Trajectories of early to mid-life adulthood BMI and incident diabetes: the China Health and Nutrition Survey
Trajectories of early to mid-life adulthood BMI and incident diabetes: the China Health and Nutrition Survey
复制标题
成年早期至中年体重指数和糖尿病发生的轨迹：中国健康与营养调查
DOI：
10.1136/bmjdrc-2019-000972
复制
发表时间：
2020-01-01
影响因子：
4.1
通讯作者：
Zhang, Tao
中科院分区：
文献类型：
--
作者：
Lv, Jiali;Fan, Bingbing;Zhang, Tao
本纵向研究旨在描述成年早期到中年不同的身体质量指数（BMI）轨迹，并探讨从青年到中年BMI变化与糖尿病发病率之间的关系。研究设计和方法本研究纳入了7289名成年人，他们在1989-2011年期间反复测量BMI 3-9次，并获得了糖尿病发生率的信息。使用潜在类别增长混合模型（LCGMM）来识别不同的BMI轨迹。Cox比例风险模型用于研究轨迹组成员与高血糖发生率之间的关系，并对协变量进行调整。高血糖组包括糖尿病前期或糖尿病患者。模型估计的BMI水平和斜率分别根据模型参数及其一阶导数在1年间隔的每个年龄点计算。使用Logistic回归分析来检验模型估计的每个年龄点的BMI水平和斜率与高血糖事件的关系。曲线下面积（AUC）是根据每个个体随访期间的纵向生长曲线模型计算的。在进行逻辑回归分析之前，计算总AUC值、基线AUC值和增量AUC值的四分位数。结果LCGMM具有低增长组（n=5136）、中增长组（n=1914）和高增长组（n=239） 3种不同的轨迹。与低增长组相比，中增长组和高增长组的校正hr和95% ci分别为1.21（0.99 ~ 1.48）和1.56（1.06 ~ 2.30）。模型估计BMI水平的调整标准化or值在20-50岁之间增加，范围从0.98（0.87 - 1.10）到1.19（1.08 - 1.32）。20 ~ 29 a期间，水平调整线性边坡的标准化ORs从1.30（1.16 ~ 1.45）逐渐增大到1.42 (1.21 ~ 1.67)，30 ~ 43 a期间从1.41（1.20 ~ 1.66）减小到1.20 (1.08 ~ 1.33)，50 a期间逐渐增大到1.20（1.04 ~ 1.38）。调整协变量后，第四个四分位数的增量AUC （OR=1.31, 95% CI 1.03至1.66）与第一个四分位数相比具有显著性。这些发现表明，成年早期的BMI轨迹与晚年糖尿病有显著相关性。青年期是糖尿病发展的关键时期，这对早期预防具有重要意义。
Introduction This longitudinal study aims to characterize distinct body mass index (BMI) trajectories during early to mid-life adulthood and to explore the association between BMI change from young adulthood to midlife and incident diabetes. Research design and methods This study included 7289 adults who had repeatedly measured BMI 3–9 times during 1989–2011 and information on incident diabetes. Latent class growth mixed model (LCGMM) was used to identify different BMI trajectories. Cox proportional hazard models were used to investigate the association between the trajectory group membership and incident hyperglycemia, adjusting for covariates. The hyperglycemia group included individuals with prediabetes or diabetes. The model-estimated BMI levels and slopes were calculated at each age point in 1-year intervals according to the model parameters and their first derivatives, respectively. Logistic regression analyses were used to examine the association of model-estimated levels and slopes of BMI at each age point with incident hyperglycemia. The area under the curve (AUC) was computed from longitudinal growth curve models during the follow-up for each individual. Prior to the logistic regression analyses, quartiles of total, baseline, and incremental AUC values were calculated. Results Three distinct trajectories were characterized by LCGMM, comprising of low-increasing group (n=5136), medium-increasing group (n=1914), and high-increasing group (n=239). Compared with the low-increasing group, adjusted HRs and 95% CIs were 1.21 (0.99 to 1.48) and 1.56 (1.06 to 2.30) for the medium-increasing and the high-increasing group, respectively. The adjusted standardized ORs of model-estimated BMI levels increased among 20–50 years, ranging from 0.98 (0.87 to 1.10) to 1.19 (1.08 to 1.32). The standardized ORs of level-adjusted linear slopes increased gradually from 1.30 (1.16 to 1.45) to 1.42 (1.21 to 1.67) during 20–29 years, then decreased from 1.41 (1.20 to 1.66) to 1.20 (1.08 to 1.33) during 30–43 years, and finally increased to 1.20 (1.04 to 1.38) until 50 years. The fourth quartile of incremental AUC (OR=1.31, 95% CI 1.03 to 1.66) was significant compared with the first quartile, after adjustment for covariates. Conclusions These findings indicate that the BMI trajectories during early adulthood were significantly associated with later-life diabetes. Young adulthood is a crucial period for the development of diabetes, which has implications for early prevention.