Abstract
OBJECTIVES: Attrition is one of the greatest difficulties in longitudinal studies on cognitive ageing because of the associated risk of underestimating declines. The aims of this paper were to characterize the magnitude and selectivity of attrition in a study of mild cognitive impairment. DESIGN: Forty two patients with multiple-domain amnestic MCI, 71 with single-domain amnestic MCI, 35 with non-amnestic MCI and 318 healthy controls were recruited from primary care centers and assessed at baseline. MEASUREMENTS: All participants underwent extensive neuropsychological evaluation, including the Mini-Mental State Examination, the Californian Verbal Learning Test, the CAMCOG-R battery, the Counting Span task and Listening Span task, and the Subjective Memory Complaints Questionnaire. RESULTS: 21.5% of the participants at baseline did not participate in the follow-up assessment. Comparison between respondents and non-returners did not reveal differences in cognitive performance in the MCI group. Data obtained at the initial assessment regarding comorbidity, social activities and attention given to memory training enabled prediction of the status of the participants in the follow-up assessment. CONCLUSION: Identification of potential non- returners is relevant, especially in MCI studies, in order to develop retention strategies to minimize attrition.