hypoglycaemia during the last 12 weeks of treatment or within 7 days after the last randomised treatment. Severe hypoglycaemia: An episode requiring assistance of another person to actively administer carbohydrate, glucagon, or other resuscitative actions. Missing HbA1c data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=348 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=173 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Responder for HbA1c (HbA1c ≤6.5%) Without Severe Hypoglycaemic Episodes Yes	105 Participants	50 Participants
Responder for HbA1c (HbA1c ≤6.5%) Without Severe Hypoglycaemic Episodes No	243 Participants	123 Participants

SECONDARY outcome

Timeframe: At week 26

Population: Full analysis set. Number of subjects analysed=subjects with SMPG values available at individual timepoints.

SMPG values were recorded at 9 time-points: before and after (90 min after the start of the meal) breakfast, lunch, main evening meal, before bedtime, at 4 am and before breakfast on the next day. Missing values were imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=361 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=182 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
9-point Profile (SMPG) After 26 Weeks of Treatment Before breakfast	5.81 mmol/L Standard Deviation 1.29	6.81 mmol/L Standard Deviation 1.90
9-point Profile (SMPG) After 26 Weeks of Treatment 90 minutes after start of breakfast	8.68 mmol/L Standard Deviation 2.81	9.56 mmol/L Standard Deviation 3.25
9-point Profile (SMPG) After 26 Weeks of Treatment Before lunch	6.47 mmol/L Standard Deviation 2.39	6.65 mmol/L Standard Deviation 2.50
9-point Profile (SMPG) After 26 Weeks of Treatment 90 minutes after start of lunch	10.22 mmol/L Standard Deviation 3.06	10.52 mmol/L Standard Deviation 3.17
9-point Profile (SMPG) After 26 Weeks of Treatment Before main evening meal	7.09 mmol/L Standard Deviation 2.35	7.97 mmol/L Standard Deviation 2.92
9-point Profile (SMPG) After 26 Weeks of Treatment 90 minutes after the start of main evening meal	9.35 mmol/L Standard Deviation 3.04	8.96 mmol/L Standard Deviation 3.26
9-point Profile (SMPG) After 26 Weeks of Treatment Before bedtime	8.65 mmol/L Standard Deviation 3.00	8.12 mmol/L Standard Deviation 2.89
9-point Profile (SMPG) After 26 Weeks of Treatment At 4:00 a.m.	6.00 mmol/L Standard Deviation 1.93	6.78 mmol/L Standard Deviation 1.95
9-point Profile (SMPG) After 26 Weeks of Treatment Before breakfast the following day	5.69 mmol/L Standard Deviation 1.49	6.87 mmol/L Standard Deviation 1.82

SECONDARY outcome

Timeframe: At week 26

Population: Full analysis set. Number of subjects analysed=subjects with available data for SMPG.

SMPG values were recorded at 9 time-points: before and after (90 min after the start of the meal) breakfast, lunch, main evening meal, before bedtime, at 4 am and before breakfast on the next day. Missing values were imputed using last observed value. The mean of 9-point profile (SMPG) was defined as the area under the profile divided by the measurement time and was calculated using the trapezoidal method.

Outcome measures

Measure	IDegAsp BID n=360 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=182 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Mean of the 9-point Profile (SMPG) After 26 Weeks of Treatment	7.70 mmol/L Standard Deviation 1.60	8.11 mmol/L Standard Deviation 1.71

SECONDARY outcome

Timeframe: At week 26

Population: Full analysis set. Number of subjects analysed=subjects with available data for SMPG.

SMPG values were recorded at 9 time-points: before and after (90 min after the start of the meal) breakfast, lunch, main evening meal, before bedtime, at 4 am and before breakfast on the next day. Missing values were imputed using last observed value. Fluctuation in 9-point SMPG profile is the average absolute difference to the mean of the profile of the 9-point SMPG measurements accumulated over the profile.

Outcome measures

Measure	IDegAsp BID n=360 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=182 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Fluctuation in the 9-point Profile (SMPG) After 26 Weeks of Treatment	1.40 mmol/L Interval 0.08 to 4.9	1.43 mmol/L Interval 0.04 to 4.18

SECONDARY outcome

Timeframe: At week 26

Population: Full analysis set. Number of subjects analysed=subjects with SMPG values available before and after (90 min after the start of the meal) breakfast, lunch, main evening meal.

