Comparing the Two Prodromal Parkinson's Disease Research Criteria—Lessons for Future Studies

Yılmaz, Rezzak; Suenkel, Ulrike; Team, Trend Study; Postuma, Ronald B.; Berg, Daniela

doi:10.1002/mds.28637

Cited by 4 publications

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While all of these participants were PD-free at the first visit, in total n = 24 incident PD cases were clinically diagnosed at follow-up based on UKBB and MDS diagnostic criteria [15]. The visit at which the conversion to PD occurred was considered.…”

Section: Overview Of the Trend Study Datamentioning

confidence: 99%

“…For each variable the probability to take a specific value, dependent on the values of its parents in the network, is inferred from the data. Notably, age (younger or older than 65 years) as well as risk and prodromal markers of PD have been discretized such that all variables indicate the presence or absence (or borderline status) of a marker in an individual TREND participant, as published previously for the TREND cohort [15,16] and suggested by the MDS research criteria for prodromal PD [3,6]. For each variable a CPT was estimated while learning the overall BN from data (See S2 Fig in S1 Appendix for CPT plots of each marker).…”

Section: Plos Onementioning

confidence: 99%

See 1 more Smart Citation

Bayesian network modeling of risk and prodromal markers of Parkinson’s disease

Sood

Suenkel²,

Thaler³

et al. 2023

PLoS ONE

Self Cite

View full text Add to dashboard Cite

Parkinson’s disease (PD) is characterized by a long prodromal phase with a multitude of markers indicating an increased PD risk prior to clinical diagnosis based on motor symptoms. Current PD prediction models do not consider interdependencies of single predictors, lack differentiation by subtypes of prodromal PD, and may be limited and potentially biased by confounding factors, unspecific assessment methods and restricted access to comprehensive marker data of prospective cohorts. We used prospective data of 18 established risk and prodromal markers of PD in 1178 healthy, PD-free individuals and 24 incident PD cases collected longitudinally in the Tübingen evaluation of Risk factors for Early detection of NeuroDegeneration (TREND) study at 4 visits over up to 10 years. We employed artificial intelligence (AI) to learn and quantify PD marker interdependencies via a Bayesian network (BN) with probabilistic confidence estimation using bootstrapping. The BN was employed to generate a synthetic cohort and individual marker profiles. Robust interdependencies were observed for BN edges from age to subthreshold parkinsonism and urinary dysfunction, sex to substantia nigra hyperechogenicity, depression, non-smoking and to constipation; depression to symptomatic hypotension and excessive daytime somnolence; solvent exposure to cognitive deficits and to physical inactivity; and non-smoking to physical inactivity. Conversion to PD was interdependent with prior subthreshold parkinsonism, sex and substantia nigra hyperechogenicity. Several additional interdependencies with lower probabilistic confidence were identified. Synthetic subjects generated via the BN based representation of the TREND study were realistic as assessed through multiple comparison approaches of real and synthetic data. Altogether our work demonstrates the potential of modern AI approaches (specifically BNs) both for modelling and understanding interdependencies between PD risk and prodromal markers, which are so far not accounted for in PD prediction models, as well as for generating realistic synthetic data.

show abstract

Section: Overview Of the Trend Study Datamentioning

confidence: 99%

Section: Plos Onementioning

confidence: 99%

Bayesian network modeling of risk and prodromal markers of Parkinson’s disease

Sood

Suenkel²,

Thaler³

et al. 2023

PLoS ONE

Self Cite

View full text Add to dashboard Cite

show abstract

“…A first update of the criteria was presented in 2019 [3]. Application of the original and updated criteria in different cohorts including our own evaluation in the Bruneck Study yielded high specificity, but variable sensitivity and positive predictivity for detecting cases of incident PD [5][6][7][8][9].…”

Section: Introductionmentioning

confidence: 99%

Comparison of different risk scores for Parkinson disease in a population‐based 10‐year study

Marini

Seppi

Kiechl

et al. 2023

Euro J of Neurology

View full text Add to dashboard Cite

Background and purposeDifferent algorithms aiming to identify individuals at risk of Parkinson disease (PD) have been proposed. Comparative studies of these scores and their recent updates in the general elder population are needed.MethodsWe have previously applied the “basic” PREDICT‐PD algorithm, designed for remote screening, and the original and updated Movement Disorder Society (MDS) criteria for prodromal PD to the longitudinal population‐based Bruneck study cohort. We have now additionally employed the “enhanced” PREDICT‐PD algorithm (which includes motor assessment, olfaction, probable rapid eye movement sleep behaviour disorder status, pesticide exposure, and diabetes as additional factors). Risk scores were calculated based on comprehensive baseline assessments (2005) in 574 subjects aged 55–94 years (290 females), and cases of incident PD were identified at 5‐year (n = 11) and 10‐year follow‐up (n = 9). We analysed the association of the different log‐transformed risk scores with incident PD at follow‐up (calculated per 1‐SD unit change).ResultsThe enhanced PREDICT‐PD algorithm was associated with incident PD over 10‐years of follow‐up, yielding higher odds for incident PD (odds ratio [OR] = 4.61, 95% confidence interval [CI] = 2.68–7.93, p < 0.001) compared with the basic PREDICT‐PD score (OR = 2.38, 95% CI = 1.49–3.79, p < 0.001). The updated MDS prodromal criteria yielded a numerically higher OR of 7.13 (95% CI = 3.49–14.54, p < 0.001) in comparison with the original criteria as well as the enhanced PREDICT‐PD algorithm, with overlapping 95% CIs.ConclusionsThe enhanced PREDICT‐PD algorithm was significantly associated with incident PD. The consistent performance of both the enhanced PREDICT‐PD algorithm and the updated MDS prodromal criteria compared to their original versions supports their use in PD risk screening.

show abstract

Challenges and Future of the International Parkinson and Movement Disorder Society Prodromal Parkinson's Disease Criteria: Are We On the Right Track?

Kulcsarova,

Skorvanek

2024

Movement Disorders

View full text Add to dashboard Cite

Comparing the Two Prodromal Parkinson's Disease Research Criteria—Lessons for Future Studies

Cited by 4 publications

References 7 publications

Bayesian network modeling of risk and prodromal markers of Parkinson’s disease

Bayesian network modeling of risk and prodromal markers of Parkinson’s disease

Comparison of different risk scores for Parkinson disease in a population‐based 10‐year study

Challenges and Future of the International Parkinson and Movement Disorder Society Prodromal Parkinson's Disease Criteria: Are We On the Right Track?

Contact Info

Product

Resources

About