Applicating Multivariate Adaptive Regression Splines (MARSplines) for Predicting Post-Traumatic Stress Disorder Risk in Military Personnel

Background. Risk prediction of post-traumatic stress disorder (PTSD), which has substantial social and economic consequences and affects military readiness, is an increasingly important challenge. Early identification of service members at elevated risk enables timely preventive strategies, including interventions based on physical culture, physical activity, sport, and rehabilitation. Objectives. This study aimed to develop and validate a predictive model for assessing PTSD risk in military personnel and to substantiate its use as a decision-support tool for pedagogical planning in physical education and sport-based prevention and rehabilitation programs. Materials and Methods. The study involved 4,403 military personnel (89.9% men) aged predominantly 21–50 years. PTSD severity was assessed using the Mississippi Scale for Combat-Related PTSD (military version), which demonstrated high internal consistency (Cronbach’s α = 0.892; Guttman split-half = 0.920). Statistical analyses included comparative and gender-based procedures, as well as machine learning modeling using Multivariate Adaptive Regression Splines (MARSplines) with progressive optimization, dataset balancing, and evaluation of classification performance. Results. A MARSplines-based model predicting PTSD risk was developed using 18 dichotomous demographic, behavioral, and service-related variables. The final model for male military personnel included 300 basis functions with an interaction order of six and demonstrated sensitivity of 71.1%, specificity of 69.5%, accuracy of 70.3%, and moderate agreement between predicted and observed risk (Cohen’s κ = 0.405). Conclusions. Machine learning–based prediction of PTSD risk represents a promising approach for early prevention and targeted intervention in military populations. The proposed model supports identification of high-risk individuals and informs the design of preventive and rehabilitative programs, including those based on physical culture, sport, and e-sports, aimed at stress reduction and resilience enhancement.

Authors

Institutions

Publication Details

Journal
Physical Education Theory and Methodology
Published
2026-01-30
DOI
https://doi.org/10.17309/tmfv.2026.1.10
Primary Topic
Posttraumatic Stress Disorder Research
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Applicating Multivariate Adaptive Regression Splines (MARSplines) for Predicting Post-Traumatic Stress Disorder Risk in Military Personnel

Kostiantyn Serhiienko, Olena O. Iakovenko, Myroslav Dutchak, Nataliia Byshevets et al.
Physical Education Theory and Methodology
Posttraumatic Stress Disorder Research
article

Applicating Multivariate Adaptive Regression Splines (MARSplines) for Predicting Post-Traumatic Stress Disorder Risk in Military Personnel

Kostiantyn Serhiienko, Olena O. Iakovenko, Myroslav Dutchak, Nataliia Byshevets, Maksym Yarmolenko, Olena Andrieieva, Oksana Shynkaruk
article en

Abstract

Background. Risk prediction of post-traumatic stress disorder (PTSD), which has substantial social and economic consequences and affects military readiness, is an increasingly important challenge. Early identification of service members at elevated risk enables timely preventive strategies, including interventions based on physical culture, physical activity, sport, and rehabilitation. Objectives. This study aimed to develop and validate a predictive model for assessing PTSD risk in military personnel and to substantiate its use as a decision-support tool for pedagogical planning in physical education and sport-based prevention and rehabilitation programs. Materials and Methods. The study involved 4,403 military personnel (89.9% men) aged predominantly 21–50 years. PTSD severity was assessed using the Mississippi Scale for Combat-Related PTSD (military version), which demonstrated high internal consistency (Cronbach’s α = 0.892; Guttman split-half = 0.920). Statistical analyses included comparative and gender-based procedures, as well as machine learning modeling using Multivariate Adaptive Regression Splines (MARSplines) with progressive optimization, dataset balancing, and evaluation of classification performance. Results. A MARSplines-based model predicting PTSD risk was developed using 18 dichotomous demographic, behavioral, and service-related variables. The final model for male military personnel included 300 basis functions with an interaction order of six and demonstrated sensitivity of 71.1%, specificity of 69.5%, accuracy of 70.3%, and moderate agreement between predicted and observed risk (Cohen’s κ = 0.405). Conclusions. Machine learning–based prediction of PTSD risk represents a promising approach for early prevention and targeted intervention in military populations. The proposed model supports identification of high-risk individuals and informs the design of preventive and rehabilitative programs, including those based on physical culture, sport, and e-sports, aimed at stress reduction and resilience enhancement.

Physical Education Theory and MethodologyVol. 26(1)
National University of Ukraine on Physical Education and Sport (UA)
Openalex Percentile: Top 87%
Posttraumatic Stress Disorder Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.