ClinicalTrials.gov
ClinicalTrials.gov Menu

Rosiglitazone (Extended Release Tablets) As Adjunctive Therapy For Subjects With Mild To Moderate Alzheimer's Disease (REFLECT-2)

The safety and scientific validity of this study is the responsibility of the study sponsor and investigators. Listing a study does not mean it has been evaluated by the U.S. Federal Government. Read our disclaimer for details.
ClinicalTrials.gov Identifier: NCT00348309
Recruitment Status : Completed
First Posted : July 4, 2006
Results First Posted : November 28, 2017
Last Update Posted : November 28, 2017
Sponsor:
Information provided by (Responsible Party):
GlaxoSmithKline

Study Type: Interventional
Study Design: Allocation: Randomized;   Intervention Model: Parallel Assignment;   Masking: Quadruple (Participant, Care Provider, Investigator, Outcomes Assessor);   Primary Purpose: Treatment
Condition: Alzheimer's Disease
Interventions: Drug: Rosiglitazone Extended Release 2mg
Drug: Rosiglitazone Extended Release 8mg
Other: Placebo
Other: Donepezil

  Participant Flow

Recruitment Details
Key information relevant to the recruitment process for the overall study, such as dates of the recruitment period and locations
The study was conducted on participants with mild to moderate Alzheimer’s disease(AD), who received donepezil for at least 6 months and who received a stable dose of donepezil for at least 2 months immediately before study entry from 06 July 2006 to 28 January 2009 across 228 centers of 19 countries.

Pre-Assignment Details
Significant events and approaches for the overall study following participant enrollment, but prior to group assignment
A total of 1496 participants were randomized for the study, out of which seventeen participants did not receive study medication. A total of 1479 were included in the Safety population.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 milligram (mg) tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Participant Flow:   Overall Study
    Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release
STARTED   496   494   489 
COMPLETED   362   396   346 
NOT COMPLETED   134   98   143 
Adverse Event                56                28                63 
Lost to Follow-up                6                7                4 
Protocol Violation                7                12                7 
Withdrawal by Subject                33                37                40 
Non-compliance                8                6                8 
Death of caregiver                1                0                1 
Lack of insight, dementia                1                0                0 
Wrong bottle allocated                1                0                0 
Caregiver sick                1                0                0 
Use of prohibited medication                5                2                2 
Caregiver withdrawal                3                0                7 
Unable to complete ET visit                1                0                0 
PI closing medical practice                1                0                0 
Principal investigator's decision                1                0                0 
No disponibility of medication                1                0                0 
Worsened neurodegenerative disease                1                0                0 
Cognitive and behaviour worsening                2                0                0 
Worsening of alzheimer                1                0                1 
Did not appear in time for Visit 3                1                0                0 
Participant died                2                1                1 
Decision of sponsor                1                0                0 
Ineffectiveness of treatment                0                1                0 
A foul of exclusion criteria                0                1                0 
Participant was hospitalized                0                1                0 
Sponsor refuseal                0                1                1 
Forgot often medication                0                1                0 
Medical Monitor decision                0                0                1 
Increased dose of Aricept                0                0                1 
Site closure                0                0                1 
Adverse event not attended                0                0                1 
ECG abnormal                0                0                2 
Cardiologist recommendations                0                0                1 
HCV career                0                0                1 



  Baseline Characteristics

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
No text entered.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 milligram (mg) tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.
Total Total of all reporting groups

Baseline Measures
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release   Total 
Overall Participants Analyzed 
[Units: Participants]
 496   494   489   1479 
Age 
[Units: Years]
Mean (Standard Deviation)
 74.2  (7.95)   74.5  (8.06)   74.4  (7.83)   74.4  (7.94) 
Sex: Female, Male 
[Units: Participants]
Count of Participants
       
Female      304  61.3%      286  57.9%      305  62.4%      895  60.5% 
Male      192  38.7%      208  42.1%      184  37.6%      584  39.5% 
Race/Ethnicity, Customized 
[Units: Participants]
       
Hispanic or Latino   97   86   82   265 
Not Hispanic or Latino   399   408   407   1214 


  Outcome Measures

1.  Primary:   Change From Baseline in Alzheimer’s Disease Assessment Scale – Cognitive Subscale (ADAS-Cog) Total Score at Week 48   [ Time Frame: Baseline (Week 0) and Week 48 ]

Measure Type Primary
Measure Title Change From Baseline in Alzheimer’s Disease Assessment Scale – Cognitive Subscale (ADAS-Cog) Total Score at Week 48
Measure Description ADAS is a performance-based test that measures specific cognitive and behavioral dysfunctions in participants with Alzheimer's Disease. The cognitive subscale of the ADAS (ADAS-Cog) comprises 11 items that are summed to a total score ranging from 0 to 70, with lower scores indicating less severe impairment. Change from baseline is calculated as Week 48 value minus the baseline value. APOE4 negative, All except E4/E4’s: comprised of APOE4 negative and E4 heterozygote and full population was analyzed for this outcome measure. A hierarchical testing procedure was used to control for the two rosiglitazone dose groups and the genetic subgroups. Least square mean is entered for adjusted mean.
Time Frame Baseline (Week 0) and Week 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population included all the participants who were randomized to treatment, who had received at least one dose of study medication and who had at least one post baseline efficacy assessment. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 milligram (mg) tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   461   473   459 
Change From Baseline in Alzheimer’s Disease Assessment Scale – Cognitive Subscale (ADAS-Cog) Total Score at Week 48 
[Units: Score on a scale]
Least Squares Mean (Standard Error)
     
APOE4 negatives       
Participants Analyzed   153   172   141 
APOE4 negatives   2.9  (0.54)   1.6  (0.42)   2.7  (0.56) 
All except E4/E4s       
Participants Analyzed   317   350   285 
All except E4/E4s   3.1  (0.36)   2.1  (0.29)   3.1  (0.37) 
Full populations       
Participants Analyzed   356   391   333 
Full populations   3.4  (0.35)   2.4  (0.30)   3.2  (0.35) 


