摘要目的:探讨中国成年人高密度脂蛋白胆固醇（HDL-C）与糖尿病患病风险的相关性。方法:本研究是一项多中心、回顾性队列研究的二次分析，数据来源于DATADRYAD数据库的中国健康体检筛查项目。2010—2016年对分布在中国的32个地点和11个城市211 833例≥20岁的中国成年人进行健康筛查，基线时无糖尿病。基线HDL-C水平为目标自变量，随访时发生糖尿病为因变量。Cox比例风险回归分析评估HDL-C水平与糖尿病患病风险之间的独立相关性。应用广义相加模型及平滑曲线拟合方法探讨非线性关系。进行亚组分析，评估不同亚组相关性的一致，进一步验证结果的可靠性。结果:调整年龄、性别、体重指数等潜在混杂因素后，HDL-C水平与糖尿病发生呈正相关（ HR=1.43，95% CI：1.08~1.90， P=0.012）。HDL-C水平与糖尿病患病风险呈非线性关系，其拐点为1.81 mmol/L。拐点左右两侧的 HR值（95% CI）分别为0.94（0.56~1.55）和2.54（1.93~3.30）。当HDL-C>1.81 mmol/L时，HDL-C与糖尿病发生呈正相关，HDL-C每增加1.00 mmol/L，糖尿病患病风险增加1.54倍（ P<0.001）；而当HDL-C<1.81 mmol/L时，HDL-C水平每增加1.00 mmol/L，糖尿病患病风险下降6%（ P=0.798）。亚组分析表明，在年龄全亚组、男性、体重指数24.5~52.7 kg/m 2亚组、收缩压全亚组、舒张压69~77和78~164 mmHg（1 mmHg=0.133 kPa）亚组、血总胆固醇0.02~4.26、5.00~17.84 mmol/L亚组、甘油三酯全亚组、低密度脂蛋白胆固醇0~2.42和2.99~12.60 mmol/L亚组、丙氨酸氨基转移酶23.4~1 508.4 U/L亚组、天冬氨酸氨基转移酶0~19.7和24.8~1 026.2 U/L亚组、尿素氮全亚组、肌酐61.5~76.9和77.0~1 116.6 μmol/L亚组、从不吸烟、经常饮酒、无糖尿病家族史中，HDL-C与中国成年人糖尿病患病风险的效应值均显示了良好的稳定性（均 HR>1.00）。 结论:高水平HDL-C与中国成年人糖尿病患病风险增加相关，当HDL-C>1.81 mmol/L时，HDL-C与糖尿病患病风险呈正相关。

abstractsObjective:To investigate the association between high-density lipoprotein cholesterol (HDL-C) and the risk of diabetes mellitus (DM) in Chinese adults.Methods:This study was a secondary analysis of a multicenter, retrospective cohort study using data from the Chinese health screening program in the DATADRYAD database. Between 2010 and 2016, 211833 Chinese adults aged 20 years or older were screened for diabetes at baseline in 32 sites and 11 cities across the country. Baseline HDL-C level was the target independent variable and the risk of DM at follow-up was the dependent variable. Cox proportional hazards regression analysis assessed the independent association between HDL-C levels and the risk of developing DM. In this paper, the generalized Additive Model (GAM) and the smoothing curve fitting method were used to study the nonlinear relationships. In addition, subgroup analyses were conducted to assess the consistency of the correlations among different subgroup and to further validate the reliability of the results.Results:After adjusting for potential confounding factors such as age, sex and body mass index, HDL-C level was positively correlated with the development of diabetes ( HR=1.43, 95% CI: 1.08-1.90, P=0.012). The level of HDL-C showed a non-linear relationship with the risk of DM, and the inflection point was 1.81 mmol/L. The HR (95% CI) of the left and right sides of the inflection point were 0.94 (0.56-1.55) and 2.54 (1.93-3.30), respectively. When HDL-C>1.81 mmol/L, HDL-C was positively correlated with the occurrence of DM. Each 1.00 mmol/L increase in HDL-C increased the risk of diabetes mellitus by 1.54 times ( P<0.001); when HDL-C<1.81 mmol/L, the risk of diabetes decreased by 6% for every 1.00 mmol/L increase in HDL-C ( P=0.798). Subgroup analysis showed that, in the age, male, BMI 24.5-52.7 kg/m 2 subgroups, all the systolic blood pressure subgroups, diastolic blood pressure 69-77 and 78-164 mmHg (1 mmHg=0.133 kPa) subgroups, total cholesterol 0.02-4.26 and 5.00-17.84 mmol/L subgroups, all the triglyceride subgroups, low-density lipoprotein 0-2.42 and 2.99-12.60 mmol/L subgroups, alanine aminotransferase 23.4-1 508.4 U/L subgroups, aspartate transaminase 0-19.7 and 24.8-1 026.2 U/L subgroups, all the urea nitrogen subgroups, creatinine 61.5-76.9, 77.0-1 116.6 μmol/L subgroups, never smoking subgroup, subgroup with frequent alcohol consumption or family history of diabetes mellitus, the effect values of HDL-C and the risk of diabetes mellitus in Chinese adults showed good stability (all HR>1.00). Conclusions:High levels of HDL-C are associated with an increased risk of DM in Chinese adults. When HDL-C is greater than 1.81 mmol/L, HDL-C is positively correlated with DM.