Developing sensitive and valid indicators of dementia and cognitive function in low literacy and low income countries is challenging and requires creative approaches to complement standard neuropsychological batteries. In this application, we aim to incorporate an innovative approach to assessing cognition and to validate it against standard cognitive batteries, medical and informant assessments in 3 of the HRS sister countries: South Africa (HAALSI), United States (HRS/COG), and Ireland (TILDA). South Africa has a low literacy rate in older people, varying degrees of use of written language and different cultural cues compared to Western industrialized countries. In order to assess cognition in this setting, we assess cognitive function leveraging non-verbal and minimally numeric assessments for a range of cognitive domains. Through administration via tablets, responses are recorded and scored immediately with less reliance on interviewer training than most other cognitive assessments. The tablet-based assessments will be validated in relation to the Harmonized Cognitive Assessment Protocol (HCAP) developed by HRS for this application across HAALSI, TILDA and HRS/COG (Aim 1). Clinical and informant interviews are added to develop a dementia risk score (Aim 2). We conduct the validations and harmonization in age and education stratified samples of 300 men and women 50 -79 years old in each of the three cohorts (n=900). These novel `next generation' instruments are just being developed and we now need information regarding their validity compared to standard cognitive assessments. Developing cross-walks between these measures and other commonly used indices is of special importance in the set of sister studies related to the Health and Retirement Study where the explicit goals are to permit comparisons across countries, especially with regard to dementia. TILDA and HAALSI offer specific advantages in calibration because they have had rich additional neuropsychological and health data from earlier waves across an educationally diverse older population. We have two aims in our application:
Aim 1. To assess the validity of the Oxford based language controlled cognitive measures in relation to the HCAP measures using identical platforms for HRS/COG, TILDA and HAALSI 1.a. to evaluate content, criterion and construct validity of Oxford assessments against HCAP 1.b. to develop a refined battery of items based on the 1.c. to develop a harmonized set of items that have good psychometric properties Aim 2. To develop a common dementia risk score in each country based on the additional information gathered from proxy/informants and a clinical evaluation. This risk score will be harmonized and validated.
Developing sensitive and valid indicators of dementia and cognitive function in low literacy and low income countries is challenging and requires creative approaches to complement standard neuropsychological batteries. We use an innovative language controlled tablet based approach to assessing cognition and validate it against standard cognitive batteries, medical and informant assessments in 3 of the HRS sister countries: South Africa (HAALSI), United States (HRS/COG), and Ireland (TILDA). We evaluate novel approaches to reduce cultural or educational bias in cognitive assessments. By using a tablet assessment with tasks that are visually-oriented and use simple language, we will be better able to disentangle the causal pathways from socioeconomic conditions to cognition. We aim to increase our ability to understand if cross-country comparisons reflect genuine differences in risk and prevalence that will give us important etiologic clues regarding dementia.
