Development and Validation of a Prognostic Model for Cognitive Impairment in Parkinson's Disease With REM Sleep Behavior Disorder

A nomogram based on neuron-specific enolase and substantia nigra hyperechogenicity for identifying cognitive impairment in Parkinson's disease

Background

One in four individuals with Parkinson's disease (PD) experience cognitive impairment (CI). However, few practical models integrating clinical and neuroimaging biomarkers have been developed to address CI in PD. This study aimed to evaluate the correlation between circulating neuron-specific enolase (NSE) levels, substantia nigra hyperechogenicity (SNH), and cognitive function in PD and to develop a nomogram based on clinical and neuroimaging biomarkers for predicting CI in patients with PD.

Methods

A total of 385 patients with PD who underwent transcranial sonography (TCS) from January 2021 to December 2022 at Beijing Tiantan Hospital, Capital Medical University, were recruited as the training cohort. For validation, 165 patients with PD treated from January 2023 to December 2023 were enrolled. Data for SNH, plasma NSE, and other clinical measures were collected, and cognitive function was assessed using the Montreal Cognitive Assessment (MoCA). Logistic regression analysis was employed to select potential risk factors and establish a nomogram. The receiver operating characteristic curve and calibration curve were generated to evaluate the performance of the nomogram.

Results

Patients with PD exhibiting CI displayed advanced age, elevated Unified PD Rating Scale-III (UPDRS-III) score, an increased percentage of SNH, higher levels of plasma NSE and homocysteine (Hcy), a larger SNH area, and lower education levels compared to PD patients without CI. Gender [odds ratio (OR) =0.561, 95% confidence interval (CI): 0.330-0.954, P=0.03], age (OR =1.039; 95% CI: 1.011-1.066; P=0.005), education level (OR =0.892; 95% CI: 0.842-0.954; P<0.001), UPDRS-III scores (OR =1.026; 95% CI: 1.009-1.043; P=0.003), plasma NSE concentration (OR =1.562; 95% CI: 1.374-1.776; P<0.001), and SNH (OR =0.545; 95% CI: 0.330-0.902; P=0.02) were independent predictors of CI in patients with PD. A nomogram developed using these six factors yielded a moderate discrimination performance with an area under the curve (AUC) of 0.823 (95% CI 0.781-0.864; P<0.001). The calibration curve demonstrated acceptable agreement between predicted outcomes and actual values. Validation further confirmed the reliability of the nomogram, with an AUC of 0.864 (95% CI: 0.805-0.922; P<0.001).

Conclusions

The level of NSE in plasma and the SNH assessed by TCS are associated with CI in patients with PD. The proposed nomogram has the potential to facilitate the detection of cognitive decline in individuals with PD.

Prediction of Cognitive Decline in Parkinson's Disease Using Clinical and DAT SPECT Imaging Features, and Hybrid Machine Learning Systems

BACKGROUND

We aimed to predict Montreal Cognitive Assessment (MoCA) scores in Parkinson's disease patients at year 4 using handcrafted radiomics (RF), deep (DF), and clinical (CF) features at year 0 (baseline) applied to hybrid machine learning systems (HMLSs).

METHODS

297 patients were selected from the Parkinson's Progressive Marker Initiative (PPMI) database. The standardized SERA radiomics software and a 3D encoder were employed to extract RFs and DFs from single-photon emission computed tomography (DAT-SPECT) images, respectively. The patients with MoCA scores over 26 were indicated as normal; otherwise, scores under 26 were indicated as abnormal. Moreover, we applied different combinations of feature sets to HMLSs, including the Analysis of Variance (ANOVA) feature selection, which was linked with eight classifiers, including Multi-Layer Perceptron (MLP), K-Neighbors Classifier (KNN), Extra Trees Classifier (ETC), and others. We employed 80% of the patients to select the best model in a 5-fold cross-validation process, and the remaining 20% were employed for hold-out testing.

RESULTS

For the sole usage of RFs and DFs, ANOVA and MLP resulted in averaged accuracies of 59 ± 3% and 65 ± 4% for 5-fold cross-validation, respectively, with hold-out testing accuracies of 59 ± 1% and 56 ± 2%, respectively. For sole CFs, a higher performance of 77 ± 8% for 5-fold cross-validation and a hold-out testing performance of 82 + 2% were obtained from ANOVA and ETC. RF+DF obtained a performance of 64 ± 7%, with a hold-out testing performance of 59 ± 2% through ANOVA and XGBC. Usage of CF+RF, CF+DF, and RF+DF+CF enabled the highest averaged accuracies of 78 ± 7%, 78 ± 9%, and 76 ± 8% for 5-fold cross-validation, and hold-out testing accuracies of 81 ± 2%, 82 ± 2%, and 83 ± 4%, respectively.

CONCLUSIONS

We demonstrated that CFs vitally contribute to predictive performance, and combining them with appropriate imaging features and HMLSs can result in the best prediction performance.

