The effectiveness of reminiscence for dementia has been claimed as an effective tool, but scientific validation and systematic assessment of this method is needed. Materials and Methods: A randomized controlled trial (RCT) was conducted to search whether a reminiscence program is associated with improvement of quality of life of demented long term care residents. The trial had three arms: interventional, comparison and control. The reminiscence program was modeled within a life-story approach, while comparison group received informal counseling to control for changes in quality of life resulting from social contacts. The Social Engagement Scale (SES) and Self-rated Quality of Life Index (SRQoL) were the outcome measures. The results were examined at baseline (T0), twelve weeks (T1), and 6 months (T2) after intervention. The sample had 135 subjects [intervention group (N=45), comparison group (N=45) and control group (N=45)].

A total number of 135 residents will be recruited from two private funded long term nursing homes, which shared equal structural and functional characteristics. Subjects were randomly assigned to one of the three groups (intervention, comparison and control). The subjects admitted for the study are diagnosed as having Alzheimer's disease according to the DSM-IV are able to communicate with a Holden Communication Scale scores>25 and have a Folstein mental Exam score above 10. Exclusion criteria are active major psychiatric disorders (schizophrenia, major affective disorders); acute or unstable chronic medical conditions including cardiac or lung diseases; blindness and deafness, even with hearing aids, both assessed with the RAI blindness and deafness scales.

Power and sample size calculation. Sample size will be estimated taking into account that the study had to test a null hypothesis whether the different participating groups were similar or different. It is assumed that a significant change in dementia nursing home resident quality of life (QoL) from baseline would be of half a standard deviation, which equals to 3 points, as found in a previous longitudinal study10. With this predicted effect size of QoL, a power of 80% and a type I error for independent groups of 0.01, the estimated sample size was 143 participants in the total study population.

Methods and materials The intervention was designed as an individual treatment condition in which each participating subject received 24 bi-weekly sessions of reminiscence therapy, lasting one hour each one, over a period of 12 weeks. The latter refers to the use of images, sentences or memorabilia which help to focus on specific segments of the life history of an individual, and stimulates the emergence of affect-laden personal recalls, which are later verbalized in the context of guided conversations. The term story life is intended to highlight samples of meaningful events of the subject's life rather than a historically structured biography. The control group was administered counseling and informal social contacts in bi-weekly sessions of one hour, but they didn't participate in reminiscence sessions. This was intended to rule out the possibility that improvement in quality of life was due only to attention received and social stimulation. The comparison group received unstructured social contacts, again in bi-weekly sessions of one hour each one. Remaining features in the design of the three arms were similar with the only exception being the structured reminiscence program participation. The study was approved by the local Ethics Review Committee. This study adopted a single-blinded, parallel-group (one intervention, one comparison, and one control [no-intervention] group) design to address the following hypothesis: a) residents with Alzheimer's disease submitted to a reminiscence program intervention will show a better quality of life as a consequence of greater sense of self identity regarding groups with no specific therapeutic intervention, b) this quality of life improvement will be sustained beyond actual therapeutic intervention, due to consolidation of self identity and the reinforcing effects of increased competence, efficacy and personal involvement in everyday activities, associated with higher sense of self.