SMPG values were recorded at 9 time-points: before and after (90 min after the start of the meal) breakfast, lunch, main evening meal, before bedtime, at 4 am and before breakfast on the next day. Missing values were imputed using last observed value. Prandial PG increment for each meal was derived from the 9-point profile (SMPG) as the difference between PG values after meal (90 min) and before meal. The reported results are mean prandial PG increment overall meals (the mean of all available meal increments).

Outcome measures

Measure	IDegAsp BID n=359 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=182 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
9-point Profile (SMPG) - Prandial Plasma Glucose (PG) Increment After 26 Weeks of Treatment	2.95 mmol/L Standard Deviation 2.10	2.60 mmol/L Standard Deviation 2.15

SECONDARY outcome

Timeframe: From randomization till achievement of titration target (up to week 26)

Population: Full analysis set

2-point profiles (SMPG): pre-breakfast and pre-dinner (main evening meal) SMPGs were to be taken for titration/dose adjustments. The pre-specified titration targets were to achieve SMPG level <5 mmol/L (90 mg/dL) both before breakfast and before main evening meal. For each target, the time from randomisation to the date a subject achieves the titration target for the first time was derived (Kaplan-Meier method). Reported results are time to all titration target (pre-breakfast and pre-dinner) was met for the first time.

Outcome measures

Measure	IDegAsp BID n=361 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=182 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
2-point Profile (SMPG) Measurements Obtained Throughout the Trial for Dose Adjustment - Time From Randomisation (Measured in Weeks) to Achieve Titration Targets	NA week Interval 24.0 to Median and upper bound of inter-quartile range could not be calculated as inadequate number of subjects achieved the pre-defined titration target at the end of trial.	NA week Median and IQR could not be calculated as inadequate number of subjects achieved the pre-defined titration target at the end of trial

SECONDARY outcome

Timeframe: At week 26

Population: Full analysis set. Number of subjects analysed=subjects who contributed to this analysis.

The logarithm transformed SMPG values available before breakfast and main evening meal were analysed separately as repeated measures in a linear mixed model with treatment, anti-diabetic therapy at screening and sex as fixed factors, age as covariate and subject as random factor. The model assumed independent within- and between-subject errors with variances depending on treatment. Within-subject variability as measured by CoV% for a treatment could be calculated from the corresponding residual variance.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=180 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
2-point Profile (SMPG) Measurements Obtained Throughout the Trial for Dose Adjustment - Within-subject Variability as Measured by Coefficient of Variance (CV)% After 26 Weeks of Treatment Before breakfast	17.26 Coefficient of variation	17.19 Coefficient of variation
2-point Profile (SMPG) Measurements Obtained Throughout the Trial for Dose Adjustment - Within-subject Variability as Measured by Coefficient of Variance (CV)% After 26 Weeks of Treatment Before main evening meal	22.45 Coefficient of variation	22.96 Coefficient of variation

SECONDARY outcome

Timeframe: Week 0-26

Population: Safety analysis set.

ADA classification of hypoglycaemia:

1. Severe: An episode requiring assistance of another person to actively administer carbohydrate, glucagon, or other resuscitative actions.

2. Asymptomatic: An episode not accompanied by typical symptoms of hypoglycaemia, but with a measured PG level ≤3.9 mmol/L.

3. Documented symptomatic: An episode during which typical symptoms of hypoglycaemia are accompanied by a measured PG level ≤3.9 mmol/L.

4. Relative: An episode during which the person with diabetes reports any of the typical symptoms of hypoglycaemia, and interprets those as indicative of hypoglycaemia, but with a measured PG level >3.9 mmol/L.

5. Probable symptomatic: An episode during which symptoms of hypoglycaemia are not accompanied by a PG determination but that was presumably caused by a PG level ≤3.9 mmol/L.

Number of treatment emergent hypoglycaemic episodes were reported in terms of rate, defined as number of events divided by PYE multiplied by 100 (1 PYE=365.25 days).

