Wandering in dementia: A longitudinal study

Objective: This article analyzes the natural history of wandering behavior throughout the course of dementia. Design: Prospective, 10-year, longitudin al study of wandering behavior in dementia, with autopsy follow-up. Setting : Participants with dementia, living at home with a carer. All lived in Oxf ordshire, UK. Participants: Eighty-six people with dementia who were living at home with a carer and who were able to walk unaided at entry to study. Measures: At 4-monthly intervals, the carers were interviewed using the Pre sent Behavioural Examination to assess wandering behavior in detail; partic ipants with dementia were assessed cognitively. Nine types of 'wandering" b ehavior were distinguished. Results: Changes in wandering behavior were not generally related to gender, age, or time since onset of dementia. Onset o f different types of wandering behavior showed some relationship with cogni tive state. Various forms of increased walking first appeared during modera te dementia, each type typically persisting for 1 to 2 years. Late dementia was characterized by decreased walking and immobility. Conclusions: Wander ing behavior in dementia can cause great problems for carers. There are dif ferent causes for such changes, some of which are related to cognitive abil ity, for example increased confusion results in ineffectual "pottering" and getting lost. Increased walking at night corresponds with disruption of di urnal rhythm.