Statistical Analysis 1 for Change From Baseline in Alzheimer’s Disease Assessment Scale – Cognitive Subscale (ADAS-Cog) Total Score at Week 48
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.049
Mean Difference (Final Values) [5] -1.3
95% Confidence Interval -2.7 to -0.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  APOE4 negatives
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change From Baseline in Alzheimer’s Disease Assessment Scale – Cognitive Subscale (ADAS-Cog) Total Score at Week 48
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.808
Mean Difference (Final Values) [5] -0.2
95% Confidence Interval -1.7 to 1.3
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  APOE4 negatives
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 3 for Change From Baseline in Alzheimer’s Disease Assessment Scale – Cognitive Subscale (ADAS-Cog) Total Score at Week 48
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.035
Mean Difference (Final Values) [5] -1.0
95% Confidence Interval -1.9 to -0.1
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  All except E4/E4s
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 4 for Change From Baseline in Alzheimer’s Disease Assessment Scale – Cognitive Subscale (ADAS-Cog) Total Score at Week 48
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.999
Mean Difference (Final Values) [5] 0.0
95% Confidence Interval -1.0 to 1.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  All except E4/E4s
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 5 for Change From Baseline in Alzheimer’s Disease Assessment Scale – Cognitive Subscale (ADAS-Cog) Total Score at Week 48
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.020
Mean Difference (Final Values) [5] -1.0
95% Confidence Interval -1.9 to -0.2
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Full population
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 6 for Change From Baseline in Alzheimer’s Disease Assessment Scale – Cognitive Subscale (ADAS-Cog) Total Score at Week 48
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.661
Mean Difference (Final Values) [5] -0.2
95% Confidence Interval -1.2 to 0.7
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Full population
[5] Other relevant estimation information:
  No text entered.



2.  Primary:   Change From Baseline in Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB) at Week 48 for APOE E4   [ Time Frame: Baseline (Week 0) and Week 48 ]

Measure Type Primary
Measure Title Change From Baseline in Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB) at Week 48 for APOE E4
Measure Description CDR-SB is a semi-structured interview of participants and their caregivers. Participant’s cognitive status is rated across 6 domains of functioning, including memory, orientation, judgment/problem solving, community affairs, home/hobbies, and personal care. Severity score assigned for each of 6 domains; Total score (SB) ranges from 0 to 18. Higher scores indicate greater disease severity. Change from baseline is calculated as Week 48 value minus the baseline value. APOE4 negative, All except E4/E4’s: comprised of APOE4 negative and E4 heterozygote and full population was analyzed for this outcome measure. A hierarchical testing procedure was used to control for the two rosiglitazone dose groups and the genetic subgroups.
Time Frame Baseline (Week 0) and Week 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   461   473   459 
Change From Baseline in Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB) at Week 48 for APOE E4 
[Units: Score on a scale]
Least Squares Mean (Standard Error)
     
APOE4 negatives       
Participants Analyzed   146   169   139 
APOE4 negatives   1.3  (0.21)   0.8  (0.16)   1.5  (0.20) 
All except E4/E4s       
Participants Analyzed   309   343   283 
All except E4/E4s   1.5  (0.14)   1.0  (0.12)   1.7  (0.14) 
Full population       
Participants Analyzed   347   384   331 
Full population   1.6  (0.14)   1.0  (0.12)   1.7  (0.13) 


Statistical Analysis 1 for Change From Baseline in Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB) at Week 48 for APOE E4
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.056
Mean Difference (Final Values) [5] -0.5
95% Confidence Interval -1.0 to 0.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  APOE4 negatives
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change From Baseline in Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB) at Week 48 for APOE E4
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.587
Mean Difference (Final Values) [5] 0.2
95% Confidence Interval -0.4 to 0.7
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  APOE4 negatives
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 3 for Change From Baseline in Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB) at Week 48 for APOE E4
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.004
Mean Difference (Final Values) [5] -0.5
95% Confidence Interval -0.9 to -0.2
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  All except E4/E4s
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 4 for Change From Baseline in Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB) at Week 48 for APOE E4
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.402
Mean Difference (Final Values) [5] 0.2
95% Confidence Interval -0.2 to 0.6
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  All except E4/E4s
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 5 for Change From Baseline in Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB) at Week 48 for APOE E4
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.002
Mean Difference (Final Values) [5] -0.5
95% Confidence Interval -0.9 to -0.2
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Full population
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 6 for Change From Baseline in Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB) at Week 48 for APOE E4
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.478
Mean Difference (Final Values) [5] 0.1
95% Confidence Interval -0.2 to 0.5
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Full population
[5] Other relevant estimation information:
  No text entered.



3.  Secondary:   Change From Baseline in Disability Assessment for Dementia Scale (DAD) Total Score   [ Time Frame: Baseline (Week 0), Week 8, 16, 24 and 48 ]

Measure Type Secondary
Measure Title Change From Baseline in Disability Assessment for Dementia Scale (DAD) Total Score
Measure Description The DAD measures instrumental and basic activities of daily living in participants with Alzheimer's Disease (AD). This scale assesses a participants’ ability to initiate, plan, and perform activities related to hygiene, dressing, continence, eating, meal preparation, telephoning, going on an outing, finance and correspondence, medications, leisure, and housework. The scale includes 23 items relating to instrumental activities of daily living and 17 items relating to basic self-care. Each item can be scored as 1 = yes, 0 = no, non applicable = NA. Total score was obtained by adding the rating for each question and converting this total score out of 100. The total score ranged from 0 to 100, where higher score indicated better function and lower score indicated greater severity of symptoms; a positive change from baseline indicated an improvement. Change from baseline is calculated as endpoint value minus the baseline value.
Time Frame Baseline (Week 0), Week 8, 16, 24 and 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   436   458   441 
Change From Baseline in Disability Assessment for Dementia Scale (DAD) Total Score 
[Units: Scores on a scale]
Least Squares Mean (Standard Error)
     
At Week 8   0.1  (0.56)   -0.5  (0.48)   -1.1  (0.47) 
At Week 16   -1.6  (0.62)   -1.6  (0.59)   -2.1  (0.56) 
At Week 24   -3.5  (0.69)   -2.2  (0.64)   -3.3  (0.69) 
At Week 48   -7.8  (0.82)   -5.7  (0.78)   -8.4  (0.84) 