Machine learning within the Parkinson's progression markers initiative: Review of the current state of affairs

The Parkinson's Progression Markers Initiative (PPMI) has collected more than a decade's worth of longitudinal and multi-modal data from patients, healthy controls, and at-risk individuals, including imaging, clinical, cognitive, and 'omics' biospecimens. Such a rich dataset presents unprecedented opportunities for biomarker discovery, patient subtyping, and prognostic prediction, but it also poses challenges that may require the development of novel methodological approaches to solve. In this review, we provide an overview of the application of machine learning methods to analyzing data from the PPMI cohort. We find that there is significant variability in the types of data, models, and validation procedures used across studies, and that much of what makes the PPMI data set unique (multi-modal and longitudinal observations) remains underutilized in most machine learning studies. We review each of these dimensions in detail and provide recommendations for future machine learning work using data from the PPMI cohort.

Predictors of cognitive impairment in newly diagnosed Parkinson's disease with normal cognition at baseline: A 5-year cohort study

Background and objective

Cognitive impairment (CI) is a substantial contributor to the disability associated with Parkinson's disease (PD). We aimed to assess the clinical features and explore the underlying biomarkers as predictors of CI in patients with newly diagnosed PD (NDPD; less than 2 years).

Methods

We evaluated the cognitive function status using the Montreal Cognitive Assessment (MoCA) and a battery of neuropsychological tests at baseline and subsequent annual follow-up for 5 years from the Parkinson's Progression Markers Initiative (PPMI) database. We assessed the baseline clinical features, apolipoprotein (APO) E status, β-glucocerebrosidase (GBA) mutation status, cerebrospinal fluid findings, and dopamine transporter imaging results. Using a diagnosis of CI (combined mild cognitive impairment and dementia) developed during the 5-year follow-up as outcome measures, we assessed the predictive values of baseline clinical variables and biomarkers. We also constructed a predictive model for the diagnosis of CI using logistic regression analysis.

Results

A total of 409 patients with NDPD with 5-year follow-up were enrolled, 232 with normal cognitive function at baseline, and 94 patients developed CI during the 5-year follow-up. In multivariate analyses, age, current diagnosis of hypertension, baseline MoCA scores, Movement disorder society Unified PD Rating Scale part III (MDS-UPDRS III) scores, and APOE status were associated with the development of CI. Predictive accuracy of CI using age alone improved by the addition of clinical variables and biomarkers (current diagnosis of hypertension, baseline MoCA scores, and MDS-UPDRS III scores, APOE status; AUC 0.80 [95% CI 0.74-0.86] vs. 0.71 [0.64-0.77], p = 0.008). Cognitive domains that had higher frequencies of impairment were found in verbal memory (12.6 vs. 16.8%) and attention/processing speed (12.7 vs. 16.9%), however, no significant difference in the prevalence of CI at annual follow-up was found during the 5-year follow-up in NDPD patients.

Conclusion

In NDPD, the development of CI during the 5-year follow-up can be predicted with good accuracy using a model combining age, current diagnosis of hypertension, baseline MoCA scores, MDS-UPDRS III scores, and APOE status. Our study underscores the need for the earlier identification of CI in NDPD patients in our clinical practice.

Acute Effects of Different Exercise Intensities on Executive Function and Oculomotor Performance in Middle-Aged and Older Adults: Moderate-Intensity Continuous Exercise vs. High-Intensity Interval Exercise

A wealth of evidence has shown that a single bout of aerobic exercise can facilitate executive function. However, none of current studies on this topic have addressed whether the magnitude of the acute-exercise benefit on executive function and oculomotor performance is influenced by different aerobic exercise modes. The present study was thus aimed toward an investigation of the acute effects of high-intensity interval exercise (HIIE) vs. moderate-intensity continuous exercise (MICE) on executive-related oculomotor performance in healthy late middle-aged and older adults. Using a within-subject design, twenty-two participants completed a single bout of 30 min of HIIE, MICE, or a non-exercise-intervention (REST) session in a counterbalanced order. The behavioral [e.g., reaction times (RTs), coefficient of variation (CV) of the RT], and oculomotor (e.g., saccade amplitude, saccade latency, and saccadic peak velocity) indices were measured when participants performed antisaccade and prosaccade tasks prior to and after an intervention mode. The results showed that a 30-min single-bout of HIIE and MICE interventions shortened the RTs in the antisaccade task, with the null effect on the CV of the RT in the late middle-aged and older adults. In terms of oculomotor metrics, although the two exercise modes could not modify the performance in terms of saccade amplitudes and saccade latencies, the participants’ saccadic peak velocities while performing the oculomotor paradigm were significantly altered only following an acute HIIE intervention. The present findings suggested that a 30-min single-bout of HIIE and MICE interventions modulated post-exercise antisaccade control on behavioral performance (e.g., RTs). Nevertheless, the HIIE relative MICE mode appears to be a more effective aerobic exercise in terms of oculomotor control (e.g., saccadic peak velocities) in late middle-aged and older adults.