Design: Prospective community cohort (ARIC-NCS)
Population: 1515 US adults with plasma biomarkers at mean ages 59 and 77
Primary outcome: 11-year change in memory, language, executive and global cognition
Effect: Midlife p-tau181: β −0.030 SD (−0.059 to −0.001) for memory; late-life p-tau181: β −0.091 (−0.123 to −0.058)

The ARIC Neurocognitive Study, reported in Neurology on 16 October, measured plasma Alzheimer biomarkers in 1515 people at a mean age of 59 and again at 77, then tracked cognition over the next 11 years.

Of the midlife markers, only p-tau181 was associated with later memory decline, and the effect was small. In late life, p-tau181, neurofilament light and GFAP were each linked to faster decline, most strongly in memory. Midlife amyloid ratio, NfL and GFAP did not predict decline.

The finding supports a long preclinical phase in which tau pathology begins decades before symptoms. It does not make midlife blood testing useful for individuals: the associations are modest at group level and there is no intervention proven to change the course.

For now, plasma biomarkers belong in the assessment of people with cognitive symptoms, not in screening healthy middle-aged adults.

Do not order plasma Alzheimer biomarkers in asymptomatic middle-aged adults; there is no action to follow.
In late life, p-tau, NfL and GFAP all tracked faster decline, most clearly in memory.
Use biomarkers to support diagnosis in symptomatic patients, alongside clinical assessment.
Midlife vascular risk control remains the evidence-based step for brain health.

Why it matters

It pushes the start of Alzheimer biology back to midlife, which shapes where prevention trials will look.

The statistics, in plain English

A coefficient of −0.030 SD (95% CI −0.059 to −0.001) per SD of midlife p-tau181 is a small effect at group level, and the upper limit is close to zero. The late-life association (−0.091 SD) was about three times stronger.