Statistical Analysis 1 for Change From Baseline in Disability Assessment for Dementia Scale (DAD) Total Score
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.340
Mean Difference (Final Values) [5] -0.6
95% Confidence Interval -2.0 to 0.7
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 8
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change From Baseline in Disability Assessment for Dementia Scale (DAD) Total Score
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.079
Mean Difference (Final Values) [5] -1.2
95% Confidence Interval -2.5 to 0.1
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 8
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 3 for Change From Baseline in Disability Assessment for Dementia Scale (DAD) Total Score
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.910
Mean Difference (Final Values) [5] -0.1
95% Confidence Interval -1.7 to 1.5
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 16
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 4 for Change From Baseline in Disability Assessment for Dementia Scale (DAD) Total Score
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.521
Mean Difference (Final Values) [5] -0.5
95% Confidence Interval -2.1 to 1.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 16
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 5 for Change From Baseline in Disability Assessment for Dementia Scale (DAD) Total Score
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.154
Mean Difference (Final Values) [5] 1.3
95% Confidence Interval -0.5 to 3.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 24
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 6 for Change From Baseline in Disability Assessment for Dementia Scale (DAD) Total Score
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.836
Mean Difference (Final Values) [5] 0.2
95% Confidence Interval -1.6 to 2.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 24
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 7 for Change From Baseline in Disability Assessment for Dementia Scale (DAD) Total Score
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.061
Mean Difference (Final Values) [5] 2.1
95% Confidence Interval -0.1 to 4.2
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 8 for Change From Baseline in Disability Assessment for Dementia Scale (DAD) Total Score
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.566
Mean Difference (Final Values) [5] -0.7
95% Confidence Interval -2.9 to 1.6
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48
[5] Other relevant estimation information:
  No text entered.



4.  Secondary:   Change From Baseline in Neuropsychiatric Inventory (NPI) Total Score   [ Time Frame: Baseline (Week 0), Week 8, 16, 24 and 48 ]

Measure Type Secondary
Measure Title Change From Baseline in Neuropsychiatric Inventory (NPI) Total Score
Measure Description The NPI is a questionnaire that quantifies behavioral changes in dementia. For each of 12 behavioral domains there are 4 scores: Frequency (scale:1=occasionally to 4=very frequently), Severity (scale:1=Mild to 3=Severe), Total (frequency x severity), Caregiver distress (scale: 0=not at all distressing to 5=extremely distressing).The NPI Psychosis Subscale consists of the two domains of Delusions and Hallucinations, calculated by adding the Individual Item Scores, to yield a possible total score of 0 to 24. Lower score=less severity. Change from baseline is calculated as endpoint value minus the baseline value.
Time Frame Baseline (Week 0), Week 8, 16, 24 and 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   461   473   459 
Change From Baseline in Neuropsychiatric Inventory (NPI) Total Score 
[Units: Score on a scale]
Least Squares Mean (Standard Error)
     
At Week 8       
Participants Analyzed   440   458   442 
At Week 8   -0.3  (0.39)   -0.7  (0.32)   -0.3  (0.39) 
At Week 16       
Participants Analyzed   425   442   415 
At Week 16   -0.3  (0.43)   -0.6  (0.36)   0.1  (0.43) 
At Week 24       
Participants Analyzed   411   425   390 
At Week 24   0.4  (0.47)   -0.2  (0.39)   0.2  (0.48) 
At Week 48       
Participants Analyzed   359   395   345 
At Week 48   1.6  (0.61)   0.1  (0.44)   1.8  (0.55) 


Statistical Analysis 1 for Change From Baseline in Neuropsychiatric Inventory (NPI) Total Score
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] =0.414
Mean Difference (Final Values) [5] -0.4
95% Confidence Interval -1.3 to 0.5
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 8
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change From Baseline in Neuropsychiatric Inventory (NPI) Total Score
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.931
Mean Difference (Final Values) [5] 0.0
95% Confidence Interval -1.0 to 1.1
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 8
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 3 for Change From Baseline in Neuropsychiatric Inventory (NPI) Total Score
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.485
Mean Difference (Final Values) [5] -0.4
95% Confidence Interval -1.4 to 0.7
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 16
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 4 for Change From Baseline in Neuropsychiatric Inventory (NPI) Total Score
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.550
Mean Difference (Final Values) [5] 0.3
95% Confidence Interval -0.8 to 1.5
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 16
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 5 for Change From Baseline in Neuropsychiatric Inventory (NPI) Total Score
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.264
Mean Difference (Final Values) [5] -0.6
95% Confidence Interval -1.8 to 0.5
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 24
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 6 for Change From Baseline in Neuropsychiatric Inventory (NPI) Total Score
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.732
Mean Difference (Final Values) [5] -0.2
95% Confidence Interval -1.5 to 1.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 24
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 7 for Change From Baseline in Neuropsychiatric Inventory (NPI) Total Score
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.043
Mean Difference (Final Values) [5] -1.5
95% Confidence Interval -2.9 to -0.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 8 for Change From Baseline in Neuropsychiatric Inventory (NPI) Total Score
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.814
Mean Difference (Final Values) [5] 0.2
95% Confidence Interval -1.4 to 1.8
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48
[5] Other relevant estimation information:
  No text entered.



5.  Secondary:   Change From Screening in Mini Mental State Examination (MMSE) Total Score   [ Time Frame: Screening (Week -4) and Week 48 ]

Measure Type Secondary
Measure Title Change From Screening in Mini Mental State Examination (MMSE) Total Score
Measure Description The MMSE consists of 11 tests of orientation, memory (recent and immediate), concentration, language and praxis. The scale was completed by the investigator, based on the performance of the participant, and took approximately 5 to 10 minutes to administer. The scores from 11 tests were combined to obtain the total score. The total scores range from 0 to 30, with lower scores indicating greater cognitive impairment and higher score indicating better outcome; a positive change from screening indicated an improvement. The total MMSE score for participants at screening was between 10 and 26, inclusive, in order to be eligible to participate in the trial. Change from screening is calculated as endpoint value minus the screening value.
Time Frame Screening (Week -4) and Week 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   348   388   335 
Change From Screening in Mini Mental State Examination (MMSE) Total Score 
[Units: Score on a scale]
Least Squares Mean (Standard Error)
 -1.6  (0.21)   -1.6  (0.20)   -1.7  (0.21) 


Statistical Analysis 1 for Change From Screening in Mini Mental State Examination (MMSE) Total Score
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.870
Mean Difference (Final Values) [5] 0.0
95% Confidence Interval -0.5 to 0.6
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change From Screening in Mini Mental State Examination (MMSE) Total Score
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.617
Mean Difference (Final Values) [5] -0.1
95% Confidence Interval -0.7 to 0.4
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.



