# Predicting re-fracture risk factors in older adult osteoporotic vertebral fractures patients with comorbidities: development and validation of nomogram

## TL;DR

This study developed a new tool to predict the risk of re-fracture in older adults with osteoporotic vertebral fractures, considering comorbidities and lifestyle factors.

## Contribution

A novel nomogram integrating comorbidities and lifestyle factors for personalized re-fracture risk prediction in OVCF patients.

## Key findings

- The nomogram showed strong discrimination with AUCs of 0.886 in training and 0.827 in testing.
- Key predictors included tumor history, scoliosis, mental disorders, alcohol use, and chronic kidney disease.
- Hypertension showed a paradoxical protective association, requiring further investigation.

## Abstract

Osteoporotic vertebral compression fractures (OVCFs) pose a significant health burden in older adult populations, with postoperative re-fracture (re.fra) complicating recovery. Existing models (e.g., FRAX, QFracture) inadequately address comorbidities and modifiable lifestyle factors. This study aimed to develop and validate a novel nomogram integrating these underrecognized yet critical predictors for personalized risk stratification.

A retrospective cohort of 560 older adult OVCF patients undergoing percutaneous vertebroplasty (PVP) was analyzed. Patients were randomly divided into training (70%, n = 392) and testing (30%, n = 168) cohorts. Univariable and backward stepwise multivariable logistic regression identified independent re.fra predictors. A nomogram was developed and internally validated using area under the curve (AUC), calibration curves (slopes, intercepts), Brier scores, and decision curve analysis (DCA) to assess discrimination, calibration, and clinical utility.

Independent predictors included tumor history [adjusted odds ratio (aOR) = 12.29, 95% CI: 2.50–60.35], scoliosis (aOR = 6.46, 95% CI: 2.97–14.03), mental disorders (aOR = 5.91, 95% CI: 2.73–12.82), alcohol use ≥10 years (aOR = 3.69, 95% CI: 1.90–7.17), and chronic kidney disease (aOR = 3.12, 95% CI: 1.61–6.06). Hypertension exhibited a paradoxical protective association (aOR = 0.50, 95% CI: 0.27–0.93). The nomogram demonstrated strong discrimination [AUC:0.886 (training), 0.827 (testing)], excellent calibration in training (slope = 1.000, Brier = 0.118) with slight deviation in testing (slope = 0.697, Brier = 0.162), and superior net benefit over treat-all/none strategies across thresholds (DCA).

This validated nomogram integrates often-overlooked comorbidities and lifestyle factors to predict post-PVP re.fra risk, providing a practical tool for personalized management and highlighting the need for multidisciplinary care in high-risk subgroups such as those with scoliosis, mental disorders, or prolonged alcohol use. The intriguing protective association of hypertension, however, requires cautious interpretation and further investigation before clinical application.