按体重指数分层的中心性肥胖与老年女性2型糖尿病发病风险
基于妇女健康倡议(WHI)133,620名基线无2型糖尿病的女性前瞻性分析,并按70%与30%随机分为训练集与测试集,采用Cox模型评估各BMI及BMI特异性腰围(WC)分类的2型糖尿病风险。结果显示腰围在BMI分层内可进一步区分风险,外部队列HRS与HANDLS方向一致,但模型区分度仅为中等(C统计量约0.60-0.75)。
要点
- BMI分层内腰围更高者2型糖尿病风险更高,可补充BMI进行风险分层
- 联合BMI-腰围分类的校正风险比范围为1.40-4.43,BMI分类为1.51-3.30
- 模型个体水平区分度中等,腰围不宜单独作为2型糖尿病预测工具
核心结论
- 在老年女性中,BMI特异性腰围分类可在BMI之外细化肥胖相关2型糖尿病风险分层,但个体水平区分度仅为中等。
- 腰围可作为BMI的补充,用于识别需要更全面代谢风险评估的女性,而非独立的2型糖尿病预测工具。
研究设计与样本
- 前瞻性分析纳入133,620名基线无2型糖尿病的妇女健康倡议(WHI)参与者。
- 参与者随机分为训练样本(70%;n=93,463)和测试样本(30%;n=40,157)。
- 外部验证纳入健康与退休研究(HRS;n=2,885)和HANDLS研究(n=1,037)。
- 采用Cox模型估计各BMI及BMI特异性腰围分类的2型糖尿病风险,并以Harrell C统计量评估区分度。
主要发现
- WHI中2型糖尿病风险随BMI分类升高而递增,校正风险比为1.51-3.30。
- 联合BMI-腰围分类的校正风险比为1.40-4.43。
- 在同一BMI分层内,腰围较高通常对应更高的2型糖尿病风险。
- HRS与HANDLS中关联方向一致。
- 年龄对关联存在修饰作用,较年轻年龄组梯度更强;种族和族裔交互作用无统计学意义。
临床怎么用
- 对老年女性可在BMI基础上加测腰围,以细化2型糖尿病风险分层。
- 腰围升高提示需要更全面的代谢风险评估,但不应单独用于预测2型糖尿病。
- 风险分层时应考虑年龄,较年轻女性中腰围相关风险梯度可能更明显。
局限性
- 模型个体水平区分度仅为中等(C统计量约0.60-0.75)。
- 摘要未报告腰围分类的具体切点、效应值置信区间及随访时长。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
CONTEXT: Body mass index (BMI) does not capture central adiposity, which may provide additional information for type 2 diabetes (T2D) risk stratification. OBJECTIVE: To evaluate BMI-specific waist circumference (WC) categories in relation to incident T2D and assess their performance across independent cohorts. METHODS: Prospective analyses included 133,620 Women's Health Initiative (WHI) participants free of T2D at baseline, randomly assigned to training (70%; n=93,463) and test (30%; n=40,157) samples. External validation included the Health and Retirement Study (HRS; n=2,885) and Healthy Aging in Neighborhoods of Diversity across the Life Span (HANDLS; n=1,037). Cox models estimated T2D risk across BMI and BMI-specific WC categories. Discrimination was assessed using Harrell's C-statistic. RESULTS: In WHI, T2D risk increased progressively across BMI categories (adjusted hazard ratios [HRs], 1.51-3.30) and joint BMI-WC categories (HRs, 1.40-4.43). Within BMI strata, higher WC generally identified greater T2D risk. Associations were directionally consistent in HRS and HANDLS. Age modified associations, with stronger gradients at younger ages, whereas race and ethnicity interactions were not statistically significant. Model discrimination was modest (C-statistics approximately 0.60-0.75). CONCLUSIONS: BMI-specific WC categories refine adiposity-related T2D risk stratification beyond BMI alone but provide modest individual-level discrimination. WC may complement BMI in identifying women warranting more comprehensive metabolic risk assessment rather than serving as a stand-alone T2D prediction tool.
引用
- Beydoun MA, Beydoun HA, Manson JE, et al. Body Mass Index-Specific Central Adiposity and Risk of Incident Type 2 Diabetes Among Aging Women. J Clin Endocrinol Metab. 2026. PMID: 42963906. doi:10.1210/clinem/dgag443
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。