Psychometric properties of the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) in a low-education and low-income population in Southern Brazil

Abstract

Introduction: Aging is spreading in the worldwide. How old are the people, it will have more chance to develop dementia and to track them, are hardly, by the way there are tool that are trying to objective this screening to make this diagnoses as fast as possible. Objectives: To study IQCODE-BR (Informant Questionnaire on Cognitive Decline in the Elderly) psychometrics characteristics in the elderly assisted by Family Health Program(ESF)-POA. Methods: Transversal study from population with low education and social level. IQCODE was applied in all the informants from the 87 elderly assisted in the AMBEC (Outpatient Brain Aging from PUCRS), in the period of March to Dec 2013, and it’s a reference for monitoring of the elderly in Family Health Program(ESF)–POA from a cross-sectional population-based study. Those elderly were assisted in the AMBEC, with a minimum a neurological and a psychiatric appointment, receiving a diagnosis of depressive disorder (DSM-IV criteria), Mild Cognitive Impairment (MCI) or dementia (criteria from NIA-AA, 2011), and part of the elderly had no such diagnosis. The application of IQCODE was blind to diagnosis. Results: There were studied 87 elderly people and their health home care. The average was 72 years old (60-90 years), most of them were women (72,4%), 31 were illiterate (35,6%). The elderly was classified in 4 groups, by the way, 30 had diagnoses of dementia (G4), 20 had MCI (G3), 21 had depression (G2), 16 did not have none diagnosis (G1). We observed that were a correlation by MCI and dementia with high age and low educational level. The mean cutoff from IQCODE were higher in G2 and G3 than G1, being similar between the first 2 groups. The mean cutoff from G4 were higher than all others. The IQCODE and the short form showed a Cronbach’s alpha of 0.906 and 0.908, respectively. The ROC curves showed for diagnosis of MCI a power of discrimination is higher when excluding patients with depression, with no significant difference in the diagnosis of dementia. The complete version and the short form of IQCODE showed similar accuracy. Conclusion: The IQCODE is an excellent tool for screening MCI and dementia. We recommend using the short form with a cutoff of 3.22 for MCI but also having to exclude depression diagnosis and 3.48 for dementia regardless of having depressive symptoms.