6.  Secondary:   Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)   [ Time Frame: Baseline (Week 0), Week 12, 24, 36 and 48 ]

Measure Type Secondary
Measure Title Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Measure Description The RUD instrument was developed as a comprehensive tool to assess the amount of resource use among demented patients. RUD assessd both formal and informal resource use of the patient and the primary caregiver, making it possible to calculate costs from a societal perspective. Q1 corresponds to the number of hours during the last month the caregiver spent assisting the patient with toilet visits, eating, dressing, grooming, walking and bathing and Q2 corresponds to the number of hours during the last month the caregiver spent assisting the patient with shopping, food preparation, housekeeping, laundry, transportation, taking medication and managing financial matters. Change from Baseline was calculated as value at scheduled time point minus Baseline value. Baseline was defined as value at Week 0. Estimated value was calculated by Active treatment minus Placebo. The adjusted means were presented.
Time Frame Baseline (Week 0), Week 12, 24, 36 and 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   461   473   459 
Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD) 
[Units: Hours]
Least Squares Mean (Standard Error)
     
Q1: Week 12       
Participants Analyzed   386   408   379 
Q1: Week 12   1.6  (2.25)   2.4  (2.50)   -0.1  (3.18) 
Q1: Week 24       
Participants Analyzed   386   380   347 
Q1: Week 24   10.7  (3.52)   3.4  (1.84)   7.0  (4.06) 
Q1: Week 36       
Participants Analyzed   342   361   327 
Q1: Week 36   16.3  (3.60)   10.0  (3.43)   15.2  (4.95) 
Q1: Week 48       
Participants Analyzed   320   351   310 
Q1: Week 48   21.7  (4.16)   17.0  (4.17)   18.1  (4.72) 
Q2: Week 12       
Participants Analyzed   171   183   162 
Q2: Week 12   -6.4  (4.64)   2.4  (4.16)   1.5  (4.50) 
Q2: Week 24       
Participants Analyzed   162   170   150 
Q2: Week 24   0.3  (4.89)   1.6  (4.17)   6.5  (4.88) 
Q2: Week 36       
Participants Analyzed   153   162   141 
Q2: Week 36   5.0  (5.34)   4.6  (4.44)   13.3  (5.35) 
Q2: Week 48       
Participants Analyzed   139   157   137 
Q2: Week 48   10.8  (5.51)   8.4  (4.88)   27.0  (6.24) 


Statistical Analysis 1 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.789
Mean Difference (Final Values) [5] 0.8
95% Confidence Interval -5.2 to 6.9
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  Week 12 Q1
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.643
Mean Difference (Final Values) [5] -1.7
95% Confidence Interval -9.0 to 5.6
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 12Q1
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 3 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.052
Mean Difference (Final Values) [5] -7.3
95% Confidence Interval -14.7 to 0.1
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 24 Q1
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 4 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.481
Mean Difference (Final Values) [5] -3.7
95% Confidence Interval -14.0 to 6.6
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 24 Q1
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 5 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.195
Mean Difference (Final Values) [5] -6.2
95% Confidence Interval -15.6 to 3.2
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 36 Q1
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 6 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.857
Mean Difference (Final Values) [5] -1.1
95% Confidence Interval -12.9 to 10.7
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 36 Q1
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 7 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.422
Mean Difference (Final Values) [5] -4.6
95% Confidence Interval -15.9 to 6.7
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48 Q1
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 8 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.562
Mean Difference (Final Values) [5] -3.6
95% Confidence Interval -15.7 to 8.6
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48 Q1
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 9 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.129
Mean Difference (Final Values) [5] 8.8
95% Confidence Interval -2.6 to 20.2
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 12 Q2
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 10 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.194
Mean Difference (Final Values) [5] 7.9
95% Confidence Interval -4.0 to 19.9
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 12 Q2
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 11 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.832
Mean Difference (Final Values) [5] 1.3
95% Confidence Interval -10.5 to 13.1
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 24 Q2
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 12 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.351
Mean Difference (Final Values) [5] 6.1
95% Confidence Interval -6.8 to 19.1
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 24 Q2
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 13 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.949
Mean Difference (Final Values) [5] -0.4
95% Confidence Interval -13.3 to 12.5
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 36 Q2
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 14 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.253
Mean Difference (Final Values) [5] 8.3
95% Confidence Interval -6.0 to 22.6
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 36 Q2
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 15 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.735
Mean Difference (Final Values) [5] -2.4
95% Confidence Interval -16.2 to 11.4
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48 Q2
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 16 for Change From Baseline in the Domains of the Resource Utilization in Dementia Scale (RUD)
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.046
Mean Difference (Final Values) [5] 16.2
95% Confidence Interval 0.3 to 32.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48 Q2
[5] Other relevant estimation information:
  No text entered.



7.  Secondary:   Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility   [ Time Frame: Baseline (Week 0), Week 12, 36 and 48 ]

Measure Type Secondary
Measure Title Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Measure Description The EQ-5D Proxy is a two part scale that evaluated the participant’s health status via Thermometer and Utility scores. The Thermometer score was the caregiver’s rating of the participant’s overall health status on a VAS (0 [“worst possible status”] to 100 [“best imaginable status”]). The Utility score was a caregiver rating of health status on dimensions of mobility, self-care, usual activities, pain/discomfort, and anxiety/depression] where ‘1’ indicated better health state (no problems); ‘3’ indicated worst health state ("confined to bed"). Total possible score was the sum of individual items, ranged from 5 to 15; lower score indicated a better health state and higher score indicated greater severity of symptoms. A positive change from baseline indicated improvement in the Thermometer score and a negative change from baseline indicated improvement in the Utility score. Change from baseline is calculated as endpoint value minus the baseline value.
Time Frame Baseline (Week 0), Week 12, 36 and 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   461   473   459 
Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility 
[Units: Score on a scale]
Least Squares Mean (Standard Error)
     
