Prevalence of preclinical Alzheimer disease: Comparison of current classification systems

To determine the prevalence of preclinical Alzheimer disease (AD) according to current classification systems by examining CSF from a representative general population sample of 70-year-olds from Gothenburg, Sweden.

The sample was derived from the population-based H70 Gothenburg Birth Cohort Studies in Gothenburg, Sweden. The participants (n = 322, age 70 years) underwent comprehensive neuropsychiatric, cognitive, and somatic examinations. CSF levels of β-amyloid (Aβ)₄₂, Aβ₄₀, total tau, and phosphorylated tau were measured. Preclinical AD was classified according to criteria of the A/T/N system, Dubois 2016, National Institute on Aging-Alzheimer's Association (NIA-AA) criteria, and International Working Group-2 (IWG-2) criteria. Individuals with Clinical Dementia Rating score >0 were excluded, leaving 259 cognitively unimpaired individuals.

The prevalence of amyloid pathology was 22.8%, of total tau pathology was 33.2%, and of phosphorylated tau pathology was 6.9%. With the A/T/N system, the prevalence of A+/T-/N- was 13.1%, A+/T-/N+ was 7.3%, A+/T+/N+ was 2.3%, A-/T-/N+ was 18.9%, and A-/T+/N+ was 4.6%. When the Dubois criteria were applied, the prevalence of asymptomatic at risk for AD was 36.7% and of preclinical AD was 9.7%. With the NIA-AA criteria, the prevalence of stage 1 was 13.1% and stage 2 was 9.7%. With the IWG-2 criteria, the prevalence of asymptomatic at risk for AD was 9.7%. The APOE ε4 allele was associated with several of the categories. Men more often had total tau pathology, A+/T-/N+, preclinical AD according to Dubois 2016, asymptomatic at risk for AD according to the IWG-2 criteria, and NIA-AA stage 2.

The prevalence of pathologic AD markers was very common (46%) in a representative population sample of 70-year-olds. The clinical implications of these findings need to be scrutinized further in longitudinal studies.