Abstract Fall-related hip fractures are a major public health issue. While individual-level risk assessment tools exist, population-level predictive models could catalyze innovation in large-scale interventions. This study presents a hierarchical probabilistic model that predicts population-level hip fracture risk based on Factor of Risk (FOR) principles. Model validation demonstrated that FOR output aligned with a published dataset categorized by sex and hip fracture status. The model predicted normalized FOR for 100000 individuals simulating the Canadian older-adult population. Predicted hip fracture risk was higher for females (by an average of 38%), and increased with age (by15% per decade). Potential applications are discussed.

中文翻译：

---

预测人群水平的髋部骨折风险：一种结合概率方法和风险因素因素的新型分层模型

摘要 与跌倒相关的髋部骨折是一个重大的公共卫生问题。虽然存在个人层面的风险评估工具，但人口层面的预测模型可以促进大规模干预措施的创新。本研究提出了一个分层概率模型，该模型根据风险因素 (FOR) 原则预测人群水平的髋部骨折风险。模型验证表明 FOR 输出与按性别和髋部骨折状态分类的已发布数据集一致。该模型预测了模拟加拿大老年人口的 100000 个人的归一化 FOR。女性髋部骨折的预测风险较高（平均为 38%），并且随着年龄的增长而增加（每十年增加 15%）。讨论了潜在的应用。