Multifactorial Approach to Dementia, Multicentric Study

Previous studies have identified multiple risk factors for dementia in a variety of settings. However personality traits have received less attention as a contributing or mediating factor. The identification of potentially modifiable risk factors has been critical in the development of effective multifactorial intervention programmes for people at risk of dementia. Studies that examined the role of personality traits as risk factor for dementia recruited participants with mild cognitive impairment or mild to moderate dementia living in institutional or community settings. In this 3 tiered study, 3 lines of research will be pursued. In lines 1 and 2 retrospective analysis of patients with dementia will be performed searching for the mediating or predisposing role of narcissistic personality disorder as a risk factor for the development of dementia. in the 3rd line of research normal participants will complete a multifactorial baseline assessment of putative predictors of mild cognitive impairment or dementia and then will complete prospective cognitive, affective and personality diaries for a period of 12 months, in order to detect association of personality traits to dementia conversion.

Methods Ethics statement This study will be submitted to Ethics Committee of INICIEN Research Authority, and participants will be requested to give written informed consent in accordance with the declaration of Helsinki.

Design Longitudinal prospective and retrospective cohort study. Participant recruitment 3 lines of research. In line 1 and 2 patients with dementia will be assessed in a retrospective approach and in the line 3 healthy cases without cognitive impairment will be recruited to determine if narcisistic personality play a role in the risk of dementia onset (AD, frontotemporal dementia). Patients files and subjects will be recruited from multicenter Neurology, Old Age Psychiatry and Geriatric Medical secondary care outpatient clinics. Cases should be referred to these clinics for assessment, diagnosis and management by their primary care physician, or referred to a Psychiatrist by their clinical physician.

Inclusion and exclusion criteria All participants should be over 50 years of age in research line 1, and no age limit will be assumed for cases in reesarch lines 1 and 2. All diagnoses of dementia will be made according to DSM V criteria. Significant medical causes of dementia should be excluded during diagnostic investigations. Diagnoses will be made by operationalised criteria for dementia validated against neuropathological diagnosis Participants were excluded if they declined participation, died or withdrew from the study before commencing follow up period, unable to perform the assessments due to other co-morbid conditions, or were too visually impaired to complete cognitive assessments. Baseline Clinical Assessment All participants received a detailed baseline assessment to quantify putative risk factors for dementia. The factors included were selected on the basis of their previous identification as risk factors in more than one high quality study in older people, and/or their relevance to clinical features of dementia postulated to be causative of neurocognive disorder conversion.

Assessments included medical history (residence, history of life stressful events, affective state, sociodemographic factors). Objective assessments included the cognitive battery, personality assessment with narcisistic personality disorder, Yesavage depression scale. Participants were assessed at baseline and during follow up period of one year.

Dementia specific scales were used to assess specific staging level and dementia specific type, depression, and behavioural and psychological symptoms of dementia.

Outcome Measures The primary outcome measure will be dementia conversion occurring during the 12 month follow up period. Secondary outcome measures will be proportional hazard ratios for time to dementia conversion, according to diagnosis and status of personality disorder narcisistic type as putative clinical predictor. If the participants did not show off at previously assigned points, they will be reminded by telephone after 2 weeks. The caregiver will be asked to assist in the completion of diaries when dementia is present.

Statistics Differences in baseline characteristics across groups will be compared using Fisher's Exact test for categorical data, ANOVA for normally distributed data and Kruskal-Wallis for non-normally distributed data. Differences between individual groups will be compared using the Chi squared test for categorical data, Student's t test for normally distributed data and Mann-Whitney U test for non-normally distributed data.

In case of censoring of data from some participants (as a result of death or withdrawal from the study), Cox regression (adjusted for age and sex) will be used to obtain a proportional hazard ratio for personality disorder as risk factor to dementia conversion. Loglinear Poisson regression models (adjusted for age and sex) will be used to obtain an incidence density ratio for dementia and its subtypes in comparison with healthy controls. without personality disorders In order to examine the associations between exposure to narcisitic personality disorder as risk factors for dementia conversion in those participants with dementia, Cox regression will be used to obtain univariate proportional hazard ratios for each risk factor, adjusted for age and sex, using time to the occurrence of at least mild cognitive impairment as the dependent variable. Hazard ratios will be given according to presence or absence of the risk factor, or per point on quantitative scales as appropriate. Analyses will be performed initially for all participants with or without dementia, and then repeated stratified by diagnosis.

Following identification in univariate models, significant and potentially modifiable risk factors will be entered into a multivariate forward stepwise Cox regression model, p 0.05 for entry, p 0.1 for removal. Age and sex will be included even if not significant. Where similar clinical features coould be described by more than one significant risk factor the factor with the higher level of significance in univariate analyses will be entered into multivariate analyses, in order to avoid co-aggregation of predictors.