Development and validation of an osteoporosis risk prediction model incorporating nighttime eating exposure
Chenxi Liu1,2, Yiping Tong2, Mingxia Pan2
1Luoyang Orthopedic Hospital of Henan Province (Henan Provincial Orthopedic Hospital), Luoyang, China.
Introduction:
Osteoporosis (OP) is a global bone metabolic disease, and its incidence continues to rise with the intensifying aging of the population. In recent years, chrono-nutrition research has uncovered the significant impact of dietary timing patterns on health, but the relationship between nighttime eating exposure (NEE) and OP remains unclear.This study aimed to confirm through multi-dimensional methodologies that NEE is an independent risk factor for OP.
Methods:
A clinical cross-sectional study (n = 186) employed LASSO regression, logistic regression, and random forest to screen for OP predictors, constructing and validating a nomogram model. Analysis of the US NHANES database (n = 18,975) assessed the association between NEE and OP and explored non-linear relationships using weighted logistic regression. Multivariable Mendelian randomization (MR) analysis inferred causal associations using genetic instruments for dietary patterns, BMI, and sleep duration.
Results:
NEE was identified as an independent risk factor for OP. In the clinical prediction model, the combination model including NEE (NEE + bone mineral density + age, etc.) demonstrated optimal predictive performance (training set AUC = 0.808). NHANES validation showed a significantly increased OP risk when NEE > 25% (OR = 1.83, 95% CI: 1.27-2.64), with a dose-effect threshold identified (steep risk increase at NEE = 10-25%). Subgroup analysis revealed that the effect size of NEE was significantly higher in individuals with osteopenia (OR = 1.12) compared to those with normal bone mass (OR = 1.06, interaction p = 0.0297), and the protective effect of bone mineral density was weakened in the osteopenic group. Multivariable MR results indicated that a genetic predisposition representing dietary disturbance was positively associated with OP risk (β = 0.0228, p = 0.0033), reflecting the potential causal role of NEE-related behavioral patterns.
Discussion:
NEE is a novel behavioral risk factor for OP, posing a significant hazard particularly for osteopenic individuals. Its mechanism may be related to circadian rhythm disruption, metabolic dysregulation, and melatonin suppression. This research provides a new direction for precision OP prevention strategies based on "chrono-nutrition."
