A Study to Test How Well Empagliflozin Works in Japanese People With Type 2 Diabetes Who Are Older Than 65 Years

Change in glycated hemoglobin (HbA1c) (in units of %) from baseline after 52 weeks of treatment was modelled using a restricted maximum likelihood (REML)-based mixed model repeated measures (MMRM) which included fixed classification effects for treatment, gender, baseline renal function, visit and visit-by-treatment interaction, and a linear covariate for baseline HbA1c and age. The term baseline refers to the last observed measurement prior to the administration of any randomised trial medication.The Least Squares Mean (Standard Error) after 52 weeks of treatment is reported.

Time frame: Change in HbA1c from baseline after 52 weeks of treatment was calculated using the MMRM model which is a longitudinal analyses and it incorporates HbA1c values from baseline and after 4 weeks, 12 weeks, 24 weeks, 36 weeks and 52 weeks of treatment.

Population: Full Analysis Set (Observed cases): This set included all patients randomised, treated with at least one dose of trial medication, and with a baseline and at least one on-treatment HbA1c value. Only the available data that were observed while the patients were on treatment were analysed.

Body fat mass was estimated via bioelectrical impedance analysis (BIA) which is a commonly used method for estimating body composition and it assesses body composition by passing a very small current through the body and assessing differences in impedance caused by the fact that fat and lean tissues have different electrical properties. To measure the body composition the patient barefoot stood on the bench evenly on the toe and heel electrodes and held the grip on each hand. Change of body fat measurement from baseline to Week 52 was modelled using an Analysis of Covariance (ANCOVA) which included baseline body fat measurement, age, baseline glycated hemoglobin (HbA1c), baseline body mass index (BMI) as linear covariates and sex, treatment as fixed effects. The term baseline refers to the last observed measurement prior to the administration of any randomised trial medication.

Time frame: At baseline and at Week 52

Population: Treated set (Observed cases, including values after rescue (OC-IR)): All patients randomized and treated with at least one dose of trial medication. Values measured after rescue medication were also used. Patients with missing values at Week 52 or at baseline were not included in the analysis.

Bone mineral content (estimated bone mass) was estimated via bioelectrical impedance analysis (BIA) which is a commonly used method for estimating body composition, and it assesses body composition by passing a very small current through the body and assessing differences in impedance caused by the fact that fat and lean tissues have different electrical properties. To measure the body composition the patient barefoot stood on the bench evenly on the toe and heel electrodes and held the grip on each hand. Change of bone mineral content from baseline to Week 52 was modelled using an Analysis of Covariance (ANCOVA) which included baseline bone mineral content, age, baseline glycated hemoglobin (HbA1c), baseline body mass index (BMI) as linear covariates and sex, treatment as fixed effects. The term baseline refers to the last observed measurement prior to the administration of any randomised trial medication.

Time frame: At baseline and at Week 52.

Population: Treated set (Observed cases, including values after rescue (OC-IR)): All patients randomized and treated with at least one dose of trial medication. Values measured after rescue medication were also used. Patients with missing values at Week 52 or at baseline were not included in the analysis.

A Smedley-type dynamometer was used to measure grip strength. The site staff instructed the patient to adjust the grip width so that the second joint of the index finger is approximately 90 degrees (almost right angle). The site staff asked the patient to be careful not to touch the body or clothes with hand while keeping arms down naturally. The site staff made sure that the patient does not wave the grip dynamometer. Grip strength was measured twice alternately left and right. Change of grip strength from baseline to Week 52 was modelled using an Analysis of Covariance (ANCOVA) which included baseline grip strength, age, baseline glycated hemoglobin (HbA1c), baseline body mass index (BMI) as linear covariates and sex, treatment as fixed effects. The term baseline refers to the last observed measurement prior to the administration of any randomised trial medication.

Time frame: At baseline and at Week 52.

Population: Treated set (Observed cases, including values after rescue (OC-IR)): All patients randomized and treated with at least one dose of trial medication. Values measured after rescue medication were also used. Patients with missing values at Week 52 or at baseline were not included in the analysis.

Lean body mass (fat-free mass) was estimated via bioelectrical impedance analysis (BIA) which is a commonly used method for estimating body composition, and it assesses body composition by passing a very small current through the body and assessing differences in impedance caused by the fact that fat and lean tissues have different electrical properties. To measure the body composition the patient barefoot stood on the bench evenly on the toe and heel electrodes and held the grip on each hand. Change of lean body mass from baseline to Week 52 was modelled using an Analysis of Covariance (ANCOVA) which included baseline lean body mass, age, baseline glycated hemoglobin (HbA1c), baseline body mass index (BMI) as linear covariates and sex, treatment as fixed effects. The term baseline refers to the last observed measurement prior to the administration of any randomised trial medication.

