Prodromal Parkinson's Disease Detection via Wrist Actigraphy

Introduction

The prodromal phase of Parkinson's disease — the period between the earliest detectable biological changes and the onset of clinical motor symptoms — can span 10–20 years. During this window, subtle changes in sleep architecture, circadian rhythm, and nocturnal motor activity may be detectable through wrist-worn accelerometers.

This article discusses the potential and limitations of actigraphy-based approaches for early neurodegenerative disease detection.

Context

The UK Biobank study collected 7-day wrist accelerometer recordings from approximately 100,000 participants. This dataset provides a unique opportunity to study the relationship between passively collected movement data and subsequent neurodegenerative disease diagnoses at population scale.

Previous work has demonstrated that REM Sleep Behavior Disorder (RBD) — characterized by loss of normal REM atonia and resultant dream-enactment behaviors — is among the strongest prodromal markers for alpha-synucleinopathies. Wrist accelerometers, while unable to detect REM sleep directly, can capture nocturnal movement patterns that may serve as proxies for sleep disruption.

Analysis

Key accelerometer-derived features relevant to prodromal neurodegeneration include:

• Sleep fragmentation index: Frequency and duration of nocturnal awakenings
• Intra-daily variability (IV): Regularity of the rest-activity cycle across 24 hours
• Nocturnal movement intensity: Acceleration magnitude during the primary sleep period
• Sleep onset and offset variability: Night-to-night consistency of sleep timing

These features can be combined with clinical variables (age, sex, family history, constipation, hyposmia) in survival analysis frameworks to estimate individualized conversion risk.

Clinical Implications

Wearable-based screening for prodromal neurodegeneration remains an active research area. Several important caveats must be acknowledged:

1. Specificity is limited. Many conditions besides prodromal PD cause sleep fragmentation and circadian disruption.
2. Clinical actionability is evolving. While disease-modifying therapies for early PD are under investigation, no approved intervention currently exists for the prodromal phase.
3. Population-level models may not translate to individual prediction. Strong cohort-level associations do not necessarily yield clinically useful individual-level predictions.

These limitations should inform how results are communicated to stakeholders and participants.

Conclusion

Wrist actigraphy offers a promising, non-invasive, and scalable approach to identifying individuals at elevated risk of neurodegenerative disease. The path from research signal to clinical utility, however, requires careful validation, transparent communication of uncertainty, and alignment with available clinical interventions.