Dementia transition in early cognitive decline trajectories (DETECT) study: a prospective longitudinal study protocol

BACKGROUND: Mild cognitive impairment is widely recognized as a high-risk state associated with a progression to dementia. Although previous studies have reported several predictors of the transition from MCI to dementia, many of these predictive factors are non-modifiable, and evidence on modifiable factors remains limited and inconsistent. Moreover, to date, social factors and their role in this transition have not been sufficiently examined. Guided by the Biopsychosocial Model of Dementia, this protocol describes a study which will examine the transition to dementia as a multidimensional process shaped by biological, psychological, and social factors. This study aims to identify the determinants associated with the transition to dementia and to develop machine learning-based prediction models for this transition among older adults with mild cognitive impairment. METHODS: This is a 3-year prospective longitudinal study which will be conducted among older adults with mild cognitive impairment in South Korea. The participants will be recruited from community-based settings, and annual home visits will be conducted to collect data on biological, psychological, and social factors. This data will be obtained through self-reported questionnaires, physical measurements, wrist-worn actigraphy, sweat patch sampling, indoor environmental sensing, and public records. The outcome will be transition to dementia, defined by clinical diagnosis or cognitive screening criteria. Machine-learning algorithms will be used to develop the prediction models, and model performance will be evaluated using classification metrics and area under the receiver operating characteristic curve analysis. DISCUSSION: This study is expected to provide evidence on the determinants associated with the transition to dementia among older adults with mild cognitive impairment. The use of home-based assessments and multiple data sources may offer a broader understanding of factors related to this transition. The findings may be useful for early risk identification and for developing community-based interventions and policies to support dementia prevention.