Premium
Determinants of cognitive decline and dementia in stage 2: The SCIENCe project
Author(s) -
van Der Flier Wiesje
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.040263
Subject(s) - dementia , medicine , cohort , proportional hazards model , neuropsychology , cognitive decline , hazard ratio , oncology , disease , neuroimaging , demographics , cohort study , cognition , psychiatry , confidence interval , demography , sociology
Background We initiated the SCIENCe cohort to study the earliest brain changes leading to Alzheimer’s disease in individuals with subjective cognitive decline (SCD). We aimed to identify which individuals are most likely to progress to dementia using imaging and non‐imaging markers. Methods SCIENCe is an ongoing memory clinic cohort, enrolling cognitively normal individuals with SCD. In addition to the standardized diagnostic work‐up (medical exam, neuropsychology, MRI, and lumbar puncture) we offer amyloid PET to participants. Patients are followed annually. Incident dementia was used as outcome. We classified 693 individuals with SCD (60±9yr, 41%F, MMSE 28±2) according to ATN based on CSF, PET and MRI. We measured plasma abeta42/40 ratio with SIMOA. We used Cox proportional hazards models to determine associations with risk of dementia, and to develop prognostic models, providing individualized risks. Results Compared to A‐T‐N‐ (56%), participants in the Alzheimer continuum (A+; 18%) had an increased risk of dementia (HR=17, 95%CI 3.6‐79), while the risk in the non‐AD pathologic change profiles (27%) was not increased (HR=3, 95%CI 0.6‐18). Within the A+ profiles, abnormal T and N further increased the risk of dementia. Using Cox regression to develop prognostic models (n=481), we found that the model including CSF biomarkers had a Harrell’s c of 0.82. Using CSF biomarkers, the prognostic model reclassified 58% of individuals with uncertain outcome based on demographics only, as either high or low risk. Finally, in 241 participants with paired CSF and plasma amyloid measurements, we found that both lower CSF abeta42 and lower plasma ab42/40 ratio were associated with steeper rate of cognitive decline. On visual inspection, effect sizes of CSF and plasma amyloid associations with cognitive decline were in the same order of magnitude. Conclusion In individuals with SCD, only 18% could be labeled as stage 2 Alzheimer’s disease, based on positive amyloid markers. Those with stage 2 AD are at increased risk of dementia, even more so when there is additional tau pathology and/or neurodegeneration. In the future, it is feasible that biomarkers could be used for individualized risk profiling, and blood based biomarkers could offer a reasonable alternative for CSF and/or PET.