Sugimori H, Miyakawa M, Yoshida K, Izuno T, Takahashi E, Tanaka C, Nakamura K, Hinohara S
Department of Hygiene and Preventive Medicine, Showa University School of Medicine, Tokyo, Japan.
J Med Syst. 1998 Feb;22(1):27-32. doi: 10.1023/a:1022650305109.
This study was designed with a follow up of 16 years to provide the epidemiological model evaluating the risk of developing diabetes mellitus in Japan. A cohort of 2573 subjects (1851 males and 722 females) from a MHTS in Tokyo, who were nondiabetic (fasting blood glucose (FBS) less than 110 mg/dl) in the initial year, were selected. This cohort was followed every year to identify the occurrence of diabetes mellitus. Diabetes Mellitus was defined as fasting blood glucose over 110 mg/dl, or the initiation of diabetic therapy. We compared two prognosis groups (a normal group and a diabetic group) in terms of age, examination findings, and prevalence of health risks (lifestyle, stress, and working form). We also assessed family history of diabetes and past histories including hypertension, hypercholesterolemia, and hyperuricemia. After assessing each variable by univariate analysis (t-test, chi 2 test), we employed Cox's proportional hazards model analysis. We used stepwise model adopting risk factors. The diabetic group had significant differences compared to the normal group in age, BMI (body mass index), FBS, smoking, drinking, not eating breakfast, dairy intake, hypertension, hypercholesterolemia, hyperuricemia, and family history by univariate analysis. According to proportional hazards model analysis, FBS, age, family history, hypertension, smoking, and BMI were incorporated into significant risk factors for diabetes in males, and not eating breakfast, FBS, age, drinking, and hypertension were incorporated in females. Diabetes seemed to be related to fixed factors (age), or genetic factors (family history and FBS) in males. For females, lifestyle (not eating breakfast and drinking habit) seemed to play an important role. It will be worthwhile to assess the risks of developing diabetes mellitus by this epidemiological model.
本研究设计了16年的随访期，以提供评估日本糖尿病发病风险的流行病学模型。从东京的一个MHTS中选取了2573名受试者（1851名男性和722名女性）组成队列，这些受试者在初始年份为非糖尿病患者（空腹血糖（FBS）低于110mg/dl）。每年对该队列进行随访以确定糖尿病的发生情况。糖尿病定义为空腹血糖超过110mg/dl，或开始糖尿病治疗。我们比较了两个预后组（正常组和糖尿病组）在年龄、检查结果以及健康风险（生活方式、压力和工作形式）患病率方面的情况。我们还评估了糖尿病家族史以及包括高血压、高胆固醇血症和高尿酸血症在内的既往病史。在通过单因素分析（t检验、卡方检验）评估每个变量后，我们采用了Cox比例风险模型分析。我们使用逐步模型纳入风险因素。通过单因素分析，糖尿病组与正常组在年龄、体重指数（BMI）、空腹血糖、吸烟、饮酒、不吃早餐、乳制品摄入量、高血压、高胆固醇血症、高尿酸血症和家族史方面存在显著差异。根据比例风险模型分析，空腹血糖、年龄、家族史、高血压、吸烟和BMI被纳入男性糖尿病的显著风险因素，而不吃早餐、空腹血糖、年龄、饮酒和高血压被纳入女性糖尿病的显著风险因素。在男性中，糖尿病似乎与固定因素（年龄）或遗传因素（家族史和空腹血糖）有关。对于女性来说，生活方式（不吃早餐和饮酒习惯）似乎起着重要作用。通过这种流行病学模型评估糖尿病发病风险将是值得的。
