Prevalence and correlates of dementia and mild cognitive impairment classified with different versions of the modified Telephone Interview for Cognitive Status (TICS-m)

Objectives The modified Telephone Interview for Cognitive Status (TICS-m) is an efficient and cost-effective screening instrument of dementia, but there is less support for its utility in the detection of mild cognitive impairment (MCI). We undertook a comprehensive evaluation of the utility of different TICS-m versions with or without an education-adjusted scoring method to classify dementia and MCI in a large population-based sample. Methods Cross-sectional assessment of cognition (TICS-m), depressive symptoms (CES-D), and apolipoprotein E (APOE) epsilon 4 status was performed on 1772 older adults (aged 71-78 y, education 5-16 y, 50% female) from the population-based older Finnish Twin Cohort. TICS-m classification methods with and without education adjustment were used to classify individuals with normal cognition, MCI, or dementia. Results The prevalence of dementia and MCI varied between education-adjusted (dementia = 3.7%, MCI = 9.3%) and unadjusted classifications (dementia = 8.5%-11%, MCI = 22.3%-41.3%). APOE epsilon 4 status was associated with dementia irrespective of education adjustment, but with MCI only when education adjustment was used. Regardless of the version, poorer continuous TICS-m scores were associated with higher age, lower education, more depressive symptoms, male sex, and being an APOE epsilon 4 carrier. Conclusions We showed that demographic factors, APOE epsilon 4 status, and depressive symptoms were similarly related to continuous TICS-m scores and dementia classifications with different versions. However, education-adjusted classification resulted in a lower prevalence of dementia and MCI and in a higher proportion of APOE epsilon 4 allele carriers among those identified as having MCI. Our results support the use of education-adjusted classification especially in the context of MCI.