Thermometer: Week 12       
Participants Analyzed   425   449   413 
Thermometer: Week 12   0.5  (0.84)   0.3  (0.82)   -0.5  (0.94) 
Thermometer: Week 36       
Participants Analyzed   381   404   362 
Thermometer: Week 36   0.8  (0.95)   -1.6  (0.97)   -2.1  (1.03) 
Thermometer: Week 48       
Participants Analyzed   352   389   337 
Thermometer: Week 48   -1.5  (1.05)   -0.3  (0.96)   -2.4  (1.08) 
Utility: Week 12       
Participants Analyzed   425   448   414 
Utility: Week 12   0.01  (0.009)   0.02  (0.009)   -0.01  (0.010) 
Utility: Week 36       
Participants Analyzed   381   403   363 
Utility: Week 36   -0.02  (0.011)   -0.01  (0.010)   -0.04  (0.012) 
Utility: Week 48       
Participants Analyzed   351   389   337 
Utility: Week 48   -0.04  (0.012)   -0.02  (0.011)   -0.05  (0.013) 


Statistical Analysis 1 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.914
Mean Difference (Final Values) [5] -0.1
95% Confidence Interval -2.3 to 2.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 12 Thermometer
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.415
Mean Difference (Final Values) [5] -1.0
95% Confidence Interval -3.3 to 1.4
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 12 Thermometer
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 3 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.063
Mean Difference (Final Values) [5] -2.4
95% Confidence Interval -4.9 to 0.1
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 36 Thermometer
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 4 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.030
Mean Difference (Final Values) [5] -2.9
95% Confidence Interval -5.6 to -0.3
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 36 Thermometer
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 5 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.392
Mean Difference (Final Values) [5] 1.2
95% Confidence Interval -1.5 to 3.8
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48 Thermometer
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 6 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.557
Mean Difference (Final Values) [5] -0.9
95% Confidence Interval -3.7 to 2.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48 Thermometer
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 7 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.320
Mean Difference (Final Values) [5] 0.01
95% Confidence Interval -0.01 to 0.03
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 12 Utility
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 8 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.180
Mean Difference (Final Values) [5] -0.02
95% Confidence Interval -0.04 to 0.01
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 12 Utility
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 9 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.498
Mean Difference (Final Values) [5] 0.01
95% Confidence Interval -0.02 to 0.04
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 36 Utility
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 10 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.081
Mean Difference (Final Values) [5] -0.03
95% Confidence Interval -0.06 to 0.00
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 36 Utility
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 11 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.077
Mean Difference (Final Values) [5] 0.03
95% Confidence Interval -0.00 to 0.06
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48 Utility
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 12 for Change From Baseline in European Quality of Life-5 Dimensions Proxy Version (EQ-5D Proxy) Scale Total Score Assessed by Thermometer (Visual Analog Scale [VAS]) and Utility
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] Mixed Models Analysis
P Value [4] 0.640
Mean Difference (Final Values) [5] -0.01
95% Confidence Interval -0.04 to 0.03
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  Week 48 Utility
[5] Other relevant estimation information:
  No text entered.



8.  Secondary:   Change From Baseline in ADAS-Cog Total Score for Observed Cases at Weeks 8, 16, 24, 36 and 48   [ Time Frame: Baseline (Week 0), Week 8, 16, 24, 36 and 48 ]

Measure Type Secondary
Measure Title Change From Baseline in ADAS-Cog Total Score for Observed Cases at Weeks 8, 16, 24, 36 and 48
Measure Description ADAS is a performance-based test that measures specific cognitive and behavioral dysfunctions in patients with Alzheimer's Disease. The cognitive subscale of the ADAS (ADAS-Cog) comprises 11 items that are summed to a total score ranging from 0 to 70, with lower scores indicating less severe impairment. Change from baseline is calculated as endpoint value minus the baseline value.
Time Frame Baseline (Week 0), Week 8, 16, 24, 36 and 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   461   473   459 
Change From Baseline in ADAS-Cog Total Score for Observed Cases at Weeks 8, 16, 24, 36 and 48 
[Units: Score on a scale]
Mean (Standard Deviation)
     
Week 8       
Participants Analyzed   435   457   437 
Week 8   -0.2  (4.21)   -0.6  (4.57)   -0.2  (4.44) 
Week 16       
Participants Analyzed   422   441   411 
Week 16   -0.1  (5.09)   -0.5  (4.76)   0.1  (5.13) 
Week 24       
Participants Analyzed   404   425   386 
Week 24   1.0  (5.86)   -0.2  (5.17)   0.8  (5.66) 
Week 36       
Participants Analyzed   378   402   362 
Week 36   1.8  (6.08)   1.3  (5.79)   2.2  (6.40) 
Week 48       
Participants Analyzed   358   393   333 
Week 48   2.9  (6.85)   2.1  (6.25)   2.6  (6.76) 


Statistical Analysis 1 for Change From Baseline in ADAS-Cog Total Score for Observed Cases at Weeks 8, 16, 24, 36 and 48
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.105
Mean Difference (Final Values) [5] -0.7
95% Confidence Interval -1.6 to 0.2
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change From Baseline in ADAS-Cog Total Score for Observed Cases at Weeks 8, 16, 24, 36 and 48
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.483
Mean Difference (Final Values) [5] -0.3
95% Confidence Interval -1.3 to 0.6
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.



9.  Secondary:   Change From Baseline in CDR-SB Score for Observed Cases at Weeks 12, 24, 36 and 48   [ Time Frame: Baseline (Week 0), Week 12, 24, 36 and 48 ]

Measure Type Secondary
Measure Title Change From Baseline in CDR-SB Score for Observed Cases at Weeks 12, 24, 36 and 48
Measure Description CDR-SB is a semi-structured interview of participants and their caregivers. Participant’s cognitive status is rated across 6 domains of functioning, including memory, orientation, judgment/problem solving, community affairs, home/hobbies, and personal care. Severity score assigned for each of 6 domains; Total score (SB) ranges from 0 to 18. Higher scores indicate greater disease severity. Change from baseline is calculated as endpoint value minus the baseline value.
Time Frame Baseline (Week 0), Week 12, 24, 36 and 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   461   473   459 
Change From Baseline in CDR-SB Score for Observed Cases at Weeks 12, 24, 36 and 48 
[Units: Score on a scale]
Mean (Standard Deviation)
     
Week 12       
Participants Analyzed   423   446   414 
Week 12   0.4  (1.55)   0.3  (1.36)   0.3  (1.40) 
Week 24       
Participants Analyzed   402   421   382 
Week 24   0.7  (2.01)   0.5  (1.67)   0.8  (1.83) 
Week 36       
Participants Analyzed   374   400   357 
Week 36   1.0  (2.26)   0.7  (2.06)   1.2  (2.30) 
Week 48       
Participants Analyzed   349   385   331 
Week 48   1.5  (2.68)   1.0  (2.28)   1.6  (2.56) 


Statistical Analysis 1 for Change From Baseline in CDR-SB Score for Observed Cases at Weeks 12, 24, 36 and 48
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.004
Mean Difference (Final Values) [5] -0.5
95% Confidence Interval -0.9 to -0.2
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change From Baseline in CDR-SB Score for Observed Cases at Weeks 12, 24, 36 and 48
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.554
Mean Difference (Final Values) [5] 0.1
95% Confidence Interval -0.3 to 0.5
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.



