Verbal memory separated prodromal neuronal alpha synuclein disease from controls
Prodromal neuronal α synuclein disease did not behave like a broad cognitive syndrome. In this longitudinal Parkinson’s Precision Medicine Initiative cohort, the clearest signal landed in verbal memory, where participants with prodromal disease performed worse at baseline than healthy controls by an adjusted difference of −0.20 standard deviations, with a 95% confidence interval from −0.34 to −0.07 and an adjusted P value of .007. NSD-negative GBA1 or LRRK2 nonmanifesting carriers looked different: their cognitive performance did not diverge from healthy controls, arguing against a simple genetic-carrier effect and pointing instead toward the biologic state of prodromal synuclein disease itself.
The cohort was large enough to give the pattern real weight. Investigators analyzed 3,031 participants with a mean age of 66.6 years, and 43.6% were female. Healthy controls numbered 244, NSD-negative genetic carriers 323, prodromal NSD participants 1,640, and sporadic Parkinson disease 824. Cognitive modeling adjusted for age, sex, education, and APOE ε4 status, which matters because each of those factors can blur the line between early neurodegenerative change and background variation in performance. Here, the signal stayed narrow: no other cognitive domain was lower than healthy controls at baseline or over time.
One sentence captures the central finding: the disease did not spread a cognitive shadow across every domain at once.
Longitudinal change favored a memory-first pattern, but the trajectory was not fully stable
The longitudinal analysis pointed in the same direction, then grew less certain when the cohort was narrowed. In the primary age-adjusted model over four years, the verbal-memory trajectory differed from healthy controls, with P = .009. In a smaller age-matched sensitivity subset followed through year 3, that trajectory difference disappeared, with P = .62, while the baseline memory gap remained intact at −0.28 standard deviations, 95% confidence interval −0.46 to −0.10, P = .003.
That split is important because it suggests the most reproducible element of the dataset is not a rapid slope of decline, but a detectable baseline deficit in a preclinical or prodromal phase. In practical neurodegenerative terms, memory performance may be showing the first measurable friction from synuclein pathology before other cognitive systems become visibly involved. The biology fits a staged process: early network disruption in medial temporal and connected associative circuits can impair retrieval and encoding before executive, visuospatial, or language measures fall away. The data do not prove that mechanism, but they are consistent with it.
Mild cognitive impairment became more common as prodromal disease advanced
The probability of mild cognitive impairment increased in the prodromal group across follow-up. Adjusted MCI probability rose from 10.2% to 15.1% in prodromal NSD, while healthy controls moved from 8.4% to 5.0% and sporadic Parkinson disease from 13.4% to 17.7%. The increase in prodromal NSD was greater than in healthy controls, with a ratio of annual odds ratios of 1.28, a 95% confidence interval of 1.04 to 1.58, and an adjusted P value of .04. Against sporadic Parkinson disease, however, the increase did not differ significantly, with P = .62.
Those numbers place prodromal NSD in an intermediate clinical zone. Cognitive risk is not absent, and it is not yet the same as established Parkinson disease. The gradual rise in MCI probability suggests that the prodromal state already carries measurable cognitive vulnerability, even when only one domain is clearly shifted. For clinicians watching patients with hyposmia or isolated rapid eye movement sleep behavior disorder, the message is not that dementia is imminent. It is that subtle cognitive change may already be underway, especially in memory tasks that depend on early network integrity.
Why one domain may move first
Selective verbal-memory impairment is biologically plausible in synuclein disease because memory encoding and retrieval depend on distributed cortical-subcortical circuits that can be disrupted before more global dysfunction becomes visible. Early α-synuclein pathology may impair attention to new information, consolidation efficiency, and strategic retrieval, all of which would depress verbal-memory scores without necessarily touching processing speed, language, executive function, or visuospatial performance in the same time window. The absence of broad domain loss also argues against a nonspecific effect from age, education, or a general illness burden, since those factors would usually erode multiple cognitive measures at once.
This pattern also helps distinguish prodromal NSD from a carrier state alone. GBA1 and LRRK2 nonmanifesting carriers without NSD did not show the same baseline cognitive deficit, which implies that carrier status by itself is not sufficient to explain the memory signal in this cohort. The combination of prodromal clinical phenotype and biomarker-supported disease status appears to matter more than genotype alone.
A second sentence matters here: the memory change came first, and it came alone.
What the preprint can and cannot establish
This is a preprint, so the findings remain preliminary and have not undergone peer review. The study’s strongest result is the reproducible baseline verbal-memory difference, while the longitudinal trajectory signal became less convincing in the age-matched sensitivity subset. That tension should temper over-reading of progression estimates. The dataset still supports a clinically useful idea: cognitive involvement in prodromal neuronal α synuclein disease may begin as a selective memory phenotype rather than as global impairment.
For now, the most defensible bedside implication is surveillance. Patients recruited through hyposmia or isolated REM sleep behavior disorder cohorts may warrant closer cognitive attention even before broader deficits emerge, especially when memory performance slips relative to age and education expectations. Future follow-up will need to show whether the memory signal persists, deepens, or broadens into other cognitive domains as conversion to overt Parkinson disease unfolds.