A nomogram with online dynamic calculator for predicting osteoporosis: development and validation based on NHANES

Area of Science:

• Gerontology
• Public Health
• Medical Informatics

Background:

• Osteoporosis presents a significant health challenge in the elderly due to its insidious onset.
• Dual-energy X-ray absorptiometry (DXA) examinations are costly, necessitating accessible screening tools.
• The National Health and Nutrition Examination Survey (NHANES) provides valuable data for developing predictive models.

Purpose of the Study:

• To develop and validate a nomogram clinical prediction model for estimating osteoporosis risk in older adults.
• To leverage NHANES data, including bone mineral density (BMD) and nutritional information, for model creation.
• To provide an accessible tool for predicting the probability of osteoporosis in the elderly population.

Main Methods:

• Utilized NHANES data, splitting participants into training (3181) and validation (1622) groups.
• Employed univariate and multivariate logistic regression analyses on clinical characteristics and BMD.
• Constructed and validated general and dynamic nomogram models using ROC, calibration, and DCA curves.

Main Results:

• A nomogram model was developed using 11 variables: age, gender, race, PIR, waist circumference, DBP, exercise, protein, carbohydrate, caffeine intake, and fracture history.
• The model demonstrated moderate predictive value (AUC=0.795) with good calibration and clinical utility.
• An interactive, user-friendly online dynamic nomogram was created for practical application.

Conclusions:

• The developed nomogram model and its online dynamic version show significant utility for the U.S. elderly population.
• External validation in diverse cohorts is recommended prior to global implementation.
• This tool can enhance early osteoporosis prediction, diagnosis, and treatment, benefiting elderly bone health and public health initiatives.