Outcome measures

Measure	IDegAsp BID n=360 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=181 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Number of Treatment Emergent Hypoglycaemic Episodes According to the American Diabetes Association (ADA) Definition During 26 Weeks of Treatment	1699.53 hypoglycaemic episodes per 100 PYE	1585.31 hypoglycaemic episodes per 100 PYE

SECONDARY outcome

Timeframe: From week 16 to end of treatment (week 26) + 1 week follow-up

Population: Safety analysis set

Confirmed hypoglycaemia was defined as either severe episodes or episodes with plasma glucose < 3.1 mmol/L (56 mg/dL) with or without symptoms. Severe: An episode requiring assistance of another person to actively administer carbohydrate, glucagon, or other resuscitative actions. Maintenance period was defined as the period from week 16 to end of treatment, including 1 week follow-up.

Number of treatment emergent hypoglycaemic episodes were reported in terms of rate, defined as number of events divided by PYE multiplied by 100 (1 PYE=365.25 days).

Outcome measures

Measure	IDegAsp BID n=360 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=181 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Number of Treatment Emergent Confirmed Hypoglycaemic Episodes in the Maintenance Period	159.08 hypoglycaemic episodes per 100 PYE	361.18 hypoglycaemic episodes per 100 PYE

SECONDARY outcome

Timeframe: From week 16 to end of treatment (week 26) + 1 week follow-up

Population: Safety analysis set

Confirmed hypoglycaemia was defined as either severe episodes or episodes with plasma glucose < 3.1 mmol/L (56 mg/dL) with or without symptoms. Severe hypoglycaemia: An episode requiring assistance of another person to actively administer carbohydrate, glucagon, or other resuscitative actions. The nocturnal period was defined as the period between 00:01 and 05:59 a.m. (both inclusive). Maintenance period was defined as the period from week 16 to end of treatment, including 1 week follow-up.

Number of treatment emergent hypoglycaemic episodes were reported in terms of rate, defined as number of events divided by PYE multiplied by 100 (1 PYE=365.25 days).

Outcome measures

Measure	IDegAsp BID n=360 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=181 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Number of Treatment Emergent Nocturnal (00:01-05:59) Confirmed Hypoglycaemic Episodes in the Maintenance Period	19.20 hypoglycaemic episodes per 100 PYE	38.60 hypoglycaemic episodes per 100 PYE

SECONDARY outcome

Timeframe: Week 0, week 26

Population: Full analysis set. Number of subjects analysed=subjects with data available for individual component of SF-36 questionnaire.

Change from baseline in SF-36 at week 26. The questionnaire contains 36 items across 8 domains and 2 summary scores. Score range: 0 (worst score) to 100 (best score). Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=361 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=182 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Change in Health Related Quality of Life Questionnaire (SF -36) Physical Score	0.19 units on a scale Standard Deviation 6.89	-0.13 units on a scale Standard Deviation 6.61
Change in Health Related Quality of Life Questionnaire (SF -36) Physical Functioning	-1.01 units on a scale Standard Deviation 6.85	-0.50 units on a scale Standard Deviation 6.58
Change in Health Related Quality of Life Questionnaire (SF -36) Role-Physical	0.41 units on a scale Standard Deviation 8.88	-0.18 units on a scale Standard Deviation 8.92
Change in Health Related Quality of Life Questionnaire (SF -36) Bodily Pain	0.93 units on a scale Standard Deviation 10.10	-0.71 units on a scale Standard Deviation 10.38
Change in Health Related Quality of Life Questionnaire (SF -36) General Health	3.00 units on a scale Standard Deviation 10.24	2.19 units on a scale Standard Deviation 9.45
Change in Health Related Quality of Life Questionnaire (SF -36) Mental Score	2.08 units on a scale Standard Deviation 9.65	0.72 units on a scale Standard Deviation 10.11
Change in Health Related Quality of Life Questionnaire (SF -36) Vitality	1.56 units on a scale Standard Deviation 9.57	0.69 units on a scale Standard Deviation 9.33
Change in Health Related Quality of Life Questionnaire (SF -36) Social Functioning	1.35 units on a scale Standard Deviation 8.46	-0.12 units on a scale Standard Deviation 9.98
Change in Health Related Quality of Life Questionnaire (SF -36) Role-Emotional	1.54 units on a scale Standard Deviation 10.60	1.01 units on a scale Standard Deviation 11.14
Change in Health Related Quality of Life Questionnaire (SF -36) Mental Health	1.39 units on a scale Standard Deviation 9.53	-0.17 units on a scale Standard Deviation 9.81

SECONDARY outcome

Timeframe: Week 0, week 26

Population: Full analysis set. Number of subjects analysed=subjects with data available for TRIM-D total score.

