Autonomic Burden in Unstaged Prodromal Participants in Ppmi Cohort

Technical Summary: Autonomic Burden in Unstaged Prodromal Participants in PPMI Cohort

Problem Statement
The Parkinson's Progression Markers Initiative (PPMI) and the emerging Neuronal Alpha-Synuclein Disease (NSD) staging system utilize alpha-synuclein seed amplification assays (SAA) to biologically define Parkinsonian synucleinopathy. However, a specific subgroup of biomarker-positive prodromal participants remains "unstaged" because they lack an analyzable NSD stage assignment. There is a critical interpretive gap: it is unclear whether these biologically positive but formally unstaged individuals carry clinically meaningful disease burden or if they represent a negligible phenotype. Specifically, the extent to which this group exhibits autonomic dysfunction—a key non-motor feature of synucleinopathy—relative to healthy controls, biomarker-negative prodromal participants, and formally staged disease (Biological Stages 2 and 3) has not been rigorously quantified.

Methodology
This retrospective observational analysis utilized merged clinical and SAA data from the PPMI cohort. The study population (N=3,151 at baseline) was stratified into five prespecified groups:

1. Healthy Controls (HC; n=315).
2. Prodromal unstaged SAA-negative (n=796).
3. Prodromal unstaged SAA-positive (n=1,089).
4. Biological Stage 2 (n=263).
5. Biological Stage 3 (n=688).

The primary outcome was the SCOPA-AUT (Scale for Outcomes in Parkinson's Disease for Autonomic Symptoms) total score, with secondary outcomes including subdomain scores (gastrointestinal, urinary, cardiovascular, thermoregulatory, pupillomotor, sexual, and orthostatic).

Statistical Approach:

• Cross-sectional Analysis: Baseline differences were assessed using HC3-robust linear regression models adjusted for age, sex, and years of education. Prespecified contrasts compared the Prodromal unstaged SAA-positive group against all other groups.
• Longitudinal Analysis: Linear mixed-effects models with participant-level random intercepts were used to estimate trajectories over time. Fixed effects included baseline group, time, group-by-time interaction, and covariates.
• Sensitivity and Robustness: The analysis included multiple sensitivity checks:
  • Adjustment for Parkinson's disease treatment status and Levodopa Equivalent Daily Dose (LEDD).
  • "Off-medication" subset analyses to rule out treatment-induced symptom suppression.
  • Broader confounder adjustments (BMI, mood, cognition, motor severity).
  • Distributional robustness checks (negative binomial, median quantile, quadratic-time GEE).
  • Pattern-mixture models to address attrition bias.
  • False-discovery-rate (FDR) correction for domain-level comparisons.

Key Results

• Baseline Autonomic Burden: Prodromal unstaged SAA-positive participants exhibited significantly higher baseline autonomic burden than Healthy Controls (Beta = 3.96, p = 1.96e-36) and Biological Stage 2 participants (Beta = 2.80, p = 3.04e-13). Their burden was statistically indistinguishable from Biological Stage 3 (Beta = 0.14, p = 0.657) in the primary model.
• Domain Specificity: The elevated burden in the SAA-positive unstaged group was selective rather than uniform. Significant differences from Healthy Controls were observed across gastrointestinal, urinary, cardiovascular, thermoregulatory, sexual, and orthostatic domains, but not pupillomotor. Notably, within the prodromal cohort, SAA-positive participants showed distinct elevations in gastrointestinal, thermoregulatory, and orthostatic domains compared to SAA-negative prodromal participants, despite the total SCOPA-AUT score difference being non-significant in that specific pairwise comparison.
• Longitudinal Trajectories: The SAA-positive unstaged group demonstrated a faster rate of worsening over time compared to Healthy Controls (Slope difference = 0.42 points/year, p = 1.15e-16) and SAA-negative prodromal participants (Slope difference = 0.48 points/year, p = 1.56e-19). Their rate of decline was slightly slower than Biological Stage 2 but not significantly different from Biological Stage 3.
• Robustness: The finding that the unstaged SAA-positive group carries higher burden than Biological Stage 2 persisted after adjusting for medication, in off-medication subsets, and across various sensitivity models. The comparison with Biological Stage 3 was more model-sensitive; while non-significant in primary models, some broader confounder-adjusted models suggested a higher burden in the unstaged group.

Significance and Claims
The paper argues that the absence of a formal NSD stage assignment should not be interpreted as an absence of clinically meaningful disease burden. The primary claim is that biomarker-positive, unstaged prodromal participants represent a distinct phenotype with autonomic dysfunction that is:

1. Clinically Significant: Substantially elevated above healthy aging and Biological Stage 2.
2. Progressive: Exhibiting a trajectory of worsening that exceeds healthy controls and SAA-negative prodromal participants.
3. Refinement, Not Redefinition: The authors explicitly state these findings support a refinement of early disease stratification rather than a biological redefinition of the staging system. They do not claim that autonomic burden alone defines a new formal stage or that all unstaged participants are equivalent to Stage 3.

The study concludes that current staging frameworks may under-capture a clinically relevant prodromal phenotype enriched for autonomic dysfunction. These participants represent a meaningful group for early non-motor progression studies and future research into biologically anchored prodromal classification. The authors caution that the results are based on a single cohort without external replication and that the relationship between biomarker positivity, autonomic dysfunction, and stage progression remains phenotypic rather than mechanistic.