Time frame: At baseline and at Week 52.

Population: Treated set (Observed cases, including values after rescue (OC-IR)): All patients randomized and treated with at least one dose of trial medication. Values measured after rescue medication were also used. Patients with missing values at Week 52 or at baseline were not included in the analysis.

Muscle mass was estimated via bioelectrical impedance analysis (BIA) which is a commonly used method for estimating body composition, and it assesses body composition by passing a very small current through the body and assessing differences in impedance caused by the fact that fat and lean tissues have different electrical properties. To measure the body composition the patient barefoot stood on the bench evenly on the toe and heel electrodes and held the grip on each hand. Change of muscle mass from baseline to Week 52 was modelled using an Analysis of Covariance (ANCOVA) which included baseline muscle mass, age, baseline glycated hemoglobin (HbA1c), baseline body mass index (BMI) as linear covariates and sex, treatment as fixed effects. The term baseline refers to the last observed measurement prior to the administration of any randomised trial medication.

Time frame: At baseline and at Week 52

Population: Treated set (Observed cases, including values after rescue (OC-IR)): All patients randomized and treated with at least one dose of trial medication. Values measured after rescue medication were also used. Patients with missing values at Week 52 or at baseline were not included in the analysis.

Skeletal muscle index is calculated by dividing the limb muscle mass (kg) by the square of the height (m2). The limb muscle mass was estimated via bioelectrical impedance analysis (BIA) which is a commonly used method for estimating body composition, and it assesses body composition by passing a very small current through the body and assessing differences in impedance caused by the fact that fat and lean tissues have different electrical properties. To measure the body composition the patient barefoot stood on the bench evenly on the toe and heel electrodes and held the grip on each hand. Change of skeletal muscle index from baseline to Week 52 was modelled using an Analysis of Covariance (ANCOVA) which included baseline skeletal muscle index, age, baseline glycated hemoglobin (HbA1c), baseline body mass index as linear covariates and sex, treatment as fixed effects. Baseline refers to the last observed measurement prior to the administration of any randomised trial medication.

Time frame: At baseline and at Week 52.

Population: Treated set (Observed cases, including values after rescue (OC-IR)): All patients randomized and treated with at least one dose of trial medication. Values measured after rescue medication were also used. Patients with missing values at Week 52 or at baseline were not included in the analysis.

For the stand test patients fold their arms across their chest and try to stand up once from a chair. If patients can stand from a chair, they repeat same action five times. It is measured the time required to perform five rise from a chair to an upright position as fast as possible without the use of arms. Change of time in the 5-time chair stand test from baseline to Week 52 was modelled using an Analysis of Covariance (ANCOVA) which included baseline 5-time chair stand test, age, baseline glycated hemoglobin (HbA1c), baseline body mass index (BMI) as linear covariates and sex, treatment as fixed effects. The term baseline refers to the last observed measurement prior to the administration of any randomised trial medication.

Time frame: At baseline and at Week 52.

Population: Treated set (Observed cases, including values after rescue (OC-IR)): All patients randomized and treated with at least one dose of trial medication. Values measured after rescue medication were also used. Patients with missing values at Week 52 or at baseline were not included in the analysis.

Total body water was estimated via bioelectrical impedance analysis (BIA) which is a commonly used method for estimating body composition, and it assesses body composition by passing a very small current through the body and assessing differences in impedance caused by the fact that fat and lean tissues have different electrical properties. To measure the body composition the patient barefoot stood on the bench evenly on the toe and heel electrodes and held the grip on each hand. Change of total body water from baseline to Week 52 was modelled using an Analysis of Covariance (ANCOVA) which included baseline total body water, age, baseline glycated hemoglobin (HbA1c), baseline body mass index (BMI) as linear covariates and sex, treatment as fixed effects. The term baseline refers to the last observed measurement prior to the administration of any randomised trial medication.

Time frame: At baseline and at Week 52.

Population: Treated set (Observed cases, including values after rescue (OC-IR)): All patients randomized and treated with at least one dose of trial medication. Values measured after rescue medication were also used. Patients with missing values at Week 52 or at baseline were not included in the analysis.