Sarma Shohinee, Bůžková Petra, Elam Rachel E, Fink Howard A, Cauley Jane A, Djoussé Luc, Barzilay Joshua, Mukamal Kenneth J
Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02215, United States.
Department of Biostatistics, University of Washington, Seattle, WA 98115, United States.
J Bone Miner Res. 2024 Nov 29;39(12):1735-1743. doi: 10.1093/jbmr/zjae142.
Type 2 diabetes mellitus and lower weight are both associated with osteoporotic fractures, but the roles of variability and trajectory are less clear. The associations of these factors among older adults with dysglycemia, who are at highest risk of fracture, with fracture risk and BMD remain uncertain. We followed 775 men and 1080 women from the Cardiovascular Health Study (mean age 77.4 years) with abnormal oral glucose tolerance testing in 1989-1990. We measured their weights yearly through 1994-1995 and derived intra-individual mean weight, weight slope, and weight variability. We also used growth mixture modeling to derive 4 latent BMI trajectories over time. We used Cox proportional hazards models to calculate hazard ratios (HRs) and 95% CI for subsequent hip fracture through 2015 and linear regression models to estimate cross-sectional associations with BMD of the hip. Each 10 kg higher mean weight was associated with a lower risk of subsequent hip fracture overall (HR 0.81; CI, 0.70-0.94) and among women (HR 0.76; CI, 0.64-0.91) and with higher BMD (p <.001). Higher weight variability was directly associated with incident hip fracture among women (HR 1.18; CI, 1.03-1.35). Compared with a stable trajectory, a "progressive overweight" trajectory was associated with lower risk of hip fracture (HR 0.66; CI, 0.44-0.99). An uncommon trajectory of "accelerating obesity" was associated with higher BMD. Among older adults with dysglycemia at high risk for fracture, lower mean weight is associated with higher fracture risk, but variability and trajectory may also contribute. These results highlight the complex effects of weight in older age.
2型糖尿病和体重较低均与骨质疏松性骨折有关,但变异性和轨迹的作用尚不清楚。在骨折风险最高的血糖异常的老年人中,这些因素与骨折风险和骨密度之间的关联仍不确定。我们对心血管健康研究中的775名男性和1080名女性(平均年龄77.4岁)进行了随访,这些人在1989 - 1990年口服葡萄糖耐量试验异常。我们在1994 - 1995年期间每年测量他们的体重,并得出个体内平均体重、体重斜率和体重变异性。我们还使用生长混合模型得出随时间变化的4种潜在体重指数轨迹。我们使用Cox比例风险模型计算至2015年后续髋部骨折的风险比(HRs)和95%置信区间(CI),并使用线性回归模型估计与髋部骨密度的横断面关联。总体而言,平均体重每增加10千克,后续髋部骨折风险较低(HR 0.81;CI,0.70 - 0.94),在女性中也是如此(HR 0.76;CI,0.64 - 0.91),且与较高的骨密度相关(p <.001)。较高的体重变异性与女性髋部骨折发生率直接相关(HR 1.18;CI,1.03 - 1.35)。与稳定轨迹相比,“渐进性超重”轨迹与较低的髋部骨折风险相关(HR 0.66;CI,0.44 - 0.99)。一种不常见的“加速肥胖”轨迹与较高的骨密度相关。在骨折风险高的血糖异常老年人中,较低的平均体重与较高的骨折风险相关,但变异性和轨迹也可能起作用。这些结果突出了体重在老年时的复杂影响。