Correlation between the physical indicators and bone mineral density in middle-aged and elderly people based on the process of generating "latent variables" in structural equation modeling
Objective To explore the correlation between five physical indicators,namely height,weight,thigh length,waist circumference,hip circumference,and bone mineral density in middle-aged and older adults aged 50 years and older,and to provide a reference for developing interventions.Methods A population of 2 072 surveyed individuals aged 50 and older with complete data on physical indicators and bone density form 2017-2020 National Health and Nutrition Examination Survey database was selected as the study population.First,five related physical indicators,namely height,weight,thigh length,waist circumference,and hip circumference,were downscaled using structural equation modeling(SEM)"latent variable"generation process.Second,Body_index,gender,and ethnicity,generated after the downscaling process,were used as the independent variables to analyze the correlation between them and total spinal bone mineral density(BMD).Results The new variable Body_index was fitted to the independent variables height,weight,thigh length,waist circumference,and hip circumference in a multiple linear fit(R2 of 0.998)with less than 2.0％loss of information.Multiple linear regression analysis revealed that the new variable Body_index could be considered to have a linear correlation with total spinal BMD(P＜0.001)and that gender and race could not yet be considered to have a linear relationship with total spinal BMD(P＞0.05).Conclusion Short stature and thinness are risk factors for low BMD in middle-aged and elderly people,suggesting that high-risk groups should increase their body weight appropriately and eat a reasonable diet to prevent osteoporosis and fractures.
Bone mineral densityPhysicalCorrelationMiddle-aged and elderly population
万方数据
