Zhang Jingyuan, Wang Yuqi, Guo Jun, Liu Heng, Lei Zhuolin, Cheng Shouquan, Cao Hong
Department of Traumatic Orthopedics, Renmin Hospital, Hubei University of Medicine, Shiyan, 442000, People's Republic of China.
Department of Traumatic Orthopedics, Weifang Yidu Central Hospital, Shiyan, 250000, People's Republic of China.
Sci Rep. 2025 Mar 31;15(1):10994. doi: 10.1038/s41598-025-94218-4.
Osteoporosis and osteopenia pose substantial public health challenges, particularly among postmenopausal women. Although various anthropometric measures have been proposed for risk assessment, their predictive performance and nonlinear relationships with bone outcomes remain unclear. We analyzed data from six consecutive cycles (2007-2018) of the National Health and Nutrition Examination Survey (NHANES). A total of 4473 postmenopausal women aged ≥ 50 years were included, each classified as having normal bone density, osteopenia, or osteoporosis based on dual-energy X-ray absorptiometry measurements. Ten anthropometric indices-waist-to-height ratio (WTHR), conicity index (CI), a body shape index (ABSI), body roundness index (BRI), lipid accumulation product, visceral adiposity index, cardiometabolic index, atherogenic index of plasma, weight-adjusted-waist index (WWI), and triglyceride-glucose (TyG) index-were examined using multiple logistic regression, restricted cubic spline curves, threshold-effect analyses, and SHAP (SHapley Additive exPlanations) analysis to assess their associations with osteoporosis and osteopenia. In fully adjusted models, individuals with higher WTHR, CI, BRI, WWI, and TyG tended to have a lower likelihood of osteoporosis or osteopenia (p < 0.01), whereas those with elevated ABSI were more likely to be diagnosed with these conditions (p < 0.01). Notably, several indices showed nonlinear effects on bone outcomes. Receiver operating characteristic (ROC) analyses indicated that WTHR and BRI had the highest discriminative capacity for differentiating osteoporosis from osteopenia. SHAP analysis further highlighted WTHR and BRI as the most influential predictors. WTHR and BRI demonstrated strong predictive utility for osteoporosis and osteopenia, suggesting their potential as noninvasive, cost-effective screening metrics for postmenopausal bone health. These findings provide insight into the varying roles of central and general obesity indicators in bone density and underscore the importance of incorporating novel anthropometric indices into early-stage risk assessment.
骨质疏松症和骨质减少给公共卫生带来了重大挑战，尤其是在绝经后女性中。尽管已经提出了各种人体测量方法用于风险评估，但其预测性能以及与骨骼健康状况的非线性关系仍不明确。我们分析了美国国家健康与营养检查调查（NHANES）连续六个周期（2007 - 2018年）的数据。总共纳入了4473名年龄≥50岁的绝经后女性，根据双能X线吸收法测量结果，将她们每个人分为骨密度正常、骨质减少或骨质疏松。使用多因素逻辑回归、受限立方样条曲线、阈值效应分析和SHAP（Shapley值加法解释）分析，对十个身体测量指标——腰高比（WTHR）、锥度指数（CI）、体型指数（ABSI）、身体圆润度指数（BRI）、脂质积聚产物、内脏脂肪指数、心脏代谢指数、血浆致动脉粥样硬化指数、体重调整腰围指数（WWI）和甘油三酯-葡萄糖（TyG）指数——进行了研究，以评估它们与骨质疏松症和骨质减少的关联。在完全调整模型中，WTHR、CI、BRI、WWI和TyG较高的个体患骨质疏松症或骨质减少的可能性往往较低（p < 0.01），而ABSI升高的个体更有可能被诊断出患有这些疾病（p < 0.01）。值得注意的是，几个指标对骨骼健康状况显示出非线性影响。受试者工作特征（ROC）分析表明，WTHR和BRI在区分骨质疏松症和骨质减少方面具有最高的判别能力。SHAP分析进一步突出了WTHR和BRI是最具影响力的预测指标。WTHR和BRI对骨质疏松症和骨质减少具有很强的预测效用，表明它们作为绝经后骨骼健康的非侵入性、具有成本效益的筛查指标的潜力。这些发现深入了解了中心性肥胖和总体肥胖指标在骨密度中的不同作用，并强调了将新的人体测量指标纳入早期风险评估的重要性。
