Predicting Sleep Quality Based on Metabolic, Body Composition, and Physical Fitness Variables in Aged People: Exploratory Analysis with a Conventional Machine Learning Model.
Pedro Forte, Samuel G Encarnação, José E Teixeira, Luís Branquinho, Tiago M Barbosa, António M Monteiro, Daniel Pecos-Martín
{"title":"Predicting Sleep Quality Based on Metabolic, Body Composition, and Physical Fitness Variables in Aged People: Exploratory Analysis with a Conventional Machine Learning Model.","authors":"Pedro Forte, Samuel G Encarnação, José E Teixeira, Luís Branquinho, Tiago M Barbosa, António M Monteiro, Daniel Pecos-Martín","doi":"10.3390/jfmk10030337","DOIUrl":null,"url":null,"abstract":"<p><p><b>Background</b>: Sleep plays a crucial role in the health of older adults, and its quality is influenced by multiple physiological and functional factors. However, the relationship between sleep quality and physical fitness, body composition, and metabolic markers remains unclear. This exploratory study aimed to investigate the associations between sleep quality and physical, metabolic, and body composition variables in older adults, and to evaluate the preliminary performance of a logistic regression model in classifying sleep quality. <b>Methods</b>: A total of 32 subjects participated in this study, with a mean age of 69. The resting arterial pressure (systolic and diastolic), resting heart rate, anthropometrics (high waist girth), body composition (by bioimpedance), and physical fitness (Functional Fitness Test) and sleep quality (Pitsburg sleep-quality index) were evaluated. Group comparisons, associative analysis and logistic regression with 5-fold stratified cross-validation was used to classify sleep quality based on selected non-sleep-related predictors. <b>Results</b>: Individuals with good sleep quality showed significantly better back stretch (t = 2.592; <i>p</i> = 0.015; η<sup>2</sup> = 0.239), lower limb strength (5TSTS; t = 2.564; <i>p</i> = 0.016; η<sup>2</sup> = 0.476), and longer total sleep time (t = 6.882; <i>p</i> < 0.001; η<sup>2</sup> = 0.675). Exploratory correlations showed that poor sleep quality was moderately associated with reduced lower-limb strength and mobility. The logistic regression model including 5TSTS and TUG achieved a mean accuracy of 0.76 ± 0.15, precision of 0.79 ± 0.18, recall of 0.83 ± 0.21, and AUC of 0.74 ± 0.16 across cross-validation folds. <b>Conclusions</b>: These preliminary findings suggest that physical fitness and clinical variables significantly influence sleep quality in older adults. Sleep-quality-dependent patterns suggest that interventions to improve lower limb strength may promote better sleep outcomes.</p>","PeriodicalId":16052,"journal":{"name":"Journal of Functional Morphology and Kinesiology","volume":"10 3","pages":""},"PeriodicalIF":2.5000,"publicationDate":"2025-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12452653/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Functional Morphology and Kinesiology","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.3390/jfmk10030337","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"SPORT SCIENCES","Score":null,"Total":0}
引用次数: 0
Abstract
Background: Sleep plays a crucial role in the health of older adults, and its quality is influenced by multiple physiological and functional factors. However, the relationship between sleep quality and physical fitness, body composition, and metabolic markers remains unclear. This exploratory study aimed to investigate the associations between sleep quality and physical, metabolic, and body composition variables in older adults, and to evaluate the preliminary performance of a logistic regression model in classifying sleep quality. Methods: A total of 32 subjects participated in this study, with a mean age of 69. The resting arterial pressure (systolic and diastolic), resting heart rate, anthropometrics (high waist girth), body composition (by bioimpedance), and physical fitness (Functional Fitness Test) and sleep quality (Pitsburg sleep-quality index) were evaluated. Group comparisons, associative analysis and logistic regression with 5-fold stratified cross-validation was used to classify sleep quality based on selected non-sleep-related predictors. Results: Individuals with good sleep quality showed significantly better back stretch (t = 2.592; p = 0.015; η2 = 0.239), lower limb strength (5TSTS; t = 2.564; p = 0.016; η2 = 0.476), and longer total sleep time (t = 6.882; p < 0.001; η2 = 0.675). Exploratory correlations showed that poor sleep quality was moderately associated with reduced lower-limb strength and mobility. The logistic regression model including 5TSTS and TUG achieved a mean accuracy of 0.76 ± 0.15, precision of 0.79 ± 0.18, recall of 0.83 ± 0.21, and AUC of 0.74 ± 0.16 across cross-validation folds. Conclusions: These preliminary findings suggest that physical fitness and clinical variables significantly influence sleep quality in older adults. Sleep-quality-dependent patterns suggest that interventions to improve lower limb strength may promote better sleep outcomes.