A TEAV for a laboratory parameter was defined as a value that was clinically significantly outside (above or below) the normal range postdose, but within the normal range prior to drug administration,or a value that represented a clinically significant exacerbation of an abnormality present prior to drug administration.Abnormal values for clinical chemistry parameters were:Sodium:Less than(<)130 milliequivalents per litre (mEq/L) or greater than(>)150 mEq/L;Potassium:<3 mEq/L or >5.5 mEq/L; Calcium: <8.4 milligram per deciliter (mg/dL) or >1.5 mg/dL;Albumin: 50% lower limit of normal (LLN); Alkaline Phosphatase:>=3*upper limit of normal (ULN);Aspartate aminotransferase (AST):>=3*ULN;Alanine aminotransferase(ALT):>=3*ULN;Total Bilirubin:>=2.0 mg/dL;Chloride:<90 mEq/L or >115 mEq/L;Creatinine:>=2.0 mg/dL;Creatine phosphokinase:>=3*ULN;Blood Urea Nitrogen (BUN):>=30 mg/dL. Percentage of participants with at least 1 abnormal TEAV (selected parameters) for clinical chemistry was reported.

The SIB evaluated the severity of cognitive dysfunction in participants with more advanced dementia. Test questions measured attention, language, orientation, memory, praxis, visuospatial ability, construction, social skills, orienting head to name. Non-verbal responses were allowed, thus decreasing the need for language output. Forty questions were included with a total possible score range of 0-100. Lower scores indicated greater cognitive impairment.

MMSE is a 30-point scale that measured orientation to time and place, registration, immediate and delayed recall, attention, language, and drawing. Scores ranged from 0 (most impaired) to 30 (no impairment). Lower score indicated more impairment.

ADCS-ADL is a comprehensive battery of ADL questions used to measure a participant's functional capabilities. The modified ADCS-ADL-severe scale is a 19-item scale that has been validated for the assessment of participants with moderate to severe dementia. It measured the most appropriate basic and instrumental abilities (such as walking, grooming, bathing, and eating) in this participant population. Response to each item was obtained by interview with the caregiver. Ratings reflected caregiver observations about the participant's actual functioning and provided an assessment of change in the functional state of the participant over time. The total score ranged from 0 to 54. Lower scores indicated greater functional impairment.

QoL-AD is a 13-item quality of life instrument developed to specifically assess QoL in participants who have dementia. Each item was scored on a 4-point scale (poor, fair, good, excellent) and a single score was calculated, ranging from 13 (low functioning) to 52 (normal function). Higher score indicated better QoL. The QoL-AD total scores for the participants and caregiver were the sum of the 13 items on each test.

EQ-5D is a health profile questionnaire assessing quality of life along 5 domains. Caregivers rated 5 domains of health (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression) by choosing from 3 answering options (1=no problems; 2=some problems; 3=extreme problems). The EQ-5D Health Utilities Index (HUI) is derived from the five dimensions of the EQ-5D, using country-specific preference weights (tariffs) to summarize how good or bad each health state is on a scale from 1 to 0. HSI total score ranged from 1 (full health) and 0 (worst health/death).

The SCB is a 25-item instrument that questions caregivers about the occurrence and degree of distress as aspects of the burden of care during the preceding two weeks. It was designed specifically for use with caregivers of Alzheimer's disease participants. Each item was assessed on a 5-point scale ranging from "0" (no occurrence) to "4" (occurrence with severe distress). The objective burden was the sum of the total numbers of (1 2 3 4) in all items. Total score ranged from 0 (low distress) to 100 (high distress).

The SCB is a 25-item instrument that questioned caregivers about the occurrence and degree of distress as aspects of the burden of care during the preceding two weeks. It was designed specifically for use with caregivers of Alzheimer's disease participants. Each item was assessed on a 5-point scale ranging from "0" (no occurrence) to "4" (occurrence with severe distress). The subjective burden was the sum of the (total number with score*score). Total score ranged from 0 (no occurrence) to 100 (occurrence with severe distress).

TOS is a pilot instrument designed to evaluate the caregiver's assessment of the participant's status. The caregiver was asked how much he/she thinks the participant has been helped by participating in the study. This instrument comprised a 7-point Likert scale in which a rating of 1=Very much improved; 2=Moderately improved; 3=Minimally improved; 4=About the same; 5=Minimally worse; 6=Moderately worse; 7=Very much worse. The total scale ranged from 1 (marked improvement) to 7 (marked worsening).

