Community-based Cohort of Functional Decline in Subjective Cognitive Complaint Elderly

Objectives of the Study:

The general objective of this study is to analyze the rate of functional decline (FD) and progression to clinical dementia and their risks factors (biomedical, imaging, psychosocial, and clinical) in a community-dwelling elderly with subjective cognitive complaint (SCC), through a population-based study. The specific objectives are to determinate (i) longitudinal evolution of biomarkers measured from blood, stool and structural and functional neuroimaging (MRI), (ii) evolution of the health-related quality of life; (iii) rate of cardiovascular events (stroke and coronary events), and (iv) mortality rates. The investigators also aim to build the capacity to undertake clinical research on brain aging and dementia disorders and create Data-Bank and Bio-Banks with an appropriate infrastructure to further studies and facilitate access to the data and samples for research.

The Gero cohort is the core clinical project of the GERO, supported by the Fund for Research Centers in Priority Areas Program (FONDAP) of the Chilean National Commission for Scientific and Technological Research (CONICYT). GERO is initially funded for 5 years, and its main aim is to establish a center for studying Brain Aging in Chile including basic and clinical research.

Methods/Design

Field work during the first contact

The recruitment process considers two steps. Firstly, a lay team contacts each home to determine the presence of eligible individuals. In positive cases, the person receives a second visit by a trained psychologist who proceeds to check the eligibility. In case of acceptance, the inclusion and exclusion criteria protocol are applied. If the subject fulfills the criteria, the psychologist schedule a medical interview. Following this evaluation, a neurologist decides if the subject fulfill the inclusion criteria of the cohort.

The fieldwork is preceded by an outreach campaign (flyers, local radio advertisements, and presentations to community-organized groups) raising awareness about the visit of interviewers and the relevance of participating in the study. Rates of contact and response are monitored permanently, and the procedures around the contact and first interview are checked in the field and also by telephone to a subsample of participants. Contact to homes is attempted up to three times on different days and hours before considering it frustrated. The fieldwork started in November 2017 and is expected to finish at the end of 2019.

The lay team and psychologists involved in the first contact and recruitment received specific training on their labour in the field. The lay team completed a whole week training, which included theoretical and practical elements. Psychologists received a twelve weeks length training, which covers several sessions of neuropsychological assessment, including performance psychology.

Sample size

The sample size had to satisfy two criteria, one concerned with the statistical power required to explore multiple associations with outcomes, and other related to the feasibility to perform a wide range of assessments to each participant assuming costs and logistics. Both criteria meant a trade-off between the tolerance to uncertainty around the parameters to be estimated and the number of assessments that would be investigated throughout the study. The final sample chosen was 300 participants. This number allows maintaining the integrity of the original protocol and permits to test associations equivalent to an odds ratio (OR) around 1.5 (Cohen's d equal to 0,22) in cases of exposition and probability of the outcome close to 50%, using a significance of 5%. It is expected to follow each participant between 2 and 3 years, accumulating roughly 750 person-years of follow up.

Follow-up and retention strategy

Socio-demographic, clinical, psychosocial, neuropsychological, neuropsychiatric, motor, neuroimaging, blood biomarkers, stool, and genetic samples will be performed as baseline evaluation and every 18 months, with the exception of genetic study that will be performed only at baseline and neuroimaging at baseline and the 36 month. Patients' health status, functionality, and involvement in the Gero cohort will be monitored every 6 months by a telephonic questionnaire.

To avoid a significant attrition of the sample a set following strategies have been considered: to recruit only people who have at least one person that can facilitate the contact with him or her, it means a person who can be contacted for asking about the location of the participant; telephone contact every six months; and domicile visit in case of absence of contact or attending to assessment appointments. Additionally, all transport costs of participants will be covered by the Gero cohort administration, as well as any food that is required during the days of assessment. Initially, the end of the follow up of the cohort is programmed for October 2022.

Assessments and measurements

The protocol considers an intensive and deep multidimensional study of factors related to the prognosis of FD and dementia development. The range of assessments includes: socio-demographic, psychosocial, neuropsychological, neuropsychiatric, motor, neuroimaging, blood biomarkers, genetic and stool samples to perform gut microbiome studies. Neuroimaging protocol will allow assessing brain atrophy, structural and functional connectivity and white matter lesions. Gero biological samples of whole blood, buffy coat, plasma, serum, and peripheral mononuclear cells are taken and processed according to the guidelines published in 2015. Samples are stored in our Gero biobank for long-term storage at -80 °C or in liquid nitrogen. Stool samples will be collected using standardized kits and DNA extracted using the protocol Q suggested by the international human microbiome standards (IHMS SOP 06 V1). Data are recorded in an ad-hoc platform developed by bioinformatics and bioengineers personal of Gero.

Data analysis plan

The Gero cohort offers a unique opportunity for multiple analyses to identify, correlate and analyze multidimensional factors related to FD and progression to dementia in elderlies with SCC.

In broad terms, a descriptive of baseline measurements (either outcomes or potential predictive factors) will be performed. The procedure will be repeated at each measurement time, every 18 months. Random effect models will be used for describing trajectories of participant subgroups and the whole cohort according to main variables, using Markov-Chain Montecarlo procedures.

The association between variables and outcomes will be explored broadly using different machine learning methods, such as elastic net procedure, random forest procedure, based-tree methods, and support vector machines. These procedures are suitable to lead with multi-collinearity and also high dimensional data (e.g. the number of predictive variables is larger than the number of participants in the cohort). Interpretation of causality will be conducted using standard random effect models and eventually structural equation modeling.

Missing data and loss of follow up of participants are common in observational studies, mainly in cohorts. Firstly, cases with missing data in any outcome will be explored and compared with cases without missing data describing any pattern. Secondly, two strategies will be followed to estimate results: (1) to analyze only cases with complete information (i.e. assuming that missing data is completely at random); and (2) imputing data according to multivariate imputation by chained equation techniques. The analysis will be performed using the statistical software R.

Coordination with local health services

The Gero cohort has been carefully design to avoid undermining the usual care of participants in their common health services facilities. Even more, a linkage between the health assessments provided by the cohort and the usual health care has been promoted.

In cases when the cohort's assessment detects a new health condition (diabetes, depression, hypertension, etc.), participants are derived to the primary healthcare centre of their territory. In case of detection of a significant neurological disorder (dementia syndrome, Parkinson, etc.) the participants are directly derived to specialized care according to their health district, communicating the decision to the primary health care.

Primary care health centres, specialized care polyclinics and the direction of the Health District involved have been informed about the study and jointly the protocol of derivation and communication were established.

Regulation of access to data/biospecimens

The access to data and biospecimens is regulated by the GERO directorate in accordance with the local Institutional Review Board authorization. A bilateral agreement must be signed before sharing of data. Access to the server will not be granted.