10.  Secondary:   Change in ADAS-Cog Total Score for Observed Cases at Week 54 Compared to Week 48   [ Time Frame: Week 48 and 54 ]

Measure Type Secondary
Measure Title Change in ADAS-Cog Total Score for Observed Cases at Week 54 Compared to Week 48
Measure Description ADAS is a performance-based test that measures specific cognitive and behavioral dysfunctions in patients with Alzheimer's Disease. The cognitive subscale of the ADAS (ADAS-Cog) comprises 11 items that are summed to a total score ranging from 0 to 70, with lower scores indicating less severe impairment. Change was calculated as endpoint value (Week 54) minus Week 48 value.
Time Frame Week 48 and 54  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   332   366   305 
Change in ADAS-Cog Total Score for Observed Cases at Week 54 Compared to Week 48 
[Units: Score on a scale]
Least Squares Mean (Standard Error)
 0.7  (0.28)   1.1  (0.27)   1.1  (0.29) 


Statistical Analysis 1 for Change in ADAS-Cog Total Score for Observed Cases at Week 54 Compared to Week 48
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.292
Mean Difference (Final Values) [5] 0.4
95% Confidence Interval -0.3 to 1.1
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change in ADAS-Cog Total Score for Observed Cases at Week 54 Compared to Week 48
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.297
Mean Difference (Final Values) [5] 0.4
95% Confidence Interval -0.3 to 1.1
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.



11.  Secondary:   Change in CDR-SB Total Score at Week 54 Compared to Week 48   [ Time Frame: Week 48 and 54 ]

Measure Type Secondary
Measure Title Change in CDR-SB Total Score at Week 54 Compared to Week 48
Measure Description CDR-SB is a semi-structured interview of participants and their caregivers. Participant’s cognitive status is rated across 6 domains of functioning, including memory, orientation, judgment/problem solving, community affairs, home/hobbies, and personal care. Severity score assigned for each of 6 domains; Total score (SB) ranges from 0 to 18. Higher scores indicate greater disease severity. Change was calculated as endpoint value (Week 54) minus Week 48 value.
Time Frame Week 48 and 54  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   328   364   302 
Change in CDR-SB Total Score at Week 54 Compared to Week 48 
[Units: Score on a scale]
Least Squares Mean (Standard Error)
 0.2  (0.07)   0.2  (0.07)   0.1  (0.07) 


Statistical Analysis 1 for Change in CDR-SB Total Score at Week 54 Compared to Week 48
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.598
Mean Difference (Final Values) [5] 0.0
95% Confidence Interval -0.2 to 0.1
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change in CDR-SB Total Score at Week 54 Compared to Week 48
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.073
Mean Difference (Final Values) [5] -0.2
95% Confidence Interval -0.4 to 0.0
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.



12.  Secondary:   Change From Baseline in Glycosylated Hemoglobin (HbA1c) at Week 48   [ Time Frame: Baseline (Week 0) and Week 48 ]

Measure Type Secondary
Measure Title Change From Baseline in Glycosylated Hemoglobin (HbA1c) at Week 48
Measure Description Blood samples of participants were collected for HbA1c assessment. HbA1c is a form of hemoglobin that is measured primarily to identify the average plasma glucose concentration over a 2- to 3-month period. Change from Baseline in HbA1c was calculated as the value at Week 48 minus the value at Baseline.
Time Frame Baseline (Week 0) and Week 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   321   352   313 
Change From Baseline in Glycosylated Hemoglobin (HbA1c) at Week 48 
[Units: Percentage of total hemoglobin]
Least Squares Mean (Standard Error)
 0.14  (0.020)   0.21  (0.020)   0.18  (0.020) 


Statistical Analysis 1 for Change From Baseline in Glycosylated Hemoglobin (HbA1c) at Week 48
Groups [1] Placebo vs. 2mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.006
Mean Difference (Final Values) [5] 0.07
95% Confidence Interval 0.02 to 0.12
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.

Statistical Analysis 2 for Change From Baseline in Glycosylated Hemoglobin (HbA1c) at Week 48
Groups [1] Placebo vs. 8mg Rosiglitazone Extended Release
Statistical Test Type [2] Superiority or Other
Statistical Method [3] ANCOVA
P Value [4] 0.140
Mean Difference (Final Values) [5] 0.04
95% Confidence Interval -0.01 to 0.09
[1] Additional details about the analysis, such as null hypothesis and power calculation:
  No text entered.
[2] Details of power calculation, definition of non-inferiority margin, and other key parameters:
  No text entered.
[3] Other relevant method information, such as adjustments or degrees of freedom:
  No text entered.
[4] Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
  No text entered.
[5] Other relevant estimation information:
  No text entered.



13.  Secondary:   Number of Participants With Treatment Emergent Adverse Events (AEs) and Serious Adverse Events (SAEs)   [ Time Frame: Up to Week 54 ]

Measure Type Secondary
Measure Title Number of Participants With Treatment Emergent Adverse Events (AEs) and Serious Adverse Events (SAEs)
Measure Description An adverse event (AE) was defined as any untoward medical occurrence that occurred during the course of the trial after study treatment had started. An adverse event was therefore any unfavourable and unintended sign, symptom, or disease temporally associated with the use of study drug, whether or not considered related to the study drug. A Serious Adverse Event (SAE) is any untoward medical occurrence that at any dose results in death, are life threatening, requires hospitalization or prolongation of hospitalization or results in disability/incapacity, and congenital anomaly/birth defect. The data was reported for prospective period.
Time Frame Up to Week 54  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
The safety population included all participants randomized to treatment and received at least one dose of study medication.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   496   494   489 
Number of Participants With Treatment Emergent Adverse Events (AEs) and Serious Adverse Events (SAEs) 
[Units: Participants]
     