GAtS is a technique that is standardized with respect to the general approach, but individualized with respect to the outcome goals of each participant. GAtS was designed to provide insight into the changes that are considered as important by the caregiver and (if feasible) by the participants. At study initiation, the participants (if capable) and, separately, the caregiver were asked to specify up to 4 goals for the participants during study participation. For each goal, a description of the current state (or one supplied with the help of the clinician if necessary) was provided to anchor the baseline assessment at 0, and other possible outcomes were described. The outcome for each goal was quantified by a 4-point scale that provided for improvement (+1, +2), no change (0) or worsening (-1, -2). Total score ranged from -30 (worsened) to 130 (most improved). Baseline score was set to be 50 (when scores of all goals are '0' [no change]).

Total of all reporting groups

A TEAV for a laboratory parameter was defined as a value that was clinically significantly outside (above or below) the normal range post-dose, but within the normal range prior to drug administration, or a value that represented a clinically significant exacerbation of an abnormality present prior to drug administration. Abnormal values for hematology parameters were: White Blood cells count: less than or equal to [<=] 2,800/per millimeter (mm) or greater than or equal to [>=] 16,000/mm; Neutrophils: <=15 percent (%); Hemoglobin: Male (<=11.5 gram per deciliter [g/dL]), Female (<=9.5 g/dL); Hematocrit: Male (<=37%), Female (<=32%); Eosinophils: >=10%; Platelet Count: <=75,000/mm or >=700,000/mm. Percentage of participants with at least 1 abnormal TEAV for hematology was reported.

The safety population included all participants who received at least one dose of the study medication during the open-label treatment period and had at least one post-baseline safety assessment. Here 'N' (overall number analyzed) are the participants who were evaluable for the outcome measure.

A TEAV for a laboratory parameter was defined as a value that was clinically significantly outside (above or below) the normal range postdose, but within the normal range prior to drug administration,or a value that represented a clinically significant exacerbation of an abnormality present prior to drug administration.Abnormal values for clinical chemistry parameters were:Sodium:Less than(<)130 milliequivalents per litre (mEq/L) or greater than(>)150 mEq/L;Potassium:<3 mEq/L or >5.5 mEq/L; Calcium: <8.4 milligram per deciliter (mg/dL) or >1.5 mg/dL;Albumin: 50% lower limit of normal (LLN); Alkaline Phosphatase:>=3*upper limit of normal (ULN);Aspartate aminotransferase (AST):>=3*ULN;Alanine aminotransferase(ALT):>=3*ULN;Total Bilirubin:>=2.0 mg/dL;Chloride:<90 mEq/L or >115 mEq/L;Creatinine:>=2.0 mg/dL;Creatine phosphokinase:>=3*ULN;Blood Urea Nitrogen (BUN):>=30 mg/dL. Percentage of participants with at least 1 abnormal TEAV (selected parameters) for clinical chemistry was reported.

The safety population included all participants who received at least one dose of the study medication during the open-label treatment period and had at least one post-baseline safety assessment. Here 'N' (overall number analyzed) are the participants who were evaluable for the outcome measure.

The SIB evaluated the severity of cognitive dysfunction in participants with more advanced dementia. Test questions measured attention, language, orientation, memory, praxis, visuospatial ability, construction, social skills, orienting head to name. Non-verbal responses were allowed, thus decreasing the need for language output. Forty questions were included with a total possible score range of 0-100. Lower scores indicated greater cognitive impairment.

The Intent-to-treat (ITT) population included all safety participants who have at least one post-baseline efficacy measurement. Here 'N' (overall number of participants analyzed) included all participants who were evaluable for this outcome measure and 'n' (number analyzed) included all participants who were evaluable at specified timepoints for this outcome measure.

MMSE is a 30-point scale that measured orientation to time and place, registration, immediate and delayed recall, attention, language, and drawing. Scores ranged from 0 (most impaired) to 30 (no impairment). Lower score indicated more impairment.

The ITT population included all safety participants who have at least one post-baseline efficacy measurement. Here 'N' (overall number of participants analyzed) included all participants who were evaluable for this outcome measure and 'n' (number analyzed) included all participants who were evaluable at specified timepoints for this outcome measure.

ADCS-ADL is a comprehensive battery of ADL questions used to measure a participant's functional capabilities. The modified ADCS-ADL-severe scale is a 19-item scale that has been validated for the assessment of participants with moderate to severe dementia. It measured the most appropriate basic and instrumental abilities (such as walking, grooming, bathing, and eating) in this participant population. Response to each item was obtained by interview with the caregiver. Ratings reflected caregiver observations about the participant's actual functioning and provided an assessment of change in the functional state of the participant over time. The total score ranged from 0 to 54. Lower scores indicated greater functional impairment.

The ITT population included all safety participants who have at least one post-baseline efficacy measurement. Here 'N' (overall number of participants analyzed) included all participants who were evaluable for this outcome measure and 'n' (number analyzed) included all participants who were evaluable at specified timepoints for this outcome measure.

