Prospective study of blood-based biomarkers and 20-year risk of clinically diagnosed Alzheimer's disease
Littlejohns, T.; Liu, W.; Maronga, C.; Tong, T. Y.; Amin, N.; Breeur, M.; Collister, J.; Parsaeian, M.; Papier, K.; Piazza, P.; Rockett, G.; Smith-Byrne, K.; Travis, R.; van Duijn, C.; Hunter, D.
Show abstract
Identifying individuals in the preclinical stages of Alzheimer's disease (AD) is necessary for inclusion into future prevention trials. AD pathology occurs in the brain 20 or more years before diagnosis. In a nested 1:1 matched case-control sample of 426 participants selected from 19,500 members of the EPIC-Oxford cohort, we found that higher blood-based brain-derived and total p-tau 181, 217, and 231, as well as GFAP, were associated with AD over up to 25 years of follow-up (median=19.4, interquartile range 16.8-21.9 years). Of these seven biomarkers, LASSO regression selected brain derived p-tau 217 as the strongest discriminator of AD cases from controls. The AUC for brain derived p-tau 217 accounting for age, sex, and time of blood draw was 0.80, which increased to 0.82, 0.83, 0.84, after further addition of 1) APOE-e4 carrier status, 2) sociodemographic and lifestyle factors, and 3) both, respectively. Blood-based biomarkers, including the novel brain-derived p-tau 217, could identify individuals at-risk of AD two decades pre-diagnosis.
Matching journals
The top 9 journals account for 50% of the predicted probability mass.