Alternative TMA-93 version demonstrates comparable performance and discriminative accuracy in early Alzheimer's disease

BackgroundThe TMA-93 is a brief memory test shown to be accurate for early Alzheimer's disease (AD) diagnosis. Normative data are available, and biomarker-based cut-offs have been established.ObjectiveTo develop an alternative version of the TMA-93 and compare its performance with the original.MethodsWe prospectively recruited cognitively-unimpaired healthy controls (HC) and patients with mild cognitive impairment or mild dementia who had available AD biomarkers, all aged ≥50 years or older. HC participants received either version A or B of the TMA-93 in alternating order. Patients were administered both versions in a counterbalanced design-one before and the other after biomarker testing (interval <4 months). We compared total scores between HC groups, assessed test-retest reliability using intraclass correlation coefficients (ICC) in patients, and evaluated the discriminative performance of both versions in distinguishing biomarker-confirmed AD from sociodemographically matched HC.Results99 HC received version A (median age 68, 64.6% female) and 96 version B (median age 67, 62.5% female), with no sociodemographic differences. Test scores did not differ significantly between HC groups. Sixteen patients were biomarker-negative (median age 72, 37.5% female) and 51 were biomarker-positive (median age 74, 51% female). ICC showed good agreement overall (0.877; 95% CI: 0.800-0.924) and in the AD group (0.860; 95% CI: 0.755-0.920). Area under the curve for detecting AD was 0.95 for version A and 0.90 for version B, with no significant differences.ConclusionsThese findings support the equivalence of the new TMA-93 version B, facilitating its use in longitudinal evaluations.