| 摘要: | 糖尿病與癌症之間有共同的危險因子，例如抽菸、喝酒、肥胖、缺乏運動、攝取高熱量或脂肪以及有關代謝方面的疾病。但是在台灣較少探討糖尿病以及癌症之間的相關研究。因此本研究主要的研究目的有二，第一為估算糖尿病病人十大癌症的發生率；第二為探討糖尿病與胰臟癌之相關。
本研究使用回顧性世代研究法，研究追蹤起始點為西元1997年至1998年且追蹤10至11年，在第一個研究目的，第二型糖尿病族群十大癌症發生率與一般沒有任何糖尿病族群作比較，由台灣健保資料庫(NHIRD)中全部糖尿病門診檔以及全部住院檔獲得第二型糖尿病族群，由納保人口基本資料檔定義一般族群。在此研究中，第二型糖尿病族群有474,686人，一般族群有9,412,056人。在第二個研究目的，一般族群是由百萬歸人門診檔以及住院檔獲得。在此研究中，依照不同的年齡、性別及居住地，一般族群與第二型糖尿病族群以1:1之頻率配對，獲得168,689第二型糖尿病個案及一般族群個案。癌症發生資料從全部的癌症門診檔以及全部住院檔獲得。使用卜瓦松迴歸分析計算糖尿病病人發生十大癌症之標準化發生比，以Cox比例危險模型預估第二型糖尿病與胰臟癌之相關，並且調整人口學因子以及合併症的干擾效應。
不論男性或女性，第二型糖尿病有較高發生肝癌與胰臟癌 (男性肝癌SIR為1.60，95%信賴區間為1.57-1.64；胰臟癌SIR為1.55，95%信賴區間為1.53-1.72。女性肝癌SIR為1.55，95%信賴區間1.51-1.60；胰臟癌SIR為1.45，95%信賴區間為1.35-1.55。)。以Cox比例危險模型探討第二型糖尿病與胰臟癌之相關，調整人口學因子以及合併症後，發現第二型糖尿病為胰臟癌的風險因子，在追蹤時間小於4.5年時，其風險比值為3.776，95% CI為2.717-5.247，在追蹤時間大於4.5年時，其風險比值為2.057，95% CI為1.72-2.46。
在此研究中發現，第二型糖尿病為肝癌、結腸直腸癌、口腔癌、女性乳癌以及胰臟癌之風險因子。但對於肺癌、食道癌，攝護腺癌以及子宮頸癌則為保護因子。第二型糖尿病和胰臟癌間的風險，則獨立於一般人口學因子及合併症。
Diabetes mellitus and cancers share the common risk factors such as smoking, alcohol consumption, obesity, physical inactivity, a high calorie intake and particularly saturated fat intake, and metabolic disorders. But there were few studies exploring the association of cancers incidence with diabetes mellitus in Taiwan. Therefore, there are two aims of this study. One is to estimate ten major cancer risks in type 2 diabetes in Taiwan, and the other is to evaluate the association between type 2 diabetes and pancreatic cancer by adjusting for sociodemographic factors and co-morbidity.
For the first aim, we conducted a retrospective cohort study of Taiwan patients with type 2 diabetes identified in 1997-1998 and being followed for 10-11 years, and compared their cancer incidences with those of the entire Taiwan general population of the same age without any diabetes. It consisted of 474,686 type 2 diabetes patients and 9,412,056 individuals without any diabetes at baseline being followed for 10-11 years. For the second aim, patients with type 2 diabetes were obtained from all outpatients and all inpatients datasets of Taiwan National Health Insurance Research Database (NHIRD) and general population were obtained from the Longitudinal Health Insurance Database 2005 (LHID2005). The LHID2005 contains all ambulatory and inpatient claims data of one million beneficiaries, randomly sampled from 25.68 million enrollees in the NHI. In order to control for age, gender, and residential area, we randomly selected 1 patients of type 2 diabetes for every person without any diabetes by frequency matching on age (5-year groups), sex and residential area from 1997 to 1998. It consisted of 168,689 type 2 diabetes patients and individuals without any diabetes at baseline being followed for 10-11 years. Cancer incidences were also identified from all inpatient and all outpatient datasets of NHIRD. Standardized incidence ratios (SIRs) were estimated for ten major cancer using Poisson regression analysis for estimating ten major cancers incidence rates study. Cox’s proportional hazards model was used to evalute the association of type 2 diabetes and pancreatic cancer by adjusting for sociodemographic factors and co-morbidity.
For estimating ten major cancer incidence rates, liver and pancreas cancers had the highest standardized incidence ratios (SIRs) both in men and women (in men, SIR=1.60, 95% CI=1.57-1.64 for liver cancer and SIR=1.62, 95% CI=1.53-1.72 for pancreas cancer; in women, SIR=1.55, 95% CI=1.51-1.60 for liver cancer and SIR=1.45, 95% CI=1.35-1.55 for pancreas cancer). For exploring the relationship between type 2 diabetes and pancreatic cancer incidence, we found type 2 diabetes was associated with pancreatic cancer after adjusting for sociodemographic factors and co-morbidity (hazard ratio: 3.776, 95% CI: 2.717-5.247 for follow-up time less than 4.5 years; hazard ratio: 2.057, 95% CI: 1.72-2.46 for follow-up time more than 4.5 years).
In summary, this study identified an increased risk for liver, colorectal, oral, breast and pancreatic cancers in patients with type 2 diabetes in Taiwan. But decreased risk of cancers of lung, esophagus, prostate and cervix were observed. In addition, type 2 diabetes was found to be associated with increased risk of pancreatic cancer and this effect was independent of sociodemographic factors and co-morbidity. |
