Association between ultra-processed food intake and osteoarthritis prevalence among U.S. adults : A cross-sectional study using NHANES data
- 1Changsha Central Hospital, Changsha, Hunan, China
- 2Xiangya Second Hospital of Central South University, Changsha, Hunan, China.
Abstract
This study examines the association between ultra-processed food (UPF) intake and osteoarthritis (OA) prevalence among U.S. adults aged 40 years and older. Data for this study were from the U.S. National Health and Nutrition Examination Survey between 2015 and 2020. Multivariable logistic regression models, adjusted for sociodemographic, lifestyle, and clinical factors, were employed to assess the association between UPF intake (categorized into quartiles using the NOVA classification) and OA prevalence. Restricted cubic spline analysis evaluated the dose-response relationship, and machine learning models were applied for predictive modeling. The final analysis included a weighted sample representing approximately 216,413,969 U.S. adults aged ≥40 years. Fully adjusted multivariable logistic regression revealed a positive association between higher UPF intake quartiles and OA prevalence (adjusted odds ratio [aOR] for highest vs lowest quartile = 1.24, 95% confidence interval: 1.05–1.48, P = .01). Restricted cubic spline analysis indicated a linear dose-response relationship (P for linearity = .02). Subgroup analysis identified a statistically significant interaction by body mass index category (P for interaction = .033), with a significant association in the obesity subgroup (odds ratio = 1.16, 95% confidence interval: 1.05–1.30; P = .006). Among machine learning models, XGBoost demonstrated superior predictive accuracy (area under the receiver operating characteristic curve = 0.96). Shapley Additive Explanation analysis identified age, body mass index, and UPF intake as the most influential predictors. Higher UPF consumption was associated with higher OA prevalence. Prospective studies are needed to examine temporal relationships and causal mechanisms.
Related
- Association between serum 25-hydroxyvitamin D and osteoarthritis prevalence in US adults: a cross-sectional study using NHANES 2007–2023
- Ultra-Processed Food Intake Is Associated with Dietary Quality, Dental Caries, and Constipation-Related Symptoms Among Preschool Children in Taiwan: A Cross-Sectional Study
- Prevalence of Mild Autonomous Cortisol Secretion Among Patients With Adrenal Incidentaloma and Its Association With Metabolic Abnormalities: A Hospital-Based, Analytical, Cross-Sectional Study
- Physical activity, sleep duration, and multimorbidity: a cross-sectional study of prevalence associations and association rule mining of multimorbidity patterns
- Association of dietary fiber intake with current asthma and recent asthma outcomes: a cross-sectional analysis of National Health and Nutrition Examination Survey 2017–March 2020 pre-pandemic data
BibTeX
@article{liang2026association,
title = {{Association between ultra-processed food intake and osteoarthritis prevalence among U.S. adults : A cross-sectional study using NHANES data}},
author = {Hongming Liang and Hongchun Peng and Li Shen and Xin Liu and Cheng Li and Liwen Xue},
journal = {Medicine},
year = {2026},
month = oct,
doi = {10.1097/md.0000000000051343},
url = {https://europepmc.org/article/PMC/PMC13727231},
note = {CC BY},
}
RIS
TY - JOUR
TI - Association between ultra-processed food intake and osteoarthritis prevalence among U.S. adults : A cross-sectional study using NHANES data
AU - Hongming Liang
AU - Hongchun Peng
AU - Li Shen
AU - Xin Liu
AU - Cheng Li
AU - Liwen Xue
PY - 2026
DA - 2026/10/30
JO - Medicine
DO - 10.1097/md.0000000000051343
UR - https://europepmc.org/article/PMC/PMC13727231
AB - This study examines the association between ultra-processed food (UPF) intake and osteoarthritis (OA) prevalence among U.S. adults aged 40 years and older. Data for this study were from the U.S. National Health and Nutrition Examination Survey between 2015 and 2020. Multivariable logistic regression models, adjusted for sociodemographic, lifestyle, and clinical factors, were employed to assess the association between UPF intake (categorized into quartiles using the NOVA classification) and OA prevalence. Restricted cubic spline analysis evaluated the dose-response relationship, and machine learning models were applied for predictive modeling. The final analysis included a weighted sample representing approximately 216,413,969 U.S. adults aged ≥40 years. Fully adjusted multivariable logistic regression revealed a positive association between higher UPF intake quartiles and OA prevalence (adjusted odds ratio [aOR] for highest vs lowest quartile = 1.24, 95% confidence interval: 1.05–1.48, P = .01). Restricted cubic spline analysis indicated a linear dose-response relationship (P for linearity = .02). Subgroup analysis identified a statistically significant interaction by body mass index category (P for interaction = .033), with a significant association in the obesity subgroup (odds ratio = 1.16, 95% confidence interval: 1.05–1.30; P = .006). Among machine learning models, XGBoost demonstrated superior predictive accuracy (area under the receiver operating characteristic curve = 0.96). Shapley Additive Explanation analysis identified age, body mass index, and UPF intake as the most influential predictors. Higher UPF consumption was associated with higher OA prevalence. Prospective studies are needed to examine temporal relationships and causal mechanisms.
SN - 0304-5412
SN - 1536-5964
SN - 0025-7974
ER -