Identification and prediction of Parkinson's disease subtypes and progression using machine learning in two cohorts.

The clinical manifestations of Parkinson's disease (PD) are characterized by heterogeneity in age at onset, disease duration, rate of progression, and the constellation of motor versus non-motor features. There is an unmet need for the characterization of distinct disease subtypes as well as improved, individualized predictions of the disease course. We used unsupervised and supervised machine learning methods on comprehensive, longitudinal clinical data from the Parkinson's Disease Progression Marker Initiative (n = 294 cases) to identify patient subtypes and to predict disease progression. The resulting models were validated in an independent, clinically well-characterized cohort from the Parkinson's Disease Biomarker Program (n = 263 cases). Our analysis distinguished three distinct disease subtypes with highly predictable progression rates, corresponding to slow, moderate, and fast disease progression. We achieved highly accurate projections of disease progression 5 years after initial diagnosis with an average area under the curve (AUC) of 0.92 (95% CI: 0.95 ± 0.01) for the slower progressing group (PDvec1), 0.87 ± 0.03 for moderate progressors, and 0.95 ± 0.02 for the fast-progressing group (PDvec3). We identified serum neurofilament light as a significant indicator of fast disease progression among other key biomarkers of interest. We replicated these findings in an independent cohort, released the analytical code, and developed models in an open science manner. Our data-driven study provides insights to deconstruct PD heterogeneity. This approach could have immediate implications for clinical trials by improving the detection of significant clinical outcomes. We anticipate that machine learning models will improve patient counseling, clinical trial design, and ultimately individualized patient care.

Frailty Risks of Prescription Analgesics and Sedatives across Frailty Models: the Health and Retirement Study.

INTRODUCTION: Limited evidence for incident frailty risks associated with prescription analgesics and sedatives in older (≥ 65 years) community-living adults prompted a more comprehensive investigation. METHODS: We used data from older Health and Retirement Study respondents and three frailty models (frailty index, functional domain, frailty phenotype with 8803, 10,470, and 6850 non-frail individuals, respectively) and estimated sub-hazard ratios of regular prescription drug use (co-use, analgesic use, and sedative use), by frailty model. We addressed confounding with covariate adjustment and propensity score matching approaches. RESULTS: The baseline prevalence of analgesic and sedative co-use, analgesic use, and sedative use among non-frail respondents was 1.8%, 12.8%, and 4.7% for the frailty index model, 4.2%, 16.2%, and 5.3% for the functional domain model, and 4.3%, 15.4%, and 6.1% for the frailty phenotype model, respectively. Cumulative frailty incidence over 10 years was 39.3%, 36.1%, and 14.2% for frailty index, functional domain, and frailty phenotype models, respectively; covariate-adjusted sub-hazard ratio estimates were 2.00 (1.63-2.45), 1.83 (1.57-2.13), and 1.68 (1.21-2.33) for co-use; 1.72 (1.56-1.89), 1.38 (1.27-1.51), and 1.51 (1.27-1.79) for analgesic use; and 1.46 (1.24-1.72), 1.25 (1.07-1.46), and 1.31 (0.97-1.76) for sedative use. Frailty risk ranking (co-use > analgesic use > sedative use) persisted across all model sensitivity analyses. DISCUSSION: Consistently significant frailty risk estimates of regular prescription analgesic and sedative co-use and of prescription analgesic use support existing clinical, public health, and regulatory guidance on opioid and benzodiazepine co-prescription, on opioid prescription, and on NSAID prescription. Frailty phenotype measurement administration limited power to detect significant frailty risks. Research into specific pharmaceutical exposures and comparison of results across cohorts will be required to contribute to the deprescribing evidence base.

Sex and Age Differences in Ambulatory Blood Pressure Monitoring Tolerability.

BACKGROUND: Clinical practice guidelines endorse using ambulatory blood pressure monitoring (ABPM) for the diagnosis and management of hypertension. However, ABPM is not always tolerated by patients, and differences between individuals according to age and sex remain unexplored. METHODS: This is a post hoc analysis of a prospective, single-arm clinical trial (NCT03920956) that evaluated the feasibility of an ABPM service provided at 2 community pharmacies. Tolerability was assessed using a previously published survey, which included 7 yes/no questions and 8 answered on a scale of 0-10. Descriptive statistics and Chi-square analyses were used to summarize the data for the patient surveys and to describe sex and age differences in device tolerability. RESULTS: Of the 52 subjects enrolled, 50 (96%) completed the survey; half were female with a mean (SD) age of 57.5 years (15.8). Chi-square analyses showed that compared with their male counterparts, females were more likely to find the monitor cumbersome to wear (76.2% vs. 40%, P = 0.014). Subjects under 55 years of age were more likely to be disturbed by the noise of the monitor during driving (38.1% vs. 4.2%, P = 0.005) and at other times (35.0% vs. 8.3%, P = 0.029), and to find the monitor embarrassing to wear (33.3% vs. 7.1%, P = 0.019). CONCLUSIONS: Although ABPM was generally well-tolerated overall, we did identify age and sex differences in tolerability. These factors should be considered to ensure patient acceptance and tolerability of ABPM.