Change from baseline in TRIM-D scores at week 26. TRIM-D score measured treatment satisfaction which included an overall score as well the subscale scores (daily life, diabetes management, compliance and psychological health). The scores were transformed to a 0-100 scale with higher scores indicating less treatment related impact. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=360 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=181 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Change in Treatment Satisfaction Questionnaire (Treatment Related Impact Measure-diabetes [TRIM-D])	4.87 units on a scale Standard Deviation 11.15	4.11 units on a scale Standard Deviation 10.49

SECONDARY outcome

Timeframe: At week 26

Population: Full analysis set. Number of subjects analysed=subjects with data available for TRIM-D device total score.

TRIM-D device is an eight item measure with two domains assessing Device Bother and Device Function. This captures information on the ease of use, convenience, and handling of the device(s) used to take diabetes medication. The measure has acceptable reliability, validity and ability to detect change. The scores were transformed to a 0-100 scale with higher scores indicating less treatment related impact. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Treatment Satisfaction Questionnaire (Treatment Related Impact Measures - Diabetes Device [TRIM-D Device])	72.31 units on a scale Standard Deviation 14.78	69.43 units on a scale Standard Deviation 14.12

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How easy or difficult is it to read the dose scale) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult and Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Easy or Difficult is it to Read the Dose Scale) Very easy	220 Participants	104 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Read the Dose Scale) Fairly easy	128 Participants	69 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Read the Dose Scale) Neither easy nor difficult	10 Participants	5 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Read the Dose Scale) Rather difficult	0 Participants	0 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How easy or difficult do you find it to hold the pen stable when injecting?) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult and Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Easy or Difficult do You Find it to Hold the Pen Stable When Injecting?) Very easy	228 Participants	99 Participants
Device Specific Questionnaires I (How Easy or Difficult do You Find it to Hold the Pen Stable When Injecting?) Fairly easy	121 Participants	72 Participants
Device Specific Questionnaires I (How Easy or Difficult do You Find it to Hold the Pen Stable When Injecting?) Neither easy nor difficult	8 Participants	7 Participants
Device Specific Questionnaires I (How Easy or Difficult do You Find it to Hold the Pen Stable When Injecting?) Very difficult	1 Participants	0 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How easy or difficult is it to hear the clicks for each unit increment?) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult and Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Easy or Difficult is it to Hear the Clicks for Each Unit Increment?) Very easy	230 Participants	108 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Hear the Clicks for Each Unit Increment?) Fairly easy	119 Participants	63 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Hear the Clicks for Each Unit Increment?) Neither easy nor difficult	9 Participants	6 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Hear the Clicks for Each Unit Increment?) Rather difficult	0 Participants	1 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How easy or difficult is it to feel the clicks for each unit increment?) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult and Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Easy or Difficult is it to Feel the Clicks for Each Unit Increment?) Very easy	217 Participants	103 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Feel the Clicks for Each Unit Increment?) Fairly easy	128 Participants	65 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Feel the Clicks for Each Unit Increment?) Neither easy nor difficult	12 Participants	9 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Feel the Clicks for Each Unit Increment?) Rather difficult	1 Participants	1 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How easy or difficult is it to push down the injection button?) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult and Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Easy or Difficult is it to Push Down the Injection Button?) Very easy	224 Participants	89 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Push Down the Injection Button?) Fairly easy	125 Participants	75 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Push Down the Injection Button?) Neither easy nor difficult	9 Participants	14 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to Push Down the Injection Button?) Rather difficult	0 Participants	0 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How easy/difficult is it to turn the dose selector when choosing the right dose?) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult and Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Easy/Difficult is it to Turn the Dose Selector When Choosing the Right Dose?) Very easy	222 Participants	99 Participants
Device Specific Questionnaires I (How Easy/Difficult is it to Turn the Dose Selector When Choosing the Right Dose?) Fairly easy	128 Participants	72 Participants
Device Specific Questionnaires I (How Easy/Difficult is it to Turn the Dose Selector When Choosing the Right Dose?) Neither easy nor difficult	7 Participants	6 Participants
Device Specific Questionnaires I (How Easy/Difficult is it to Turn the Dose Selector When Choosing the Right Dose?) Rather difficult	1 Participants	0 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How easy/difficult is it to know if the push button has been pushed completely down?) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult or Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Easy/Difficult is it to Know if the Push Button Has Been Pushed Completely Down?) Neither easy nor difficult	11 Participants	9 Participants
Device Specific Questionnaires I (How Easy/Difficult is it to Know if the Push Button Has Been Pushed Completely Down?) Very easy	213 Participants	92 Participants
Device Specific Questionnaires I (How Easy/Difficult is it to Know if the Push Button Has Been Pushed Completely Down?) Fairly easy	134 Participants	77 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How easy or difficult is it to see the dose scale when injecting?) was measured on following categories: Very confident, Very easy, Fairly easy, Neither easy nor difficult, Rather difficult, and Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Easy or Difficult is it to See the Dose Scale When Injecting?) Very confident	1 Participants	0 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to See the Dose Scale When Injecting?) Very easy	215 Participants	94 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to See the Dose Scale When Injecting?) Fairly easy	126 Participants	78 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to See the Dose Scale When Injecting?) Neither easy nor difficult	14 Participants	4 Participants
