Programme in Health Services and Systems Research, Duke-NUS Medical School, Singapore, Republic of Singapore; School of Population Health & Environmental Sciences, Faculty of Life Sciences & Medicine, King's College London, London, United Kingdom; National Institute for Health Research (NIHR) Biomedical Research Centre, Guy's and St Thomas' NHS Foundation Trust and King's College London, United Kingdom.
SingHealth Regional Health System, Singapore Health Services, Singapore, Republic of Singapore.
Diabetes Res Clin Pract. 2021 Jan;171:108551. doi: 10.1016/j.diabres.2020.108551. Epub 2020 Nov 22.
To examine whether the excess risks of coronary heart disease (CHD), stroke, dialysis, and lower extremity amputation (LEA) associated with type 2 diabetes mellitus (T2DM) differ across age, sex, and ethnicity in Singapore.
Using a 10-year administrative data, we matched individuals with T2DM using nearest neighbour matching, to those without, in 1:2 ratio. To examine whether the excess risks were heterogeneous across age, sex, and ethnicity, we generated interaction terms of age, sex, and ethnicity with T2DM status in Cox proportional hazard (PH) models.
The main analyses included ~1 million person years, comprising 66,099 and 120,485 individuals with and without T2DM, respectively. The associations of T2DM with CHD and dialysis, split into two time periods to address violation of PH assumption, were higher with older age in short-term but lower with older age in long-term. The association of T2DM with stroke and LEA were lower with older age. The associations of T2DM with CHD and stroke were also consistently higher in women than men. The associations of T2DM with LEA were higher in ethnic Malays than ethnic Chinese.
The excess risks of CHD, stroke, dialysis, and LEA associated with T2DM were heterogeneous across some demographic subgroups.
在新加坡人群中，检验 2 型糖尿病（T2DM）与冠心病（CHD）、中风、透析和下肢截肢（LEA）风险增加的相关性是否因年龄、性别和种族而异。
我们使用 10 年的行政数据，采用 1:2 的比例，通过最近邻匹配法，将 T2DM 患者与无 T2DM 患者进行匹配。为了检验 T2DM 状态与年龄、性别和种族之间的超额风险是否存在异质性，我们在 Cox 比例风险（PH）模型中生成了年龄、性别和种族与 T2DM 状态的交互项。
主要分析包括约 100 万人年的数据，其中分别有 66099 人和 120485 人患有和未患有 T2DM。将 T2DM 与 CHD 和透析的关联分为两个时间段，以解决 PH 假设的违反问题，结果显示在短期，T2DM 与 CHD 和透析的关联随着年龄的增加而增加，但在长期，随着年龄的增加，T2DM 与 CHD 和透析的关联降低。T2DM 与中风和 LEA 的关联则随着年龄的增加而降低。T2DM 与 CHD 和中风的关联在女性中也始终高于男性。T2DM 与 LEA 的关联在马来族中高于华族。
T2DM 与 CHD、中风、透析和 LEA 风险增加的相关性在一些人口统计学亚组中存在异质性。
