第二型糖尿病延遲診斷預測因子之研究：以感染性疾病為例

A study on predicted factors for a delay in diagnosis of type 2 diabetes：infectious disease as an example

陳怡慧(Yi-Huei Chen)

指導教授 ： 邱弘毅

臺北醫學大學/公共衛生暨營養學院/公共衛生學研究所/碩士(2012年)

https://doi.org/10.6831/TMU.2012.00023

摘要

研究目的
過去研究指出，第二型糖尿病可能在被臨床確診之前，已存在至少4-7年，最長可為9-12年。第二型糖尿病的延遲診斷，不但會影響到病患的治療與預後，同時也會造成醫療體系以及社會負擔的增加。目前已有許多研究指出高血糖會對個體的免疫系統產生負面影響，以致罹患感染性疾病的危險性上升，且有關於第二型糖尿病延遲診斷與感染性疾病的相關研究資料仍有限。因此，本研究使用台灣全民健保資料庫，1)探討研究對象在確診為第二型糖尿病之前，感染性疾病的
罹患時間、頻率以及疾病別的總數與之後確診為第二型糖尿病之相關性；2)分析在確診為第二型糖尿病之前，哪些感染性疾病可作為糖尿病延遲診斷的預測因子；3)評估利用感染性疾病作為預測指標，將可縮短多久第二型糖尿病延遲診斷的時間。

材料與方法
本研究為病例對照研究，以2005 年承保抽樣歸人檔中2003 年至2009 年間，選取73,393 位第二型糖尿病患者，從中排除非新發的糖尿病患者、第一型的糖尿病患者、年紀小於18 歲者、性別不詳者以及妊娠糖尿病患者後，共有34,842位新發的第二型糖尿病患者為病例組(case 組)。另外，從承保檔(ID)中，排除第一型及第二型糖尿病患者、年紀小於18 歲者、性別不詳以及妊娠糖尿病患者後，採一比一的方式與病例組配對年齡及性別，共有34,842 位非糖尿病患者為對照組(control 組)。以新發第二型糖尿病患者診斷日期為index date，探討兩組在index date 前三年內，感染性疾病的罹患時間、頻率以及疾病別的總數與之後確診第二型糖尿病之相關性。
進一步從69,684 位研究對象中，排除有任一危險因子者(包含癌症、慢性腎臟病、自體免疫疾病、高血壓、血脂異常、C 肝、心血管疾病史、多囊性卵巢症候群以及胰臟發炎)，採一比二的方式配對兩組的年齡及性別，探討兩組在index date 前三年內，感染性疾病的罹患時間、頻率以及疾病別的總數與之後確診第二型糖尿病之相關性。
以Chi-square test 檢驗case 組及control 組在基本人口學及臨床特性有無差異；conditional logistic regression 分析case 組及control 組，罹患感染性疾病(以每三個月為一時期)與之後確診第二型糖尿病的相關性；trend test 分析確診(index date)前罹患感染性疾病與之後確診第二型糖尿病的劑量效應關係是否達統計上顯著
意義。

研究結果
69,684位研究對象結果顯示，在調整了其它因子後，在感染性疾病罹患時間的部分，在第二型糖尿病確診前12個月內，罹患感染性疾病者之後確診第二型糖尿病的危險性，顯著高於無罹患感染性疾病者，且隨著第二型糖尿病確診時間的接近，罹患感染性疾病與之後確診第二型糖尿病的危險性隨之增高。在感染頻率的部分，結果指出在第二型糖尿病確診前27個月內，感染頻率越高之後確診第二型糖尿病的危險性顯著越高。在感染性疾病別罹患總數的部分，結果指出在第二型糖尿病確診前21個月內，罹患總數越多，之後確診第二型糖尿病的危險性就越高。而10,971位無危險因子者的研究結果則與有任一危險因子者相似。針對所有研究對象以及無危險因子者的特定感染性疾病分析結果顯示，蜂窩性組織炎、真菌病以及腎、尿道、膀胱感染與之後確診為第二型糖尿病皆具顯著相關，較適合作為確診第二型糖尿病的預測指標。

結論
研究結果指出，以感染性疾病作為第二型糖尿病延遲診斷之預測因子，可縮短約一年延遲診斷的時間。特定的感染性疾病，如蜂窩性組織炎、真菌病以及腎、尿道、膀胱感染，對於縮短糖尿病之延遲診斷有較高的預測性，較適合作為醫生在診療時的輔助指標，找出尚未被診斷的第二型糖尿病個案，以縮短第二型糖尿病延遲診斷時間。

