Development and Evaluation of Multiscale Signal and Complexity Measures From ABPM Data for Health Risk Prediction

Author

• Roberto Arias, The University of Texas Rio Grande Valley

Date of Award

7-1-2026

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Statistics

First Advisor

Kristina Vatcheva

Second Advisor

Wilson-Zuniga Galindo

Third Advisor

Zhijun Qiao

Abstract

Ambulatory blood pressure (ABP) monitoring provides a 24-hour assessment of blood pressure and is widely used for hypertension diagnosis and risk stratification. Conventional ABP measures primarily rely on averages and linear summaries, potentially overlooking the dynamic, nonlinear, and multiscale characteristics of blood pressure regulation. This study developed novel features from 24-hour ABP recordings using Maximal Overlap Discrete Wavelet Transform (MODWT), entropy measures derived from reconstructed phase spaces, recurrence quantification analysis (RQA), and complementary distance-matrix-based approaches applied across MODWT scales.

The novel measures were evaluated alongside conventional ABP measures in both cross-sectional and longitudinal analyses of hypertension, dipping status, diabetes mellitus (DM), and cognitive outcomes, including Mini-Mental State Examination (MMSE) scores. Across outcomes, the extracted features formed three distinct groups: measures that underperformed conventional measures, measures that extended conventional measures by incorporating temporal information, and measures that captured independent information not reflected in traditional ABP summaries. Several multiscale features demonstrated improved discriminative and predictive performance relative to conventional measures. Distinct scale-dependent patterns emerged across outcomes, with hypertension and diabetes associated primarily with finer-scale dynamics, dipping status with coarser-scale vascular regulation, and cognitive outcomes exhibiting contributions from both short- and long-term dynamics. These findings suggest that multiscale ABP features capture clinically relevant information beyond conventional measures and may improve future cardiometabolic and cognitive risk stratification.

Comments

Copyright 2026 Roberto Arias. All rights reserved. https://proquest.com/docview/3409019623

Recommended Citation

Arias, R. (2026). Development and Evaluation of Multiscale Signal and Complexity Measures From ABPM Data for Health Risk Prediction [Doctoral dissertation, The University of Texas Rio Grande Valley]. ScholarWorks @ UTRGV. https://scholarworks.utrgv.edu/etd/2040