Any TEAEs   304   273   327 
Any SAEs   62   45   50 

No statistical analysis provided for Number of Participants With Treatment Emergent Adverse Events (AEs) and Serious Adverse Events (SAEs)



14.  Secondary:   Mean Change From Baseline in Systolic and Diastolic Blood Pressure (BP)   [ Time Frame: Baseline (Week 0), Week 4, 8, 12, 16, 24, 36, 48 and 56 ]

Measure Type Secondary
Measure Title Mean Change From Baseline in Systolic and Diastolic Blood Pressure (BP)
Measure Description The plethysmographic method was used to measure BP throughout the study. Change in Systolic and Diastolic BP was calculated as endpoint value minus the baseline value.
Time Frame Baseline (Week 0), Week 4, 8, 12, 16, 24, 36, 48 and 56  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
Safety population. Only those participants available at the specified time points were analyzed. Data for Site 040449 was not included in analysis due to audit finding.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   477   481   470 
Mean Change From Baseline in Systolic and Diastolic Blood Pressure (BP) 
[Units: Millimeter of mercury (mmHg)]
Mean (Standard Deviation)
     
Systolic BP Week 4       
Participants Analyzed   457   469   457 
Systolic BP Week 4   -0.7  (15.21)   -1.0  (13.21)   -2.3  (13.43) 
Systolic BP Week 8       
Participants Analyzed   436   460   434 
Systolic BP Week 8   -0.3  (15.49)   -0.7  (14.55)   -2.9  (14.49) 
Systolic BP Week 12       
Participants Analyzed   431   454   415 
Systolic BP Week 12   -0.5  (16.48)   -2.5  (14.49)   -4.2  (15.25) 
Systolic BP Week 16       
Participants Analyzed   425   442   411 
Systolic BP Week 16   -0.8  (15.80)   -1.7  (14.80)   -3.7  (15.75) 
Systolic BP Week 24       
Participants Analyzed   413   431   392 
Systolic BP Week 24   -1.2  (15.95)   -1.0  (15.22)   -3.5  (14.74) 
Systolic BP Week 36       
Participants Analyzed   384   406   367 
Systolic BP Week 36   -1.6  (16.65)   -0.3  (15.79)   -2.7  (15.97) 
Systolic BP Week 48       
Participants Analyzed   363   397   345 
Systolic BP Week 48   -1.0  (17.08)   -0.6  (15.05)   -2.5  (16.55) 
Systolic BP Week 54       
Participants Analyzed   345   385   321 
Systolic BP Week 54   0.0  (17.24)   -1.1  (14.95)   -2.3  (15.56) 
Diastolic BP Week 4       
Participants Analyzed   457   469   457 
Diastolic BP Week 4   -0.4  (9.54)   -0.7  (9.21)   -1.2  (8.51) 
Diastolic BP Week 8       
Participants Analyzed   436   460   434 
Diastolic BP Week 8   -0.0  (9.55)   -0.9  (9.20)   -2.1  (9.21) 
Diastolic BP Week 12       
Participants Analyzed   431   454   415 
Diastolic BP Week 12   0.0  (9.63)   -1.3  (9.25)   -2.5  (9.49) 
Diastolic BP Week 16       
Participants Analyzed   425   442   411 
Diastolic BP Week 16   -0.7  (9.78)   -1.3  (9.16)   -3.0  (9.62) 
Diastolic BP Week 24       
Participants Analyzed   413   431   392 
Diastolic BP Week 24   -1.1  (9.87)   -1.2  (9.55)   -2.6  (9.48) 
Diastolic BP Week 36       
Participants Analyzed   384   406   367 
Diastolic BP Week 36   -1.2  (10.27)   -0.6  (9.94)   -3.3  (10.68) 
Diastolic BP Week 48       
Participants Analyzed   363   397   345 
Diastolic BP Week 48   -0.8  (10.85)   -0.5  (10.02)   -1.9  (11.07) 
Diastolic BP Week 54       
Participants Analyzed   345   385   321 
Diastolic BP Week 54   -0.9  (10.35)   -0.4  (9.62)   -1.8  (10.04) 

No statistical analysis provided for Mean Change From Baseline in Systolic and Diastolic Blood Pressure (BP)



15.  Secondary:   Mean Change From Baseline in Heart Rate   [ Time Frame: Baseline (Week 0), Week 4, 8, 12, 16, 24, 36, 48 and 56 ]

Measure Type Secondary
Measure Title Mean Change From Baseline in Heart Rate
Measure Description Mean Change From Baseline in heart rate was calculated as endpoint value minus the baseline value.
Time Frame Baseline (Week 0), Week 4, 8, 12, 16, 24, 36, 48 and 56  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
Safety population. Only those participants available at the specified time points were analyzed. Data for Site 040449 was not included in analysis due to audit finding.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   477   481   470 
Mean Change From Baseline in Heart Rate 
[Units: Beats per min (bpm)]
Mean (Standard Deviation)
     
Week 4       
Participants Analyzed   457   468   457 
Week 4   0.8  (8.42)   1.1  (7.95)   0.6  (8.36) 
Week 8       
Participants Analyzed   435   459   434 
Week 8   1.5  (8.98)   1.3  (9.19)   1.0  (8.91) 
Week 12       
Participants Analyzed   430   453   414 
Week 12   1.6  (9.06)   1.8  (9.04)   0.7  (9.68) 
Week 16       
Participants Analyzed   425   441   411 
Week 16   1.0  (9.23)   1.3  (9.18)   0.8  (9.52) 
Week 24       
Participants Analyzed   413   431   392 
Week 24   0.9  (9.41)   1.3  (8.70)   0.8  (10.19) 
Week 36       
Participants Analyzed   382   406   367 
Week 36   0.9  (9.30)   1.8  (9.42)   1.1  (9.58) 
Week 48       
Participants Analyzed   361   397   344 
Week 48   1.2  (9.10)   1.4  (8.92)   0.7  (9.43) 
Week 54       
Participants Analyzed   344   384   321 
Week 54   0.5  (10.10)   1.4  (9.55)   0.0  (11.05) 

No statistical analysis provided for Mean Change From Baseline in Heart Rate



16.  Secondary:   Mean Change From Baseline in Weight   [ Time Frame: Baseline (Week 0), Week 4, 8, 12, 16, 24, 36, 48 and 56 ]