QoL-AD is a 13-item quality of life instrument developed to specifically assess QoL in participants who have dementia. Each item was scored on a 4-point scale (poor, fair, good, excellent) and a single score was calculated, ranging from 13 (low functioning) to 52 (normal function). Higher score indicated better QoL. The QoL-AD total scores for the participants and caregiver were the sum of the 13 items on each test.

The ITT population included all safety participants who have at least one post-baseline efficacy measurement. Here 'N' (overall number of participants analyzed) included all participants who were evaluable for this outcome measure and 'n' (number analyzed) included all participants who were evaluable at specified timepoints for this outcome measure.

EQ-5D is a health profile questionnaire assessing quality of life along 5 domains. Caregivers rated 5 domains of health (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression) by choosing from 3 answering options (1=no problems; 2=some problems; 3=extreme problems). The EQ-5D Health Utilities Index (HUI) is derived from the five dimensions of the EQ-5D, using country-specific preference weights (tariffs) to summarize how good or bad each health state is on a scale from 1 to 0. HSI total score ranged from 1 (full health) and 0 (worst health/death).

The ITT population included all safety participants who have at least one post-baseline efficacy measurement. Here 'N' (overall number of participants analyzed) included all participants who were evaluable for this outcome measure and 'n' (number analyzed) included all participants who were evaluable at specified timepoints for this outcome measure.

The SCB is a 25-item instrument that questions caregivers about the occurrence and degree of distress as aspects of the burden of care during the preceding two weeks. It was designed specifically for use with caregivers of Alzheimer's disease participants. Each item was assessed on a 5-point scale ranging from "0" (no occurrence) to "4" (occurrence with severe distress). The objective burden was the sum of the total numbers of (1 2 3 4) in all items. Total score ranged from 0 (low distress) to 100 (high distress).

The ITT population included all safety participants who have at least one post-baseline efficacy measurement. Here 'N' (overall number of participants analyzed) included all participants who were evaluable for this outcome measure and 'n' (number analyzed) included all participants who were evaluable at specified timepoints for this outcome measure.

The SCB is a 25-item instrument that questioned caregivers about the occurrence and degree of distress as aspects of the burden of care during the preceding two weeks. It was designed specifically for use with caregivers of Alzheimer's disease participants. Each item was assessed on a 5-point scale ranging from "0" (no occurrence) to "4" (occurrence with severe distress). The subjective burden was the sum of the (total number with score*score). Total score ranged from 0 (no occurrence) to 100 (occurrence with severe distress).

The ITT population included all safety participants who have at least one post-baseline efficacy measurement. Here 'N' (overall number of participants analyzed) included all participants who were evaluable for this outcome measure and 'n' (number analyzed) included all participants who were evaluable at specified timepoints for this outcome measure.

TOS is a pilot instrument designed to evaluate the caregiver's assessment of the participant's status. The caregiver was asked how much he/she thinks the participant has been helped by participating in the study. This instrument comprised a 7-point Likert scale in which a rating of 1=Very much improved; 2=Moderately improved; 3=Minimally improved; 4=About the same; 5=Minimally worse; 6=Moderately worse; 7=Very much worse. The total scale ranged from 1 (marked improvement) to 7 (marked worsening).

The ITT population included all safety participants who have at least one post-baseline efficacy measurement. Here 'N' (overall number of participants analyzed) included all participants who were evaluable for this outcome measure and 'n' (number analyzed) included all participants who were evaluable at specified timepoints for this outcome measure.

GAtS is a technique that is standardized with respect to the general approach, but individualized with respect to the outcome goals of each participant. GAtS was designed to provide insight into the changes that are considered as important by the caregiver and (if feasible) by the participants. At study initiation, the participants (if capable) and, separately, the caregiver were asked to specify up to 4 goals for the participants during study participation. For each goal, a description of the current state (or one supplied with the help of the clinician if necessary) was provided to anchor the baseline assessment at 0, and other possible outcomes were described. The outcome for each goal was quantified by a 4-point scale that provided for improvement (+1, +2), no change (0) or worsening (-1, -2). Total score ranged from -30 (worsened) to 130 (most improved). Baseline score was set to be 50 (when scores of all goals are '0' [no change]).

The ITT population included all safety participants who have at least one post-baseline efficacy measurement. Here 'N' (overall number of participants analyzed) included all participants who were evaluable for this outcome measure and 'n' (number analyzed) included all participants who were evaluable at specified timepoints for this outcome measure.

Donepezil SR 23 mg once daily orally for 12 months to participants who received donepezil 10 mg IR in the preceding double-blind study E2020-G000-326 (NCT00478205).