關鍵字

第二型糖尿病；延遲診斷；感染性疾病；台灣全民健保資料庫；病例對照研究

並列摘要

OBJECTIVE
Previous studies have shown that type II diabetes (DM here after) may be present for at least 4~7 years or more before its clinical diagnosis. A delayed diagnosis of DM, not only affects the treatment and prognosis of the patients, but also increases the burden of healthcare and society. Many studies have indicated that hyperglycemia may influence the immunity system and increase the occurrence of infectious diseases. However, the association between the risk of infectious diseases and a delayed diagnosis of DM is still unclear. Therefore, the aim of this study is 1) to examine the association between infectious diseases and DM diagnosis prior to clinical onset by using National Health Insurance Research Database (NHIRD) in Taiwan； 2) to evaluate the association between specific infectious diseases and DM diagnosis prior to clinical onset； 3) to assess whether infectious disease diagnoses could be a clinical tool to shorten the time between the biological onset and clinical onset of DM.

METHODS
We conducted a 1:1 matched case-control study. Data was obtained from NHIRD between 2000 and 2010. A total of 73,393 patients with DM was selected between 2003 and 2009.Those who were diagnosed with type I DM or gestational DM , age less than 18 and unknown sex were excluded. We additionally excluded patients with DM diagnosis prior to year of 2003. A total of 34,842 new incidence cases were included and treated as a case group. From the reminding samples, we selected a comparison group without any DM diagnosis and matched cases in terms of age and sex.
We had a subset of sample that cases were selected if they did not have any coexisting medical condition(including cancer, chronic renal disease, hypertension, polycystic ovarian syndrome, autoimmune disorder, dyslipidemia, history of cardiovascular disease, hepatitis C and pancreatitis) and comparison controls were selected and matched the age and sex of cases(match ratio1:2). Index date of DM diagnosis as index diagnosis (here after).Both cases and controls were tracked their medical claims prior to three years of index diangosis. Outcome of study interest were the time, frequency and type of infectious diseases.
Baseline difference of cases and controls were compared by using Chi-square test. Conditional logistic regressions were used to examine the risk of infectious diseases (every three-month prior to index diagnosis as a window) on the incidence of DM. Trend test was used to assess the frequency of infectious disease visits on the probability of DM diagnosis.

RESULTS
A total of 69,684 subjects and 10,971 subjects without any risk factors were included in the study. The results showed that patients with infectious diseases prior to index diagnosis have a higher risk of being diagnosed with DM. We found the significance prior to 12 months of index diagnosis after adjust for other risk factors. The association between infectious diseases and DM diagnosis is stronger as the diagnosis of infectious diseases is closed to the index diagnosis. In term of frequency of infectious disease diagnoses, we found the association prior to 27 month of index diagnosis. A total number of types of infectious diseases prior to index diagnosis was also positively associated with the risk of DM.
The results of subjects without any risk factors were consistent with the findings of those with risk factors.
The results of the analyses of specific infectious diseases showed that cellulitis, mycoses and kidney, urinary tract, and bladder infections could be used as a predicted factor for shortening the time between the biological onset and clinical onset of DM.

CONCLUSIONS
In sum, this study found that the time between biological onset and clinical onset of DM can be shortened 1 year by using infectious disease diagnosis as an indicator to predict the risk of DM.
Especially, cellulitis, mycoses and kidney, urinary tract, and bladder infections could be used as a predicted factor for shortening the delayed diagnosis time of DM.

並列關鍵字

type II diabetes；delayed diagnosis；infectious disease；Taiwan National Health Insurance Research Database；case-control study

參考文獻

Lin, C. C., M. S. Lai, et al. (2005). "Accuracy of diabetes diagnosis in health insurance claims data in Taiwan." J Formos Med Assoc 104(3): 157-163.

Abdul-Ghani, M. A. and R. A. DeFronzo (2009). "Pathophysiology of prediabetes." Curr Diab Rep 9(3): 193-199.

Agyemang, C., I. Vaartjes, et al. (2009). "Risk of death after first admission for cardiovascular diseases by country of birth in The Netherlands: a nationwide record-linked retrospective cohort study." Heart 95(9): 747-753.

Ahluwalia, A., A. Sood, et al. (2000). "Nasal colonization with Staphylococcus aureus in patients with diabetes mellitus." Diabet Med 17(6): 487-488.

Alexiewicz, J. M., D. Kumar, et al. (1995). "Polymorphonuclear leukocytes in non-insulin-dependent diabetes mellitus: abnormalities in metabolism and function." Ann Intern Med 123(12): 919-924.