Measure Type Secondary
Measure Title Mean Change From Baseline in Weight
Measure Description Body weight was measured at all visits, without shoes and wearing light clothing. Mean Change From Baseline in Weight was calculated as endpoint value minus the baseline value.
Time Frame Baseline (Week 0), Week 4, 8, 12, 16, 24, 36, 48 and 56  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
Safety population. Here, n=number of participants with observed data contributing to the analysis. Data for Site 040449 was not included in analysis due to audit finding.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   477   481   470 
Mean Change From Baseline in Weight 
[Units: Kilogram (kg)]
Mean (Standard Deviation)
     
Week 4       
Participants Analyzed   456   471   458 
Week 4   0.1  (2.04)   0.2  (1.66)   0.3  (1.88) 
Week 8       
Participants Analyzed   436   460   435 
Week 8   0.2  (2.35)   0.3  (1.95)   0.7  (2.37) 
Week 12       
Participants Analyzed   430   453   414 
Week 12   0.2  (2.50)   0.3  (2.09)   0.9  (2.33) 
Week 16       
Participants Analyzed   425   441   412 
Week 16   0.1  (2.88)   0.3  (2.44)   1.0  (2.56) 
Week 24       
Participants Analyzed   413   430   391 
Week 24   0.1  (2.88)   0.6  (2.77)   0.9  (3.26) 
Week 36       
Participants Analyzed   384   405   367 
Week 36   0.0  (3.37)   0.7  (3.00)   1.3  (3.92) 
Week 48       
Participants Analyzed   363   396   346 
Week 48   0.4  (3.54)   0.8  (3.47)   1.3  (4.13) 
Week 54       
Participants Analyzed   346   384   322 
Week 54   0.3  (3.74)   0.7  (3.69)   0.8  (4.53) 

No statistical analysis provided for Mean Change From Baseline in Weight



17.  Secondary:   Change From Baseline in Hemoglobin Values   [ Time Frame: Baseline (Week 0), Week 4, 16, 36 and 48 ]

Measure Type Secondary
Measure Title Change From Baseline in Hemoglobin Values
Measure Description Blood samples of participants were collected for Hemoglobin. Change from baseline in Hemoglobin was calculated as endpoint value minus the baseline value.
Time Frame Baseline (Week 0), Week 4, 16, 36 and 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
Safety population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   496   494   489 
Change From Baseline in Hemoglobin Values 
[Units: Grams per litre (g/L)]
Mean (Standard Deviation)
     
Week 4       
Participants Analyzed   453   457   457 
Week 4   -0.4  (6.45)   -2.5  (6.92)   -3.7  (6.19) 
Week 16       
Participants Analyzed   423   444   407 
Week 16   -0.6  (7.01)   -6.2  (7.76)   -11.9  (8.65) 
Week 36       
Participants Analyzed   382   398   363 
Week 36   -2.0  (7.25)   -6.4  (8.04)   -12.5  (9.38) 
Week 48       
Participants Analyzed   346   377   334 
Week 48   -1.9  (7.86)   -6.5  (8.45)   -12.2  (10.62) 

No statistical analysis provided for Change From Baseline in Hemoglobin Values



18.  Secondary:   Change From Baseline in Hematocrit Values   [ Time Frame: Baseline (Week 0), Week 4, 8, 12, 16, 36 and 48 ]

Measure Type Secondary
Measure Title Change From Baseline in Hematocrit Values
Measure Description Blood samples of participants were collected for Hematocrit . Change from baseline in Hematocrit was calculated as endpoint value minus the baseline value.
Time Frame Baseline (Week 0), Week 4, 8, 12, 16, 36 and 48  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
Safety population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   496   494   489 
Change From Baseline in Hematocrit Values 
[Units: Litre]
Mean (Standard Deviation)
     
Week 4       
Participants Analyzed   453   457   457 
Week 4   -0.0007  (0.02133)   -0.0068  (0.02189)   -0.0115  (0.01994) 
Week 16       
Participants Analyzed   423   444   407 
Week 16   -0.0003  (0.02318)   -0.0174  (0.02460)   -0.0339  (0.02697) 
Week 36       
Participants Analyzed   382   398   363 
Week 36   -0.0037  (0.02246)   -0.0167  (0.02471)   -0.0352  (0.02834) 
Week 48       
Participants Analyzed   346   377   334 
Week 48   -0.0029  (0.02447)   -0.0177  (0.02658)   -0.0346  (0.03177) 

No statistical analysis provided for Change From Baseline in Hematocrit Values



19.  Secondary:   Mean Change From Baseline in Short Term Memory Assessment Score   [ Time Frame: Baseline (Week 0), Week 8, 16, 24, 36, 48 and 56 ]

Measure Type Secondary
Measure Title Mean Change From Baseline in Short Term Memory Assessment Score
Measure Description Short term memory assessment score was based on ADAS-Cog questionnaire (Question 1 and 7). ADAS is a performance-based test that measures specific cognitive and behavioral dysfunctions in participants with AD. Question 1 (Word Recall) and Question 7 (Word Recognition) of the ADAS-Cog questionnaire were summed to get a short term memory assessment score. Word recall task consist of the participants score was the mean number of words not recalled on three trials (maximum score 10) and word recognition task, to score this item the number of incorrect responses was counted (maximum error score was 12). The total score ranged from 0 to 22 with 0 indicating absence of symptoms and higher scores indicating greater dysfunction; a negative change from baseline indicated improvement. Change from Baseline in short term memory assessment was calculated as endpoint value minus the baseline value.
Time Frame Baseline (Week 0), Week 8, 16, 24, 36, 48 and 56  

Population Description
Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
ITT population. Only those participants available at the specified time points were analyzed.

Reporting Groups
  Description
Placebo Participants received rosiglitazone extended release matching placebo tablet orally, once daily until Week 54.
2mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 2 mg tablet orally, once daily until Week 48 followed by placebo tablet orally once daily until Week 54.
8mg Rosiglitazone Extended Release Participants received rosiglitazone extended release 4 mg tablet orally, once daily until Week 4 followed by 8 mg tablet orally once daily from Week 4 until Week 48 followed by placebo tablet orally once daily until Week 54.

Measured Values
   Placebo   2mg Rosiglitazone Extended Release   8mg Rosiglitazone Extended Release 
Participants Analyzed   461   473   459 
Mean Change From Baseline in Short Term Memory Assessment Score 
[Units: Score on a scale]
Mean (Standard Deviation)
     
Week 8       
Participants Analyzed