Demographic and clinical data of residents, including age, gender, marital status, level of education, religion, length of dementia, length of stay in the nursing home, associated medical problems, Mini-mental State Examination (MMSE) score, Cognitive Performance Scale level (CPS)15, ability to communicate, functional abilities, use of psychotropic medications, fitness programs, physical restrictions, number of visits per month from families and friends, and caregiver burden with the Burden Interview (short version ) will be collected. The severity of dementia will be staged with the CDR. The functional performance of the residents will be assessed using an index of physical function for level of independence in eating, dressing, toileting, transferring, and walking by using magnitude estimation weights. Each level of disability on each activity is given a weight, rather than a simple count. The resulting score ranges from 0 (no limitation) to 3.77 (completely disabled on five activities of daily living) and has ratio scale properties. To assess physical restraints the investigators used an indicator of daily use of full bed rails, trunk or limb restraints, or a chair that prevents rising. Physical restraints are not an aspect of the individual resident, but they are a clinical care process that is modifiable by the facility. Restraint use is considered an indicator of poor quality of care and an infringement on individual autonomy that diminishes QOL. Cognitive performance was assessed with the CPS. This is a clinically derived scale to predict MMSE and Test for Severe Impairment scores18. While MMSE has a floor effect with minimal scores suggesting questionable validity for more cognitively impaired elders, the TSI achieves meaningful variations, minimizes reliance on language skills and permitting subjects to answer correctly through nonverbal as well as verbal responses. Is composed of twenty-one items covering six cognitive areas: well-learned motor performance, language comprehension, language production, immediate and delayed memory, conceptualization, and general knowledge. The best score is 24, the lower is zero. For persons with an MMSE score of 11 or more, they will have TSI scores of 22 or higher. The CPS is composed by five items: 1-short-term memory, 2-cognitive skills for daily decision making, 3-coma (or persistent vegetative state), 4-making self understood and 5-eating. The scale has an average inter-rater reliability of 0.85 and a sensibility of 0.92 and specificity of 0.87. The CPS classifies residents into seven cognitive performance levels, from level 0 (Intact) with a mean MMSE score of 25, to Level 6 (Very Severe Impairment) with a near to zero. CPS Levels 2 and 3 (Mild and Moderate Impairment) corresponds to a MMSE score of 10 or higher, averaging 10.3 and 13.8, and to a TPI of 21 (SD=3.6). The Social Engagement Scale19 rates the resident status during the last seven days in areas such as ease interaction with others and doing planned or structured activities. Each item is ranked on a binary basis, as yes (1) or no (0) by the caregiver. The highest score is 6 and the lowest is 0. It has high internal consistency (intra-class correlation: 0.51-0.64), and the items shows reliability across different groups of residents with variable levels of functional and cognitive status. The resident self-reported SRQoL was measured using a multidimensional self-report instrument20. It measures 11 dimensions of QOL relative to a resident's experience: comfort, functional competence, privacy, dignity, autonomy, meaningful activities, relationships, food enjoyment, spiritual well-being, security, and individuality. Each dimension is scored on a 4-point Likert scale, with 4 meaning often, 3 sometimes, 2 rarely and 1 never. Residents unable to use the 4-point scale could answer ''generally yes'' or ''generally no.'' These responses are scored as 3.8 and 1.5, respectively, based on a z score approximation method. Reliability scores range between Cronbach's alpha values of 0.78 and 0.8521. Anxiety will be assessed using the Rating of Anxiety in Dementia (RAID): an 18-item scale with scores >11 indicating significant anxiety symptoms. Depression will be ruled out using the Minimum Data Set Depression Rating Scale. This is a standardized screening instrument for detecting depression among nursing home residents. Its comprises seven core Minimum Data Set mood items with a sensibility of 69% and a sensitivity of 91%, and a Cronbach α measure of internal consistency of 0,75. It has a score range of 0-14 with a cut-off point of 3. The Burden Interview short version (ZBI)24 will be used to measure the strain and burden experienced by caregivers on a 12 items scale, each one is scored on a 5 point Likert scale from 0 (never) to 4 (always). It has a range of results from 0 to 48, a Cronbach's alpha of 0.88, and a cutoff score of 17. Well being of residents will be assessed using the Well-being/Ill-being Scale (WIB)25 which includes positive components such as "being able to express wishes in an acceptable way," "bodily relaxation," and "creative self expression" (such as singing, dancing or painting), and negative ones such as "unattended sadness or grief," "sustained anger" or "anxiety." The WIB scale rates each category of behavior observed every five minutes for a minimum of six hours. After five minutes, the rater quantifies the nature of the observed behavior category by assigning a WIB value to it. The six-point WIB scale ranges from very negative to very positive (-5, -3, -1, +1, +3, +5). Values will be calculated at the end of the observation period to extract a mean score.

Inclusion Criteria:

• Alzheimer's disease
• Able to communicate

Exclusion Criteria:

• Active major psychiatric disorders
• Acute or unstable chronic medical conditions
• Blindness
• Deafness