Device Specific Questionnaires I (How Easy or Difficult is it to See the Dose Scale When Injecting?) Rather difficult	2 Participants	2 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How confident are you that you set the insulin dose correctly every time?) was measured on following categories: Very confident, Rather confident, Fairly confident, Not very confident and Not at all confident. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Confident Are You That You Set the Insulin Dose Correctly Every Time?) Very confident	209 Participants	85 Participants
Device Specific Questionnaires I (How Confident Are You That You Set the Insulin Dose Correctly Every Time?) Fairly confident	43 Participants	26 Participants
Device Specific Questionnaires I (How Confident Are You That You Set the Insulin Dose Correctly Every Time?) Rather confident	106 Participants	65 Participants
Device Specific Questionnaires I (How Confident Are You That You Set the Insulin Dose Correctly Every Time?) Not very confident	0 Participants	2 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How confident are you that you inject the correct amount of insulin every time?) was measured on following categories: Very confident, Rather confident, Fairly confident, Not very confident and Not at all confident. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Confident Are You That You Inject the Correct Amount of Insulin Every Time?) Very confident	202 Participants	83 Participants
Device Specific Questionnaires I (How Confident Are You That You Inject the Correct Amount of Insulin Every Time?) Fairly confident	52 Participants	28 Participants
Device Specific Questionnaires I (How Confident Are You That You Inject the Correct Amount of Insulin Every Time?) Rather confident	104 Participants	67 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (Overall, how confident are you in your management of your daily insulin injections using this pen?) was measured on following categories: Very confident, Fairly confident, Rather confident, Not very confident and Not at all confident. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (Overall, How Confident Are You in Your Management of Your Daily Insulin Injections Using This Pen?) Very confident	197 Participants	82 Participants
Device Specific Questionnaires I (Overall, How Confident Are You in Your Management of Your Daily Insulin Injections Using This Pen?) Fairly confident	59 Participants	39 Participants
Device Specific Questionnaires I (Overall, How Confident Are You in Your Management of Your Daily Insulin Injections Using This Pen?) Rather confident	102 Participants	57 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (Overall, how confident are you in controlling your blood sugar level using this pen?) was measured on following categories: Very confident, Fairly confident, Rather confident, Not very confident and Not at all confident. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (Overall, How Confident Are You in Controlling Your Blood Sugar Level Using This Pen?) Very confident	180 Participants	71 Participants
Device Specific Questionnaires I (Overall, How Confident Are You in Controlling Your Blood Sugar Level Using This Pen?) Fairly confident	85 Participants	60 Participants
Device Specific Questionnaires I (Overall, How Confident Are You in Controlling Your Blood Sugar Level Using This Pen?) Rather confident	92 Participants	42 Participants
Device Specific Questionnaires I (Overall, How Confident Are You in Controlling Your Blood Sugar Level Using This Pen?) Not very confident	1 Participants	3 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How suitable is the pen to use in public?) was measured on following scale: Very suitable, Fairly suitable, Rather suitable, Not very suitable and Not at all suitable. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Suitable is the Pen to Use in Public?) Very suitable	129 Participants	60 Participants
Device Specific Questionnaires I (How Suitable is the Pen to Use in Public?) Fairly suitable	117 Participants	64 Participants
Device Specific Questionnaires I (How Suitable is the Pen to Use in Public?) Rather confident	0 Participants	1 Participants
Device Specific Questionnaires I (How Suitable is the Pen to Use in Public?) Rather suitable	94 Participants	35 Participants
Device Specific Questionnaires I (How Suitable is the Pen to Use in Public?) Not very suitable	18 Participants	18 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How confident are you that the air shot has been done correctly?) was measured on following categories: Very confident, Fairly confident, Rather confident, Not very confident and Not at all confident. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Confident Are You That the Air Shot Has Been Done Correctly?) Very confident	157 Participants	73 Participants
Device Specific Questionnaires I (How Confident Are You That the Air Shot Has Been Done Correctly?) Fairly confident	85 Participants	55 Participants
Device Specific Questionnaires I (How Confident Are You That the Air Shot Has Been Done Correctly?) Rather confident	109 Participants	47 Participants
Device Specific Questionnaires I (How Confident Are You That the Air Shot Has Been Done Correctly?) Not very confident	7 Participants	3 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How convenient do you find the size of the pen?) was measured on following categories: Very convenient, Fairly convenient, Rather convenient, Not very convenient and Not at all convenient. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Convenient do You Find the Size of the Pen?) Very convenient	143 Participants	63 Participants
Device Specific Questionnaires I (How Convenient do You Find the Size of the Pen?) Fairly convenient	109 Participants	69 Participants
Device Specific Questionnaires I (How Convenient do You Find the Size of the Pen?) Rather convenient	101 Participants	42 Participants
Device Specific Questionnaires I (How Convenient do You Find the Size of the Pen?) Not very convenient	5 Participants	4 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Device Specific Questionnaires I (How comfortable do you find the handling of the pen?) was measured on following categories: Very comfortable, Fairly comfortable, Rather comfortable, Not very comfortable and Not at all comfortable. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (How Comfortable do You Find the Handling of the Pen?) Very comfortable	146 Participants	66 Participants
Device Specific Questionnaires I (How Comfortable do You Find the Handling of the Pen?) Fairly comfortable	107 Participants	68 Participants
Device Specific Questionnaires I (How Comfortable do You Find the Handling of the Pen?) Rather comfortable	104 Participants	44 Participants
Device Specific Questionnaires I (How Comfortable do You Find the Handling of the Pen?) Not very comfortable	1 Participants	0 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 1.

