OBJECTIVES Literature regarding dietary protein intake and risk of falls is limited to a few studies with relatively small sample sizes and short follow-ups, which have reported contrasting findings. Thus, we investigated whether dietary protein intake is associated with risk of falls in a large cohort of North American adults. DESIGN Data were drawn from the Osteoarthritis Initiative, a cohort study, with 8 years of follow-up. SETTING AND PARTICIPANTS Community-dwelling adults with knee osteoarthritis or at high risk for this condition. METHODS Dietary protein intake was recorded using the Block Brief 2000 food frequency questionnaire and categorized using gender-specific quartiles (Q). Falls were self-reported in response to the question "Did you fall during the past year?" categorized as yes vs no and made during the 6 visits over 8 years of follow-up. Results are reported as relative risks (RRs), with their 95% confidence intervals (CIs), using a multivariable Poisson regression. RESULTS The final sample consisted of 4450 adults (mean age 61.2 years, females = 59.6%). Higher dietary protein intake was significantly associated with higher frequency of falls during the year before baseline. After adjusting for 17 potential confounders, people with the greatest amount of protein intake (Q4) had a significantly higher risk of falling over the 8-year follow-up period (RR 1.112, 95% CI 1.027-1.211, P = .009) than those with the lowest protein intake (Q1). CONCLUSIONS/IMPLICATIONS In this cohort of people affected by knee osteoarthritis or at high risk for this condition, high dietary protein intake may increase the risk of falls in older people, but further research is needed to confirm or refute these findings.

中文翻译：

---

老年人的膳食蛋白质摄入量和跌倒：来自骨关节炎倡议的纵向分析

目标关于膳食蛋白质摄入量和跌倒风险的文献仅限于少数样本量相对较小且随访时间较短的研究，这些研究报告了对比结果。因此，我们调查了大量北美成年人的膳食蛋白质摄入量是否与跌倒风险相关。设计数据来自骨关节炎倡议，一项队列研究，随访 8 年。地点和参与者 患有膝骨关节炎或这种情况的高危人群的社区成人。方法 使用 Block Brief 2000 食物频率问卷记录膳食蛋白质摄入量，并使用特定性别的四分位数 (Q) 进行分类。跌倒是在回答“您在过去一年跌倒了吗？”这个问题时自我报告的。分类为是与否，并在 8 年随访期间的 6 次访问中进行。使用多变量泊松回归将结果报告为相对风险 (RR) 及其 95% 置信区间 (CI)。结果 最终样本包括 4450 名成年人（平均年龄 61.2 岁，女性 = 59.6%）。较高的膳食蛋白质摄入量与基线前一年中较高的跌倒频率显着相关。在调整了 17 个潜在混杂因素后，蛋白质摄入量最大的人（Q4）在 8 年随访期间跌倒的风险显着更高（RR 1.112，95% CI 1.027-1.211，P = .009）比那些蛋白质摄入量最低的人（Q1）。结论/意义 在这群受膝骨关节炎影响或有此病症高风险的人群中，