P3-430: Impact of cognitive impairment on the reliability and construct validation of the QOL-AD scale

Abstract

Background: To assess the impact of cognitive impairment on the internal consistency and construct validity of the Brazilian version of the Quality of Life (QOL) in Alzheimer’s disease (AD) evaluation scale (QOL-AD) (Logsdon, 1999). Methods: The scale was administered to a sample of 60 patients with probable AD, with mild (n 30) and moderate (n 30) dementia, according to NINCDS-ADRDA and DSM-III-R criteria, respectively. Internal consistency was evaluated through Cronbach’s alpha coefficient. Construct validity was determined as follows: 1) Correlation between total scores of patients’ and caregivers’ reports about patients’ QOL with the results of the following measures: MMSE, Geriatric depression screening scale (GDS), Cornell scale for depression in dementia, Physical and Instrumental-Self Maintenance scales (IADLs and BADLs), Neuropsychiatric inventory (NPI) and socio-economic level and; 2) Correlation of the total score of the scale with a generic QOL instrument (WHOQOLbrief). All analyses were performed for both groups of severity. Results: The reliability was excellent for both groups of dementia severity. In mild dementia the patient’s and the relative’s versions in relation to the patient’s QOL (alpha 0.81 and 0.75, respectively). In moderate dementia the alpha values were 0.80 for patient’s report and 0.83 for relative’s report in relation to the patient’s QOL. In the construct validation of the patient’s version for mild dementia, we obtained significant correlations with GDS (-0.76, p 0.01), Cornell (-0.53, p 0.01), NPI (-0.46, p 0.05) and the total score of WHOQOL-brief (0.89, p 0.01). In the caregiver’s version about patient’s quality of life we registered significant correlations with the following measures: GDS (-0.48, p 0.01), Cornell (-0.57, p 0.01), IADLs(-0.36, p 0.05), NPI (-0.62, p 0.01) and total score of WHOQOLbrief(1.00, p 0.01). In the patient’s version for moderate dementia we obtained significant correlations with GDS (-0.45, p 0.05), Cornell (-0.46, p 0.05) and total score of WHOQOL-brief (0.90, p 0.01). In the caregiver’s version about patient’s QOL, significant correlations emerged with Cornell (-0.68, p 0.01), IADLs (-0.41, p 0.05), BADLs (-0.49, p 0.01), NPI (-0.67, p 0.01) and total score of WHOQOL-brief (0.69, p 0.01). Conclusions: The reliability and construct validity of QOL-AD were kept independent from the severity of the disease in this sample of AD patients with mild and moderate dementia.