Participants were asked to report whether they had any problems using the pen. Number of participants reporting "yes" and "no" are presented. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires I (Did You Have Any Problems Using the Pen?) Yes	12 Participants	8 Participants
Device Specific Questionnaires I (Did You Have Any Problems Using the Pen?) No	346 Participants	169 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 2.

Device Specific Questionnaires II (How easy or difficult was it to learn how to use this pen?) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult and Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires II (How Easy or Difficult Was it to Learn How to Use This Pen?) Very easy	206 Participants	92 Participants
Device Specific Questionnaires II (How Easy or Difficult Was it to Learn How to Use This Pen?) Fairly easy	145 Participants	79 Participants
Device Specific Questionnaires II (How Easy or Difficult Was it to Learn How to Use This Pen?) Neither easy nor difficult	7 Participants	7 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 2.

Device Specific Questionnaires II (How easy or difficult is it to distinguish between dialling up and down?) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult and Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires II (How Easy or Difficult is it to Distinguish Between Dialling up and Down?) Very easy	203 Participants	91 Participants
Device Specific Questionnaires II (How Easy or Difficult is it to Distinguish Between Dialling up and Down?) Fairly easy	144 Participants	81 Participants
Device Specific Questionnaires II (How Easy or Difficult is it to Distinguish Between Dialling up and Down?) Neither easy nor difficult	10 Participants	6 Participants
Device Specific Questionnaires II (How Easy or Difficult is it to Distinguish Between Dialling up and Down?) Rather difficult	1 Participants	0 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 2.

Device Specific Questionnaires II (How easy or difficult is it to inject your usual insulin dose?) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult and Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires II (How Easy or Difficult is it to Inject Your Usual Insulin Dose?) Very easy	193 Participants	93 Participants
Device Specific Questionnaires II (How Easy or Difficult is it to Inject Your Usual Insulin Dose?) Fairly easy	154 Participants	81 Participants
Device Specific Questionnaires II (How Easy or Difficult is it to Inject Your Usual Insulin Dose?) Neither easy nor difficult	11 Participants	3 Participants
Device Specific Questionnaires II (How Easy or Difficult is it to Inject Your Usual Insulin Dose?) Rather difficult	0 Participants	1 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 2.

