RESUMEN
BACKGROUND: Disease modifying treatments for dementia are only just surfacing, and their development is still significantly hindered by the lack of validated tools for identification of subjects with subclinical disease. Much interest has been taken in developing accessible non-invasive serum biomarkers of neurodegeneration. Recent studies have identified caspase-3-cleaved tau (Tau-C), and ADAM-10 cleaved tau (Tau-A) as possible markers of preclinical neurodegenerative disease. OBJECTIVES: To explore if serum levels of Tau-A and Tau-C change as a consequence of neurodegeneration. DESIGN AND SETTING: Cohort study with measurement of biomarkers and genome sequencing at baseline with follow-up after an average of 14 years. PARTICIPANTS: Postmenopausal Danish women from the Prospective Epidemiological Risk Factor (PERF) cohort (n=4968) Methods: Genotyping data was used to perform a Mendelian randomization analysis of serum levels of Tau-A and Tau-C in relation to a diagnosis of dementia at follow-up. A dementia diagnosis was defined as a composite of an all-cause dementia diagnosis derived from the Danish National Health registries, a self-reported diagnosis of dementia and/or cognitive test scores suggestive of dementia. Serum levels of Tau-A and Tau-C were measured blinded in samples from baseline in a CAP certified lab. The association with dementia was assessed using bi-directional one- and two- sample Mendelian randomization. RESULTS: A lead single nucleotide polymorphism (SNP) was identified for Tau-A (rs10414043) and Tau-C (rs429358), respectively were identified. Both were located in the APOE/C1 cluster on chromosome 19. APOE and EPOC1 variants were associated with lower levels of Tau-A and Tau-C levels - effect size -0.13, 95%CI [-0.17 - -0.09] log2 (ng/mL), p=7.05e-11 for rs10414043 association with Tau-A and effect size -0.12, 95%CI [-0.15 - -0.08] log2 (ng/mL), p=2e-11 for rs429358 association with Tau-C. When incorporating genetic data from a larger genetic study we found that Alzheimer's disease was marginally associated with a decreased Tau-A and Tau-C levels (Odds Ratio 0.97, 95%CI [0.93 - 1.00]. No association was found in the forward Mendelian randomization analysis. CONCLUSIONS: By combining genotype data with serum measurements of the novel biomarkers Tau-A and Tau-C, we conclude that Tau-A and Tau-C levels change because of neurodegeneration. We also conclude that lower serum-values of the biomarkers are associated with the presence of genetic variants commonly found in individuals suffering from late-onset Alzheimer's Dementia. These findings add to the growing data pointing towards Tau-A and Tau-C as valuable biomarkers for neurodegeneration.