People with preclinical Alzheimer's disease do not all slide downhill at the same pace. Analyzing participants who entered one of two related studies without symptoms, researchers at the Keck School of Medicine of USC identified three distinct cognitive trajectories: stable, slow decline and fast decline. Roughly 70 percent of participants stayed stable across the study period of about six years. The work, funded in part by the National Institutes of Health, was published in Alzheimer's & Dementia: The Journal of the Alzheimer's Association.
That variability tends to disappear in conventional analyses. As corresponding author Michael Donohue, professor of neurology and associate director of biostatistics at the USC Epstein Family Alzheimer's Therapeutic Research Institute, put it, most studies report the average across participants, which creates the impression that everyone is worsening at a uniform rate. Averaging, he said, masks major differences between individuals and makes the disease look more predictable than it is.
Earlier research had hinted that decline rates differ from person to person, but this is among the first efforts to connect those patterns to biomarker data. The team tested whether specific blood tests and brain scans could sort people into the likely-stable group versus those likely to worsen. Their models did so with about 70 percent accuracy - useful, but far from definitive.
The practical stakes are twofold. With refinement, a tool like this could give someone newly diagnosed a more realistic sense of what to expect rather than a generic prognosis. It could also sharpen clinical trials: if current trials assume every participant is on the same downward path, they may be testing drugs against an oversimplified version of the disease, potentially obscuring real effects in the subgroups most likely to decline.