Device Specific Questionnaires II (How easy/difficult is it to reach the dose button when inject your insulin dose?) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult and Very difficult. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires II (How Easy/Difficult is it to Reach the Dose Button When Inject Your Insulin Dose?) Very easy	193 Participants	90 Participants
Device Specific Questionnaires II (How Easy/Difficult is it to Reach the Dose Button When Inject Your Insulin Dose?) Fairly easy	153 Participants	82 Participants
Device Specific Questionnaires II (How Easy/Difficult is it to Reach the Dose Button When Inject Your Insulin Dose?) Neither easy nor difficult	11 Participants	6 Participants
Device Specific Questionnaires II (How Easy/Difficult is it to Reach the Dose Button When Inject Your Insulin Dose?) Rather difficult	1 Participants	0 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 2.

Device Specific Questionnaires II (How easy or difficult is it to inject yourself in different places of the body using this pen?) was measured on following categories: Very easy, Fairly easy, Neither easy nor difficult, Rather difficult, Very difficult and Not applicable. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires II (How Easy or Difficult is it to Inject Yourself in Different Places of the Body Using This Pen?) Very easy	167 Participants	77 Participants
Device Specific Questionnaires II (How Easy or Difficult is it to Inject Yourself in Different Places of the Body Using This Pen?) Fairly easy	169 Participants	85 Participants
Device Specific Questionnaires II (How Easy or Difficult is it to Inject Yourself in Different Places of the Body Using This Pen?) Neither easy nor difficult	20 Participants	16 Participants
Device Specific Questionnaires II (How Easy or Difficult is it to Inject Yourself in Different Places of the Body Using This Pen?) Rather difficult	1 Participants	0 Participants
Device Specific Questionnaires II (How Easy or Difficult is it to Inject Yourself in Different Places of the Body Using This Pen?) Very difficult	1 Participants	0 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 2.

Device Specific Questionnaires II (How confident are you that the full dose has been delivered?) was measured on following categories: Very confident, Fairly confident, Fairly easy, Rather confident, Not very confident and Not at all confident. Reported results are number of participants reporting the individual category of the question at week 26. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires II (How Confident Are You That the Full Dose Has Been Delivered?) Not at all confident	1 Participants	0 Participants
Device Specific Questionnaires II (How Confident Are You That the Full Dose Has Been Delivered?) Very confident	184 Participants	79 Participants
Device Specific Questionnaires II (How Confident Are You That the Full Dose Has Been Delivered?) Fairly confident	54 Participants	42 Participants
Device Specific Questionnaires II (How Confident Are You That the Full Dose Has Been Delivered?) Fairly easy	0 Participants	1 Participants
Device Specific Questionnaires II (How Confident Are You That the Full Dose Has Been Delivered?) Rather confident	117 Participants	56 Participants
Device Specific Questionnaires II (How Confident Are You That the Full Dose Has Been Delivered?) Not very confident	2 Participants	0 Participants

SECONDARY outcome

Timeframe: week 26

Population: Full analysis set. Number of participants analysed=subjects with data available for Device specific questionnaire 2.

Participants were asked the question "would you recommend the pen?" Number of participants reporting "yes" and "no" are presented. Missing data was imputed using last observed value.

Outcome measures

Measure	IDegAsp BID n=358 Participants Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=178 Participants Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Device Specific Questionnaires II (Would You Recommend the Pen?) Yes	338 Participants	161 Participants
Device Specific Questionnaires II (Would You Recommend the Pen?) No	19 Participants	17 Participants

Adverse Events

IDegAsp BID

Serious events: 16 serious events

Other events: 109 other events

Deaths: 0 deaths

BIAsp 30 BID

Serious events: 14 serious events

Other events: 65 other events

Deaths: 0 deaths

Serious adverse events

Measure	IDegAsp BID n=360 participants at risk Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=181 participants at risk Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Cardiac disorders Acute myocardial infarction	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Cardiac disorders Angina pectoris	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Injury, poisoning and procedural complications Ankle fracture	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
General disorders Asthenia	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Surgical and medical procedures Cataract operation	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Nervous system disorders Cerebral infarction	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	1.1% 2/181 • Number of events 2 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Nervous system disorders Cerebral ischaemia	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Hepatobiliary disorders Cholecystitis acute	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Hepatobiliary disorders Cholelithiasis	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Nervous system disorders Diabetic neuropathy	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Gastrointestinal disorders Dysbacteriosis	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Pregnancy, puerperium and perinatal conditions Ectopic pregnancy	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Infections and infestations Gastroenteritis	0.56% 2/360 • Number of events 2 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Gastrointestinal disorders Gastrooesophageal reflux disease	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Injury, poisoning and procedural complications Humerus fracture	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Vascular disorders Hypertension	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Nervous system disorders Hypoglycaemic coma	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	1.1% 2/181 • Number of events 3 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Injury, poisoning and procedural complications Joint dislocation	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Nervous system disorders Lacunar infarction	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Neoplasms benign, malignant and unspecified (incl cysts and polyps) Lung adenocarcinoma	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Gastrointestinal disorders Pancreatitis acute	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Injury, poisoning and procedural complications Patella fracture	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Musculoskeletal and connective tissue disorders Patellofemoral pain syndrome	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Infections and infestations Pharyngitis	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Infections and infestations Pneumonia	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Injury, poisoning and procedural complications Soft tissue injury	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Musculoskeletal and connective tissue disorders Spinal osteoarthritis	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Neoplasms benign, malignant and unspecified (incl cysts and polyps) Uterine leiomyoma	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
General disorders Vascular stent occlusion	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Ear and labyrinth disorders Vertigo	0.00% 0/360 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.55% 1/181 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Ear and labyrinth disorders Vertigo positional	0.28% 1/360 • Number of events 1 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	0.00% 0/181 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.

Other adverse events

Measure	IDegAsp BID n=360 participants at risk Subjects received subcutaneous (s.c.) injections of insulin degludec/insulin aspart (IDegAsp) twice daily (BID) for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner self-measured plasma glucose (SMPG) ≤ 5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on the day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.	BIAsp 30 BID n=181 participants at risk Subjects received s.c. injections of biphasic insulin aspart 30 (BIAsp 30, a mixture of soluble insulin aspart \[IAsp\] 30% and protaminated IAsp 70%) BID for 26 weeks. Subjects started insulin treatment at dose corresponding to their pre-trial insulin treatment. The insulin dose adjustments aimed to reach pre-breakfast and pre-dinner SMPG ≤5.0 mmol/L, based on the subject's pre-breakfast and pre-dinner SMPG. The dose taken with main evening meal was adjusted according to mean pre-breakfast plasma glucose measured on day of adjustment and the two preceding days. The dose taken at breakfast was adjusted according to mean pre-main evening meal plasma glucose measured on the three days preceding the adjustment day. If one of the PG values was below target (\< 4.0 mmol/L), the corresponding dose was to be reduced. If one or more SMPG values were missing, the adjustment was to be performed on the remaining SMPG values. Subjects continued the pre-trial treatment with metformin at stable dose.
Eye disorders Diabetic retinopathy	10.0% 36/360 • Number of events 37 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	6.6% 12/181 • Number of events 12 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Infections and infestations Upper respiratory tract infection	21.1% 76/360 • Number of events 90 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	25.4% 46/181 • Number of events 52 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.
Infections and infestations Viral upper respiratory tract infection	2.5% 9/360 • Number of events 10 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.	5.5% 10/181 • Number of events 12 • Week 0 to week 26 (+ 1 week of follow-up) A treatment emergent adverse event was defined as an event that had onset date on or after the first day of exposure to randomised treatment and no later than 7 days after the last day of randomised treatment. Summary is based on the safety analysis set.

Additional Information

Clinical Reporting Anchor and Disclosure (1452)

Novo Nordisk A/S

Phone: (+1) 866-867-7178

Email: clinicaltrials@novonordisk.com

Results disclosure agreements

• Principal investigator is a sponsor employee At the end of the trial, one or more scientific publications may be prepared collaboratively by the investigator(s) and Novo Nordisk. Novo Nordisk reserves the right to postpone publication and/or communication for up to 60 days to protect intellectual property.
• Publication restrictions are in place

Restriction type: OTHER